diff --git a/2.7/amp.html b/2.7/amp.html index 161834c67a8..6d90db42da8 100644 --- a/2.7/amp.html +++ b/2.7/amp.html @@ -646,7 +646,7 @@

Automatic Mixed Precision package - torch.amp

Autocasting

-torch.amp.autocast_mode.is_autocast_available(device_type)[source][source]
+torch.amp.autocast_mode.is_autocast_available(device_type)[source][source]

Return a bool indicating if autocast is available on device_type.

Parameters
@@ -662,7 +662,7 @@

Automatic Mixed Precision package - torch.amp
-class torch.autocast(device_type, dtype=None, enabled=True, cache_enabled=None)[source][source]
+class torch.autocast(device_type, dtype=None, enabled=True, cache_enabled=None)[source][source]

Instances of autocast serve as context managers or decorators that allow regions of your script to run in mixed precision.

In these regions, ops run in an op-specific dtype chosen by autocast @@ -828,7 +828,7 @@

Automatic Mixed Precision package - torch.amp
-torch.amp.custom_fwd(fwd=None, *, device_type, cast_inputs=None)[source][source]
+torch.amp.custom_fwd(fwd=None, *, device_type, cast_inputs=None)[source][source]

Create a helper decorator for forward methods of custom autograd functions.

Autograd functions are subclasses of torch.autograd.Function. See the example page for more detail.

@@ -855,7 +855,7 @@

Automatic Mixed Precision package - torch.amp
-torch.amp.custom_bwd(bwd=None, *, device_type)[source][source]
+torch.amp.custom_bwd(bwd=None, *, device_type)[source][source]

Create a helper decorator for backward methods of custom autograd functions.

Autograd functions are subclasses of torch.autograd.Function. Ensures that backward executes with the same autocast state as forward. @@ -871,7 +871,7 @@

Automatic Mixed Precision package - torch.amp
-class torch.cuda.amp.autocast(enabled=True, dtype=torch.float16, cache_enabled=True)[source][source]
+class torch.cuda.amp.autocast(enabled=True, dtype=torch.float16, cache_enabled=True)[source][source]

See torch.autocast.

torch.cuda.amp.autocast(args...) is deprecated. Please use torch.amp.autocast("cuda", args...) instead.

@@ -880,21 +880,21 @@

Automatic Mixed Precision package - torch.amp
-torch.cuda.amp.custom_fwd(fwd=None, *, cast_inputs=None)[source][source]
+torch.cuda.amp.custom_fwd(fwd=None, *, cast_inputs=None)[source][source]

torch.cuda.amp.custom_fwd(args...) is deprecated. Please use torch.amp.custom_fwd(args..., device_type='cuda') instead.

-torch.cuda.amp.custom_bwd(bwd)[source][source]
+torch.cuda.amp.custom_bwd(bwd)[source][source]

torch.cuda.amp.custom_bwd(args...) is deprecated. Please use torch.amp.custom_bwd(args..., device_type='cuda') instead.

-class torch.cpu.amp.autocast(enabled=True, dtype=torch.bfloat16, cache_enabled=True)[source][source]
+class torch.cpu.amp.autocast(enabled=True, dtype=torch.bfloat16, cache_enabled=True)[source][source]

See torch.autocast. torch.cpu.amp.autocast(args...) is deprecated. Please use torch.amp.autocast("cpu", args...) instead.

@@ -925,7 +925,7 @@

Automatic Mixed Precision package - torch.amp
-class torch.cuda.amp.GradScaler(init_scale=65536.0, growth_factor=2.0, backoff_factor=0.5, growth_interval=2000, enabled=True)[source][source]
+class torch.cuda.amp.GradScaler(init_scale=65536.0, growth_factor=2.0, backoff_factor=0.5, growth_interval=2000, enabled=True)[source][source]

See torch.amp.GradScaler. torch.cuda.amp.GradScaler(args...) is deprecated. Please use torch.amp.GradScaler("cuda", args...) instead.

@@ -934,7 +934,7 @@

Automatic Mixed Precision package - torch.amp
-class torch.cpu.amp.GradScaler(init_scale=65536.0, growth_factor=2.0, backoff_factor=0.5, growth_interval=2000, enabled=True)[source][source]
+class torch.cpu.amp.GradScaler(init_scale=65536.0, growth_factor=2.0, backoff_factor=0.5, growth_interval=2000, enabled=True)[source][source]

See torch.amp.GradScaler. torch.cpu.amp.GradScaler(args...) is deprecated. Please use torch.amp.GradScaler("cpu", args...) instead.

diff --git a/2.7/autograd.html b/2.7/autograd.html index 2951d84d7ff..2f0a29d0a09 100644 --- a/2.7/autograd.html +++ b/2.7/autograd.html @@ -807,7 +807,7 @@

Tensor autograd functionsFunction

-class torch.autograd.Function(*args, **kwargs)[source][source]
+class torch.autograd.Function(*args, **kwargs)[source][source]

Base class to create custom autograd.Function.

To create a custom autograd.Function, subclass this class and implement the forward() and backward() static methods. Then, to use your custom @@ -925,7 +925,7 @@

Profileremit_itt.

-class torch.autograd.profiler.profile(enabled=True, *, use_cuda=False, use_device=None, record_shapes=False, with_flops=False, profile_memory=False, with_stack=False, with_modules=False, use_kineto=False, use_cpu=True, experimental_config=None, acc_events=False, custom_trace_id_callback=None)[source][source]
+class torch.autograd.profiler.profile(enabled=True, *, use_cuda=False, use_device=None, record_shapes=False, with_flops=False, profile_memory=False, with_stack=False, with_modules=False, use_kineto=False, use_cpu=True, experimental_config=None, acc_events=False, custom_trace_id_callback=None)[source][source]

Context manager that manages autograd profiler state and holds a summary of results.

Under the hood it just records events of functions being executed in C++ and exposes those events to Python. You can wrap any code into it and it will @@ -1034,7 +1034,7 @@

Profiler
-class torch.autograd.profiler.emit_nvtx(enabled=True, record_shapes=False)[source][source]
+class torch.autograd.profiler.emit_nvtx(enabled=True, record_shapes=False)[source][source]

Context manager that makes every autograd operation emit an NVTX range.

It is useful when running the program under nvprof:

nvprof --profile-from-start off -o trace_name.prof -- <regular command here>
@@ -1103,7 +1103,7 @@ 
Profiler
 

-class torch.autograd.profiler.emit_itt(enabled=True, record_shapes=False)[source][source]

+class torch.autograd.profiler.emit_itt(enabled=True, record_shapes=False)[source][source]
 
Context manager that makes every autograd operation emit an ITT range.

 
It is useful when running the program under Intel(R) VTune Profiler:

 
vtune <--vtune-flags> <regular command here>
@@ -1149,7 +1149,7 @@ 
Profiler

 

 

-class torch.autograd.detect_anomaly(check_nan=True)[source][source]

+class torch.autograd.detect_anomaly(check_nan=True)[source][source]
 
Context-manager that enable anomaly detection for the autograd engine.

 
This does two things:

 
    
@@ -1219,7 +1219,7 @@ 
    Debugging and anomaly detection
 
    
-class torch.autograd.set_detect_anomaly(mode, check_nan=True)[source][source]
    
+class torch.autograd.set_detect_anomaly(mode, check_nan=True)[source][source]
 
    Context-manager that sets the anomaly detection for the autograd engine on or off.
    
 
    set_detect_anomaly will enable or disable the autograd anomaly detection
 based on its argument mode.
@@ -1295,7 +1295,7 @@ 
    Autograd graphHooks for saved tensors.
    
 
    
 
    
-class torch.autograd.graph.saved_tensors_hooks(pack_hook, unpack_hook)[source][source]
    
+class torch.autograd.graph.saved_tensors_hooks(pack_hook, unpack_hook)[source][source]
 
    Context-manager that sets a pair of pack / unpack hooks for saved tensors.
    
 
    Use this context-manager to define how intermediary results of an operation
 should be packed before saving, and unpacked on retrieval.
    
@@ -1356,7 +1356,7 @@ 
    Autograd graph
 
    
-class torch.autograd.graph.save_on_cpu(pin_memory=False, device_type='cuda')[source][source]
    
+class torch.autograd.graph.save_on_cpu(pin_memory=False, device_type='cuda')[source][source]
 
    Context manager under which tensors saved by the forward pass will be stored on cpu, then retrieved for backward.
    
 
    When performing operations within this context manager, intermediary
 results saved in the graph during the forward pass will be moved to CPU,
@@ -1396,7 +1396,7 @@ 
    Autograd graph
 
    
-class torch.autograd.graph.disable_saved_tensors_hooks(error_message)[source][source]
    
+class torch.autograd.graph.disable_saved_tensors_hooks(error_message)[source][source]
 
    Context-manager that disables the saved tensors default hooks feature.
    
 
    Useful for if you are creating a feature that does not work with saved
 tensors default hooks.
    
@@ -1422,7 +1422,7 @@ 
    Autograd graph
 
    
-class torch.autograd.graph.register_multi_grad_hook(tensors, fn, *, mode='all')[source][source]
    
+class torch.autograd.graph.register_multi_grad_hook(tensors, fn, *, mode='all')[source][source]
 
    Register a multi-grad backward hook.
    
 
    There are two supported modes: "all" and "any".
    
 
    Under the "all" mode, the hook will be called after gradients with respect to every tensor in
@@ -1473,7 +1473,7 @@ 
    Autograd graph
 
    
-class torch.autograd.graph.allow_mutation_on_saved_tensors[source][source]
    
+class torch.autograd.graph.allow_mutation_on_saved_tensors[source][source]
 
    Context manager under which mutating tensors saved for backward is allowed.
    
 
    Under this context manager, tensors saved for backward are cloned on mutation,
 so the original version can still be used during backward. Normally, mutating a tensor
@@ -1509,7 +1509,7 @@ 
    Autograd graph
 
    
-class torch.autograd.graph.GradientEdge(node, output_nr)[source][source]
    
+class torch.autograd.graph.GradientEdge(node, output_nr)[source][source]
 
    Object representing a given gradient edge within the autograd graph.
    
 
    To get the gradient edge where a given Tensor gradient will be computed,
 you can do edge = autograd.graph.get_gradient_edge(tensor).
    
@@ -1519,7 +1519,7 @@ 
    Autograd graph
 
    
-torch.autograd.graph.get_gradient_edge(tensor)[source][source]
    
+torch.autograd.graph.get_gradient_edge(tensor)[source][source]
 
    Get the gradient edge for computing the gradient of the given Tensor.
    
 
    In particular, it is equivalent to call
 g = autograd.grad(loss, input) and g = autograd.grad(loss, get_gradient_edge(input)).
    
diff --git a/2.7/backends.html b/2.7/backends.html
index c42844f0ede..87c64cfd636 100644
--- a/2.7/backends.html
+++ b/2.7/backends.html
@@ -612,7 +612,7 @@
 
    torch.backends.cpu
    
 
    
 
    
-torch.backends.cpu.get_cpu_capability()[source][source]
    
+torch.backends.cpu.get_cpu_capability()[source][source]
 
    Return cpu capability as a string value.
    
 
    Possible values:
 - “DEFAULT”
@@ -634,7 +634,7 @@
 
    torch.backends.cuda
    
 
    
 
    
-torch.backends.cuda.is_built()[source][source]
    
+torch.backends.cuda.is_built()[source][source]
 
    Return whether PyTorch is built with CUDA support.
    
 
    Note that this doesn’t necessarily mean CUDA is available; just that if this PyTorch
 binary were run on a machine with working CUDA drivers and devices, we would be able to use it.
    
@@ -686,7 +686,7 @@
 
 
    
 
    
-torch.backends.cuda.preferred_blas_library(backend=None)[source][source]
    
+torch.backends.cuda.preferred_blas_library(backend=None)[source][source]
 
    Override the library PyTorch uses for BLAS operations. Choose between cuBLAS, cuBLASLt, and CK [ROCm-only].
    
 
    
 
    Warning
    
@@ -719,7 +719,7 @@
 
 
    
 
    
-torch.backends.cuda.preferred_rocm_fa_library(backend=None)[source][source]
    
+torch.backends.cuda.preferred_rocm_fa_library(backend=None)[source][source]
 
    [ROCm-only]
 Override the backend PyTorch uses in ROCm environments for Flash Attention. Choose between AOTriton and CK
    
 
    
@@ -749,7 +749,7 @@
 
 
    
 
    
-torch.backends.cuda.preferred_linalg_library(backend=None)[source][source]
    
+torch.backends.cuda.preferred_linalg_library(backend=None)[source][source]
 
    Override the heuristic PyTorch uses to choose between cuSOLVER and MAGMA for CUDA linear algebra operations.
    
 
    
 
    Warning
    
@@ -804,7 +804,7 @@
 
 
    
 
    
-torch.backends.cuda.flash_sdp_enabled()[source][source]
    
+torch.backends.cuda.flash_sdp_enabled()[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -814,7 +814,7 @@
 
 
    
 
    
-torch.backends.cuda.enable_mem_efficient_sdp(enabled)[source][source]
    
+torch.backends.cuda.enable_mem_efficient_sdp(enabled)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -826,7 +826,7 @@
 
 
    
 
    
-torch.backends.cuda.mem_efficient_sdp_enabled()[source][source]
    
+torch.backends.cuda.mem_efficient_sdp_enabled()[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -836,7 +836,7 @@
 
 
    
 
    
-torch.backends.cuda.enable_flash_sdp(enabled)[source][source]
    
+torch.backends.cuda.enable_flash_sdp(enabled)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -848,7 +848,7 @@
 
 
    
 
    
-torch.backends.cuda.math_sdp_enabled()[source][source]
    
+torch.backends.cuda.math_sdp_enabled()[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -858,7 +858,7 @@
 
 
    
 
    
-torch.backends.cuda.enable_math_sdp(enabled)[source][source]
    
+torch.backends.cuda.enable_math_sdp(enabled)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -870,7 +870,7 @@
 
 
    
 
    
-torch.backends.cuda.fp16_bf16_reduction_math_sdp_allowed()[source][source]
    
+torch.backends.cuda.fp16_bf16_reduction_math_sdp_allowed()[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -880,7 +880,7 @@
 
 
    
 
    
-torch.backends.cuda.allow_fp16_bf16_reduction_math_sdp(enabled)[source][source]
    
+torch.backends.cuda.allow_fp16_bf16_reduction_math_sdp(enabled)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -892,7 +892,7 @@
 
 
    
 
    
-torch.backends.cuda.cudnn_sdp_enabled()[source][source]
    
+torch.backends.cuda.cudnn_sdp_enabled()[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -902,7 +902,7 @@
 
 
    
 
    
-torch.backends.cuda.enable_cudnn_sdp(enabled)[source][source]
    
+torch.backends.cuda.enable_cudnn_sdp(enabled)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -914,7 +914,7 @@
 
 
    
 
    
-torch.backends.cuda.is_flash_attention_available()[source][source]
    
+torch.backends.cuda.is_flash_attention_available()[source][source]
 
    Check if PyTorch was built with FlashAttention for scaled_dot_product_attention.
    
 
    
 
    Returns
    
@@ -933,7 +933,7 @@
 
 
    
 
    
-torch.backends.cuda.can_use_flash_attention(params, debug=False)[source][source]
    
+torch.backends.cuda.can_use_flash_attention(params, debug=False)[source][source]
 
    Check if FlashAttention can be utilized in scaled_dot_product_attention.
    
 
    
 
    Parameters
    
@@ -961,7 +961,7 @@
 
 
    
 
    
-torch.backends.cuda.can_use_efficient_attention(params, debug=False)[source][source]
    
+torch.backends.cuda.can_use_efficient_attention(params, debug=False)[source][source]
 
    Check if efficient_attention can be utilized in scaled_dot_product_attention.
    
 
    
 
    Parameters
    
@@ -989,7 +989,7 @@
 
 
    
 
    
-torch.backends.cuda.can_use_cudnn_attention(params, debug=False)[source][source]
    
+torch.backends.cuda.can_use_cudnn_attention(params, debug=False)[source][source]
 
    Check if cudnn_attention can be utilized in scaled_dot_product_attention.
    
 
    
 
    Parameters
    
@@ -1017,7 +1017,7 @@
 
 
    
 
    
-torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=True, enable_mem_efficient=True, enable_cudnn=True)[source][source]
    
+torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=True, enable_mem_efficient=True, enable_cudnn=True)[source][source]
 
    
 
    Warning
    
 
    This flag is beta and subject to change.
    
@@ -1033,13 +1033,13 @@
 
    torch.backends.cudnn
    
 
    
 
    
-torch.backends.cudnn.version()[source][source]
    
+torch.backends.cudnn.version()[source][source]
 
    Return the version of cuDNN.
    
 
    
 
 
    
 
    
-torch.backends.cudnn.is_available()[source][source]
    
+torch.backends.cudnn.is_available()[source][source]
 
    Return a bool indicating if CUDNN is currently available.
    
 
    
 
@@ -1085,7 +1085,7 @@
 
    torch.backends.cusparselt
    
 
    
 
    
-torch.backends.cusparselt.version()[source][source]
    
+torch.backends.cusparselt.version()[source][source]
 
    Return the version of cuSPARSELt
    
 
    
 
    Return type
    
@@ -1096,7 +1096,7 @@
 
 
    
 
    
-torch.backends.cusparselt.is_available()[source][source]
    
+torch.backends.cusparselt.is_available()[source][source]
 
    Return a bool indicating if cuSPARSELt is currently available.
    
 
    
 
    Return type
    
@@ -1110,7 +1110,7 @@
 
    torch.backends.mha
    
 
    
 
    
-torch.backends.mha.get_fastpath_enabled()[source][source]
    
+torch.backends.mha.get_fastpath_enabled()[source][source]
 
    Returns whether fast path for TransformerEncoder and MultiHeadAttention
 is enabled, or True if jit is scripting.
    
 
    
@@ -1127,7 +1127,7 @@
 
 
    
 
    
-torch.backends.mha.set_fastpath_enabled(value)[source][source]
    
+torch.backends.mha.set_fastpath_enabled(value)[source][source]
 
    Sets whether fast path is enabled
    
 
    
 
    
@@ -1138,7 +1138,7 @@
 
    torch.backends.mps
    
 
    
 
    
-torch.backends.mps.is_available()[source][source]
    
+torch.backends.mps.is_available()[source][source]
 
    Return a bool indicating if MPS is currently available.
    
 
    
 
    Return type
    
@@ -1149,7 +1149,7 @@
 
 
    
 
    
-torch.backends.mps.is_built()[source][source]
    
+torch.backends.mps.is_built()[source][source]
 
    Return whether PyTorch is built with MPS support.
    
 
    Note that this doesn’t necessarily mean MPS is available; just that
 if this PyTorch binary were run a machine with working MPS drivers
@@ -1166,13 +1166,13 @@
 
    torch.backends.mkl
    
 
    
 
    
-torch.backends.mkl.is_available()[source][source]
    
+torch.backends.mkl.is_available()[source][source]
 
    Return whether PyTorch is built with MKL support.
    
 
    
 
 
    
 
    
-class torch.backends.mkl.verbose(enable)[source][source]
    
+class torch.backends.mkl.verbose(enable)[source][source]
 
    On-demand oneMKL verbosing functionality.
    
 
    To make it easier to debug performance issues, oneMKL can dump verbose
 messages containing execution information like duration while executing
@@ -1204,13 +1204,13 @@
 
    torch.backends.mkldnn
    
 
    
 
    
-torch.backends.mkldnn.is_available()[source][source]
    
+torch.backends.mkldnn.is_available()[source][source]
 
    Return whether PyTorch is built with MKL-DNN support.
    
 
    
 
 
    
 
    
-class torch.backends.mkldnn.verbose(level)[source][source]
    
+class torch.backends.mkldnn.verbose(level)[source][source]
 
    On-demand oneDNN (former MKL-DNN) verbosing functionality.
    
 
    To make it easier to debug performance issues, oneDNN can dump verbose
 messages containing information like kernel size, input data size and
@@ -1243,19 +1243,19 @@
 
    torch.backends.nnpack
    
 
    
 
    
-torch.backends.nnpack.is_available()[source][source]
    
+torch.backends.nnpack.is_available()[source][source]
 
    Return whether PyTorch is built with NNPACK support.
    
 
    
 
 
    
 
    
-torch.backends.nnpack.flags(enabled=False)[source][source]
    
+torch.backends.nnpack.flags(enabled=False)[source][source]
 
    Context manager for setting if nnpack is enabled globally
    
 
    
 
 
    
 
    
-torch.backends.nnpack.set_flags(_enabled)[source][source]
    
+torch.backends.nnpack.set_flags(_enabled)[source][source]
 
    Set if nnpack is enabled globally
    
 
    
 
@@ -1264,7 +1264,7 @@
 
    torch.backends.openmp
    
 
    
 
    
-torch.backends.openmp.is_available()[source][source]
    
+torch.backends.openmp.is_available()[source][source]
 
    Return whether PyTorch is built with OpenMP support.
    
 
    
 
@@ -1273,7 +1273,7 @@
 
    torch.backends.opt_einsum
    
 
    
 
    
-torch.backends.opt_einsum.is_available()[source][source]
    
+torch.backends.opt_einsum.is_available()[source][source]
 
    Return a bool indicating if opt_einsum is currently available.
    
 
    You must install opt-einsum in order for torch to automatically optimize einsum. To
 make opt-einsum available, you can install it along with torch: pip install torch[opt-einsum]
@@ -1289,7 +1289,7 @@
 
 
    
 
    
-torch.backends.opt_einsum.get_opt_einsum()[source][source]
    
+torch.backends.opt_einsum.get_opt_einsum()[source][source]
 
    Return the opt_einsum package if opt_einsum is currently available, else None.
    
 
    
 
    Return type
    
diff --git a/2.7/benchmark_utils.html b/2.7/benchmark_utils.html
index 790eff7a542..3f4509e0228 100644
--- a/2.7/benchmark_utils.html
+++ b/2.7/benchmark_utils.html
@@ -594,7 +594,7 @@
 
    Benchmark Utils - torch.utils.benchmark
    
 
    
 
    
-class torch.utils.benchmark.Timer(stmt='pass', setup='pass', global_setup='', timer=<built-in function perf_counter>, globals=None, label=None, sub_label=None, description=None, env=None, num_threads=1, language=Language.PYTHON)[source][source]
    
+class torch.utils.benchmark.Timer(stmt='pass', setup='pass', global_setup='', timer=<built-in function perf_counter>, globals=None, label=None, sub_label=None, description=None, env=None, num_threads=1, language=Language.PYTHON)[source][source]
 
    Helper class for measuring execution time of PyTorch statements.
    
 
    For a full tutorial on how to use this class, see:
 https://pytorch.org/tutorials/recipes/recipes/benchmark.html
    
@@ -698,7 +698,7 @@
 
    
 
    
 
    
-adaptive_autorange(threshold=0.1, *, min_run_time=0.01, max_run_time=10.0, callback=None)[source][source]
    
+adaptive_autorange(threshold=0.1, *, min_run_time=0.01, max_run_time=10.0, callback=None)[source][source]
 
    Similar to blocked_autorange but also checks for variablility in measurements
 and repeats until iqr/median is smaller than threshold or max_run_time is reached.
    
 
    At a high level, adaptive_autorange executes the following pseudo-code:
    
@@ -739,7 +739,7 @@
 
 
    
 
    
-blocked_autorange(callback=None, min_run_time=0.2)[source][source]
    
+blocked_autorange(callback=None, min_run_time=0.2)[source][source]
 
    Measure many replicates while keeping timer overhead to a minimum.
    
 
    At a high level, blocked_autorange executes the following pseudo-code:
    
 
    `setup`
@@ -784,7 +784,7 @@
 
 
    
 
    
-collect_callgrind(number: int, *, repeats: None, collect_baseline: bool, retain_out_file: bool) → CallgrindStats[source][source]
    
+collect_callgrind(number: int, *, repeats: None, collect_baseline: bool, retain_out_file: bool) → CallgrindStats[source][source]
 
    
 collect_callgrind(number: int, *, repeats: int, collect_baseline: bool, retain_out_file: bool) → tuple[torch.utils.benchmark.utils.valgrind_wrapper.timer_interface.CallgrindStats, ...]
    
 
    Collect instruction counts using Callgrind.
    
@@ -819,7 +819,7 @@
 
 
    
 
    
-timeit(number=1000000)[source][source]
    
+timeit(number=1000000)[source][source]
 
    Mirrors the semantics of timeit.Timer.timeit().
    
 
    Execute the main statement (stmt) number times.
 https://docs.python.org/3/library/timeit.html#timeit.Timer.timeit
    
@@ -834,7 +834,7 @@
 
 
    
 
    
-class torch.utils.benchmark.Measurement(number_per_run, raw_times, task_spec, metadata=None)[source][source]
    
+class torch.utils.benchmark.Measurement(number_per_run, raw_times, task_spec, metadata=None)[source][source]
 
    The result of a Timer measurement.
    
 
    This class stores one or more measurements of a given statement. It is
 serializable and provides several convenience methods
@@ -843,7 +843,7 @@
 
    
 
    
 
    
-static merge(measurements)[source][source]
    
+static merge(measurements)[source][source]
 
    Convenience method for merging replicates.
    
 
    Merge will extrapolate times to number_per_run=1 and will not
 transfer any metadata. (Since it might differ between replicates)
    
@@ -873,7 +873,7 @@
 
 
    
 
    
-class torch.utils.benchmark.CallgrindStats(task_spec, number_per_run, built_with_debug_symbols, baseline_inclusive_stats, baseline_exclusive_stats, stmt_inclusive_stats, stmt_exclusive_stats, stmt_callgrind_out)[source][source]
    
+class torch.utils.benchmark.CallgrindStats(task_spec, number_per_run, built_with_debug_symbols, baseline_inclusive_stats, baseline_exclusive_stats, stmt_inclusive_stats, stmt_exclusive_stats, stmt_callgrind_out)[source][source]
 
    Top level container for Callgrind results collected by Timer.
    
 
    Manipulation is generally done using the FunctionCounts class, which is
 obtained by calling CallgrindStats.stats(…). Several convenience
@@ -883,7 +883,7 @@
 
    
 
    
 
    
-as_standardized()[source][source]
    
+as_standardized()[source][source]
 
    Strip library names and some prefixes from function strings.
    
 
    When comparing two different sets of instruction counts, on stumbling
 block can be path prefixes. Callgrind includes the full filepath
@@ -912,7 +912,7 @@
 
 
    
 
    
-counts(*, denoise=False)[source][source]
    
+counts(*, denoise=False)[source][source]
 
    Returns the total number of instructions executed.
    
 
    See FunctionCounts.denoise() for an explanation of the denoise arg.
    
 
    
@@ -924,7 +924,7 @@
 
 
    
 
    
-delta(other, inclusive=False)[source][source]
    
+delta(other, inclusive=False)[source][source]
 
    Diff two sets of counts.
    
 
    One common reason to collect instruction counts is to determine the
 the effect that a particular change will have on the number of instructions
@@ -942,7 +942,7 @@
 
 
    
 
    
-stats(inclusive=False)[source][source]
    
+stats(inclusive=False)[source][source]
 
    Returns detailed function counts.
    
 
    Conceptually, the FunctionCounts returned can be thought of as a tuple
 of (count, path_and_function_name) tuples.
    
@@ -962,7 +962,7 @@
 
 
    
 
    
-class torch.utils.benchmark.FunctionCounts(_data, inclusive, truncate_rows=True, _linewidth=None)[source][source]
    
+class torch.utils.benchmark.FunctionCounts(_data, inclusive, truncate_rows=True, _linewidth=None)[source][source]
 
    Container for manipulating Callgrind results.
    
 
    
 
    It supports:
      
@@ -979,7 +979,7 @@
 
    
 
    
 
    
-denoise()[source][source]
    
+denoise()[source][source]
 
    Remove known noisy instructions.
    
 
    Several instructions in the CPython interpreter are rather noisy. These
 instructions involve unicode to dictionary lookups which Python uses to
@@ -995,7 +995,7 @@
 
 
    
 
    
-filter(filter_fn)[source][source]
    
+filter(filter_fn)[source][source]
 
    Keep only the elements where filter_fn applied to function name returns True.
    
 
    
 
    Return type
    
@@ -1006,7 +1006,7 @@
 
 
    
 
    
-transform(map_fn)[source][source]
    
+transform(map_fn)[source][source]
 
    Apply map_fn to all of the function names.
    
 
    This can be used to regularize function names (e.g. stripping irrelevant
 parts of the file path), coalesce entries by mapping multiple functions
@@ -1022,7 +1022,7 @@
 
 
    
 
    
-class torch.utils.benchmark.Compare(results)[source][source]
    
+class torch.utils.benchmark.Compare(results)[source][source]
 
    Helper class for displaying the results of many measurements in a
 formatted table.
    
 
    The table format is based on the information fields provided in
@@ -1038,33 +1038,33 @@
 
    
 
    
 
    
-colorize(rowwise=False)[source][source]
    
+colorize(rowwise=False)[source][source]
 
    Colorize formatted table.
    
 
    Colorize columnwise by default.
    
 
    
 
 
    
 
    
-extend_results(results)[source][source]
    
+extend_results(results)[source][source]
 
    Append results to already stored ones.
    
 
    All added results must be instances of Measurement.
    
 
    
 
 
    
 
    
-highlight_warnings()[source][source]
    
+highlight_warnings()[source][source]
 
    Enables warning highlighting when building formatted table.
    
 
    
 
 
    
 
    
-print()[source][source]
    
+print()[source][source]
 
    Print formatted table
    
 
    
 
 
    
 
    
-trim_significant_figures()[source][source]
    
+trim_significant_figures()[source][source]
 
    Enables trimming of significant figures when building the formatted table.
    
 
    
 
diff --git a/2.7/checkpoint.html b/2.7/checkpoint.html
index a4d586a773c..7a98e2c4391 100644
--- a/2.7/checkpoint.html
+++ b/2.7/checkpoint.html
@@ -624,7 +624,7 @@ 
    torch.utils.checkpoint
 
    
 
    
-torch.utils.checkpoint.checkpoint(function, *args, use_reentrant=None, context_fn=<function noop_context_fn>, determinism_check='default', debug=False, **kwargs)[source][source]
    
+torch.utils.checkpoint.checkpoint(function, *args, use_reentrant=None, context_fn=<function noop_context_fn>, determinism_check='default', debug=False, **kwargs)[source][source]
 
    Checkpoint a model or part of the model.
    
 
    Activation checkpointing is a technique that trades compute for memory.
 Instead of keeping tensors needed for backward alive until they are used in
@@ -736,7 +736,7 @@ 
    torch.utils.checkpoint
 
    
-torch.utils.checkpoint.checkpoint_sequential(functions, segments, input, use_reentrant=None, **kwargs)[source][source]
    
+torch.utils.checkpoint.checkpoint_sequential(functions, segments, input, use_reentrant=None, **kwargs)[source][source]
 
    Checkpoint a sequential model to save memory.
    
 
    Sequential models execute a list of modules/functions in order
 (sequentially). Therefore, we can divide such a model in various segments
@@ -786,7 +786,7 @@ 
    torch.utils.checkpoint
 
    
-torch.utils.checkpoint.set_checkpoint_debug_enabled(enabled)[source][source]
    
+torch.utils.checkpoint.set_checkpoint_debug_enabled(enabled)[source][source]
 
    Context manager that sets whether checkpoint should print additional debug
 information when running. See the debug flag for
 checkpoint() for more information. Note that
@@ -802,7 +802,7 @@ 
    torch.utils.checkpoint
 
    
-class torch.utils.checkpoint.CheckpointPolicy(value)[source][source]
    
+class torch.utils.checkpoint.CheckpointPolicy(value)[source][source]
 
    Enum for specifying the policy for checkpointing during backpropagation.
    
 
    The following policies are supported:
    
 
      
@@ -826,7 +826,7 @@ 
      torch.utils.checkpoint
 
      
-class torch.utils.checkpoint.SelectiveCheckpointContext(*, is_recompute)[source][source]
      
+class torch.utils.checkpoint.SelectiveCheckpointContext(*, is_recompute)[source][source]
 
      Context passed to policy function during selective checkpointing.
      
 
      This class is used to pass relevant metadata to the policy function during
 selective checkpointing. The metadata includes whether the current invocation
@@ -849,7 +849,7 @@ 
      torch.utils.checkpoint
 
      
-torch.utils.checkpoint.create_selective_checkpoint_contexts(policy_fn_or_list, allow_cache_entry_mutation=False)[source][source]
      
+torch.utils.checkpoint.create_selective_checkpoint_contexts(policy_fn_or_list, allow_cache_entry_mutation=False)[source][source]
 
      Helper to avoid recomputing certain ops during activation checkpointing.
      
 
      Use this with torch.utils.checkpoint.checkpoint to control which
 operations are recomputed during the backward pass.
      
diff --git a/2.7/cond.html b/2.7/cond.html
index 8e343bf2ab8..d30187996fa 100644
--- a/2.7/cond.html
+++ b/2.7/cond.html
@@ -760,7 +760,7 @@ 
      Invariants of torch.ops.higher_order.cond
      
 
      
 
      
-torch._higher_order_ops.cond.cond(pred, true_fn, false_fn, operands=())[source]
      
+torch._higher_order_ops.cond.cond(pred, true_fn, false_fn, operands=())[source]
 
      Conditionally applies true_fn or false_fn.
      
 
      
 
      Warning
      
diff --git a/2.7/config_mod.html b/2.7/config_mod.html
index e9d00a02013..4eb5669c95e 100644
--- a/2.7/config_mod.html
+++ b/2.7/config_mod.html
@@ -594,7 +594,7 @@
 
      torch.__config__
      
 
      
 
      
-torch.__config__.show()[source][source]
      
+torch.__config__.show()[source][source]
 
      Return a human-readable string with descriptions of the
 configuration of PyTorch.
      
 
      
@@ -606,7 +606,7 @@
 
 
      
 
      
-torch.__config__.parallel_info()[source][source]
      
+torch.__config__.parallel_info()[source][source]
 
      Returns detailed string with parallelization settings
      
 
      
 
      Return type
      
diff --git a/2.7/cpp_extension.html b/2.7/cpp_extension.html
index d13c71708ba..8e2a6054058 100644
--- a/2.7/cpp_extension.html
+++ b/2.7/cpp_extension.html
@@ -594,7 +594,7 @@
 
      torch.utils.cpp_extension
      
 
      
 
      
-torch.utils.cpp_extension.CppExtension(name, sources, *args, **kwargs)[source][source]
      
+torch.utils.cpp_extension.CppExtension(name, sources, *args, **kwargs)[source][source]
 
      Create a setuptools.Extension for C++.
      
 
      Convenience method that creates a setuptools.Extension with the
 bare minimum (but often sufficient) arguments to build a C++ extension.
      
@@ -639,7 +639,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.CUDAExtension(name, sources, *args, **kwargs)[source][source]
      
+torch.utils.cpp_extension.CUDAExtension(name, sources, *args, **kwargs)[source][source]
 
      Create a setuptools.Extension for CUDA/C++.
      
 
      Convenience method that creates a setuptools.Extension with the
 bare minimum (but often sufficient) arguments to build a CUDA/C++
@@ -744,7 +744,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.SyclExtension(name, sources, *args, **kwargs)[source][source]
      
+torch.utils.cpp_extension.SyclExtension(name, sources, *args, **kwargs)[source][source]
 
      Creates a setuptools.Extension for SYCL/C++.
      
 
      Convenience method that creates a setuptools.Extension with the
 bare minimum (but often sufficient) arguments to build a SYCL/C++
@@ -796,7 +796,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.BuildExtension(*args, **kwargs)[source][source]
      
+torch.utils.cpp_extension.BuildExtension(*args, **kwargs)[source][source]
 
      A custom setuptools build extension .
      
 
      This setuptools.build_ext subclass takes care of passing the
 minimum required compiler flags (e.g. -std=c++17) as well as mixed
@@ -824,7 +824,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.load(name, sources, extra_cflags=None, extra_cuda_cflags=None, extra_sycl_cflags=None, extra_ldflags=None, extra_include_paths=None, build_directory=None, verbose=False, with_cuda=None, with_sycl=None, is_python_module=True, is_standalone=False, keep_intermediates=True)[source][source]
      
+torch.utils.cpp_extension.load(name, sources, extra_cflags=None, extra_cuda_cflags=None, extra_sycl_cflags=None, extra_ldflags=None, extra_include_paths=None, build_directory=None, verbose=False, with_cuda=None, with_sycl=None, is_python_module=True, is_standalone=False, keep_intermediates=True)[source][source]
 
      Load a PyTorch C++ extension just-in-time (JIT).
      
 
      To load an extension, a Ninja build file is emitted, which is used to
 compile the given sources into a dynamic library. This library is
@@ -924,7 +924,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.load_inline(name, cpp_sources, cuda_sources=None, sycl_sources=None, functions=None, extra_cflags=None, extra_cuda_cflags=None, extra_sycl_cflags=None, extra_ldflags=None, extra_include_paths=None, build_directory=None, verbose=False, with_cuda=None, with_sycl=None, is_python_module=True, with_pytorch_error_handling=True, keep_intermediates=True, use_pch=False)[source][source]
      
+torch.utils.cpp_extension.load_inline(name, cpp_sources, cuda_sources=None, sycl_sources=None, functions=None, extra_cflags=None, extra_cuda_cflags=None, extra_sycl_cflags=None, extra_ldflags=None, extra_include_paths=None, build_directory=None, verbose=False, with_cuda=None, with_sycl=None, is_python_module=True, with_pytorch_error_handling=True, keep_intermediates=True, use_pch=False)[source][source]
 
      Load a PyTorch C++ extension just-in-time (JIT) from string sources.
      
 
      This function behaves exactly like load(), but takes its sources as
 strings rather than filenames. These strings are stored to files in the
@@ -1017,7 +1017,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.include_paths(device_type='cpu')[source][source]
      
+torch.utils.cpp_extension.include_paths(device_type='cpu')[source][source]
 
      Get the include paths required to build a C++ or CUDA or SYCL extension.
      
 
      
 
      Parameters
      
@@ -1034,7 +1034,7 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.get_compiler_abi_compatibility_and_version(compiler)[source][source]
      
+torch.utils.cpp_extension.get_compiler_abi_compatibility_and_version(compiler)[source][source]
 
      Determine if the given compiler is ABI-compatible with PyTorch alongside its version.
      
 
      
 
      Parameters
      
@@ -1053,13 +1053,13 @@ 
      torch.utils.cpp_extension
 
      
-torch.utils.cpp_extension.verify_ninja_availability()[source][source]
      
+torch.utils.cpp_extension.verify_ninja_availability()[source][source]
 
      Raise RuntimeError if ninja build system is not available on the system, does nothing otherwise.
      
 
      
 
 
      
 
      
-torch.utils.cpp_extension.is_ninja_available()[source][source]
      
+torch.utils.cpp_extension.is_ninja_available()[source][source]
 
      Return True if the ninja build system is available on the system, False otherwise.
      
 
      
 
diff --git a/2.7/cuda._sanitizer.html b/2.7/cuda._sanitizer.html
index 356881c30d0..ff1c5837343 100644
--- a/2.7/cuda._sanitizer.html
+++ b/2.7/cuda._sanitizer.html
@@ -696,7 +696,7 @@ 
      Usage
 
      API Reference
      
 
      
 
      
-torch.cuda._sanitizer.enable_cuda_sanitizer()[source][source]
      
+torch.cuda._sanitizer.enable_cuda_sanitizer()[source][source]
 
      Enable CUDA Sanitizer.
      
 
      The sanitizer will begin to analyze low-level CUDA calls invoked by torch functions
 for synchronization errors. All data races found will be printed to the standard
diff --git a/2.7/cuda.html b/2.7/cuda.html
index dddd9188f48..b1a1f0542cb 100644
--- a/2.7/cuda.html
+++ b/2.7/cuda.html
@@ -891,7 +891,7 @@ 
      Graphs (beta)
 
      
-class torch.cuda.use_mem_pool(pool, device=None)[source][source]
      
+class torch.cuda.use_mem_pool(pool, device=None)[source][source]
 
      A context manager that routes allocations to a given pool.
      
 
      
 
      Parameters
      
diff --git a/2.7/cuda.tunable.html b/2.7/cuda.tunable.html
index 7e64d0816c1..157b5ed2a55 100644
--- a/2.7/cuda.tunable.html
+++ b/2.7/cuda.tunable.html
@@ -761,7 +761,7 @@ 
      Environment Variable Interface
      
 
      
 
      
-torch.cuda.tunable.enable(val=True)[source][source]
      
+torch.cuda.tunable.enable(val=True)[source][source]
 
      This is the big on/off switch for all TunableOp implementations.
      
 
      
 
      
@@ -769,7 +769,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.is_enabled()[source][source]
      
+torch.cuda.tunable.is_enabled()[source][source]
 
      Returns whether the TunableOp feature is enabled.
      
 
      
 
      Return type
      
@@ -780,7 +780,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.tuning_enable(val=True)[source][source]
      
+torch.cuda.tunable.tuning_enable(val=True)[source][source]
 
      Enable tuning of TunableOp implementations.
      
 
      When enabled, if a tuned entry isn’t found, run the tuning step and record
 the entry.
      
@@ -790,7 +790,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.tuning_is_enabled()[source][source]
      
+torch.cuda.tunable.tuning_is_enabled()[source][source]
 
      Returns whether TunableOp implementations can be tuned.
      
 
      
 
      Return type
      
@@ -801,7 +801,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.record_untuned_enable(val=True)[source][source]
      
+torch.cuda.tunable.record_untuned_enable(val=True)[source][source]
 
      Enable recording untuned of TunableOp perations for offline tuning.
      
 
      When enabled, if a tuned entry isn’t found, write it to the untuned file.
      
 
      
@@ -810,7 +810,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.record_untuned_is_enabled()[source][source]
      
+torch.cuda.tunable.record_untuned_is_enabled()[source][source]
 
      Returns whether TunableOp operations are recorded for offline tuning.
      
 
      
 
      Return type
      
@@ -821,7 +821,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.set_max_tuning_duration(duration)[source][source]
      
+torch.cuda.tunable.set_max_tuning_duration(duration)[source][source]
 
      Set max time in milliseconds to spend tuning a given solution.
      
 
      If both max tuning duration and iterations are set, the smaller of the two
 will be honored. At minimum 1 tuning iteration will always be run.
      
@@ -831,7 +831,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_max_tuning_duration()[source][source]
      
+torch.cuda.tunable.get_max_tuning_duration()[source][source]
 
      Get max time to spend tuning a given solution.
      
 
      
 
      Return type
      
@@ -842,7 +842,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.set_max_tuning_iterations(iterations)[source][source]
      
+torch.cuda.tunable.set_max_tuning_iterations(iterations)[source][source]
 
      Set max number of iterations to spend tuning a given solution.
      
 
      If both max tuning duration and iterations are set, the smaller of the two
 will be honored. At minimum 1 tuning iteration will always be run.
      
@@ -852,7 +852,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_max_tuning_iterations()[source][source]
      
+torch.cuda.tunable.get_max_tuning_iterations()[source][source]
 
      Get max iterations to spend tuning a given solution.
      
 
      
 
      Return type
      
@@ -863,7 +863,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.set_filename(filename, insert_device_ordinal=False)[source][source]
      
+torch.cuda.tunable.set_filename(filename, insert_device_ordinal=False)[source][source]
 
      Set the filename to use for input/output of tuning results.
      
 
      If insert_device_ordinal is True then the current device ordinal
 will be added to the given filename automatically. This can be used in a
@@ -874,7 +874,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_filename()[source][source]
      
+torch.cuda.tunable.get_filename()[source][source]
 
      Get the results filename.
      
 
      
 
      Return type
      
@@ -885,7 +885,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_results()[source][source]
      
+torch.cuda.tunable.get_results()[source][source]
 
      Return all TunableOp results.
      
 
      
 
      Return type
      
@@ -896,7 +896,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_validators()[source][source]
      
+torch.cuda.tunable.get_validators()[source][source]
 
      Return the TunableOp validators.
      
 
      
 
      Return type
      
@@ -907,7 +907,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.write_file_on_exit(val)[source][source]
      
+torch.cuda.tunable.write_file_on_exit(val)[source][source]
 
      During Tuning Context destruction, write file to disk.
      
 
      This is useful as a final flush of your results to disk if your application
 terminates as result of normal operation or an error. Manual flushing of
@@ -918,7 +918,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.write_file(filename=None)[source][source]
      
+torch.cuda.tunable.write_file(filename=None)[source][source]
 
      Write results to a CSV file.
      
 
      If filename is not given, get_filename() is called.
      
 
      
@@ -930,7 +930,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.read_file(filename=None)[source][source]
      
+torch.cuda.tunable.read_file(filename=None)[source][source]
 
      Read results from a TunableOp CSV file.
      
 
      If filename is not given, get_filename() is called.
      
 
      
@@ -942,7 +942,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.tune_gemm_in_file(filename)[source][source]
      
+torch.cuda.tunable.tune_gemm_in_file(filename)[source][source]
 
      tune GEMM in file.
      
 
      
 
      
@@ -950,7 +950,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.mgpu_tune_gemm_in_file(filename_pattern, num_gpus)[source][source]
      
+torch.cuda.tunable.mgpu_tune_gemm_in_file(filename_pattern, num_gpus)[source][source]
 
      Process one or more files and distribute work over one or more GPUs.
      
 
      
 
      
@@ -958,7 +958,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.set_rotating_buffer_size(buffer_size)[source][source]
      
+torch.cuda.tunable.set_rotating_buffer_size(buffer_size)[source][source]
 
      Set rotating buffer size to this value in MB, if the buffer size is greater than zero.
      
 
      If less than zero, query L2 cache size. If equal to zero, means deactivate rotating buffer.
      
 
      
@@ -967,7 +967,7 @@ 
      API Reference
 
      
-torch.cuda.tunable.get_rotating_buffer_size()[source][source]
      
+torch.cuda.tunable.get_rotating_buffer_size()[source][source]
 
      Get the rotating buffer size in kilobytes.
      
 
      
 
      Return type
      
diff --git a/2.7/data.html b/2.7/data.html
index c1cd8cc69b7..779a2f329c9 100644
--- a/2.7/data.html
+++ b/2.7/data.html
@@ -960,7 +960,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.DataLoader(dataset, batch_size=1, shuffle=None, sampler=None, batch_sampler=None, num_workers=0, collate_fn=None, pin_memory=False, drop_last=False, timeout=0, worker_init_fn=None, multiprocessing_context=None, generator=None, *, prefetch_factor=None, persistent_workers=False, pin_memory_device='', in_order=True)[source][source]
      
+class torch.utils.data.DataLoader(dataset, batch_size=1, shuffle=None, sampler=None, batch_sampler=None, num_workers=0, collate_fn=None, pin_memory=False, drop_last=False, timeout=0, worker_init_fn=None, multiprocessing_context=None, generator=None, *, prefetch_factor=None, persistent_workers=False, pin_memory_device='', in_order=True)[source][source]
 
      Data loader combines a dataset and a sampler, and provides an iterable over the given dataset.
      
 
      The DataLoader supports both map-style and
 iterable-style datasets with single- or multi-process loading, customizing
@@ -1061,7 +1061,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.Dataset[source][source]
      
+class torch.utils.data.Dataset[source][source]
 
      An abstract class representing a Dataset.
      
 
      All datasets that represent a map from keys to data samples should subclass
 it. All subclasses should overwrite __getitem__(), supporting fetching a
@@ -1082,7 +1082,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.IterableDataset[source][source]
      
+class torch.utils.data.IterableDataset[source][source]
 
      An iterable Dataset.
      
 
      All datasets that represent an iterable of data samples should subclass it.
 Such form of datasets is particularly useful when data come from a stream.
      
@@ -1184,7 +1184,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.TensorDataset(*tensors)[source][source]
      
+class torch.utils.data.TensorDataset(*tensors)[source][source]
 
      Dataset wrapping tensors.
      
 
      Each sample will be retrieved by indexing tensors along the first dimension.
      
 
      
@@ -1196,7 +1196,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.StackDataset(*args, **kwargs)[source][source]
      
+class torch.utils.data.StackDataset(*args, **kwargs)[source][source]
 
      Dataset as a stacking of multiple datasets.
      
 
      This class is useful to assemble different parts of complex input data, given as datasets.
      
 
      Example
      
@@ -1220,7 +1220,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.ConcatDataset(datasets)[source][source]
      
+class torch.utils.data.ConcatDataset(datasets)[source][source]
 
      Dataset as a concatenation of multiple datasets.
      
 
      This class is useful to assemble different existing datasets.
      
 
      
@@ -1232,7 +1232,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.ChainDataset(datasets)[source][source]
      
+class torch.utils.data.ChainDataset(datasets)[source][source]
 
      Dataset for chaining multiple IterableDataset s.
      
 
      This class is useful to assemble different existing dataset streams. The
 chaining operation is done on-the-fly, so concatenating large-scale
@@ -1246,7 +1246,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.Subset(dataset, indices)[source][source]
      
+class torch.utils.data.Subset(dataset, indices)[source][source]
 
      Subset of a dataset at specified indices.
      
 
      
 
      Parameters
      
@@ -1260,7 +1260,7 @@ 
      Memory Pinning
 
      
-torch.utils.data._utils.collate.collate(batch, *, collate_fn_map=None)[source][source]
      
+torch.utils.data._utils.collate.collate(batch, *, collate_fn_map=None)[source][source]
 
      General collate function that handles collection type of element within each batch.
      
 
      The function also opens function registry to deal with specific element types. default_collate_fn_map
 provides default collate functions for tensors, numpy arrays, numbers and strings.
      
@@ -1295,7 +1295,7 @@ 
      Memory Pinning
 
      
-torch.utils.data.default_collate(batch)[source][source]
      
+torch.utils.data.default_collate(batch)[source][source]
 
      Take in a batch of data and put the elements within the batch into a tensor with an additional outer dimension - batch size.
      
 
      The exact output type can be a torch.Tensor, a Sequence of torch.Tensor, a
 Collection of torch.Tensor, or left unchanged, depending on the input type.
@@ -1361,7 +1361,7 @@ 
      Memory Pinning
 
      
-torch.utils.data.default_convert(data)[source][source]
      
+torch.utils.data.default_convert(data)[source][source]
 
      Convert each NumPy array element into a torch.Tensor.
      
 
      If the input is a Sequence, Collection, or Mapping, it tries to convert each element inside to a torch.Tensor.
 If the input is not an NumPy array, it is left unchanged.
@@ -1396,7 +1396,7 @@ 
      Memory Pinning
 
      
-torch.utils.data.get_worker_info()[source][source]
      
+torch.utils.data.get_worker_info()[source][source]
 
      Returns the information about the current
 DataLoader iterator worker process.
      
 
      When called in a worker, this returns an object guaranteed to have the
@@ -1430,7 +1430,7 @@ 
      Memory Pinning
 
      
-torch.utils.data.random_split(dataset, lengths, generator=<torch._C.Generator object>)[source][source]
      
+torch.utils.data.random_split(dataset, lengths, generator=<torch._C.Generator object>)[source][source]
 
      Randomly split a dataset into non-overlapping new datasets of given lengths.
      
 
      If a list of fractions that sum up to 1 is given,
 the lengths will be computed automatically as
@@ -1462,7 +1462,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.Sampler(data_source=None)[source][source]
      
+class torch.utils.data.Sampler(data_source=None)[source][source]
 
      Base class for all Samplers.
      
 
      Every Sampler subclass has to provide an __iter__() method, providing a
 way to iterate over indices or lists of indices (batches) of dataset elements,
@@ -1509,7 +1509,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.SequentialSampler(data_source)[source][source]
      
+class torch.utils.data.SequentialSampler(data_source)[source][source]
 
      Samples elements sequentially, always in the same order.
      
 
      
 
      Parameters
      
@@ -1520,7 +1520,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.RandomSampler(data_source, replacement=False, num_samples=None, generator=None)[source][source]
      
+class torch.utils.data.RandomSampler(data_source, replacement=False, num_samples=None, generator=None)[source][source]
 
      Samples elements randomly. If without replacement, then sample from a shuffled dataset.
      
 
      If with replacement, then user can specify num_samples to draw.
      
 
      
@@ -1537,7 +1537,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.SubsetRandomSampler(indices, generator=None)[source][source]
      
+class torch.utils.data.SubsetRandomSampler(indices, generator=None)[source][source]
 
      Samples elements randomly from a given list of indices, without replacement.
      
 
      
 
      Parameters
      
@@ -1551,7 +1551,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.WeightedRandomSampler(weights, num_samples, replacement=True, generator=None)[source][source]
      
+class torch.utils.data.WeightedRandomSampler(weights, num_samples, replacement=True, generator=None)[source][source]
 
      Samples elements from [0,..,len(weights)-1] with given probabilities (weights).
      
 
      
 
      Parameters
      
@@ -1576,7 +1576,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.BatchSampler(sampler, batch_size, drop_last)[source][source]
      
+class torch.utils.data.BatchSampler(sampler, batch_size, drop_last)[source][source]
 
      Wraps another sampler to yield a mini-batch of indices.
      
 
      
 
      Parameters
      
@@ -1599,7 +1599,7 @@ 
      Memory Pinning
 
      
-class torch.utils.data.distributed.DistributedSampler(dataset, num_replicas=None, rank=None, shuffle=True, seed=0, drop_last=False)[source][source]
      
+class torch.utils.data.distributed.DistributedSampler(dataset, num_replicas=None, rank=None, shuffle=True, seed=0, drop_last=False)[source][source]
 
      Sampler that restricts data loading to a subset of the dataset.
      
 
      It is especially useful in conjunction with
 torch.nn.parallel.DistributedDataParallel. In such a case, each
diff --git a/2.7/ddp_comm_hooks.html b/2.7/ddp_comm_hooks.html
index ffffe2972cc..01f0b6d23f1 100644
--- a/2.7/ddp_comm_hooks.html
+++ b/2.7/ddp_comm_hooks.html
@@ -694,7 +694,7 @@ 
      Default Communication Hooksbucket is a torch.distributed.GradBucket object.
      
 
      
 
      
-torch.distributed.algorithms.ddp_comm_hooks.default_hooks.allreduce_hook(process_group, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.default_hooks.allreduce_hook(process_group, bucket)[source][source]
 
      Call allreduce using GradBucket tensors.
      
 
      Once gradient tensors are aggregated across all workers, its then
 callback takes the mean and returns the result.
      
@@ -718,7 +718,7 @@ 
      Default Communication Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.default_hooks.fp16_compress_hook(process_group, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.default_hooks.fp16_compress_hook(process_group, bucket)[source][source]
 
      Compress by casting GradBucket to torch.float16 divided by process group size.
      
 
      This DDP communication hook implements a simple gradient compression
 approach that casts GradBucket tensor to half-precision floating-point format (torch.float16)
@@ -740,7 +740,7 @@ 
      Default Communication Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.default_hooks.bf16_compress_hook(process_group, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.default_hooks.bf16_compress_hook(process_group, bucket)[source][source]
 
      Warning: This API is experimental, and it requires NCCL version later than 2.9.6.
      
 
      This DDP communication hook implements a simple gradient compression
 approach that casts GradBucket tensor to half-precision
@@ -765,7 +765,7 @@ 
      Default Communication Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.default_hooks.fp16_compress_wrapper(hook)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.default_hooks.fp16_compress_wrapper(hook)[source][source]
 
      Cast input tensor to torch.float16, cast result of hook back to input dtype.
      
 
      This wrapper casts the input gradient tensor of a given DDP communication hook to half-precision
 floating point format (torch.float16), and casts the resulting tensor of the given hook back to
@@ -787,7 +787,7 @@ 
      Default Communication Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.default_hooks.bf16_compress_wrapper(hook)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.default_hooks.bf16_compress_wrapper(hook)[source][source]
 
      Warning: This API is experimental, and it requires NCCL version later than 2.9.6.
      
 
      This wrapper casts the input gradient tensor of a given DDP communication hook to half-precision
 Brain floating point format <https://en.wikipedia.org/wiki/Bfloat16_floating-point_format> `_  (``torch.bfloat16`),
@@ -820,7 +820,7 @@ 
      PowerSGD Communication Hook
      
 
      
 
      
-class torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.PowerSGDState(process_group, matrix_approximation_rank=1, start_powerSGD_iter=1000, min_compression_rate=2, use_error_feedback=True, warm_start=True, orthogonalization_epsilon=0, random_seed=0, compression_stats_logging_frequency=10000, batch_tensors_with_same_shape=False)[source][source]
      
+class torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.PowerSGDState(process_group, matrix_approximation_rank=1, start_powerSGD_iter=1000, min_compression_rate=2, use_error_feedback=True, warm_start=True, orthogonalization_epsilon=0, random_seed=0, compression_stats_logging_frequency=10000, batch_tensors_with_same_shape=False)[source][source]
 
      Store both the algorithm’s hyperparameters and internal state for all gradients during training.
      
 
      Particularly, matrix_approximation_rank and start_powerSGD_iter are the main hyperparameters that should be tuned by the user.
 For performance, we suggest to keep binary hyperparameters use_error_feedback and warm_start on.
      
@@ -872,7 +872,7 @@ 
      PowerSGD Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.powerSGD_hook(state, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.powerSGD_hook(state, bucket)[source][source]
 
      Implement PowerSGD algorithm.
      
 
      This DDP communication hook implements PowerSGD gradient compression
 algorithm described in the paper.
@@ -937,7 +937,7 @@ 
      PowerSGD Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.batched_powerSGD_hook(state, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.batched_powerSGD_hook(state, bucket)[source][source]
 
      Implement simplified PowerSGD algorithm.
      
 
      This DDP communication hook implements a simplified PowerSGD gradient compression
 algorithm described in the paper.
@@ -1005,7 +1005,7 @@ 
      Debugging Communication Hooks
 
      
-torch.distributed.algorithms.ddp_comm_hooks.debugging_hooks.noop_hook(_, bucket)[source][source]
      
+torch.distributed.algorithms.ddp_comm_hooks.debugging_hooks.noop_hook(_, bucket)[source][source]
 
      Return a future that wraps the input, so it is a no-op that does not incur any communication overheads.
      
 
      This hook should only be used for headroom analysis of allreduce optimization,
 instead of the normal gradient synchronization.
@@ -1043,12 +1043,12 @@ 
      Checkpointing of Communication HooksPowerSGDState has __setstate__ and __getstate__ implemented and can be used as a reference.
      
 
      
 
      
-class torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.PowerSGDState[source][source]
      
+class torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook.PowerSGDState[source][source]
 
      
 
      
 
      
 
      
-__getstate__()[source][source]
      
+__getstate__()[source][source]
 
      Return a Dict[str, Any] which will be pickled and saved.
      
 
      process_group is not serializable and excluded from
 a returned state.
      
@@ -1056,7 +1056,7 @@ 
      Checkpointing of Communication Hooks
 
      
-__setstate__(state)[source][source]
      
+__setstate__(state)[source][source]
 
      Take a provided state and set to this PowerSGDState instance.
      
 
      process_group is set to default.
      
 
      
diff --git a/2.7/distributed.algorithms.join.html b/2.7/distributed.algorithms.join.html
index d69a8a21535..f7e61b30e77 100644
--- a/2.7/distributed.algorithms.join.html
+++ b/2.7/distributed.algorithms.join.html
@@ -598,7 +598,7 @@ 
      Generic Join Context ManagerDistributed Training with Uneven Inputs Using the Join Context Manager.
      
 
      
 
      
-class torch.distributed.algorithms.Join(joinables, enable=True, throw_on_early_termination=False, **kwargs)[source][source]
      
+class torch.distributed.algorithms.Join(joinables, enable=True, throw_on_early_termination=False, **kwargs)[source][source]
 
      This class defines the generic join context manager, which allows custom hooks to be called after a process joins.
      
 
      These hooks should shadow the
 collective communications of non-joined processes to prevent hanging and
@@ -661,7 +661,7 @@ 
      Generic Join Context Manager
 
      
-static notify_join_context(joinable)[source][source]
      
+static notify_join_context(joinable)[source][source]
 
      Notifies the join context manager that the calling process has not yet joined.
      
 
      Then, if throw_on_early_termination=True, checks if uneven inputs have been detected
 (i.e. if one process has already joined) and throws an exception if so.
      
@@ -689,7 +689,7 @@ 
      Generic Join Context Manager
 
      
-class torch.distributed.algorithms.Joinable[source][source]
      
+class torch.distributed.algorithms.Joinable[source][source]
 
      This defines an abstract base class for joinable classes.
      
 
      A joinable class
 (inheriting from Joinable) should implement join_hook(),
@@ -706,7 +706,7 @@ 
      Generic Join Context Manager
 
      
-abstract join_hook(**kwargs)[source][source]
      
+abstract join_hook(**kwargs)[source][source]
 
      Return a JoinHook instance for the given Joinable.
      
 
      
 
      Parameters
      
@@ -731,7 +731,7 @@ 
      Generic Join Context Manager
 
      
-class torch.distributed.algorithms.JoinHook[source][source]
      
+class torch.distributed.algorithms.JoinHook[source][source]
 
      This defines a join hook, which provides two entry points in the join context manager.
      
 
      Entry points : a main hook, which is called repeatedly while there exists a non-joined
 process, and a post-hook, which is called once all processes have joined.
      
@@ -740,7 +740,7 @@ 
      Generic Join Context Managerpost_hook() as appropriate.
      
 
      
 
      
-main_hook()[source][source]
      
+main_hook()[source][source]
 
      Call this hook while there exists a non-joined process to shadow collective communications in a training iteration.
      
 
      Training iteration i.e., in one forward pass, backward pass, and optimizer step.
      
 
      
@@ -749,7 +749,7 @@ 
      Generic Join Context Manager
 
      
-post_hook(is_last_joiner)[source][source]
      
+post_hook(is_last_joiner)[source][source]
 
      Call hook after all processes have joined.
      
 
      It is passed an additional bool argument is_last_joiner, which indicates if the rank is one of the last to join.
      
 
      
diff --git a/2.7/distributed.checkpoint.html b/2.7/distributed.checkpoint.html
index 7781ccdc430..57db76a746a 100644
--- a/2.7/distributed.checkpoint.html
+++ b/2.7/distributed.checkpoint.html
@@ -610,13 +610,13 @@ 
      Additional resources:
 
      
 
      
-class torch.distributed.checkpoint.state_dict_saver.AsyncCheckpointerType(value)[source][source]
      
+class torch.distributed.checkpoint.state_dict_saver.AsyncCheckpointerType(value)[source][source]
 
      Enum for async checkpointer type.
      
 
      
 
 
      
 
      
-torch.distributed.checkpoint.state_dict_saver.save(state_dict, *, checkpoint_id=None, storage_writer=None, planner=None, process_group=None, no_dist=False)[source][source]
      
+torch.distributed.checkpoint.state_dict_saver.save(state_dict, *, checkpoint_id=None, storage_writer=None, planner=None, process_group=None, no_dist=False)[source][source]
 
      Save a distributed model in SPMD style.
      
 
      This function is different from torch.save() as it handles
 ShardedTensor , and DTensor by having each rank only save their local shards.
      
@@ -702,7 +702,7 @@ 
      Additional resources:
 
      
-torch.distributed.checkpoint.state_dict_saver.async_save(state_dict, *, checkpoint_id=None, storage_writer=None, planner=None, process_group=None, async_checkpointer_type=AsyncCheckpointerType.THREAD)[source][source]
      
+torch.distributed.checkpoint.state_dict_saver.async_save(state_dict, *, checkpoint_id=None, storage_writer=None, planner=None, process_group=None, async_checkpointer_type=AsyncCheckpointerType.THREAD)[source][source]
 
      Asynchronous version of save. This code first de-stages the state_dict on to the
 staging storage (defaults to CPU memory), and then calls the save in a separate thread.
      
 
      
@@ -758,7 +758,7 @@ 
      Additional resources:
 
      
-torch.distributed.checkpoint.state_dict_saver.save_state_dict(state_dict, storage_writer, process_group=None, coordinator_rank=0, no_dist=False, planner=None)[source][source]
      
+torch.distributed.checkpoint.state_dict_saver.save_state_dict(state_dict, storage_writer, process_group=None, coordinator_rank=0, no_dist=False, planner=None)[source][source]
 
      This method is deprecated. Please switch to ‘save’.
      
 
      
 
      Return type
      
@@ -769,7 +769,7 @@ 
      Additional resources:
 
      
-torch.distributed.checkpoint.state_dict_loader.load(state_dict, *, checkpoint_id=None, storage_reader=None, planner=None, process_group=None, no_dist=False)[source][source]
      
+torch.distributed.checkpoint.state_dict_loader.load(state_dict, *, checkpoint_id=None, storage_reader=None, planner=None, process_group=None, no_dist=False)[source][source]
 
      Load a checkpoint into a distributed state dict in SPMD style.
      
 
      Each rank must have the same keys in their state_dict provided to this
 API. Mismatched keys may result in hangs or errors. If unsure, you can use
@@ -858,7 +858,7 @@ 
      Additional resources:
 
      
-torch.distributed.checkpoint.state_dict_loader.load_state_dict(state_dict, storage_reader, process_group=None, coordinator_rank=0, no_dist=False, planner=None)[source][source]
      
+torch.distributed.checkpoint.state_dict_loader.load_state_dict(state_dict, storage_reader, process_group=None, coordinator_rank=0, no_dist=False, planner=None)[source][source]
 
      This method is deprecated. Please switch to ‘load’.
      
 
      
 
      
@@ -867,7 +867,7 @@ 
      Additional resources:The following module is also useful for additional customization of the staging mechanisms used for asynchronous checkpointing (torch.distributed.checkpoint.async_save):
      
 
      
 
      
-class torch.distributed.checkpoint.staging.AsyncStager(*args, **kwargs)[source][source]
      
+class torch.distributed.checkpoint.staging.AsyncStager(*args, **kwargs)[source][source]
 
      This protocol is meant to provide customization and extensibility for dcp.async_save, allowing users
 to customize how data is staged previous to executing the usual dcp.save path in parallel.
 The expected order of operations (concretely defined in torch.distributed.state_dict_saver.async_save)
@@ -906,7 +906,7 @@ 
      Additional resources:
 
      
-stage(state_dict)[source][source]
      
+stage(state_dict)[source][source]
 
      Returns a “staged” copy of state_dict. The expectation of the staged copy is that it is
 innoculated from any updates incurred after the stage call is complete.
      
 
      
@@ -918,7 +918,7 @@ 
      Additional resources:
 
      
-synchronize_staging()[source][source]
      
+synchronize_staging()[source][source]
 
      In the case stage is async in some way, this method should be called to ensure staging
 is complete and it is safe to begin modifying the original state_dict
      
 
      
@@ -929,7 +929,7 @@ 
      Additional resources:
 
      
-class torch.distributed.checkpoint.staging.BlockingAsyncStager(cache_staged_state_dict=False, type_check=False)[source][source]
      
+class torch.distributed.checkpoint.staging.BlockingAsyncStager(cache_staged_state_dict=False, type_check=False)[source][source]
 
      An implementation of AsyncStager which stages the state_dict on CPU RAM and blocks until the copy is complete.
 This implementation also provides an option to optimize stage latency using pinned memory.
      
 
      N.B. synchronize_staging is a no-op in this case.
      
@@ -937,7 +937,7 @@ 
      Additional resources:
 
      
 
      
-stage(state_dict)[source][source]
      
+stage(state_dict)[source][source]
 
      Returns a copy of state_dict on the CPU.
      
 
      
 
      Return type
      
@@ -948,7 +948,7 @@ 
      Additional resources:
 
      
-synchronize_staging()[source][source]
      
+synchronize_staging()[source][source]
 
      No-op function, since staging is blocking.
      
 
      
 
      
@@ -960,11 +960,11 @@ 
      Additional resources:
 
      
 
      
-class torch.distributed.checkpoint.stateful.Stateful(*args, **kwargs)[source][source]
      
+class torch.distributed.checkpoint.stateful.Stateful(*args, **kwargs)[source][source]
 
      Stateful protocol for objects that can be checkpointed and restored.
      
 
      
 
      
-load_state_dict(state_dict)[source][source]
      
+load_state_dict(state_dict)[source][source]
 
      Restore the object’s state from the provided state_dict.
      
 
      
 
      Parameters
      
@@ -975,7 +975,7 @@ 
      Additional resources:
 
      
-state_dict()[source][source]
      
+state_dict()[source][source]
 
      Objects should return their state_dict representation as a dictionary.
 The output of this function will be checkpointed, and later restored in
 load_state_dict().
      
@@ -1000,7 +1000,7 @@ 
      Additional resources:The following types define the IO interface used during checkpoint:
      
 
      
 
      
-class torch.distributed.checkpoint.StorageReader[source][source]
      
+class torch.distributed.checkpoint.StorageReader[source][source]
 
      Interface used by load_state_dict to read from storage.
      
 
      One StorageReader instance acts as both the coordinator and the follower
 in a distributed checkpoint. As part of initialization, each instance
@@ -1016,7 +1016,7 @@ 
      Additional resources:
 
      
 
      
-abstract prepare_global_plan(plans)[source][source]
      
+abstract prepare_global_plan(plans)[source][source]
 
      Perform centralized planning of storage loading.
      
 
      This method is only called on the coordinator instance.
      
 
      While this method can produce a completely different plan, the preferred
@@ -1036,7 +1036,7 @@ 
      Additional resources:
 
      
-abstract prepare_local_plan(plan)[source][source]
      
+abstract prepare_local_plan(plan)[source][source]
 
      Perform storage-specific local planning.
      
 
      While this method can produce a completely different plan, the recommended
 way is to store storage specific data in LoadPlan::storage_data.
      
@@ -1055,7 +1055,7 @@ 
      Additional resources:
 
      
-abstract read_data(plan, planner)[source][source]
      
+abstract read_data(plan, planner)[source][source]
 
      Read all items from plan using planner to resolve the data.
      
 
      A subclass should call LoadPlanner::load_bytes to deserialize a BytesIO
 object into the right place.
      
@@ -1081,7 +1081,7 @@ 
      Additional resources:
 
      
-abstract read_metadata()[source][source]
      
+abstract read_metadata()[source][source]
 
      Read the checkpoint metadata.
      
 
      
 
      Returns
      
@@ -1095,7 +1095,7 @@ 
      Additional resources:
 
      
-abstract reset(checkpoint_id=None)[source][source]
      
+abstract reset(checkpoint_id=None)[source][source]
 
      Calls to indicates a brand new checkpoint read is going to happen.
 A checkpoint_id may be present if users set the checkpoint_id for
 this checkpoint read. The meaning of the checkpiont_id is
@@ -1113,7 +1113,7 @@ 
      Additional resources:
 
      
-abstract set_up_storage_reader(metadata, is_coordinator)[source][source]
      
+abstract set_up_storage_reader(metadata, is_coordinator)[source][source]
 
      Initialize this instance.
      
 
      
 
      Parameters
      
@@ -1128,7 +1128,7 @@ 
      Additional resources:
 
      
-abstract classmethod validate_checkpoint_id(checkpoint_id)[source][source]
      
+abstract classmethod validate_checkpoint_id(checkpoint_id)[source][source]
 
      Check if the given checkpoint_id is supported by the stroage. This allow
 us to enable automatic storage selection.
      
 
      
@@ -1142,7 +1142,7 @@ 
      Additional resources:
 
      
-class torch.distributed.checkpoint.StorageWriter[source][source]
      
+class torch.distributed.checkpoint.StorageWriter[source][source]
 
      Interface used by save_state_dict to write to storage.
      
 
      One StorageWriter instance acts as both the coordinator and the follower
 in a distributed checkpoint. As part of initialization, each instance
@@ -1158,7 +1158,7 @@ 
      Additional resources:
 
      
 
      
-abstract finish(metadata, results)[source][source]
      
+abstract finish(metadata, results)[source][source]
 
      Write the metadata and marks the current checkpoint as successful.
      
 
      The actual format/schema used for serializing metadata is an
 implementation detail. The only requirement is that it’s recoverable
@@ -1181,7 +1181,7 @@ 
      Additional resources:
 
      
-abstract prepare_global_plan(plans)[source][source]
      
+abstract prepare_global_plan(plans)[source][source]
 
      Perform centralized planning of storage.
      
 
      This method is only called on the coordinator instance.
      
 
      While this method can produce a completely different plan, the preferred
@@ -1201,7 +1201,7 @@ 
      Additional resources:
 
      
-abstract prepare_local_plan(plan)[source][source]
      
+abstract prepare_local_plan(plan)[source][source]
 
      Perform storage-specific local planning.
      
 
      While this method can produce a completely different plan, the recommended
 way is to store storage specific data in SavePlan::storage_data.
      
@@ -1220,7 +1220,7 @@ 
      Additional resources:
 
      
-abstract reset(checkpoint_id=None)[source][source]
      
+abstract reset(checkpoint_id=None)[source][source]
 
      Calls to indicates a brand new checkpoint write is going to happen.
 A checkpoint_id may be present if users set the checkpoint_id for
 this checkpoint write. The meaning of the checkpiont_id is
@@ -1238,7 +1238,7 @@ 
      Additional resources:
 
      
-abstract set_up_storage_writer(is_coordinator)[source][source]
      
+abstract set_up_storage_writer(is_coordinator)[source][source]
 
      Initialize this instance.
      
 
      
 
      Parameters
      
@@ -1250,7 +1250,7 @@ 
      Additional resources:
 
      
-storage_meta()[source][source]
      
+storage_meta()[source][source]
 
      Return the storage-specific metadata. This is used to store additional information
 in a checkpoint that can be useful for providing request-level observability. StorageMeta
 is passed to the SavePlanner during save calls. Returns None by default.
      
@@ -1264,7 +1264,7 @@ 
      Additional resources:
 
      
-abstract classmethod validate_checkpoint_id(checkpoint_id)[source][source]
      
+abstract classmethod validate_checkpoint_id(checkpoint_id)[source][source]
 
      Check if the given checkpoint_id is supported by the stroage. This allow
 us to enable automatic storage selection.
      
 
      
@@ -1276,7 +1276,7 @@ 
      Additional resources:
 
      
-abstract write_data(plan, planner)[source][source]
      
+abstract write_data(plan, planner)[source][source]
 
      Write all items from plan using planner to resolve the data.
      
 
      A subclass should call SavePlanner::resolve_data on each item
 from the plan to get access to the underlying object to write.
      
@@ -1307,7 +1307,7 @@ 
      Additional resources:The following types define the planner interface used during checkpoint:
      
 
      
 
      
-class torch.distributed.checkpoint.LoadPlanner[source][source]
      
+class torch.distributed.checkpoint.LoadPlanner[source][source]
 
      Abstract class defining the protocol used by load_state_dict to plan the load process.
      
 
      LoadPlanner are stateful objects that can be used to customize the whole load process.
      
 
      LoadPlanner acts as an access proxy to the state_dict, so any transformation done to it
@@ -1384,7 +1384,7 @@ 
      Additional resources:
 
      
 
      
-abstract commit_tensor(read_item, tensor)[source][source]
      
+abstract commit_tensor(read_item, tensor)[source][source]
 
      Call once the StorageReader finished loading data into tensor.
      
 
      The provided tensor is the same one returned by the call to resolve_tensor.
 This method is only needed if this LoadPlanner needs to post process tensor prior to
@@ -1396,7 +1396,7 @@ 
      Additional resources:
 
      
-abstract create_global_plan(global_plan)[source][source]
      
+abstract create_global_plan(global_plan)[source][source]
 
      Compute the global load plan and return plans for each rank.
      
 
      . N.B. This is called on the coordinator rank only
      
 
      
@@ -1408,7 +1408,7 @@ 
      Additional resources:
 
      
-abstract create_local_plan()[source][source]
      
+abstract create_local_plan()[source][source]
 
      Create a LoadPlan based on state_dict and metadata provided by set_up_planner.
      
 
      . N.B. This is called on every rank.
      
 
      
@@ -1420,7 +1420,7 @@ 
      Additional resources:
 
      
-abstract finish_plan(central_plan)[source][source]
      
+abstract finish_plan(central_plan)[source][source]
 
      Accept the plan from coordinator and return final LoadPlan.
      
 
      
 
      Return type
      
@@ -1431,7 +1431,7 @@ 
      Additional resources:
 
      
-abstract load_bytes(read_item, value)[source][source]
      
+abstract load_bytes(read_item, value)[source][source]
 
      Load the item described by read_item``and ``value.
      
 
      This method is expected to modify in-place the underlying state_dict.
      
 
      The contents of value are defined by the SavePlanner used to produce
@@ -1442,7 +1442,7 @@ 
      Additional resources:
 
      
-resolve_bytes(read_item)[source][source]
      
+resolve_bytes(read_item)[source][source]
 
      Return the BytesIO to be used by the StorageReader to load read_item.
      
 
      The BytesIO should alias with one on the underlying state_dict as StorageReader will replace its contents.
      
 
      
@@ -1454,7 +1454,7 @@ 
      Additional resources:
 
      
-abstract resolve_tensor(read_item)[source][source]
      
+abstract resolve_tensor(read_item)[source][source]
 
      Return the tensor described by read_item to be used by the StorageReader to load read_item.
      
 
      The tensor should alias with one on the underlying state_dict as StorageReader will replace its contents.
 If, for any reason, that’s not possible, the planner can use the commit_tensor method to copy the data
@@ -1468,7 +1468,7 @@ 
      Additional resources:
 
      
-abstract set_up_planner(state_dict, metadata=None, is_coordinator=False)[source][source]
      
+abstract set_up_planner(state_dict, metadata=None, is_coordinator=False)[source][source]
 
      Initialize this instance to load data into state_dict.
      
 
      . N.B. This is called on every rank.
      
 
      
@@ -1479,21 +1479,21 @@ 
      Additional resources:
 
      
-class torch.distributed.checkpoint.LoadPlan(items: list[torch.distributed.checkpoint.planner.ReadItem], storage_data: Any = None, planner_data: Any = None)[source][source]
      
+class torch.distributed.checkpoint.LoadPlan(items: list[torch.distributed.checkpoint.planner.ReadItem], storage_data: Any = None, planner_data: Any = None)[source][source]
 
      
 
      
 
      
 
 
      
 
      
-class torch.distributed.checkpoint.ReadItem(type: torch.distributed.checkpoint.planner.LoadItemType, dest_index: torch.distributed.checkpoint.metadata.MetadataIndex, dest_offsets: torch.Size, storage_index: torch.distributed.checkpoint.metadata.MetadataIndex, storage_offsets: torch.Size, lengths: torch.Size)[source][source]
      
+class torch.distributed.checkpoint.ReadItem(type: torch.distributed.checkpoint.planner.LoadItemType, dest_index: torch.distributed.checkpoint.metadata.MetadataIndex, dest_offsets: torch.Size, storage_index: torch.distributed.checkpoint.metadata.MetadataIndex, storage_offsets: torch.Size, lengths: torch.Size)[source][source]
 
      
 
      
 
      
 
 
      
 
      
-class torch.distributed.checkpoint.SavePlanner[source][source]
      
+class torch.distributed.checkpoint.SavePlanner[source][source]
 
      Abstract class defining the protocol used by save_state_dict to plan the save process.
      
 
      SavePlanners are stateful objects that can be used to customize the whole save process.
      
 
      SavePlanner acts as an access proxy to the state_dict, so any transformation done to it
@@ -1592,7 +1592,7 @@ 
      Additional resources:
 
      
 
      
-abstract create_global_plan(all_plans)[source][source]
      
+abstract create_global_plan(all_plans)[source][source]
 
      Compute the global checkpoint plan and return the local plan of each rank.
      
 
      This is called on the coordinator rank only.
      
 
      
@@ -1604,7 +1604,7 @@ 
      Additional resources:
 
      
-abstract create_local_plan()[source][source]
      
+abstract create_local_plan()[source][source]
 
      Compute the save plan for the current rank.
      
 
      This will be aggregated and passed to create_global_plan.
 Planner specific data can be passed through SavePlan::planner_data.
      
@@ -1618,7 +1618,7 @@ 
      Additional resources:
 
      
-abstract finish_plan(new_plan)[source][source]
      
+abstract finish_plan(new_plan)[source][source]
 
      Merge the plan created by create_local_plan and the result of create_global_plan.
      
 
      This is called on all ranks.
      
 
      
@@ -1630,7 +1630,7 @@ 
      Additional resources:
 
      
-abstract resolve_data(write_item)[source][source]
      
+abstract resolve_data(write_item)[source][source]
 
      Transform and prepare write_item from state_dict for storage, ensuring idempotency and thread-safety.
      
 
      Lookup the object associated with write_item in state_dict and apply any
 transformation (such as serialization) prior to the storage layer consuming it.
      
@@ -1650,7 +1650,7 @@ 
      Additional resources:
 
      
-abstract set_up_planner(state_dict, storage_meta=None, is_coordinator=False)[source][source]
      
+abstract set_up_planner(state_dict, storage_meta=None, is_coordinator=False)[source][source]
 
      Initialize this planner to save state_dict.
      
 
      Implementations should save those values as they won’t be provided lated in the save process.
      
 
      This is called on all ranks.
      
@@ -1662,20 +1662,20 @@ 
      Additional resources:
 
      
-class torch.distributed.checkpoint.SavePlan(items: list[torch.distributed.checkpoint.planner.WriteItem], storage_data: Any = None, planner_data: Any = None, usable: bool = True)[source][source]
      
+class torch.distributed.checkpoint.SavePlan(items: list[torch.distributed.checkpoint.planner.WriteItem], storage_data: Any = None, planner_data: Any = None, usable: bool = True)[source][source]
 
      
 
      
 
      
 
 
      
 
      
-class torch.distributed.checkpoint.planner.WriteItem(index, type, tensor_data=None)[source][source]
      
+class torch.distributed.checkpoint.planner.WriteItem(index, type, tensor_data=None)[source][source]
 
      Dataclass which holds information about what needs to be written to storage.
      
 
      
 
      
 
      
 
      
-tensor_storage_size()[source][source]
      
+tensor_storage_size()[source][source]
 
      Calculates the storage size of the underlying tensor, or None if this is not a tensor write.
      
 
      
 
      Returns
      
@@ -1692,7 +1692,7 @@ 
      Additional resources:We provide a filesystem based storage layer:
      
 
      
 
      
-class torch.distributed.checkpoint.FileSystemReader(path, _extension_registry=None)[source][source]
      
+class torch.distributed.checkpoint.FileSystemReader(path, _extension_registry=None)[source][source]
 
      
 
      
 
      
@@ -1705,7 +1705,7 @@ 
      Additional resources:
 
      
-class torch.distributed.checkpoint.FileSystemWriter(path, single_file_per_rank=True, sync_files=True, thread_count=1, per_thread_copy_ahead=10000000, cache_staged_state_dict=False, overwrite=True, _extensions=None)[source][source]
      
+class torch.distributed.checkpoint.FileSystemWriter(path, single_file_per_rank=True, sync_files=True, thread_count=1, per_thread_copy_ahead=10000000, cache_staged_state_dict=False, overwrite=True, _extensions=None)[source][source]
 
      Basic implementation of StorageWriter using file IO.
      
 
      This implementation makes the following assumptions and simplifications:
      
 
        
@@ -1718,7 +1718,7 @@ 
        Additional resources:
 
        
 
        
-stage(state_dict)[source][source]
        
+stage(state_dict)[source][source]
 
        Override of AsyncStager.stage
        
 
        
 
        Return type
        
@@ -1733,12 +1733,12 @@ 
        Additional resources:
 
        
 
        
-class torch.distributed.checkpoint.DefaultSavePlanner(flatten_state_dict=True, flatten_sharded_tensors=True, dedup_replicated_tensors=None, dedup_save_to_lowest_rank=False, enable_plan_caching=False)[source][source]
        
+class torch.distributed.checkpoint.DefaultSavePlanner(flatten_state_dict=True, flatten_sharded_tensors=True, dedup_replicated_tensors=None, dedup_save_to_lowest_rank=False, enable_plan_caching=False)[source][source]
 
        
 
        
 
        
 
        
-lookup_object(index)[source][source]
        
+lookup_object(index)[source][source]
 
        Extension from the planner interface to make it easy to extend the default planner.
        
 
        
 
        Return type
        
@@ -1749,7 +1749,7 @@ 
        Additional resources:
 
        
-transform_object(write_item, object)[source][source]
        
+transform_object(write_item, object)[source][source]
 
        Extension from the planner interface to make it easy to extend the default planner.
        
 
        
 
        
@@ -1759,7 +1759,7 @@ 
        Additional resources:
 
        
-class torch.distributed.checkpoint.DefaultLoadPlanner(flatten_state_dict=True, flatten_sharded_tensors=True, allow_partial_load=False)[source][source]
        
+class torch.distributed.checkpoint.DefaultLoadPlanner(flatten_state_dict=True, flatten_sharded_tensors=True, allow_partial_load=False)[source][source]
 
        DefaultLoadPlanner that adds multiple features on top of LoadPlanner.
        
 
        In particular it adds the following:
        
 
        flatten_state_dict: Handle state_dict with nested dicts
@@ -1769,7 +1769,7 @@ 
        Additional resources:
 
        
 
        
-lookup_tensor(index)[source][source]
        
+lookup_tensor(index)[source][source]
 
        Extension from the planner interface to make it easy to extend the default planner.
        
 
        
 
        Return type
        
@@ -1780,7 +1780,7 @@ 
        Additional resources:
 
        
-transform_tensor(read_item, tensor)[source][source]
        
+transform_tensor(read_item, tensor)[source][source]
 
        Extension from the planner interface to make it easy to extend the default planner.
        
 
        
 
        
@@ -1796,7 +1796,7 @@ 
        Additional resources:Note that this feature is experimental, and API signatures might change in the future.
        
 
        
 
        
-torch.distributed.checkpoint.state_dict.get_state_dict(model, optimizers, *, submodules=None, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.get_state_dict(model, optimizers, *, submodules=None, options=None)[source][source]
 
        Return the model state_dict and optimizers state_dict.
        
 
        get_state_dict can process any module that is parallelized by PyTorch
 FSDP/fully_shard, DDP/replicate, tensor_parallel/parallelize_module, and any
@@ -1865,7 +1865,7 @@ 
        Additional resources:
 
        
-torch.distributed.checkpoint.state_dict.get_model_state_dict(model, *, submodules=None, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.get_model_state_dict(model, *, submodules=None, options=None)[source][source]
 
        Return the model state_dict of model.
        
 
        See get_state_dict for the detail usage.
        
 
        
@@ -1890,7 +1890,7 @@ 
        Additional resources:
 
        
-torch.distributed.checkpoint.state_dict.get_optimizer_state_dict(model, optimizers, *, submodules=None, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.get_optimizer_state_dict(model, optimizers, *, submodules=None, options=None)[source][source]
 
        Return the combined state_dict for optimizers.
        
 
        See get_state_dict for the detail usage.
        
 
        
@@ -1916,7 +1916,7 @@ 
        Additional resources:
 
        
-torch.distributed.checkpoint.state_dict.set_state_dict(model, optimizers, *, model_state_dict, optim_state_dict, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.set_state_dict(model, optimizers, *, model_state_dict, optim_state_dict, options=None)[source][source]
 
        Load the model state_dict and optimizers state_dict.
        
 
        The counterpart of get_state_dict to set the state_dict to the model and
 optimizers.  The given model_state_dict and optim_state_dict do not
@@ -1962,7 +1962,7 @@ 
        Additional resources:
 
        
-torch.distributed.checkpoint.state_dict.set_model_state_dict(model, model_state_dict, *, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.set_model_state_dict(model, model_state_dict, *, options=None)[source][source]
 
        Load the model state_dict.
        
 
        The counterpart of get_model_state_dict to set the state_dict to the
 model. See set_state_dict for the detail usage.
        
@@ -1995,7 +1995,7 @@ 
        Additional resources:
 
        
-torch.distributed.checkpoint.state_dict.set_optimizer_state_dict(model, optimizers, optim_state_dict, *, options=None)[source][source]
        
+torch.distributed.checkpoint.state_dict.set_optimizer_state_dict(model, optimizers, optim_state_dict, *, options=None)[source][source]
 
        Load the optimizers state_dict.
        
 
        The counterpart of get_optimizer_state_dict to set the state_dict to the
 optimizers. See set_state_dict for the detail usage.
        
@@ -2027,7 +2027,7 @@ 
        Additional resources:
 
        
-class torch.distributed.checkpoint.state_dict.StateDictOptions(full_state_dict=False, cpu_offload=False, ignore_frozen_params=False, keep_submodule_prefixes=True, strict=True, broadcast_from_rank0=False, flatten_optimizer_state_dict=False, dsd_fqn_modifiers='_fqn_modifiers')[source][source]
        
+class torch.distributed.checkpoint.state_dict.StateDictOptions(full_state_dict=False, cpu_offload=False, ignore_frozen_params=False, keep_submodule_prefixes=True, strict=True, broadcast_from_rank0=False, flatten_optimizer_state_dict=False, dsd_fqn_modifiers='_fqn_modifiers')[source][source]
 
        This dataclass specifies how get_state_dict/set_state_dict will work.
        
 
          
 
        • full_state_dict: if this is set to True, all the tensors in the
@@ -2067,7 +2067,7 @@ 
          Additional resources:For users which are used to using and sharing models in the torch.save format, the following methods are provided which provide offline utilities for converting betweeing formats.
          
 
          
 
          
-torch.distributed.checkpoint.format_utils.dcp_to_torch_save(dcp_checkpoint_dir, torch_save_path)[source][source]
          
+torch.distributed.checkpoint.format_utils.dcp_to_torch_save(dcp_checkpoint_dir, torch_save_path)[source][source]
 
          Given a directory containing a DCP checkpoint, this function will convert it into a
 Torch save file.
          
 
          
@@ -2086,7 +2086,7 @@ 
          Additional resources:
 
          
-torch.distributed.checkpoint.format_utils.torch_save_to_dcp(torch_save_path, dcp_checkpoint_dir)[source][source]
          
+torch.distributed.checkpoint.format_utils.torch_save_to_dcp(torch_save_path, dcp_checkpoint_dir)[source][source]
 
          Given the location of a torch save file, converts it into a DCP checkpoint.
          
 
          
 
          Parameters
          
@@ -2105,7 +2105,7 @@ 
          Additional resources:The following classes can also be utilized for online loading and resharding of models from the torch.save format.
          
 
          
 
          
-class torch.distributed.checkpoint.format_utils.BroadcastingTorchSaveReader(checkpoint_id=None, coordinator_rank=0)[source][source]
          
+class torch.distributed.checkpoint.format_utils.BroadcastingTorchSaveReader(checkpoint_id=None, coordinator_rank=0)[source][source]
 
          StorageReader for reading a Torch Save file. This reader will read the entire checkpoint
 on the coordinator rank, and then broadcast and shard each tensor to all ranks.
          
 
          . N.B. Intended to be used with DynamicMetaLoadPlanner
          
@@ -2126,7 +2126,7 @@ 
          Additional resources:
 
          
 
          
-prepare_global_plan(global_plan)[source][source]
          
+prepare_global_plan(global_plan)[source][source]
 
          Implementation of the StorageReader method
          
 
          
 
          Return type
          
@@ -2137,7 +2137,7 @@ 
          Additional resources:
 
          
-prepare_local_plan(plan)[source][source]
          
+prepare_local_plan(plan)[source][source]
 
          Implementation of the StorageReader method
          
 
          
 
          Return type
          
@@ -2148,7 +2148,7 @@ 
          Additional resources:
 
          
-read_data(plan, planner)[source][source]
          
+read_data(plan, planner)[source][source]
 
          Reads torch save data on the coordinator rank, and broadcast afterwards
 this incurrs a communication cost, but avoids having to load
 the entire checkpoint on each rank, hopefully preventing OOM issues
          
@@ -2161,7 +2161,7 @@ 
          Additional resources:
 
          
-read_metadata()[source][source]
          
+read_metadata()[source][source]
 
          Extends the default StorageReader to support building the metadata file
          
 
          
 
          Return type
          
@@ -2172,7 +2172,7 @@ 
          Additional resources:
 
          
-reset(checkpoint_id=None)[source][source]
          
+reset(checkpoint_id=None)[source][source]
 
          Implementation of the StorageReader method
          
 
          
 
          
@@ -2180,7 +2180,7 @@ 
          Additional resources:
 
          
-set_up_storage_reader(metadata, is_coordinator)[source][source]
          
+set_up_storage_reader(metadata, is_coordinator)[source][source]
 
          Implementation of the StorageReader method
          
 
          
 
          
@@ -2188,7 +2188,7 @@ 
          Additional resources:
 
          
-classmethod validate_checkpoint_id(checkpoint_id)[source][source]
          
+classmethod validate_checkpoint_id(checkpoint_id)[source][source]
 
          Implementation of the StorageReader method
          
 
          
 
          Return type
          
@@ -2201,7 +2201,7 @@ 
          Additional resources:
 
          
-class torch.distributed.checkpoint.format_utils.DynamicMetaLoadPlanner(flatten_state_dict=True, flatten_sharded_tensors=True, allow_partial_load=False)[source][source]
          
+class torch.distributed.checkpoint.format_utils.DynamicMetaLoadPlanner(flatten_state_dict=True, flatten_sharded_tensors=True, allow_partial_load=False)[source][source]
 
          Extension of DefaultLoadPlanner, which creates a new Metadata object based on the passed in state dict,
 avoiding the need to read metadata from disk. This is useful when reading formats which don’t have a
 metadata file, like Torch Save files.
          
@@ -2223,7 +2223,7 @@ 
          Additional resources:
 
          
 
          
-set_up_planner(state_dict, metadata=None, is_coordinator=False)[source][source]
          
+set_up_planner(state_dict, metadata=None, is_coordinator=False)[source][source]
 
          Setups of the planner, extnding default behavior by creating the Metadata object from the state dict
          
 
          
 
          
diff --git a/2.7/distributed.fsdp.fully_shard.html b/2.7/distributed.fsdp.fully_shard.html
index 043abfeeb13..d741dedacae 100644
--- a/2.7/distributed.fsdp.fully_shard.html
+++ b/2.7/distributed.fsdp.fully_shard.html
@@ -644,7 +644,7 @@ 
          PyTorch FSDP2 (
 
          The frontend API is fully_shard that can be called on a module:
          
 
          
 
          
-torch.distributed.fsdp.fully_shard(module, *, mesh=None, reshard_after_forward=True, shard_placement_fn=None, mp_policy=MixedPrecisionPolicy(param_dtype=None, reduce_dtype=None, output_dtype=None, cast_forward_inputs=True), offload_policy=OffloadPolicy(), ignored_params=None)[source]
          
+torch.distributed.fsdp.fully_shard(module, *, mesh=None, reshard_after_forward=True, shard_placement_fn=None, mp_policy=MixedPrecisionPolicy(param_dtype=None, reduce_dtype=None, output_dtype=None, cast_forward_inputs=True), offload_policy=OffloadPolicy(), ignored_params=None)[source]
 
          Apply fully sharded data parallelism (FSDP) to module, where FSDP
 shards module parameters, gradients, and optimizer states across data
 parallel workers to save memory at the cost of communication.
          
@@ -752,7 +752,7 @@ 
          PyTorch FSDP2 (
 class torch.distributed.fsdp.FSDPModule(*args, **kwargs)
 
          
 
          
-reshard()[source][source]
          
+reshard()[source][source]
 
          Reshards the module’s parameters, freeing the unsharded parameters if
 they are allocated and registering the sharded parameters to the
 module. This method is not recursive.
          
@@ -762,7 +762,7 @@ 
          PyTorch FSDP2 (
 
 
          
 
          
-set_all_reduce_hook(hook, *, stream=None)[source][source]
          
+set_all_reduce_hook(hook, *, stream=None)[source][source]
 
          
 
          Parameters
          
 
            
@@ -781,7 +781,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_is_last_backward(is_last_backward)[source][source]
            
+set_is_last_backward(is_last_backward)[source][source]
 
            Sets whether the next backward is the last one. On the last backward,
 FSDP waits on pending gradient reduction and clears internal data
 data structures for backward prefetching. This can be useful for
@@ -792,7 +792,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_modules_to_backward_prefetch(modules)[source][source]
            
+set_modules_to_backward_prefetch(modules)[source][source]
 
            Sets the FSDP modules for which this FSDP module should explicitly
 prefetch all-gathers in backward. This overrides the default backward
 pretching implementation that prefetches the next FSDP module based on
@@ -810,7 +810,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_modules_to_forward_prefetch(modules)[source][source]
            
+set_modules_to_forward_prefetch(modules)[source][source]
 
            Sets the FSDP modules for which this FSDP module should explicitly
 prefetch all-gathers in forward. The prefetching runs after this
 module’s all-gather copy-out.
            
@@ -828,7 +828,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_post_optim_event(event)[source][source]
            
+set_post_optim_event(event)[source][source]
 
            Sets a post-optimizer-step event for the root FSDP module to wait the
 all-gather streams on.
            
 
            By default, the root FSDP module waits the all-gather streams on the
@@ -848,7 +848,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_reduce_scatter_divide_factor(factor)[source][source]
            
+set_reduce_scatter_divide_factor(factor)[source][source]
 
            Sets a custom divide factor for the reduce-scatter. This becomes a
 custom reduce op using NCCL’s PreMulSum, which allows multiplying by
 the factor before reduction.
            
@@ -861,7 +861,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_requires_all_reduce(requires_all_reduce, *, recurse=True)[source][source]
            
+set_requires_all_reduce(requires_all_reduce, *, recurse=True)[source][source]
 
            Sets if the module should all-reduce gradients. This can be used to
 implement gradient accumulation with only reduce-scatter but not
 all-reduce for HSDP.
            
@@ -871,7 +871,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_requires_gradient_sync(requires_gradient_sync, *, recurse=True)[source][source]
            
+set_requires_gradient_sync(requires_gradient_sync, *, recurse=True)[source][source]
 
            Sets if the module should sync gradients. This can be used to implement
 gradient accumulation without communication. For HSDP, this controls
 both reduce-scatter and all-reduce together. This is the equivalence of
@@ -890,7 +890,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_reshard_after_backward(reshard_after_backward, *, recurse=True)[source][source]
            
+set_reshard_after_backward(reshard_after_backward, *, recurse=True)[source][source]
 
            Sets if the module should reshard parameters after backward. This can
 be used during gradient accumulation to trade off higher memory for
 reduced communication since the unsharded parameters do not need to be
@@ -909,7 +909,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-set_unshard_in_backward(unshard_in_backward)[source][source]
            
+set_unshard_in_backward(unshard_in_backward)[source][source]
 
            Sets whether the FSDP module’s parameters need to be unsharded in
 backward. This can be used in expert cases when the user knows that all
 parameters in this FSDP module’s parameter group are not needed for
@@ -920,7 +920,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-unshard(async_op=False)[source][source]
            
+unshard(async_op=False)[source][source]
 
            Unshards the module’s parameters by allocating memory and all-gathering
 the parameters. This method is not recursive. The unshard follows the
 MixedPrecisionPolicy, so it will all-gather following
@@ -953,7 +953,7 @@ 
            PyTorch FSDP2 (
 
            A handle to wait on a FSDPModule.unshard() op.
            
 
            
 
            
-wait()[source][source]
            
+wait()[source][source]
 
            Waits on the unshard op. This ensures that the current stream can use
 the unsharded parameters, which are now registered to the module.
            
 
            
@@ -964,7 +964,7 @@ 
            PyTorch FSDP2 (
 
 
            
 
            
-torch.distributed.fsdp.register_fsdp_forward_method(module, method_name)[source]
            
+torch.distributed.fsdp.register_fsdp_forward_method(module, method_name)[source]
 
            Registers a method on module to be considered a forward method for
 FSDP.
            
 
            FSDP all-gathers parameters pre-forward and optionally frees parameters
diff --git a/2.7/distributed.html b/2.7/distributed.html
index 152b457c110..d533be00348 100644
--- a/2.7/distributed.html
+++ b/2.7/distributed.html
@@ -844,7 +844,7 @@ 
            Initialization
 
            
-torch.distributed.is_available()[source][source]
            
+torch.distributed.is_available()[source][source]
 
            Return True if the distributed package is available.
            
 
            Otherwise,
 torch.distributed does not expose any other APIs. Currently,
@@ -861,7 +861,7 @@ 
            Initialization
 
            
-torch.distributed.init_process_group(backend=None, init_method=None, timeout=None, world_size=-1, rank=-1, store=None, group_name='', pg_options=None, device_id=None)[source][source]
            
+torch.distributed.init_process_group(backend=None, init_method=None, timeout=None, world_size=-1, rank=-1, store=None, group_name='', pg_options=None, device_id=None)[source][source]
 
            Initialize the default distributed process group.
            
 
            This will also initialize the distributed package.
            
 
            
@@ -952,7 +952,7 @@ 
            Initialization
 
            
-torch.distributed.device_mesh.init_device_mesh(device_type, mesh_shape, *, mesh_dim_names=None)[source][source]
            
+torch.distributed.device_mesh.init_device_mesh(device_type, mesh_shape, *, mesh_dim_names=None)[source][source]
 
            Initializes a DeviceMesh based on device_type, mesh_shape, and mesh_dim_names parameters.
            
 
            This creates a DeviceMesh with an n-dimensional array layout, where n is the length of mesh_shape.
 If mesh_dim_names is provided, each dimension is labeled as mesh_dim_names[i].
            
@@ -999,7 +999,7 @@ 
            Initialization
 
            
-torch.distributed.is_initialized()[source][source]
            
+torch.distributed.is_initialized()[source][source]
 
            Check if the default process group has been initialized.
            
 
            
 
            Return type
            
@@ -1010,7 +1010,7 @@ 
            Initialization
 
            
-torch.distributed.is_mpi_available()[source][source]
            
+torch.distributed.is_mpi_available()[source][source]
 
            Check if the MPI backend is available.
            
 
            
 
            Return type
            
@@ -1021,7 +1021,7 @@ 
            Initialization
 
            
-torch.distributed.is_nccl_available()[source][source]
            
+torch.distributed.is_nccl_available()[source][source]
 
            Check if the NCCL backend is available.
            
 
            
 
            Return type
            
@@ -1032,7 +1032,7 @@ 
            Initialization
 
            
-torch.distributed.is_gloo_available()[source][source]
            
+torch.distributed.is_gloo_available()[source][source]
 
            Check if the Gloo backend is available.
            
 
            
 
            Return type
            
@@ -1043,7 +1043,7 @@ 
            Initialization
 
            
-torch.distributed.distributed_c10d.is_xccl_available()[source][source]
            
+torch.distributed.distributed_c10d.is_xccl_available()[source][source]
 
            Check if the XCCL backend is available.
            
 
            
 
            Return type
            
@@ -1054,7 +1054,7 @@ 
            Initialization
 
            
-torch.distributed.is_torchelastic_launched()[source][source]
            
+torch.distributed.is_torchelastic_launched()[source][source]
 
            Check whether this process was launched with torch.distributed.elastic (aka torchelastic).
            
 
            The existence of TORCHELASTIC_RUN_ID environment
 variable is used as a proxy to determine whether the current process
@@ -1149,7 +1149,7 @@ 
            Post-Initializationtorch.distributed.is_initialized().
            
 
            
 
            
-class torch.distributed.Backend(name)[source][source]
            
+class torch.distributed.Backend(name)[source][source]
 
            An enum-like class for backends.
            
 
            Available backends: GLOO, NCCL, UCC, MPI, XCCL, and other registered backends.
            
 
            The values of this class are lowercase strings, e.g., "gloo". They can
@@ -1168,7 +1168,7 @@ 
            Post-Initialization
 
            
-classmethod register_backend(name, func, extended_api=False, devices=None)[source][source]
            
+classmethod register_backend(name, func, extended_api=False, devices=None)[source][source]
 
            Register a new backend with the given name and instantiating function.
            
 
            This class method is used by 3rd party ProcessGroup extension to
 register new backends.
            
@@ -1201,7 +1201,7 @@ 
            Post-Initialization
 
            
-torch.distributed.get_backend(group=None)[source][source]
            
+torch.distributed.get_backend(group=None)[source][source]
 
            Return the backend of the given process group.
            
 
            
 
            Parameters
            
@@ -1220,7 +1220,7 @@ 
            Post-Initialization
 
            
-torch.distributed.get_rank(group=None)[source][source]
            
+torch.distributed.get_rank(group=None)[source][source]
 
            Return the rank of the current process in the provided group, default otherwise.
            
 
            Rank is a unique identifier assigned to each process within a distributed
 process group. They are always consecutive integers ranging from 0 to
@@ -1242,7 +1242,7 @@ 
            Post-Initialization
 
            
-torch.distributed.get_world_size(group=None)[source][source]
            
+torch.distributed.get_world_size(group=None)[source][source]
 
            Return the number of processes in the current process group.
            
 
            
 
            Parameters
            
@@ -1298,7 +1298,7 @@ 
            Groups
 
            
 
            
-torch.distributed.new_group(ranks=None, timeout=None, backend=None, pg_options=None, use_local_synchronization=False, group_desc=None, device_id=None)[source][source]
            
+torch.distributed.new_group(ranks=None, timeout=None, backend=None, pg_options=None, use_local_synchronization=False, group_desc=None, device_id=None)[source][source]
 
            Create a new distributed group.
            
 
            This function requires that all processes in the main group (i.e. all
 processes that are part of the distributed job) enter this function, even
@@ -1366,7 +1366,7 @@ 
            Groups
 
            
-torch.distributed.get_group_rank(group, global_rank)[source][source]
            
+torch.distributed.get_group_rank(group, global_rank)[source][source]
 
            Translate a global rank into a group rank.
            
 
            global_rank must be part of group otherwise this raises RuntimeError.
            
 
            
@@ -1388,7 +1388,7 @@ 
            Groups
 
            
-torch.distributed.get_global_rank(group, group_rank)[source][source]
            
+torch.distributed.get_global_rank(group, group_rank)[source][source]
 
            Translate a group rank into a global rank.
            
 
            group_rank must be part of group otherwise this raises RuntimeError.
            
 
            
@@ -1410,7 +1410,7 @@ 
            Groups
 
            
-torch.distributed.get_process_group_ranks(group)[source][source]
            
+torch.distributed.get_process_group_ranks(group)[source][source]
 
            Get all ranks associated with group.
            
 
            
 
            Parameters
            
@@ -1435,7 +1435,7 @@ 
            DeviceMesh
 
            
-class torch.distributed.device_mesh.DeviceMesh(device_type, mesh, *, mesh_dim_names=None, _init_backend=True)[source][source]
            
+class torch.distributed.device_mesh.DeviceMesh(device_type, mesh, *, mesh_dim_names=None, _init_backend=True)[source][source]
 
            DeviceMesh represents a mesh of devices, where layout of devices could be
 represented as a n-d dimension array, and each value of the n-d dimensional
 array is the global id of the default process group ranks.
            
@@ -1481,7 +1481,7 @@ 
            DeviceMesh
 
            
-static from_group(group, device_type, mesh=None, *, mesh_dim_names=None)[source][source]
            
+static from_group(group, device_type, mesh=None, *, mesh_dim_names=None)[source][source]
 
            Constructs a DeviceMesh with device_type from an
 existing ProcessGroup or a list of existing ProcessGroup.
            
 
            The constructed device mesh has number of dimensions equal to the
@@ -1522,7 +1522,7 @@ 
            DeviceMesh
 
            
-get_all_groups()[source][source]
            
+get_all_groups()[source][source]
 
            Returns a list of ProcessGroups for all mesh dimensions.
            
 
            
 
            Returns
            
@@ -1536,7 +1536,7 @@ 
            DeviceMesh
 
            
-get_coordinate()[source][source]
            
+get_coordinate()[source][source]
 
            Return the relative indices of this rank relative to all
 dimensions of the mesh. If this rank is not part of the mesh, return None.
            
 
            
@@ -1548,7 +1548,7 @@ 
            DeviceMesh
 
            
-get_group(mesh_dim=None)[source][source]
            
+get_group(mesh_dim=None)[source][source]
 
            Returns the single ProcessGroup specified by mesh_dim, or, if mesh_dim is not specified and the
 DeviceMesh is 1-dimensional, returns the only ProcessGroup in the mesh.
            
 
            
@@ -1569,7 +1569,7 @@ 
            DeviceMesh
 
            
-get_local_rank(mesh_dim=None)[source][source]
            
+get_local_rank(mesh_dim=None)[source][source]
 
            Returns the local rank of the given mesh_dim of the DeviceMesh.
            
 
            
 
            Parameters
            
@@ -1607,7 +1607,7 @@ 
            DeviceMesh
 
            
-get_rank()[source][source]
            
+get_rank()[source][source]
 
            Returns the current global rank.
            
 
            
 
            Return type
            
@@ -1623,7 +1623,7 @@ 
            DeviceMesh
            
 
            
 
            
-torch.distributed.send(tensor, dst=None, group=None, tag=0, group_dst=None)[source][source]
            
+torch.distributed.send(tensor, dst=None, group=None, tag=0, group_dst=None)[source][source]
 
            Send a tensor synchronously.
            
 
            
 
            Warning
            
@@ -1646,7 +1646,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.recv(tensor, src=None, group=None, tag=0, group_src=None)[source][source]
            
+torch.distributed.recv(tensor, src=None, group=None, tag=0, group_src=None)[source][source]
 
            Receives a tensor synchronously.
            
 
            
 
            Warning
            
@@ -1684,7 +1684,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.isend(tensor, dst=None, group=None, tag=0, group_dst=None)[source][source]
            
+torch.distributed.isend(tensor, dst=None, group=None, tag=0, group_dst=None)[source][source]
 
            Send a tensor asynchronously.
            
 
            
 
            Warning
            
@@ -1719,7 +1719,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.irecv(tensor, src=None, group=None, tag=0, group_src=None)[source][source]
            
+torch.distributed.irecv(tensor, src=None, group=None, tag=0, group_src=None)[source][source]
 
            Receives a tensor asynchronously.
            
 
            
 
            Warning
            
@@ -1750,7 +1750,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.send_object_list(object_list, dst=None, group=None, device=None, group_dst=None)[source][source]
            
+torch.distributed.send_object_list(object_list, dst=None, group=None, device=None, group_dst=None)[source][source]
 
            Sends picklable objects in object_list synchronously.
            
 
            Similar to send(), but Python objects can be passed in.
 Note that all objects in object_list must be picklable in order to be
@@ -1819,7 +1819,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.recv_object_list(object_list, src=None, group=None, device=None, group_src=None)[source][source]
            
+torch.distributed.recv_object_list(object_list, src=None, group=None, device=None, group_src=None)[source][source]
 
            Receives picklable objects in object_list synchronously.
            
 
            Similar to recv(), but can receive Python objects.
            
 
            
@@ -1886,7 +1886,7 @@ 
            Point-to-point communication
 
            
-torch.distributed.batch_isend_irecv(p2p_op_list)[source][source]
            
+torch.distributed.batch_isend_irecv(p2p_op_list)[source][source]
 
            Send or Receive a batch of tensors asynchronously and return a list of requests.
            
 
            Process each of the operations in p2p_op_list and return the corresponding
 requests. NCCL, Gloo, and UCC backend are currently supported.
            
@@ -1935,7 +1935,7 @@ 
            Point-to-point communication
 
            
-class torch.distributed.P2POp(op, tensor, peer=None, group=None, tag=0, group_peer=None)[source][source]
            
+class torch.distributed.P2POp(op, tensor, peer=None, group=None, tag=0, group_peer=None)[source][source]
 
            A class to build point-to-point operations for batch_isend_irecv.
            
 
            This class builds the type of P2P operation, communication buffer, peer rank,
 Process Group, and tag. Instances of this class will be passed to
@@ -2010,7 +2010,7 @@ 
            Synchronous and asynchronous collective operationsCollective functions
            
 
            
 
            
-torch.distributed.broadcast(tensor, src=None, group=None, async_op=False, group_src=None)[source][source]
            
+torch.distributed.broadcast(tensor, src=None, group=None, async_op=False, group_src=None)[source][source]
 
            Broadcasts the tensor to the whole group.
            
 
            tensor must have the same number of elements in all processes
 participating in the collective.
            
@@ -2036,7 +2036,7 @@ 
            Collective functions
 
            
-torch.distributed.broadcast_object_list(object_list, src=None, group=None, device=None, group_src=None)[source][source]
            
+torch.distributed.broadcast_object_list(object_list, src=None, group=None, device=None, group_src=None)[source][source]
 
            Broadcasts picklable objects in object_list to the whole group.
            
 
            Similar to broadcast(), but Python objects can be passed in.
 Note that all objects in object_list must be picklable in order to be
@@ -2112,7 +2112,7 @@ 
            Collective functions
 
            
-torch.distributed.all_reduce(tensor, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
            
+torch.distributed.all_reduce(tensor, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
 
            Reduces the tensor data across all machines in a way that all get the final result.
            
 
            After the call tensor is going to be bitwise identical in all processes.
            
 
            Complex tensors are supported.
            
@@ -2166,7 +2166,7 @@ 
            Collective functions
 
            
-torch.distributed.reduce(tensor, dst=None, op=<RedOpType.SUM: 0>, group=None, async_op=False, group_dst=None)[source][source]
            
+torch.distributed.reduce(tensor, dst=None, op=<RedOpType.SUM: 0>, group=None, async_op=False, group_dst=None)[source][source]
 
            Reduces the tensor data across all machines.
            
 
            Only the process with rank dst is going to receive the final result.
            
 
            
@@ -2194,7 +2194,7 @@ 
            Collective functions
 
            
-torch.distributed.all_gather(tensor_list, tensor, group=None, async_op=False)[source][source]
            
+torch.distributed.all_gather(tensor_list, tensor, group=None, async_op=False)[source][source]
 
            Gathers tensors from the whole group in a list.
            
 
            Complex and uneven sized tensors are supported.
            
 
            
@@ -2258,7 +2258,7 @@ 
            Collective functions
 
            
-torch.distributed.all_gather_into_tensor(output_tensor, input_tensor, group=None, async_op=False)[source][source]
            
+torch.distributed.all_gather_into_tensor(output_tensor, input_tensor, group=None, async_op=False)[source][source]
 
            Gather tensors from all ranks and put them in a single output tensor.
            
 
            This function requires all tensors to be the same size on each process.
            
 
            
@@ -2317,7 +2317,7 @@ 
            Collective functions
 
            
-torch.distributed.all_gather_object(object_list, obj, group=None)[source][source]
            
+torch.distributed.all_gather_object(object_list, obj, group=None)[source][source]
 
            Gathers picklable objects from the whole group into a list.
            
 
            Similar to all_gather(), but Python objects can be passed in.
 Note that the object must be picklable in order to be gathered.
            
@@ -2383,7 +2383,7 @@ 
            Collective functions
 
            
-torch.distributed.gather(tensor, gather_list=None, dst=None, group=None, async_op=False, group_dst=None)[source][source]
            
+torch.distributed.gather(tensor, gather_list=None, dst=None, group=None, async_op=False, group_dst=None)[source][source]
 
            Gathers a list of tensors in a single process.
            
 
            This function requires all tensors to be the same size on each process.
            
 
            
@@ -2432,7 +2432,7 @@ 
            Collective functions
 
            
-torch.distributed.gather_object(obj, object_gather_list=None, dst=None, group=None, group_dst=None)[source][source]
            
+torch.distributed.gather_object(obj, object_gather_list=None, dst=None, group=None, group_dst=None)[source][source]
 
            Gathers picklable objects from the whole group in a single process.
            
 
            Similar to gather(), but Python objects can be passed in. Note that the
 object must be picklable in order to be gathered.
            
@@ -2506,7 +2506,7 @@ 
            Collective functions
 
            
-torch.distributed.scatter(tensor, scatter_list=None, src=None, group=None, async_op=False, group_src=None)[source][source]
            
+torch.distributed.scatter(tensor, scatter_list=None, src=None, group=None, async_op=False, group_src=None)[source][source]
 
            Scatters a list of tensors to all processes in a group.
            
 
            Each process will receive exactly one tensor and store its data in the
 tensor argument.
            
@@ -2561,7 +2561,7 @@ 
            Collective functions
 
            
-torch.distributed.scatter_object_list(scatter_object_output_list, scatter_object_input_list=None, src=None, group=None, group_src=None)[source][source]
            
+torch.distributed.scatter_object_list(scatter_object_output_list, scatter_object_input_list=None, src=None, group=None, group_src=None)[source][source]
 
            Scatters picklable objects in scatter_object_input_list to the whole group.
            
 
            Similar to scatter(), but Python objects can be passed in. On
 each rank, the scattered object will be stored as the first element of
@@ -2629,7 +2629,7 @@ 
            Collective functions
 
            
-torch.distributed.reduce_scatter(output, input_list, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
            
+torch.distributed.reduce_scatter(output, input_list, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
 
            Reduces, then scatters a list of tensors to all processes in a group.
            
 
            
 
            Parameters
            
@@ -2653,7 +2653,7 @@ 
            Collective functions
 
            
-torch.distributed.reduce_scatter_tensor(output, input, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
            
+torch.distributed.reduce_scatter_tensor(output, input, op=<RedOpType.SUM: 0>, group=None, async_op=False)[source][source]
 
            Reduces, then scatters a tensor to all ranks in a group.
            
 
            
 
            Parameters
            
@@ -2713,7 +2713,7 @@ 
            Collective functions
 
            
-torch.distributed.all_to_all_single(output, input, output_split_sizes=None, input_split_sizes=None, group=None, async_op=False)[source][source]
            
+torch.distributed.all_to_all_single(output, input, output_split_sizes=None, input_split_sizes=None, group=None, async_op=False)[source][source]
 
            Split input tensor and then scatter the split list to all processes in a group.
            
 
            Later the received tensors are concatenated from all the processes in the group
 and returned as a single output tensor.
            
@@ -2813,7 +2813,7 @@ 
            Collective functions
 
            
-torch.distributed.all_to_all(output_tensor_list, input_tensor_list, group=None, async_op=False)[source][source]
            
+torch.distributed.all_to_all(output_tensor_list, input_tensor_list, group=None, async_op=False)[source][source]
 
            Scatters list of input tensors to all processes in a group and return gathered list of tensors in output list.
            
 
            Complex tensors are supported.
            
 
            
@@ -2913,7 +2913,7 @@ 
            Collective functions
 
            
-torch.distributed.barrier(group=None, async_op=False, device_ids=None)[source][source]
            
+torch.distributed.barrier(group=None, async_op=False, device_ids=None)[source][source]
 
            Synchronize all processes.
            
 
            This collective blocks processes until the whole group enters this function,
 if async_op is False, or if async work handle is called on wait().
            
@@ -2939,7 +2939,7 @@ 
            Collective functions
 
            
-torch.distributed.monitored_barrier(group=None, timeout=None, wait_all_ranks=False)[source][source]
            
+torch.distributed.monitored_barrier(group=None, timeout=None, wait_all_ranks=False)[source][source]
 
            Synchronize processes similar to torch.distributed.barrier, but consider a configurable timeout.
            
 
            It is able to report ranks that did not pass this barrier within the provided timeout.
 Specifically, for non-zero ranks, will block until a send/recv is processed from rank 0.
@@ -4185,7 +4185,7 @@ 
            Logging
 
            
-torch.distributed.breakpoint(rank=0, skip=0)[source][source]
            
+torch.distributed.breakpoint(rank=0, skip=0)[source][source]
 
            Set a breakpoint, but only on a single rank.  All other ranks will wait for you to be
 done with the breakpoint before continuing.
            
 
            
diff --git a/2.7/distributed.optim.html b/2.7/distributed.optim.html
index 9c6f709b68f..22379271609 100644
--- a/2.7/distributed.optim.html
+++ b/2.7/distributed.optim.html
@@ -603,7 +603,7 @@ 
            Distributed Optimizers
 
            
 
            
-class torch.distributed.optim.DistributedOptimizer(optimizer_class, params_rref, *args, **kwargs)[source][source]
            
+class torch.distributed.optim.DistributedOptimizer(optimizer_class, params_rref, *args, **kwargs)[source][source]
 
            DistributedOptimizer takes remote references to parameters scattered
 across workers and applies the given optimizer locally for each parameter.
            
 
            This class uses get_gradients() in order
@@ -663,7 +663,7 @@ 
            Distributed Optimizers
 
            
 
            
-step(context_id)[source][source]
            
+step(context_id)[source][source]
 
            Performs a single optimization step.
            
 
            This will call torch.optim.Optimizer.step() on each worker
 containing parameters to be optimized, and will block until all workers
@@ -682,7 +682,7 @@ 
            Distributed Optimizers
 
            
-class torch.distributed.optim.PostLocalSGDOptimizer(optim, averager)[source][source]
            
+class torch.distributed.optim.PostLocalSGDOptimizer(optim, averager)[source][source]
 
            Wraps an arbitrary torch.optim.Optimizer and runs post-local SGD,
 This optimizer runs local optimizer at every step.
 After the warm-up stage, it averages parameters periodically afer the local optimizer is applied.
            
@@ -734,7 +734,7 @@ 
            Distributed Optimizers
 
            
 
            
-load_state_dict(state_dict)[source][source]
            
+load_state_dict(state_dict)[source][source]
 
            This is the same as torch.optim.Optimizer load_state_dict(),
 but also restores model averager’s step value to the one
 saved in the provided state_dict.
            
@@ -744,7 +744,7 @@ 
            Distributed Optimizers
 
            
-state_dict()[source][source]
            
+state_dict()[source][source]
 
            This is the same as torch.optim.Optimizer state_dict(),
 but adds an extra entry to record model averager’s step to the checkpoint
 to ensure reload does not cause unnecessary warm up again.
            
@@ -752,7 +752,7 @@ 
            Distributed Optimizers
 
            
-step()[source][source]
            
+step()[source][source]
 
            Performs a single optimization step (parameter update).
            
 
            
 
@@ -760,7 +760,7 @@ 
            Distributed Optimizers
 
            
-class torch.distributed.optim.ZeroRedundancyOptimizer(params, optimizer_class, process_group=None, parameters_as_bucket_view=False, overlap_with_ddp=False, **defaults)[source][source]
            
+class torch.distributed.optim.ZeroRedundancyOptimizer(params, optimizer_class, process_group=None, parameters_as_bucket_view=False, overlap_with_ddp=False, **defaults)[source][source]
 
            Wrap an arbitrary optim.Optimizer and shards its states across ranks in the group.
            
 
            The sharing is done as described by ZeRO.
            
 
            The local optimizer instance in each rank is only
@@ -850,7 +850,7 @@ 
            Distributed Optimizers
 
            
 
            
-add_param_group(param_group)[source][source]
            
+add_param_group(param_group)[source][source]
 
            Add a parameter group to the Optimizer ‘s param_groups.
            
 
            This can be useful when fine tuning a pre-trained network, as frozen
 layers can be made trainable and added to the Optimizer as
@@ -873,7 +873,7 @@ 
            Distributed Optimizers
 
            
-consolidate_state_dict(to=0)[source][source]
            
+consolidate_state_dict(to=0)[source][source]
 
            Consolidate a list of state_dict s (one per rank) on the target rank.
            
 
            
 
            Parameters
            
@@ -901,7 +901,7 @@ 
            Distributed Optimizers
 
            
-join_hook(**kwargs)[source][source]
            
+join_hook(**kwargs)[source][source]
 
            Return the ZeRO join hook.
            
 
            It enables training on uneven inputs by
 shadowing the collective communications in the optimizer step.
            
@@ -926,7 +926,7 @@ 
            Distributed Optimizers
 
            
-load_state_dict(state_dict)[source][source]
            
+load_state_dict(state_dict)[source][source]
 
            Load the state pertaining to the given rank from the input state_dict, updating the local optimizer as needed.
            
 
            
 
            Parameters
            
@@ -945,7 +945,7 @@ 
            Distributed Optimizers
 
            
-state_dict()[source][source]
            
+state_dict()[source][source]
 
            Return the last global optimizer state known to this rank.
            
 
            
 
            Raises
            
@@ -964,7 +964,7 @@ 
            Distributed Optimizers
 
            
-step(closure=None, **kwargs)[source][source]
            
+step(closure=None, **kwargs)[source][source]
 
            Perform a single optimizer step and syncs parameters across all ranks.
            
 
            
 
            Parameters
            
diff --git a/2.7/distributed.pipelining.html b/2.7/distributed.pipelining.html
index 81fb1d2133e..7daa5d2097f 100644
--- a/2.7/distributed.pipelining.html
+++ b/2.7/distributed.pipelining.html
@@ -960,7 +960,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.SplitPoint(value)[source][source]
            
+class torch.distributed.pipelining.SplitPoint(value)[source][source]
 
            Enum representing the points at which a split can occur in the execution of a submodule.
 :ivar BEGINNING: Represents adding a split point before the execution of a certain submodule in the forward function.
 :ivar END: Represents adding a split point after the execution of a certain submodule in the forward function.
            
@@ -968,7 +968,7 @@ 
            Model Split APIs
 
            
-torch.distributed.pipelining.pipeline(module, mb_args, mb_kwargs=None, split_spec=None, split_policy=None)[source][source]
            
+torch.distributed.pipelining.pipeline(module, mb_args, mb_kwargs=None, split_spec=None, split_policy=None)[source][source]
 
            Split a module based on a specification.
            
 
            See Pipe for more details.
            
 
            
@@ -989,14 +989,14 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.Pipe(split_gm, num_stages, has_loss_and_backward, loss_spec)[source][source]
            
+class torch.distributed.pipelining.Pipe(split_gm, num_stages, has_loss_and_backward, loss_spec)[source][source]
 
            
 
            
 
            
 
 
            
 
            
-torch.distributed.pipelining.pipe_split()[source][source]
            
+torch.distributed.pipelining.pipe_split()[source][source]
 
            pipe_split is a special operator that is used to mark the boundary between
 stages in a module. It is used to split the module into stages. It is a
 no-op if your annotated module is run eagerly.
            
@@ -1017,7 +1017,7 @@ 
            Model Split APIs
            Microbatch Utilities
            
 
            
 
            
-class torch.distributed.pipelining.microbatch.TensorChunkSpec(split_dim)[source][source]
            
+class torch.distributed.pipelining.microbatch.TensorChunkSpec(split_dim)[source][source]
 
            Class used to specify chunking of inputs
            
 
            
 
            
@@ -1025,7 +1025,7 @@ 
            Model Split APIs
 
            
-torch.distributed.pipelining.microbatch.split_args_kwargs_into_chunks(args, kwargs, chunks, args_chunk_spec=None, kwargs_chunk_spec=None)[source][source]
            
+torch.distributed.pipelining.microbatch.split_args_kwargs_into_chunks(args, kwargs, chunks, args_chunk_spec=None, kwargs_chunk_spec=None)[source][source]
 
            Given a sequence of args and kwargs, split them into a number of chunks
 according to  their respective chunking specs.
            
 
            
@@ -1050,7 +1050,7 @@ 
            Model Split APIs
 
            
-torch.distributed.pipelining.microbatch.merge_chunks(chunks, chunk_spec)[source][source]
            
+torch.distributed.pipelining.microbatch.merge_chunks(chunks, chunk_spec)[source][source]
 
            Given a list of chunks, merge them into a single value according to
 the chunk spec.
            
 
            
@@ -1074,7 +1074,7 @@ 
            Model Split APIs
            Pipeline Stages
            
 
            
 
            
-class torch.distributed.pipelining.stage.PipelineStage(submodule, stage_index, num_stages, device, input_args=None, output_args=None, group=None, dw_builder=None)[source][source]
            
+class torch.distributed.pipelining.stage.PipelineStage(submodule, stage_index, num_stages, device, input_args=None, output_args=None, group=None, dw_builder=None)[source][source]
 
            A class representing a pipeline stage in a pipeline parallelism setup.
            
 
            PipelineStage assumes sequential partitioning of the model, i.e. the model is split into chunks where outputs from
 one chunk feed into inputs of the next chunk, with no skip connections.
            
@@ -1100,7 +1100,7 @@ 
            Model Split APIs
 
            
-torch.distributed.pipelining.stage.build_stage(stage_module, stage_index, pipe_info, device, group=None)[source][source]
            
+torch.distributed.pipelining.stage.build_stage(stage_module, stage_index, pipe_info, device, group=None)[source][source]
 
            Create a pipeline stage given a stage_module to be wrapped by this stage
 and pipeline information.
            
 
            
@@ -1127,7 +1127,7 @@ 
            Model Split APIs
            Pipeline Schedules
            
 
            
 
            
-class torch.distributed.pipelining.schedules.ScheduleGPipe(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.ScheduleGPipe(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            The GPipe schedule.
 Will go through all the microbatches in a fill-drain manner.
            
 
            
@@ -1136,7 +1136,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.Schedule1F1B(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.Schedule1F1B(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            The 1F1B schedule.
 Will perform one forward and one backward on the microbatches in steady state.
            
 
            
@@ -1145,7 +1145,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.ScheduleInterleaved1F1B(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.ScheduleInterleaved1F1B(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            The Interleaved 1F1B schedule.
 See https://arxiv.org/pdf/2104.04473 for details.
 Will perform one forward and one backward on the microbatches in steady
@@ -1166,7 +1166,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.ScheduleLoopedBFS(stages, n_microbatches, loss_fn=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.ScheduleLoopedBFS(stages, n_microbatches, loss_fn=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            Breadth-First Pipeline Parallelism.
 See https://arxiv.org/abs/2211.05953 for details.
 Simliar to Interleaved 1F1B, Looped BFS supports multiple stages per rank.
@@ -1179,7 +1179,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.ScheduleInterleavedZeroBubble(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.ScheduleInterleavedZeroBubble(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            The Interleaved Zero Bubble schedule.
 See https://arxiv.org/pdf/2401.10241 for details.
 Will perform one forward and one backward on inputs for the microbatches in steady
@@ -1192,7 +1192,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.ScheduleZBVZeroBubble(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.ScheduleZBVZeroBubble(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            The Zero Bubble schedule (ZBV variant).
 See https://arxiv.org/pdf/2401.10241 Section 6 for details.
            
 
            This schedules requires exactly two stages per rank.
            
@@ -1208,7 +1208,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.PipelineScheduleSingle(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.PipelineScheduleSingle(stage, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, scale_grads=True)[source][source]
 
            Base class for single-stage schedules.
 Implements the step method.
 Derived classes should implement _step_microbatches.
            
@@ -1219,7 +1219,7 @@ 
            Model Split APIs
 
            
-step(*args, target=None, losses=None, **kwargs)[source][source]
            
+step(*args, target=None, losses=None, **kwargs)[source][source]
 
            Run one iteration of the pipeline schedule with whole-batch input.
 Will chunk the input into microbatches automatically, and go through the
 microbatches according to the schedule implementation.
            
@@ -1235,7 +1235,7 @@ 
            Model Split APIs
 
            
-class torch.distributed.pipelining.schedules.PipelineScheduleMulti(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, use_full_backward=None, scale_grads=True)[source][source]
            
+class torch.distributed.pipelining.schedules.PipelineScheduleMulti(stages, n_microbatches, loss_fn=None, args_chunk_spec=None, kwargs_chunk_spec=None, output_merge_spec=None, use_full_backward=None, scale_grads=True)[source][source]
 
            Base class for multi-stage schedules.
 Implements the step method.
            
 
            Gradients are scaled by num_microbatches depending on the scale_grads argument, defaulting to True.  This setting
@@ -1245,7 +1245,7 @@ 
            Model Split APIs
 
            
-step(*args, target=None, losses=None, **kwargs)[source][source]
            
+step(*args, target=None, losses=None, **kwargs)[source][source]
 
            Run one iteration of the pipeline schedule with whole-batch input.
 Will chunk the input into microbatches automatically, and go through the
 microbatches according to the schedule implementation.
            
diff --git a/2.7/distributed.tensor.html b/2.7/distributed.tensor.html
index 9a6f63a9261..7c42ac4bad2 100644
--- a/2.7/distributed.tensor.html
+++ b/2.7/distributed.tensor.html
@@ -656,7 +656,7 @@ 
            DTensor Class APIs
 
            
-__create_chunk_list__()[source][source]
            
+__create_chunk_list__()[source][source]
 
            Return a list of ChunkStorageMetadata, which is a dataclass that describes the size/offset of the local shard/replica
 on current rank. For DTensor, each rank will have a single local shard/replica, so the returned list usually only
 has one element.
            
@@ -670,7 +670,7 @@ 
            DTensor Class APIs
 
            
-static from_local(local_tensor, device_mesh=None, placements=None, *, run_check=False, shape=None, stride=None)[source][source]
            
+static from_local(local_tensor, device_mesh=None, placements=None, *, run_check=False, shape=None, stride=None)[source][source]
 
            Create a DTensor from a local torch.Tensor on each rank
 according to the device_mesh and placements specified.
            
 
            
@@ -725,7 +725,7 @@ 
            DTensor Class APIs
 
            
-full_tensor(*, grad_placements=None)[source][source]
            
+full_tensor(*, grad_placements=None)[source][source]
 
            Return the full tensor of this DTensor. It will perform necessary collectives
 to gather the local tensors from other ranks in its DeviceMesh and concatenate
 them together. It’s a syntatic sugar of the following code:
            
@@ -756,7 +756,7 @@ 
            DTensor Class APIs
 
            
-redistribute(device_mesh=None, placements=None, *, async_op=False)[source][source]
            
+redistribute(device_mesh=None, placements=None, *, async_op=False)[source][source]
 
            redistribute performs necessary collective operations that redistribute the current
 DTensor from its current placements to a new placements, or from is current DeviceMesh
 to a new DeviceMesh. i.e. we can turn a Sharded DTensor to a Replicated DTensor by
@@ -809,7 +809,7 @@ 
            DTensor Class APIs
 
            
-to_local(*, grad_placements=None)[source][source]
            
+to_local(*, grad_placements=None)[source][source]
 
            Get the local tensor of this DTensor on its current rank. For sharding it returns
 a local shard of the logical tensor view, for replication it returns the replica on
 its current rank.
            
@@ -877,7 +877,7 @@ 
            DeviceMesh as the distributed communicatorPlacement on each DeviceMesh dimension:
            
 
            
 
            
-class torch.distributed.tensor.placement_types.Shard(dim)[source][source]
            
+class torch.distributed.tensor.placement_types.Shard(dim)[source][source]
 
            The Shard(dim) placement describes the DTensor sharding on tensor dimension
 dim over a corresponding DeviceMesh dimension, where each rank on the
 DeviceMesh dimension only holds a shard/piece of the global Tensor. The
@@ -905,7 +905,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-class torch.distributed.tensor.placement_types.Replicate[source][source]
            
+class torch.distributed.tensor.placement_types.Replicate[source][source]
 
            The Replicate() placement describes the DTensor replicating on a corresponding
 DeviceMesh dimension, where each rank on the DeviceMesh dimension holds a
 replica of the global Tensor. The Replicate placement can be used by all
@@ -916,7 +916,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-class torch.distributed.tensor.placement_types.Partial(reduce_op='sum')[source][source]
            
+class torch.distributed.tensor.placement_types.Partial(reduce_op='sum')[source][source]
 
            The Partial(reduce_op) placement describes the DTensor that is pending
 reduction on a specified DeviceMesh dimension, where each rank on the
 DeviceMesh dimension holds the partial value of the global Tensor. User can
@@ -945,7 +945,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-class torch.distributed.tensor.placement_types.Placement[source][source]
            
+class torch.distributed.tensor.placement_types.Placement[source][source]
 
            The base class for the Placement type, where it describes how a DTensor is placed onto the
 DeviceMesh. Placement and DeviceMesh together could describe the DTensor Layout.
 It is the base class of the three main DTensor Placement types: Shard, Replicate,
@@ -953,7 +953,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-is_partial()[source][source]
            
+is_partial()[source][source]
 
            
 
            Return type
            
 
            bool
            
@@ -963,7 +963,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-is_replicate()[source][source]
            
+is_replicate()[source][source]
 
            
 
            Return type
            
 
            bool
            
@@ -973,7 +973,7 @@ 
            DeviceMesh as the distributed communicator
 
            
-is_shard(dim=None)[source][source]
            
+is_shard(dim=None)[source][source]
 
            
 
            Return type
            
 
            bool
            
@@ -1013,7 +1013,7 @@ 
            Create DTensor from a logical torch.TensorDTensor s to comply with the single device semantic, which is critical for numerical correctness.
            
 
            
 
            
-torch.distributed.tensor.distribute_tensor(tensor, device_mesh=None, placements=None, *, src_data_rank=0)[source]
            
+torch.distributed.tensor.distribute_tensor(tensor, device_mesh=None, placements=None, *, src_data_rank=0)[source]
 
            Distribute a leaf torch.Tensor (i.e. nn.Parameter/buffers) to the device_mesh according
 to the placements specified. The rank of device_mesh and placements must be the
 same. The tensor to distribute is the logical or “global” tensor, and the API would use
@@ -1065,7 +1065,7 @@ 
            Create DTensor from a logical torch.Tensornn.Module level
            
 
            
 
            
-torch.distributed.tensor.distribute_module(module, device_mesh=None, partition_fn=None, input_fn=None, output_fn=None)[source]
            
+torch.distributed.tensor.distribute_module(module, device_mesh=None, partition_fn=None, input_fn=None, output_fn=None)[source]
 
            This function expose three functions to control the parameters/inputs/outputs of the module:
            
 
            1. To perform sharding on the module before runtime execution by specifying the
 partition_fn (i.e. allow user to convert Module parameters to DTensor
@@ -1113,7 +1113,7 @@ 
            DTensor Factory FunctionsDeviceMesh and Placement for the DTensor created:
            
 
            
 
            
-torch.distributed.tensor.zeros(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.zeros(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with the scalar value 0.
            
 
            
 
            Parameters
            
@@ -1144,7 +1144,7 @@ 
            DTensor Factory Functions
 
            
-torch.distributed.tensor.ones(*size, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.ones(*size, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with the scalar value 1, with the shape defined
 by the variable argument size.
            
 
            
@@ -1176,7 +1176,7 @@ 
            DTensor Factory Functions
 
            
-torch.distributed.tensor.empty(*size, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.empty(*size, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with uninitialized data. The shape of the DTensor
 is defined by the variable argument size.
            
 
            
@@ -1207,7 +1207,7 @@ 
            DTensor Factory Functions
 
            
-torch.distributed.tensor.full(size, fill_value, *, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.full(size, fill_value, *, dtype=None, layout=torch.strided, requires_grad=False, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with fill_value according to device_mesh and
 placements, with the shape defined by the argument size.
            
 
            
@@ -1242,7 +1242,7 @@ 
            DTensor Factory Functions
 
            
-torch.distributed.tensor.rand(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.rand(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with random numbers from a uniform distribution
 on the interval [0, 1). The shape of the tensor is defined by the variable
 argument size.
            
@@ -1275,7 +1275,7 @@ 
            DTensor Factory Functions
 
            
-torch.distributed.tensor.randn(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
            
+torch.distributed.tensor.randn(*size, requires_grad=False, dtype=None, layout=torch.strided, device_mesh=None, placements=None)[source]
 
            Returns a DTensor filled with random numbers from a normal distribution
 with mean 0 and variance 1. The shape of the tensor is defined by the variable
 argument size.
            
@@ -1344,7 +1344,7 @@ 
            Debugging Tools
 
            
-generate_comm_debug_tracing_table(noise_level=3)[source][source]
            
+generate_comm_debug_tracing_table(noise_level=3)[source][source]
 
            Generates detailed table displaying operations and collective tracing information
 on a module level. Amount of information is dependent on noise_level
            
 
              
@@ -1357,7 +1357,7 @@ 
              Debugging Tools
 
              
-generate_json_dump(file_name='comm_mode_log.json', noise_level=3)[source][source]
              
+generate_json_dump(file_name='comm_mode_log.json', noise_level=3)[source][source]
 
              Creates json file used to build browser visual
 0. prints module-level collective counts
 1. prints dTensor operations not included in trivial operations
@@ -1367,7 +1367,7 @@ 
              Debugging Tools
 
              
-get_comm_counts()[source][source]
              
+get_comm_counts()[source][source]
 
              Returns the communication counts as a dictionary.
              
 
              
 
              Returns
              
@@ -1381,7 +1381,7 @@ 
              Debugging Tools
 
              
-get_parameter_info()[source][source]
              
+get_parameter_info()[source][source]
 
              
 
              Return type
              
 
              dict[str, dict[str, Any]]
              
@@ -1391,7 +1391,7 @@ 
              Debugging Tools
 
              
-get_sharding_info()[source][source]
              
+get_sharding_info()[source][source]
 
              
 
              Return type
              
 
              dict[str, dict[str, Any]]
              
@@ -1401,7 +1401,7 @@ 
              Debugging Tools
 
              
-get_total_counts()[source][source]
              
+get_total_counts()[source][source]
 
              
 
              Return type
              
 
              int
              
@@ -1411,7 +1411,7 @@ 
              Debugging Tools
 
              
-log_comm_debug_tracing_table_to_file(file_name='comm_mode_log.txt', noise_level=3)[source][source]
              
+log_comm_debug_tracing_table_to_file(file_name='comm_mode_log.txt', noise_level=3)[source][source]
 
              Alternative to console CommDebugMode output, writes to file specified by the user
              
 
              
 
@@ -1420,7 +1420,7 @@ 
              Debugging Toolsvisualize_sharding():
              
 
              
 
              
-torch.distributed.tensor.debug.visualize_sharding(dtensor, header='')[source]
              
+torch.distributed.tensor.debug.visualize_sharding(dtensor, header='')[source]
 
              Visualizes sharding in the terminal for DTensor that are 1D or 2D.
              
 
              
 
              Note
              
@@ -1437,7 +1437,7 @@ 
              Experimental Features
 
              
 
              
-torch.distributed.tensor.experimental.context_parallel(mesh, *, buffers=None, buffer_seq_dims=None, no_restore_buffers=None)[source]
              
+torch.distributed.tensor.experimental.context_parallel(mesh, *, buffers=None, buffer_seq_dims=None, no_restore_buffers=None)[source]
 
              context_parallel is an experimental API to enable context
 parallelism (CP). This API performs two actions: 1) patch the SDPA
 (torch.nn.functional.scaled_dot_product_attention) with the CP-enabled
@@ -1476,7 +1476,7 @@ 
              Experimental Features
 
              
-torch.distributed.tensor.experimental.local_map(func, out_placements, in_placements=None, device_mesh=None, *, redistribute_inputs=False)[source]
              
+torch.distributed.tensor.experimental.local_map(func, out_placements, in_placements=None, device_mesh=None, *, redistribute_inputs=False)[source]
 
              local_map() is an experimental API that allows users to pass DTensor s
 to a function that is written to be applied on torch.Tensor s. It is done by extracting
 the local components of DTensor, call the function, and wrap the outputs to
@@ -1576,7 +1576,7 @@ 
              Experimental Features
 
              
-torch.distributed.tensor.experimental.register_sharding(op)[source]
              
+torch.distributed.tensor.experimental.register_sharding(op)[source]
 
              register_sharding() is an experimental API that allows users to register sharding
 strategies for an operator when the tensor inputs and outputs are DTensor.
 It can be useful when: (1) there doesn’t exist a default sharding strategy for op,
diff --git a/2.7/distributed.tensor.parallel.html b/2.7/distributed.tensor.parallel.html
index ab7bdf94553..054a6cc8d62 100644
--- a/2.7/distributed.tensor.parallel.html
+++ b/2.7/distributed.tensor.parallel.html
@@ -602,7 +602,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallelThe entrypoint to parallelize your nn.Module using Tensor Parallelism is:
              
 
              
 
              
-torch.distributed.tensor.parallel.parallelize_module(module, device_mesh=None, parallelize_plan=None, *, src_data_rank=0)[source][source]
              
+torch.distributed.tensor.parallel.parallelize_module(module, device_mesh=None, parallelize_plan=None, *, src_data_rank=0)[source][source]
 
              Apply Tensor Parallelism in PyTorch by parallelizing modules or sub-modules based on a user-specified plan.
              
 
              We parallelize module or sub_modules based on a parallelize_plan. The parallelize_plan contains
 ParallelStyle, which indicates how user wants the module or sub_module
@@ -661,7 +661,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallelTensor Parallelism supports the following parallel styles:
              
 
              
 
              
-class torch.distributed.tensor.parallel.ColwiseParallel(*, input_layouts=None, output_layouts=None, use_local_output=True)[source][source]
              
+class torch.distributed.tensor.parallel.ColwiseParallel(*, input_layouts=None, output_layouts=None, use_local_output=True)[source][source]
 
              Partition a compatible nn.Module in a column-wise fashion. Currently supports nn.Linear and nn.Embedding.
 Users can compose it together with RowwiseParallel to achieve the sharding of more complicated modules.
 (i.e. MLP, Attention)
              
@@ -705,7 +705,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallel
 
              
-class torch.distributed.tensor.parallel.RowwiseParallel(*, input_layouts=None, output_layouts=None, use_local_output=True)[source][source]
              
+class torch.distributed.tensor.parallel.RowwiseParallel(*, input_layouts=None, output_layouts=None, use_local_output=True)[source][source]
 
              Partition a compatible nn.Module in a row-wise fashion. Currently supports nn.Linear and nn.Embedding.
 Users can compose it with ColwiseParallel to achieve the sharding of more complicated modules.
 (i.e. MLP, Attention)
              
@@ -743,7 +743,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallel
 
              
-class torch.distributed.tensor.parallel.SequenceParallel(*, sequence_dim=1, use_local_output=False)[source][source]
              
+class torch.distributed.tensor.parallel.SequenceParallel(*, sequence_dim=1, use_local_output=False)[source][source]
 
              SequenceParallel replicates a compatible nn.Module parameters and runs the sharded computation with
 input sharded on the sequence dimension. This currently supports nn.LayerNorm, nn.Dropout, and the
 RMSNorm python implementation
              
@@ -797,7 +797,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallelparallelize_plan when calling parallelize_module:
              
 
              
 
              
-class torch.distributed.tensor.parallel.PrepareModuleInput(*, input_layouts=None, desired_input_layouts=None, input_kwarg_layouts=None, desired_input_kwarg_layouts=None, use_local_output=False)[source][source]
              
+class torch.distributed.tensor.parallel.PrepareModuleInput(*, input_layouts=None, desired_input_layouts=None, input_kwarg_layouts=None, desired_input_kwarg_layouts=None, use_local_output=False)[source][source]
 
              Configure the nn.Module’s inputs to convert the input tensors of the nn.Module to DTensors at runtime according to
 input_layouts, and perform layout redistribution according to the desired_input_layouts.
              
 
              
@@ -847,7 +847,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallel
 
              
-class torch.distributed.tensor.parallel.PrepareModuleOutput(*, output_layouts, desired_output_layouts, use_local_output=True)[source][source]
              
+class torch.distributed.tensor.parallel.PrepareModuleOutput(*, output_layouts, desired_output_layouts, use_local_output=True)[source][source]
 
              Configure the nn.Module’s outputs to convert the output tensors of the nn.Module to DTensors at runtime according to
 output_layouts, and perform layout redistribution according to the desired_output_layouts.
              
 
              
@@ -905,7 +905,7 @@ 
              Tensor Parallelism - torch.distributed.tensor.parallelParallelized cross-entropy loss computation (loss parallelism), is supported via the following context manager:
              
 
              
 
              
-torch.distributed.tensor.parallel.loss_parallel()[source][source]
              
+torch.distributed.tensor.parallel.loss_parallel()[source][source]
 
              A context manager that enables loss parallelism, where efficient parallelized loss computation
 can be performed when the input is sharded on the class dimension. Currently only the cross-entropy
 loss is supported.
              
diff --git a/2.7/distributions.html b/2.7/distributions.html
index f3d8847a740..74e960e710f 100644
--- a/2.7/distributions.html
+++ b/2.7/distributions.html
@@ -654,7 +654,7 @@ 
              Pathwise derivativeDistribution
              
 
              
 
              
-class torch.distributions.distribution.Distribution(batch_shape=torch.Size([]), event_shape=torch.Size([]), validate_args=None)[source][source]
              
+class torch.distributions.distribution.Distribution(batch_shape=torch.Size([]), event_shape=torch.Size([]), validate_args=None)[source][source]
 
              Bases: object
              
 
              Distribution is the abstract base class for probability distributions.
              
 
              
@@ -676,7 +676,7 @@ 
              Distribution
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              Returns the cumulative density/mass function evaluated at
 value.
              
 
              
@@ -691,7 +691,7 @@ 
              Distribution
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              Returns entropy of distribution, batched over batch_shape.
              
 
              
 
              Returns
              
@@ -705,7 +705,7 @@ 
              Distribution
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              Returns tensor containing all values supported by a discrete
 distribution. The result will enumerate over dimension 0, so the shape
 of the result will be (cardinality,) + batch_shape + event_shape
@@ -738,7 +738,7 @@ 
              Distribution
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              Returns a new distribution instance (or populates an existing instance
 provided by a derived class) with batch dimensions expanded to
 batch_shape. This method calls expand on
@@ -763,7 +763,7 @@ 
              Distribution
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              Returns the inverse cumulative density/mass function evaluated at
 value.
              
 
              
@@ -778,7 +778,7 @@ 
              Distribution
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              Returns the log of the probability density/mass function evaluated at
 value.
              
 
              
@@ -805,7 +805,7 @@ 
              Distribution
 
              
-perplexity()[source][source]
              
+perplexity()[source][source]
 
              Returns perplexity of distribution, batched over batch_shape.
              
 
              
 
              Returns
              
@@ -819,7 +819,7 @@ 
              Distribution
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              Generates a sample_shape shaped reparameterized sample or sample_shape
 shaped batch of reparameterized samples if the distribution parameters
 are batched.
              
@@ -832,7 +832,7 @@ 
              Distribution
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              Generates a sample_shape shaped sample or sample_shape shaped batch of
 samples if the distribution parameters are batched.
              
 
              
@@ -844,7 +844,7 @@ 
              Distribution
 
              
-sample_n(n)[source][source]
              
+sample_n(n)[source][source]
 
              Generates n samples or n batches of samples if the distribution
 parameters are batched.
              
 
              
@@ -856,7 +856,7 @@ 
              Distribution
 
              
-static set_default_validate_args(value)[source][source]
              
+static set_default_validate_args(value)[source][source]
 
              Sets whether validation is enabled or disabled.
              
 
              The default behavior mimics Python’s assert statement: validation
 is on by default, but is disabled if Python is run in optimized mode
@@ -895,7 +895,7 @@ 
              DistributionExponentialFamily
              
 
              
 
              
-class torch.distributions.exp_family.ExponentialFamily(batch_shape=torch.Size([]), event_shape=torch.Size([]), validate_args=None)[source][source]
              
+class torch.distributions.exp_family.ExponentialFamily(batch_shape=torch.Size([]), event_shape=torch.Size([]), validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              ExponentialFamily is the abstract base class for probability distributions belonging to an
 exponential family, whose probability mass/density function has the form is defined below
              
@@ -915,7 +915,7 @@ 
              ExponentialFamily
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              Method to compute the entropy using Bregman divergence of the log normalizer.
              
 
              
 
@@ -926,7 +926,7 @@ 
              ExponentialFamilyBernoulli
              
 
              
 
              
-class torch.distributions.bernoulli.Bernoulli(probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.bernoulli.Bernoulli(probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Bernoulli distribution parameterized by probs
 or logits (but not both).
              
@@ -953,17 +953,17 @@ 
              Bernoulli
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -973,7 +973,7 @@ 
              Bernoulli
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1003,7 +1003,7 @@ 
              Bernoulli
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -1023,7 +1023,7 @@ 
              BernoulliBeta
              
 
              
 
              
-class torch.distributions.beta.Beta(concentration1, concentration0, validate_args=None)[source][source]
              
+class torch.distributions.beta.Beta(concentration1, concentration0, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Beta distribution parameterized by concentration1 and concentration0.
              
 
              Example:
              
@@ -1059,12 +1059,12 @@ 
              Beta
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1074,7 +1074,7 @@ 
              Beta
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1089,7 +1089,7 @@ 
              Beta
 
              
-rsample(sample_shape=())[source][source]
              
+rsample(sample_shape=())[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1114,7 +1114,7 @@ 
              BetaBinomial
              
 
              
 
              
-class torch.distributions.binomial.Binomial(total_count=1, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.binomial.Binomial(total_count=1, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Binomial distribution parameterized by total_count and
 either probs or logits (but not both). total_count must be
@@ -1146,17 +1146,17 @@ 
              Binomial
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1166,7 +1166,7 @@ 
              Binomial
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1196,7 +1196,7 @@ 
              Binomial
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -1221,7 +1221,7 @@ 
              BinomialCategorical
              
 
              
 
              
-class torch.distributions.categorical.Categorical(probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.categorical.Categorical(probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a categorical distribution parameterized by either probs or
 logits (but not both).
              
@@ -1267,17 +1267,17 @@ 
              Categorical
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1287,7 +1287,7 @@ 
              Categorical
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1317,7 +1317,7 @@ 
              Categorical
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -1342,7 +1342,7 @@ 
              CategoricalCauchy
              
 
              
 
              
-class torch.distributions.cauchy.Cauchy(loc, scale, validate_args=None)[source][source]
              
+class torch.distributions.cauchy.Cauchy(loc, scale, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Samples from a Cauchy (Lorentz) distribution. The distribution of the ratio of
 independent normally distributed random variables with means 0 follows a
@@ -1368,17 +1368,17 @@ 
              Cauchy
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1388,12 +1388,12 @@ 
              Cauchy
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1408,7 +1408,7 @@ 
              Cauchy
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1433,7 +1433,7 @@ 
              CauchyChi2
              
 
              
 
              
-class torch.distributions.chi2.Chi2(df, validate_args=None)[source][source]
              
+class torch.distributions.chi2.Chi2(df, validate_args=None)[source][source]
 
              Bases: Gamma
              
 
              Creates a Chi-squared distribution parameterized by shape parameter df.
 This is exactly equivalent to Gamma(alpha=0.5*df, beta=0.5)
              
@@ -1460,7 +1460,7 @@ 
              Chi2
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1470,7 +1470,7 @@ 
              Chi2ContinuousBernoulli
              
 
              
 
              
-class torch.distributions.continuous_bernoulli.ContinuousBernoulli(probs=None, logits=None, lims=(0.499, 0.501), validate_args=None)[source][source]
              
+class torch.distributions.continuous_bernoulli.ContinuousBernoulli(probs=None, logits=None, lims=(0.499, 0.501), validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a continuous Bernoulli distribution parameterized by probs
 or logits (but not both).
              
@@ -1503,17 +1503,17 @@ 
              ContinuousBernoulli
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1523,12 +1523,12 @@ 
              ContinuousBernoulli
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1553,7 +1553,7 @@ 
              ContinuousBernoulli
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1563,7 +1563,7 @@ 
              ContinuousBernoulli
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -1588,7 +1588,7 @@ 
              ContinuousBernoulliDirichlet
              
 
              
 
              
-class torch.distributions.dirichlet.Dirichlet(concentration, validate_args=None)[source][source]
              
+class torch.distributions.dirichlet.Dirichlet(concentration, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Dirichlet distribution parameterized by concentration concentration.
              
 
              Example:
              
@@ -1610,12 +1610,12 @@ 
              Dirichlet
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1625,7 +1625,7 @@ 
              Dirichlet
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1640,7 +1640,7 @@ 
              Dirichlet
 
              
-rsample(sample_shape=())[source][source]
              
+rsample(sample_shape=())[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1665,7 +1665,7 @@ 
              DirichletExponential
              
 
              
 
              
-class torch.distributions.exponential.Exponential(rate, validate_args=None)[source][source]
              
+class torch.distributions.exponential.Exponential(rate, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Exponential distribution parameterized by rate.
              
 
              Example:
              
@@ -1686,17 +1686,17 @@ 
              Exponential
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1706,12 +1706,12 @@ 
              Exponential
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1726,7 +1726,7 @@ 
              Exponential
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1756,7 +1756,7 @@ 
              ExponentialFisherSnedecor
              
 
              
 
              
-class torch.distributions.fishersnedecor.FisherSnedecor(df1, df2, validate_args=None)[source][source]
              
+class torch.distributions.fishersnedecor.FisherSnedecor(df1, df2, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Fisher-Snedecor distribution parameterized by df1 and df2.
              
 
              Example:
              
@@ -1780,7 +1780,7 @@ 
              FisherSnedecor
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1790,7 +1790,7 @@ 
              FisherSnedecor
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1805,7 +1805,7 @@ 
              FisherSnedecor
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1830,7 +1830,7 @@ 
              FisherSnedecorGamma
              
 
              
 
              
-class torch.distributions.gamma.Gamma(concentration, rate, validate_args=None)[source][source]
              
+class torch.distributions.gamma.Gamma(concentration, rate, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Gamma distribution parameterized by shape concentration and rate.
              
 
              Example:
              
@@ -1856,17 +1856,17 @@ 
              Gamma
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -1876,7 +1876,7 @@ 
              Gamma
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1891,7 +1891,7 @@ 
              Gamma
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -1916,7 +1916,7 @@ 
              GammaGeometric
              
 
              
 
              
-class torch.distributions.geometric.Geometric(probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.geometric.Geometric(probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Geometric distribution parameterized by probs,
 where probs is the probability of success of Bernoulli trials.
              
@@ -1948,17 +1948,17 @@ 
              Geometric
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -1983,7 +1983,7 @@ 
              Geometric
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -2003,7 +2003,7 @@ 
              GeometricGumbel
              
 
              
 
              
-class torch.distributions.gumbel.Gumbel(loc, scale, validate_args=None)[source][source]
              
+class torch.distributions.gumbel.Gumbel(loc, scale, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Samples from a Gumbel Distribution.
              
 
              Examples:
              
@@ -2027,17 +2027,17 @@ 
              Gumbel
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2072,7 +2072,7 @@ 
              GumbelHalfCauchy
              
 
              
 
              
-class torch.distributions.half_cauchy.HalfCauchy(scale, validate_args=None)[source][source]
              
+class torch.distributions.half_cauchy.HalfCauchy(scale, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates a half-Cauchy distribution parameterized by scale where:
              
 
              X ~ Cauchy(0, scale)
@@ -2097,17 +2097,17 @@ 
              HalfCauchy
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2117,12 +2117,12 @@ 
              HalfCauchy
 
              
-icdf(prob)[source][source]
              
+icdf(prob)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2157,7 +2157,7 @@ 
              HalfCauchyHalfNormal
              
 
              
 
              
-class torch.distributions.half_normal.HalfNormal(scale, validate_args=None)[source][source]
              
+class torch.distributions.half_normal.HalfNormal(scale, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates a half-normal distribution parameterized by scale where:
              
 
              X ~ Normal(0, scale)
@@ -2182,17 +2182,17 @@ 
              HalfNormal
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2202,12 +2202,12 @@ 
              HalfNormal
 
              
-icdf(prob)[source][source]
              
+icdf(prob)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2242,7 +2242,7 @@ 
              HalfNormalIndependent
              
 
              
 
              
-class torch.distributions.independent.Independent(base_distribution, reinterpreted_batch_ndims, validate_args=None)[source][source]
              
+class torch.distributions.independent.Independent(base_distribution, reinterpreted_batch_ndims, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Reinterprets some of the batch dims of a distribution as event dims.
              
 
              This is mainly useful for changing the shape of the result of
@@ -2281,17 +2281,17 @@ 
              Independent
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2306,7 +2306,7 @@ 
              Independent
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2321,7 +2321,7 @@ 
              Independent
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -2331,7 +2331,7 @@ 
              Independent
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -2361,7 +2361,7 @@ 
              IndependentInverseGamma
              
 
              
 
              
-class torch.distributions.inverse_gamma.InverseGamma(concentration, rate, validate_args=None)[source][source]
              
+class torch.distributions.inverse_gamma.InverseGamma(concentration, rate, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates an inverse gamma distribution parameterized by concentration and rate
 where:
              
@@ -2397,12 +2397,12 @@ 
              InverseGamma
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2442,7 +2442,7 @@ 
              InverseGammaKumaraswamy
              
 
              
 
              
-class torch.distributions.kumaraswamy.Kumaraswamy(concentration1, concentration0, validate_args=None)[source][source]
              
+class torch.distributions.kumaraswamy.Kumaraswamy(concentration1, concentration0, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Samples from a Kumaraswamy distribution.
              
 
              Example:
              
@@ -2468,12 +2468,12 @@ 
              Kumaraswamy
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2508,7 +2508,7 @@ 
              KumaraswamyLKJCholesky
              
 
              
 
              
-class torch.distributions.lkj_cholesky.LKJCholesky(dim, concentration=1.0, validate_args=None)[source][source]
              
+class torch.distributions.lkj_cholesky.LKJCholesky(dim, concentration=1.0, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              LKJ distribution for lower Cholesky factor of correlation matrices.
 The distribution is controlled by concentration parameter η\eta
@@ -2553,17 +2553,17 @@ 
              LKJCholesky
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -2578,7 +2578,7 @@ 
              LKJCholeskyLaplace
              
 
              
 
              
-class torch.distributions.laplace.Laplace(loc, scale, validate_args=None)[source][source]
              
+class torch.distributions.laplace.Laplace(loc, scale, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Laplace distribution parameterized by loc and scale.
              
 
              Example:
              
@@ -2602,17 +2602,17 @@ 
              Laplace
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2622,12 +2622,12 @@ 
              Laplace
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2642,7 +2642,7 @@ 
              Laplace
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -2672,7 +2672,7 @@ 
              LaplaceLogNormal
              
 
              
 
              
-class torch.distributions.log_normal.LogNormal(loc, scale, validate_args=None)[source][source]
              
+class torch.distributions.log_normal.LogNormal(loc, scale, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates a log-normal distribution parameterized by
 loc and scale where:
              
@@ -2701,12 +2701,12 @@ 
              LogNormal
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2751,7 +2751,7 @@ 
              LogNormalLowRankMultivariateNormal
              
 
              
 
              
-class torch.distributions.lowrank_multivariate_normal.LowRankMultivariateNormal(loc, cov_factor, cov_diag, validate_args=None)[source][source]
              
+class torch.distributions.lowrank_multivariate_normal.LowRankMultivariateNormal(loc, cov_factor, cov_diag, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a multivariate normal distribution with covariance matrix having a low-rank form
 parameterized by cov_factor and cov_diag:
              
@@ -2800,12 +2800,12 @@ 
              LowRankMultivariateNormal
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2815,7 +2815,7 @@ 
              LowRankMultivariateNormal
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -2835,7 +2835,7 @@ 
              LowRankMultivariateNormal
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -2865,7 +2865,7 @@ 
              LowRankMultivariateNormalMixtureSameFamily
              
 
              
 
              
-class torch.distributions.mixture_same_family.MixtureSameFamily(mixture_distribution, component_distribution, validate_args=None)[source][source]
              
+class torch.distributions.mixture_same_family.MixtureSameFamily(mixture_distribution, component_distribution, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              The MixtureSameFamily distribution implements a (batch of) mixture
 distribution where all component are from different parameterizations of
@@ -2916,7 +2916,7 @@ 
              MixtureSameFamily
 
              
-cdf(x)[source][source]
              
+cdf(x)[source][source]
 
              
 
 
              
@@ -2926,7 +2926,7 @@ 
              MixtureSameFamily
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -2936,7 +2936,7 @@ 
              MixtureSameFamily
 
              
-log_prob(x)[source][source]
              
+log_prob(x)[source][source]
 
              
 
 
              
@@ -2951,7 +2951,7 @@ 
              MixtureSameFamily
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -2976,7 +2976,7 @@ 
              MixtureSameFamilyMultinomial
              
 
              
 
              
-class torch.distributions.multinomial.Multinomial(total_count=1, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.multinomial.Multinomial(total_count=1, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Multinomial distribution parameterized by total_count and
 either probs or logits (but not both). The innermost dimension of
@@ -3024,17 +3024,17 @@ 
              Multinomial
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3059,7 +3059,7 @@ 
              Multinomial
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -3089,7 +3089,7 @@ 
              MultinomialMultivariateNormal
              
 
              
 
              
-class torch.distributions.multivariate_normal.MultivariateNormal(loc, covariance_matrix=None, precision_matrix=None, scale_tril=None, validate_args=None)[source][source]
              
+class torch.distributions.multivariate_normal.MultivariateNormal(loc, covariance_matrix=None, precision_matrix=None, scale_tril=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a multivariate normal (also called Gaussian) distribution
 parameterized by a mean vector and a covariance matrix.
              
@@ -3137,12 +3137,12 @@ 
              MultivariateNormal
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3152,7 +3152,7 @@ 
              MultivariateNormal
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3172,7 +3172,7 @@ 
              MultivariateNormal
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -3202,7 +3202,7 @@ 
              MultivariateNormalNegativeBinomial
              
 
              
 
              
-class torch.distributions.negative_binomial.NegativeBinomial(total_count, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.negative_binomial.NegativeBinomial(total_count, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Negative Binomial distribution, i.e. distribution
 of the number of successful independent and identical Bernoulli trials
@@ -3226,12 +3226,12 @@ 
              NegativeBinomial
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3261,7 +3261,7 @@ 
              NegativeBinomial
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -3281,7 +3281,7 @@ 
              NegativeBinomialNormal
              
 
              
 
              
-class torch.distributions.normal.Normal(loc, scale, validate_args=None)[source][source]
              
+class torch.distributions.normal.Normal(loc, scale, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a normal (also called Gaussian) distribution parameterized by
 loc and scale.
              
@@ -3307,17 +3307,17 @@ 
              Normal
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3327,12 +3327,12 @@ 
              Normal
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3347,7 +3347,7 @@ 
              Normal
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -3357,7 +3357,7 @@ 
              Normal
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -3382,7 +3382,7 @@ 
              NormalOneHotCategorical
              
 
              
 
              
-class torch.distributions.one_hot_categorical.OneHotCategorical(probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.one_hot_categorical.OneHotCategorical(probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a one-hot categorical distribution parameterized by probs or
 logits.
              
@@ -3420,17 +3420,17 @@ 
              OneHotCategorical
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-enumerate_support(expand=True)[source][source]
              
+enumerate_support(expand=True)[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3440,7 +3440,7 @@ 
              OneHotCategorical
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3470,7 +3470,7 @@ 
              OneHotCategorical
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -3490,7 +3490,7 @@ 
              OneHotCategoricalPareto
              
 
              
 
              
-class torch.distributions.pareto.Pareto(scale, alpha, validate_args=None)[source][source]
              
+class torch.distributions.pareto.Pareto(scale, alpha, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Samples from a Pareto Type 1 distribution.
              
 
              Example:
              
@@ -3514,7 +3514,7 @@ 
              Pareto
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -3524,7 +3524,7 @@ 
              Pareto
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
              Return type
              
 
              Pareto
              
@@ -3564,7 +3564,7 @@ 
              ParetoPoisson
              
 
              
 
              
-class torch.distributions.poisson.Poisson(rate, validate_args=None)[source][source]
              
+class torch.distributions.poisson.Poisson(rate, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Poisson distribution parameterized by rate, the rate parameter.
              
 
              Samples are nonnegative integers, with a pmf given by
              
@@ -3589,12 +3589,12 @@ 
              Poisson
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3609,7 +3609,7 @@ 
              Poisson
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              
 
 
              
@@ -3629,7 +3629,7 @@ 
              PoissonRelaxedBernoulli
              
 
              
 
              
-class torch.distributions.relaxed_bernoulli.RelaxedBernoulli(temperature, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.relaxed_bernoulli.RelaxedBernoulli(temperature, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates a RelaxedBernoulli distribution, parametrized by
 temperature, and either probs or logits
@@ -3658,7 +3658,7 @@ 
              RelaxedBernoulli
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3693,7 +3693,7 @@ 
              RelaxedBernoulliLogitRelaxedBernoulli
              
 
              
 
              
-class torch.distributions.relaxed_bernoulli.LogitRelaxedBernoulli(temperature, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.relaxed_bernoulli.LogitRelaxedBernoulli(temperature, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a LogitRelaxedBernoulli distribution parameterized by probs
 or logits (but not both), which is the logit of a RelaxedBernoulli
@@ -3719,12 +3719,12 @@ 
              LogitRelaxedBernoulli
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3744,7 +3744,7 @@ 
              LogitRelaxedBernoulli
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -3764,7 +3764,7 @@ 
              LogitRelaxedBernoulliRelaxedOneHotCategorical
              
 
              
 
              
-class torch.distributions.relaxed_categorical.RelaxedOneHotCategorical(temperature, probs=None, logits=None, validate_args=None)[source][source]
              
+class torch.distributions.relaxed_categorical.RelaxedOneHotCategorical(temperature, probs=None, logits=None, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Creates a RelaxedOneHotCategorical distribution parametrized by
 temperature, and either probs or logits.
@@ -3793,7 +3793,7 @@ 
              RelaxedOneHotCategorical
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3828,7 +3828,7 @@ 
              RelaxedOneHotCategoricalStudentT
              
 
              
 
              
-class torch.distributions.studentT.StudentT(df, loc=0.0, scale=1.0, validate_args=None)[source][source]
              
+class torch.distributions.studentT.StudentT(df, loc=0.0, scale=1.0, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Creates a Student’s t-distribution parameterized by degree of
 freedom df, mean loc and scale scale.
              
@@ -3854,12 +3854,12 @@ 
              StudentT
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3869,7 +3869,7 @@ 
              StudentT
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -3884,7 +3884,7 @@ 
              StudentT
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -3909,7 +3909,7 @@ 
              StudentTTransformedDistribution
              
 
              
 
              
-class torch.distributions.transformed_distribution.TransformedDistribution(base_distribution, transforms, validate_args=None)[source][source]
              
+class torch.distributions.transformed_distribution.TransformedDistribution(base_distribution, transforms, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Extension of the Distribution class, which applies a sequence of Transforms
 to a base distribution.  Let f be the composition of transforms applied:
              
@@ -3947,14 +3947,14 @@ 
              TransformedDistribution
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              Computes the cumulative distribution function by inverting the
 transform(s) and computing the score of the base distribution.
              
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -3964,21 +3964,21 @@ 
              TransformedDistribution
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              Computes the inverse cumulative distribution function using
 transform(s) and computing the score of the base distribution.
              
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              Scores the sample by inverting the transform(s) and computing the score
 using the score of the base distribution and the log abs det jacobian.
              
 
              
 
 
              
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              Generates a sample_shape shaped reparameterized sample or sample_shape
 shaped batch of reparameterized samples if the distribution parameters
 are batched. Samples first from base distribution and applies
@@ -3992,7 +3992,7 @@ 
              TransformedDistribution
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              Generates a sample_shape shaped sample or sample_shape shaped batch of
 samples if the distribution parameters are batched. Samples first from
 base distribution and applies transform() for every transform in the
@@ -4016,7 +4016,7 @@ 
              TransformedDistributionUniform
              
 
              
 
              
-class torch.distributions.uniform.Uniform(low, high, validate_args=None)[source][source]
              
+class torch.distributions.uniform.Uniform(low, high, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              Generates uniformly distributed random samples from the half-open interval
 [low, high).
              
@@ -4041,17 +4041,17 @@ 
              Uniform
 
              
-cdf(value)[source][source]
              
+cdf(value)[source][source]
 
              
 
 
              
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -4061,12 +4061,12 @@ 
              Uniform
 
              
-icdf(value)[source][source]
              
+icdf(value)[source][source]
 
              
 
 
              
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -4081,7 +4081,7 @@ 
              Uniform
 
              
-rsample(sample_shape=torch.Size([]))[source][source]
              
+rsample(sample_shape=torch.Size([]))[source][source]
 
              
 
              Return type
              
 
              Tensor
              
@@ -4116,7 +4116,7 @@ 
              UniformVonMises
              
 
              
 
              
-class torch.distributions.von_mises.VonMises(loc, concentration, validate_args=None)[source][source]
              
+class torch.distributions.von_mises.VonMises(loc, concentration, validate_args=None)[source][source]
 
              Bases: Distribution
              
 
              A circular von Mises distribution.
              
 
              This implementation uses polar coordinates. The loc and value args
@@ -4145,7 +4145,7 @@ 
              VonMises
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -4155,7 +4155,7 @@ 
              VonMises
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -4171,7 +4171,7 @@ 
              VonMises
 
              
-sample(sample_shape=torch.Size([]))[source][source]
              
+sample(sample_shape=torch.Size([]))[source][source]
 
              The sampling algorithm for the von Mises distribution is based on the
 following paper: D.J. Best and N.I. Fisher, “Efficient simulation of the
 von Mises distribution.” Applied Statistics (1979): 152-157.
              
@@ -4198,7 +4198,7 @@ 
              VonMisesWeibull
              
 
              
 
              
-class torch.distributions.weibull.Weibull(scale, concentration, validate_args=None)[source][source]
              
+class torch.distributions.weibull.Weibull(scale, concentration, validate_args=None)[source][source]
 
              Bases: TransformedDistribution
              
 
              Samples from a two-parameter Weibull distribution.
              
 
              Example
              
@@ -4222,12 +4222,12 @@ 
              Weibull
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -4257,7 +4257,7 @@ 
              WeibullWishart
              
 
              
 
              
-class torch.distributions.wishart.Wishart(df, covariance_matrix=None, precision_matrix=None, scale_tril=None, validate_args=None)[source][source]
              
+class torch.distributions.wishart.Wishart(df, covariance_matrix=None, precision_matrix=None, scale_tril=None, validate_args=None)[source][source]
 
              Bases: ExponentialFamily
              
 
              Creates a Wishart distribution parameterized by a symmetric positive definite matrix Σ\Sigma,
 or its Cholesky decomposition Σ=LL⊤\mathbf{\Sigma} = \mathbf{L}\mathbf{L}^\top
              
@@ -4305,12 +4305,12 @@ 
              Wishart
 
              
-entropy()[source][source]
              
+entropy()[source][source]
 
              
 
 
              
 
              
-expand(batch_shape, _instance=None)[source][source]
              
+expand(batch_shape, _instance=None)[source][source]
 
              
 
 
              
@@ -4320,7 +4320,7 @@ 
              Wishart
 
              
-log_prob(value)[source][source]
              
+log_prob(value)[source][source]
 
              
 
 
              
@@ -4340,7 +4340,7 @@ 
              Wishart
 
              
-rsample(sample_shape=torch.Size([]), max_try_correction=None)[source][source]
              
+rsample(sample_shape=torch.Size([]), max_try_correction=None)[source][source]
 
              
 
              Warning
              
 
              In some cases, sampling algorithm based on Bartlett decomposition may return singular matrix samples.
@@ -4378,7 +4378,7 @@ 
              Wishart
              KL Divergence
              
 
              
 
              
-torch.distributions.kl.kl_divergence(p, q)[source][source]
              
+torch.distributions.kl.kl_divergence(p, q)[source][source]
 
              Compute Kullback-Leibler divergence KL(p∥q)KL(p \| q) between two distributions.
              
 
              
 KL(p∥q)=∫p(x)log⁡p(x)q(x) dxKL(p \| q) = \int p(x) \log\frac {p(x)} {q(x)} \,dx
              
@@ -4495,7 +4495,7 @@ 
              Wishart
 
              
-torch.distributions.kl.register_kl(type_p, type_q)[source][source]
              
+torch.distributions.kl.register_kl(type_p, type_q)[source][source]
 
              Decorator to register a pairwise function with kl_divergence().
 Usage:
              
 
              @register_kl(Normal, Normal)
@@ -4531,13 +4531,13 @@ 
              Wishart
              Transforms
              
 
              
 
              
-class torch.distributions.transforms.AbsTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.AbsTransform(cache_size=0)[source][source]
 
              Transform via the mapping y=∣x∣y = |x|.
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.AffineTransform(loc, scale, event_dim=0, cache_size=0)[source][source]
              
+class torch.distributions.transforms.AffineTransform(loc, scale, event_dim=0, cache_size=0)[source][source]
 
              Transform via the pointwise affine mapping y=loc+scale×xy = \text{loc} + \text{scale} \times x.
              
 
              
 
              Parameters
              
@@ -4554,7 +4554,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.CatTransform(tseq, dim=0, lengths=None, cache_size=0)[source][source]
              
+class torch.distributions.transforms.CatTransform(tseq, dim=0, lengths=None, cache_size=0)[source][source]
 
              Transform functor that applies a sequence of transforms tseq
 component-wise to each submatrix at dim, of length lengths[dim],
 in a way compatible with torch.cat().
              
@@ -4570,7 +4570,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.ComposeTransform(parts, cache_size=0)[source][source]
              
+class torch.distributions.transforms.ComposeTransform(parts, cache_size=0)[source][source]
 
              Composes multiple transforms in a chain.
 The transforms being composed are responsible for caching.
              
 
              
@@ -4586,7 +4586,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.CorrCholeskyTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.CorrCholeskyTransform(cache_size=0)[source][source]
 
              Transforms an uncontrained real vector xx with length D∗(D−1)/2D*(D-1)/2 into the
 Cholesky factor of a D-dimension correlation matrix. This Cholesky factor is a lower
 triangular matrix with positive diagonals and unit Euclidean norm for each row.
@@ -4618,7 +4618,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.CumulativeDistributionTransform(distribution, cache_size=0)[source][source]
              
+class torch.distributions.transforms.CumulativeDistributionTransform(distribution, cache_size=0)[source][source]
 
              Transform via the cumulative distribution function of a probability distribution.
              
 
              
 
              Parameters
              
@@ -4640,13 +4640,13 @@ 
              Wishart
 
              
-class torch.distributions.transforms.ExpTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.ExpTransform(cache_size=0)[source][source]
 
              Transform via the mapping y=exp⁡(x)y = \exp(x).
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.IndependentTransform(base_transform, reinterpreted_batch_ndims, cache_size=0)[source][source]
              
+class torch.distributions.transforms.IndependentTransform(base_transform, reinterpreted_batch_ndims, cache_size=0)[source][source]
 
              Wrapper around another transform to treat
 reinterpreted_batch_ndims-many extra of the right most dimensions as
 dependent. This has no effect on the forward or backward transforms, but
@@ -4665,7 +4665,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.LowerCholeskyTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.LowerCholeskyTransform(cache_size=0)[source][source]
 
              Transform from unconstrained matrices to lower-triangular matrices with
 nonnegative diagonal entries.
              
 
              This is useful for parameterizing positive definite matrices in terms of
@@ -4674,19 +4674,19 @@ 
              Wishart
 
              
-class torch.distributions.transforms.PositiveDefiniteTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.PositiveDefiniteTransform(cache_size=0)[source][source]
 
              Transform from unconstrained matrices to positive-definite matrices.
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.PowerTransform(exponent, cache_size=0)[source][source]
              
+class torch.distributions.transforms.PowerTransform(exponent, cache_size=0)[source][source]
 
              Transform via the mapping y=xexponenty = x^{\text{exponent}}.
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.ReshapeTransform(in_shape, out_shape, cache_size=0)[source][source]
              
+class torch.distributions.transforms.ReshapeTransform(in_shape, out_shape, cache_size=0)[source][source]
 
              Unit Jacobian transform to reshape the rightmost part of a tensor.
              
 
              Note that in_shape and out_shape must have the same number of
 elements, just as for torch.Tensor.reshape().
              
@@ -4704,20 +4704,20 @@ 
              Wishart
 
              
-class torch.distributions.transforms.SigmoidTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.SigmoidTransform(cache_size=0)[source][source]
 
              Transform via the mapping y=11+exp⁡(−x)y = \frac{1}{1 + \exp(-x)} and x=logit(y)x = \text{logit}(y).
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.SoftplusTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.SoftplusTransform(cache_size=0)[source][source]
 
              Transform via the mapping Softplus(x)=log⁡(1+exp⁡(x))\text{Softplus}(x) = \log(1 + \exp(x)).
 The implementation reverts to the linear function when x>20x > 20.
              
 
              
 
 
              
 
              
-class torch.distributions.transforms.TanhTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.TanhTransform(cache_size=0)[source][source]
 
              Transform via the mapping y=tanh⁡(x)y = \tanh(x).
              
 
              It is equivalent to
              
 
              ComposeTransform(
@@ -4736,7 +4736,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.SoftmaxTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.SoftmaxTransform(cache_size=0)[source][source]
 
              Transform from unconstrained space to the simplex via y=exp⁡(x)y = \exp(x) then
 normalizing.
              
 
              This is not bijective and cannot be used for HMC. However this acts mostly
@@ -4746,7 +4746,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.StackTransform(tseq, dim=0, cache_size=0)[source][source]
              
+class torch.distributions.transforms.StackTransform(tseq, dim=0, cache_size=0)[source][source]
 
              Transform functor that applies a sequence of transforms tseq
 component-wise to each submatrix at dim
 in a way compatible with torch.stack().
              
@@ -4760,7 +4760,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.StickBreakingTransform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.StickBreakingTransform(cache_size=0)[source][source]
 
              Transform from unconstrained space to the simplex of one additional
 dimension via a stick-breaking process.
              
 
              This transform arises as an iterated sigmoid transform in a stick-breaking
@@ -4773,7 +4773,7 @@ 
              Wishart
 
              
-class torch.distributions.transforms.Transform(cache_size=0)[source][source]
              
+class torch.distributions.transforms.Transform(cache_size=0)[source][source]
 
              Abstract class for invertable transformations with computable log
 det jacobians. They are primarily used in
 torch.distributions.TransformedDistribution.
              
@@ -4832,20 +4832,20 @@ 
              Wishart
 
              
-log_abs_det_jacobian(x, y)[source][source]
              
+log_abs_det_jacobian(x, y)[source][source]
 
              Computes the log det jacobian log |dy/dx| given input and output.
              
 
              
 
 
              
 
              
-forward_shape(shape)[source][source]
              
+forward_shape(shape)[source][source]
 
              Infers the shape of the forward computation, given the input shape.
 Defaults to preserving shape.
              
 
              
 
 
              
 
              
-inverse_shape(shape)[source][source]
              
+inverse_shape(shape)[source][source]
 
              Infers the shapes of the inverse computation, given the output shape.
 Defaults to preserving shape.
              
 
              
@@ -4857,7 +4857,7 @@ 
              Wishart
              Constraints
              
 
              
 
              
-class torch.distributions.constraints.Constraint[source][source]
              
+class torch.distributions.constraints.Constraint[source][source]
 
              Abstract base class for constraints.
              
 
              A constraint object represents a region over which a variable is valid,
 e.g. within which a variable can be optimized.
              
@@ -4874,7 +4874,7 @@ 
              Wishart
 
              
-check(value)[source][source]
              
+check(value)[source][source]
 
              Returns a byte tensor of sample_shape + batch_shape indicating
 whether each event in value satisfies this constraint.
              
 
              
@@ -4883,55 +4883,55 @@ 
              Wishart
 
              
-torch.distributions.constraints.cat[source]
              
+torch.distributions.constraints.cat[source]
 
              alias of _Cat
              
 
              
 
 
              
 
              
-torch.distributions.constraints.dependent_property[source]
              
+torch.distributions.constraints.dependent_property[source]
 
              alias of _DependentProperty
              
 
              
 
 
              
 
              
-torch.distributions.constraints.greater_than[source]
              
+torch.distributions.constraints.greater_than[source]
 
              alias of _GreaterThan
              
 
              
 
 
              
 
              
-torch.distributions.constraints.greater_than_eq[source]
              
+torch.distributions.constraints.greater_than_eq[source]
 
              alias of _GreaterThanEq
              
 
              
 
 
              
 
              
-torch.distributions.constraints.independent[source]
              
+torch.distributions.constraints.independent[source]
 
              alias of _IndependentConstraint
              
 
              
 
 
              
 
              
-torch.distributions.constraints.integer_interval[source]
              
+torch.distributions.constraints.integer_interval[source]
 
              alias of _IntegerInterval
              
 
              
 
 
              
 
              
-torch.distributions.constraints.interval[source]
              
+torch.distributions.constraints.interval[source]
 
              alias of _Interval
              
 
              
 
 
              
 
              
-torch.distributions.constraints.half_open_interval[source]
              
+torch.distributions.constraints.half_open_interval[source]
 
              alias of _HalfOpenInterval
              
 
              
 
 
              
 
              
-torch.distributions.constraints.is_dependent(constraint)[source][source]
              
+torch.distributions.constraints.is_dependent(constraint)[source][source]
 
              Checks if constraint is a _Dependent object.
              
 
              
 
              Parameters
              
@@ -4964,19 +4964,19 @@ 
              Wishart
 
              
-torch.distributions.constraints.less_than[source]
              
+torch.distributions.constraints.less_than[source]
 
              alias of _LessThan
              
 
              
 
 
              
 
              
-torch.distributions.constraints.multinomial[source]
              
+torch.distributions.constraints.multinomial[source]
 
              alias of _Multinomial
              
 
              
 
 
              
 
              
-torch.distributions.constraints.stack[source]
              
+torch.distributions.constraints.stack[source]
 
              alias of _Stack
              
 
              
 
@@ -5049,11 +5049,11 @@ 
              Wishart
 
              
-class torch.distributions.constraint_registry.ConstraintRegistry[source][source]
              
+class torch.distributions.constraint_registry.ConstraintRegistry[source][source]
 
              Registry to link constraints to transforms.
              
 
              
 
              
-register(constraint, factory=None)[source][source]
              
+register(constraint, factory=None)[source][source]
 
              Registers a Constraint
 subclass in this registry. Usage:
              
 
              @my_registry.register(MyConstraintClass)
diff --git a/2.7/dlpack.html b/2.7/dlpack.html
index 2e9de4f9069..0cbf41890ec 100644
--- a/2.7/dlpack.html
+++ b/2.7/dlpack.html
@@ -594,7 +594,7 @@
 
              torch.utils.dlpack
              
 
              
 
              
-torch.utils.dlpack.from_dlpack(ext_tensor) → Tensor[source][source]
              
+torch.utils.dlpack.from_dlpack(ext_tensor) → Tensor[source][source]
 
              Converts a tensor from an external library into a torch.Tensor.
              
 
              The returned PyTorch tensor will share the memory with the input tensor
 (which may have come from another library). Note that in-place operations
diff --git a/2.7/elastic/agent.html b/2.7/elastic/agent.html
index a9afd164652..c4809c8e218 100644
--- a/2.7/elastic/agent.html
+++ b/2.7/elastic/agent.html
@@ -622,7 +622,7 @@ 
              Conceptsagent in torchelastic.
              
 
              
 
              
-class torch.distributed.elastic.agent.server.ElasticAgent[source][source]
              
+class torch.distributed.elastic.agent.server.ElasticAgent[source][source]
 
              An agent process responsible for managing one or more worker processes.
              
 
              The worker processes are assumed to be regular distributed PyTorch scripts.
 When the worker process is created by the agent, the agent provides the
@@ -654,7 +654,7 @@ 
              Concepts
 
              
-abstract get_worker_group(role='default')[source][source]
              
+abstract get_worker_group(role='default')[source][source]
 
              Return the WorkerGroup for the given role.
              
 
              Note that the worker group is a mutable object and hence in a
 multi-threaded/process environment it may change state.
@@ -669,7 +669,7 @@ 
              Concepts
 
              
-abstract run(role='default')[source][source]
              
+abstract run(role='default')[source][source]
 
              Run the agent.
              
 
              Supports retrying the worker group on failures up to max_restarts.
              
 
              
@@ -690,7 +690,7 @@ 
              Concepts
 
              
-class torch.distributed.elastic.agent.server.WorkerSpec(role, local_world_size, rdzv_handler, fn=None, entrypoint=None, args=(), max_restarts=3, monitor_interval=0.1, master_port=None, master_addr=None, local_addr=None)[source][source]
              
+class torch.distributed.elastic.agent.server.WorkerSpec(role, local_world_size, rdzv_handler, fn=None, entrypoint=None, args=(), max_restarts=3, monitor_interval=0.1, master_port=None, master_addr=None, local_addr=None)[source][source]
 
              Blueprint information about a particular type of worker.
              
 
              For a given role, there must only exist a single worker spec.
 Worker spec is expected to be homogeneous across all nodes (machine),
@@ -721,7 +721,7 @@ 
              Concepts
 
              
-get_entrypoint_name()[source][source]
              
+get_entrypoint_name()[source][source]
 
              Get the entry point name.
              
 
              If the entrypoint is a function (e.g. Callable) returns its __qualname__
 else if the entrypoint is a binary (e.g. str), returns the binary name.
              
@@ -731,7 +731,7 @@ 
              Concepts
 
              
-class torch.distributed.elastic.agent.server.WorkerState(value)[source][source]
              
+class torch.distributed.elastic.agent.server.WorkerState(value)[source][source]
 
              A state of the WorkerGroup.
              
 
              Workers in a worker group change state as a unit. If a single worker
 in a worker group fails the entire set is considered failed:
              
@@ -763,7 +763,7 @@ 
              Concepts
 
              
-static is_running(state)[source][source]
              
+static is_running(state)[source][source]
 
              Return the state of the Worker.
              
 
              
 
              Returns
              
@@ -780,7 +780,7 @@ 
              Concepts
 
              
-class torch.distributed.elastic.agent.server.Worker(local_rank, global_rank=-1, role_rank=-1, world_size=-1, role_world_size=-1)[source][source]
              
+class torch.distributed.elastic.agent.server.Worker(local_rank, global_rank=-1, role_rank=-1, world_size=-1, role_world_size=-1)[source][source]
 
              A worker instance.
              
 
              Contrast this with WorkerSpec that represents the specifications of a
 worker. A Worker is created from a WorkerSpec. A Worker is to
@@ -805,7 +805,7 @@ 
              Concepts
 
              
-class torch.distributed.elastic.agent.server.WorkerGroup(spec)[source][source]
              
+class torch.distributed.elastic.agent.server.WorkerGroup(spec)[source][source]
 
              A set of Worker instances.
              
 
              The class defines a set of Worker instances for the given WorkerSpec managed by ElasticAgent. Whether the worker
 group contains cross instance workers or not depends on the implementation of the agent.
              
@@ -819,7 +819,7 @@ 
              Implementations
 
              
-class torch.distributed.elastic.agent.server.local_elastic_agent.LocalElasticAgent(spec, logs_specs, start_method='spawn', exit_barrier_timeout=300, log_line_prefix_template=None)[source][source]
              
+class torch.distributed.elastic.agent.server.local_elastic_agent.LocalElasticAgent(spec, logs_specs, start_method='spawn', exit_barrier_timeout=300, log_line_prefix_template=None)[source][source]
 
              An implementation of torchelastic.agent.server.ElasticAgent that handles host-local workers.
              
 
              This agent is deployed per host and is configured to spawn n workers.
 When using GPUs, n maps to the number of GPUs available on the host.
              
@@ -905,7 +905,7 @@ 
              Extending the Agent
 
              
-class torch.distributed.elastic.agent.server.SimpleElasticAgent(spec, exit_barrier_timeout=300)[source][source]
              
+class torch.distributed.elastic.agent.server.SimpleElasticAgent(spec, exit_barrier_timeout=300)[source][source]
 
              An ElasticAgent that manages one particular type of worker role.
              
 
              An ElasticAgent that manages workers (WorkerGroup) for a single WorkerSpec
 such as one particular type of worker role.
              
@@ -913,7 +913,7 @@ 
              Extending the Agent
 
              
-_assign_worker_ranks(store, group_rank, group_world_size, spec)[source][source]
              
+_assign_worker_ranks(store, group_rank, group_world_size, spec)[source][source]
 
              Determine proper ranks for worker processes.
              
 
              Fast Path: when all workers have the same role and world size. We calculate
 the global rank to be group_rank * group_world_size + local_rank. And the
@@ -949,7 +949,7 @@ 
              Extending the Agent
 
              
-_exit_barrier()[source][source]
              
+_exit_barrier()[source][source]
 
              Define a barrier that keeps the agent process alive until all workers finish.
              
 
              Wait for exit_barrier_timeout seconds for all agents to finish
 executing their local workers (either successfully or not). This
@@ -959,7 +959,7 @@ 
              Extending the Agent
 
              
-_initialize_workers(worker_group)[source][source]
              
+_initialize_workers(worker_group)[source][source]
 
              Start a fresh set of workers for the worker_group.
              
 
              Essentially, a rendezvous followed by a start_workers.
 The caller should first call _stop_workers() to stop running workers
@@ -973,7 +973,7 @@ 
              Extending the Agent
 
              
-abstract _monitor_workers(worker_group)[source][source]
              
+abstract _monitor_workers(worker_group)[source][source]
 
              Check on the workers for the worker_group.
              
 
              This function also returns the new state of the worker group.
              
 
              
@@ -985,7 +985,7 @@ 
              Extending the Agent
 
              
-_rendezvous(worker_group)[source][source]
              
+_rendezvous(worker_group)[source][source]
 
              Run rendezvous for the workers specified by the worker spec.
              
 
              Assigns workers a new global rank and world size.
 Updates the rendezvous store for the worker group.
              
@@ -995,7 +995,7 @@ 
              Extending the Agent
 
              
-_restart_workers(worker_group)[source][source]
              
+_restart_workers(worker_group)[source][source]
 
              Restart (stops, rendezvous, starts) all local workers in the group.
              
 
              
 
              
@@ -1003,7 +1003,7 @@ 
              Extending the Agent
 
              
-abstract _shutdown(death_sig=Signals.SIGTERM, is_restart=False)[source][source]
              
+abstract _shutdown(death_sig=Signals.SIGTERM, is_restart=False)[source][source]
 
              Clean up any resources that were allocated during the agent’s work.
              
 
              
 
              Parameters
              
@@ -1014,7 +1014,7 @@ 
              Extending the Agent
 
              
-abstract _start_workers(worker_group)[source][source]
              
+abstract _start_workers(worker_group)[source][source]
 
              Start worker_group.spec.local_world_size number of workers.
              
 
              This is according to worker spec for the worker group .
 Returns a map of local_rank to worker id.
              
@@ -1027,7 +1027,7 @@ 
              Extending the Agent
 
              
-abstract _stop_workers(worker_group, is_restart=False)[source][source]
              
+abstract _stop_workers(worker_group, is_restart=False)[source][source]
 
              Stop all workers in the given worker group.
              
 
              Implementors must deal with workers in all states defined by
 WorkerState. That is, it must gracefully handle stopping
@@ -1040,7 +1040,7 @@ 
              Extending the Agent
 
              
-class torch.distributed.elastic.agent.server.api.RunResult(state, return_values=<factory>, failures=<factory>)[source][source]
              
+class torch.distributed.elastic.agent.server.api.RunResult(state, return_values=<factory>, failures=<factory>)[source][source]
 
              Return results of the worker executions.
              
 
              Run results follow an “all-or-nothing” policy where the run is successful if and
 only if ALL local workers managed by this agent complete successfully.
              
@@ -1089,7 +1089,7 @@ 
              Health Check Server
              
 
              
-class torch.distributed.elastic.agent.server.health_check_server.HealthCheckServer(alive_callback, port, timeout)[source][source]
              
+class torch.distributed.elastic.agent.server.health_check_server.HealthCheckServer(alive_callback, port, timeout)[source][source]
 
              Interface for health check monitoring server, which can be extended
 by starting tcp/http server on the specified port.
              
 
              
@@ -1103,7 +1103,7 @@ 
              Health Check Server
 
              
-start()[source][source]
              
+start()[source][source]
 
              Unsupported functionality for Pytorch, doesn’t start any health check server
              
 
              
 
              
@@ -1111,7 +1111,7 @@ 
              Health Check Server
 
              
-stop()[source][source]
              
+stop()[source][source]
 
              Function to stop health check server
              
 
              
 
              
@@ -1121,7 +1121,7 @@ 
              Health Check Server
 
              
-torch.distributed.elastic.agent.server.health_check_server.create_healthcheck_server(alive_callback, port, timeout)[source][source]
              
+torch.distributed.elastic.agent.server.health_check_server.create_healthcheck_server(alive_callback, port, timeout)[source][source]
 
              creates health check server object
              
 
              
 
              Return type
              
diff --git a/2.7/elastic/control_plane.html b/2.7/elastic/control_plane.html
index 83b826652d0..aafc366875d 100644
--- a/2.7/elastic/control_plane.html
+++ b/2.7/elastic/control_plane.html
@@ -598,7 +598,7 @@
 into your application.
              
 
              
 
              
-torch.distributed.elastic.control_plane.worker_main()[source][source]
              
+torch.distributed.elastic.control_plane.worker_main()[source][source]
 
              This is a context manager that wraps your main entry function. This combines
 the existing errors.record logic as well as a new _WorkerServer that
 exposes handlers via a unix socket specified by
diff --git a/2.7/elastic/errors.html b/2.7/elastic/errors.html
index 2b95cf8a56e..7984d42a684 100644
--- a/2.7/elastic/errors.html
+++ b/2.7/elastic/errors.html
@@ -645,7 +645,7 @@
 
              Methods and Classes
              
 
              
 
              
-torch.distributed.elastic.multiprocessing.errors.record(fn, error_handler=None)[source][source]
              
+torch.distributed.elastic.multiprocessing.errors.record(fn, error_handler=None)[source][source]
 
              Syntactic sugar to record errors/exceptions that happened in the decorated
 function using the provided error_handler.
              
 
              Using this decorator is equivalent to:
              
@@ -686,7 +686,7 @@ 
              Methods and Classes
 
              
-class torch.distributed.elastic.multiprocessing.errors.ChildFailedError(name, failures)[source][source]
              
+class torch.distributed.elastic.multiprocessing.errors.ChildFailedError(name, failures)[source][source]
 
              Special exception type that can be raised from a function annotated with the
 @record decorator to have the child process’ (root exception) propagate
 up the stack as-is (e.g. without being wrapped in the parent’s traceback).
              
@@ -723,7 +723,7 @@ 
              Methods and Classes
 
              
-class torch.distributed.elastic.multiprocessing.errors.ErrorHandler[source][source]
              
+class torch.distributed.elastic.multiprocessing.errors.ErrorHandler[source][source]
 
              Write the provided exception object along with some other metadata about
 the error in a structured way in JSON format to an error file specified by the
 environment variable: TORCHELASTIC_ERROR_FILE. If this environment
@@ -735,7 +735,7 @@ 
              Methods and Classes
 
              
-class torch.distributed.elastic.multiprocessing.errors.ProcessFailure(local_rank, pid, exitcode, error_file)[source][source]
              
+class torch.distributed.elastic.multiprocessing.errors.ProcessFailure(local_rank, pid, exitcode, error_file)[source][source]
 
              Represent the failed process result. When the worker process fails, it may record failure root cause into the file.
              
 
              Tries to read the failure timestamp from the provided error_file,
 if the error_file does not exist, the timestamp is the current
diff --git a/2.7/elastic/events.html b/2.7/elastic/events.html
index 3e328517241..45647c85f99 100644
--- a/2.7/elastic/events.html
+++ b/2.7/elastic/events.html
@@ -608,14 +608,14 @@
 
              API Methods
              
 
              
 
              
-torch.distributed.elastic.events.record(event, destination='null')[source][source]
              
+torch.distributed.elastic.events.record(event, destination='null')[source][source]
 
              
 
              
 
              
 
 
              
 
              
-torch.distributed.elastic.events.construct_and_record_rdzv_event(run_id, message, node_state, name='', hostname='', pid=None, master_endpoint='', local_id=None, rank=None)[source][source]
              
+torch.distributed.elastic.events.construct_and_record_rdzv_event(run_id, message, node_state, name='', hostname='', pid=None, master_endpoint='', local_id=None, rank=None)[source][source]
 
              Initialize rendezvous event object and record its operations.
              
 
              
 
              Parameters
              
@@ -662,7 +662,7 @@ 
              API Methods
 
              
-torch.distributed.elastic.events.get_logging_handler(destination='null')[source][source]
              
+torch.distributed.elastic.events.get_logging_handler(destination='null')[source][source]
 
              
 
              Return type
              
 
              Handler
              
@@ -675,7 +675,7 @@ 
              API Methods
              
 
              
 
              
-class torch.distributed.elastic.events.api.Event(name, source, timestamp=0, metadata=<factory>)[source][source]
              
+class torch.distributed.elastic.events.api.Event(name, source, timestamp=0, metadata=<factory>)[source][source]
 
              The class represents the generic event that occurs during the torchelastic job execution.
              
 
              The event can be any kind of meaningful action.
              
 
              
@@ -692,7 +692,7 @@ 
              Event Objects
 
              
-class torch.distributed.elastic.events.api.EventSource(value)[source][source]
              
+class torch.distributed.elastic.events.api.EventSource(value)[source][source]
 
              Known identifiers of the event producers.
              
 
              
 
diff --git a/2.7/elastic/metrics.html b/2.7/elastic/metrics.html
index 6179e6f5ab6..a4f13e0cd7d 100644
--- a/2.7/elastic/metrics.html
+++ b/2.7/elastic/metrics.html
@@ -709,17 +709,17 @@ 
              Metric Handlers
 
              
-class torch.distributed.elastic.metrics.api.MetricHandler[source][source]
              
+class torch.distributed.elastic.metrics.api.MetricHandler[source][source]
 
              
 
 
              
 
              
-class torch.distributed.elastic.metrics.api.ConsoleMetricHandler[source][source]
              
+class torch.distributed.elastic.metrics.api.ConsoleMetricHandler[source][source]
 
              
 
 
              
 
              
-class torch.distributed.elastic.metrics.api.NullMetricHandler[source][source]
              
+class torch.distributed.elastic.metrics.api.NullMetricHandler[source][source]
 
              
 
 
              
@@ -727,14 +727,14 @@ 
              Metric Handlers
              
 
              
 
              
-torch.distributed.elastic.metrics.configure(handler, group=None)[source][source]
              
+torch.distributed.elastic.metrics.configure(handler, group=None)[source][source]
 
              
 
              
 
              
 
 
              
 
              
-torch.distributed.elastic.metrics.prof(fn=None, group='torchelastic')[source][source]
              
+torch.distributed.elastic.metrics.prof(fn=None, group='torchelastic')[source][source]
 
              @profile decorator publishes duration.ms, count, success, failure metrics for the function that it decorates.
              
 
              The metric name defaults to the qualified name (class_name.def_name) of the function.
 If the function does not belong to a class, it uses the leaf module name instead.
              
@@ -755,7 +755,7 @@ 
              Methods
 
              
-torch.distributed.elastic.metrics.put_metric(metric_name, metric_value, metric_group='torchelastic')[source][source]
              
+torch.distributed.elastic.metrics.put_metric(metric_name, metric_value, metric_group='torchelastic')[source][source]
 
              Publish a metric data point.
              
 
              Usage
              
 
              put_metric("metric_name", 1)
diff --git a/2.7/elastic/multiprocessing.html b/2.7/elastic/multiprocessing.html
index 93616d45824..cc269ea1d14 100644
--- a/2.7/elastic/multiprocessing.html
+++ b/2.7/elastic/multiprocessing.html
@@ -650,7 +650,7 @@
 
              Starting Multiple Workers
              
 
              
 
              
-torch.distributed.elastic.multiprocessing.start_processes(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None, start_method='spawn')[source][source]
              
+torch.distributed.elastic.multiprocessing.start_processes(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None, start_method='spawn')[source][source]
 
              Start n copies of entrypoint processes with the provided options.
              
 
              entrypoint is either a Callable (function) or a str (binary).
 The number of copies is determined by the number of entries for args and
@@ -754,7 +754,7 @@ 
              Starting Multiple Workers
              
 
              
 
              
-class torch.distributed.elastic.multiprocessing.api.PContext(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None)[source][source]
              
+class torch.distributed.elastic.multiprocessing.api.PContext(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None)[source][source]
 
              The base class that standardizes operations over a set of processes that are launched via different mechanisms.
              
 
              The name PContext is intentional to disambiguate with torch.multiprocessing.ProcessContext.
              
 
              
@@ -769,7 +769,7 @@ 
              Process Context
 
              
-class torch.distributed.elastic.multiprocessing.api.MultiprocessContext(name, entrypoint, args, envs, start_method, logs_specs, log_line_prefixes=None)[source][source]
              
+class torch.distributed.elastic.multiprocessing.api.MultiprocessContext(name, entrypoint, args, envs, start_method, logs_specs, log_line_prefixes=None)[source][source]
 
              PContext holding worker processes invoked as a function.
              
 
              
 
              
@@ -777,7 +777,7 @@ 
              Process Context
 
              
-class torch.distributed.elastic.multiprocessing.api.SubprocessContext(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None)[source][source]
              
+class torch.distributed.elastic.multiprocessing.api.SubprocessContext(name, entrypoint, args, envs, logs_specs, log_line_prefixes=None)[source][source]
 
              PContext holding worker processes invoked as a binary.
              
 
              
 
              
@@ -785,7 +785,7 @@ 
              Process Context
 
              
-class torch.distributed.elastic.multiprocessing.api.RunProcsResult(return_values=<factory>, failures=<factory>, stdouts=<factory>, stderrs=<factory>)[source][source]
              
+class torch.distributed.elastic.multiprocessing.api.RunProcsResult(return_values=<factory>, failures=<factory>, stdouts=<factory>, stderrs=<factory>)[source][source]
 
              Results of a completed run of processes started with start_processes(). Returned by PContext.
              
 
              Note the following:
              
 
                
@@ -800,7 +800,7 @@ 
                Process Context
 
                
-class torch.distributed.elastic.multiprocessing.api.DefaultLogsSpecs(log_dir=None, redirects=Std.NONE, tee=Std.NONE, local_ranks_filter=None)[source][source]
                
+class torch.distributed.elastic.multiprocessing.api.DefaultLogsSpecs(log_dir=None, redirects=Std.NONE, tee=Std.NONE, local_ranks_filter=None)[source][source]
 
                Default LogsSpecs implementation:
                
 
                  
 
                • log_dir will be created if it doesn’t exist
                  • 
@@ -810,7 +810,7 @@ 
                  Process Context
 
                  
-reify(envs)[source][source]
                  
+reify(envs)[source][source]
 
                  Uses following scheme to build log destination paths:
                  
 
                    
 
                  • <log_dir>/<rdzv_run_id>/attempt_<attempt>/<rank>/stdout.log
                    • 
@@ -828,7 +828,7 @@ 
                    Process Context
 
                    
-class torch.distributed.elastic.multiprocessing.api.LogsDest(stdouts=<factory>, stderrs=<factory>, tee_stdouts=<factory>, tee_stderrs=<factory>, error_files=<factory>)[source][source]
                    
+class torch.distributed.elastic.multiprocessing.api.LogsDest(stdouts=<factory>, stderrs=<factory>, tee_stdouts=<factory>, tee_stderrs=<factory>, error_files=<factory>)[source][source]
 
                    For each log type, holds mapping of local rank ids to file paths.
                    
 
                    
 
                    
@@ -836,7 +836,7 @@ 
                    Process Context
 
                    
-class torch.distributed.elastic.multiprocessing.api.LogsSpecs(log_dir=None, redirects=Std.NONE, tee=Std.NONE, local_ranks_filter=None)[source][source]
                    
+class torch.distributed.elastic.multiprocessing.api.LogsSpecs(log_dir=None, redirects=Std.NONE, tee=Std.NONE, local_ranks_filter=None)[source][source]
 
                    Defines logs processing and redirection for each worker process.
                    
 
                    
 
                    Parameters
                    
@@ -853,7 +853,7 @@ 
                    Process Context
 
                    
-abstract reify(envs)[source][source]
                    
+abstract reify(envs)[source][source]
 
                    Given the environment variables, builds destination of log files for each of the local ranks.
                    
 
                    Envs parameter contains env variables dict for each of the local ranks, where entries are defined in:
 _start_workers().
                    
diff --git a/2.7/elastic/rendezvous.html b/2.7/elastic/rendezvous.html
index 4321b780a35..0dd7bd2672c 100644
--- a/2.7/elastic/rendezvous.html
+++ b/2.7/elastic/rendezvous.html
@@ -693,7 +693,7 @@
 
                    Registry
                    
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.RendezvousParameters(backend, endpoint, run_id, min_nodes, max_nodes, local_addr=None, **kwargs)[source][source]
                    
+class torch.distributed.elastic.rendezvous.RendezvousParameters(backend, endpoint, run_id, min_nodes, max_nodes, local_addr=None, **kwargs)[source][source]
 
                    Hold the parameters to construct a RendezvousHandler.
                    
 
                    
 
                    Parameters
                    
@@ -710,7 +710,7 @@ 
                    Registry
 
                    
-get(key, default=None)[source][source]
                    
+get(key, default=None)[source][source]
 
                    Return the value for key if key exists, else default.
                    
 
                    
 
                    Return type
                    
@@ -721,7 +721,7 @@ 
                    Registry
 
                    
-get_as_bool(key, default=None)[source][source]
                    
+get_as_bool(key, default=None)[source][source]
 
                    Return the value for key as a bool.
                    
 
                    
 
                    Return type
                    
@@ -732,7 +732,7 @@ 
                    Registry
 
                    
-get_as_int(key, default=None)[source][source]
                    
+get_as_int(key, default=None)[source][source]
 
                    Return the value for key as an int.
                    
 
                    
 
                    Return type
                    
@@ -745,7 +745,7 @@ 
                    Registry
 
                    
-class torch.distributed.elastic.rendezvous.RendezvousHandlerRegistry[source][source]
                    
+class torch.distributed.elastic.rendezvous.RendezvousHandlerRegistry[source][source]
 
                    Represent a registry of RendezvousHandler backends.
                    
 
                    
 
                    
@@ -756,7 +756,7 @@ 
                    Registry
                    
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.RendezvousHandler[source][source]
                    
+class torch.distributed.elastic.rendezvous.RendezvousHandler[source][source]
 
                    Main rendezvous interface.
                    
 
                    
 
                    Note
                    
@@ -766,7 +766,7 @@ 
                    Handler
 
                    
-abstract get_backend()[source][source]
                    
+abstract get_backend()[source][source]
 
                    Return the name of the rendezvous backend.
                    
 
                    
 
                    Return type
                    
@@ -777,7 +777,7 @@ 
                    Handler
 
                    
-abstract get_run_id()[source][source]
                    
+abstract get_run_id()[source][source]
 
                    Return the run id of the rendezvous.
                    
 
                    The run id is a user-defined id that uniquely identifies an instance of
 a distributed application. It typically maps to a job id and is used to
@@ -791,7 +791,7 @@ 
                    Handler
 
                    
-abstract is_closed()[source][source]
                    
+abstract is_closed()[source][source]
 
                    Check whether the rendezvous has been closed.
                    
 
                    A closed rendezvous means all future attempts to re-rendezvous within
 same job will fail.
                    
@@ -809,7 +809,7 @@ 
                    Handler
 
                    
-abstract next_rendezvous()[source][source]
                    
+abstract next_rendezvous()[source][source]
 
                    Main entry-point into the rendezvous barrier.
                    
 
                    Blocks until the rendezvous is complete and the current process is
 included in the formed worker group, or a timeout occurs, or the
@@ -834,7 +834,7 @@ 
                    Handler
 
                    
-abstract num_nodes_waiting()[source][source]
                    
+abstract num_nodes_waiting()[source][source]
 
                    Return the number of nodes who arrived late at the rendezvous
 barrier, hence were not included in the current worker group.
                    
 
                    Callers should periodically call this method to check whether new
@@ -849,13 +849,13 @@ 
                    Handler
 
                    
-abstract set_closed()[source][source]
                    
+abstract set_closed()[source][source]
 
                    Mark the rendezvous as closed.
                    
 
                    
 
 
                    
 
                    
-abstract shutdown()[source][source]
                    
+abstract shutdown()[source][source]
 
                    Close all resources that were open for the rendezvous.
                    
 
                    Example:
                    
 
                    rdzv_handler = ...
@@ -888,7 +888,7 @@ 
                    Handler
                    
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.RendezvousInfo(store, rank, world_size, bootstrap_store_info)[source][source]
                    
+class torch.distributed.elastic.rendezvous.RendezvousInfo(store, rank, world_size, bootstrap_store_info)[source][source]
 
                    Holds the information about the rendezvous.
                    
 
                    
 
                    
@@ -896,13 +896,13 @@ 
                    Dataclasses
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousStoreInfo(master_addr, master_port)[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousStoreInfo(master_addr, master_port)[source][source]
 
                    Store address and port that can be used to bootstrap trainer distributed comms
                    
 
                    
 
                    
 
                    
 
                    
-static build(rank, store)[source][source]
                    
+static build(rank, store)[source][source]
 
                    Factory method, finds unused new port on rank0 host and addr/port info with all ranks.
                    
 
                    If master_addr/master_port is knowns (useful when sharing existing tcp store server) use the constructor.
                    
 
                    
@@ -927,37 +927,37 @@ 
                    Dataclasses
                    
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousError[source][source]
 
                    Represents the base type for rendezvous errors.
                    
 
                    
 
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousClosedError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousClosedError[source][source]
 
                    Raised when a rendezvous is closed.
                    
 
                    
 
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousTimeoutError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousTimeoutError[source][source]
 
                    Raised when a rendezvous did not complete on time.
                    
 
                    
 
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousConnectionError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousConnectionError[source][source]
 
                    Raised when the connection to a rendezvous backend has failed.
                    
 
                    
 
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousStateError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousStateError[source][source]
 
                    Raised when the state of a rendezvous is corrupt.
                    
 
                    
 
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.api.RendezvousGracefulExitError[source][source]
                    
+class torch.distributed.elastic.rendezvous.api.RendezvousGracefulExitError[source][source]
 
                    Raised when node wasn’t not included in rendezvous and gracefully exits.
                    
 
                    Exception is a mechanism to exit the stack, however does not mean a failure.
                    
 
                    
@@ -969,7 +969,7 @@ 
                    Implementations
                    
 
                    
 
                    
-torch.distributed.elastic.rendezvous.dynamic_rendezvous.create_handler(store, backend, params)[source][source]
                    
+torch.distributed.elastic.rendezvous.dynamic_rendezvous.create_handler(store, backend, params)[source][source]
 
                    Create a new DynamicRendezvousHandler from the specified parameters.
                    
 
                    
 
                    Parameters
                    
@@ -1016,13 +1016,13 @@ 
                    Dynamic Rendezvous
 
                    
-class torch.distributed.elastic.rendezvous.dynamic_rendezvous.DynamicRendezvousHandler[source][source]
                    
+class torch.distributed.elastic.rendezvous.dynamic_rendezvous.DynamicRendezvousHandler[source][source]
 
                    Represent a handler that sets up a rendezvous among a set of nodes.
                    
 
                    
 
                    
 
                    
 
                    
-classmethod from_backend(run_id, store, backend, min_nodes, max_nodes, local_addr=None, timeout=None, keep_alive_interval=5, keep_alive_max_attempt=3)[source][source]
                    
+classmethod from_backend(run_id, store, backend, min_nodes, max_nodes, local_addr=None, timeout=None, keep_alive_interval=5, keep_alive_max_attempt=3)[source][source]
 
                    Create a new DynamicRendezvousHandler.
                    
 
                    
 
                    Parameters
                    
@@ -1047,11 +1047,11 @@ 
                    Dynamic Rendezvous
 
                    
-class torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend[source][source]
                    
+class torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend[source][source]
 
                    Represent a backend that holds the rendezvous state.
                    
 
                    
 
                    
-abstract get_state()[source][source]
                    
+abstract get_state()[source][source]
 
                    Get the rendezvous state.
                    
 
                    
 
                    Returns
                    
@@ -1078,7 +1078,7 @@ 
                    Dynamic Rendezvous
 
                    
-abstract set_state(state, token=None)[source][source]
                    
+abstract set_state(state, token=None)[source][source]
 
                    Set the rendezvous state.
                    
 
                    The new rendezvous state is set conditionally:
                    
 
                    
@@ -1123,7 +1123,7 @@ 
                    Dynamic Rendezvous
 
                    
-class torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout(join=None, last_call=None, close=None, heartbeat=None)[source][source]
                    
+class torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout(join=None, last_call=None, close=None, heartbeat=None)[source][source]
 
                    Hold the timeout configuration of a rendezvous.
                    
 
                    
 
                    Parameters
                    
@@ -1169,7 +1169,7 @@ 
                    Dynamic Rendezvous
                    
 
                    
 
                    
-torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.create_backend(params)[source][source]
                    
+torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.create_backend(params)[source][source]
 
                    Create a new C10dRendezvousBackend from the specified parameters.
                    
 
@@ -1219,7 +1219,7 @@ 
                    C10d Backend
 
                    
-class torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend(store, run_id)[source][source]
                    
+class torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend(store, run_id)[source][source]
 
                    Represents a C10d-backed rendezvous backend.
                    
 
                    
 
                    Parameters
                    
@@ -1232,7 +1232,7 @@ 
                    C10d Backend
 
                    
-get_state()[source][source]
                    
+get_state()[source][source]
 
                    See base class.
                    
 
                    
 
                    Return type
                    
@@ -1249,7 +1249,7 @@ 
                    C10d Backend
 
                    
-set_state(state, token=None)[source][source]
                    
+set_state(state, token=None)[source][source]
 
                    See base class.
                    
 
                    
 
                    Return type
                    
@@ -1265,7 +1265,7 @@ 
                    C10d Backend
                    
 
                    
 
                    
-torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.create_backend(params)[source][source]
                    
+torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.create_backend(params)[source][source]
 
                    Create a new EtcdRendezvousBackend from the specified parameters.
                    
 
                    
 
                    
@@ -1305,7 +1305,7 @@ 
                    Etcd Backend
 
                    
-class torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend(client, run_id, key_prefix=None, ttl=None)[source][source]
                    
+class torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend(client, run_id, key_prefix=None, ttl=None)[source][source]
 
                    Represents an etcd-based rendezvous backend.
                    
 
                    
 
                    Parameters
                    
@@ -1319,7 +1319,7 @@ 
                    Etcd Backend
 
                    
-get_state()[source][source]
                    
+get_state()[source][source]
 
                    See base class.
                    
 
                    
 
                    Return type
                    
@@ -1336,7 +1336,7 @@ 
                    Etcd Backend
 
                    
-set_state(state, token=None)[source][source]
                    
+set_state(state, token=None)[source][source]
 
                    See base class.
                    
 
                    
 
                    Return type
                    
@@ -1359,7 +1359,7 @@ 
                    Etcd Rendezvous (Legacy)
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.etcd_rendezvous.EtcdRendezvousHandler(rdzv_impl, local_addr)[source][source]
                    
+class torch.distributed.elastic.rendezvous.etcd_rendezvous.EtcdRendezvousHandler(rdzv_impl, local_addr)[source][source]
 
                    Implements a
 torch.distributed.elastic.rendezvous.RendezvousHandler interface
 backed by
@@ -1440,14 +1440,14 @@ 
                    Etcd Storenext_rendezvous() when etcd is used as the rendezvous backend.
                    
 
                    
 
                    
-class torch.distributed.elastic.rendezvous.etcd_store.EtcdStore(etcd_client, etcd_store_prefix, timeout=None)[source][source]
                    
+class torch.distributed.elastic.rendezvous.etcd_store.EtcdStore(etcd_client, etcd_store_prefix, timeout=None)[source][source]
 
                    Implement a c10 Store interface by piggybacking on the rendezvous etcd instance.
                    
 
                    This is the store object returned by EtcdRendezvous.
                    
 
                    
 
                    
 
                    
 
                    
-add(key, num)[source][source]
                    
+add(key, num)[source][source]
 
                    Atomically increment a value by an integer amount.
                    
 
                    The integer is represented as a string using base 10. If key is not present,
 a default value of 0 will be assumed.
                    
@@ -1463,7 +1463,7 @@ 
                    Etcd Store
 
                    
-check(keys)[source][source]
                    
+check(keys)[source][source]
 
                    Check if all of the keys are immediately present (without waiting).
                    
 
                    
 
                    Return type
                    
@@ -1474,7 +1474,7 @@ 
                    Etcd Store
 
                    
-get(key)[source][source]
                    
+get(key)[source][source]
 
                    Get a value by key, possibly doing a blocking wait.
                    
 
                    If key is not immediately present, will do a blocking wait
 for at most timeout duration or until the key is published.
                    
@@ -1493,14 +1493,14 @@ 
                    Etcd Store
 
                    
-set(key, value)[source][source]
                    
+set(key, value)[source][source]
 
                    Write a key/value pair into EtcdStore.
                    
 
                    Both key and value may be either Python str or bytes.
                    
 
                    
 
 
                    
 
                    
-wait(keys, override_timeout=None)[source][source]
                    
+wait(keys, override_timeout=None)[source][source]
 
                    Wait until all of the keys are published, or until timeout.
                    
 
                    
 
                    Raises
                    
@@ -1526,7 +1526,7 @@ 
                    Etcd Server
 
                    
-class torch.distributed.elastic.rendezvous.etcd_server.EtcdServer(data_dir=None)[source][source]
                    
+class torch.distributed.elastic.rendezvous.etcd_server.EtcdServer(data_dir=None)[source][source]
 
                    
 
                    Note
                    
 
                    tested on etcd server v3.4.3.
                    
diff --git a/2.7/elastic/subprocess_handler.html b/2.7/elastic/subprocess_handler.html
index 0818a69caea..101ef0b6d28 100644
--- a/2.7/elastic/subprocess_handler.html
+++ b/2.7/elastic/subprocess_handler.html
@@ -598,7 +598,7 @@
 
                    Retrieve SubprocessHandler
                    
 
                    
 
                    
-torch.distributed.elastic.multiprocessing.subprocess_handler.handlers.get_subprocess_handler(entrypoint, args, env, stdout, stderr, local_rank_id)[source][source]
                    
+torch.distributed.elastic.multiprocessing.subprocess_handler.handlers.get_subprocess_handler(entrypoint, args, env, stdout, stderr, local_rank_id)[source][source]
 
                    
 
                    Return type
                    
 
                    SubprocessHandler
                    
@@ -611,7 +611,7 @@ 
                    Retrieve SubprocessHandler
                    
 
                    
 
                    
-class torch.distributed.elastic.multiprocessing.subprocess_handler.subprocess_handler.SubprocessHandler(entrypoint, args, env, stdout, stderr, local_rank_id)[source][source]
                    
+class torch.distributed.elastic.multiprocessing.subprocess_handler.subprocess_handler.SubprocessHandler(entrypoint, args, env, stdout, stderr, local_rank_id)[source][source]
 
                    Convenience wrapper around python’s subprocess.Popen. Keeps track of
 meta-objects associated to the process (e.g. stdout and stderr redirect fds).
                    
 
                    
diff --git a/2.7/elastic/timer.html b/2.7/elastic/timer.html
index 6669a3b1281..92f6f5c830d 100644
--- a/2.7/elastic/timer.html
+++ b/2.7/elastic/timer.html
@@ -629,7 +629,7 @@
 
                    Client Methods
                    
 
                    
 
                    
-torch.distributed.elastic.timer.configure(timer_client)[source][source]
                    
+torch.distributed.elastic.timer.configure(timer_client)[source][source]
 
                    Configures a timer client. Must be called before using expires.
                    
 
                    
 
                    
@@ -637,7 +637,7 @@ 
                    Client Methods
 
                    
-torch.distributed.elastic.timer.expires(after, scope=None, client=None)[source][source]
                    
+torch.distributed.elastic.timer.expires(after, scope=None, client=None)[source][source]
 
                    Acquires a countdown timer that expires in after seconds from now,
 unless the code-block that it wraps is finished within the timeframe.
 When the timer expires, this worker is eligible to be reaped. The
@@ -671,7 +671,7 @@ 
                    Server/Client Implementationsmultiprocess.Queue.
                    
 
                    
 
                    
-class torch.distributed.elastic.timer.LocalTimerServer(mp_queue, max_interval=60, daemon=True)[source][source]
                    
+class torch.distributed.elastic.timer.LocalTimerServer(mp_queue, max_interval=60, daemon=True)[source][source]
 
                    Server that works with LocalTimerClient. Clients are expected to be
 subprocesses to the parent process that is running this server. Each host
 in the job is expected to start its own timer server locally and each
@@ -683,7 +683,7 @@ 
                    Server/Client Implementations
 
                    
-class torch.distributed.elastic.timer.LocalTimerClient(mp_queue)[source][source]
                    
+class torch.distributed.elastic.timer.LocalTimerClient(mp_queue)[source][source]
 
                    Client side of LocalTimerServer. This client is meant to be used
 on the same host that the LocalTimerServer is running on and uses
 pid to uniquely identify a worker. This is particularly useful in situations
@@ -695,7 +695,7 @@ 
                    Server/Client Implementations
 
                    
-class torch.distributed.elastic.timer.FileTimerServer(file_path, run_id, max_interval=10, daemon=True, log_event=None)[source][source]
                    
+class torch.distributed.elastic.timer.FileTimerServer(file_path, run_id, max_interval=10, daemon=True, log_event=None)[source][source]
 
                    Server that works with FileTimerClient. Clients are expected to be
 running on the same host as the process that is running this server.
 Each host in the job is expected to start its own timer server locally
@@ -717,7 +717,7 @@ 
                    Server/Client Implementations
 
                    
-class torch.distributed.elastic.timer.FileTimerClient(file_path, signal=Signals.SIGKILL)[source][source]
                    
+class torch.distributed.elastic.timer.FileTimerClient(file_path, signal=Signals.SIGKILL)[source][source]
 
                    Client side of FileTimerServer. This client is meant to be used
 on the same host that the FileTimerServer is running on and uses
 pid to uniquely identify a worker.
@@ -747,7 +747,7 @@ 
                    Writing a custom timer server/client
 
                    
-class torch.distributed.elastic.timer.TimerRequest(worker_id, scope_id, expiration_time)[source][source]
                    
+class torch.distributed.elastic.timer.TimerRequest(worker_id, scope_id, expiration_time)[source][source]
 
                    Data object representing a countdown timer acquisition and release
 that is used between the TimerClient and TimerServer.
 A negative expiration_time should be interpreted as a “release”
@@ -764,7 +764,7 @@ 
                    Writing a custom timer server/client
 
                    
-class torch.distributed.elastic.timer.TimerServer(request_queue, max_interval, daemon=True)[source][source]
                    
+class torch.distributed.elastic.timer.TimerServer(request_queue, max_interval, daemon=True)[source][source]
 
                    Entity that monitors active timers and expires them
 in a timely fashion. This server is responsible for
 reaping workers that have expired timers.
                    
@@ -772,7 +772,7 @@ 
                    Writing a custom timer server/client
 
                    
-abstract clear_timers(worker_ids)[source][source]
                    
+abstract clear_timers(worker_ids)[source][source]
 
                    Clears all timers for the given worker_ids.
                    
 
                    
 
                    
@@ -780,7 +780,7 @@ 
                    Writing a custom timer server/client
 
                    
-abstract get_expired_timers(deadline)[source][source]
                    
+abstract get_expired_timers(deadline)[source][source]
 
                    Returns all expired timers for each worker_id. An expired timer
 is a timer for which the expiration_time is less than or equal to
 the provided deadline.
                    
@@ -793,7 +793,7 @@ 
                    Writing a custom timer server/client
 
                    
-abstract register_timers(timer_requests)[source][source]
                    
+abstract register_timers(timer_requests)[source][source]
 
                    Processes the incoming timer requests and registers them with the server.
 The timer request can either be a acquire-timer or release-timer request.
 Timer requests with a negative expiration_time should be interpreted
@@ -806,12 +806,12 @@ 
                    Writing a custom timer server/client
 
                    
-class torch.distributed.elastic.timer.TimerClient[source][source]
                    
+class torch.distributed.elastic.timer.TimerClient[source][source]
 
                    Client library to acquire and release countdown timers by communicating
 with the TimerServer.
                    
 
                    
 
                    
-abstract acquire(scope_id, expiration_time)[source][source]
                    
+abstract acquire(scope_id, expiration_time)[source][source]
 
                    Acquires a timer for the worker that holds this client object
 given the scope_id and expiration_time. Typically registers
 the timer with the TimerServer.
                    
@@ -821,7 +821,7 @@ 
                    Writing a custom timer server/client
 
                    
-abstract release(scope_id)[source][source]
                    
+abstract release(scope_id)[source][source]
 
                    Releases the timer for the scope_id on the worker this
 client represents. After this method is
 called, the countdown timer on the scope is no longer in effect.
                    
@@ -836,7 +836,7 @@ 
                    Writing a custom timer server/client
                    Debug info logging
                    
 
                    
 
                    
-torch.distributed.elastic.timer.debug_info_logging.log_debug_info_for_expired_timers(run_id, expired_timers)[source][source]
                    
+torch.distributed.elastic.timer.debug_info_logging.log_debug_info_for_expired_timers(run_id, expired_timers)[source][source]
 
                    
 
                    
 
                    
diff --git a/2.7/export.html b/2.7/export.html
index bd5ae28829a..e4e92643896 100644
--- a/2.7/export.html
+++ b/2.7/export.html
@@ -1277,7 +1277,7 @@ 
                    Read More
                    API Reference
                    
 
                    
 
                    
-torch.export.export(mod, args, kwargs=None, *, dynamic_shapes=None, strict=True, preserve_module_call_signature=())[source][source]
                    
+torch.export.export(mod, args, kwargs=None, *, dynamic_shapes=None, strict=True, preserve_module_call_signature=())[source][source]
 
                    export() takes any nn.Module along with example inputs, and produces a traced graph representing
 only the Tensor computation of the function in an Ahead-of-Time (AOT) fashion,
 which can subsequently be executed with different inputs or serialized.  The
@@ -1361,7 +1361,7 @@ 
                    Read More
 
                    
-torch.export.save(ep, f, *, extra_files=None, opset_version=None, pickle_protocol=2)[source][source]
                    
+torch.export.save(ep, f, *, extra_files=None, opset_version=None, pickle_protocol=2)[source][source]
 
                    
 
                    Warning
                    
 
                    Under active development, saved files may not be usable in newer versions
@@ -1408,7 +1408,7 @@ 
                    Read More
 
                    
-torch.export.load(f, *, extra_files=None, expected_opset_version=None)[source][source]
                    
+torch.export.load(f, *, extra_files=None, expected_opset_version=None)[source][source]
 
                    
 
                    Warning
                    
 
                    Under active development, saved files may not be usable in newer versions
@@ -1459,7 +1459,7 @@ 
                    Read More
 
                    
-torch.export.register_dataclass(cls, *, serialized_type_name=None)[source][source]
                    
+torch.export.register_dataclass(cls, *, serialized_type_name=None)[source][source]
 
                    Registers a dataclass as a valid input/output type for torch.export.export().
                    
 
                    
 
                    Parameters
                    
@@ -1500,7 +1500,7 @@ 
                    Read More
 
                    
-torch.export.dynamic_shapes.Dim(name, *, min=None, max=None)[source][source]
                    
+torch.export.dynamic_shapes.Dim(name, *, min=None, max=None)[source][source]
 
                    Dim() constructs a type analogous to a named symbolic integer with a range.
 It can be used to describe multiple possible values of a dynamic tensor dimension.
 Note that different dynamic dimensions of the same tensor, or of different tensors,
@@ -1521,7 +1521,7 @@ 
                    Read More
 
                    
-torch.export.exported_program.default_decompositions()[source][source]
                    
+torch.export.exported_program.default_decompositions()[source][source]
 
                    This is the default decomposition table which contains decomposition of
 all ATEN operators to core aten opset. Use this API together with
 run_decompositions()
                    
@@ -1534,7 +1534,7 @@ 
                    Read More
 
                    
-torch.export.dims(*names, min=None, max=None)[source][source]
                    
+torch.export.dims(*names, min=None, max=None)[source][source]
 
                    Util to create multiple Dim() types.
                    
 
                    
 
                    Returns
                    
@@ -1548,7 +1548,7 @@ 
                    Read More
 
                    
-class torch.export.dynamic_shapes.ShapesCollection[source][source]
                    
+class torch.export.dynamic_shapes.ShapesCollection[source][source]
 
                    Builder for dynamic_shapes.
 Used to assign dynamic shape specifications to tensors that appear in inputs.
                    
 
                    This is useful particularly when args() is a nested input structure, and it’s
@@ -1569,7 +1569,7 @@ 
                    Read More
 
                    
-dynamic_shapes(m, args, kwargs=None)[source][source]
                    
+dynamic_shapes(m, args, kwargs=None)[source][source]
 
                    Generates the dynamic_shapes() pytree structure according to args() and kwargs().
                    
 
                    
 
@@ -1577,7 +1577,7 @@ 
                    Read More
 
                    
-torch.export.dynamic_shapes.refine_dynamic_shapes_from_suggested_fixes(msg, dynamic_shapes)[source][source]
                    
+torch.export.dynamic_shapes.refine_dynamic_shapes_from_suggested_fixes(msg, dynamic_shapes)[source][source]
 
                    When exporting with dynamic_shapes(), export may fail with a ConstraintViolation error if the specification
 doesn’t match the constraints inferred from tracing the model. The error message may provide suggested fixes -
 changes that can be made to dynamic_shapes() to export successfully.
                    
@@ -1616,7 +1616,7 @@ 
                    Read More
 
                    
-class torch.export.ExportedProgram(root, graph, graph_signature, state_dict, range_constraints, module_call_graph, example_inputs=None, constants=None, *, verifiers=None)[source][source]
                    
+class torch.export.ExportedProgram(root, graph, graph_signature, state_dict, range_constraints, module_call_graph, example_inputs=None, constants=None, *, verifiers=None)[source][source]
 
                    Package of a program from export(). It contains
 an torch.fx.Graph that represents Tensor computation, a state_dict containing
 tensor values of all lifted parameters and buffers, and various metadata.
                    
@@ -1631,7 +1631,7 @@ 
                    Read More
 
                    
-module()[source][source]
                    
+module()[source][source]
 
                    Returns a self contained GraphModule with all the parameters/buffers inlined.
                    
 
                    
 
                    Return type
                    
@@ -1642,7 +1642,7 @@ 
                    Read More
 
                    
-buffers()[source][source]
                    
+buffers()[source][source]
 
                    Returns an iterator over original module buffers.
                    
 
                    
 
                    Warning
                    
@@ -1657,7 +1657,7 @@ 
                    Read More
 
                    
-named_buffers()[source][source]
                    
+named_buffers()[source][source]
 
                    Returns an iterator over original module buffers, yielding
 both the name of the buffer as well as the buffer itself.
                    
 
                    
@@ -1673,7 +1673,7 @@ 
                    Read More
 
                    
-parameters()[source][source]
                    
+parameters()[source][source]
 
                    Returns an iterator over original module’s parameters.
                    
 
                    
 
                    Warning
                    
@@ -1688,7 +1688,7 @@ 
                    Read More
 
                    
-named_parameters()[source][source]
                    
+named_parameters()[source][source]
 
                    Returns an iterator over original module parameters, yielding
 both the name of the parameter as well as the parameter itself.
                    
 
                    
@@ -1704,7 +1704,7 @@ 
                    Read More
 
                    
-run_decompositions(decomp_table=None, decompose_custom_triton_ops=False)[source][source]
                    
+run_decompositions(decomp_table=None, decompose_custom_triton_ops=False)[source][source]
 
                    Run a set of decompositions on the exported program and returns a new
 exported program. By default we will run the Core ATen decompositions to
 get operators in the
@@ -1739,14 +1739,14 @@ 
                    Read More
 
                    
-class torch.export.ExportBackwardSignature(gradients_to_parameters: dict[str, str], gradients_to_user_inputs: dict[str, str], loss_output: str)[source][source]
                    
+class torch.export.ExportBackwardSignature(gradients_to_parameters: dict[str, str], gradients_to_user_inputs: dict[str, str], loss_output: str)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.ExportGraphSignature(input_specs, output_specs)[source][source]
                    
+class torch.export.ExportGraphSignature(input_specs, output_specs)[source][source]
 
                    ExportGraphSignature models the input/output signature of Export Graph,
 which is a fx.Graph with stronger invariants gurantees.
                    
 
                    Export Graph is functional and does not access “states” like parameters
@@ -1819,21 +1819,21 @@ 
                    Read More
 
                    
-class torch.export.ModuleCallSignature(inputs: list[Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument]], outputs: list[Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument]], in_spec: torch.utils._pytree.TreeSpec, out_spec: torch.utils._pytree.TreeSpec, forward_arg_names: Optional[list[str]] = None)[source][source]
                    
+class torch.export.ModuleCallSignature(inputs: list[Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument]], outputs: list[Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument]], in_spec: torch.utils._pytree.TreeSpec, out_spec: torch.utils._pytree.TreeSpec, forward_arg_names: Optional[list[str]] = None)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.ModuleCallEntry(fqn: str, signature: Optional[torch.export.exported_program.ModuleCallSignature] = None)[source][source]
                    
+class torch.export.ModuleCallEntry(fqn: str, signature: Optional[torch.export.exported_program.ModuleCallSignature] = None)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.decomp_utils.CustomDecompTable[source][source]
                    
+class torch.export.decomp_utils.CustomDecompTable[source][source]
 
                    This is a custom dictionary that is specifically used for handling decomp_table in export.
 The reason we need this is because in the new world, you can only delete an op from decomp
 table to preserve it. This is problematic for custom ops because we don’t know when the custom
@@ -1851,7 +1851,7 @@ 
                    Read More
 
                    
-copy()[source][source]
                    
+copy()[source][source]
 
                    
 
                    Return type
                    
 
                    CustomDecompTable
                    
@@ -1861,17 +1861,17 @@ 
                    Read More
 
                    
-items()[source][source]
                    
+items()[source][source]
 
                    
 
 
                    
 
                    
-keys()[source][source]
                    
+keys()[source][source]
 
                    
 
 
                    
 
                    
-materialize()[source][source]
                    
+materialize()[source][source]
 
                    
 
                    Return type
                    
 
                    dict[torch._ops.OperatorBase, Callable]
                    
@@ -1881,66 +1881,66 @@ 
                    Read More
 
                    
-pop(*args)[source][source]
                    
+pop(*args)[source][source]
 
                    
 
 
                    
 
                    
-update(other_dict)[source][source]
                    
+update(other_dict)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.InputKind(value)[source][source]
                    
+class torch.export.graph_signature.InputKind(value)[source][source]
 
                    An enumeration.
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.InputSpec(kind: torch.export.graph_signature.InputKind, arg: Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument], target: Optional[str], persistent: Optional[bool] = None)[source][source]
                    
+class torch.export.graph_signature.InputSpec(kind: torch.export.graph_signature.InputKind, arg: Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument], target: Optional[str], persistent: Optional[bool] = None)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.OutputKind(value)[source][source]
                    
+class torch.export.graph_signature.OutputKind(value)[source][source]
 
                    An enumeration.
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.OutputSpec(kind: torch.export.graph_signature.OutputKind, arg: Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument], target: Optional[str])[source][source]
                    
+class torch.export.graph_signature.OutputSpec(kind: torch.export.graph_signature.OutputKind, arg: Union[torch.export.graph_signature.TensorArgument, torch.export.graph_signature.SymIntArgument, torch.export.graph_signature.SymFloatArgument, torch.export.graph_signature.SymBoolArgument, torch.export.graph_signature.ConstantArgument, torch.export.graph_signature.CustomObjArgument, torch.export.graph_signature.TokenArgument], target: Optional[str])[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.SymIntArgument(name: str)[source][source]
                    
+class torch.export.graph_signature.SymIntArgument(name: str)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.SymBoolArgument(name: str)[source][source]
                    
+class torch.export.graph_signature.SymBoolArgument(name: str)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.SymFloatArgument(name: str)[source][source]
                    
+class torch.export.graph_signature.SymFloatArgument(name: str)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.ExportGraphSignature(input_specs, output_specs)[source][source]
                    
+class torch.export.graph_signature.ExportGraphSignature(input_specs, output_specs)[source][source]
 
                    ExportGraphSignature models the input/output signature of Export Graph,
 which is a fx.Graph with stronger invariants gurantees.
                    
 
                    Export Graph is functional and does not access “states” like parameters
@@ -2011,7 +2011,7 @@ 
                    Read More
 
                    
-replace_all_uses(old, new)[source][source]
                    
+replace_all_uses(old, new)[source][source]
 
                    Replace all uses of the old name with new name in the signature.
                    
 
                    
 
                    
@@ -2019,25 +2019,25 @@ 
                    Read More
 
                    
-get_replace_hook(replace_inputs=False)[source][source]
                    
+get_replace_hook(replace_inputs=False)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.export.graph_signature.CustomObjArgument(name: str, class_fqn: str, fake_val: Optional[torch._library.fake_class_registry.FakeScriptObject] = None)[source][source]
                    
+class torch.export.graph_signature.CustomObjArgument(name: str, class_fqn: str, fake_val: Optional[torch._library.fake_class_registry.FakeScriptObject] = None)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-class torch.export.unflatten.FlatArgsAdapter[source][source]
                    
+class torch.export.unflatten.FlatArgsAdapter[source][source]
 
                    Adapts input arguments with input_spec to align target_spec.
                    
 
                    
 
                    
-abstract adapt(target_spec, input_spec, input_args, metadata=None)[source][source]
                    
+abstract adapt(target_spec, input_spec, input_args, metadata=None)[source][source]
 
                    NOTE: This adapter may mutate given input_args_with_path.
                    
 
                    
 
                    Return type
                    
@@ -2050,7 +2050,7 @@ 
                    Read More
 
                    
-class torch.export.unflatten.InterpreterModule(graph, ty=None)[source][source]
                    
+class torch.export.unflatten.InterpreterModule(graph, ty=None)[source][source]
 
                    A module that uses torch.fx.Interpreter to execute instead of the usual
 codegen that GraphModule uses. This provides better stack trace information
 and makes it easier to debug execution.
                    
@@ -2060,7 +2060,7 @@ 
                    Read More
 
                    
-class torch.export.unflatten.InterpreterModuleDispatcher(attrs, call_modules)[source][source]
                    
+class torch.export.unflatten.InterpreterModuleDispatcher(attrs, call_modules)[source][source]
 
                    A module that carries a sequence of InterpreterModules corresponding to
 a sequence of calls of that module. Each call to the module dispatches
 to the next InterpreterModule, and wraps back around after the last.
                    
@@ -2070,7 +2070,7 @@ 
                    Read More
 
                    
-torch.export.unflatten.unflatten(module, flat_args_adapter=None)[source][source]
                    
+torch.export.unflatten.unflatten(module, flat_args_adapter=None)[source][source]
 
                    Unflatten an ExportedProgram, producing a module with the same module
 hierarchy as the original eager module. This can be useful if you are trying
 to use torch.export with another system that expects a module
@@ -2102,7 +2102,7 @@ 
                    Read More
                    
 
                    
-torch.export.passes.move_to_device_pass(ep, location)[source][source]
                    
+torch.export.passes.move_to_device_pass(ep, location)[source][source]
 
                    Move the exported program to the given device.
                    
 
                    
 
                    Parameters
                    
diff --git a/2.7/fsdp.html b/2.7/fsdp.html
index c0887e51aef..0e5cd883e4e 100644
--- a/2.7/fsdp.html
+++ b/2.7/fsdp.html
@@ -594,7 +594,7 @@
 
                    FullyShardedDataParallel
                    
 
                    
 
                    
-class torch.distributed.fsdp.FullyShardedDataParallel(module, process_group=None, sharding_strategy=None, cpu_offload=None, auto_wrap_policy=None, backward_prefetch=BackwardPrefetch.BACKWARD_PRE, mixed_precision=None, ignored_modules=None, param_init_fn=None, device_id=None, sync_module_states=False, forward_prefetch=False, limit_all_gathers=True, use_orig_params=False, ignored_states=None, device_mesh=None)[source][source]
                    
+class torch.distributed.fsdp.FullyShardedDataParallel(module, process_group=None, sharding_strategy=None, cpu_offload=None, auto_wrap_policy=None, backward_prefetch=BackwardPrefetch.BACKWARD_PRE, mixed_precision=None, ignored_modules=None, param_init_fn=None, device_id=None, sync_module_states=False, forward_prefetch=False, limit_all_gathers=True, use_orig_params=False, ignored_states=None, device_mesh=None)[source][source]
 
                    A wrapper for sharding module parameters across data parallel workers.
                    
 
                    This is inspired by Xu et al. as well as the ZeRO Stage 3 from DeepSpeed.
 FullyShardedDataParallel is commonly shortened to FSDP.
                    
@@ -852,7 +852,7 @@
 
                    
 
                    
 
                    
-apply(fn)[source][source]
                    
+apply(fn)[source][source]
 
                    Apply fn recursively to every submodule (as returned by .children()) as well as self.
                    
 
                    Typical use includes initializing the parameters of a model (see also torch.nn.init).
                    
 
                    Compared to torch.nn.Module.apply, this version additionally gathers
@@ -873,7 +873,7 @@
 
 
                    
 
                    
-check_is_root()[source][source]
                    
+check_is_root()[source][source]
 
                    Check if this instance is a root FSDP module.
                    
 
                    
 
                    Return type
                    
@@ -884,7 +884,7 @@
 
 
                    
 
                    
-clip_grad_norm_(max_norm, norm_type=2.0)[source][source]
                    
+clip_grad_norm_(max_norm, norm_type=2.0)[source][source]
 
                    Clip the gradient norm of all parameters.
                    
 
                    The norm is computed over all parameters’ gradients as viewed as a single vector, and the
 gradients are modified in-place.
                    
@@ -925,7 +925,7 @@
 
 
                    
 
                    
-static flatten_sharded_optim_state_dict(sharded_optim_state_dict, model, optim)[source][source]
                    
+static flatten_sharded_optim_state_dict(sharded_optim_state_dict, model, optim)[source][source]
 
                    Flatten a sharded optimizer state-dict.
                    
 
                    The API is similar to shard_full_optim_state_dict(). The only
 difference is that the input sharded_optim_state_dict should be
@@ -953,7 +953,7 @@
 
 
                    
 
                    
-forward(*args, **kwargs)[source][source]
                    
+forward(*args, **kwargs)[source][source]
 
                    Run the forward pass for the wrapped module, inserting FSDP-specific pre- and post-forward sharding logic.
                    
 
                    
 
                    Return type
                    
@@ -964,7 +964,7 @@
 
 
                    
 
                    
-static fsdp_modules(module, root_only=False)[source][source]
                    
+static fsdp_modules(module, root_only=False)[source][source]
 
                    Return all nested FSDP instances.
                    
 
                    This possibly includes module itself and only includes FSDP root modules if root_only=True.
                    
 
                    
@@ -988,7 +988,7 @@
 
 
                    
 
                    
-static full_optim_state_dict(model, optim, optim_input=None, rank0_only=True, group=None)[source][source]
                    
+static full_optim_state_dict(model, optim, optim_input=None, rank0_only=True, group=None)[source][source]
 
                    Return the full optimizer state-dict.
                    
 
                    Consolidates the full optimizer state on rank 0 and returns it
 as a dict following the convention of
@@ -1041,7 +1041,7 @@
 
 
                    
 
                    
-static get_state_dict_type(module)[source][source]
                    
+static get_state_dict_type(module)[source][source]
 
                    Get the state_dict_type and the corresponding configurations for the FSDP modules rooted at module.
                    
 
                    The target module does not have to be an FSDP module.
                    
 
                    
@@ -1069,7 +1069,7 @@
 
 
                    
 
                    
-named_buffers(*args, **kwargs)[source][source]
                    
+named_buffers(*args, **kwargs)[source][source]
 
                    Return an iterator over module buffers, yielding both the name of the buffer and the buffer itself.
                    
 
                    Intercepts buffer names and removes all occurrences of the FSDP-specific flattened buffer prefix
 when inside the summon_full_params() context manager.
                    
@@ -1082,7 +1082,7 @@
 
 
                    
 
                    
-named_parameters(*args, **kwargs)[source][source]
                    
+named_parameters(*args, **kwargs)[source][source]
 
                    Return an iterator over module parameters, yielding both the name of the parameter and the parameter itself.
                    
 
                    Intercepts parameter names and removes all occurrences of the FSDP-specific flattened parameter prefix
 when inside the summon_full_params() context manager.
                    
@@ -1095,7 +1095,7 @@
 
 
                    
 
                    
-no_sync()[source][source]
                    
+no_sync()[source][source]
 
                    Disable gradient synchronizations across FSDP instances.
                    
 
                    Within this context, gradients will be accumulated in module
 variables, which will later be synchronized in the first
@@ -1123,7 +1123,7 @@
 
 
                    
 
                    
-static optim_state_dict(model, optim, optim_state_dict=None, group=None)[source][source]
                    
+static optim_state_dict(model, optim, optim_state_dict=None, group=None)[source][source]
 
                    Transform the state-dict of an optimizer corresponding to a sharded model.
                    
 
                    The given state-dict can be transformed to one of three types:
 1) full optimizer state_dict, 2) sharded optimizer state_dict, 3) local optimizer state_dict.
                    
@@ -1197,7 +1197,7 @@
 
 
                    
 
                    
-static optim_state_dict_to_load(model, optim, optim_state_dict, is_named_optimizer=False, load_directly=False, group=None)[source][source]
                    
+static optim_state_dict_to_load(model, optim, optim_state_dict, is_named_optimizer=False, load_directly=False, group=None)[source][source]
 
                    Convert an optimizer state-dict so that it can be loaded into the optimizer associated with the FSDP model.
                    
 
                    Given a optim_state_dict that is transformed through
 optim_state_dict(), it gets converted to the flattened optimizer
@@ -1268,7 +1268,7 @@
 
 
                    
 
                    
-register_comm_hook(state, hook)[source][source]
                    
+register_comm_hook(state, hook)[source][source]
 
                    Register a communication hook.
                    
 
                    This is an enhancement that provides a flexible hook to users where they can specify how FSDP aggregates
 gradients across multiple workers.
@@ -1314,7 +1314,7 @@
 
 
                    
 
                    
-static rekey_optim_state_dict(optim_state_dict, optim_state_key_type, model, optim_input=None, optim=None)[source][source]
                    
+static rekey_optim_state_dict(optim_state_dict, optim_state_key_type, model, optim_input=None, optim=None)[source][source]
 
                    Re-keys the optimizer state dict optim_state_dict to use the key type optim_state_key_type.
                    
 
                    This can be used to achieve compatibility between optimizer state dicts from models with FSDP
 instances and ones without.
                    
@@ -1351,7 +1351,7 @@
 
 
                    
 
                    
-static scatter_full_optim_state_dict(full_optim_state_dict, model, optim_input=None, optim=None, group=None)[source][source]
                    
+static scatter_full_optim_state_dict(full_optim_state_dict, model, optim_input=None, optim=None, group=None)[source][source]
 
                    Scatter the full optimizer state dict from rank 0 to all other ranks.
                    
 
                    Returns the sharded optimizer state dict on each rank.
 The return value is the same as shard_full_optim_state_dict(), and on rank
@@ -1416,7 +1416,7 @@
 
 
                    
 
                    
-static set_state_dict_type(module, state_dict_type, state_dict_config=None, optim_state_dict_config=None)[source][source]
                    
+static set_state_dict_type(module, state_dict_type, state_dict_config=None, optim_state_dict_config=None)[source][source]
 
                    Set the state_dict_type of all the descendant FSDP modules of the target module.
                    
 
                    Also takes (optional) configuration for the model’s and optimizer’s state dict.
 The target module does not have to be a FSDP module. If the target
@@ -1470,7 +1470,7 @@
 
 
                    
 
                    
-static shard_full_optim_state_dict(full_optim_state_dict, model, optim_input=None, optim=None)[source][source]
                    
+static shard_full_optim_state_dict(full_optim_state_dict, model, optim_input=None, optim=None)[source][source]
 
                    Shard a full optimizer state-dict.
                    
 
                    Remaps the state in full_optim_state_dict to flattened parameters instead of unflattened
 parameters and restricts to only this rank’s part of the optimizer state.
@@ -1533,7 +1533,7 @@
 
 
                    
 
                    
-static sharded_optim_state_dict(model, optim, group=None)[source][source]
                    
+static sharded_optim_state_dict(model, optim, group=None)[source][source]
 
                    Return the optimizer state-dict in its sharded form.
                    
 
                    The API is similar to full_optim_state_dict() but this API chunks
 all non-zero-dimension states to ShardedTensor to save memory.
@@ -1554,7 +1554,7 @@
 
 
                    
 
                    
-static state_dict_type(module, state_dict_type, state_dict_config=None, optim_state_dict_config=None)[source][source]
                    
+static state_dict_type(module, state_dict_type, state_dict_config=None, optim_state_dict_config=None)[source][source]
 
                    Set the state_dict_type of all the descendant FSDP modules of the target module.
                    
 
                    This context manager has the same functions as set_state_dict_type(). Read the document of
 set_state_dict_type() for the detail.
                    
@@ -1586,7 +1586,7 @@
 
 
                    
 
                    
-static summon_full_params(module, recurse=True, writeback=True, rank0_only=False, offload_to_cpu=False, with_grads=False)[source][source]
                    
+static summon_full_params(module, recurse=True, writeback=True, rank0_only=False, offload_to_cpu=False, with_grads=False)[source][source]
 
                    Expose full params for FSDP instances with this context manager.
                    
 
                    Can be useful after forward/backward for a model to get
 the params for additional processing or checking. It can take a non-FSDP
@@ -1676,7 +1676,7 @@
 
 
                    
 
                    
-class torch.distributed.fsdp.BackwardPrefetch(value)[source][source]
                    
+class torch.distributed.fsdp.BackwardPrefetch(value)[source][source]
 
                    This configures explicit backward prefetching, which improves throughput by
 enabling communication and computation overlap in the backward pass at the
 cost of slightly increased memory usage.
                    
@@ -1708,7 +1708,7 @@
 
 
                    
 
                    
-class torch.distributed.fsdp.ShardingStrategy(value)[source][source]
                    
+class torch.distributed.fsdp.ShardingStrategy(value)[source][source]
 
                    This specifies the sharding strategy to be used for distributed training by
 FullyShardedDataParallel.
                    
 
                      
@@ -1743,7 +1743,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.MixedPrecision(param_dtype=None, reduce_dtype=None, buffer_dtype=None, keep_low_precision_grads=False, cast_forward_inputs=False, cast_root_forward_inputs=True, _module_classes_to_ignore=(<class 'torch.nn.modules.batchnorm._BatchNorm'>, ))[source][source]
                      
+class torch.distributed.fsdp.MixedPrecision(param_dtype=None, reduce_dtype=None, buffer_dtype=None, keep_low_precision_grads=False, cast_forward_inputs=False, cast_root_forward_inputs=True, _module_classes_to_ignore=(<class 'torch.nn.modules.batchnorm._BatchNorm'>, ))[source][source]
 
                      This configures FSDP-native mixed precision training.
                      
 
                      
 
                      Variables
                      
@@ -1867,7 +1867,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.CPUOffload(offload_params=False)[source][source]
                      
+class torch.distributed.fsdp.CPUOffload(offload_params=False)[source][source]
 
                      This configures CPU offloading.
                      
 
                      
 
                      Variables
                      
@@ -1881,7 +1881,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.StateDictConfig(offload_to_cpu=False)[source][source]
                      
+class torch.distributed.fsdp.StateDictConfig(offload_to_cpu=False)[source][source]
 
                      StateDictConfig is the base class for all state_dict configuration
 classes. Users should instantiate a child class (e.g.
 FullStateDictConfig) in order to configure settings for the
@@ -1897,7 +1897,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.FullStateDictConfig(offload_to_cpu=False, rank0_only=False)[source][source]
                      
+class torch.distributed.fsdp.FullStateDictConfig(offload_to_cpu=False, rank0_only=False)[source][source]
 
                      FullStateDictConfig is a config class meant to be used with
 StateDictType.FULL_STATE_DICT. We recommend enabling both
 offload_to_cpu=True and rank0_only=True when saving full state
@@ -1938,7 +1938,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.ShardedStateDictConfig(offload_to_cpu=False, _use_dtensor=False)[source][source]
                      
+class torch.distributed.fsdp.ShardedStateDictConfig(offload_to_cpu=False, _use_dtensor=False)[source][source]
 
                      ShardedStateDictConfig is a config class meant to be used with
 StateDictType.SHARDED_STATE_DICT.
                      
 
                      
@@ -1958,14 +1958,14 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.LocalStateDictConfig(offload_to_cpu: bool = False)[source][source]
                      
+class torch.distributed.fsdp.LocalStateDictConfig(offload_to_cpu: bool = False)[source][source]
 
                      
 
                      
 
                      
 
 
                      
 
                      
-class torch.distributed.fsdp.OptimStateDictConfig(offload_to_cpu=True)[source][source]
                      
+class torch.distributed.fsdp.OptimStateDictConfig(offload_to_cpu=True)[source][source]
 
                      OptimStateDictConfig is the base class for all optim_state_dict
 configuration classes.  Users should instantiate a child class (e.g.
 FullOptimStateDictConfig) in order to configure settings for the
@@ -1982,7 +1982,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.FullOptimStateDictConfig(offload_to_cpu=True, rank0_only=False)[source][source]
                      
+class torch.distributed.fsdp.FullOptimStateDictConfig(offload_to_cpu=True, rank0_only=False)[source][source]
 
                      
 
                      Variables
                      
 
                      rank0_only (bool) – If True, then only rank 0 saves the full state
@@ -1994,7 +1994,7 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.ShardedOptimStateDictConfig(offload_to_cpu=True, _use_dtensor=False)[source][source]
                      
+class torch.distributed.fsdp.ShardedOptimStateDictConfig(offload_to_cpu=True, _use_dtensor=False)[source][source]
 
                      ShardedOptimStateDictConfig is a config class meant to be used with
 StateDictType.SHARDED_STATE_DICT.
                      
 
                      
@@ -2014,14 +2014,14 @@
 
 
                      
 
                      
-class torch.distributed.fsdp.LocalOptimStateDictConfig(offload_to_cpu: bool = False)[source][source]
                      
+class torch.distributed.fsdp.LocalOptimStateDictConfig(offload_to_cpu: bool = False)[source][source]
 
                      
 
                      
 
                      
 
 
                      
 
                      
-class torch.distributed.fsdp.StateDictSettings(state_dict_type: torch.distributed.fsdp.api.StateDictType, state_dict_config: torch.distributed.fsdp.api.StateDictConfig, optim_state_dict_config: torch.distributed.fsdp.api.OptimStateDictConfig)[source][source]
                      
+class torch.distributed.fsdp.StateDictSettings(state_dict_type: torch.distributed.fsdp.api.StateDictType, state_dict_config: torch.distributed.fsdp.api.StateDictConfig, optim_state_dict_config: torch.distributed.fsdp.api.OptimStateDictConfig)[source][source]
 
                      
 
                      
 
                      
diff --git a/2.7/future_mod.html b/2.7/future_mod.html
index baf8068f559..4019aeb14f9 100644
--- a/2.7/future_mod.html
+++ b/2.7/future_mod.html
@@ -594,7 +594,7 @@
 
                      torch.__future__
                      
 
                      
 
                      
-torch.__future__.set_overwrite_module_params_on_conversion(value)[source][source]
                      
+torch.__future__.set_overwrite_module_params_on_conversion(value)[source][source]
 
                      Sets whether to assign new tensors to the parameters instead of changing the
 existing parameters in-place when converting an nn.Module.
                      
 
                      When enabled, the following methods will assign new parameters to the module:
                      
@@ -613,7 +613,7 @@
 
 
                      
 
                      
-torch.__future__.get_overwrite_module_params_on_conversion()[source][source]
                      
+torch.__future__.get_overwrite_module_params_on_conversion()[source][source]
 
                      Returns whether to assign new tensors to the parameters instead of changing the
 existing parameters in-place when converting an torch.nn.Module. Defaults to False.
                      
 
                      See set_overwrite_module_params_on_conversion() for more information.
                      
@@ -626,7 +626,7 @@
 
 
                      
 
                      
-torch.__future__.set_swap_module_params_on_conversion(value)[source][source]
                      
+torch.__future__.set_swap_module_params_on_conversion(value)[source][source]
 
                      Sets whether to use swap_tensors() instead of setting .data to
 change the existing parameters in-place when converting an nn.Module and instead
 of param.copy_(state_dict[key]) when loading a state dict into an nn.Module.
                      
@@ -659,7 +659,7 @@
 
 
                      
 
                      
-torch.__future__.get_swap_module_params_on_conversion()[source][source]
                      
+torch.__future__.get_swap_module_params_on_conversion()[source][source]
 
                      Returns whether to use swap_tensors() instead of setting .data to
 change the existing parameters in-place when converting an nn.Module. Defaults to False.
                      
 
                      See set_swap_module_params_on_conversion() for more information.
                      
diff --git a/2.7/futures.html b/2.7/futures.html
index 91873de5238..89daea62671 100644
--- a/2.7/futures.html
+++ b/2.7/futures.html
@@ -611,7 +611,7 @@
 
                      
 
                      
 
                      
-add_done_callback(callback)[source][source]
                      
+add_done_callback(callback)[source][source]
 
                      Append the given callback function to this Future, which will be run
 when the Future is completed.  Multiple callbacks can be added to
 the same Future, but the order in which they will be executed cannot
@@ -659,7 +659,7 @@
 
 
                      
 
                      
-done()[source][source]
                      
+done()[source][source]
 
                      Return True if this Future is done. A Future is done if it
 has a result or an exception.
                      
 
                      If the value contains tensors that reside on GPUs, Future.done()
@@ -676,7 +676,7 @@
 
 
                      
 
                      
-set_exception(result)[source][source]
                      
+set_exception(result)[source][source]
 
                      Set an exception for this Future, which will mark this Future as
 completed with an error and trigger all attached callbacks. Note that
 when calling wait()/value() on this Future, the exception set here
@@ -701,7 +701,7 @@
 
 
                      
 
                      
-set_result(result)[source][source]
                      
+set_result(result)[source][source]
 
                      Set the result for this Future, which will mark this Future as
 completed and trigger all attached callbacks. Note that a Future
 cannot be marked completed twice.
                      
@@ -743,7 +743,7 @@
 
 
                      
 
                      
-then(callback)[source][source]
                      
+then(callback)[source][source]
 
                      Append the given callback function to this Future, which will be run
 when the Future is completed.  Multiple callbacks can be added to
 the same Future, but the order in which they will be executed cannot
@@ -814,7 +814,7 @@
 
 
                      
 
                      
-value()[source][source]
                      
+value()[source][source]
 
                      Obtain the value of an already-completed future.
                      
 
                      This method should only be called after a call to wait() has
 completed, or inside a callback function passed to then(). In
@@ -838,7 +838,7 @@
 
 
                      
 
                      
-wait()[source][source]
                      
+wait()[source][source]
 
                      Block until the value of this Future is ready.
                      
 
                      If the value contains tensors that reside on GPUs, then an additional
 synchronization is performed with the kernels (executing on the device)
@@ -865,7 +865,7 @@
 
 
                      
 
                      
-torch.futures.collect_all(futures)[source][source]
                      
+torch.futures.collect_all(futures)[source][source]
 
                      Collects the provided Future objects into a single
 combined Future that is completed when all of the
 sub-futures are completed.
                      
@@ -900,7 +900,7 @@
 
 
                      
 
                      
-torch.futures.wait_all(futures)[source][source]
                      
+torch.futures.wait_all(futures)[source][source]
 
                      Waits for all provided futures to be complete, and returns
 the list of completed values. If any of the futures encounters an error,
 the method will exit early and report the error not waiting for other
diff --git a/2.7/fx.html b/2.7/fx.html
index 016fd7005c5..c3615a70c71 100644
--- a/2.7/fx.html
+++ b/2.7/fx.html
@@ -1721,7 +1721,7 @@ 
                      Miscellanea
                      
 
                      
 
                      
-torch.fx.symbolic_trace(root, concrete_args=None)[source][source]
                      
+torch.fx.symbolic_trace(root, concrete_args=None)[source][source]
 
                      Symbolic tracing API
                      
 
                      Given an nn.Module or function instance root, this function will return a GraphModule
 constructed by recording operations seen while tracing through root.
                      
@@ -1780,7 +1780,7 @@ 
                      API Reference
 
                      
-torch.fx.wrap(fn_or_name)[source][source]
                      
+torch.fx.wrap(fn_or_name)[source][source]
 
                      This function can be called at module-level scope to register fn_or_name as a “leaf function”.
 A “leaf function” will be preserved as a CallFunction node in the FX trace instead of being
 traced through:
                      
@@ -1822,7 +1822,7 @@ 
                      API Reference
 
                      
-class torch.fx.GraphModule(*args, **kwargs)[source][source]
                      
+class torch.fx.GraphModule(*args, **kwargs)[source][source]
 
                      GraphModule is an nn.Module generated from an fx.Graph. Graphmodule has a
 graph attribute, as well as code and forward attributes generated
 from that graph.
                      
@@ -1839,7 +1839,7 @@ 
                      API Reference
 
                      
-__init__(root, graph, class_name='GraphModule')[source][source]
                      
+__init__(root, graph, class_name='GraphModule')[source][source]
 
                      Construct a GraphModule.
                      
 
                      
 
                      Parameters
                      
@@ -1866,7 +1866,7 @@ 
                      API Reference
 
                      
-add_submodule(target, m)[source][source]
                      
+add_submodule(target, m)[source][source]
 
                      Adds the given submodule to self.
                      
 
                      This installs empty Modules where none exist yet if they are
 subpaths of target.
                      
@@ -1909,7 +1909,7 @@ 
                      API Reference
 
                      
-delete_all_unused_submodules()[source][source]
                      
+delete_all_unused_submodules()[source][source]
 
                      Deletes all unused submodules from self.
                      
 
                      A Module is considered “used” if any one of the following is
 true:
@@ -1929,7 +1929,7 @@ 
                      API Reference
 
                      
-delete_submodule(target)[source][source]
                      
+delete_submodule(target)[source][source]
 
                      Deletes the given submodule from self.
                      
 
                      The module will not be deleted if target is not a valid
 target.
                      
@@ -1966,7 +1966,7 @@ 
                      API Reference
 
                      
-print_readable(print_output=True, include_stride=False, include_device=False, colored=False)[source][source]
                      
+print_readable(print_output=True, include_stride=False, include_device=False, colored=False)[source][source]
 
                      Return the Python code generated for current GraphModule and its children GraphModules
                      
 
                      
 
                      Warning
                      
@@ -1976,7 +1976,7 @@ 
                      API Reference
 
                      
-recompile()[source][source]
                      
+recompile()[source][source]
 
                      Recompile this GraphModule from its graph attribute. This should be
 called after editing the contained graph, otherwise the generated
 code of this GraphModule will be out of date.
                      
@@ -1993,7 +1993,7 @@ 
                      API Reference
 
                      
-to_folder(folder, module_name='FxModule')[source][source]
                      
+to_folder(folder, module_name='FxModule')[source][source]
 
                      
 
                      Dumps out module to folder with module_name so that it can be
                      imported with from <folder> import <module_name>
                      
 
                      Args:
                      
@@ -2018,7 +2018,7 @@ 
                      API Reference
 
                      
-class torch.fx.Graph(owning_module=None, tracer_cls=None, tracer_extras=None)[source][source]
                      
+class torch.fx.Graph(owning_module=None, tracer_cls=None, tracer_extras=None)[source][source]
 
                      Graph is the main data structure used in the FX Intermediate Representation.
 It consists of a series of Node s, each representing callsites (or other
 syntactic constructs). The list of Node s, taken together, constitute a
@@ -2067,7 +2067,7 @@ 
                      API Reference
 
                      
-__init__(owning_module=None, tracer_cls=None, tracer_extras=None)[source][source]
                      
+__init__(owning_module=None, tracer_cls=None, tracer_extras=None)[source][source]
 
                      Construct an empty Graph.
                      
 
                      
 
                      Note
                      
@@ -2079,7 +2079,7 @@ 
                      API Reference
 
                      
-call_function(the_function, args=None, kwargs=None, type_expr=None)[source][source]
                      
+call_function(the_function, args=None, kwargs=None, type_expr=None)[source][source]
 
                      Insert a call_function Node into the Graph. A call_function node
 represents a call to a Python callable, specified by the_function.
                      
 
                      
@@ -2116,7 +2116,7 @@ 
                      API Reference
 
                      
-call_method(method_name, args=None, kwargs=None, type_expr=None)[source][source]
                      
+call_method(method_name, args=None, kwargs=None, type_expr=None)[source][source]
 
                      Insert a call_method Node into the Graph. A call_method node
 represents a call to a given method on the 0th element of args.
                      
 
                      
@@ -2153,7 +2153,7 @@ 
                      API Reference
 
                      
-call_module(module_name, args=None, kwargs=None, type_expr=None)[source][source]
                      
+call_module(module_name, args=None, kwargs=None, type_expr=None)[source][source]
 
                      Insert a call_module Node into the Graph. A call_module node
 represents a call to the forward() function of a Module in the Module
 hierarchy.
                      
@@ -2193,7 +2193,7 @@ 
                      API Reference
 
                      
-create_node(op, target, args=None, kwargs=None, name=None, type_expr=None)[source][source]
                      
+create_node(op, target, args=None, kwargs=None, name=None, type_expr=None)[source][source]
 
                      Create a Node and add it to the Graph at the current insert-point.
 Note that the current insert-point can be set via Graph.inserting_before()
 and Graph.inserting_after().
                      
@@ -2227,7 +2227,7 @@ 
                      API Reference
 
                      
-eliminate_dead_code(is_impure_node=None)[source][source]
                      
+eliminate_dead_code(is_impure_node=None)[source][source]
 
                      Remove all dead code from the graph, based on each node’s number of
 users, and whether the nodes have any side effects. The graph must be
 topologically sorted before calling.
                      
@@ -2278,7 +2278,7 @@ 
                      API Reference
 
                      
-erase_node(to_erase)[source][source]
                      
+erase_node(to_erase)[source][source]
 
                      Erases a Node from the Graph. Throws an exception if
 there are still users of that node in the Graph.
                      
 
                      
@@ -2294,7 +2294,7 @@ 
                      API Reference
 
                      
-find_nodes(*, op, target=None, sort=True)[source][source]
                      
+find_nodes(*, op, target=None, sort=True)[source][source]
 
                      Allows for fast query of nodes
                      
 
                      
 
                      Parameters
                      
@@ -2318,7 +2318,7 @@ 
                      API Reference
 
                      
-get_attr(qualified_name, type_expr=None)[source][source]
                      
+get_attr(qualified_name, type_expr=None)[source][source]
 
                      Insert a get_attr node into the Graph. A get_attr Node represents the
 fetch of an attribute from the Module hierarchy.
                      
 
                      
@@ -2352,7 +2352,7 @@ 
                      API Reference
 
                      
-graph_copy(g, val_map, return_output_node=False)[source][source]
                      
+graph_copy(g, val_map, return_output_node=False)[source][source]
 
                      Copy all nodes from a given graph into self.
                      
 
                      
 
                      Parameters
                      
@@ -2379,7 +2379,7 @@ 
                      API Reference
 
                      
-inserting_after(n=None)[source][source]
                      
+inserting_after(n=None)[source][source]
 
                      
 
                      Set the point at which create_node and companion methods will insert into the graph.
                      When used within a ‘with’ statement, this will temporary set the insert point and
 then restore it when the with statement exits:
                      
@@ -2412,7 +2412,7 @@ 
                      API Reference
 
                      
-inserting_before(n=None)[source][source]
                      
+inserting_before(n=None)[source][source]
 
                      
 
                      Set the point at which create_node and companion methods will insert into the graph.
                      When used within a ‘with’ statement, this will temporary set the insert point and
 then restore it when the with statement exits:
                      
@@ -2445,7 +2445,7 @@ 
                      API Reference
 
                      
-lint()[source][source]
                      
+lint()[source][source]
 
                      Runs various checks on this Graph to make sure it is well-formed. In
 particular:
 - Checks Nodes have correct ownership (owned by this graph)
@@ -2460,7 +2460,7 @@ 
                      API Reference
 
                      
-node_copy(node, arg_transform=<function Graph.<lambda>>)[source][source]
                      
+node_copy(node, arg_transform=<function Graph.<lambda>>)[source][source]
 
                      Copy a node from one graph into another. arg_transform needs to transform arguments from
 the graph of node to the graph of self. Example:
                      
 
                      # Copying all the nodes in `g` into `new_graph`
@@ -2508,7 +2508,7 @@ 
                      API Reference
 
                      
-on_generate_code(make_transformer)[source][source]
                      
+on_generate_code(make_transformer)[source][source]
 
                      Register a transformer function when python code is generated
                      
 
                      
 
                      
@@ -2581,7 +2581,7 @@ 
                      API Reference
 
                      
-output(result, type_expr=None)[source][source]
                      
+output(result, type_expr=None)[source][source]
 
                      Insert an output Node into the Graph. An output node represents
 a return statement in Python code. result is the value that should
 be returned.
                      
@@ -2607,7 +2607,7 @@ 
                      API Reference
 
                      
-output_node()[source][source]
                      
+output_node()[source][source]
 
                      
 
                      Warning
                      
 
                      This API is experimental and is NOT backward-compatible.
                      
@@ -2621,7 +2621,7 @@ 
                      API Reference
 
                      
-placeholder(name, type_expr=None, default_value)[source][source]
                      
+placeholder(name, type_expr=None, default_value)[source][source]
 
                      Insert a placeholder node into the Graph. A placeholder represents
 a function input.
                      
 
                      
@@ -2656,7 +2656,7 @@ 
                      API Reference
 
                      
-print_tabular()[source][source]
                      
+print_tabular()[source][source]
 
                      Prints the intermediate representation of the graph in tabular
 format. Note that this API requires the tabulate module to be
 installed.
                      
@@ -2668,7 +2668,7 @@ 
                      API Reference
 
                      
-process_inputs(*args)[source][source]
                      
+process_inputs(*args)[source][source]
 
                      Processes args so that they can be passed to the FX graph.
                      
 
                      
 
                      Warning
                      
@@ -2678,7 +2678,7 @@ 
                      API Reference
 
                      
-process_outputs(out)[source][source]
                      
+process_outputs(out)[source][source]
 
                      
 
                      Warning
                      
 
                      This API is experimental and is NOT backward-compatible.
                      
@@ -2687,7 +2687,7 @@ 
                      API Reference
 
                      
-python_code(root_module, *, verbose=False, include_stride=False, include_device=False, colored=False)[source][source]
                      
+python_code(root_module, *, verbose=False, include_stride=False, include_device=False, colored=False)[source][source]
 
                      Turn this Graph into valid Python code.
                      
 
                      
 
                      Parameters
                      
@@ -2710,7 +2710,7 @@ 
                      API Reference
 
                      
-set_codegen(codegen)[source][source]
                      
+set_codegen(codegen)[source][source]
 
                      
 
                      Warning
                      
 
                      This API is experimental and is NOT backward-compatible.
                      
@@ -2723,7 +2723,7 @@ 
                      API Reference
 
                      
-class torch.fx.Node(graph, name, op, target, args, kwargs, return_type=None)[source][source]
                      
+class torch.fx.Node(graph, name, op, target, args, kwargs, return_type=None)[source][source]
 
                      Node is the data structure that represents individual operations within
 a Graph. For the most part, Nodes represent callsites to various entities,
 such as operators, methods, and Modules (some exceptions include nodes that
@@ -2771,7 +2771,7 @@ 
                      API Reference
 
                      
-append(x)[source][source]
                      
+append(x)[source][source]
 
                      Insert x after this node in the list of nodes in the graph.
 Equivalent to self.next.prepend(x)
                      
 
                      
@@ -2797,7 +2797,7 @@ 
                      API Reference
 
                      
-format_node(placeholder_names=None, maybe_return_typename=None)[source][source]
                      
+format_node(placeholder_names=None, maybe_return_typename=None)[source][source]
 
                      Return a descriptive string representation of self.
                      
 
                      This method can be used with no arguments as a debugging
 utility.
                      
@@ -2840,7 +2840,7 @@ 
                      API Reference
 
                      
-insert_arg(idx, arg)[source][source]
                      
+insert_arg(idx, arg)[source][source]
 
                      Insert an positional argument to the argument list with given index.
                      
 
                      
 
                      Parameters
                      
@@ -2858,7 +2858,7 @@ 
                      API Reference
 
                      
-is_impure()[source][source]
                      
+is_impure()[source][source]
 
                      Returns whether this op is impure, i.e. if its op is a placeholder or
 output, or if a call_function or call_module which is impure.
                      
 
                      
@@ -2898,7 +2898,7 @@ 
                      API Reference
 
                      
-normalized_arguments(root, arg_types=None, kwarg_types=None, normalize_to_only_use_kwargs=False)[source][source]
                      
+normalized_arguments(root, arg_types=None, kwarg_types=None, normalize_to_only_use_kwargs=False)[source][source]
 
                      Returns normalized arguments to Python targets. This means that
 args/kwargs will be matched up to the module/functional’s
 signature and return exclusively kwargs in positional order
@@ -2931,7 +2931,7 @@ 
                      API Reference
 
                      
-prepend(x)[source][source]
                      
+prepend(x)[source][source]
 
                      Insert x before this node in the list of nodes in the graph. Example:
                      
 
                      Before: p -> self
         bx -> x -> ax
@@ -2963,7 +2963,7 @@ 
                      API Reference
 
                      
-replace_all_uses_with(replace_with, delete_user_cb=<function Node.<lambda>>, *, propagate_meta=False)[source][source]
                      
+replace_all_uses_with(replace_with, delete_user_cb=<function Node.<lambda>>, *, propagate_meta=False)[source][source]
 
                      Replace all uses of self in the Graph with the Node replace_with.
                      
 
                      
 
                      Parameters
                      
@@ -2992,7 +2992,7 @@ 
                      API Reference
 
                      
-replace_input_with(old_input, new_input)[source][source]
                      
+replace_input_with(old_input, new_input)[source][source]
 
                      Loop through input nodes of self, and replace all instances of
 old_input with new_input.
                      
 
                      
@@ -3023,7 +3023,7 @@ 
                      API Reference
 
                      
-update_arg(idx, arg)[source][source]
                      
+update_arg(idx, arg)[source][source]
 
                      Update an existing positional argument to contain the new value
 arg. After calling, self.args[idx] == arg.
                      
 
                      
@@ -3042,7 +3042,7 @@ 
                      API Reference
 
                      
-update_kwarg(key, arg)[source][source]
                      
+update_kwarg(key, arg)[source][source]
 
                      Update an existing keyword argument to contain the new value
 arg. After calling, self.kwargs[key] == arg.
                      
 
                      
@@ -3063,7 +3063,7 @@ 
                      API Reference
 
                      
-class torch.fx.Tracer(autowrap_modules=(math,), autowrap_functions=())[source][source]
                      
+class torch.fx.Tracer(autowrap_modules=(math,), autowrap_functions=())[source][source]
 
                      
 
                      Tracer is the class that implements the symbolic tracing functionality
 of torch.fx.symbolic_trace. A call to symbolic_trace(m) is equivalent
@@ -3080,7 +3080,7 @@ 
                      API Reference
 
                      
-call_module(m, forward, args, kwargs)[source][source]
                      
+call_module(m, forward, args, kwargs)[source][source]
 
                      Method that specifies the behavior of this Tracer when it encounters
 a call to an nn.Module instance.
                      
 
                      By default, the behavior is to check if the called module is a leaf module
@@ -3116,7 +3116,7 @@ 
                      API Reference
 
                      
-create_arg(a)[source][source]
                      
+create_arg(a)[source][source]
 
                      A method to specify the behavior of tracing when preparing values to
 be used as arguments to nodes in the Graph.
                      
 
                      By default, the behavior includes:
                      
@@ -3154,7 +3154,7 @@ 
                      API Reference
 
                      
-create_args_for_root(root_fn, is_module, concrete_args=None)[source][source]
                      
+create_args_for_root(root_fn, is_module, concrete_args=None)[source][source]
 
                      Create placeholder nodes corresponding to the signature of the root
 Module. This method introspects root’s signature and emits those
 nodes accordingly, also supporting *args and **kwargs.
                      
@@ -3166,7 +3166,7 @@ 
                      API Reference
 
                      
-create_node(kind, target, args, kwargs, name=None, type_expr=None)[source]
                      
+create_node(kind, target, args, kwargs, name=None, type_expr=None)[source]
 
                      Inserts a graph node given target, args, kwargs, and name.
                      
 
                      This method can be overridden to do extra checking, validation, or
 modification of values used in node creation. For example, one might
@@ -3184,7 +3184,7 @@ 
                      API Reference
 
                      
-create_proxy(kind, target, args, kwargs, name=None, type_expr=None, proxy_factory_fn=None)[source]
                      
+create_proxy(kind, target, args, kwargs, name=None, type_expr=None, proxy_factory_fn=None)[source]
 
                      Create a Node from the given arguments, then return the Node
 wrapped in a Proxy object.
                      
 
                      If kind = ‘placeholder’, then we’re creating a Node that
@@ -3201,7 +3201,7 @@ 
                      API Reference
 
                      
-get_fresh_qualname(prefix)[source][source]
                      
+get_fresh_qualname(prefix)[source][source]
 
                      Gets a fresh name for a prefix and returns it. This function ensures
 that it will not clash with an existing attribute on the graph.
                      
 
                      
@@ -3217,7 +3217,7 @@ 
                      API Reference
 
                      
-getattr(attr, attr_val, parameter_proxy_cache)[source][source]
                      
+getattr(attr, attr_val, parameter_proxy_cache)[source][source]
 
                      Method that specifies the behavior of this Tracer when we call getattr
 on a call to an nn.Module instance.
                      
 
                      By default, the behavior is to return a proxy value for the attribute. It
@@ -3245,7 +3245,7 @@ 
                      API Reference
 
                      
-is_leaf_module(m, module_qualified_name)[source][source]
                      
+is_leaf_module(m, module_qualified_name)[source][source]
 
                      A method to specify whether a given nn.Module is a “leaf” module.
                      
 
                      Leaf modules are the atomic units that appear in
 the IR, referenced by call_module calls. By default,
@@ -3275,7 +3275,7 @@ 
                      API Reference
 
                      
-iter(obj)[source]
                      
+iter(obj)[source]
 
                      
 
                      Called when a proxy object is being iterated over, such as
                      when used in control flow.  Normally we don’t know what to do because
 we don’t know the value of the proxy, but a custom tracer can attach more
@@ -3295,7 +3295,7 @@ 
                      API Reference
 
                      
-keys(obj)[source]
                      
+keys(obj)[source]
 
                      
 
                      Called when a proxy object is has the keys() method called.
                      This is what happens when ** is called on a proxy. This should return an
 iterator it ** is suppose to work in your custom tracer.
                      
@@ -3314,7 +3314,7 @@ 
                      API Reference
 
                      
-path_of_module(mod)[source][source]
                      
+path_of_module(mod)[source][source]
 
                      Helper method to find the qualified name of mod in the Module hierarchy
 of root. For example, if root has a submodule named foo, which has
 a submodule named bar, passing bar into this function will return
@@ -3335,7 +3335,7 @@ 
                      API Reference
 
                      
-proxy(node)[source]
                      
+proxy(node)[source]
 
                      
 
                      Note
                      
 
                      Backwards-compatibility for this API is guaranteed.
                      
@@ -3349,7 +3349,7 @@ 
                      API Reference
 
                      
-to_bool(obj)[source]
                      
+to_bool(obj)[source]
 
                      
 
                      Called when a proxy object is being converted to a boolean, such as
                      when used in control flow.  Normally we don’t know what to do because
 we don’t know the value of the proxy, but a custom tracer can attach more
@@ -3369,7 +3369,7 @@ 
                      API Reference
 
                      
-trace(root, concrete_args=None)[source][source]
                      
+trace(root, concrete_args=None)[source][source]
 
                      Trace root and return the corresponding FX Graph representation. root
 can either be an nn.Module instance or a Python callable.
                      
 
                      Note that after this call, self.root may be different from the root passed
@@ -3404,7 +3404,7 @@ 
                      API Reference
 
                      
-class torch.fx.Proxy(node, tracer=None)[source][source]
                      
+class torch.fx.Proxy(node, tracer=None)[source][source]
 
                      Proxy objects are Node wrappers that flow through the
 program during symbolic tracing and record all the operations
 (torch function calls, method calls, operators) that they touch
@@ -3437,7 +3437,7 @@ 
                      API Reference
 
                      
-class torch.fx.Interpreter(module, garbage_collect_values=True, graph=None)[source][source]
                      
+class torch.fx.Interpreter(module, garbage_collect_values=True, graph=None)[source][source]
 
                      An Interpreter executes an FX graph Node-by-Node. This pattern
 can be useful for many things, including writing code
 transformations as well as analysis passes.
                      
@@ -3501,7 +3501,7 @@ 
                      API Reference
 
                      
-boxed_run(args_list)[source][source]
                      
+boxed_run(args_list)[source][source]
 
                      Run module via interpretation and return the result.  This uses the “boxed”
 calling convention, where you pass a list of arguments, which will be cleared
 by the interpreter.  This ensures that input tensors are promptly deallocated.
                      
@@ -3513,7 +3513,7 @@ 
                      API Reference
 
                      
-call_function(target, args, kwargs)[source][source]
                      
+call_function(target, args, kwargs)[source][source]
 
                      Execute a call_function node and return the result.
                      
 
                      
 
                      Parameters
                      
@@ -3541,7 +3541,7 @@ 
                      API Reference
 
                      
-call_method(target, args, kwargs)[source][source]
                      
+call_method(target, args, kwargs)[source][source]
 
                      Execute a call_method node and return the result.
                      
 
                      
 
                      Parameters
                      
@@ -3569,7 +3569,7 @@ 
                      API Reference
 
                      
-call_module(target, args, kwargs)[source][source]
                      
+call_module(target, args, kwargs)[source][source]
 
                      Execute a call_module node and return the result.
                      
 
                      
 
                      Parameters
                      
@@ -3597,7 +3597,7 @@ 
                      API Reference
 
                      
-fetch_args_kwargs_from_env(n)[source][source]
                      
+fetch_args_kwargs_from_env(n)[source][source]
 
                      Fetch the concrete values of args and kwargs of node n
 from the current execution environment.
                      
 
                      
@@ -3619,7 +3619,7 @@ 
                      API Reference
 
                      
-fetch_attr(target)[source][source]
                      
+fetch_attr(target)[source][source]
 
                      Fetch an attribute from the Module hierarchy of self.module.
                      
 
                      
 
                      Parameters
                      
@@ -3640,7 +3640,7 @@ 
                      API Reference
 
                      
-get_attr(target, args, kwargs)[source][source]
                      
+get_attr(target, args, kwargs)[source][source]
 
                      Execute a get_attr node. Will retrieve an attribute
 value from the Module hierarchy of self.module.
                      
 
                      
@@ -3668,7 +3668,7 @@ 
                      API Reference
 
                      
-map_nodes_to_values(args, n)[source][source]
                      
+map_nodes_to_values(args, n)[source][source]
 
                      Recursively descend through args and look up the concrete value
 for each Node in the current execution environment.
                      
 
                      
@@ -3690,7 +3690,7 @@ 
                      API Reference
 
                      
-output(target, args, kwargs)[source][source]
                      
+output(target, args, kwargs)[source][source]
 
                      Execute an output node. This really just retrieves
 the value referenced by the output node and returns it.
                      
 
                      
@@ -3718,7 +3718,7 @@ 
                      API Reference
 
                      
-placeholder(target, args, kwargs)[source][source]
                      
+placeholder(target, args, kwargs)[source][source]
 
                      Execute a placeholder node. Note that this is stateful:
 Interpreter maintains an internal iterator over
 arguments passed to run and this method returns
@@ -3748,7 +3748,7 @@ 
                      API Reference
 
                      
-run(*args, initial_env=None, enable_io_processing=True)[source][source]
                      
+run(*args, initial_env=None, enable_io_processing=True)[source][source]
 
                      Run module via interpretation and return the result.
                      
 
                      
 
                      Parameters
                      
@@ -3777,7 +3777,7 @@ 
                      API Reference
 
                      
-run_node(n)[source][source]
                      
+run_node(n)[source][source]
 
                      Run a specific node n and return the result.
 Calls into placeholder, get_attr, call_function,
 call_method, call_module, or output depending
@@ -3803,7 +3803,7 @@ 
                      API Reference
 
                      
-class torch.fx.Transformer(module)[source][source]
                      
+class torch.fx.Transformer(module)[source][source]
 
                      Transformer is a special type of interpreter that produces a
 new Module. It exposes a transform() method that returns
 the transformed Module. Transformer does not require
@@ -3852,7 +3852,7 @@ 
                      API Reference
 
                      
-call_function(target, args, kwargs)[source][source]
                      
+call_function(target, args, kwargs)[source][source]
 
                      
 
                      Note
                      
 
                      Backwards-compatibility for this API is guaranteed.
                      
@@ -3866,7 +3866,7 @@ 
                      API Reference
 
                      
-call_module(target, args, kwargs)[source][source]
                      
+call_module(target, args, kwargs)[source][source]
 
                      
 
                      Note
                      
 
                      Backwards-compatibility for this API is guaranteed.
                      
@@ -3880,7 +3880,7 @@ 
                      API Reference
 
                      
-get_attr(target, args, kwargs)[source][source]
                      
+get_attr(target, args, kwargs)[source][source]
 
                      Execute a get_attr node. In Transformer, this is
 overridden to insert a new get_attr node into the output
 graph.
                      
@@ -3906,7 +3906,7 @@ 
                      API Reference
 
                      
-placeholder(target, args, kwargs)[source][source]
                      
+placeholder(target, args, kwargs)[source][source]
 
                      Execute a placeholder node. In Transformer, this is
 overridden to insert a new placeholder into the output
 graph.
                      
@@ -3932,7 +3932,7 @@ 
                      API Reference
 
                      
-transform()[source][source]
                      
+transform()[source][source]
 
                      Transform self.module and return the transformed
 GraphModule.
                      
 
                      
@@ -3950,7 +3950,7 @@ 
                      API Reference
 
                      
-torch.fx.replace_pattern(gm, pattern, replacement)[source][source]
                      
+torch.fx.replace_pattern(gm, pattern, replacement)[source][source]
 
                      Matches all possible non-overlapping sets of operators and their
 data dependencies (pattern) in the Graph of a GraphModule
 (gm), then replaces each of these matched subgraphs with another
diff --git a/2.7/generated/torch.Tensor.backward.html b/2.7/generated/torch.Tensor.backward.html
index 7c5ef5bc51f..24f7a9636a8 100644
--- a/2.7/generated/torch.Tensor.backward.html
+++ b/2.7/generated/torch.Tensor.backward.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.backward
                      
 
                      
 
                      
-Tensor.backward(gradient=None, retain_graph=None, create_graph=False, inputs=None)[source][source]
                      
+Tensor.backward(gradient=None, retain_graph=None, create_graph=False, inputs=None)[source][source]
 
                      Computes the gradient of current tensor wrt graph leaves.
                      
 
                      The graph is differentiated using the chain rule. If the tensor is
 non-scalar (i.e. its data has more than one element) and requires
diff --git a/2.7/generated/torch.Tensor.dim_order.html b/2.7/generated/torch.Tensor.dim_order.html
index a5eb5c06de7..c3940b87b9f 100644
--- a/2.7/generated/torch.Tensor.dim_order.html
+++ b/2.7/generated/torch.Tensor.dim_order.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.dim_order
                      
 
                      
 
                      
-Tensor.dim_order(ambiguity_check=False) → tuple[source][source]
                      
+Tensor.dim_order(ambiguity_check=False) → tuple[source][source]
 
                      Returns the uniquely determined tuple of int describing the dim order or
 physical layout of self.
                      
 
                      The dim order represents how dimensions are laid out in memory of dense tensors,
diff --git a/2.7/generated/torch.Tensor.is_shared.html b/2.7/generated/torch.Tensor.is_shared.html
index d89783e15ef..f89e9f79903 100644
--- a/2.7/generated/torch.Tensor.is_shared.html
+++ b/2.7/generated/torch.Tensor.is_shared.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.is_shared
                      
 
                      
 
                      
-Tensor.is_shared()[source][source]
                      
+Tensor.is_shared()[source][source]
 
                      Checks if tensor is in shared memory.
                      
 
                      This is always True for CUDA tensors.
                      
 
                      
diff --git a/2.7/generated/torch.Tensor.istft.html b/2.7/generated/torch.Tensor.istft.html
index 67af2b17812..2a6cdeeef58 100644
--- a/2.7/generated/torch.Tensor.istft.html
+++ b/2.7/generated/torch.Tensor.istft.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.istft
                      
 
                      
 
                      
-Tensor.istft(n_fft, hop_length=None, win_length=None, window=None, center=True, normalized=False, onesided=None, length=None, return_complex=False)[source][source]
                      
+Tensor.istft(n_fft, hop_length=None, win_length=None, window=None, center=True, normalized=False, onesided=None, length=None, return_complex=False)[source][source]
 
                      See torch.istft()
                      
 
                      
 
                      
diff --git a/2.7/generated/torch.Tensor.lu.html b/2.7/generated/torch.Tensor.lu.html
index 52093e87401..6f68cacf7f2 100644
--- a/2.7/generated/torch.Tensor.lu.html
+++ b/2.7/generated/torch.Tensor.lu.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.lu
                      
 
                      
 
                      
-Tensor.lu(pivot=True, get_infos=False)[source][source]
                      
+Tensor.lu(pivot=True, get_infos=False)[source][source]
 
                      See torch.lu()
                      
 
                      
 
diff --git a/2.7/generated/torch.Tensor.module_load.html b/2.7/generated/torch.Tensor.module_load.html
index 14930171161..8f5d84ca137 100644
--- a/2.7/generated/torch.Tensor.module_load.html
+++ b/2.7/generated/torch.Tensor.module_load.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.module_load
                      
 
                      
 
                      
-Tensor.module_load(other, assign=False)[source][source]
                      
+Tensor.module_load(other, assign=False)[source][source]
 
                      Defines how to transform other when loading it into self in load_state_dict().
                      
 
                      Used when get_swap_module_params_on_conversion() is True.
                      
 
                      It is expected that self is a parameter or buffer in an nn.Module and other is the
diff --git a/2.7/generated/torch.Tensor.norm.html b/2.7/generated/torch.Tensor.norm.html
index aa81f397080..97cedbbfbb9 100644
--- a/2.7/generated/torch.Tensor.norm.html
+++ b/2.7/generated/torch.Tensor.norm.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.norm
                      
 
                      
 
                      
-Tensor.norm(p='fro', dim=None, keepdim=False, dtype=None)[source][source]
                      
+Tensor.norm(p='fro', dim=None, keepdim=False, dtype=None)[source][source]
 
                      See torch.norm()
                      
 
                      
 
                      
diff --git a/2.7/generated/torch.Tensor.register_hook.html b/2.7/generated/torch.Tensor.register_hook.html
index e782c124015..b6de6c4d9ac 100644
--- a/2.7/generated/torch.Tensor.register_hook.html
+++ b/2.7/generated/torch.Tensor.register_hook.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.register_hook
                      
 
                      
 
                      
-Tensor.register_hook(hook)[source][source]
                      
+Tensor.register_hook(hook)[source][source]
 
                      Registers a backward hook.
                      
 
                      The hook will be called every time a gradient with respect to the
 Tensor is computed. The hook should have the following signature:
                      
diff --git a/2.7/generated/torch.Tensor.register_post_accumulate_grad_hook.html b/2.7/generated/torch.Tensor.register_post_accumulate_grad_hook.html
index 91715a83a45..d6262a70b6b 100644
--- a/2.7/generated/torch.Tensor.register_post_accumulate_grad_hook.html
+++ b/2.7/generated/torch.Tensor.register_post_accumulate_grad_hook.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.register_post_accumulate_grad_hook
                      
 
                      
 
                      
-Tensor.register_post_accumulate_grad_hook(hook)[source][source]
                      
+Tensor.register_post_accumulate_grad_hook(hook)[source][source]
 
                      Registers a backward hook that runs after grad accumulation.
                      
 
                      The hook will be called after all gradients for a tensor have been accumulated,
 meaning that the .grad field has been updated on that tensor. The post
diff --git a/2.7/generated/torch.Tensor.share_memory_.html b/2.7/generated/torch.Tensor.share_memory_.html
index 2ef9d56bb9b..4212f72da90 100644
--- a/2.7/generated/torch.Tensor.share_memory_.html
+++ b/2.7/generated/torch.Tensor.share_memory_.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.share_memory_
                      
 
                      
 
                      
-Tensor.share_memory_()[source][source]
                      
+Tensor.share_memory_()[source][source]
 
                      Moves the underlying storage to shared memory.
                      
 
                      This is a no-op if the underlying storage is already in shared memory
 and for CUDA tensors. Tensors in shared memory cannot be resized.
                      
diff --git a/2.7/generated/torch.Tensor.split.html b/2.7/generated/torch.Tensor.split.html
index 98b255c5181..dc91bb818b9 100644
--- a/2.7/generated/torch.Tensor.split.html
+++ b/2.7/generated/torch.Tensor.split.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.split
                      
 
                      
 
                      
-Tensor.split(split_size, dim=0)[source][source]
                      
+Tensor.split(split_size, dim=0)[source][source]
 
                      See torch.split()
                      
 
                      
 
diff --git a/2.7/generated/torch.Tensor.stft.html b/2.7/generated/torch.Tensor.stft.html
index 6eb4f8af520..2331a454278 100644
--- a/2.7/generated/torch.Tensor.stft.html
+++ b/2.7/generated/torch.Tensor.stft.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.stft
                      
 
                      
 
                      
-Tensor.stft(n_fft, hop_length=None, win_length=None, window=None, center=True, pad_mode='reflect', normalized=False, onesided=None, return_complex=None, align_to_window=None)[source][source]
                      
+Tensor.stft(n_fft, hop_length=None, win_length=None, window=None, center=True, pad_mode='reflect', normalized=False, onesided=None, return_complex=None, align_to_window=None)[source][source]
 
                      See torch.stft()
                      
 
                      
 
                      Warning
                      
diff --git a/2.7/generated/torch.Tensor.storage.html b/2.7/generated/torch.Tensor.storage.html
index 6b094b72ce4..75fae00a261 100644
--- a/2.7/generated/torch.Tensor.storage.html
+++ b/2.7/generated/torch.Tensor.storage.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.storage
                      
 
                      
 
                      
-Tensor.storage() → torch.TypedStorage[source][source]
                      
+Tensor.storage() → torch.TypedStorage[source][source]
 
                      Returns the underlying TypedStorage.
                      
 
                      
 
                      Warning
                      
diff --git a/2.7/generated/torch.Tensor.storage_type.html b/2.7/generated/torch.Tensor.storage_type.html
index 6600cd2b89e..945c12cc0dc 100644
--- a/2.7/generated/torch.Tensor.storage_type.html
+++ b/2.7/generated/torch.Tensor.storage_type.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.storage_type
                      
 
                      
 
                      
-Tensor.storage_type() → type[source][source]
                      
+Tensor.storage_type() → type[source][source]
 
                      Returns the type of the underlying storage.
                      
 
                      
 
diff --git a/2.7/generated/torch.Tensor.to_sparse_coo.html b/2.7/generated/torch.Tensor.to_sparse_coo.html
index 80209d4c05a..445824c7faa 100644
--- a/2.7/generated/torch.Tensor.to_sparse_coo.html
+++ b/2.7/generated/torch.Tensor.to_sparse_coo.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.to_sparse_coo
                      
 
                      
 
                      
-Tensor.to_sparse_coo()[source][source]
                      
+Tensor.to_sparse_coo()[source][source]
 
                      Convert a tensor to coordinate format.
                      
 
                      Examples:
                      
 
                      >>> dense = torch.randn(5, 5)
diff --git a/2.7/generated/torch.Tensor.unflatten.html b/2.7/generated/torch.Tensor.unflatten.html
index 9b3e6769fa3..8ae53bb332e 100644
--- a/2.7/generated/torch.Tensor.unflatten.html
+++ b/2.7/generated/torch.Tensor.unflatten.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.unflatten
                      
 
                      
 
                      
-Tensor.unflatten(dim, sizes) → Tensor[source][source]
                      
+Tensor.unflatten(dim, sizes) → Tensor[source][source]
 
                      See torch.unflatten().
                      
 
                      
 
diff --git a/2.7/generated/torch.Tensor.unique.html b/2.7/generated/torch.Tensor.unique.html
index 40d21f056fd..9a5f4a3bb5f 100644
--- a/2.7/generated/torch.Tensor.unique.html
+++ b/2.7/generated/torch.Tensor.unique.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.unique
                      
 
                      
 
                      
-Tensor.unique(sorted=True, return_inverse=False, return_counts=False, dim=None)[source][source]
                      
+Tensor.unique(sorted=True, return_inverse=False, return_counts=False, dim=None)[source][source]
 
                      Returns the unique elements of the input tensor.
                      
 
                      See torch.unique()
                      
 
                      
diff --git a/2.7/generated/torch.Tensor.unique_consecutive.html b/2.7/generated/torch.Tensor.unique_consecutive.html
index e94c6bf23c0..cbd0dd414b1 100644
--- a/2.7/generated/torch.Tensor.unique_consecutive.html
+++ b/2.7/generated/torch.Tensor.unique_consecutive.html
@@ -596,7 +596,7 @@
 
                      torch.Tensor.unique_consecutive
                      
 
                      
 
                      
-Tensor.unique_consecutive(return_inverse=False, return_counts=False, dim=None)[source][source]
                      
+Tensor.unique_consecutive(return_inverse=False, return_counts=False, dim=None)[source][source]
 
                      Eliminates all but the first element from every consecutive group of equivalent elements.
                      
 
                      See torch.unique_consecutive()
                      
 
                      
diff --git a/2.7/generated/torch._assert.html b/2.7/generated/torch._assert.html
index cfb1781e260..a5c2357b009 100644
--- a/2.7/generated/torch._assert.html
+++ b/2.7/generated/torch._assert.html
@@ -596,7 +596,7 @@
 
                      torch._assert
                      
 
                      
 
                      
-torch._assert(condition, message)[source][source]
                      
+torch._assert(condition, message)[source][source]
 
                      A wrapper around Python’s assert which is symbolically traceable.
                      
 
                      
 
diff --git a/2.7/generated/torch._logging.set_logs.html b/2.7/generated/torch._logging.set_logs.html
index 9ef7f0919b2..612ec2c8197 100644
--- a/2.7/generated/torch._logging.set_logs.html
+++ b/2.7/generated/torch._logging.set_logs.html
@@ -600,7 +600,7 @@
 
                      torch._logging.set_logs
                      
 
                      
 
                      
-torch._logging.set_logs(*, all=None, dynamo=None, aot=None, autograd=None, dynamic=None, inductor=None, distributed=None, c10d=None, ddp=None, fsdp=None, dtensor=None, onnx=None, bytecode=False, aot_graphs=False, aot_joint_graph=False, ddp_graphs=False, graph=False, graph_code=False, graph_breaks=False, graph_sizes=False, guards=False, recompiles=False, recompiles_verbose=False, trace_source=False, trace_call=False, trace_bytecode=False, output_code=False, kernel_code=False, schedule=False, perf_hints=False, pre_grad_graphs=False, post_grad_graphs=False, ir_pre_fusion=False, ir_post_fusion=False, onnx_diagnostics=False, fusion=False, overlap=False, export=None, modules=None, cudagraphs=False, sym_node=False, compiled_autograd=False, compiled_autograd_verbose=False, cudagraph_static_inputs=False, benchmarking=False, autotuning=False, graph_region_expansion=False)[source][source]
                      
+torch._logging.set_logs(*, all=None, dynamo=None, aot=None, autograd=None, dynamic=None, inductor=None, distributed=None, c10d=None, ddp=None, fsdp=None, dtensor=None, onnx=None, bytecode=False, aot_graphs=False, aot_joint_graph=False, ddp_graphs=False, graph=False, graph_code=False, graph_breaks=False, graph_sizes=False, guards=False, recompiles=False, recompiles_verbose=False, trace_source=False, trace_call=False, trace_bytecode=False, output_code=False, kernel_code=False, schedule=False, perf_hints=False, pre_grad_graphs=False, post_grad_graphs=False, ir_pre_fusion=False, ir_post_fusion=False, onnx_diagnostics=False, fusion=False, overlap=False, export=None, modules=None, cudagraphs=False, sym_node=False, compiled_autograd=False, compiled_autograd_verbose=False, cudagraph_static_inputs=False, benchmarking=False, autotuning=False, graph_region_expansion=False)[source][source]
 
                      Sets the log level for individual components and toggles individual log
 artifact types.
                      
 
                      
diff --git a/2.7/generated/torch.accelerator.current_accelerator.html b/2.7/generated/torch.accelerator.current_accelerator.html
index c616edf7753..d6b00b44653 100644
--- a/2.7/generated/torch.accelerator.current_accelerator.html
+++ b/2.7/generated/torch.accelerator.current_accelerator.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.current_accelerator
                      
 
                      
 
                      
-torch.accelerator.current_accelerator(check_available=False)[source][source]
                      
+torch.accelerator.current_accelerator(check_available=False)[source][source]
 
                      Return the device of the accelerator available at compilation time.
 If no accelerator were available at compilation time, returns None.
 See accelerator for details.
                      
diff --git a/2.7/generated/torch.accelerator.current_device_idx.html b/2.7/generated/torch.accelerator.current_device_idx.html
index b72de762ba7..09b3927b7fa 100644
--- a/2.7/generated/torch.accelerator.current_device_idx.html
+++ b/2.7/generated/torch.accelerator.current_device_idx.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.current_device_idx
                      
 
                      
 
                      
-torch.accelerator.current_device_idx()[source]
                      
+torch.accelerator.current_device_idx()[source]
 
                      Return the index of a currently selected device for the current accelerator.
                      
 
                      
 
                      Returns
                      
diff --git a/2.7/generated/torch.accelerator.current_device_index.html b/2.7/generated/torch.accelerator.current_device_index.html
index b7349c8552c..52de13a404b 100644
--- a/2.7/generated/torch.accelerator.current_device_index.html
+++ b/2.7/generated/torch.accelerator.current_device_index.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.current_device_index
                      
 
                      
 
                      
-torch.accelerator.current_device_index()[source][source]
                      
+torch.accelerator.current_device_index()[source][source]
 
                      Return the index of a currently selected device for the current accelerator.
                      
 
                      
 
                      Returns
                      
diff --git a/2.7/generated/torch.accelerator.current_stream.html b/2.7/generated/torch.accelerator.current_stream.html
index bd93be98220..8ef109d3c0d 100644
--- a/2.7/generated/torch.accelerator.current_stream.html
+++ b/2.7/generated/torch.accelerator.current_stream.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.current_stream
                      
 
                      
 
                      
-torch.accelerator.current_stream(device=None, /)[source][source]
                      
+torch.accelerator.current_stream(device=None, /)[source][source]
 
                      Return the currently selected stream for a given device.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.accelerator.device_count.html b/2.7/generated/torch.accelerator.device_count.html
index c2424307e8e..9a656e49980 100644
--- a/2.7/generated/torch.accelerator.device_count.html
+++ b/2.7/generated/torch.accelerator.device_count.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.device_count
                      
 
                      
 
                      
-torch.accelerator.device_count()[source][source]
                      
+torch.accelerator.device_count()[source][source]
 
                      Return the number of current accelerator available.
                      
 
                      
 
                      Returns
                      
diff --git a/2.7/generated/torch.accelerator.is_available.html b/2.7/generated/torch.accelerator.is_available.html
index 154c1ba303c..218782297bd 100644
--- a/2.7/generated/torch.accelerator.is_available.html
+++ b/2.7/generated/torch.accelerator.is_available.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.is_available
                      
 
                      
 
                      
-torch.accelerator.is_available()[source][source]
                      
+torch.accelerator.is_available()[source][source]
 
                      Check if the current accelerator is available at runtime: it was build, all the
 required drivers are available and at least one device is visible.
 See accelerator for details.
                      
diff --git a/2.7/generated/torch.accelerator.set_device_idx.html b/2.7/generated/torch.accelerator.set_device_idx.html
index 1bdd78381ae..b3f15a69b75 100644
--- a/2.7/generated/torch.accelerator.set_device_idx.html
+++ b/2.7/generated/torch.accelerator.set_device_idx.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.set_device_idx
                      
 
                      
 
                      
-torch.accelerator.set_device_idx(device, /)[source]
                      
+torch.accelerator.set_device_idx(device, /)[source]
 
                      Set the current device index to a given device.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.accelerator.set_device_index.html b/2.7/generated/torch.accelerator.set_device_index.html
index 6635344ec3e..0498d94357c 100644
--- a/2.7/generated/torch.accelerator.set_device_index.html
+++ b/2.7/generated/torch.accelerator.set_device_index.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.set_device_index
                      
 
                      
 
                      
-torch.accelerator.set_device_index(device, /)[source][source]
                      
+torch.accelerator.set_device_index(device, /)[source][source]
 
                      Set the current device index to a given device.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.accelerator.set_stream.html b/2.7/generated/torch.accelerator.set_stream.html
index ddf3792b975..32b55d7977b 100644
--- a/2.7/generated/torch.accelerator.set_stream.html
+++ b/2.7/generated/torch.accelerator.set_stream.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.set_stream
                      
 
                      
 
                      
-torch.accelerator.set_stream(stream)[source][source]
                      
+torch.accelerator.set_stream(stream)[source][source]
 
                      Set the current stream to a given stream.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.accelerator.synchronize.html b/2.7/generated/torch.accelerator.synchronize.html
index 735629a55bf..ddf93c3cf12 100644
--- a/2.7/generated/torch.accelerator.synchronize.html
+++ b/2.7/generated/torch.accelerator.synchronize.html
@@ -596,7 +596,7 @@
 
                      torch.accelerator.synchronize
                      
 
                      
 
                      
-torch.accelerator.synchronize(device=None, /)[source][source]
                      
+torch.accelerator.synchronize(device=None, /)[source][source]
 
                      Wait for all kernels in all streams on the given device to complete.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.BNReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.BNReLU2d.html
index c7165f0af1a..1496c45c347 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.BNReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.BNReLU2d.html
@@ -598,7 +598,7 @@
 
                      BNReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.BNReLU2d(batch_norm, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.BNReLU2d(batch_norm, relu)[source][source]
 
                      This is a sequential container which calls the BatchNorm 2d and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.BNReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.BNReLU3d.html
index 42446ab80ce..38c191ad968 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.BNReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.BNReLU3d.html
@@ -598,7 +598,7 @@
 
                      BNReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.BNReLU3d(batch_norm, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.BNReLU3d(batch_norm, relu)[source][source]
 
                      This is a sequential container which calls the BatchNorm 3d and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBn1d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBn1d.html
index 98c076f04eb..2533e0672e2 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBn1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBn1d.html
@@ -598,7 +598,7 @@
 
                      ConvBn1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBn1d(conv, bn)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBn1d(conv, bn)[source][source]
 
                      This is a sequential container which calls the Conv 1d and Batch Norm 1d modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBn2d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBn2d.html
index b2b376190cb..ee7da07e791 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBn2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBn2d.html
@@ -598,7 +598,7 @@
 
                      ConvBn2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBn2d(conv, bn)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBn2d(conv, bn)[source][source]
 
                      This is a sequential container which calls the Conv 2d and Batch Norm 2d modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBn3d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBn3d.html
index 1cf959b2efe..748b3ee47e2 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBn3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBn3d.html
@@ -598,7 +598,7 @@
 
                      ConvBn3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBn3d(conv, bn)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBn3d(conv, bn)[source][source]
 
                      This is a sequential container which calls the Conv 3d and Batch Norm 3d modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU1d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU1d.html
index 83f424607b8..6fbe7c83d0e 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU1d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBnReLU1d(conv, bn, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBnReLU1d(conv, bn, relu)[source][source]
 
                      This is a sequential container which calls the Conv 1d, Batch Norm 1d, and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU2d.html
index b91cdcd67f2..8c685b8eb08 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU2d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBnReLU2d(conv, bn, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBnReLU2d(conv, bn, relu)[source][source]
 
                      This is a sequential container which calls the Conv 2d, Batch Norm 2d, and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU3d.html
index 3265e30da56..5fe396478a4 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvBnReLU3d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvBnReLU3d(conv, bn, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvBnReLU3d(conv, bn, relu)[source][source]
 
                      This is a sequential container which calls the Conv 3d, Batch Norm 3d, and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU1d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU1d.html
index bf6e1795337..b3f5383746a 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU1d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvReLU1d(conv, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvReLU1d(conv, relu)[source][source]
 
                      This is a sequential container which calls the Conv1d and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU2d.html
index 28d50785cdb..2e0205df205 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU2d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvReLU2d(conv, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvReLU2d(conv, relu)[source][source]
 
                      This is a sequential container which calls the Conv2d and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU3d.html
index 05f1b5a21d5..72ea3d5ffbc 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.ConvReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.ConvReLU3d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.ConvReLU3d(conv, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.ConvReLU3d(conv, relu)[source][source]
 
                      This is a sequential container which calls the Conv3d and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.LinearReLU.html b/2.7/generated/torch.ao.nn.intrinsic.LinearReLU.html
index 65daa1c6e49..1383660d7df 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.LinearReLU.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.LinearReLU.html
@@ -598,7 +598,7 @@
 
                      LinearReLU
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.LinearReLU(linear, relu)[source][source]
                      
+class torch.ao.nn.intrinsic.LinearReLU(linear, relu)[source][source]
 
                      This is a sequential container which calls the Linear and ReLU modules.
 During quantization this will be replaced with the corresponding fused module.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn1d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn1d.html
index b106b35ffe1..8f2f2de1f1d 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn1d.html
@@ -598,7 +598,7 @@
 
                      ConvBn1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBn1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBn1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBn1d module is a module fused from Conv1d and BatchNorm1d,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn2d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn2d.html
index 4eebcff63ca..067a08ad578 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn2d.html
@@ -598,7 +598,7 @@
 
                      ConvBn2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBn2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBn2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBn2d module is a module fused from Conv2d and BatchNorm2d,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn3d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn3d.html
index d43c7009584..3ddd9d44a74 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBn3d.html
@@ -598,7 +598,7 @@
 
                      ConvBn3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBn3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBn3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBn3d module is a module fused from Conv3d and BatchNorm3d,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU1d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU1d.html
index 8dbb4dbd455..f5477d1c411 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU1d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBnReLU1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBnReLU1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBnReLU1d module is a module fused from Conv1d, BatchNorm1d and ReLU,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU2d.html
index 1ac4bd579a9..c09558676bf 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU2d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBnReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBnReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBnReLU2d module is a module fused from Conv2d, BatchNorm2d and ReLU,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU3d.html
index 42448a00c93..bc82d3f5885 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvBnReLU3d.html
@@ -598,7 +598,7 @@
 
                      ConvBnReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvBnReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvBnReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=None, padding_mode='zeros', eps=1e-05, momentum=0.1, freeze_bn=False, qconfig=None)[source][source]
 
                      A ConvBnReLU3d module is a module fused from Conv3d, BatchNorm3d and ReLU,
 attached with FakeQuantize modules for weight,
 used in quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU2d.html
index 09531a6a4a2..85c50e1bfb5 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU2d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None)[source][source]
 
                      A ConvReLU2d module is a fused module of Conv2d and ReLU, attached with
 FakeQuantize modules for weight for
 quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU3d.html
index 2be81e741ed..b880c1637d5 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.ConvReLU3d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.ConvReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.ConvReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None)[source][source]
 
                      A ConvReLU3d module is a fused module of Conv3d and ReLU, attached with
 FakeQuantize modules for weight for
 quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.LinearReLU.html b/2.7/generated/torch.ao.nn.intrinsic.qat.LinearReLU.html
index 57c893e1486..83fa2279e75 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.LinearReLU.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.LinearReLU.html
@@ -598,7 +598,7 @@
 
                      LinearReLU
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.LinearReLU(in_features, out_features, bias=True, qconfig=None)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.LinearReLU(in_features, out_features, bias=True, qconfig=None)[source][source]
 
                      A LinearReLU module fused from Linear and ReLU modules, attached with
 FakeQuantize modules for weight, used in
 quantization aware training.
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.freeze_bn_stats.html b/2.7/generated/torch.ao.nn.intrinsic.qat.freeze_bn_stats.html
index ae694b005f8..203fbc90f32 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.freeze_bn_stats.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.freeze_bn_stats.html
@@ -598,7 +598,7 @@
 
                      freeze_bn_stats
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.freeze_bn_stats(mod)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.freeze_bn_stats(mod)[source][source]
 
                      
 
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.qat.update_bn_stats.html b/2.7/generated/torch.ao.nn.intrinsic.qat.update_bn_stats.html
index 55639e657ae..6438684af75 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.qat.update_bn_stats.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.qat.update_bn_stats.html
@@ -598,7 +598,7 @@
 
                      update_bn_stats
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.qat.update_bn_stats(mod)[source][source]
                      
+class torch.ao.nn.intrinsic.qat.update_bn_stats(mod)[source][source]
 
                      
 
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU2d.html
index 7a94dac79d7..34252b1544d 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU2d.html
@@ -598,7 +598,7 @@
 
                      BNReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.BNReLU2d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.BNReLU2d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
 
                      A BNReLU2d module is a fused module of BatchNorm2d and ReLU
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.BatchNorm2d.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU3d.html
index 9890d169e63..58d53fa0d69 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.BNReLU3d.html
@@ -598,7 +598,7 @@
 
                      BNReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.BNReLU3d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.BNReLU3d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
 
                      A BNReLU3d module is a fused module of BatchNorm3d and ReLU
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.BatchNorm3d.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU1d.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU1d.html
index e5964ad346b..72f44026737 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU1d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU1d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU1d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.ConvReLU1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.ConvReLU1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      A ConvReLU1d module is a fused module of Conv1d and ReLU
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.Conv1d.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU2d.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU2d.html
index ced962c1bed..7e5e3ab9661 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU2d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU2d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU2d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.ConvReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.ConvReLU2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      A ConvReLU2d module is a fused module of Conv2d and ReLU
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.Conv2d.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU3d.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU3d.html
index 83ddf681025..95e5194d9cc 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU3d.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.ConvReLU3d.html
@@ -598,7 +598,7 @@
 
                      ConvReLU3d
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.ConvReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.ConvReLU3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      A ConvReLU3d module is a fused module of Conv3d and ReLU
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.Conv3d.
                      
 
                      Attributes: Same as torch.ao.nn.quantized.Conv3d
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.LinearReLU.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.LinearReLU.html
index e3f788091b7..b898b2e5ea1 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.LinearReLU.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.LinearReLU.html
@@ -598,7 +598,7 @@
 
                      LinearReLU
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.LinearReLU(in_features, out_features, bias=True, dtype=torch.qint8)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.LinearReLU(in_features, out_features, bias=True, dtype=torch.qint8)[source][source]
 
                      A LinearReLU module fused from Linear and ReLU modules
                      
 
                      We adopt the same interface as torch.ao.nn.quantized.Linear.
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU.html b/2.7/generated/torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU.html
index 01f29792492..e78caa0e763 100644
--- a/2.7/generated/torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU.html
+++ b/2.7/generated/torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU.html
@@ -598,7 +598,7 @@
 
                      LinearReLU
                      
 
                      
 
                      
-class torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU(in_features, out_features, bias=True, dtype=torch.qint8)[source][source]
                      
+class torch.ao.nn.intrinsic.quantized.dynamic.LinearReLU(in_features, out_features, bias=True, dtype=torch.qint8)[source][source]
 
                      A LinearReLU module fused from Linear and ReLU modules that can be used
 for dynamic quantization.
 Supports both, FP16 and INT8 quantization.
                      
diff --git a/2.7/generated/torch.ao.nn.qat.Conv2d.html b/2.7/generated/torch.ao.nn.qat.Conv2d.html
index 4a5b146071c..5493ecedef9 100644
--- a/2.7/generated/torch.ao.nn.qat.Conv2d.html
+++ b/2.7/generated/torch.ao.nn.qat.Conv2d.html
@@ -598,7 +598,7 @@
 
                      Conv2d
                      
 
                      
 
                      
-class torch.ao.nn.qat.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.qat.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None, device=None, dtype=None)[source][source]
 
                      A Conv2d module attached with FakeQuantize modules for weight,
 used for quantization aware training.
                      
 
                      We adopt the same interface as torch.nn.Conv2d, please see
diff --git a/2.7/generated/torch.ao.nn.qat.Conv3d.html b/2.7/generated/torch.ao.nn.qat.Conv3d.html
index e228800aaa2..f55af9b38b5 100644
--- a/2.7/generated/torch.ao.nn.qat.Conv3d.html
+++ b/2.7/generated/torch.ao.nn.qat.Conv3d.html
@@ -598,7 +598,7 @@
 
                      Conv3d
                      
 
                      
 
                      
-class torch.ao.nn.qat.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.qat.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None, device=None, dtype=None)[source][source]
 
                      A Conv3d module attached with FakeQuantize modules for weight,
 used for quantization aware training.
                      
 
                      We adopt the same interface as torch.nn.Conv3d, please see
diff --git a/2.7/generated/torch.ao.nn.qat.Linear.html b/2.7/generated/torch.ao.nn.qat.Linear.html
index 22925c7ead3..9f0ce6442fe 100644
--- a/2.7/generated/torch.ao.nn.qat.Linear.html
+++ b/2.7/generated/torch.ao.nn.qat.Linear.html
@@ -598,7 +598,7 @@
 
                      Linear
                      
 
                      
 
                      
-class torch.ao.nn.qat.Linear(in_features, out_features, bias=True, qconfig=None, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.qat.Linear(in_features, out_features, bias=True, qconfig=None, device=None, dtype=None)[source][source]
 
                      A linear module attached with FakeQuantize modules for weight,
 used for quantization aware training.
                      
 
                      We adopt the same interface as torch.nn.Linear, please see
@@ -613,7 +613,7 @@ 
                      Linear
 
                      
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Create a qat module from a float module or qparams_dict
 Args: mod a float module, either produced by torch.ao.quantization utilities
 or directly from user
                      
diff --git a/2.7/generated/torch.ao.nn.qat.dynamic.Linear.html b/2.7/generated/torch.ao.nn.qat.dynamic.Linear.html
index 55c00f4eed6..8ce443cfd81 100644
--- a/2.7/generated/torch.ao.nn.qat.dynamic.Linear.html
+++ b/2.7/generated/torch.ao.nn.qat.dynamic.Linear.html
@@ -598,7 +598,7 @@
 
                      Linear
                      
 
                      
 
                      
-class torch.ao.nn.qat.dynamic.Linear(in_features, out_features, bias=True, qconfig=None, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.qat.dynamic.Linear(in_features, out_features, bias=True, qconfig=None, device=None, dtype=None)[source][source]
 
                      A linear module attached with FakeQuantize modules for weight,
 used for dynamic quantization aware training.
                      
 
                      We adopt the same interface as torch.nn.Linear, please see
diff --git a/2.7/generated/torch.ao.nn.quantizable.LSTM.html b/2.7/generated/torch.ao.nn.quantizable.LSTM.html
index 678c7bf6b6a..5e0f7fc595e 100644
--- a/2.7/generated/torch.ao.nn.quantizable.LSTM.html
+++ b/2.7/generated/torch.ao.nn.quantizable.LSTM.html
@@ -598,7 +598,7 @@
 
                      LSTM
                      
 
                      
 
                      
-class torch.ao.nn.quantizable.LSTM(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None, *, split_gates=False)[source][source]
                      
+class torch.ao.nn.quantizable.LSTM(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None, *, split_gates=False)[source][source]
 
                      A quantizable long short-term memory (LSTM).
                      
 
                      For the description and the argument types, please, refer to LSTM
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.quantizable.MultiheadAttention.html b/2.7/generated/torch.ao.nn.quantizable.MultiheadAttention.html
index 994db80fb00..135e18f9a7f 100644
--- a/2.7/generated/torch.ao.nn.quantizable.MultiheadAttention.html
+++ b/2.7/generated/torch.ao.nn.quantizable.MultiheadAttention.html
@@ -598,12 +598,12 @@
 
                      MultiheadAttention
                      
 
                      
 
                      
-class torch.ao.nn.quantizable.MultiheadAttention(embed_dim, num_heads, dropout=0.0, bias=True, add_bias_kv=False, add_zero_attn=False, kdim=None, vdim=None, batch_first=False, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantizable.MultiheadAttention(embed_dim, num_heads, dropout=0.0, bias=True, add_bias_kv=False, add_zero_attn=False, kdim=None, vdim=None, batch_first=False, device=None, dtype=None)[source][source]
 
                      
 
                      
 
                      
 
                      
-dequantize()[source][source]
                      
+dequantize()[source][source]
 
                      Utility to convert the quantized MHA back to float.
                      
 
                      The motivation for this is that it is not trivial to convert the weights
 from the format that is used in the quantized version back to the
@@ -612,7 +612,7 @@ 
                      MultiheadAttention
 
                      
-forward(query, key, value, key_padding_mask=None, need_weights=True, attn_mask=None, average_attn_weights=True, is_causal=False)[source][source]
                      
+forward(query, key, value, key_padding_mask=None, need_weights=True, attn_mask=None, average_attn_weights=True, is_causal=False)[source][source]
 
                      
 
                      Note::
                      Please, refer to forward() for more
 information
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.BatchNorm2d.html b/2.7/generated/torch.ao.nn.quantized.BatchNorm2d.html
index 33add3e9798..dafb3b960e2 100644
--- a/2.7/generated/torch.ao.nn.quantized.BatchNorm2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.BatchNorm2d.html
@@ -598,7 +598,7 @@
 
                      BatchNorm2d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.BatchNorm2d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.BatchNorm2d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
 
                      This is the quantized version of BatchNorm2d.
                      
 
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.BatchNorm3d.html b/2.7/generated/torch.ao.nn.quantized.BatchNorm3d.html
index b835a2f0c87..8b7ea6c9fa4 100644
--- a/2.7/generated/torch.ao.nn.quantized.BatchNorm3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.BatchNorm3d.html
@@ -598,7 +598,7 @@
 
                      BatchNorm3d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.BatchNorm3d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.BatchNorm3d(num_features, eps=1e-05, momentum=0.1, device=None, dtype=None)[source][source]
 
                      This is the quantized version of BatchNorm3d.
                      
 
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.Conv1d.html b/2.7/generated/torch.ao.nn.quantized.Conv1d.html
index 22d6d2ec788..5c4d678efab 100644
--- a/2.7/generated/torch.ao.nn.quantized.Conv1d.html
+++ b/2.7/generated/torch.ao.nn.quantized.Conv1d.html
@@ -598,7 +598,7 @@
 
                      Conv1d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.Conv1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.Conv1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 1D convolution over a quantized input signal composed of
 several quantized input planes.
                      
 
                      For details on input arguments, parameters, and implementation see
@@ -633,7 +633,7 @@ 
                      Conv1d
 
                      
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Creates a quantized module from a float module or qparams_dict.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.Conv2d.html b/2.7/generated/torch.ao.nn.quantized.Conv2d.html
index 7a737e1451c..368673844d3 100644
--- a/2.7/generated/torch.ao.nn.quantized.Conv2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.Conv2d.html
@@ -598,7 +598,7 @@
 
                      Conv2d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 2D convolution over a quantized input signal composed of
 several quantized input planes.
                      
 
                      For details on input arguments, parameters, and implementation see
@@ -637,7 +637,7 @@ 
                      Conv2d
 
                      
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Creates a quantized module from a float module or qparams_dict.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.Conv3d.html b/2.7/generated/torch.ao.nn.quantized.Conv3d.html
index ee4a0dced45..e5f51ebd4d1 100644
--- a/2.7/generated/torch.ao.nn.quantized.Conv3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.Conv3d.html
@@ -598,7 +598,7 @@
 
                      Conv3d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 3D convolution over a quantized input signal composed of
 several quantized input planes.
                      
 
                      For details on input arguments, parameters, and implementation see
@@ -637,7 +637,7 @@ 
                      Conv3d
 
                      
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Creates a quantized module from a float module or qparams_dict.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.ConvTranspose1d.html b/2.7/generated/torch.ao.nn.quantized.ConvTranspose1d.html
index dcdc14eda72..32d80ee9a30 100644
--- a/2.7/generated/torch.ao.nn.quantized.ConvTranspose1d.html
+++ b/2.7/generated/torch.ao.nn.quantized.ConvTranspose1d.html
@@ -598,7 +598,7 @@
 
                      ConvTranspose1d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.ConvTranspose1d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.ConvTranspose1d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 1D transposed convolution operator over an input image
 composed of several input planes.
 For details on input arguments, parameters, and implementation see
diff --git a/2.7/generated/torch.ao.nn.quantized.ConvTranspose2d.html b/2.7/generated/torch.ao.nn.quantized.ConvTranspose2d.html
index 27211690e13..a53fae32cf9 100644
--- a/2.7/generated/torch.ao.nn.quantized.ConvTranspose2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.ConvTranspose2d.html
@@ -598,7 +598,7 @@
 
                      ConvTranspose2d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.ConvTranspose2d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.ConvTranspose2d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 2D transposed convolution operator over an input image
 composed of several input planes.
 For details on input arguments, parameters, and implementation see
diff --git a/2.7/generated/torch.ao.nn.quantized.ConvTranspose3d.html b/2.7/generated/torch.ao.nn.quantized.ConvTranspose3d.html
index 37c093e1af7..a5526436ecf 100644
--- a/2.7/generated/torch.ao.nn.quantized.ConvTranspose3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.ConvTranspose3d.html
@@ -598,7 +598,7 @@
 
                      ConvTranspose3d
                      
 
                      
 
                      
-class torch.ao.nn.quantized.ConvTranspose3d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.ConvTranspose3d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                      Applies a 3D transposed convolution operator over an input image
 composed of several input planes.
 For details on input arguments, parameters, and implementation see
diff --git a/2.7/generated/torch.ao.nn.quantized.ELU.html b/2.7/generated/torch.ao.nn.quantized.ELU.html
index f4216ef9957..d40f4129163 100644
--- a/2.7/generated/torch.ao.nn.quantized.ELU.html
+++ b/2.7/generated/torch.ao.nn.quantized.ELU.html
@@ -598,7 +598,7 @@
 
                      ELU
                      
 
                      
 
                      
-class torch.ao.nn.quantized.ELU(scale, zero_point, alpha=1.0)[source][source]
                      
+class torch.ao.nn.quantized.ELU(scale, zero_point, alpha=1.0)[source][source]
 
                      This is the quantized equivalent of ELU.
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.Embedding.html b/2.7/generated/torch.ao.nn.quantized.Embedding.html
index 1107edbdf97..9c7f589a3bb 100644
--- a/2.7/generated/torch.ao.nn.quantized.Embedding.html
+++ b/2.7/generated/torch.ao.nn.quantized.Embedding.html
@@ -598,7 +598,7 @@
 
                      Embedding
                      
 
                      
 
                      
-class torch.ao.nn.quantized.Embedding(num_embeddings, embedding_dim, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False, _weight=None, dtype=torch.quint8)[source][source]
                      
+class torch.ao.nn.quantized.Embedding(num_embeddings, embedding_dim, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False, _weight=None, dtype=torch.quint8)[source][source]
 
                      A quantized Embedding module with quantized packed weights as inputs.
 We adopt the same interface as torch.nn.Embedding, please see
 https://pytorch.org/docs/stable/generated/torch.nn.Embedding.html for documentation.
                      
@@ -622,7 +622,7 @@ 
                      Embedding
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Create a quantized embedding module from a float module
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.EmbeddingBag.html b/2.7/generated/torch.ao.nn.quantized.EmbeddingBag.html
index 2e7171ed687..1c86ec2a9cf 100644
--- a/2.7/generated/torch.ao.nn.quantized.EmbeddingBag.html
+++ b/2.7/generated/torch.ao.nn.quantized.EmbeddingBag.html
@@ -598,7 +598,7 @@
 
                      EmbeddingBag
                      
 
                      
 
                      
-class torch.ao.nn.quantized.EmbeddingBag(num_embeddings, embedding_dim, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='sum', sparse=False, _weight=None, include_last_offset=False, dtype=torch.quint8)[source][source]
                      
+class torch.ao.nn.quantized.EmbeddingBag(num_embeddings, embedding_dim, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='sum', sparse=False, _weight=None, include_last_offset=False, dtype=torch.quint8)[source][source]
 
                      A quantized EmbeddingBag module with quantized packed weights as inputs.
 We adopt the same interface as torch.nn.EmbeddingBag, please see
 https://pytorch.org/docs/stable/generated/torch.nn.EmbeddingBag.html for documentation.
                      
@@ -623,7 +623,7 @@ 
                      EmbeddingBag
 
                      
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                      
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                      Create a quantized embedding_bag module from a float module
                      
 
                      
 
                      Parameters
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.FXFloatFunctional.html b/2.7/generated/torch.ao.nn.quantized.FXFloatFunctional.html
index 4c85f5d1738..84feb0865f9 100644
--- a/2.7/generated/torch.ao.nn.quantized.FXFloatFunctional.html
+++ b/2.7/generated/torch.ao.nn.quantized.FXFloatFunctional.html
@@ -598,7 +598,7 @@
 
                      FXFloatFunctional
                      
 
                      
 
                      
-class torch.ao.nn.quantized.FXFloatFunctional(*args, **kwargs)[source][source]
                      
+class torch.ao.nn.quantized.FXFloatFunctional(*args, **kwargs)[source][source]
 
                      module to replace FloatFunctional module before FX graph mode quantization,
 since activation_post_process will be inserted in top level module directly
                      
 
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.FloatFunctional.html b/2.7/generated/torch.ao.nn.quantized.FloatFunctional.html
index 3b3767f9adb..d3ed26e835a 100644
--- a/2.7/generated/torch.ao.nn.quantized.FloatFunctional.html
+++ b/2.7/generated/torch.ao.nn.quantized.FloatFunctional.html
@@ -598,7 +598,7 @@
 
                      FloatFunctional
                      
 
                      
 
                      
-class torch.ao.nn.quantized.FloatFunctional[source][source]
                      
+class torch.ao.nn.quantized.FloatFunctional[source][source]
 
                      State collector class for float operations.
                      
 
                      The instance of this class can be used instead of the torch. prefix for
 some operations. See example usage below.
                      
diff --git a/2.7/generated/torch.ao.nn.quantized.GroupNorm.html b/2.7/generated/torch.ao.nn.quantized.GroupNorm.html
index cc9359987a1..617fd7ef4f6 100644
--- a/2.7/generated/torch.ao.nn.quantized.GroupNorm.html
+++ b/2.7/generated/torch.ao.nn.quantized.GroupNorm.html
@@ -598,7 +598,7 @@
 
                      GroupNorm
                      
 
                      
 
                      
-class torch.ao.nn.quantized.GroupNorm(num_groups, num_channels, weight, bias, scale, zero_point, eps=1e-05, affine=True, device=None, dtype=None)[source][source]
                      
+class torch.ao.nn.quantized.GroupNorm(num_groups, num_channels, weight, bias, scale, zero_point, eps=1e-05, affine=True, device=None, dtype=None)[source][source]
 
                      This is the quantized version of GroupNorm.
                      
 
                      
 
                      Additional args:
                        
diff --git a/2.7/generated/torch.ao.nn.quantized.Hardswish.html b/2.7/generated/torch.ao.nn.quantized.Hardswish.html
index 12bf9b14c26..594df518585 100644
--- a/2.7/generated/torch.ao.nn.quantized.Hardswish.html
+++ b/2.7/generated/torch.ao.nn.quantized.Hardswish.html
@@ -598,7 +598,7 @@
 
                        Hardswish
                        
 
                        
 
                        
-class torch.ao.nn.quantized.Hardswish(scale, zero_point, device=None, dtype=None)[source][source]
                        
+class torch.ao.nn.quantized.Hardswish(scale, zero_point, device=None, dtype=None)[source][source]
 
                        This is the quantized version of Hardswish.
                        
 
                        
 
                        Parameters
                        
diff --git a/2.7/generated/torch.ao.nn.quantized.InstanceNorm1d.html b/2.7/generated/torch.ao.nn.quantized.InstanceNorm1d.html
index bc46fbb35e4..59ba684c550 100644
--- a/2.7/generated/torch.ao.nn.quantized.InstanceNorm1d.html
+++ b/2.7/generated/torch.ao.nn.quantized.InstanceNorm1d.html
@@ -598,7 +598,7 @@
 
                        InstanceNorm1d
                        
 
                        
 
                        
-class torch.ao.nn.quantized.InstanceNorm1d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                        
+class torch.ao.nn.quantized.InstanceNorm1d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                        This is the quantized version of InstanceNorm1d.
                        
 
                        
 
                        Additional args:
                          
diff --git a/2.7/generated/torch.ao.nn.quantized.InstanceNorm2d.html b/2.7/generated/torch.ao.nn.quantized.InstanceNorm2d.html
index 0d1cdc4cfc6..de7e1eee256 100644
--- a/2.7/generated/torch.ao.nn.quantized.InstanceNorm2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.InstanceNorm2d.html
@@ -598,7 +598,7 @@
 
                          InstanceNorm2d
                          
 
                          
 
                          
-class torch.ao.nn.quantized.InstanceNorm2d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                          
+class torch.ao.nn.quantized.InstanceNorm2d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                          This is the quantized version of InstanceNorm2d.
                          
 
                          
 
                          Additional args:
                            
diff --git a/2.7/generated/torch.ao.nn.quantized.InstanceNorm3d.html b/2.7/generated/torch.ao.nn.quantized.InstanceNorm3d.html
index f6c83504efa..948a3cc7b5d 100644
--- a/2.7/generated/torch.ao.nn.quantized.InstanceNorm3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.InstanceNorm3d.html
@@ -598,7 +598,7 @@
 
                            InstanceNorm3d
                            
 
                            
 
                            
-class torch.ao.nn.quantized.InstanceNorm3d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                            
+class torch.ao.nn.quantized.InstanceNorm3d(num_features, weight, bias, scale, zero_point, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                            This is the quantized version of InstanceNorm3d.
                            
 
                            
 
                            Additional args:
                              
diff --git a/2.7/generated/torch.ao.nn.quantized.LayerNorm.html b/2.7/generated/torch.ao.nn.quantized.LayerNorm.html
index 56a27be2783..2e2a0eaf450 100644
--- a/2.7/generated/torch.ao.nn.quantized.LayerNorm.html
+++ b/2.7/generated/torch.ao.nn.quantized.LayerNorm.html
@@ -598,7 +598,7 @@
 
                              LayerNorm
                              
 
                              
 
                              
-class torch.ao.nn.quantized.LayerNorm(normalized_shape, weight, bias, scale, zero_point, eps=1e-05, elementwise_affine=True, device=None, dtype=None)[source][source]
                              
+class torch.ao.nn.quantized.LayerNorm(normalized_shape, weight, bias, scale, zero_point, eps=1e-05, elementwise_affine=True, device=None, dtype=None)[source][source]
 
                              This is the quantized version of LayerNorm.
                              
 
                              
 
                              Additional args:
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.LeakyReLU.html b/2.7/generated/torch.ao.nn.quantized.LeakyReLU.html
index 5aa5e5e6d5f..a11ffb10e22 100644
--- a/2.7/generated/torch.ao.nn.quantized.LeakyReLU.html
+++ b/2.7/generated/torch.ao.nn.quantized.LeakyReLU.html
@@ -598,7 +598,7 @@
 
                                LeakyReLU
                                
 
                                
 
                                
-class torch.ao.nn.quantized.LeakyReLU(scale, zero_point, negative_slope=0.01, inplace=False, device=None, dtype=None)[source][source]
                                
+class torch.ao.nn.quantized.LeakyReLU(scale, zero_point, negative_slope=0.01, inplace=False, device=None, dtype=None)[source][source]
 
                                This is the quantized equivalent of LeakyReLU.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.Linear.html b/2.7/generated/torch.ao.nn.quantized.Linear.html
index 5a7cd5181b0..e4b1da649f0 100644
--- a/2.7/generated/torch.ao.nn.quantized.Linear.html
+++ b/2.7/generated/torch.ao.nn.quantized.Linear.html
@@ -598,7 +598,7 @@
 
                                Linear
                                
 
                                
 
                                
-class torch.ao.nn.quantized.Linear(in_features, out_features, bias_=True, dtype=torch.qint8)[source][source]
                                
+class torch.ao.nn.quantized.Linear(in_features, out_features, bias_=True, dtype=torch.qint8)[source][source]
 
                                A quantized linear module with quantized tensor as inputs and outputs.
 We adopt the same interface as torch.nn.Linear, please see
 https://pytorch.org/docs/stable/nn.html#torch.nn.Linear for documentation.
                                
@@ -627,7 +627,7 @@ 
                                Linear
 
                                
 
                                
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                                
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                                Create a quantized module from an observed float module
                                
 
                                
 
                                Parameters
                                
@@ -643,7 +643,7 @@ 
                                Linear
 
                                
-classmethod from_reference(ref_qlinear, output_scale, output_zero_point)[source][source]
                                
+classmethod from_reference(ref_qlinear, output_scale, output_zero_point)[source][source]
 
                                Create a (fbgemm/qnnpack) quantized module from a reference quantized module
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.QFunctional.html b/2.7/generated/torch.ao.nn.quantized.QFunctional.html
index 8d7f4d85ff9..401f636fdb1 100644
--- a/2.7/generated/torch.ao.nn.quantized.QFunctional.html
+++ b/2.7/generated/torch.ao.nn.quantized.QFunctional.html
@@ -598,7 +598,7 @@
 
                                QFunctional
                                
 
                                
 
                                
-class torch.ao.nn.quantized.QFunctional[source][source]
                                
+class torch.ao.nn.quantized.QFunctional[source][source]
 
                                Wrapper class for quantized operations.
                                
 
                                The instance of this class can be used instead of the
 torch.ops.quantized prefix. See example usage below.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.ReLU6.html b/2.7/generated/torch.ao.nn.quantized.ReLU6.html
index 7cb5aa5f9b4..fe7ab07ce3a 100644
--- a/2.7/generated/torch.ao.nn.quantized.ReLU6.html
+++ b/2.7/generated/torch.ao.nn.quantized.ReLU6.html
@@ -598,7 +598,7 @@
 
                                ReLU6
                                
 
                                
 
                                
-class torch.ao.nn.quantized.ReLU6(inplace=False)[source][source]
                                
+class torch.ao.nn.quantized.ReLU6(inplace=False)[source][source]
 
                                Applies the element-wise function:
                                
 
                                ReLU6(x)=min⁡(max⁡(x0,x),q(6))\text{ReLU6}(x) = \min(\max(x_0, x), q(6)), where x0x_0 is the
 zero_point, and q(6)q(6) is the quantized representation of number 6.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.Sigmoid.html b/2.7/generated/torch.ao.nn.quantized.Sigmoid.html
index a0bb337e932..c36bab412c8 100644
--- a/2.7/generated/torch.ao.nn.quantized.Sigmoid.html
+++ b/2.7/generated/torch.ao.nn.quantized.Sigmoid.html
@@ -598,7 +598,7 @@
 
                                Sigmoid
                                
 
                                
 
                                
-class torch.ao.nn.quantized.Sigmoid(output_scale, output_zero_point)[source][source]
                                
+class torch.ao.nn.quantized.Sigmoid(output_scale, output_zero_point)[source][source]
 
                                This is the quantized equivalent of Sigmoid.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.GRU.html b/2.7/generated/torch.ao.nn.quantized.dynamic.GRU.html
index 7a30eaa96b4..41714ca142e 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.GRU.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.GRU.html
@@ -598,7 +598,7 @@
 
                                GRU
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.GRU(*args, **kwargs)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.GRU(*args, **kwargs)[source][source]
 
                                Applies a multi-layer gated recurrent unit (GRU) RNN to an input sequence.
                                
 
                                For each element in the input sequence, each layer computes the following
 function:
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.GRUCell.html b/2.7/generated/torch.ao.nn.quantized.dynamic.GRUCell.html
index bfaedfd2ec0..3301c24a4f7 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.GRUCell.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.GRUCell.html
@@ -598,7 +598,7 @@
 
                                GRUCell
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.GRUCell(input_size, hidden_size, bias=True, dtype=torch.qint8)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.GRUCell(input_size, hidden_size, bias=True, dtype=torch.qint8)[source][source]
 
                                A gated recurrent unit (GRU) cell
                                
 
                                A dynamic quantized GRUCell module with floating point tensor as inputs and outputs.
 Weights are quantized to 8 bits. We adopt the same interface as torch.nn.GRUCell,
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.LSTM.html b/2.7/generated/torch.ao.nn.quantized.dynamic.LSTM.html
index 1292886c67b..8613d3865e3 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.LSTM.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.LSTM.html
@@ -598,7 +598,7 @@
 
                                LSTM
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.LSTM(*args, **kwargs)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.LSTM(*args, **kwargs)[source][source]
 
                                A dynamic quantized LSTM module with floating point tensor as inputs and outputs.
 We adopt the same interface as torch.nn.LSTM, please see
 https://pytorch.org/docs/stable/nn.html#torch.nn.LSTM for documentation.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.LSTMCell.html b/2.7/generated/torch.ao.nn.quantized.dynamic.LSTMCell.html
index fff39444598..ef9acebb4bb 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.LSTMCell.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.LSTMCell.html
@@ -598,7 +598,7 @@
 
                                LSTMCell
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.LSTMCell(*args, **kwargs)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.LSTMCell(*args, **kwargs)[source][source]
 
                                A long short-term memory (LSTM) cell.
                                
 
                                A dynamic quantized LSTMCell module with floating point tensor as inputs and outputs.
 Weights are quantized to 8 bits. We adopt the same interface as torch.nn.LSTMCell,
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.Linear.html b/2.7/generated/torch.ao.nn.quantized.dynamic.Linear.html
index 3b85faaecd3..acb7ec8e7e2 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.Linear.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.Linear.html
@@ -598,7 +598,7 @@
 
                                Linear
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.Linear(in_features, out_features, bias_=True, dtype=torch.qint8)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.Linear(in_features, out_features, bias_=True, dtype=torch.qint8)[source][source]
 
                                A dynamic quantized linear module with floating point tensor as inputs and outputs.
 We adopt the same interface as torch.nn.Linear, please see
 https://pytorch.org/docs/stable/nn.html#torch.nn.Linear for documentation.
                                
@@ -625,7 +625,7 @@ 
                                Linear
 
                                
 
                                
-classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
                                
+classmethod from_float(mod, use_precomputed_fake_quant=False)[source][source]
 
                                Create a dynamic quantized module from a float module or qparams_dict
                                
 
                                
 
                                Parameters
                                
@@ -637,7 +637,7 @@ 
                                Linear
 
                                
-classmethod from_reference(ref_qlinear)[source][source]
                                
+classmethod from_reference(ref_qlinear)[source][source]
 
                                Create a (fbgemm/qnnpack) dynamic quantized module from a reference quantized
 module
 :param ref_qlinear: a reference quantized  module, either produced by
diff --git a/2.7/generated/torch.ao.nn.quantized.dynamic.RNNCell.html b/2.7/generated/torch.ao.nn.quantized.dynamic.RNNCell.html
index 5ce9943e5d9..e194d1eef01 100644
--- a/2.7/generated/torch.ao.nn.quantized.dynamic.RNNCell.html
+++ b/2.7/generated/torch.ao.nn.quantized.dynamic.RNNCell.html
@@ -598,7 +598,7 @@
 
                                RNNCell
                                
 
                                
 
                                
-class torch.ao.nn.quantized.dynamic.RNNCell(input_size, hidden_size, bias=True, nonlinearity='tanh', dtype=torch.qint8)[source][source]
                                
+class torch.ao.nn.quantized.dynamic.RNNCell(input_size, hidden_size, bias=True, nonlinearity='tanh', dtype=torch.qint8)[source][source]
 
                                An Elman RNN cell with tanh or ReLU non-linearity.
 A dynamic quantized RNNCell module with floating point tensor as inputs and outputs.
 Weights are quantized to 8 bits. We adopt the same interface as torch.nn.RNNCell,
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool2d.html b/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool2d.html
index 7ce2f294bc9..a4fd6b4263b 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool2d.html
@@ -598,7 +598,7 @@
 
                                adaptive_avg_pool2d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.adaptive_avg_pool2d(input, output_size)[source][source]
                                
+class torch.ao.nn.quantized.functional.adaptive_avg_pool2d(input, output_size)[source][source]
 
                                Applies a 2D adaptive average pooling over a quantized input signal composed
 of several quantized input planes.
                                
 
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool3d.html b/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool3d.html
index e2bdc28ac81..1ea2f7eefd5 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.adaptive_avg_pool3d.html
@@ -598,7 +598,7 @@
 
                                adaptive_avg_pool3d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.adaptive_avg_pool3d(input, output_size)[source][source]
                                
+class torch.ao.nn.quantized.functional.adaptive_avg_pool3d(input, output_size)[source][source]
 
                                Applies a 3D adaptive average pooling over a quantized input signal composed
 of several quantized input planes.
                                
 
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.avg_pool2d.html b/2.7/generated/torch.ao.nn.quantized.functional.avg_pool2d.html
index 8c321867d57..2244a2a7ed5 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.avg_pool2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.avg_pool2d.html
@@ -598,7 +598,7 @@
 
                                avg_pool2d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.avg_pool2d(input, kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.avg_pool2d(input, kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
 
                                Applies 2D average-pooling operation in kH×kWkH \times kW regions by step size
 sH×sWsH \times sW steps. The number of output features is equal to the number of
 input planes.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.avg_pool3d.html b/2.7/generated/torch.ao.nn.quantized.functional.avg_pool3d.html
index ed20a3fdd78..d29428b5df2 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.avg_pool3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.avg_pool3d.html
@@ -598,7 +598,7 @@
 
                                avg_pool3d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.avg_pool3d(input, kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.avg_pool3d(input, kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
 
                                Applies 3D average-pooling operation in kD timeskH×kWkD \ times kH \times kW regions by step size
 sD×sH×sWsD \times sH \times sW steps. The number of output features is equal to the number of
 input planes.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.celu.html b/2.7/generated/torch.ao.nn.quantized.functional.celu.html
index 2b0afed8652..6dce6d87f9e 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.celu.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.celu.html
@@ -598,7 +598,7 @@
 
                                celu
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.celu(input, scale, zero_point, alpha=1.)[source][source]
                                
+class torch.ao.nn.quantized.functional.celu(input, scale, zero_point, alpha=1.)[source][source]
 
                                Applies the quantized CELU function element-wise.
                                
 
                                
 CELU(x)=max⁡(0,x)+min⁡(0,α∗(exp⁡(x/α)−1))\text{CELU}(x) = \max(0,x) + \min(0, \alpha * (\exp(x / \alpha) - 1))
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.clamp.html b/2.7/generated/torch.ao.nn.quantized.functional.clamp.html
index 9ca3cd7c1f9..04c942d56aa 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.clamp.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.clamp.html
@@ -598,7 +598,7 @@
 
                                clamp
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.clamp(input, min_, max_)[source][source]
                                
+class torch.ao.nn.quantized.functional.clamp(input, min_, max_)[source][source]
 
                                float(input, min_, max_) -> Tensor
                                
 
                                Applies the clamp function element-wise.
 See clamp for more details.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.conv1d.html b/2.7/generated/torch.ao.nn.quantized.functional.conv1d.html
index b51ca20b386..9f641785cbf 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.conv1d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.conv1d.html
@@ -598,7 +598,7 @@
 
                                conv1d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.conv1d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
                                
+class torch.ao.nn.quantized.functional.conv1d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
 
                                Applies a 1D convolution over a quantized 1D input composed of several input
 planes.
                                
 
                                See Conv1d for details and output shape.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.conv2d.html b/2.7/generated/torch.ao.nn.quantized.functional.conv2d.html
index 0f75fec87c2..f5c3c8b7d62 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.conv2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.conv2d.html
@@ -598,7 +598,7 @@
 
                                conv2d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.conv2d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
                                
+class torch.ao.nn.quantized.functional.conv2d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
 
                                Applies a 2D convolution over a quantized 2D input composed of several input
 planes.
                                
 
                                See Conv2d for details and output shape.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.conv3d.html b/2.7/generated/torch.ao.nn.quantized.functional.conv3d.html
index d09860f1596..6fb1d562109 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.conv3d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.conv3d.html
@@ -598,7 +598,7 @@
 
                                conv3d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.conv3d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
                                
+class torch.ao.nn.quantized.functional.conv3d(input, weight, bias, stride=1, padding=0, dilation=1, groups=1, padding_mode='zeros', scale=1.0, zero_point=0, dtype=torch.quint8)[source][source]
 
                                Applies a 3D convolution over a quantized 3D input composed of several input
 planes.
                                
 
                                See Conv3d for details and output shape.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.elu.html b/2.7/generated/torch.ao.nn.quantized.functional.elu.html
index 540d23584d4..4fa6340c1cb 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.elu.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.elu.html
@@ -598,7 +598,7 @@
 
                                elu
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.elu(input, scale, zero_point, alpha=1.0)[source][source]
                                
+class torch.ao.nn.quantized.functional.elu(input, scale, zero_point, alpha=1.0)[source][source]
 
                                This is the quantized version of elu().
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.hardsigmoid.html b/2.7/generated/torch.ao.nn.quantized.functional.hardsigmoid.html
index ec840eb7620..c14d528e2f0 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.hardsigmoid.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.hardsigmoid.html
@@ -598,7 +598,7 @@
 
                                hardsigmoid
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.hardsigmoid(input, inplace=False)[source][source]
                                
+class torch.ao.nn.quantized.functional.hardsigmoid(input, inplace=False)[source][source]
 
                                This is the quantized version of hardsigmoid().
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.hardswish.html b/2.7/generated/torch.ao.nn.quantized.functional.hardswish.html
index 42093405cc4..561d7068835 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.hardswish.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.hardswish.html
@@ -598,7 +598,7 @@
 
                                hardswish
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.hardswish(input, scale, zero_point)[source][source]
                                
+class torch.ao.nn.quantized.functional.hardswish(input, scale, zero_point)[source][source]
 
                                This is the quantized version of hardswish().
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.hardtanh.html b/2.7/generated/torch.ao.nn.quantized.functional.hardtanh.html
index 52471fe3875..02195626ca5 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.hardtanh.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.hardtanh.html
@@ -598,7 +598,7 @@
 
                                hardtanh
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.hardtanh(input, min_val=-1.0, max_val=1.0, inplace=False)[source][source]
                                
+class torch.ao.nn.quantized.functional.hardtanh(input, min_val=-1.0, max_val=1.0, inplace=False)[source][source]
 
                                This is the quantized version of hardtanh().
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.interpolate.html b/2.7/generated/torch.ao.nn.quantized.functional.interpolate.html
index 3d788661213..5555d1a370a 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.interpolate.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.interpolate.html
@@ -598,7 +598,7 @@
 
                                interpolate
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.interpolate(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.interpolate(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
 
                                Down/up samples the input to either the given size or the given
 scale_factor
                                
 
                                See torch.nn.functional.interpolate() for implementation details.
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.leaky_relu.html b/2.7/generated/torch.ao.nn.quantized.functional.leaky_relu.html
index bddc0d14a93..a6fd1ec81c9 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.leaky_relu.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.leaky_relu.html
@@ -598,7 +598,7 @@
 
                                leaky_relu
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.leaky_relu(input, negative_slope=0.01, inplace=False, scale=None, zero_point=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.leaky_relu(input, negative_slope=0.01, inplace=False, scale=None, zero_point=None)[source][source]
 
                                Quantized version of the.
 leaky_relu(input, negative_slope=0.01, inplace=False, scale, zero_point) -> Tensor
                                
 
                                Applies element-wise,
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.linear.html b/2.7/generated/torch.ao.nn.quantized.functional.linear.html
index e379bab8984..9b8cd172e55 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.linear.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.linear.html
@@ -598,7 +598,7 @@
 
                                linear
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.linear(input, weight, bias=None, scale=None, zero_point=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.linear(input, weight, bias=None, scale=None, zero_point=None)[source][source]
 
                                Applies a linear transformation to the incoming quantized data:
 y=xAT+by = xA^T + b.
 See Linear
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.max_pool1d.html b/2.7/generated/torch.ao.nn.quantized.functional.max_pool1d.html
index 158c0d0981b..31d3a784222 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.max_pool1d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.max_pool1d.html
@@ -598,7 +598,7 @@
 
                                max_pool1d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.max_pool1d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source][source]
                                
+class torch.ao.nn.quantized.functional.max_pool1d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source][source]
 
                                Applies a 1D max pooling over a quantized input signal composed of
 several quantized input planes.
                                
 
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.max_pool2d.html b/2.7/generated/torch.ao.nn.quantized.functional.max_pool2d.html
index 5b7dc2cccec..319fb018c3e 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.max_pool2d.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.max_pool2d.html
@@ -598,7 +598,7 @@
 
                                max_pool2d
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.max_pool2d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source][source]
                                
+class torch.ao.nn.quantized.functional.max_pool2d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source][source]
 
                                Applies a 2D max pooling over a quantized input signal composed of
 several quantized input planes.
                                
 
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.threshold.html b/2.7/generated/torch.ao.nn.quantized.functional.threshold.html
index 9860c7effb3..fb4bd21eadf 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.threshold.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.threshold.html
@@ -598,7 +598,7 @@
 
                                threshold
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.threshold(input, threshold, value)[source][source]
                                
+class torch.ao.nn.quantized.functional.threshold(input, threshold, value)[source][source]
 
                                Applies the quantized version of the threshold function element-wise:
                                
 
                                
 x={xif x>thresholdvalueotherwisex = \begin{cases}
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.upsample.html b/2.7/generated/torch.ao.nn.quantized.functional.upsample.html
index 2f40bde6cd1..60f1dc6329d 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.upsample.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.upsample.html
@@ -598,7 +598,7 @@
 
                                upsample
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.upsample(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.upsample(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
 
                                Upsamples the input to either the given size or the given
 scale_factor
                                
 
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.upsample_bilinear.html b/2.7/generated/torch.ao.nn.quantized.functional.upsample_bilinear.html
index 2df4b327a74..c90ae9e6db7 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.upsample_bilinear.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.upsample_bilinear.html
@@ -598,7 +598,7 @@
 
                                upsample_bilinear
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.upsample_bilinear(input, size=None, scale_factor=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.upsample_bilinear(input, size=None, scale_factor=None)[source][source]
 
                                Upsamples the input, using bilinear upsampling.
                                
 
                                
 
                                Warning
                                
diff --git a/2.7/generated/torch.ao.nn.quantized.functional.upsample_nearest.html b/2.7/generated/torch.ao.nn.quantized.functional.upsample_nearest.html
index 65df1e085cb..db894ab35ef 100644
--- a/2.7/generated/torch.ao.nn.quantized.functional.upsample_nearest.html
+++ b/2.7/generated/torch.ao.nn.quantized.functional.upsample_nearest.html
@@ -598,7 +598,7 @@
 
                                upsample_nearest
                                
 
                                
 
                                
-class torch.ao.nn.quantized.functional.upsample_nearest(input, size=None, scale_factor=None)[source][source]
                                
+class torch.ao.nn.quantized.functional.upsample_nearest(input, size=None, scale_factor=None)[source][source]
 
                                Upsamples the input, using nearest neighbours’ pixel values.
                                
 
                                
 
                                Warning
                                
diff --git a/2.7/generated/torch.ao.quantization.DeQuantStub.html b/2.7/generated/torch.ao.quantization.DeQuantStub.html
index cad01f94dc1..3341c5420f5 100644
--- a/2.7/generated/torch.ao.quantization.DeQuantStub.html
+++ b/2.7/generated/torch.ao.quantization.DeQuantStub.html
@@ -598,7 +598,7 @@
 
                                DeQuantStub
                                
 
                                
 
                                
-class torch.ao.quantization.DeQuantStub(qconfig=None)[source][source]
                                
+class torch.ao.quantization.DeQuantStub(qconfig=None)[source][source]
 
                                Dequantize stub module, before calibration, this is same as identity,
 this will be swapped as nnq.DeQuantize in convert.
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.QuantStub.html b/2.7/generated/torch.ao.quantization.QuantStub.html
index bc655303665..c7f6cef4889 100644
--- a/2.7/generated/torch.ao.quantization.QuantStub.html
+++ b/2.7/generated/torch.ao.quantization.QuantStub.html
@@ -598,7 +598,7 @@
 
                                QuantStub
                                
 
                                
 
                                
-class torch.ao.quantization.QuantStub(qconfig=None)[source][source]
                                
+class torch.ao.quantization.QuantStub(qconfig=None)[source][source]
 
                                Quantize stub module, before calibration, this is same as an observer,
 it will be swapped as nnq.Quantize in convert.
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.QuantWrapper.html b/2.7/generated/torch.ao.quantization.QuantWrapper.html
index d882e61bc63..2d21252ac64 100644
--- a/2.7/generated/torch.ao.quantization.QuantWrapper.html
+++ b/2.7/generated/torch.ao.quantization.QuantWrapper.html
@@ -598,7 +598,7 @@
 
                                QuantWrapper
                                
 
                                
 
                                
-class torch.ao.quantization.QuantWrapper(module)[source][source]
                                
+class torch.ao.quantization.QuantWrapper(module)[source][source]
 
                                A wrapper class that wraps the input module, adds QuantStub and
 DeQuantStub and surround the call to module with call to quant and dequant
 modules.
                                
diff --git a/2.7/generated/torch.ao.quantization.add_quant_dequant.html b/2.7/generated/torch.ao.quantization.add_quant_dequant.html
index 38e45888b74..72ba12d2222 100644
--- a/2.7/generated/torch.ao.quantization.add_quant_dequant.html
+++ b/2.7/generated/torch.ao.quantization.add_quant_dequant.html
@@ -598,7 +598,7 @@
 
                                add_quant_dequant
                                
 
                                
 
                                
-class torch.ao.quantization.add_quant_dequant(module)[source][source]
                                
+class torch.ao.quantization.add_quant_dequant(module)[source][source]
 
                                Wrap the leaf child module in QuantWrapper if it has a valid qconfig
 Note that this function will modify the children of module inplace and it
 can return a new module which wraps the input module as well.
                                
diff --git a/2.7/generated/torch.ao.quantization.backend_config.BackendConfig.html b/2.7/generated/torch.ao.quantization.backend_config.BackendConfig.html
index 0daaef1d0a3..d1603308927 100644
--- a/2.7/generated/torch.ao.quantization.backend_config.BackendConfig.html
+++ b/2.7/generated/torch.ao.quantization.backend_config.BackendConfig.html
@@ -598,7 +598,7 @@
 
                                BackendConfig
                                
 
                                
 
                                
-class torch.ao.quantization.backend_config.BackendConfig(name='')[source][source]
                                
+class torch.ao.quantization.backend_config.BackendConfig(name='')[source][source]
 
                                Config that defines the set of patterns that can be quantized on a given backend, and how reference
 quantized models can be produced from these patterns.
                                
 
                                A pattern in this context refers to a module, a functional, an operator, or a directed acyclic graph
@@ -651,7 +651,7 @@ 
                                BackendConfig
 
                                
-classmethod from_dict(backend_config_dict)[source][source]
                                
+classmethod from_dict(backend_config_dict)[source][source]
 
                                Create a BackendConfig from a dictionary with the following items:
                                
 
                                
 
                                “name”: the name of the target backend
                                
@@ -666,7 +666,7 @@ 
                                BackendConfig
 
                                
-set_backend_pattern_config(config)[source][source]
                                
+set_backend_pattern_config(config)[source][source]
 
                                Set the config for an pattern that can be run on the target backend.
 This overrides any existing config for the given pattern.
                                
 
                                
@@ -678,7 +678,7 @@ 
                                BackendConfig
 
                                
-set_backend_pattern_configs(configs)[source][source]
                                
+set_backend_pattern_configs(configs)[source][source]
 
                                Set the configs for patterns that can be run on the target backend.
 This overrides any existing config for a given pattern if it was previously registered already.
                                
 
                                
@@ -690,7 +690,7 @@ 
                                BackendConfig
 
                                
-set_name(name)[source][source]
                                
+set_name(name)[source][source]
 
                                Set the name of the target backend.
                                
 
                                
 
                                Return type
                                
@@ -701,7 +701,7 @@ 
                                BackendConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this BackendConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.backend_config.BackendPatternConfig.html b/2.7/generated/torch.ao.quantization.backend_config.BackendPatternConfig.html
index 2e67746ed82..40d683f18ae 100644
--- a/2.7/generated/torch.ao.quantization.backend_config.BackendPatternConfig.html
+++ b/2.7/generated/torch.ao.quantization.backend_config.BackendPatternConfig.html
@@ -598,14 +598,14 @@
 
                                BackendPatternConfig
                                
 
                                
 
                                
-class torch.ao.quantization.backend_config.BackendPatternConfig(pattern=None)[source][source]
                                
+class torch.ao.quantization.backend_config.BackendPatternConfig(pattern=None)[source][source]
 
                                Config object that specifies quantization behavior for a given operator pattern.
 For a detailed example usage, see BackendConfig.
                                
 
                                
 
                                
 
                                
 
                                
-add_dtype_config(dtype_config)[source][source]
                                
+add_dtype_config(dtype_config)[source][source]
 
                                Add a set of supported data types passed as arguments to quantize ops in the
 reference model spec.
                                
 
                                
@@ -617,7 +617,7 @@ 
                                BackendPatternConfig
 
                                
-classmethod from_dict(backend_pattern_config_dict)[source][source]
                                
+classmethod from_dict(backend_pattern_config_dict)[source][source]
 
                                Create a BackendPatternConfig from a dictionary with the following items:
                                
 
                                
 
                                “pattern”: the pattern being configured
@@ -641,7 +641,7 @@ 
                                BackendPatternConfig
 
                                
-set_dtype_configs(dtype_configs)[source][source]
                                
+set_dtype_configs(dtype_configs)[source][source]
 
                                Set the supported data types passed as arguments to quantize ops in the
 reference model spec, overriding all previously registered data types.
                                
 
                                
@@ -653,7 +653,7 @@ 
                                BackendPatternConfig
 
                                
-set_fused_module(fused_module)[source][source]
                                
+set_fused_module(fused_module)[source][source]
 
                                Set the module that represents the fused implementation for this pattern.
                                
 
                                
 
                                Return type
                                
@@ -664,7 +664,7 @@ 
                                BackendPatternConfig
 
                                
-set_fuser_method(fuser_method)[source][source]
                                
+set_fuser_method(fuser_method)[source][source]
 
                                Set the function that specifies how to fuse this BackendPatternConfig’s pattern.
                                
 
                                The first argument of this function should be is_qat, and the rest of the arguments
 should be the items in the tuple pattern. The return value of this function should be
@@ -686,7 +686,7 @@ 
                                BackendPatternConfig
 
                                
-set_observation_type(observation_type)[source][source]
                                
+set_observation_type(observation_type)[source][source]
 
                                Set how observers should be inserted in the graph for this pattern.
                                
 
                                Observation type here refers to how observers (or quant-dequant ops) will be placed
 in the graph. This is used to produce the desired reference patterns understood by
@@ -711,7 +711,7 @@ 
                                BackendPatternConfig
 
                                
-set_pattern(pattern)[source][source]
                                
+set_pattern(pattern)[source][source]
 
                                Set the pattern to configure.
                                
 
                                The pattern can be a float module, functional operator, pytorch operator, or a tuple
 combination of the above. Tuple patterns are treated as sequential patterns, and
@@ -725,7 +725,7 @@ 
                                BackendPatternConfig
 
                                
-set_qat_module(qat_module)[source][source]
                                
+set_qat_module(qat_module)[source][source]
 
                                Set the module that represents the QAT implementation for this pattern.
                                
 
                                
 
                                Return type
                                
@@ -736,7 +736,7 @@ 
                                BackendPatternConfig
 
                                
-set_reference_quantized_module(reference_quantized_module)[source][source]
                                
+set_reference_quantized_module(reference_quantized_module)[source][source]
 
                                Set the module that represents the reference quantized implementation for
 this pattern’s root module.
                                
 
                                For more detail, see set_root_module().
                                
@@ -749,7 +749,7 @@ 
                                BackendPatternConfig
 
                                
-set_root_module(root_module)[source][source]
                                
+set_root_module(root_module)[source][source]
 
                                Set the module that represents the root for this pattern.
                                
 
                                When we construct the reference quantized model during the convert phase,
 the root modules (e.g. torch.nn.Linear for torch.ao.nn.intrinsic.LinearReLU)
@@ -768,7 +768,7 @@ 
                                BackendPatternConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this BackendPatternConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.backend_config.DTypeConfig.html b/2.7/generated/torch.ao.quantization.backend_config.DTypeConfig.html
index 3db3c509e30..1fd8dbfd3ac 100644
--- a/2.7/generated/torch.ao.quantization.backend_config.DTypeConfig.html
+++ b/2.7/generated/torch.ao.quantization.backend_config.DTypeConfig.html
@@ -598,7 +598,7 @@
 
                                DTypeConfig
                                
 
                                
 
                                
-class torch.ao.quantization.backend_config.DTypeConfig(input_dtype=None, output_dtype=None, weight_dtype=None, bias_dtype=None, is_dynamic=None)[source][source]
                                
+class torch.ao.quantization.backend_config.DTypeConfig(input_dtype=None, output_dtype=None, weight_dtype=None, bias_dtype=None, is_dynamic=None)[source][source]
 
                                Config object that specifies the supported data types passed as arguments to
 quantize ops in the reference model spec, for input and output activations,
 weights, and biases.
                                
@@ -664,7 +664,7 @@ 
                                DTypeConfig
 
                                
-classmethod from_dict(dtype_config_dict)[source][source]
                                
+classmethod from_dict(dtype_config_dict)[source][source]
 
                                
 
                                Create a DTypeConfig from a dictionary with the following items (all optional):
                                “input_dtype”: torch.dtype or DTypeWithConstraints
 “output_dtype”: torch.dtype or DTypeWithConstraints
@@ -682,7 +682,7 @@ 
                                DTypeConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this DTypeConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.backend_config.DTypeWithConstraints.html b/2.7/generated/torch.ao.quantization.backend_config.DTypeWithConstraints.html
index ad37ba3ce72..fe6157ac7b5 100644
--- a/2.7/generated/torch.ao.quantization.backend_config.DTypeWithConstraints.html
+++ b/2.7/generated/torch.ao.quantization.backend_config.DTypeWithConstraints.html
@@ -598,7 +598,7 @@
 
                                DTypeWithConstraints
                                
 
                                
 
                                
-class torch.ao.quantization.backend_config.DTypeWithConstraints(dtype=None, quant_min_lower_bound=None, quant_max_upper_bound=None, scale_min_lower_bound=None, scale_max_upper_bound=None, scale_exact_match=None, zero_point_exact_match=None)[source][source]
                                
+class torch.ao.quantization.backend_config.DTypeWithConstraints(dtype=None, quant_min_lower_bound=None, quant_max_upper_bound=None, scale_min_lower_bound=None, scale_max_upper_bound=None, scale_exact_match=None, zero_point_exact_match=None)[source][source]
 
                                Config for specifying additional constraints for a given dtype, such as quantization
 value ranges, scale value ranges, and fixed quantization params, to be used in
 DTypeConfig.
                                
diff --git a/2.7/generated/torch.ao.quantization.backend_config.ObservationType.html b/2.7/generated/torch.ao.quantization.backend_config.ObservationType.html
index 93c80a0b7b8..9773b0ecb7e 100644
--- a/2.7/generated/torch.ao.quantization.backend_config.ObservationType.html
+++ b/2.7/generated/torch.ao.quantization.backend_config.ObservationType.html
@@ -598,7 +598,7 @@
 
                                ObservationType
                                
 
                                
 
                                
-class torch.ao.quantization.backend_config.ObservationType(value)[source][source]
                                
+class torch.ao.quantization.backend_config.ObservationType(value)[source][source]
 
                                An enum that represents different ways of how an operator/operator pattern
 should be observed
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.compare_results.html b/2.7/generated/torch.ao.quantization.compare_results.html
index db0b23e7f68..e78078d2951 100644
--- a/2.7/generated/torch.ao.quantization.compare_results.html
+++ b/2.7/generated/torch.ao.quantization.compare_results.html
@@ -598,7 +598,7 @@
 
                                compare_results
                                
 
                                
 
                                
-class torch.ao.quantization.compare_results(ref_results, actual_results)[source]
                                
+class torch.ao.quantization.compare_results(ref_results, actual_results)[source]
 
                                Given two dict mapping from debug_handle_id (int) to list of tensors
 return a map from debug_handle_id to NodeAccuracySummary that contains
 comparison information like SQNR, MSE etc.
                                
diff --git a/2.7/generated/torch.ao.quantization.convert.html b/2.7/generated/torch.ao.quantization.convert.html
index c9994ce1bd8..51d1f353334 100644
--- a/2.7/generated/torch.ao.quantization.convert.html
+++ b/2.7/generated/torch.ao.quantization.convert.html
@@ -598,7 +598,7 @@
 
                                convert
                                
 
                                
 
                                
-class torch.ao.quantization.convert(module, mapping=None, inplace=False, remove_qconfig=True, is_reference=False, convert_custom_config_dict=None, use_precomputed_fake_quant=False)[source][source]
                                
+class torch.ao.quantization.convert(module, mapping=None, inplace=False, remove_qconfig=True, is_reference=False, convert_custom_config_dict=None, use_precomputed_fake_quant=False)[source][source]
 
                                Converts submodules in input module to a different module according to mapping
 by calling from_float method on the target module class. And remove qconfig at the
 end if remove_qconfig is set to True.
                                
diff --git a/2.7/generated/torch.ao.quantization.default_eval_fn.html b/2.7/generated/torch.ao.quantization.default_eval_fn.html
index 912c23a9294..2ba478a2675 100644
--- a/2.7/generated/torch.ao.quantization.default_eval_fn.html
+++ b/2.7/generated/torch.ao.quantization.default_eval_fn.html
@@ -598,7 +598,7 @@
 
                                default_eval_fn
                                
 
                                
 
                                
-class torch.ao.quantization.default_eval_fn(model, calib_data)[source][source]
                                
+class torch.ao.quantization.default_eval_fn(model, calib_data)[source][source]
 
                                Define the default evaluation function.
                                
 
                                Default evaluation function takes a torch.utils.data.Dataset or a list of
 input Tensors and run the model on the dataset
                                
diff --git a/2.7/generated/torch.ao.quantization.extract_results_from_loggers.html b/2.7/generated/torch.ao.quantization.extract_results_from_loggers.html
index fb115cf9522..e909348cced 100644
--- a/2.7/generated/torch.ao.quantization.extract_results_from_loggers.html
+++ b/2.7/generated/torch.ao.quantization.extract_results_from_loggers.html
@@ -598,7 +598,7 @@
 
                                extract_results_from_loggers
                                
 
                                
 
                                
-class torch.ao.quantization.extract_results_from_loggers(model)[source]
                                
+class torch.ao.quantization.extract_results_from_loggers(model)[source]
 
                                For a given model, extract the tensors stats and related information for each debug handle.
 The reason we have a list of object, instead of Tensor is because the output of node may not be
 a Tensor, it could be (nested) list, tuple or dict as well.
                                
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantize.html b/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantize.html
index d72e3db78a7..b8a0cad65e8 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantize.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantize.html
@@ -598,7 +598,7 @@
 
                                FakeQuantize
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.FakeQuantize(observer=<class 'torch.ao.quantization.observer.MovingAverageMinMaxObserver'>, quant_min=None, quant_max=None, is_dynamic=False, **observer_kwargs)[source][source]
                                
+class torch.ao.quantization.fake_quantize.FakeQuantize(observer=<class 'torch.ao.quantization.observer.MovingAverageMinMaxObserver'>, quant_min=None, quant_max=None, is_dynamic=False, **observer_kwargs)[source][source]
 
                                Simulate the quantize and dequantize operations in training time.
                                
 
                                The output of this module is given by:
                                
 
                                x_out = (
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantizeBase.html b/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantizeBase.html
index dcc25f98687..6baef7b7073 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantizeBase.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.FakeQuantizeBase.html
@@ -598,7 +598,7 @@
 
                                FakeQuantizeBase
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.FakeQuantizeBase[source][source]
                                
+class torch.ao.quantization.fake_quantize.FakeQuantizeBase[source][source]
 
                                Base fake quantize module.
                                
 
                                Base fake quantize module
 Any fake quantize implementation should derive from this class.
                                
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize.html b/2.7/generated/torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize.html
index c95dddf6b74..845b9290386 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize.html
@@ -598,14 +598,14 @@
 
                                FixedQParamsFakeQuantize
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize(observer)[source][source]
                                
+class torch.ao.quantization.fake_quantize.FixedQParamsFakeQuantize(observer)[source][source]
 
                                Simulate quantize and dequantize in training time.
                                
 
                                Simulate quantize and dequantize with fixed quantization
 parameters in training time. Only per tensor quantization
 is supported.
                                
 
                                
 
                                
-extra_repr()[source][source]
                                
+extra_repr()[source][source]
 
                                Define a string representation of the object’s attributes.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize.html b/2.7/generated/torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize.html
index 494c2ff37e0..6a2aacd8309 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize.html
@@ -598,7 +598,7 @@
 
                                FusedMovingAvgObsFakeQuantize
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize(observer=<class 'torch.ao.quantization.observer.MovingAverageMinMaxObserver'>, quant_min=0, quant_max=255, **observer_kwargs)[source][source]
                                
+class torch.ao.quantization.fake_quantize.FusedMovingAvgObsFakeQuantize(observer=<class 'torch.ao.quantization.observer.MovingAverageMinMaxObserver'>, quant_min=0, quant_max=255, **observer_kwargs)[source][source]
 
                                Define a fused module to observe the tensor.
                                
 
                                Fused module that is used to observe the input tensor (compute min/max), compute
 scale/zero_point and fake_quantize the tensor.
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.disable_fake_quant.html b/2.7/generated/torch.ao.quantization.fake_quantize.disable_fake_quant.html
index 671f59b85e8..1e0171b6117 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.disable_fake_quant.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.disable_fake_quant.html
@@ -598,7 +598,7 @@
 
                                disable_fake_quant
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.disable_fake_quant(mod)[source][source]
                                
+class torch.ao.quantization.fake_quantize.disable_fake_quant(mod)[source][source]
 
                                Disable fake quantization for the module.
                                
 
                                Disable fake quantization for this module, if applicable. Example usage:
                                
 
                                # model is any PyTorch model
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.disable_observer.html b/2.7/generated/torch.ao.quantization.fake_quantize.disable_observer.html
index 049060e22c8..6092dbc548f 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.disable_observer.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.disable_observer.html
@@ -598,7 +598,7 @@
 
                                disable_observer
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.disable_observer(mod)[source][source]
                                
+class torch.ao.quantization.fake_quantize.disable_observer(mod)[source][source]
 
                                Disable observation for this module.
                                
 
                                Disable observation for this module, if applicable. Example usage:
                                
 
                                # model is any PyTorch model
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.enable_fake_quant.html b/2.7/generated/torch.ao.quantization.fake_quantize.enable_fake_quant.html
index 48e84800ab4..5118e88d2d1 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.enable_fake_quant.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.enable_fake_quant.html
@@ -598,7 +598,7 @@
 
                                enable_fake_quant
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.enable_fake_quant(mod)[source][source]
                                
+class torch.ao.quantization.fake_quantize.enable_fake_quant(mod)[source][source]
 
                                Enable fake quantization for the module.
                                
 
                                Enable fake quantization for this module, if applicable. Example usage:
                                
 
                                # model is any PyTorch model
diff --git a/2.7/generated/torch.ao.quantization.fake_quantize.enable_observer.html b/2.7/generated/torch.ao.quantization.fake_quantize.enable_observer.html
index 76094ef947a..47df3aec9d4 100644
--- a/2.7/generated/torch.ao.quantization.fake_quantize.enable_observer.html
+++ b/2.7/generated/torch.ao.quantization.fake_quantize.enable_observer.html
@@ -598,7 +598,7 @@
 
                                enable_observer
                                
 
                                
 
                                
-class torch.ao.quantization.fake_quantize.enable_observer(mod)[source][source]
                                
+class torch.ao.quantization.fake_quantize.enable_observer(mod)[source][source]
 
                                Enable observation for this module.
                                
 
                                Enable observation for this module, if applicable. Example usage:
                                
 
                                # model is any PyTorch model
diff --git a/2.7/generated/torch.ao.quantization.fuse_modules.fuse_modules.html b/2.7/generated/torch.ao.quantization.fuse_modules.fuse_modules.html
index 8b3d472a6b1..4c1ba94b631 100644
--- a/2.7/generated/torch.ao.quantization.fuse_modules.fuse_modules.html
+++ b/2.7/generated/torch.ao.quantization.fuse_modules.fuse_modules.html
@@ -598,7 +598,7 @@
 
                                fuse_modules
                                
 
                                
 
                                
-class torch.ao.quantization.fuse_modules.fuse_modules(model, modules_to_fuse, inplace=False, fuser_func=<function fuse_known_modules>, fuse_custom_config_dict=None)[source][source]
                                
+class torch.ao.quantization.fuse_modules.fuse_modules(model, modules_to_fuse, inplace=False, fuser_func=<function fuse_known_modules>, fuse_custom_config_dict=None)[source][source]
 
                                Fuse a list of modules into a single module.
                                
 
                                Fuses only the following sequence of modules:
 conv, bn
diff --git a/2.7/generated/torch.ao.quantization.fx.custom_config.ConvertCustomConfig.html b/2.7/generated/torch.ao.quantization.fx.custom_config.ConvertCustomConfig.html
index 7066eae7f79..1484207446d 100644
--- a/2.7/generated/torch.ao.quantization.fx.custom_config.ConvertCustomConfig.html
+++ b/2.7/generated/torch.ao.quantization.fx.custom_config.ConvertCustomConfig.html
@@ -598,7 +598,7 @@
 
                                ConvertCustomConfig
                                
 
                                
 
                                
-class torch.ao.quantization.fx.custom_config.ConvertCustomConfig[source][source]
                                
+class torch.ao.quantization.fx.custom_config.ConvertCustomConfig[source][source]
 
                                Custom configuration for convert_fx().
                                
 
                                Example usage:
                                
 
                                convert_custom_config = ConvertCustomConfig()             .set_observed_to_quantized_mapping(ObservedCustomModule, QuantizedCustomModule)             .set_preserved_attributes(["attr1", "attr2"])
@@ -608,7 +608,7 @@ 
                                ConvertCustomConfig
 
                                
-classmethod from_dict(convert_custom_config_dict)[source][source]
                                
+classmethod from_dict(convert_custom_config_dict)[source][source]
 
                                Create a ConvertCustomConfig from a dictionary with the following items:
                                
 
                                
 
                                “observed_to_quantized_custom_module_class”: a nested dictionary mapping from quantization
@@ -630,7 +630,7 @@ 
                                ConvertCustomConfig
 
                                
-set_observed_to_quantized_mapping(observed_class, quantized_class, quant_type=QuantType.STATIC)[source][source]
                                
+set_observed_to_quantized_mapping(observed_class, quantized_class, quant_type=QuantType.STATIC)[source][source]
 
                                Set the mapping from a custom observed module class to a custom quantized module class.
                                
 
                                The quantized module class must have a from_observed class method that converts the observed module class
 to the quantized module class.
                                
@@ -643,7 +643,7 @@ 
                                ConvertCustomConfig
 
                                
-set_preserved_attributes(attributes)[source][source]
                                
+set_preserved_attributes(attributes)[source][source]
 
                                Set the names of the attributes that will persist in the graph module even if they are not used in
 the model’s forward method.
                                
 
                                
@@ -655,7 +655,7 @@ 
                                ConvertCustomConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this ConvertCustomConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.fx.custom_config.FuseCustomConfig.html b/2.7/generated/torch.ao.quantization.fx.custom_config.FuseCustomConfig.html
index db2fee47759..b52b0f20647 100644
--- a/2.7/generated/torch.ao.quantization.fx.custom_config.FuseCustomConfig.html
+++ b/2.7/generated/torch.ao.quantization.fx.custom_config.FuseCustomConfig.html
@@ -598,7 +598,7 @@
 
                                FuseCustomConfig
                                
 
                                
 
                                
-class torch.ao.quantization.fx.custom_config.FuseCustomConfig[source][source]
                                
+class torch.ao.quantization.fx.custom_config.FuseCustomConfig[source][source]
 
                                Custom configuration for fuse_fx().
                                
 
                                Example usage:
                                
 
                                fuse_custom_config = FuseCustomConfig().set_preserved_attributes(["attr1", "attr2"])
@@ -608,7 +608,7 @@ 
                                FuseCustomConfig
 
                                
-classmethod from_dict(fuse_custom_config_dict)[source][source]
                                
+classmethod from_dict(fuse_custom_config_dict)[source][source]
 
                                Create a ConvertCustomConfig from a dictionary with the following items:
                                
 
                                
 
                                “preserved_attributes”: a list of attributes that persist even if they are not used in forward
                                
@@ -623,7 +623,7 @@ 
                                FuseCustomConfig
 
                                
-set_preserved_attributes(attributes)[source][source]
                                
+set_preserved_attributes(attributes)[source][source]
 
                                Set the names of the attributes that will persist in the graph module even if they are not used in
 the model’s forward method.
                                
 
                                
@@ -635,7 +635,7 @@ 
                                FuseCustomConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this FuseCustomConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.fx.custom_config.PrepareCustomConfig.html b/2.7/generated/torch.ao.quantization.fx.custom_config.PrepareCustomConfig.html
index 102b19476ee..7d94ac7cd50 100644
--- a/2.7/generated/torch.ao.quantization.fx.custom_config.PrepareCustomConfig.html
+++ b/2.7/generated/torch.ao.quantization.fx.custom_config.PrepareCustomConfig.html
@@ -598,7 +598,7 @@
 
                                PrepareCustomConfig
                                
 
                                
 
                                
-class torch.ao.quantization.fx.custom_config.PrepareCustomConfig[source][source]
                                
+class torch.ao.quantization.fx.custom_config.PrepareCustomConfig[source][source]
 
                                Custom configuration for prepare_fx() and
 prepare_qat_fx().
                                
 
                                Example usage:
                                
@@ -609,7 +609,7 @@ 
                                PrepareCustomConfig
 
                                
-classmethod from_dict(prepare_custom_config_dict)[source][source]
                                
+classmethod from_dict(prepare_custom_config_dict)[source][source]
 
                                Create a PrepareCustomConfig from a dictionary with the following items:
                                
 
                                
 
                                “standalone_module_name”: a list of (module_name, qconfig_mapping, example_inputs,
@@ -635,7 +635,7 @@ 
                                PrepareCustomConfig
 
                                
-set_float_to_observed_mapping(float_class, observed_class, quant_type=QuantType.STATIC)[source][source]
                                
+set_float_to_observed_mapping(float_class, observed_class, quant_type=QuantType.STATIC)[source][source]
 
                                Set the mapping from a custom float module class to a custom observed module class.
                                
 
                                The observed module class must have a from_float class method that converts the float module class
 to the observed module class. This is currently only supported for static quantization.
                                
@@ -648,7 +648,7 @@ 
                                PrepareCustomConfig
 
                                
-set_input_quantized_indexes(indexes)[source][source]
                                
+set_input_quantized_indexes(indexes)[source][source]
 
                                Set the indexes of the inputs of the graph that should be quantized.
 Inputs are otherwise assumed to be in fp32 by default instead.
                                
 
                                
@@ -660,7 +660,7 @@ 
                                PrepareCustomConfig
 
                                
-set_non_traceable_module_classes(module_classes)[source][source]
                                
+set_non_traceable_module_classes(module_classes)[source][source]
 
                                Set the modules that are not symbolically traceable, identified by class.
                                
 
                                
 
                                Return type
                                
@@ -671,7 +671,7 @@ 
                                PrepareCustomConfig
 
                                
-set_non_traceable_module_names(module_names)[source][source]
                                
+set_non_traceable_module_names(module_names)[source][source]
 
                                Set the modules that are not symbolically traceable, identified by name.
                                
 
                                
 
                                Return type
                                
@@ -682,7 +682,7 @@ 
                                PrepareCustomConfig
 
                                
-set_output_quantized_indexes(indexes)[source][source]
                                
+set_output_quantized_indexes(indexes)[source][source]
 
                                Set the indexes of the outputs of the graph that should be quantized.
 Outputs are otherwise assumed to be in fp32 by default instead.
                                
 
                                
@@ -694,7 +694,7 @@ 
                                PrepareCustomConfig
 
                                
-set_preserved_attributes(attributes)[source][source]
                                
+set_preserved_attributes(attributes)[source][source]
 
                                Set the names of the attributes that will persist in the graph module even if they are not used in
 the model’s forward method.
                                
 
                                
@@ -706,7 +706,7 @@ 
                                PrepareCustomConfig
 
                                
-set_standalone_module_class(module_class, qconfig_mapping, example_inputs, prepare_custom_config, backend_config)[source][source]
                                
+set_standalone_module_class(module_class, qconfig_mapping, example_inputs, prepare_custom_config, backend_config)[source][source]
 
                                Set the configuration for running a standalone module identified by module_class.
                                
 
                                If qconfig_mapping is None, the parent qconfig_mapping will be used instead.
 If prepare_custom_config is None, an empty PrepareCustomConfig will be used.
@@ -720,7 +720,7 @@ 
                                PrepareCustomConfig
 
                                
-set_standalone_module_name(module_name, qconfig_mapping, example_inputs, prepare_custom_config, backend_config)[source][source]
                                
+set_standalone_module_name(module_name, qconfig_mapping, example_inputs, prepare_custom_config, backend_config)[source][source]
 
                                Set the configuration for running a standalone module identified by module_name.
                                
 
                                If qconfig_mapping is None, the parent qconfig_mapping will be used instead.
 If prepare_custom_config is None, an empty PrepareCustomConfig will be used.
@@ -734,7 +734,7 @@ 
                                PrepareCustomConfig
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this PrepareCustomConfig to a dictionary with the items described in
 from_dict().
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry.html b/2.7/generated/torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry.html
index f2135583405..7bae8c7f424 100644
--- a/2.7/generated/torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry.html
+++ b/2.7/generated/torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry.html
@@ -598,7 +598,7 @@
 
                                StandaloneModuleConfigEntry
                                
 
                                
 
                                
-class torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry(qconfig_mapping: 'Optional[QConfigMapping]', example_inputs: 'tuple[Any, ...]', prepare_custom_config: 'Optional[PrepareCustomConfig]', backend_config: 'Optional[BackendConfig]')[source][source]
                                
+class torch.ao.quantization.fx.custom_config.StandaloneModuleConfigEntry(qconfig_mapping: 'Optional[QConfigMapping]', example_inputs: 'tuple[Any, ...]', prepare_custom_config: 'Optional[PrepareCustomConfig]', backend_config: 'Optional[BackendConfig]')[source][source]
 
                                
 
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.generate_numeric_debug_handle.html b/2.7/generated/torch.ao.quantization.generate_numeric_debug_handle.html
index d5655bab46a..775dc60552a 100644
--- a/2.7/generated/torch.ao.quantization.generate_numeric_debug_handle.html
+++ b/2.7/generated/torch.ao.quantization.generate_numeric_debug_handle.html
@@ -598,7 +598,7 @@
 
                                generate_numeric_debug_handle
                                
 
                                
 
                                
-class torch.ao.quantization.generate_numeric_debug_handle(ep)[source]
                                
+class torch.ao.quantization.generate_numeric_debug_handle(ep)[source]
 
                                Attach numeric_debug_handle_id for all nodes in the graph module of the given
 ExportedProgram, like conv2d, squeeze, conv1d, etc, except for placeholder.
 Notice that nodes like getattr are out of scope since they are not in the graph.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.AffineQuantizedObserverBase.html b/2.7/generated/torch.ao.quantization.observer.AffineQuantizedObserverBase.html
index 15fd23a1ca8..448b5adb683 100644
--- a/2.7/generated/torch.ao.quantization.observer.AffineQuantizedObserverBase.html
+++ b/2.7/generated/torch.ao.quantization.observer.AffineQuantizedObserverBase.html
@@ -598,7 +598,7 @@
 
                                AffineQuantizedObserverBase
                                
 
                                
 
                                
-class torch.ao.quantization.observer.AffineQuantizedObserverBase(mapping_type, target_dtype, granularity, quant_min=None, quant_max=None, eps=None, scale_dtype=None, zero_point_dtype=None, preserve_zero=True, zero_point_domain=ZeroPointDomain.INT, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.AffineQuantizedObserverBase(mapping_type, target_dtype, granularity, quant_min=None, quant_max=None, eps=None, scale_dtype=None, zero_point_dtype=None, preserve_zero=True, zero_point_domain=ZeroPointDomain.INT, **kwargs)[source][source]
 
                                Observer module for affine quantization (https://github.com/pytorch/ao/tree/main/torchao/quantization#affine-quantization)
                                
 
                                
 
                                Parameters
                                
@@ -612,7 +612,7 @@ 
                                AffineQuantizedObserverBase
 
                                
-abstract calculate_qparams()[source][source]
                                
+abstract calculate_qparams()[source][source]
 
                                Calculate quantization parameter based on the stats attached to the observer module
 and returns a tuple of scale and zero_point Tensor
                                
 
                                
@@ -624,7 +624,7 @@ 
                                AffineQuantizedObserverBase
 
                                
-abstract forward(input)[source][source]
                                
+abstract forward(input)[source][source]
 
                                forward function should take the input tensor
 and updates internal stats and return the original input Tensor
                                
 
                                
@@ -636,7 +636,7 @@ 
                                AffineQuantizedObserverBase
 
                                
-classmethod with_args(**kwargs)[source]
                                
+classmethod with_args(**kwargs)[source]
 
                                Wrapper that allows creation of class factories.
                                
 
                                This can be useful when there is a need to create classes with the same
 constructor arguments, but different instances. Can be used in conjunction with
diff --git a/2.7/generated/torch.ao.quantization.observer.Granularity.html b/2.7/generated/torch.ao.quantization.observer.Granularity.html
index ce217c03930..d470ab78245 100644
--- a/2.7/generated/torch.ao.quantization.observer.Granularity.html
+++ b/2.7/generated/torch.ao.quantization.observer.Granularity.html
@@ -598,7 +598,7 @@
 
                                Granularity
                                
 
                                
 
                                
-class torch.ao.quantization.observer.Granularity[source][source]
                                
+class torch.ao.quantization.observer.Granularity[source][source]
 
                                Base class for representing the granularity of quantization.
                                
 
                                This class serves as a parent for specific granularity types used in
 quantization operations, such as per-tensor or per-axis quantization.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.HistogramObserver.html b/2.7/generated/torch.ao.quantization.observer.HistogramObserver.html
index 52e73a29263..45f4f94e872 100644
--- a/2.7/generated/torch.ao.quantization.observer.HistogramObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.HistogramObserver.html
@@ -598,7 +598,7 @@
 
                                HistogramObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.HistogramObserver(bins=2048, dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.HistogramObserver(bins=2048, dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
 
                                The module records the running histogram of tensor values along with
 min/max values. calculate_qparams will calculate scale and zero_point.
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.MappingType.html b/2.7/generated/torch.ao.quantization.observer.MappingType.html
index 299de09d27f..d8c9d65177a 100644
--- a/2.7/generated/torch.ao.quantization.observer.MappingType.html
+++ b/2.7/generated/torch.ao.quantization.observer.MappingType.html
@@ -598,7 +598,7 @@
 
                                MappingType
                                
 
                                
 
                                
-class torch.ao.quantization.observer.MappingType(value)[source][source]
                                
+class torch.ao.quantization.observer.MappingType(value)[source][source]
 
                                How floating point number is mapped to integer number
                                
 
                                symmetric mapping means floating point range is symmetrically mapped to integer range
 let’s say we have floating point range (-3.5, 10.2) and integer range (-8, 7) (int4)
diff --git a/2.7/generated/torch.ao.quantization.observer.MinMaxObserver.html b/2.7/generated/torch.ao.quantization.observer.MinMaxObserver.html
index 12af585810f..360453df9a1 100644
--- a/2.7/generated/torch.ao.quantization.observer.MinMaxObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.MinMaxObserver.html
@@ -598,7 +598,7 @@
 
                                MinMaxObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.MinMaxObserver(dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.MinMaxObserver(dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
 
                                Observer module for computing the quantization parameters based on the
 running min and max values.
                                
 
                                This observer uses the tensor min/max statistics to compute the quantization
@@ -658,19 +658,19 @@ 
                                MinMaxObserver
 
                                
-calculate_qparams()[source][source]
                                
+calculate_qparams()[source][source]
 
                                Calculates the quantization parameters.
                                
 
                                
 
 
                                
 
                                
-forward(x_orig)[source][source]
                                
+forward(x_orig)[source][source]
 
                                Records the running minimum and maximum of x.
                                
 
                                
 
 
                                
 
                                
-reset_min_max_vals()[source][source]
                                
+reset_min_max_vals()[source][source]
 
                                Resets the min/max values.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.observer.MovingAverageMinMaxObserver.html b/2.7/generated/torch.ao.quantization.observer.MovingAverageMinMaxObserver.html
index 4b42260fd3f..637d8326a6e 100644
--- a/2.7/generated/torch.ao.quantization.observer.MovingAverageMinMaxObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.MovingAverageMinMaxObserver.html
@@ -598,7 +598,7 @@
 
                                MovingAverageMinMaxObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.MovingAverageMinMaxObserver(averaging_constant=0.01, dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.MovingAverageMinMaxObserver(averaging_constant=0.01, dtype=torch.quint8, qscheme=torch.per_tensor_affine, reduce_range=False, quant_min=None, quant_max=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
 
                                Observer module for computing the quantization parameters based on the
 moving average of the min and max values.
                                
 
                                This observer computes the quantization parameters based on the moving
diff --git a/2.7/generated/torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver.html b/2.7/generated/torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver.html
index 847f5dd0655..d46bb2b4ce5 100644
--- a/2.7/generated/torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver.html
@@ -598,7 +598,7 @@
 
                                MovingAveragePerChannelMinMaxObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver(averaging_constant=0.01, ch_axis=0, dtype=torch.quint8, qscheme=torch.per_channel_affine, reduce_range=False, quant_min=None, quant_max=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.MovingAveragePerChannelMinMaxObserver(averaging_constant=0.01, ch_axis=0, dtype=torch.quint8, qscheme=torch.per_channel_affine, reduce_range=False, quant_min=None, quant_max=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
 
                                Observer module for computing the quantization parameters based on the
 running per channel min and max values.
                                
 
                                This observer uses the tensor min/max statistics to compute the per channel
diff --git a/2.7/generated/torch.ao.quantization.observer.NoopObserver.html b/2.7/generated/torch.ao.quantization.observer.NoopObserver.html
index 5ca61c7910f..4dcc7f3f93e 100644
--- a/2.7/generated/torch.ao.quantization.observer.NoopObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.NoopObserver.html
@@ -598,7 +598,7 @@
 
                                NoopObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.NoopObserver(dtype=torch.float16, custom_op_name='')[source][source]
                                
+class torch.ao.quantization.observer.NoopObserver(dtype=torch.float16, custom_op_name='')[source][source]
 
                                Observer that doesn’t do anything and just passes its configuration to the
 quantized module’s .from_float().
                                
 
                                Primarily used for quantization to float16 which doesn’t require determining
diff --git a/2.7/generated/torch.ao.quantization.observer.ObserverBase.html b/2.7/generated/torch.ao.quantization.observer.ObserverBase.html
index 0afdb23cea7..bb411dfcc66 100644
--- a/2.7/generated/torch.ao.quantization.observer.ObserverBase.html
+++ b/2.7/generated/torch.ao.quantization.observer.ObserverBase.html
@@ -598,7 +598,7 @@
 
                                ObserverBase
                                
 
                                
 
                                
-class torch.ao.quantization.observer.ObserverBase(dtype, is_dynamic=False)[source][source]
                                
+class torch.ao.quantization.observer.ObserverBase(dtype, is_dynamic=False)[source][source]
 
                                Base observer Module.
 Any observer implementation should derive from this class.
                                
 
                                Concrete observers should follow the same API. In forward, they will update
@@ -617,7 +617,7 @@ 
                                ObserverBase
 
                                
-classmethod with_args(**kwargs)[source]
                                
+classmethod with_args(**kwargs)[source]
 
                                Wrapper that allows creation of class factories.
                                
 
                                This can be useful when there is a need to create classes with the same
 constructor arguments, but different instances. Can be used in conjunction with
@@ -635,7 +635,7 @@ 
                                ObserverBase
 
                                
-classmethod with_callable_args(**kwargs)[source]
                                
+classmethod with_callable_args(**kwargs)[source]
 
                                Wrapper that allows creation of class factories args that need to be
 called at construction time.
                                
 
                                This can be useful when there is a need to create classes with the same
diff --git a/2.7/generated/torch.ao.quantization.observer.PerAxis.html b/2.7/generated/torch.ao.quantization.observer.PerAxis.html
index 2f40ff0e96a..a859387cb31 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerAxis.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerAxis.html
@@ -598,7 +598,7 @@
 
                                PerAxis
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerAxis(axis)[source][source]
                                
+class torch.ao.quantization.observer.PerAxis(axis)[source][source]
 
                                Represents per-axis granularity in quantization.
                                
 
                                This granularity type calculates different quantization parameters
 along a specified axis of the tensor.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.PerBlock.html b/2.7/generated/torch.ao.quantization.observer.PerBlock.html
index 5844c770d00..a9cd2901ed0 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerBlock.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerBlock.html
@@ -598,7 +598,7 @@
 
                                PerBlock
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerBlock(block_size)[source][source]
                                
+class torch.ao.quantization.observer.PerBlock(block_size)[source][source]
 
                                Represents per-block granularity in quantization. See
 quantize_affine() for docs for
 block_size
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.PerChannelMinMaxObserver.html b/2.7/generated/torch.ao.quantization.observer.PerChannelMinMaxObserver.html
index ca44459c672..4e8d6741d90 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerChannelMinMaxObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerChannelMinMaxObserver.html
@@ -598,7 +598,7 @@
 
                                PerChannelMinMaxObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerChannelMinMaxObserver(ch_axis=0, dtype=torch.quint8, qscheme=torch.per_channel_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
                                
+class torch.ao.quantization.observer.PerChannelMinMaxObserver(ch_axis=0, dtype=torch.quint8, qscheme=torch.per_channel_affine, reduce_range=False, quant_min=None, quant_max=None, factory_kwargs=None, eps=1.1920928955078125e-07, is_dynamic=False, **kwargs)[source][source]
 
                                Observer module for computing the quantization parameters based on the
 running per channel min and max values.
                                
 
                                This observer uses the tensor min/max statistics to compute the per channel
@@ -630,7 +630,7 @@ 
                                PerChannelMinMaxObserver
 
                                
-reset_min_max_vals()[source][source]
                                
+reset_min_max_vals()[source][source]
 
                                Resets the min/max values.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.observer.PerGroup.html b/2.7/generated/torch.ao.quantization.observer.PerGroup.html
index 9ca322f6b0d..b969e92148e 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerGroup.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerGroup.html
@@ -598,7 +598,7 @@
 
                                PerGroup
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerGroup(group_size)[source][source]
                                
+class torch.ao.quantization.observer.PerGroup(group_size)[source][source]
 
                                Represents per-channel group granularity in quantization.
                                
 
                                This granularity type calculates different quantization parameters
 for each group of <group_size> elements.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.PerRow.html b/2.7/generated/torch.ao.quantization.observer.PerRow.html
index c243882c868..98418cf6648 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerRow.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerRow.html
@@ -598,7 +598,7 @@
 
                                PerRow
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerRow[source][source]
                                
+class torch.ao.quantization.observer.PerRow[source][source]
 
                                Represents row-wise granularity in quantization.
                                
 
                                This is a special case of per-axis quantization and is unique to Float8 matmuls
 where the input is quantized with a block_size of (1, …, input.shape[-1]). And the weight
diff --git a/2.7/generated/torch.ao.quantization.observer.PerTensor.html b/2.7/generated/torch.ao.quantization.observer.PerTensor.html
index 2babfbc6d96..3f820ce882d 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerTensor.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerTensor.html
@@ -598,7 +598,7 @@
 
                                PerTensor
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerTensor[source][source]
                                
+class torch.ao.quantization.observer.PerTensor[source][source]
 
                                Represents per-tensor granularity in quantization.
                                
 
                                This granularity type calculates the quantization parameters
 based off the entire tensor.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.PerToken.html b/2.7/generated/torch.ao.quantization.observer.PerToken.html
index 2b21554694d..0398ab02565 100644
--- a/2.7/generated/torch.ao.quantization.observer.PerToken.html
+++ b/2.7/generated/torch.ao.quantization.observer.PerToken.html
@@ -598,7 +598,7 @@
 
                                PerToken
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PerToken[source][source]
                                
+class torch.ao.quantization.observer.PerToken[source][source]
 
                                Represents per-token granularity in quantization.
                                
 
                                This granularity type calculates a different set of quantization parameters
 for each token, which is represented as the last dimension of the tensor.
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.PlaceholderObserver.html b/2.7/generated/torch.ao.quantization.observer.PlaceholderObserver.html
index 6adc9865177..f84c396fe50 100644
--- a/2.7/generated/torch.ao.quantization.observer.PlaceholderObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.PlaceholderObserver.html
@@ -598,7 +598,7 @@
 
                                PlaceholderObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.PlaceholderObserver(dtype=torch.float32, custom_op_name='', compute_dtype=None, quant_min=None, quant_max=None, qscheme=None, eps=None, is_dynamic=False)[source][source]
                                
+class torch.ao.quantization.observer.PlaceholderObserver(dtype=torch.float32, custom_op_name='', compute_dtype=None, quant_min=None, quant_max=None, qscheme=None, eps=None, is_dynamic=False)[source][source]
 
                                Observer that doesn’t do anything and just passes its configuration to the
 quantized module’s .from_float().
                                
 
                                Can be used for quantization to float16 which doesn’t require determining
diff --git a/2.7/generated/torch.ao.quantization.observer.RecordingObserver.html b/2.7/generated/torch.ao.quantization.observer.RecordingObserver.html
index a33f184c337..d0aecd8d612 100644
--- a/2.7/generated/torch.ao.quantization.observer.RecordingObserver.html
+++ b/2.7/generated/torch.ao.quantization.observer.RecordingObserver.html
@@ -598,7 +598,7 @@
 
                                RecordingObserver
                                
 
                                
 
                                
-class torch.ao.quantization.observer.RecordingObserver(dtype=torch.quint8)[source][source]
                                
+class torch.ao.quantization.observer.RecordingObserver(dtype=torch.quint8)[source][source]
 
                                The module is mainly for debug and records the tensor values during runtime.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.TorchAODType.html b/2.7/generated/torch.ao.quantization.observer.TorchAODType.html
index fd450653d26..51075f31c97 100644
--- a/2.7/generated/torch.ao.quantization.observer.TorchAODType.html
+++ b/2.7/generated/torch.ao.quantization.observer.TorchAODType.html
@@ -598,7 +598,7 @@
 
                                TorchAODType
                                
 
                                
 
                                
-class torch.ao.quantization.observer.TorchAODType(value)[source][source]
                                
+class torch.ao.quantization.observer.TorchAODType(value)[source][source]
 
                                Placeholder for dtypes that do not exist in PyTorch core yet.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.observer.ZeroPointDomain.html b/2.7/generated/torch.ao.quantization.observer.ZeroPointDomain.html
index bd87b74deb0..049a8eec39a 100644
--- a/2.7/generated/torch.ao.quantization.observer.ZeroPointDomain.html
+++ b/2.7/generated/torch.ao.quantization.observer.ZeroPointDomain.html
@@ -598,7 +598,7 @@
 
                                ZeroPointDomain
                                
 
                                
 
                                
-class torch.ao.quantization.observer.ZeroPointDomain(value)[source][source]
                                
+class torch.ao.quantization.observer.ZeroPointDomain(value)[source][source]
 
                                Enum that indicate whether zero_point is in integer domain or floating point domain
                                
 
                                integer domain: quantized_val = (float_val / scale) (integer) + zero_point (integer)
 float domain: quantized_val = (float_val - (zero_point (float) - scale * mid_point)) / scale
diff --git a/2.7/generated/torch.ao.quantization.observer.default_debug_observer.html b/2.7/generated/torch.ao.quantization.observer.default_debug_observer.html
index 41a5add258e..d9a3fab2ab6 100644
--- a/2.7/generated/torch.ao.quantization.observer.default_debug_observer.html
+++ b/2.7/generated/torch.ao.quantization.observer.default_debug_observer.html
@@ -598,7 +598,7 @@
 
                                default_debug_observer
                                
 
                                
 
                                
-torch.ao.quantization.observer.default_debug_observer[source]
                                
+torch.ao.quantization.observer.default_debug_observer[source]
 
                                Default debug-only observer.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.observer.default_placeholder_observer.html b/2.7/generated/torch.ao.quantization.observer.default_placeholder_observer.html
index 4a9dcde55fa..954269f8b05 100644
--- a/2.7/generated/torch.ao.quantization.observer.default_placeholder_observer.html
+++ b/2.7/generated/torch.ao.quantization.observer.default_placeholder_observer.html
@@ -598,7 +598,7 @@
 
                                default_placeholder_observer
                                
 
                                
 
                                
-torch.ao.quantization.observer.default_placeholder_observer[source]
                                
+torch.ao.quantization.observer.default_placeholder_observer[source]
 
                                Default placeholder observer, usually used for quantization to torch.float16.
                                
 
                                
 
diff --git a/2.7/generated/torch.ao.quantization.observer.get_block_size.html b/2.7/generated/torch.ao.quantization.observer.get_block_size.html
index 67c5cee5183..5f043438faf 100644
--- a/2.7/generated/torch.ao.quantization.observer.get_block_size.html
+++ b/2.7/generated/torch.ao.quantization.observer.get_block_size.html
@@ -598,7 +598,7 @@
 
                                get_block_size
                                
 
                                
 
                                
-class torch.ao.quantization.observer.get_block_size(input_shape, granularity)[source][source]
                                
+class torch.ao.quantization.observer.get_block_size(input_shape, granularity)[source][source]
 
                                Get the block size based on the input shape and granularity type.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.get_observer_state_dict.html b/2.7/generated/torch.ao.quantization.observer.get_observer_state_dict.html
index faf661d2aad..06988c871ed 100644
--- a/2.7/generated/torch.ao.quantization.observer.get_observer_state_dict.html
+++ b/2.7/generated/torch.ao.quantization.observer.get_observer_state_dict.html
@@ -598,7 +598,7 @@
 
                                get_observer_state_dict
                                
 
                                
 
                                
-class torch.ao.quantization.observer.get_observer_state_dict(mod)[source][source]
                                
+class torch.ao.quantization.observer.get_observer_state_dict(mod)[source][source]
 
                                Returns the state dict corresponding to the observer stats.
 Traverse the model state_dict and extract out the stats.
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.observer.load_observer_state_dict.html b/2.7/generated/torch.ao.quantization.observer.load_observer_state_dict.html
index 60e89556007..c621bd516f4 100644
--- a/2.7/generated/torch.ao.quantization.observer.load_observer_state_dict.html
+++ b/2.7/generated/torch.ao.quantization.observer.load_observer_state_dict.html
@@ -598,7 +598,7 @@
 
                                load_observer_state_dict
                                
 
                                
 
                                
-class torch.ao.quantization.observer.load_observer_state_dict(mod, obs_dict)[source][source]
                                
+class torch.ao.quantization.observer.load_observer_state_dict(mod, obs_dict)[source][source]
 
                                Given input model and a state_dict containing model observer stats,
 load the stats back into the model. The observer state_dict can be saved
 using torch.ao.quantization.get_observer_state_dict
                                
diff --git a/2.7/generated/torch.ao.quantization.prepare.html b/2.7/generated/torch.ao.quantization.prepare.html
index d40fdb06cad..4faf283c04a 100644
--- a/2.7/generated/torch.ao.quantization.prepare.html
+++ b/2.7/generated/torch.ao.quantization.prepare.html
@@ -598,7 +598,7 @@
 
                                prepare
                                
 
                                
 
                                
-class torch.ao.quantization.prepare(model, inplace=False, allow_list=None, observer_non_leaf_module_list=None, prepare_custom_config_dict=None)[source][source]
                                
+class torch.ao.quantization.prepare(model, inplace=False, allow_list=None, observer_non_leaf_module_list=None, prepare_custom_config_dict=None)[source][source]
 
                                Prepares a copy of the model for quantization calibration or quantization-aware training.
                                
 
                                Quantization configuration should be assigned preemptively
 to individual submodules in .qconfig attribute.
                                
diff --git a/2.7/generated/torch.ao.quantization.prepare_for_propagation_comparison.html b/2.7/generated/torch.ao.quantization.prepare_for_propagation_comparison.html
index ee45a0a185c..28df4944a86 100644
--- a/2.7/generated/torch.ao.quantization.prepare_for_propagation_comparison.html
+++ b/2.7/generated/torch.ao.quantization.prepare_for_propagation_comparison.html
@@ -598,7 +598,7 @@
 
                                prepare_for_propagation_comparison
                                
 
                                
 
                                
-class torch.ao.quantization.prepare_for_propagation_comparison(model)[source]
                                
+class torch.ao.quantization.prepare_for_propagation_comparison(model)[source]
 
                                Add output loggers to node that has numeric_debug_handle
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.prepare_qat.html b/2.7/generated/torch.ao.quantization.prepare_qat.html
index c2986b979f5..daded4b6a93 100644
--- a/2.7/generated/torch.ao.quantization.prepare_qat.html
+++ b/2.7/generated/torch.ao.quantization.prepare_qat.html
@@ -598,7 +598,7 @@
 
                                prepare_qat
                                
 
                                
 
                                
-class torch.ao.quantization.prepare_qat(model, mapping=None, inplace=False)[source][source]
                                
+class torch.ao.quantization.prepare_qat(model, mapping=None, inplace=False)[source][source]
 
                                Prepares a copy of the model for quantization calibration or
 quantization-aware training and converts it to quantized version.
                                
 
                                Quantization configuration should be assigned preemptively
diff --git a/2.7/generated/torch.ao.quantization.propagate_qconfig_.html b/2.7/generated/torch.ao.quantization.propagate_qconfig_.html
index 073c63892b0..69624d4a0cb 100644
--- a/2.7/generated/torch.ao.quantization.propagate_qconfig_.html
+++ b/2.7/generated/torch.ao.quantization.propagate_qconfig_.html
@@ -598,7 +598,7 @@
 
                                propagate_qconfig
                                
 
                                
 
                                
-class torch.ao.quantization.propagate_qconfig_(module, qconfig_dict=None, prepare_custom_config_dict=None)[source][source]
                                
+class torch.ao.quantization.propagate_qconfig_(module, qconfig_dict=None, prepare_custom_config_dict=None)[source][source]
 
                                Propagate qconfig through the module hierarchy and assign qconfig
 attribute on each leaf module
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.pt2e.export_utils.model_is_exported.html b/2.7/generated/torch.ao.quantization.pt2e.export_utils.model_is_exported.html
index 5a42f9b4bcb..0ab57e04022 100644
--- a/2.7/generated/torch.ao.quantization.pt2e.export_utils.model_is_exported.html
+++ b/2.7/generated/torch.ao.quantization.pt2e.export_utils.model_is_exported.html
@@ -598,7 +598,7 @@
 
                                model_is_exported
                                
 
                                
 
                                
-class torch.ao.quantization.pt2e.export_utils.model_is_exported(m)[source][source]
                                
+class torch.ao.quantization.pt2e.export_utils.model_is_exported(m)[source][source]
 
                                Return True if the torch.nn.Module was exported, False otherwise
 (e.g. if the model was FX symbolically traced or not traced at all).
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.qconfig.QConfig.html b/2.7/generated/torch.ao.quantization.qconfig.QConfig.html
index 0fbf9976ce8..6ee5790adc4 100644
--- a/2.7/generated/torch.ao.quantization.qconfig.QConfig.html
+++ b/2.7/generated/torch.ao.quantization.qconfig.QConfig.html
@@ -598,7 +598,7 @@
 
                                QConfig
                                
 
                                
 
                                
-class torch.ao.quantization.qconfig.QConfig(activation, weight)[source][source]
                                
+class torch.ao.quantization.qconfig.QConfig(activation, weight)[source][source]
 
                                Describes how to quantize a layer or a part of the network by providing
 settings (observer classes) for activations and weights respectively.
                                
 
                                Note that QConfig needs to contain observer classes (like MinMaxObserver) or a callable that returns
diff --git a/2.7/generated/torch.ao.quantization.qconfig_mapping.QConfigMapping.html b/2.7/generated/torch.ao.quantization.qconfig_mapping.QConfigMapping.html
index ca7cfd048f4..e7c5be27d2d 100644
--- a/2.7/generated/torch.ao.quantization.qconfig_mapping.QConfigMapping.html
+++ b/2.7/generated/torch.ao.quantization.qconfig_mapping.QConfigMapping.html
@@ -598,7 +598,7 @@
 
                                QConfigMapping
                                
 
                                
 
                                
-class torch.ao.quantization.qconfig_mapping.QConfigMapping[source][source]
                                
+class torch.ao.quantization.qconfig_mapping.QConfigMapping[source][source]
 
                                Mapping from model ops to torch.ao.quantization.QConfig s.
                                
 
                                The user can specify QConfigs using the following methods (in increasing match priority):
                                
 
                                
@@ -625,7 +625,7 @@ 
                                QConfigMapping
 
                                
-classmethod from_dict(qconfig_dict)[source][source]
                                
+classmethod from_dict(qconfig_dict)[source][source]
 
                                Create a QConfigMapping from a dictionary with the following keys (all optional):
                                
 
                                
 
                                “” (for global QConfig)
                                
@@ -644,7 +644,7 @@ 
                                QConfigMapping
 
                                
-set_global(global_qconfig)[source][source]
                                
+set_global(global_qconfig)[source][source]
 
                                Set the global (default) QConfig.
                                
 
                                
 
                                Return type
                                
@@ -655,7 +655,7 @@ 
                                QConfigMapping
 
                                
-set_module_name(module_name, qconfig)[source][source]
                                
+set_module_name(module_name, qconfig)[source][source]
 
                                Set the QConfig for modules matching the given module name.
 If the QConfig for an existing module name was already set, the new QConfig will override the old one.
                                
 
                                
@@ -667,7 +667,7 @@ 
                                QConfigMapping
 
                                
-set_module_name_object_type_order(module_name, object_type, index, qconfig)[source][source]
                                
+set_module_name_object_type_order(module_name, object_type, index, qconfig)[source][source]
 
                                Set the QConfig for modules matching a combination of the given module name, object type,
 and the index at which the module appears.
                                
 
                                If the QConfig for an existing (module name, object type, index)  was already set, the new QConfig
@@ -681,7 +681,7 @@ 
                                QConfigMapping
 
                                
-set_module_name_regex(module_name_regex, qconfig)[source][source]
                                
+set_module_name_regex(module_name_regex, qconfig)[source][source]
 
                                Set the QConfig for modules matching the given regex string.
                                
 
                                Regexes will be matched in the order in which they are registered through this method.
 Thus, the caller should register more specific patterns first, e.g.:
                                
@@ -704,7 +704,7 @@ 
                                QConfigMapping
 
                                
-set_object_type(object_type, qconfig)[source][source]
                                
+set_object_type(object_type, qconfig)[source][source]
 
                                Set the QConfig for a given module type, function, or method name.
 If the QConfig for an existing object type was already set, the new QConfig will override the old one.
                                
 
                                
@@ -716,7 +716,7 @@ 
                                QConfigMapping
 
                                
-to_dict()[source][source]
                                
+to_dict()[source][source]
 
                                Convert this QConfigMapping to a dictionary with the following keys:
                                
 
                                
 
                                “” (for global QConfig)
                                
diff --git a/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping.html b/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping.html
index 9856792f9f4..08160005ab8 100644
--- a/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping.html
+++ b/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping.html
@@ -598,7 +598,7 @@
 
                                get_default_qat_qconfig_mapping
                                
 
                                
 
                                
-class torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping(backend='x86', version=1)[source][source]
                                
+class torch.ao.quantization.qconfig_mapping.get_default_qat_qconfig_mapping(backend='x86', version=1)[source][source]
 
                                Return the default QConfigMapping for quantization aware training.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping.html b/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping.html
index e8d57be2db2..1cbb5f69adb 100644
--- a/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping.html
+++ b/2.7/generated/torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping.html
@@ -598,7 +598,7 @@
 
                                get_default_qconfig_mapping
                                
 
                                
 
                                
-class torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping(backend='x86', version=0)[source][source]
                                
+class torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping(backend='x86', version=0)[source][source]
 
                                Return the default QConfigMapping for post training quantization.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.quantize.html b/2.7/generated/torch.ao.quantization.quantize.html
index 3bf2333fd7f..9f504428e28 100644
--- a/2.7/generated/torch.ao.quantization.quantize.html
+++ b/2.7/generated/torch.ao.quantization.quantize.html
@@ -598,7 +598,7 @@
 
                                quantize
                                
 
                                
 
                                
-class torch.ao.quantization.quantize(model, run_fn, run_args, mapping=None, inplace=False)[source][source]
                                
+class torch.ao.quantization.quantize(model, run_fn, run_args, mapping=None, inplace=False)[source][source]
 
                                Quantize the input float model with post training static quantization.
                                
 
                                First it will prepare the model for calibration, then it calls
 run_fn which will run the calibration step, after that we will
diff --git a/2.7/generated/torch.ao.quantization.quantize_dynamic.html b/2.7/generated/torch.ao.quantization.quantize_dynamic.html
index 2d03fefec8e..e01b9198dd7 100644
--- a/2.7/generated/torch.ao.quantization.quantize_dynamic.html
+++ b/2.7/generated/torch.ao.quantization.quantize_dynamic.html
@@ -598,7 +598,7 @@
 
                                quantize_dynamic
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_dynamic(model, qconfig_spec=None, dtype=torch.qint8, mapping=None, inplace=False)[source][source]
                                
+class torch.ao.quantization.quantize_dynamic(model, qconfig_spec=None, dtype=torch.qint8, mapping=None, inplace=False)[source][source]
 
                                Converts a float model to dynamic (i.e. weights-only) quantized model.
                                
 
                                Replaces specified modules with dynamic weight-only quantized versions and output the quantized model.
                                
 
                                For simplest usage provide dtype argument that can be float16 or qint8. Weight-only quantization
diff --git a/2.7/generated/torch.ao.quantization.quantize_fx.convert_fx.html b/2.7/generated/torch.ao.quantization.quantize_fx.convert_fx.html
index e3cdb7d64f7..f2d44a1935a 100644
--- a/2.7/generated/torch.ao.quantization.quantize_fx.convert_fx.html
+++ b/2.7/generated/torch.ao.quantization.quantize_fx.convert_fx.html
@@ -598,7 +598,7 @@
 
                                convert_fx
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_fx.convert_fx(graph_module, convert_custom_config=None, _remove_qconfig=True, qconfig_mapping=None, backend_config=None, keep_original_weights=False)[source][source]
                                
+class torch.ao.quantization.quantize_fx.convert_fx(graph_module, convert_custom_config=None, _remove_qconfig=True, qconfig_mapping=None, backend_config=None, keep_original_weights=False)[source][source]
 
                                Convert a calibrated or trained model to a quantized model
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.quantize_fx.fuse_fx.html b/2.7/generated/torch.ao.quantization.quantize_fx.fuse_fx.html
index 8afdb5ede06..3c2ee259886 100644
--- a/2.7/generated/torch.ao.quantization.quantize_fx.fuse_fx.html
+++ b/2.7/generated/torch.ao.quantization.quantize_fx.fuse_fx.html
@@ -598,7 +598,7 @@
 
                                fuse_fx
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_fx.fuse_fx(model, fuse_custom_config=None, backend_config=None)[source][source]
                                
+class torch.ao.quantization.quantize_fx.fuse_fx(model, fuse_custom_config=None, backend_config=None)[source][source]
 
                                Fuse modules like conv+bn, conv+bn+relu etc, model must be in eval mode.
 Fusion rules are defined in torch.ao.quantization.fx.fusion_pattern.py
                                
 
                                
diff --git a/2.7/generated/torch.ao.quantization.quantize_fx.prepare_fx.html b/2.7/generated/torch.ao.quantization.quantize_fx.prepare_fx.html
index 9123a840dc7..0cc6af9cacd 100644
--- a/2.7/generated/torch.ao.quantization.quantize_fx.prepare_fx.html
+++ b/2.7/generated/torch.ao.quantization.quantize_fx.prepare_fx.html
@@ -598,7 +598,7 @@
 
                                prepare_fx
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_fx.prepare_fx(model, qconfig_mapping, example_inputs, prepare_custom_config=None, _equalization_config=None, backend_config=None)[source][source]
                                
+class torch.ao.quantization.quantize_fx.prepare_fx(model, qconfig_mapping, example_inputs, prepare_custom_config=None, _equalization_config=None, backend_config=None)[source][source]
 
                                Prepare a model for post training quantization
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.quantize_fx.prepare_qat_fx.html b/2.7/generated/torch.ao.quantization.quantize_fx.prepare_qat_fx.html
index de8847e2c20..760b464090d 100644
--- a/2.7/generated/torch.ao.quantization.quantize_fx.prepare_qat_fx.html
+++ b/2.7/generated/torch.ao.quantization.quantize_fx.prepare_qat_fx.html
@@ -598,7 +598,7 @@
 
                                prepare_qat_fx
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_fx.prepare_qat_fx(model, qconfig_mapping, example_inputs, prepare_custom_config=None, backend_config=None)[source][source]
                                
+class torch.ao.quantization.quantize_fx.prepare_qat_fx(model, qconfig_mapping, example_inputs, prepare_custom_config=None, backend_config=None)[source][source]
 
                                Prepare a model for quantization aware training
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.quantize_qat.html b/2.7/generated/torch.ao.quantization.quantize_qat.html
index bca2aa4f667..1cb884211b9 100644
--- a/2.7/generated/torch.ao.quantization.quantize_qat.html
+++ b/2.7/generated/torch.ao.quantization.quantize_qat.html
@@ -598,7 +598,7 @@
 
                                quantize_qat
                                
 
                                
 
                                
-class torch.ao.quantization.quantize_qat(model, run_fn, run_args, inplace=False)[source][source]
                                
+class torch.ao.quantization.quantize_qat(model, run_fn, run_args, inplace=False)[source][source]
 
                                Do quantization aware training and output a quantized model
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.ao.quantization.swap_module.html b/2.7/generated/torch.ao.quantization.swap_module.html
index a41761a8bfa..dab82649350 100644
--- a/2.7/generated/torch.ao.quantization.swap_module.html
+++ b/2.7/generated/torch.ao.quantization.swap_module.html
@@ -598,7 +598,7 @@
 
                                swap_module
                                
 
                                
 
                                
-class torch.ao.quantization.swap_module(mod, mapping, custom_module_class_mapping, use_precomputed_fake_quant=False)[source][source]
                                
+class torch.ao.quantization.swap_module(mod, mapping, custom_module_class_mapping, use_precomputed_fake_quant=False)[source][source]
 
                                Swaps the module if it has a quantized counterpart and it has an
 observer attached.
                                
 
                                
diff --git a/2.7/generated/torch.are_deterministic_algorithms_enabled.html b/2.7/generated/torch.are_deterministic_algorithms_enabled.html
index d052820489a..4a5d2c43b9f 100644
--- a/2.7/generated/torch.are_deterministic_algorithms_enabled.html
+++ b/2.7/generated/torch.are_deterministic_algorithms_enabled.html
@@ -596,7 +596,7 @@
 
                                torch.are_deterministic_algorithms_enabled
                                
 
                                
 
                                
-torch.are_deterministic_algorithms_enabled()[source][source]
                                
+torch.are_deterministic_algorithms_enabled()[source][source]
 
                                Returns True if the global deterministic flag is turned on. Refer to
 torch.use_deterministic_algorithms() documentation for more details.
                                
 
                                
diff --git a/2.7/generated/torch.atleast_1d.html b/2.7/generated/torch.atleast_1d.html
index fdc02778f48..6cd10bba941 100644
--- a/2.7/generated/torch.atleast_1d.html
+++ b/2.7/generated/torch.atleast_1d.html
@@ -596,7 +596,7 @@
 
                                torch.atleast_1d
                                
 
                                
 
                                
-torch.atleast_1d(*tensors)[source][source]
                                
+torch.atleast_1d(*tensors)[source][source]
 
                                Returns a 1-dimensional view of each input tensor with zero dimensions.
 Input tensors with one or more dimensions are returned as-is.
                                
 
                                
diff --git a/2.7/generated/torch.atleast_2d.html b/2.7/generated/torch.atleast_2d.html
index 843407f42b1..1b868f97322 100644
--- a/2.7/generated/torch.atleast_2d.html
+++ b/2.7/generated/torch.atleast_2d.html
@@ -596,7 +596,7 @@
 
                                torch.atleast_2d
                                
 
                                
 
                                
-torch.atleast_2d(*tensors)[source][source]
                                
+torch.atleast_2d(*tensors)[source][source]
 
                                Returns a 2-dimensional view of each input tensor with zero dimensions.
 Input tensors with two or more dimensions are returned as-is.
                                
 
                                
diff --git a/2.7/generated/torch.atleast_3d.html b/2.7/generated/torch.atleast_3d.html
index 5be9fd4be9e..dba16d3df4c 100644
--- a/2.7/generated/torch.atleast_3d.html
+++ b/2.7/generated/torch.atleast_3d.html
@@ -596,7 +596,7 @@
 
                                torch.atleast_3d
                                
 
                                
 
                                
-torch.atleast_3d(*tensors)[source][source]
                                
+torch.atleast_3d(*tensors)[source][source]
 
                                Returns a 3-dimensional view of each input tensor with zero dimensions.
 Input tensors with three or more dimensions are returned as-is.
                                
 
                                
diff --git a/2.7/generated/torch.autograd.Function.backward.html b/2.7/generated/torch.autograd.Function.backward.html
index 7a215106a20..7b447e2e43f 100644
--- a/2.7/generated/torch.autograd.Function.backward.html
+++ b/2.7/generated/torch.autograd.Function.backward.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.Function.backward
                                
 
                                
 
                                
-static Function.backward(ctx, *grad_outputs)[source]
                                
+static Function.backward(ctx, *grad_outputs)[source]
 
                                Define a formula for differentiating the operation with backward mode automatic differentiation.
                                
 
                                This function is to be overridden by all subclasses.
 (Defining this function is equivalent to defining the vjp function.)
                                
diff --git a/2.7/generated/torch.autograd.Function.forward.html b/2.7/generated/torch.autograd.Function.forward.html
index e6bcd531760..b94446c9f36 100644
--- a/2.7/generated/torch.autograd.Function.forward.html
+++ b/2.7/generated/torch.autograd.Function.forward.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.Function.forward
                                
 
                                
 
                                
-static Function.forward(*args, **kwargs)[source]
                                
+static Function.forward(*args, **kwargs)[source]
 
                                Define the forward of the custom autograd Function.
                                
 
                                This function is to be overridden by all subclasses.
 There are two ways to define forward:
                                
diff --git a/2.7/generated/torch.autograd.Function.jvp.html b/2.7/generated/torch.autograd.Function.jvp.html
index 71c0c886bd2..1ea043dcfdf 100644
--- a/2.7/generated/torch.autograd.Function.jvp.html
+++ b/2.7/generated/torch.autograd.Function.jvp.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.Function.jvp
                                
 
                                
 
                                
-static Function.jvp(ctx, *grad_inputs)[source]
                                
+static Function.jvp(ctx, *grad_inputs)[source]
 
                                Define a formula for differentiating the operation with forward mode automatic differentiation.
                                
 
                                This function is to be overridden by all subclasses.
 It must accept a context ctx as the first argument, followed by
diff --git a/2.7/generated/torch.autograd.Function.vmap.html b/2.7/generated/torch.autograd.Function.vmap.html
index 1bc4de3eaad..c1af44a0ec3 100644
--- a/2.7/generated/torch.autograd.Function.vmap.html
+++ b/2.7/generated/torch.autograd.Function.vmap.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.Function.vmap
                                
 
                                
 
                                
-static Function.vmap(info, in_dims, *args)[source][source]
                                
+static Function.vmap(info, in_dims, *args)[source][source]
 
                                Define the behavior for this autograd.Function underneath torch.vmap().
                                
 
                                For a torch.autograd.Function() to support
 torch.vmap(), you must either override this static method, or set
diff --git a/2.7/generated/torch.autograd.backward.html b/2.7/generated/torch.autograd.backward.html
index 64f2639c76a..2b72c1b353a 100644
--- a/2.7/generated/torch.autograd.backward.html
+++ b/2.7/generated/torch.autograd.backward.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.backward
                                
 
                                
 
                                
-torch.autograd.backward(tensors, grad_tensors=None, retain_graph=None, create_graph=False, grad_variables=None, inputs=None)[source][source]
                                
+torch.autograd.backward(tensors, grad_tensors=None, retain_graph=None, create_graph=False, grad_variables=None, inputs=None)[source][source]
 
                                Compute the sum of gradients of given tensors with respect to graph leaves.
                                
 
                                The graph is differentiated using the chain rule. If any of tensors
 are non-scalar (i.e. their data has more than one element) and require
diff --git a/2.7/generated/torch.autograd.forward_ad.UnpackedDualTensor.html b/2.7/generated/torch.autograd.forward_ad.UnpackedDualTensor.html
index 0c62d69ab82..b9ecce409a7 100644
--- a/2.7/generated/torch.autograd.forward_ad.UnpackedDualTensor.html
+++ b/2.7/generated/torch.autograd.forward_ad.UnpackedDualTensor.html
@@ -596,7 +596,7 @@
 
                                UnpackedDualTensor
                                
 
                                
 
                                
-class torch.autograd.forward_ad.UnpackedDualTensor(primal, tangent)[source][source]
                                
+class torch.autograd.forward_ad.UnpackedDualTensor(primal, tangent)[source][source]
 
                                Namedtuple returned by unpack_dual() containing the primal and tangent components of the dual tensor.
                                
 
                                See unpack_dual() for more details.
                                
 
                                
diff --git a/2.7/generated/torch.autograd.forward_ad.dual_level.html b/2.7/generated/torch.autograd.forward_ad.dual_level.html
index e7cbf09f196..340cca803de 100644
--- a/2.7/generated/torch.autograd.forward_ad.dual_level.html
+++ b/2.7/generated/torch.autograd.forward_ad.dual_level.html
@@ -596,7 +596,7 @@
 
                                dual_level
                                
 
                                
 
                                
-class torch.autograd.forward_ad.dual_level[source][source]
                                
+class torch.autograd.forward_ad.dual_level[source][source]
 
                                Context-manager for forward AD, where all forward AD computation must occur within the dual_level context.
                                
 
                                
 
                                Note
                                
diff --git a/2.7/generated/torch.autograd.forward_ad.enter_dual_level.html b/2.7/generated/torch.autograd.forward_ad.enter_dual_level.html
index 7b63837fb8d..b2cfa8bea5f 100644
--- a/2.7/generated/torch.autograd.forward_ad.enter_dual_level.html
+++ b/2.7/generated/torch.autograd.forward_ad.enter_dual_level.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.forward_ad.enter_dual_level
                                
 
                                
 
                                
-torch.autograd.forward_ad.enter_dual_level()[source][source]
                                
+torch.autograd.forward_ad.enter_dual_level()[source][source]
 
                                Enter a new forward grad level.
                                
 
                                This level can be used to make and unpack dual Tensors to compute
 forward gradients.
                                
diff --git a/2.7/generated/torch.autograd.forward_ad.exit_dual_level.html b/2.7/generated/torch.autograd.forward_ad.exit_dual_level.html
index 951f75da863..d9c1412e4b6 100644
--- a/2.7/generated/torch.autograd.forward_ad.exit_dual_level.html
+++ b/2.7/generated/torch.autograd.forward_ad.exit_dual_level.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.forward_ad.exit_dual_level
                                
 
                                
 
                                
-torch.autograd.forward_ad.exit_dual_level(*, level=None)[source][source]
                                
+torch.autograd.forward_ad.exit_dual_level(*, level=None)[source][source]
 
                                Exit a forward grad level.
                                
 
                                This function deletes all the gradients associated with this
 level. Only deleting the latest entered level is allowed.
                                
diff --git a/2.7/generated/torch.autograd.forward_ad.make_dual.html b/2.7/generated/torch.autograd.forward_ad.make_dual.html
index 3e4516ea828..2ea3c4e9eb8 100644
--- a/2.7/generated/torch.autograd.forward_ad.make_dual.html
+++ b/2.7/generated/torch.autograd.forward_ad.make_dual.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.forward_ad.make_dual
                                
 
                                
 
                                
-torch.autograd.forward_ad.make_dual(tensor, tangent, *, level=None)[source][source]
                                
+torch.autograd.forward_ad.make_dual(tensor, tangent, *, level=None)[source][source]
 
                                Associate a tensor value with its tangent to create a “dual tensor” for forward AD gradient computation.
                                
 
                                The result is a new tensor aliased to tensor with tangent embedded
 as an attribute as-is if it has the same storage layout or copied otherwise.
diff --git a/2.7/generated/torch.autograd.forward_ad.unpack_dual.html b/2.7/generated/torch.autograd.forward_ad.unpack_dual.html
index f33fcc3c4cf..603809fcde2 100644
--- a/2.7/generated/torch.autograd.forward_ad.unpack_dual.html
+++ b/2.7/generated/torch.autograd.forward_ad.unpack_dual.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.forward_ad.unpack_dual
                                
 
                                
 
                                
-torch.autograd.forward_ad.unpack_dual(tensor, *, level=None)[source][source]
                                
+torch.autograd.forward_ad.unpack_dual(tensor, *, level=None)[source][source]
 
                                Unpack a “dual tensor” to get both its Tensor value and its forward AD gradient.
                                
 
                                The result is a namedtuple (primal, tangent) where primal is a view of
 tensor’s primal and tangent is tensor’s tangent as-is.
diff --git a/2.7/generated/torch.autograd.function.BackwardCFunction.html b/2.7/generated/torch.autograd.function.BackwardCFunction.html
index deacabd1c58..e7d052ba05a 100644
--- a/2.7/generated/torch.autograd.function.BackwardCFunction.html
+++ b/2.7/generated/torch.autograd.function.BackwardCFunction.html
@@ -596,23 +596,23 @@
 
                                BackwardCFunction
                                
 
                                
 
                                
-class torch.autograd.function.BackwardCFunction[source][source]
                                
+class torch.autograd.function.BackwardCFunction[source][source]
 
                                This class is used for internal autograd work. Do not use.
                                
 
                                
 
                                
-apply(*args)[source][source]
                                
+apply(*args)[source][source]
 
                                Apply method used when executing this Node during the backward
                                
 
                                
 
 
                                
 
                                
-apply_jvp(*args)[source][source]
                                
+apply_jvp(*args)[source][source]
 
                                Apply method used when executing forward mode AD during the forward
                                
 
                                
 
 
                                
 
                                
-mark_dirty(*args)[source]
                                
+mark_dirty(*args)[source]
 
                                Mark given tensors as modified in an in-place operation.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be inputs.
                                
@@ -650,7 +650,7 @@ 
                                BackwardCFunction
 
                                
-mark_non_differentiable(*args)[source]
                                
+mark_non_differentiable(*args)[source]
 
                                Mark outputs as non-differentiable.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be tensor outputs.
                                
@@ -685,7 +685,7 @@ 
                                BackwardCFunction
 
                                
-save_for_backward(*tensors)[source]
                                
+save_for_backward(*tensors)[source]
 
                                Save given tensors for a future call to backward().
                                
 
                                save_for_backward should be called at most once, in either the
 setup_context() or forward() methods, and only with tensors.
                                
@@ -742,7 +742,7 @@ 
                                BackwardCFunction
 
                                
-save_for_forward(*tensors)[source]
                                
+save_for_forward(*tensors)[source]
 
                                Save given tensors for a future call to jvp().
                                
 
                                save_for_forward should be called at most once, in either the
 setup_context() or forward() methods, and all arguments
@@ -790,7 +790,7 @@ 
                                BackwardCFunction
 
                                
-set_materialize_grads(value)[source]
                                
+set_materialize_grads(value)[source]
 
                                Set whether to materialize grad tensors. Default is True.
                                
 
                                This should be called only from either the setup_context() or
 forward() methods.
                                
diff --git a/2.7/generated/torch.autograd.function.FunctionCtx.mark_dirty.html b/2.7/generated/torch.autograd.function.FunctionCtx.mark_dirty.html
index 577ecfeb937..e65f5771fed 100644
--- a/2.7/generated/torch.autograd.function.FunctionCtx.mark_dirty.html
+++ b/2.7/generated/torch.autograd.function.FunctionCtx.mark_dirty.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.function.FunctionCtx.mark_dirty
                                
 
                                
 
                                
-FunctionCtx.mark_dirty(*args)[source][source]
                                
+FunctionCtx.mark_dirty(*args)[source][source]
 
                                Mark given tensors as modified in an in-place operation.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be inputs.
                                
diff --git a/2.7/generated/torch.autograd.function.FunctionCtx.mark_non_differentiable.html b/2.7/generated/torch.autograd.function.FunctionCtx.mark_non_differentiable.html
index a850686561f..ae71ed2fdd2 100644
--- a/2.7/generated/torch.autograd.function.FunctionCtx.mark_non_differentiable.html
+++ b/2.7/generated/torch.autograd.function.FunctionCtx.mark_non_differentiable.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.function.FunctionCtx.mark_non_differentiable
                                
 
                                
 
                                
-FunctionCtx.mark_non_differentiable(*args)[source][source]
                                
+FunctionCtx.mark_non_differentiable(*args)[source][source]
 
                                Mark outputs as non-differentiable.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be tensor outputs.
                                
diff --git a/2.7/generated/torch.autograd.function.FunctionCtx.save_for_backward.html b/2.7/generated/torch.autograd.function.FunctionCtx.save_for_backward.html
index 83549d16a42..8a79ceb28f4 100644
--- a/2.7/generated/torch.autograd.function.FunctionCtx.save_for_backward.html
+++ b/2.7/generated/torch.autograd.function.FunctionCtx.save_for_backward.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.function.FunctionCtx.save_for_backward
                                
 
                                
 
                                
-FunctionCtx.save_for_backward(*tensors)[source][source]
                                
+FunctionCtx.save_for_backward(*tensors)[source][source]
 
                                Save given tensors for a future call to backward().
                                
 
                                save_for_backward should be called at most once, in either the
 setup_context() or forward() methods, and only with tensors.
                                
diff --git a/2.7/generated/torch.autograd.function.FunctionCtx.set_materialize_grads.html b/2.7/generated/torch.autograd.function.FunctionCtx.set_materialize_grads.html
index 3718783110a..45d8af918cd 100644
--- a/2.7/generated/torch.autograd.function.FunctionCtx.set_materialize_grads.html
+++ b/2.7/generated/torch.autograd.function.FunctionCtx.set_materialize_grads.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.function.FunctionCtx.set_materialize_grads
                                
 
                                
 
                                
-FunctionCtx.set_materialize_grads(value)[source][source]
                                
+FunctionCtx.set_materialize_grads(value)[source][source]
 
                                Set whether to materialize grad tensors. Default is True.
                                
 
                                This should be called only from either the setup_context() or
 forward() methods.
                                
diff --git a/2.7/generated/torch.autograd.function.InplaceFunction.html b/2.7/generated/torch.autograd.function.InplaceFunction.html
index 9613b8c0a44..0bb15519452 100644
--- a/2.7/generated/torch.autograd.function.InplaceFunction.html
+++ b/2.7/generated/torch.autograd.function.InplaceFunction.html
@@ -596,12 +596,12 @@
 
                                InplaceFunction
                                
 
                                
 
                                
-class torch.autograd.function.InplaceFunction(inplace=False)[source][source]
                                
+class torch.autograd.function.InplaceFunction(inplace=False)[source][source]
 
                                This class is here only for backward compatibility reasons.
 Use Function instead of this for any new use case.
                                
 
                                
 
                                
-static backward(ctx, *grad_outputs)[source]
                                
+static backward(ctx, *grad_outputs)[source]
 
                                Define a formula for differentiating the operation with backward mode automatic differentiation.
                                
 
                                This function is to be overridden by all subclasses.
 (Defining this function is equivalent to defining the vjp function.)
                                
@@ -628,7 +628,7 @@ 
                                InplaceFunction
 
                                
-static forward(*args, **kwargs)[source]
                                
+static forward(*args, **kwargs)[source]
 
                                Define the forward of the custom autograd Function.
                                
 
                                This function is to be overridden by all subclasses.
 There are two ways to define forward:
                                
@@ -677,7 +677,7 @@ 
                                InplaceFunction
 
                                
-static jvp(ctx, *grad_inputs)[source]
                                
+static jvp(ctx, *grad_inputs)[source]
 
                                Define a formula for differentiating the operation with forward mode automatic differentiation.
                                
 
                                This function is to be overridden by all subclasses.
 It must accept a context ctx as the first argument, followed by
@@ -700,7 +700,7 @@ 
                                InplaceFunction
 
                                
-mark_dirty(*args)[source]
                                
+mark_dirty(*args)[source]
 
                                Mark given tensors as modified in an in-place operation.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be inputs.
                                
@@ -738,7 +738,7 @@ 
                                InplaceFunction
 
                                
-mark_non_differentiable(*args)[source]
                                
+mark_non_differentiable(*args)[source]
 
                                Mark outputs as non-differentiable.
                                
 
                                This should be called at most once, in either the setup_context()
 or forward() methods, and all arguments should be tensor outputs.
                                
@@ -773,7 +773,7 @@ 
                                InplaceFunction
 
                                
-save_for_backward(*tensors)[source]
                                
+save_for_backward(*tensors)[source]
 
                                Save given tensors for a future call to backward().
                                
 
                                save_for_backward should be called at most once, in either the
 setup_context() or forward() methods, and only with tensors.
                                
@@ -830,7 +830,7 @@ 
                                InplaceFunction
 
                                
-save_for_forward(*tensors)[source]
                                
+save_for_forward(*tensors)[source]
 
                                Save given tensors for a future call to jvp().
                                
 
                                save_for_forward should be called at most once, in either the
 setup_context() or forward() methods, and all arguments
@@ -878,7 +878,7 @@ 
                                InplaceFunction
 
                                
-set_materialize_grads(value)[source]
                                
+set_materialize_grads(value)[source]
 
                                Set whether to materialize grad tensors. Default is True.
                                
 
                                This should be called only from either the setup_context() or
 forward() methods.
                                
@@ -926,7 +926,7 @@ 
                                InplaceFunction
 
                                
-static setup_context(ctx, inputs, output)[source]
                                
+static setup_context(ctx, inputs, output)[source]
 
                                There are two ways to define the forward pass of an autograd.Function.
                                
 
                                Either:
                                
 
                                  
@@ -947,7 +947,7 @@ 
                                  InplaceFunction
 
                                  
-static vjp(ctx, *grad_outputs)[source]
                                  
+static vjp(ctx, *grad_outputs)[source]
 
                                  Define a formula for differentiating the operation with backward mode automatic differentiation.
                                  
 
                                  This function is to be overridden by all subclasses.
 (Defining this function is equivalent to defining the vjp function.)
                                  
@@ -974,7 +974,7 @@ 
                                  InplaceFunction
 
                                  
-static vmap(info, in_dims, *args)[source]
                                  
+static vmap(info, in_dims, *args)[source]
 
                                  Define the behavior for this autograd.Function underneath torch.vmap().
                                  
 
                                  For a torch.autograd.Function() to support
 torch.vmap(), you must either override this static method, or set
diff --git a/2.7/generated/torch.autograd.function.NestedIOFunction.html b/2.7/generated/torch.autograd.function.NestedIOFunction.html
index af5d99299f9..b4ea378aa7b 100644
--- a/2.7/generated/torch.autograd.function.NestedIOFunction.html
+++ b/2.7/generated/torch.autograd.function.NestedIOFunction.html
@@ -596,12 +596,12 @@
 
                                  NestedIOFunction
                                  
 
                                  
 
                                  
-class torch.autograd.function.NestedIOFunction(*args, **kwargs)[source][source]
                                  
+class torch.autograd.function.NestedIOFunction(*args, **kwargs)[source][source]
 
                                  This class is here only for backward compatibility reasons.
 Use Function instead of this for any new use case.
                                  
 
                                  
 
                                  
-backward(*gradients)[source][source]
                                  
+backward(*gradients)[source][source]
 
                                  Shared backward utility.
                                  
 
                                  
 
                                  Return type
                                  
@@ -612,7 +612,7 @@ 
                                  NestedIOFunction
 
                                  
-backward_extended(*grad_output)[source][source]
                                  
+backward_extended(*grad_output)[source][source]
 
                                  User defined backward.
                                  
 
                                  
 
                                  
@@ -620,7 +620,7 @@ 
                                  NestedIOFunction
 
                                  
-forward(*args)[source][source]
                                  
+forward(*args)[source][source]
 
                                  Shared forward utility.
                                  
 
                                  
 
                                  Return type
                                  
@@ -631,7 +631,7 @@ 
                                  NestedIOFunction
 
                                  
-forward_extended(*input)[source][source]
                                  
+forward_extended(*input)[source][source]
 
                                  User defined forward.
                                  
 
                                  
 
                                  
@@ -639,7 +639,7 @@ 
                                  NestedIOFunction
 
                                  
-static jvp(ctx, *grad_inputs)[source]
                                  
+static jvp(ctx, *grad_inputs)[source]
 
                                  Define a formula for differentiating the operation with forward mode automatic differentiation.
                                  
 
                                  This function is to be overridden by all subclasses.
 It must accept a context ctx as the first argument, followed by
@@ -662,7 +662,7 @@ 
                                  NestedIOFunction
 
                                  
-mark_dirty(*args, **kwargs)[source][source]
                                  
+mark_dirty(*args, **kwargs)[source][source]
 
                                  See Function.mark_dirty().
                                  
 
                                  
 
                                  
@@ -670,7 +670,7 @@ 
                                  NestedIOFunction
 
                                  
-mark_non_differentiable(*args, **kwargs)[source][source]
                                  
+mark_non_differentiable(*args, **kwargs)[source][source]
 
                                  See Function.mark_non_differentiable().
                                  
 
                                  
 
                                  
@@ -678,7 +678,7 @@ 
                                  NestedIOFunction
 
                                  
-save_for_backward(*args)[source][source]
                                  
+save_for_backward(*args)[source][source]
 
                                  See Function.save_for_backward().
                                  
 
                                  
 
                                  
@@ -686,7 +686,7 @@ 
                                  NestedIOFunction
 
                                  
-save_for_forward(*tensors)[source]
                                  
+save_for_forward(*tensors)[source]
 
                                  Save given tensors for a future call to jvp().
                                  
 
                                  save_for_forward should be called at most once, in either the
 setup_context() or forward() methods, and all arguments
@@ -740,7 +740,7 @@ 
                                  NestedIOFunction
 
                                  
-set_materialize_grads(value)[source]
                                  
+set_materialize_grads(value)[source]
 
                                  Set whether to materialize grad tensors. Default is True.
                                  
 
                                  This should be called only from either the setup_context() or
 forward() methods.
                                  
@@ -788,7 +788,7 @@ 
                                  NestedIOFunction
 
                                  
-static setup_context(ctx, inputs, output)[source]
                                  
+static setup_context(ctx, inputs, output)[source]
 
                                  There are two ways to define the forward pass of an autograd.Function.
                                  
 
                                  Either:
                                  
 
                                    
@@ -809,7 +809,7 @@ 
                                    NestedIOFunction
 
                                    
-static vjp(ctx, *grad_outputs)[source]
                                    
+static vjp(ctx, *grad_outputs)[source]
 
                                    Define a formula for differentiating the operation with backward mode automatic differentiation.
                                    
 
                                    This function is to be overridden by all subclasses.
 (Defining this function is equivalent to defining the vjp function.)
                                    
@@ -836,7 +836,7 @@ 
                                    NestedIOFunction
 
                                    
-static vmap(info, in_dims, *args)[source]
                                    
+static vmap(info, in_dims, *args)[source]
 
                                    Define the behavior for this autograd.Function underneath torch.vmap().
                                    
 
                                    For a torch.autograd.Function() to support
 torch.vmap(), you must either override this static method, or set
diff --git a/2.7/generated/torch.autograd.function.once_differentiable.html b/2.7/generated/torch.autograd.function.once_differentiable.html
index c88b536eea7..173b134733f 100644
--- a/2.7/generated/torch.autograd.function.once_differentiable.html
+++ b/2.7/generated/torch.autograd.function.once_differentiable.html
@@ -596,7 +596,7 @@
 
                                    torch.autograd.function.once_differentiable
                                    
 
                                    
 
                                    
-torch.autograd.function.once_differentiable(fn)[source][source]
                                    
+torch.autograd.function.once_differentiable(fn)[source][source]
 
                                    
 
 
                                
diff --git a/2.7/generated/torch.autograd.functional.hessian.html b/2.7/generated/torch.autograd.functional.hessian.html
index 66c8293c9df..bc8373e6232 100644
--- a/2.7/generated/torch.autograd.functional.hessian.html
+++ b/2.7/generated/torch.autograd.functional.hessian.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.hessian
                                
 
                                
 
                                
-torch.autograd.functional.hessian(func, inputs, create_graph=False, strict=False, vectorize=False, outer_jacobian_strategy='reverse-mode')[source][source]
                                
+torch.autograd.functional.hessian(func, inputs, create_graph=False, strict=False, vectorize=False, outer_jacobian_strategy='reverse-mode')[source][source]
 
                                Compute the Hessian of a given scalar function.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.functional.hvp.html b/2.7/generated/torch.autograd.functional.hvp.html
index 939c042b5ab..01741f5f67c 100644
--- a/2.7/generated/torch.autograd.functional.hvp.html
+++ b/2.7/generated/torch.autograd.functional.hvp.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.hvp
                                
 
                                
 
                                
-torch.autograd.functional.hvp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
                                
+torch.autograd.functional.hvp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
 
                                Compute the dot product between the scalar function’s Hessian and a vector v at a specified point.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.functional.jacobian.html b/2.7/generated/torch.autograd.functional.jacobian.html
index 32b361484e7..1c8c87a9ab8 100644
--- a/2.7/generated/torch.autograd.functional.jacobian.html
+++ b/2.7/generated/torch.autograd.functional.jacobian.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.jacobian
                                
 
                                
 
                                
-torch.autograd.functional.jacobian(func, inputs, create_graph=False, strict=False, vectorize=False, strategy='reverse-mode')[source][source]
                                
+torch.autograd.functional.jacobian(func, inputs, create_graph=False, strict=False, vectorize=False, strategy='reverse-mode')[source][source]
 
                                Compute the Jacobian of a given function.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.functional.jvp.html b/2.7/generated/torch.autograd.functional.jvp.html
index c2931ddd360..2ea0b1f2293 100644
--- a/2.7/generated/torch.autograd.functional.jvp.html
+++ b/2.7/generated/torch.autograd.functional.jvp.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.jvp
                                
 
                                
 
                                
-torch.autograd.functional.jvp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
                                
+torch.autograd.functional.jvp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
 
                                Compute the dot product between the Jacobian of the given function at the point given by the inputs and a vector v.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.functional.vhp.html b/2.7/generated/torch.autograd.functional.vhp.html
index 3976d2ff3b9..08b85a3e1a7 100644
--- a/2.7/generated/torch.autograd.functional.vhp.html
+++ b/2.7/generated/torch.autograd.functional.vhp.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.vhp
                                
 
                                
 
                                
-torch.autograd.functional.vhp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
                                
+torch.autograd.functional.vhp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
 
                                Compute the dot product between vector v and Hessian of a  given scalar function at a specified point.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.functional.vjp.html b/2.7/generated/torch.autograd.functional.vjp.html
index 0b7cd2fabe9..2e18e8ee3ff 100644
--- a/2.7/generated/torch.autograd.functional.vjp.html
+++ b/2.7/generated/torch.autograd.functional.vjp.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.functional.vjp
                                
 
                                
 
                                
-torch.autograd.functional.vjp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
                                
+torch.autograd.functional.vjp(func, inputs, v=None, create_graph=False, strict=False)[source][source]
 
                                Compute the dot product between a vector v and the Jacobian of the given function at the point given by the inputs.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.grad.html b/2.7/generated/torch.autograd.grad.html
index 123107d33e5..c127463ebe1 100644
--- a/2.7/generated/torch.autograd.grad.html
+++ b/2.7/generated/torch.autograd.grad.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.grad
                                
 
                                
 
                                
-torch.autograd.grad(outputs, inputs, grad_outputs=None, retain_graph=None, create_graph=False, only_inputs=True, allow_unused=None, is_grads_batched=False, materialize_grads=False)[source][source]
                                
+torch.autograd.grad(outputs, inputs, grad_outputs=None, retain_graph=None, create_graph=False, only_inputs=True, allow_unused=None, is_grads_batched=False, materialize_grads=False)[source][source]
 
                                Compute and return the sum of gradients of outputs with respect to the inputs.
                                
 
                                grad_outputs should be a sequence of length matching output
 containing the “vector” in vector-Jacobian product, usually the pre-computed
diff --git a/2.7/generated/torch.autograd.grad_mode.inference_mode.html b/2.7/generated/torch.autograd.grad_mode.inference_mode.html
index d778c47b204..93193d255f6 100644
--- a/2.7/generated/torch.autograd.grad_mode.inference_mode.html
+++ b/2.7/generated/torch.autograd.grad_mode.inference_mode.html
@@ -596,7 +596,7 @@
 
                                inference_mode
                                
 
                                
 
                                
-class torch.autograd.grad_mode.inference_mode(mode=True)[source][source]
                                
+class torch.autograd.grad_mode.inference_mode(mode=True)[source][source]
 
                                Context-manager that enables or disables inference mode.
                                
 
                                InferenceMode is a context manager analogous to no_grad
 to be used when you are certain your operations will have no interactions
@@ -649,7 +649,7 @@ 
                                inference_mode
 
                                
-clone()[source][source]
                                
+clone()[source][source]
 
                                Create a copy of this class
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.autograd.grad_mode.set_grad_enabled.html b/2.7/generated/torch.autograd.grad_mode.set_grad_enabled.html
index a9777a1538f..8eff06931a2 100644
--- a/2.7/generated/torch.autograd.grad_mode.set_grad_enabled.html
+++ b/2.7/generated/torch.autograd.grad_mode.set_grad_enabled.html
@@ -596,7 +596,7 @@
 
                                set_grad_enabled
                                
 
                                
 
                                
-class torch.autograd.grad_mode.set_grad_enabled(mode)[source][source]
                                
+class torch.autograd.grad_mode.set_grad_enabled(mode)[source][source]
 
                                Context-manager that sets gradient calculation on or off.
                                
 
                                set_grad_enabled will enable or disable grads based on its argument mode.
 It can be used as a context-manager or as a function.
                                
@@ -640,7 +640,7 @@ 
                                set_grad_enabled
 
                                
-clone()[source][source]
                                
+clone()[source][source]
 
                                Create a copy of this class
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.autograd.grad_mode.set_multithreading_enabled.html b/2.7/generated/torch.autograd.grad_mode.set_multithreading_enabled.html
index aa4f9365b4d..f0b17a12fa8 100644
--- a/2.7/generated/torch.autograd.grad_mode.set_multithreading_enabled.html
+++ b/2.7/generated/torch.autograd.grad_mode.set_multithreading_enabled.html
@@ -596,7 +596,7 @@
 
                                set_multithreading_enabled
                                
 
                                
 
                                
-class torch.autograd.grad_mode.set_multithreading_enabled(mode)[source][source]
                                
+class torch.autograd.grad_mode.set_multithreading_enabled(mode)[source][source]
 
                                Context-manager that sets multithreaded backwards on or off.
                                
 
                                set_multithreading_enabled will enable or disable multithreaded backwards based on its argument mode.
 It can be used as a context-manager or as a function.
                                
@@ -614,7 +614,7 @@ 
                                set_multithreading_enabled
 
                                
-clone()[source][source]
                                
+clone()[source][source]
 
                                Create a copy of this class
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.autograd.gradcheck.GradcheckError.html b/2.7/generated/torch.autograd.gradcheck.GradcheckError.html
index 5bdad4d1d27..ea1dfb47b63 100644
--- a/2.7/generated/torch.autograd.gradcheck.GradcheckError.html
+++ b/2.7/generated/torch.autograd.gradcheck.GradcheckError.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.gradcheck.GradcheckError
                                
 
                                
 
                                
-exception torch.autograd.gradcheck.GradcheckError[source][source]
                                
+exception torch.autograd.gradcheck.GradcheckError[source][source]
 
                                Error raised by gradcheck() and gradgradcheck().
                                
 
                                
 
diff --git a/2.7/generated/torch.autograd.gradcheck.gradcheck.html b/2.7/generated/torch.autograd.gradcheck.gradcheck.html
index 4586f3890c1..d7540c2859f 100644
--- a/2.7/generated/torch.autograd.gradcheck.gradcheck.html
+++ b/2.7/generated/torch.autograd.gradcheck.gradcheck.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.gradcheck.gradcheck
                                
 
                                
 
                                
-torch.autograd.gradcheck.gradcheck(func, inputs, *, eps=1e-06, atol=1e-05, rtol=0.001, raise_exception=True, nondet_tol=0.0, check_undefined_grad=True, check_grad_dtypes=False, check_batched_grad=False, check_batched_forward_grad=False, check_forward_ad=False, check_backward_ad=True, fast_mode=False, masked=None)[source][source]
                                
+torch.autograd.gradcheck.gradcheck(func, inputs, *, eps=1e-06, atol=1e-05, rtol=0.001, raise_exception=True, nondet_tol=0.0, check_undefined_grad=True, check_grad_dtypes=False, check_batched_grad=False, check_batched_forward_grad=False, check_forward_ad=False, check_backward_ad=True, fast_mode=False, masked=None)[source][source]
 
                                Check gradients computed via small finite differences against analytical
 gradients wrt tensors in inputs that are of floating point or complex type
 and with requires_grad=True.
                                
diff --git a/2.7/generated/torch.autograd.gradcheck.gradgradcheck.html b/2.7/generated/torch.autograd.gradcheck.gradgradcheck.html
index c0124cb6b49..72ce4f38062 100644
--- a/2.7/generated/torch.autograd.gradcheck.gradgradcheck.html
+++ b/2.7/generated/torch.autograd.gradcheck.gradgradcheck.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.gradcheck.gradgradcheck
                                
 
                                
 
                                
-torch.autograd.gradcheck.gradgradcheck(func, inputs, grad_outputs=None, *, eps=1e-06, atol=1e-05, rtol=0.001, gen_non_contig_grad_outputs=False, raise_exception=True, nondet_tol=0.0, check_undefined_grad=True, check_grad_dtypes=False, check_batched_grad=False, check_fwd_over_rev=False, check_rev_over_rev=True, fast_mode=False, masked=False)[source][source]
                                
+torch.autograd.gradcheck.gradgradcheck(func, inputs, grad_outputs=None, *, eps=1e-06, atol=1e-05, rtol=0.001, gen_non_contig_grad_outputs=False, raise_exception=True, nondet_tol=0.0, check_undefined_grad=True, check_grad_dtypes=False, check_batched_grad=False, check_fwd_over_rev=False, check_rev_over_rev=True, fast_mode=False, masked=False)[source][source]
 
                                Check gradients of gradients computed via small finite differences
 against analytical gradients wrt tensors in inputs and
 grad_outputs that are of floating point or complex type and with
diff --git a/2.7/generated/torch.autograd.graph.Node.metadata.html b/2.7/generated/torch.autograd.graph.Node.metadata.html
index 761130c3752..f2fea2812d9 100644
--- a/2.7/generated/torch.autograd.graph.Node.metadata.html
+++ b/2.7/generated/torch.autograd.graph.Node.metadata.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.graph.Node.metadata
                                
 
                                
 
                                
-abstract Node.metadata()[source][source]
                                
+abstract Node.metadata()[source][source]
 
                                Return the metadata.
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.autograd.graph.Node.name.html b/2.7/generated/torch.autograd.graph.Node.name.html
index ec154d44e41..2dc2b22743e 100644
--- a/2.7/generated/torch.autograd.graph.Node.name.html
+++ b/2.7/generated/torch.autograd.graph.Node.name.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.graph.Node.name
                                
 
                                
 
                                
-abstract Node.name()[source][source]
                                
+abstract Node.name()[source][source]
 
                                Return the name.
                                
 
                                Example:
                                
 
                                >>> import torch
diff --git a/2.7/generated/torch.autograd.graph.Node.register_hook.html b/2.7/generated/torch.autograd.graph.Node.register_hook.html
index 7c154f4e0a6..166e7a3424d 100644
--- a/2.7/generated/torch.autograd.graph.Node.register_hook.html
+++ b/2.7/generated/torch.autograd.graph.Node.register_hook.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.graph.Node.register_hook
                                
 
                                
 
                                
-abstract Node.register_hook(fn)[source][source]
                                
+abstract Node.register_hook(fn)[source][source]
 
                                Register a backward hook.
                                
 
                                The hook will be called every time a gradient with respect to the
 Node is computed. The hook should have the following signature:
                                
diff --git a/2.7/generated/torch.autograd.graph.Node.register_prehook.html b/2.7/generated/torch.autograd.graph.Node.register_prehook.html
index e82ff7b443b..01d5ca069fe 100644
--- a/2.7/generated/torch.autograd.graph.Node.register_prehook.html
+++ b/2.7/generated/torch.autograd.graph.Node.register_prehook.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.graph.Node.register_prehook
                                
 
                                
 
                                
-abstract Node.register_prehook(fn)[source][source]
                                
+abstract Node.register_prehook(fn)[source][source]
 
                                Register a backward pre-hook.
                                
 
                                The hook will be called every time a gradient with respect to the
 Node is computed. The hook should have the following signature:
                                
diff --git a/2.7/generated/torch.autograd.graph.increment_version.html b/2.7/generated/torch.autograd.graph.increment_version.html
index 73eb6ea3522..1328cda5d1a 100644
--- a/2.7/generated/torch.autograd.graph.increment_version.html
+++ b/2.7/generated/torch.autograd.graph.increment_version.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.graph.increment_version
                                
 
                                
 
                                
-torch.autograd.graph.increment_version(tensor)[source][source]
                                
+torch.autograd.graph.increment_version(tensor)[source][source]
 
                                Update autograd metadata tracking whether the given Tensor was modified in place.
                                
 
                                This is to enable more accurate error checking within the autograd engine.
 It is already done automatically by PyTorch functions and within custom Function
diff --git a/2.7/generated/torch.autograd.profiler.EnforceUnique.html b/2.7/generated/torch.autograd.profiler.EnforceUnique.html
index 9fb3421d637..67592cf1ec4 100644
--- a/2.7/generated/torch.autograd.profiler.EnforceUnique.html
+++ b/2.7/generated/torch.autograd.profiler.EnforceUnique.html
@@ -596,11 +596,11 @@
 
                                EnforceUnique
                                
 
                                
 
                                
-class torch.autograd.profiler.EnforceUnique[source][source]
                                
+class torch.autograd.profiler.EnforceUnique[source][source]
 
                                Raises an error if a key is seen more than once.
                                
 
                                
 
                                
-see(*key)[source][source]
                                
+see(*key)[source][source]
 
                                Observe a key and raise an error if it is seen multiple times.
                                
 
                                
 
diff --git a/2.7/generated/torch.autograd.profiler.KinetoStepTracker.html b/2.7/generated/torch.autograd.profiler.KinetoStepTracker.html
index 208927b5bb1..4a9dc53dab9 100644
--- a/2.7/generated/torch.autograd.profiler.KinetoStepTracker.html
+++ b/2.7/generated/torch.autograd.profiler.KinetoStepTracker.html
@@ -596,7 +596,7 @@
 
                                KinetoStepTracker
                                
 
                                
 
                                
-class torch.autograd.profiler.KinetoStepTracker[source][source]
                                
+class torch.autograd.profiler.KinetoStepTracker[source][source]
 
                                Provides an abstraction for incrementing the step count globally.
                                
 
                                Previously, we only had one place to mark that a step() has occurred
 in the program via pytorch profiler step(). We will now add step hooks
@@ -631,7 +631,7 @@ 
                                KinetoStepTracker
 
                                
-classmethod current_step()[source][source]
                                
+classmethod current_step()[source][source]
 
                                Get the latest step for any requester
                                
 
                                
 
                                Return type
                                
@@ -642,7 +642,7 @@ 
                                KinetoStepTracker
 
                                
-classmethod erase_step_count(requester)[source][source]
                                
+classmethod erase_step_count(requester)[source][source]
 
                                Remove a given requester.
                                
 
                                
 
                                Return type
                                
@@ -653,7 +653,7 @@ 
                                KinetoStepTracker
 
                                
-classmethod increment_step(requester)[source][source]
                                
+classmethod increment_step(requester)[source][source]
 
                                Increments the step count for the requester.
                                
 
                                Additionally if the max over all step counts has incremented then
 trigger the _kineto_step() returns global step count
                                
@@ -666,7 +666,7 @@ 
                                KinetoStepTracker
 
                                
-classmethod init_step_count(requester)[source][source]
                                
+classmethod init_step_count(requester)[source][source]
 
                                Initialize for a given requester.
                                
 
                                
 
                                
diff --git a/2.7/generated/torch.autograd.profiler.load_nvprof.html b/2.7/generated/torch.autograd.profiler.load_nvprof.html
index 2c0a7bdf276..5a55eb7de2d 100644
--- a/2.7/generated/torch.autograd.profiler.load_nvprof.html
+++ b/2.7/generated/torch.autograd.profiler.load_nvprof.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.profiler.load_nvprof
                                
 
                                
 
                                
-torch.autograd.profiler.load_nvprof(path)[source][source]
                                
+torch.autograd.profiler.load_nvprof(path)[source][source]
 
                                Open an nvprof trace file and parses autograd annotations.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.profiler.parse_nvprof_trace.html b/2.7/generated/torch.autograd.profiler.parse_nvprof_trace.html
index 1da064f44fd..64f7e4fee98 100644
--- a/2.7/generated/torch.autograd.profiler.parse_nvprof_trace.html
+++ b/2.7/generated/torch.autograd.profiler.parse_nvprof_trace.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.profiler.parse_nvprof_trace
                                
 
                                
 
                                
-torch.autograd.profiler.parse_nvprof_trace(path)[source][source]
                                
+torch.autograd.profiler.parse_nvprof_trace(path)[source][source]
 
                                
 
 
                                
diff --git a/2.7/generated/torch.autograd.profiler.profile.export_chrome_trace.html b/2.7/generated/torch.autograd.profiler.profile.export_chrome_trace.html
index dcea857dd84..121077666f3 100644
--- a/2.7/generated/torch.autograd.profiler.profile.export_chrome_trace.html
+++ b/2.7/generated/torch.autograd.profiler.profile.export_chrome_trace.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.profiler.profile.export_chrome_trace
                                
 
                                
 
                                
-profile.export_chrome_trace(path)[source][source]
                                
+profile.export_chrome_trace(path)[source][source]
 
                                Export an EventList as a Chrome tracing tools file.
                                
 
                                The checkpoint can be later loaded and inspected under chrome://tracing URL.
                                
 
                                
diff --git a/2.7/generated/torch.autograd.profiler.profile.key_averages.html b/2.7/generated/torch.autograd.profiler.profile.key_averages.html
index 3f90416063a..f89cba10b93 100644
--- a/2.7/generated/torch.autograd.profiler.profile.key_averages.html
+++ b/2.7/generated/torch.autograd.profiler.profile.key_averages.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.profiler.profile.key_averages
                                
 
                                
 
                                
-profile.key_averages(group_by_input_shape=False, group_by_stack_n=0, group_by_overload_name=False)[source][source]
                                
+profile.key_averages(group_by_input_shape=False, group_by_stack_n=0, group_by_overload_name=False)[source][source]
 
                                Averages all function events over their keys.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.autograd.profiler.profile.total_average.html b/2.7/generated/torch.autograd.profiler.profile.total_average.html
index b84d2388e18..65647a43cce 100644
--- a/2.7/generated/torch.autograd.profiler.profile.total_average.html
+++ b/2.7/generated/torch.autograd.profiler.profile.total_average.html
@@ -596,7 +596,7 @@
 
                                torch.autograd.profiler.profile.total_average
                                
 
                                
 
                                
-profile.total_average()[source][source]
                                
+profile.total_average()[source][source]
 
                                Averages all events.
                                
 
                                
 
                                Returns
                                
diff --git a/2.7/generated/torch.autograd.profiler.record_function.html b/2.7/generated/torch.autograd.profiler.record_function.html
index 46b76322898..c9c68511bc0 100644
--- a/2.7/generated/torch.autograd.profiler.record_function.html
+++ b/2.7/generated/torch.autograd.profiler.record_function.html
@@ -596,7 +596,7 @@
 
                                record_function
                                
 
                                
 
                                
-class torch.autograd.profiler.record_function(name, args=None)[source][source]
                                
+class torch.autograd.profiler.record_function(name, args=None)[source][source]
 
                                Context manager/function decorator that adds a label to a code block/function when running autograd profiler.
 Label will only appear if CPU activity tracing is enabled.
                                
 
                                It is useful when tracing the code profile.
                                
diff --git a/2.7/generated/torch.autograd.profiler_util.Interval.html b/2.7/generated/torch.autograd.profiler_util.Interval.html
index aaced5f788c..3bbef34b4e6 100644
--- a/2.7/generated/torch.autograd.profiler_util.Interval.html
+++ b/2.7/generated/torch.autograd.profiler_util.Interval.html
@@ -596,10 +596,10 @@
 
                                Interval
                                
 
                                
 
                                
-class torch.autograd.profiler_util.Interval(start, end)[source][source]
                                
+class torch.autograd.profiler_util.Interval(start, end)[source][source]
 
                                
 
                                
-elapsed_us()[source][source]
                                
+elapsed_us()[source][source]
 
                                Returns the length of the interval
                                
 
                                
 
diff --git a/2.7/generated/torch.autograd.profiler_util.MemRecordsAcc.html b/2.7/generated/torch.autograd.profiler_util.MemRecordsAcc.html
index 415a37aa04c..4493572c035 100644
--- a/2.7/generated/torch.autograd.profiler_util.MemRecordsAcc.html
+++ b/2.7/generated/torch.autograd.profiler_util.MemRecordsAcc.html
@@ -596,11 +596,11 @@
 
                                MemRecordsAcc
                                
 
                                
 
                                
-class torch.autograd.profiler_util.MemRecordsAcc(mem_records)[source][source]
                                
+class torch.autograd.profiler_util.MemRecordsAcc(mem_records)[source][source]
 
                                Acceleration structure for accessing mem_records in interval.
                                
 
                                
 
                                
-in_interval(start_us, end_us)[source][source]
                                
+in_interval(start_us, end_us)[source][source]
 
                                Return all records in the given interval
 To maintain backward compatibility, convert us to ns in function
                                
 
                                
diff --git a/2.7/generated/torch.autograd.profiler_util.StringTable.html b/2.7/generated/torch.autograd.profiler_util.StringTable.html
index e76dd0375f9..0d61228fd3b 100644
--- a/2.7/generated/torch.autograd.profiler_util.StringTable.html
+++ b/2.7/generated/torch.autograd.profiler_util.StringTable.html
@@ -596,7 +596,7 @@
 
                                StringTable
                                
 
                                
 
                                
-class torch.autograd.profiler_util.StringTable[source][source]
                                
+class torch.autograd.profiler_util.StringTable[source][source]
 
                                
 
                                
 clear() → None.  Remove all items from D.
                                
diff --git a/2.7/generated/torch.block_diag.html b/2.7/generated/torch.block_diag.html
index 8506539739d..b16e1b0d9ff 100644
--- a/2.7/generated/torch.block_diag.html
+++ b/2.7/generated/torch.block_diag.html
@@ -596,7 +596,7 @@
 
                                torch.block_diag
                                
 
                                
 
                                
-torch.block_diag(*tensors)[source][source]
                                
+torch.block_diag(*tensors)[source][source]
 
                                Create a block diagonal matrix from provided tensors.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.broadcast_shapes.html b/2.7/generated/torch.broadcast_shapes.html
index e3a7e71b4a7..82775303205 100644
--- a/2.7/generated/torch.broadcast_shapes.html
+++ b/2.7/generated/torch.broadcast_shapes.html
@@ -596,7 +596,7 @@
 
                                torch.broadcast_shapes
                                
 
                                
 
                                
-torch.broadcast_shapes(*shapes) → Size[source][source]
                                
+torch.broadcast_shapes(*shapes) → Size[source][source]
 
                                Similar to broadcast_tensors() but for shapes.
                                
 
                                This is equivalent to
 torch.broadcast_tensors(*map(torch.empty, shapes))[0].shape
diff --git a/2.7/generated/torch.broadcast_tensors.html b/2.7/generated/torch.broadcast_tensors.html
index 102685f53b0..bda69a412a9 100644
--- a/2.7/generated/torch.broadcast_tensors.html
+++ b/2.7/generated/torch.broadcast_tensors.html
@@ -596,7 +596,7 @@
 
                                torch.broadcast_tensors
                                
 
                                
 
                                
-torch.broadcast_tensors(*tensors) → List of Tensors[source][source]
                                
+torch.broadcast_tensors(*tensors) → List of Tensors[source][source]
 
                                Broadcasts the given tensors according to Broadcasting semantics.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.cartesian_prod.html b/2.7/generated/torch.cartesian_prod.html
index 7a2f91016ef..c25d556c662 100644
--- a/2.7/generated/torch.cartesian_prod.html
+++ b/2.7/generated/torch.cartesian_prod.html
@@ -596,7 +596,7 @@
 
                                torch.cartesian_prod
                                
 
                                
 
                                
-torch.cartesian_prod(*tensors)[source][source]
                                
+torch.cartesian_prod(*tensors)[source][source]
 
                                Do cartesian product of the given sequence of tensors. The behavior is similar to
 python’s itertools.product.
                                
 
                                
diff --git a/2.7/generated/torch.cdist.html b/2.7/generated/torch.cdist.html
index 8617b10d41a..89084627fb8 100644
--- a/2.7/generated/torch.cdist.html
+++ b/2.7/generated/torch.cdist.html
@@ -596,7 +596,7 @@
 
                                torch.cdist
                                
 
                                
 
                                
-torch.cdist(x1, x2, p=2.0, compute_mode='use_mm_for_euclid_dist_if_necessary')[source][source]
                                
+torch.cdist(x1, x2, p=2.0, compute_mode='use_mm_for_euclid_dist_if_necessary')[source][source]
 
                                Computes batched the p-norm distance between each pair of the two collections of row vectors.
                                
 
                                
 
                                Parameters
                                
diff --git a/2.7/generated/torch.chain_matmul.html b/2.7/generated/torch.chain_matmul.html
index acc3e5437a0..d8baa60a989 100644
--- a/2.7/generated/torch.chain_matmul.html
+++ b/2.7/generated/torch.chain_matmul.html
@@ -596,7 +596,7 @@
 
                                torch.chain_matmul
                                
 
                                
 
                                
-torch.chain_matmul(*matrices, out=None)[source][source]
                                
+torch.chain_matmul(*matrices, out=None)[source][source]
 
                                Returns the matrix product of the NN 2-D tensors. This product is efficiently computed
 using the matrix chain order algorithm which selects the order in which incurs the lowest cost in terms
 of arithmetic operations ([CLRS]). Note that since this is a function to compute the product, NN
diff --git a/2.7/generated/torch.compile.html b/2.7/generated/torch.compile.html
index ae708328d89..62f01739721 100644
--- a/2.7/generated/torch.compile.html
+++ b/2.7/generated/torch.compile.html
@@ -596,7 +596,7 @@
 
                                torch.compile
                                
 
                                
 
                                
-torch.compile(model: Callable[[_InputT], _RetT], *, fullgraph: bool = False, dynamic: Optional[bool] = None, backend: Union[str, Callable] = 'inductor', mode: Optional[str] = None, options: Optional[dict[str, Union[str, int, bool]]] = None, disable: bool = False) → Callable[[_InputT], _RetT][source][source]
                                
+torch.compile(model: Callable[[_InputT], _RetT], *, fullgraph: bool = False, dynamic: Optional[bool] = None, backend: Union[str, Callable] = 'inductor', mode: Optional[str] = None, options: Optional[dict[str, Union[str, int, bool]]] = None, disable: bool = False) → Callable[[_InputT], _RetT][source][source]
 
                                
 torch.compile(model: None = None, *, fullgraph: bool = False, dynamic: Optional[bool] = None, backend: Union[str, Callable] = 'inductor', mode: Optional[str] = None, options: Optional[dict[str, Union[str, int, bool]]] = None, disable: bool = False) → Callable[[Callable[[_InputT], _RetT]], Callable[[_InputT], _RetT]]
                                
 
                                Optimizes given model/function using TorchDynamo and specified backend.
diff --git a/2.7/generated/torch.compiled_with_cxx11_abi.html b/2.7/generated/torch.compiled_with_cxx11_abi.html
index e638db4940f..ba7b751a599 100644
--- a/2.7/generated/torch.compiled_with_cxx11_abi.html
+++ b/2.7/generated/torch.compiled_with_cxx11_abi.html
@@ -596,7 +596,7 @@
 
                                torch.compiled_with_cxx11_abi
                                
 
                                
 
                                
-torch.compiled_with_cxx11_abi()[source][source]
                                
+torch.compiled_with_cxx11_abi()[source][source]
 
                                Returns whether PyTorch was built with _GLIBCXX_USE_CXX11_ABI=1
                                
 
                                
 
                                Return type
                                
diff --git a/2.7/generated/torch.compiler.allow_in_graph.html b/2.7/generated/torch.compiler.allow_in_graph.html
index 4f757661ea9..3b015dd929f 100644
--- a/2.7/generated/torch.compiler.allow_in_graph.html
+++ b/2.7/generated/torch.compiler.allow_in_graph.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.allow_in_graph
                                
 
                                
 
                                
-torch.compiler.allow_in_graph(fn)[source][source]
                                
+torch.compiler.allow_in_graph(fn)[source][source]
 
                                Tells the compiler frontend (Dynamo) to skip symbolic introspection of the function
 and instead directly write it to the graph when encountered.
                                
 
                                If you are using torch.compile() (with backend=”inductor” (the default)), or
diff --git a/2.7/generated/torch.compiler.assume_constant_result.html b/2.7/generated/torch.compiler.assume_constant_result.html
index 53aab9eea41..2b09ec21430 100644
--- a/2.7/generated/torch.compiler.assume_constant_result.html
+++ b/2.7/generated/torch.compiler.assume_constant_result.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.assume_constant_result
                                
 
                                
 
                                
-torch.compiler.assume_constant_result(fn)[source][source]
                                
+torch.compiler.assume_constant_result(fn)[source][source]
 
                                This function is used to mark a function fn as having a constant result.
 This allows the compiler to optimize away your function.
 Returns The same function fn
                                
diff --git a/2.7/generated/torch.compiler.compile.html b/2.7/generated/torch.compiler.compile.html
index 723e19fc0a8..ea6808fe0f7 100644
--- a/2.7/generated/torch.compiler.compile.html
+++ b/2.7/generated/torch.compiler.compile.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.compile
                                
 
                                
 
                                
-torch.compiler.compile(*args, **kwargs)[source][source]
                                
+torch.compiler.compile(*args, **kwargs)[source][source]
 
                                See torch.compile() for details on the arguments for this function.
                                
 
                                
 
diff --git a/2.7/generated/torch.compiler.cudagraph_mark_step_begin.html b/2.7/generated/torch.compiler.cudagraph_mark_step_begin.html
index 4a2d7f6926c..9699558eef5 100644
--- a/2.7/generated/torch.compiler.cudagraph_mark_step_begin.html
+++ b/2.7/generated/torch.compiler.cudagraph_mark_step_begin.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.cudagraph_mark_step_begin
                                
 
                                
 
                                
-torch.compiler.cudagraph_mark_step_begin()[source][source]
                                
+torch.compiler.cudagraph_mark_step_begin()[source][source]
 
                                Indicates that a new iteration of inference or training is about to begin.
                                
 
                                CUDA Graphs will free tensors of a prior iteration. A new iteration is started on each invocation of
 torch.compile, so long as there is not a pending backward that has not been called.
                                
diff --git a/2.7/generated/torch.compiler.disable.html b/2.7/generated/torch.compiler.disable.html
index 42935988269..8b1795ed19b 100644
--- a/2.7/generated/torch.compiler.disable.html
+++ b/2.7/generated/torch.compiler.disable.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.disable
                                
 
                                
 
                                
-torch.compiler.disable(fn=None, recursive=True)[source][source]
                                
+torch.compiler.disable(fn=None, recursive=True)[source][source]
 
                                This function provides a decorator to disable compilation on a function.
 It also provides the option of recursively disabling called functions.
                                
 
                                
diff --git a/2.7/generated/torch.compiler.is_compiling.html b/2.7/generated/torch.compiler.is_compiling.html
index f3b72c73000..bc8f1fa4081 100644
--- a/2.7/generated/torch.compiler.is_compiling.html
+++ b/2.7/generated/torch.compiler.is_compiling.html
@@ -598,7 +598,7 @@
 
                                torch.compiler.is_compiling
                                
 
                                
 
                                
-torch.compiler.is_compiling()[source][source]
                                
+torch.compiler.is_compiling()[source][source]
 
                                Indicates whether a graph is executed/traced as part of torch.compile() or torch.export().
                                
 
                                
 
                                Note that there are 2 other related flags that should deprecated eventually:
                                  
diff --git a/2.7/generated/torch.compiler.is_dynamo_compiling.html b/2.7/generated/torch.compiler.is_dynamo_compiling.html
index 84ece3d5037..af6d5f04c72 100644
--- a/2.7/generated/torch.compiler.is_dynamo_compiling.html
+++ b/2.7/generated/torch.compiler.is_dynamo_compiling.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.is_dynamo_compiling
                                  
 
                                  
 
                                  
-torch.compiler.is_dynamo_compiling()[source][source]
                                  
+torch.compiler.is_dynamo_compiling()[source][source]
 
                                  Indicates whether a graph is traced via TorchDynamo.
                                  
 
                                  It’s stricter than is_compiling() flag, as it would only be set to True when
 TorchDynamo is used.
                                  
diff --git a/2.7/generated/torch.compiler.is_exporting.html b/2.7/generated/torch.compiler.is_exporting.html
index 78652c662a0..07136ffdc65 100644
--- a/2.7/generated/torch.compiler.is_exporting.html
+++ b/2.7/generated/torch.compiler.is_exporting.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.is_exporting
                                  
 
                                  
 
                                  
-torch.compiler.is_exporting()[source][source]
                                  
+torch.compiler.is_exporting()[source][source]
 
                                  Indicated whether we’re under exporting.
                                  
 
                                  It’s stricter than is_compiling() flag, as it would only be set to True when
 torch.export is used.
                                  
diff --git a/2.7/generated/torch.compiler.list_backends.html b/2.7/generated/torch.compiler.list_backends.html
index dcdb60f6d14..4fa5aa0ddd6 100644
--- a/2.7/generated/torch.compiler.list_backends.html
+++ b/2.7/generated/torch.compiler.list_backends.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.list_backends
                                  
 
                                  
 
                                  
-torch.compiler.list_backends(exclude_tags=('debug', 'experimental'))[source][source]
                                  
+torch.compiler.list_backends(exclude_tags=('debug', 'experimental'))[source][source]
 
                                  Return valid strings that can be passed to torch.compile(…, backend=”name”).
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.compiler.reset.html b/2.7/generated/torch.compiler.reset.html
index 0d73fb92dae..7fbee929d12 100644
--- a/2.7/generated/torch.compiler.reset.html
+++ b/2.7/generated/torch.compiler.reset.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.reset
                                  
 
                                  
 
                                  
-torch.compiler.reset()[source][source]
                                  
+torch.compiler.reset()[source][source]
 
                                  This function clears all compilation caches and restores the system to its initial state.
 It is recommended to call this function, especially after using operations like torch.compile(…)
 to ensure a clean state before another unrelated compilation
                                  
diff --git a/2.7/generated/torch.compiler.set_stance.html b/2.7/generated/torch.compiler.set_stance.html
index f63023a600a..d0ff7dab7ca 100644
--- a/2.7/generated/torch.compiler.set_stance.html
+++ b/2.7/generated/torch.compiler.set_stance.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.set_stance
                                  
 
                                  
 
                                  
-torch.compiler.set_stance(stance='default', *, skip_guard_eval_unsafe=False, force_backend=None)[source][source]
                                  
+torch.compiler.set_stance(stance='default', *, skip_guard_eval_unsafe=False, force_backend=None)[source][source]
 
                                  Set the current stance of the compiler.
 Can be used as a function, context manager, or decorator.
 Do not use this function inside a torch.compile region - an error will be raised otherwise.
                                  
diff --git a/2.7/generated/torch.compiler.substitute_in_graph.html b/2.7/generated/torch.compiler.substitute_in_graph.html
index 3f38d542f0d..d9ae1b3ee4d 100644
--- a/2.7/generated/torch.compiler.substitute_in_graph.html
+++ b/2.7/generated/torch.compiler.substitute_in_graph.html
@@ -598,7 +598,7 @@
 
                                  torch.compiler.substitute_in_graph
                                  
 
                                  
 
                                  
-torch.compiler.substitute_in_graph(original_fn, *, can_constant_fold_through=False, skip_signature_check=False)[source][source]
                                  
+torch.compiler.substitute_in_graph(original_fn, *, can_constant_fold_through=False, skip_signature_check=False)[source][source]
 
                                  Register a polyfill handler for a function, usually a C function from the C extension, to be
 used in place of the original function when inlining the original function in the graph.
                                  
 
                                  
diff --git a/2.7/generated/torch.cond.html b/2.7/generated/torch.cond.html
index d47e6f56111..dc04304cd1c 100644
--- a/2.7/generated/torch.cond.html
+++ b/2.7/generated/torch.cond.html
@@ -596,7 +596,7 @@
 
                                  torch.cond
                                  
 
                                  
 
                                  
-torch.cond(pred, true_fn, false_fn, operands=())[source]
                                  
+torch.cond(pred, true_fn, false_fn, operands=())[source]
 
                                  Conditionally applies true_fn or false_fn.
                                  
 
                                  
 
                                  Warning
                                  
diff --git a/2.7/generated/torch.cpu.Stream.html b/2.7/generated/torch.cpu.Stream.html
index 217fefdd1c4..c6c8627c695 100644
--- a/2.7/generated/torch.cpu.Stream.html
+++ b/2.7/generated/torch.cpu.Stream.html
@@ -596,7 +596,7 @@
 
                                  Stream
                                  
 
                                  
 
                                  
-class torch.cpu.Stream(priority=-1)[source][source]
                                  
+class torch.cpu.Stream(priority=-1)[source][source]
 
                                  N.B. This class only exists to facilitate device-agnostic code
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.StreamContext.html b/2.7/generated/torch.cpu.StreamContext.html
index dfe88f3259e..d18779755c3 100644
--- a/2.7/generated/torch.cpu.StreamContext.html
+++ b/2.7/generated/torch.cpu.StreamContext.html
@@ -596,7 +596,7 @@
 
                                  StreamContext
                                  
 
                                  
 
                                  
-class torch.cpu.StreamContext(stream)[source][source]
                                  
+class torch.cpu.StreamContext(stream)[source][source]
 
                                  Context-manager that selects a given stream.
                                  
 
                                  N.B. This class only exists to facilitate device-agnostic code
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.current_device.html b/2.7/generated/torch.cpu.current_device.html
index c27db192ce9..5ec2fecd7b4 100644
--- a/2.7/generated/torch.cpu.current_device.html
+++ b/2.7/generated/torch.cpu.current_device.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.current_device
                                  
 
                                  
 
                                  
-torch.cpu.current_device()[source][source]
                                  
+torch.cpu.current_device()[source][source]
 
                                  Returns current device for cpu. Always ‘cpu’.
                                  
 
                                  N.B. This function only exists to facilitate device-agnostic code
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.current_stream.html b/2.7/generated/torch.cpu.current_stream.html
index b2fb63b440a..65120e0b70c 100644
--- a/2.7/generated/torch.cpu.current_stream.html
+++ b/2.7/generated/torch.cpu.current_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.current_stream
                                  
 
                                  
 
                                  
-torch.cpu.current_stream(device=None)[source][source]
                                  
+torch.cpu.current_stream(device=None)[source][source]
 
                                  Returns the currently selected Stream for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cpu.device_count.html b/2.7/generated/torch.cpu.device_count.html
index be91ed0c9fe..ce1a1233874 100644
--- a/2.7/generated/torch.cpu.device_count.html
+++ b/2.7/generated/torch.cpu.device_count.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.device_count
                                  
 
                                  
 
                                  
-torch.cpu.device_count()[source][source]
                                  
+torch.cpu.device_count()[source][source]
 
                                  Returns number of CPU devices (not cores). Always 1.
                                  
 
                                  N.B. This function only exists to facilitate device-agnostic code
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.is_available.html b/2.7/generated/torch.cpu.is_available.html
index 18f54a0cfba..4c7a448c169 100644
--- a/2.7/generated/torch.cpu.is_available.html
+++ b/2.7/generated/torch.cpu.is_available.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.is_available
                                  
 
                                  
 
                                  
-torch.cpu.is_available()[source][source]
                                  
+torch.cpu.is_available()[source][source]
 
                                  Returns a bool indicating if CPU is currently available.
                                  
 
                                  N.B. This function only exists to facilitate device-agnostic code
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.set_device.html b/2.7/generated/torch.cpu.set_device.html
index 81ddb86d746..a0c770f0309 100644
--- a/2.7/generated/torch.cpu.set_device.html
+++ b/2.7/generated/torch.cpu.set_device.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.set_device
                                  
 
                                  
 
                                  
-torch.cpu.set_device(device)[source][source]
                                  
+torch.cpu.set_device(device)[source][source]
 
                                  Sets the current device, in CPU we do nothing.
                                  
 
                                  N.B. This function only exists to facilitate device-agnostic code
                                  
 
                                  
diff --git a/2.7/generated/torch.cpu.stream.html b/2.7/generated/torch.cpu.stream.html
index 6bfa96ca6bf..54df3d3d9ed 100644
--- a/2.7/generated/torch.cpu.stream.html
+++ b/2.7/generated/torch.cpu.stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.stream
                                  
 
                                  
 
                                  
-torch.cpu.stream(stream)[source][source]
                                  
+torch.cpu.stream(stream)[source][source]
 
                                  Wrapper around the Context-manager StreamContext that
 selects a given stream.
                                  
 
                                  N.B. This function only exists to facilitate device-agnostic code
                                  
diff --git a/2.7/generated/torch.cpu.synchronize.html b/2.7/generated/torch.cpu.synchronize.html
index b1344dbae74..8969dbb1e46 100644
--- a/2.7/generated/torch.cpu.synchronize.html
+++ b/2.7/generated/torch.cpu.synchronize.html
@@ -596,7 +596,7 @@
 
                                  torch.cpu.synchronize
                                  
 
                                  
 
                                  
-torch.cpu.synchronize(device=None)[source][source]
                                  
+torch.cpu.synchronize(device=None)[source][source]
 
                                  Waits for all kernels in all streams on the CPU device to complete.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.CUDAGraph.html b/2.7/generated/torch.cuda.CUDAGraph.html
index e03393095c1..ee11991f0b7 100644
--- a/2.7/generated/torch.cuda.CUDAGraph.html
+++ b/2.7/generated/torch.cuda.CUDAGraph.html
@@ -596,7 +596,7 @@
 
                                  CUDAGraph
                                  
 
                                  
 
                                  
-class torch.cuda.CUDAGraph[source][source]
                                  
+class torch.cuda.CUDAGraph[source][source]
 
                                  Wrapper around a CUDA graph.
                                  
 
                                  
 
                                  Warning
                                  
@@ -604,7 +604,7 @@ 
                                  CUDAGraph
 
                                  
-capture_begin(pool=None, capture_error_mode='global')[source][source]
                                  
+capture_begin(pool=None, capture_error_mode='global')[source][source]
 
                                  Begin capturing CUDA work on the current stream.
                                  
 
                                  Typically, you shouldn’t call capture_begin yourself.
 Use graph or make_graphed_callables(),
@@ -627,7 +627,7 @@ 
                                  CUDAGraph
 
                                  
-capture_end()[source][source]
                                  
+capture_end()[source][source]
 
                                  End CUDA graph capture on the current stream.
                                  
 
                                  After capture_end, replay may be called on this instance.
                                  
 
                                  Typically, you shouldn’t call capture_end yourself.
@@ -637,7 +637,7 @@ 
                                  CUDAGraph
 
                                  
-debug_dump(debug_path)[source][source]
                                  
+debug_dump(debug_path)[source][source]
 
                                  
 
                                  Parameters
                                  
 
                                  debug_path (required) – Path to dump the graph to.
                                  
@@ -649,13 +649,13 @@ 
                                  CUDAGraph
 
                                  
-enable_debug_mode()[source][source]
                                  
+enable_debug_mode()[source][source]
 
                                  Enable debugging mode for CUDAGraph.debug_dump.
                                  
 
                                  
 
 
                                  
 
                                  
-pool()[source][source]
                                  
+pool()[source][source]
 
                                  Return an opaque token representing the id of this graph’s memory pool.
                                  
 
                                  This id can optionally be passed to another graph’s capture_begin,
 which hints the other graph may share the same memory pool.
                                  
@@ -663,13 +663,13 @@ 
                                  CUDAGraph
 
                                  
-replay()[source][source]
                                  
+replay()[source][source]
 
                                  Replay the CUDA work captured by this graph.
                                  
 
                                  
 
 
                                  
 
                                  
-reset()[source][source]
                                  
+reset()[source][source]
 
                                  Delete the graph currently held by this instance.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.cuda.CUDAPluggableAllocator.html b/2.7/generated/torch.cuda.CUDAPluggableAllocator.html
index eed51dad816..72e3fd86719 100644
--- a/2.7/generated/torch.cuda.CUDAPluggableAllocator.html
+++ b/2.7/generated/torch.cuda.CUDAPluggableAllocator.html
@@ -596,7 +596,7 @@
 
                                  CUDAPluggableAllocator
                                  
 
                                  
 
                                  
-class torch.cuda.CUDAPluggableAllocator(path_to_so_file, alloc_fn_name, free_fn_name)[source][source]
                                  
+class torch.cuda.CUDAPluggableAllocator(path_to_so_file, alloc_fn_name, free_fn_name)[source][source]
 
                                  CUDA memory allocator loaded from a so file.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.Event.html b/2.7/generated/torch.cuda.Event.html
index 8710c3a7eb9..cdcb7837ded 100644
--- a/2.7/generated/torch.cuda.Event.html
+++ b/2.7/generated/torch.cuda.Event.html
@@ -596,7 +596,7 @@
 
                                  Event
                                  
 
                                  
 
                                  
-class torch.cuda.Event(enable_timing=False, blocking=False, interprocess=False)[source][source]
                                  
+class torch.cuda.Event(enable_timing=False, blocking=False, interprocess=False)[source][source]
 
                                  Wrapper around a CUDA event.
                                  
 
                                  CUDA events are synchronization markers that can be used to monitor the
 device’s progress, to accurately measure timing, and to synchronize CUDA
@@ -618,7 +618,7 @@ 
                                  Event
 
                                  
 
                                  
 
                                  
-elapsed_time(end_event)[source][source]
                                  
+elapsed_time(end_event)[source][source]
 
                                  Return the time elapsed.
                                  
 
                                  Time reported in milliseconds after the event was recorded and
 before the end_event was recorded.
                                  
@@ -626,20 +626,20 @@ 
                                  Event
 
 
                                  
 
                                  
-classmethod from_ipc_handle(device, handle)[source][source]
                                  
+classmethod from_ipc_handle(device, handle)[source][source]
 
                                  Reconstruct an event from an IPC handle on the given device.
                                  
 
                                  
 
 
                                  
 
                                  
-ipc_handle()[source][source]
                                  
+ipc_handle()[source][source]
 
                                  Return an IPC handle of this event.
                                  
 
                                  If not recorded yet, the event will use the current device.
                                  
 
                                  
 
 
                                  
 
                                  
-query()[source][source]
                                  
+query()[source][source]
 
                                  Check if all work currently captured by event has completed.
                                  
 
                                  
 
                                  Returns
                                  
@@ -651,7 +651,7 @@ 
                                  Event
 
 
                                  
 
                                  
-record(stream=None)[source][source]
                                  
+record(stream=None)[source][source]
 
                                  Record the event in a given stream.
                                  
 
                                  Uses torch.cuda.current_stream() if no stream is specified. The
 stream’s device must match the event’s device.
                                  
@@ -659,7 +659,7 @@ 
                                  Event
 
 
                                  
 
                                  
-synchronize()[source][source]
                                  
+synchronize()[source][source]
 
                                  Wait for the event to complete.
                                  
 
                                  Waits until the completion of all work currently captured in this event.
 This prevents the CPU thread from proceeding until the event completes.
                                  
@@ -676,7 +676,7 @@ 
                                  Event
 
 
                                  
 
                                  
-wait(stream=None)[source][source]
                                  
+wait(stream=None)[source][source]
 
                                  Make all future work submitted to the given stream wait for this event.
                                  
 
                                  Use torch.cuda.current_stream() if no stream is specified.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.ExternalStream.html b/2.7/generated/torch.cuda.ExternalStream.html
index 2d3aaa52126..9fa12bcb655 100644
--- a/2.7/generated/torch.cuda.ExternalStream.html
+++ b/2.7/generated/torch.cuda.ExternalStream.html
@@ -596,7 +596,7 @@
 
                                  ExternalStream
                                  
 
                                  
 
                                  
-class torch.cuda.ExternalStream(stream_ptr, device=None, **kwargs)[source][source]
                                  
+class torch.cuda.ExternalStream(stream_ptr, device=None, **kwargs)[source][source]
 
                                  Wrapper around an externally allocated CUDA stream.
                                  
 
                                  This class is used to wrap streams allocated in other libraries in order
 to facilitate data exchange and multi-library interactions.
                                  
@@ -619,7 +619,7 @@ 
                                  ExternalStream
 
                                  
-query()[source]
                                  
+query()[source]
 
                                  Check if all the work submitted has been completed.
                                  
 
                                  
 
                                  Returns
                                  
@@ -633,7 +633,7 @@ 
                                  ExternalStream
 
                                  
-record_event(event=None)[source]
                                  
+record_event(event=None)[source]
 
                                  Record an event.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -648,7 +648,7 @@ 
                                  ExternalStream
 
                                  
-synchronize()[source]
                                  
+synchronize()[source]
 
                                  Wait for all the kernels in this stream to complete.
                                  
 
                                  
 
                                  Note
                                  
@@ -661,7 +661,7 @@ 
                                  ExternalStream
 
                                  
-wait_event(event)[source]
                                  
+wait_event(event)[source]
 
                                  Make all future work submitted to the stream wait for an event.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -679,7 +679,7 @@ 
                                  ExternalStream
 
                                  
-wait_stream(stream)[source]
                                  
+wait_stream(stream)[source]
 
                                  Synchronize with another stream.
                                  
 
                                  All future work submitted to this stream will wait until all kernels
 submitted to a given stream at the time of call complete.
                                  
diff --git a/2.7/generated/torch.cuda.MemPool.html b/2.7/generated/torch.cuda.MemPool.html
index 0e52fab8624..b9c85ff4315 100644
--- a/2.7/generated/torch.cuda.MemPool.html
+++ b/2.7/generated/torch.cuda.MemPool.html
@@ -596,7 +596,7 @@
 
                                  MemPool
                                  
 
                                  
 
                                  
-class torch.cuda.MemPool(*args, **kwargs)[source][source]
                                  
+class torch.cuda.MemPool(*args, **kwargs)[source][source]
 
                                  MemPool represents a pool of memory in a caching allocator. Currently,
 it’s just the ID of the pool object maintained in the CUDACachingAllocator.
                                  
 
                                  
@@ -622,7 +622,7 @@ 
                                  MemPool
 
                                  
-snapshot()[source][source]
                                  
+snapshot()[source][source]
 
                                  Return a snapshot of the CUDA memory allocator pool state across all
 devices.
                                  
 
                                  Interpreting the output of this function requires familiarity with the
@@ -636,7 +636,7 @@ 
                                  MemPool
 
                                  
-use_count()[source][source]
                                  
+use_count()[source][source]
 
                                  Returns the reference count of this pool.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.MemPoolContext.html b/2.7/generated/torch.cuda.MemPoolContext.html
index c8105b2e161..cbce9b4b8ed 100644
--- a/2.7/generated/torch.cuda.MemPoolContext.html
+++ b/2.7/generated/torch.cuda.MemPoolContext.html
@@ -596,7 +596,7 @@
 
                                  MemPoolContext
                                  
 
                                  
 
                                  
-class torch.cuda.MemPoolContext(*args, **kwargs)[source][source]
                                  
+class torch.cuda.MemPoolContext(*args, **kwargs)[source][source]
 
                                  MemPoolContext holds the currently active pool and stashes the previous
 pool. On deletion it makes the previous pool active.
                                  
 
                                  
@@ -609,7 +609,7 @@ 
                                  MemPoolContext
 
                                  
-static active_pool()[source][source]
                                  
+static active_pool()[source][source]
 
                                  Returns the active MemPool
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.Stream.html b/2.7/generated/torch.cuda.Stream.html
index 8f8860d96bd..23a7d1a37d4 100644
--- a/2.7/generated/torch.cuda.Stream.html
+++ b/2.7/generated/torch.cuda.Stream.html
@@ -596,7 +596,7 @@
 
                                  Stream
                                  
 
                                  
 
                                  
-class torch.cuda.Stream(device=None, priority=0, **kwargs)[source][source]
                                  
+class torch.cuda.Stream(device=None, priority=0, **kwargs)[source][source]
 
                                  Wrapper around a CUDA stream.
                                  
 
                                  A CUDA stream is a linear sequence of execution that belongs to a specific
 device, independent from other streams. It supports with statement as a
@@ -618,7 +618,7 @@ 
                                  Stream
 
                                  
 
                                  
-query()[source][source]
                                  
+query()[source][source]
 
                                  Check if all the work submitted has been completed.
                                  
 
                                  
 
                                  Returns
                                  
@@ -632,7 +632,7 @@ 
                                  Stream
 
                                  
-record_event(event=None)[source][source]
                                  
+record_event(event=None)[source][source]
 
                                  Record an event.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -647,7 +647,7 @@ 
                                  Stream
 
                                  
-synchronize()[source][source]
                                  
+synchronize()[source][source]
 
                                  Wait for all the kernels in this stream to complete.
                                  
 
                                  
 
                                  Note
                                  
@@ -660,7 +660,7 @@ 
                                  Stream
 
                                  
-wait_event(event)[source][source]
                                  
+wait_event(event)[source][source]
 
                                  Make all future work submitted to the stream wait for an event.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -678,7 +678,7 @@ 
                                  Stream
 
                                  
-wait_stream(stream)[source][source]
                                  
+wait_stream(stream)[source][source]
 
                                  Synchronize with another stream.
                                  
 
                                  All future work submitted to this stream will wait until all kernels
 submitted to a given stream at the time of call complete.
                                  
diff --git a/2.7/generated/torch.cuda.StreamContext.html b/2.7/generated/torch.cuda.StreamContext.html
index b5304380059..5134b5d1beb 100644
--- a/2.7/generated/torch.cuda.StreamContext.html
+++ b/2.7/generated/torch.cuda.StreamContext.html
@@ -596,7 +596,7 @@
 
                                  StreamContext
                                  
 
                                  
 
                                  
-class torch.cuda.StreamContext(stream)[source][source]
                                  
+class torch.cuda.StreamContext(stream)[source][source]
 
                                  Context-manager that selects a given stream.
                                  
 
                                  All CUDA kernels queued within its context will be enqueued on a selected
 stream.
                                  
diff --git a/2.7/generated/torch.cuda.caching_allocator_alloc.html b/2.7/generated/torch.cuda.caching_allocator_alloc.html
index 028bc38b7cd..7cf0b3b2000 100644
--- a/2.7/generated/torch.cuda.caching_allocator_alloc.html
+++ b/2.7/generated/torch.cuda.caching_allocator_alloc.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.caching_allocator_alloc
                                  
 
                                  
 
                                  
-torch.cuda.caching_allocator_alloc(size, device=None, stream=None)[source][source]
                                  
+torch.cuda.caching_allocator_alloc(size, device=None, stream=None)[source][source]
 
                                  Perform a memory allocation using the CUDA memory allocator.
                                  
 
                                  Memory is allocated for a given device and a stream, this
 function is intended to be used for interoperability with other
diff --git a/2.7/generated/torch.cuda.caching_allocator_delete.html b/2.7/generated/torch.cuda.caching_allocator_delete.html
index b4f54e510cf..6703e97c361 100644
--- a/2.7/generated/torch.cuda.caching_allocator_delete.html
+++ b/2.7/generated/torch.cuda.caching_allocator_delete.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.caching_allocator_delete
                                  
 
                                  
 
                                  
-torch.cuda.caching_allocator_delete(mem_ptr)[source][source]
                                  
+torch.cuda.caching_allocator_delete(mem_ptr)[source][source]
 
                                  Delete memory allocated using the CUDA memory allocator.
                                  
 
                                  Memory allocated with caching_allocator_alloc().
 is freed here. The associated device and stream are tracked inside
diff --git a/2.7/generated/torch.cuda.can_device_access_peer.html b/2.7/generated/torch.cuda.can_device_access_peer.html
index 91b690a4aba..beb439b18cc 100644
--- a/2.7/generated/torch.cuda.can_device_access_peer.html
+++ b/2.7/generated/torch.cuda.can_device_access_peer.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.can_device_access_peer
                                  
 
                                  
 
                                  
-torch.cuda.can_device_access_peer(device, peer_device)[source][source]
                                  
+torch.cuda.can_device_access_peer(device, peer_device)[source][source]
 
                                  Check if peer access between two devices is possible.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.change_current_allocator.html b/2.7/generated/torch.cuda.change_current_allocator.html
index 015b939ce6a..39a74cdf039 100644
--- a/2.7/generated/torch.cuda.change_current_allocator.html
+++ b/2.7/generated/torch.cuda.change_current_allocator.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.change_current_allocator
                                  
 
                                  
 
                                  
-torch.cuda.change_current_allocator(allocator)[source][source]
                                  
+torch.cuda.change_current_allocator(allocator)[source][source]
 
                                  Change the currently used memory allocator to be the one provided.
                                  
 
                                  If the current allocator has already been used/initialized, this function will error.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.clock_rate.html b/2.7/generated/torch.cuda.clock_rate.html
index d2b6d19f471..b6097f020d9 100644
--- a/2.7/generated/torch.cuda.clock_rate.html
+++ b/2.7/generated/torch.cuda.clock_rate.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.clock_rate
                                  
 
                                  
 
                                  
-torch.cuda.clock_rate(device=None)[source][source]
                                  
+torch.cuda.clock_rate(device=None)[source][source]
 
                                  Return the clock speed of the GPU SM in MHz (megahertz) over the past sample period as given by nvidia-smi.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.comm.broadcast.html b/2.7/generated/torch.cuda.comm.broadcast.html
index 4cb5fc1ee18..79e66b97955 100644
--- a/2.7/generated/torch.cuda.comm.broadcast.html
+++ b/2.7/generated/torch.cuda.comm.broadcast.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.comm.broadcast
                                  
 
                                  
 
                                  
-torch.cuda.comm.broadcast(tensor, devices=None, *, out=None)[source][source]
                                  
+torch.cuda.comm.broadcast(tensor, devices=None, *, out=None)[source][source]
 
                                  Broadcasts a tensor to specified GPU devices.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.comm.broadcast_coalesced.html b/2.7/generated/torch.cuda.comm.broadcast_coalesced.html
index a18249a801f..436365c8e27 100644
--- a/2.7/generated/torch.cuda.comm.broadcast_coalesced.html
+++ b/2.7/generated/torch.cuda.comm.broadcast_coalesced.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.comm.broadcast_coalesced
                                  
 
                                  
 
                                  
-torch.cuda.comm.broadcast_coalesced(tensors, devices, buffer_size=10485760)[source][source]
                                  
+torch.cuda.comm.broadcast_coalesced(tensors, devices, buffer_size=10485760)[source][source]
 
                                  Broadcast a sequence of tensors to the specified GPUs.
                                  
 
                                  Small tensors are first coalesced into a buffer to reduce the number of synchronizations.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.comm.gather.html b/2.7/generated/torch.cuda.comm.gather.html
index ffdfddc3301..9ac04d99d15 100644
--- a/2.7/generated/torch.cuda.comm.gather.html
+++ b/2.7/generated/torch.cuda.comm.gather.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.comm.gather
                                  
 
                                  
 
                                  
-torch.cuda.comm.gather(tensors, dim=0, destination=None, *, out=None)[source][source]
                                  
+torch.cuda.comm.gather(tensors, dim=0, destination=None, *, out=None)[source][source]
 
                                  Gathers tensors from multiple GPU devices.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.comm.reduce_add.html b/2.7/generated/torch.cuda.comm.reduce_add.html
index 122541fadfb..6359c64b00c 100644
--- a/2.7/generated/torch.cuda.comm.reduce_add.html
+++ b/2.7/generated/torch.cuda.comm.reduce_add.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.comm.reduce_add
                                  
 
                                  
 
                                  
-torch.cuda.comm.reduce_add(inputs, destination=None)[source][source]
                                  
+torch.cuda.comm.reduce_add(inputs, destination=None)[source][source]
 
                                  Sum tensors from multiple GPUs.
                                  
 
                                  All inputs should have matching shapes, dtype, and layout. The output tensor
 will be of the same shape, dtype, and layout.
                                  
diff --git a/2.7/generated/torch.cuda.comm.scatter.html b/2.7/generated/torch.cuda.comm.scatter.html
index 5d826dfae40..666dc30aaaa 100644
--- a/2.7/generated/torch.cuda.comm.scatter.html
+++ b/2.7/generated/torch.cuda.comm.scatter.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.comm.scatter
                                  
 
                                  
 
                                  
-torch.cuda.comm.scatter(tensor, devices=None, chunk_sizes=None, dim=0, streams=None, *, out=None)[source][source]
                                  
+torch.cuda.comm.scatter(tensor, devices=None, chunk_sizes=None, dim=0, streams=None, *, out=None)[source][source]
 
                                  Scatters tensor across multiple GPUs.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.cudart.html b/2.7/generated/torch.cuda.cudart.html
index 98a349d5376..ea9d1f8d4fb 100644
--- a/2.7/generated/torch.cuda.cudart.html
+++ b/2.7/generated/torch.cuda.cudart.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.cudart
                                  
 
                                  
 
                                  
-torch.cuda.cudart()[source][source]
                                  
+torch.cuda.cudart()[source][source]
 
                                  Retrieves the CUDA runtime API module.
                                  
 
                                  This function initializes the CUDA runtime environment if it is not already
 initialized and returns the CUDA runtime API module (_cudart). The CUDA
diff --git a/2.7/generated/torch.cuda.current_blas_handle.html b/2.7/generated/torch.cuda.current_blas_handle.html
index db330c99656..45dcf35f7da 100644
--- a/2.7/generated/torch.cuda.current_blas_handle.html
+++ b/2.7/generated/torch.cuda.current_blas_handle.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.current_blas_handle
                                  
 
                                  
 
                                  
-torch.cuda.current_blas_handle()[source][source]
                                  
+torch.cuda.current_blas_handle()[source][source]
 
                                  Return cublasHandle_t pointer to current cuBLAS handle
                                  
 
                                  
 
diff --git a/2.7/generated/torch.cuda.current_device.html b/2.7/generated/torch.cuda.current_device.html
index b15779851da..e1c4b55d8d8 100644
--- a/2.7/generated/torch.cuda.current_device.html
+++ b/2.7/generated/torch.cuda.current_device.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.current_device
                                  
 
                                  
 
                                  
-torch.cuda.current_device()[source][source]
                                  
+torch.cuda.current_device()[source][source]
 
                                  Return the index of a currently selected device.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.current_stream.html b/2.7/generated/torch.cuda.current_stream.html
index bf032ed91ed..910e37825bb 100644
--- a/2.7/generated/torch.cuda.current_stream.html
+++ b/2.7/generated/torch.cuda.current_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.current_stream
                                  
 
                                  
 
                                  
-torch.cuda.current_stream(device=None)[source][source]
                                  
+torch.cuda.current_stream(device=None)[source][source]
 
                                  Return the currently selected Stream for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.default_stream.html b/2.7/generated/torch.cuda.default_stream.html
index c159be42278..75d6f86ad7e 100644
--- a/2.7/generated/torch.cuda.default_stream.html
+++ b/2.7/generated/torch.cuda.default_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.default_stream
                                  
 
                                  
 
                                  
-torch.cuda.default_stream(device=None)[source][source]
                                  
+torch.cuda.default_stream(device=None)[source][source]
 
                                  Return the default Stream for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.device.html b/2.7/generated/torch.cuda.device.html
index a2a8b620cc1..dd8a54de784 100644
--- a/2.7/generated/torch.cuda.device.html
+++ b/2.7/generated/torch.cuda.device.html
@@ -596,7 +596,7 @@
 
                                  device
                                  
 
                                  
 
                                  
-class torch.cuda.device(device)[source][source]
                                  
+class torch.cuda.device(device)[source][source]
 
                                  Context-manager that changes the selected device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.device_count.html b/2.7/generated/torch.cuda.device_count.html
index 82747fc02b5..6b54d680dbe 100644
--- a/2.7/generated/torch.cuda.device_count.html
+++ b/2.7/generated/torch.cuda.device_count.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.device_count
                                  
 
                                  
 
                                  
-torch.cuda.device_count()[source][source]
                                  
+torch.cuda.device_count()[source][source]
 
                                  Return the number of GPUs available.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.device_memory_used.html b/2.7/generated/torch.cuda.device_memory_used.html
index 479e2d167a1..36aa69be7a3 100644
--- a/2.7/generated/torch.cuda.device_memory_used.html
+++ b/2.7/generated/torch.cuda.device_memory_used.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.device_memory_used
                                  
 
                                  
 
                                  
-torch.cuda.device_memory_used(device=None)[source][source]
                                  
+torch.cuda.device_memory_used(device=None)[source][source]
 
                                  Return used global (device) memory in bytes as given by nvidia-smi or amd-smi.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.device_of.html b/2.7/generated/torch.cuda.device_of.html
index 730ee9fdbdc..a9e1d7ccf7f 100644
--- a/2.7/generated/torch.cuda.device_of.html
+++ b/2.7/generated/torch.cuda.device_of.html
@@ -596,7 +596,7 @@
 
                                  device_of
                                  
 
                                  
 
                                  
-class torch.cuda.device_of(obj)[source][source]
                                  
+class torch.cuda.device_of(obj)[source][source]
 
                                  Context-manager that changes the current device to that of given object.
                                  
 
                                  You can use both tensors and storages as arguments. If a given object is
 not allocated on a GPU, this is a no-op.
                                  
diff --git a/2.7/generated/torch.cuda.empty_cache.html b/2.7/generated/torch.cuda.empty_cache.html
index b2ec0376bca..537bef434aa 100644
--- a/2.7/generated/torch.cuda.empty_cache.html
+++ b/2.7/generated/torch.cuda.empty_cache.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.empty_cache
                                  
 
                                  
 
                                  
-torch.cuda.empty_cache()[source][source]
                                  
+torch.cuda.empty_cache()[source][source]
 
                                  Release all unoccupied cached memory currently held by the caching
 allocator so that those can be used in other GPU application and visible in
 nvidia-smi.
                                  
diff --git a/2.7/generated/torch.cuda.gds.GdsFile.html b/2.7/generated/torch.cuda.gds.GdsFile.html
index baae7296780..418ea2830d8 100644
--- a/2.7/generated/torch.cuda.gds.GdsFile.html
+++ b/2.7/generated/torch.cuda.gds.GdsFile.html
@@ -596,7 +596,7 @@
 
                                  GdsFile
                                  
 
                                  
 
                                  
-class torch.cuda.gds.GdsFile(filename, flags)[source][source]
                                  
+class torch.cuda.gds.GdsFile(filename, flags)[source][source]
 
                                  Wrapper around cuFile.
                                  
 
                                  cuFile is a file-like interface to the GPUDirect Storage (GDS) API.
                                  
 
                                  See the cufile docs
@@ -628,7 +628,7 @@ 
                                  GdsFile
 
                                  
-deregister_handle()[source][source]
                                  
+deregister_handle()[source][source]
 
                                  Deregisters file descriptor from cuFile Driver.
                                  
 
                                  This is a wrapper around cuFileHandleDeregister.
                                  
 
                                  
@@ -637,7 +637,7 @@ 
                                  GdsFile
 
                                  
-load_storage(storage, offset=0)[source][source]
                                  
+load_storage(storage, offset=0)[source][source]
 
                                  Loads data from the file into the storage.
                                  
 
                                  This is a wrapper around cuFileRead. storage.nbytes() of data
 will be loaded from the file at offset into the storage.
                                  
@@ -653,7 +653,7 @@ 
                                  GdsFile
 
                                  
-register_handle()[source][source]
                                  
+register_handle()[source][source]
 
                                  Registers file descriptor to cuFile Driver.
                                  
 
                                  This is a wrapper around cuFileHandleRegister.
                                  
 
                                  
@@ -662,7 +662,7 @@ 
                                  GdsFile
 
                                  
-save_storage(storage, offset=0)[source][source]
                                  
+save_storage(storage, offset=0)[source][source]
 
                                  Saves data from the storage into the file.
                                  
 
                                  This is a wrapper around cuFileWrite. All bytes of the storage
 will be written to the file at offset.
                                  
diff --git a/2.7/generated/torch.cuda.gds.gds_deregister_buffer.html b/2.7/generated/torch.cuda.gds.gds_deregister_buffer.html
index 1e041341e07..1ea62ed360d 100644
--- a/2.7/generated/torch.cuda.gds.gds_deregister_buffer.html
+++ b/2.7/generated/torch.cuda.gds.gds_deregister_buffer.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.gds.gds_deregister_buffer
                                  
 
                                  
 
                                  
-torch.cuda.gds.gds_deregister_buffer(s)[source][source]
                                  
+torch.cuda.gds.gds_deregister_buffer(s)[source][source]
 
                                  Deregisters a previously registered storage on a CUDA device as a cufile buffer.
                                  
 
                                  Example:
                                  
 
                                  >>> src = torch.randn(1024, device="cuda")
diff --git a/2.7/generated/torch.cuda.gds.gds_register_buffer.html b/2.7/generated/torch.cuda.gds.gds_register_buffer.html
index aab4372d52d..caadb77c9e5 100644
--- a/2.7/generated/torch.cuda.gds.gds_register_buffer.html
+++ b/2.7/generated/torch.cuda.gds.gds_register_buffer.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.gds.gds_register_buffer
                                  
 
                                  
 
                                  
-torch.cuda.gds.gds_register_buffer(s)[source][source]
                                  
+torch.cuda.gds.gds_register_buffer(s)[source][source]
 
                                  Registers a storage on a CUDA device as a cufile buffer.
                                  
 
                                  Example:
                                  
 
                                  >>> src = torch.randn(1024, device="cuda")
diff --git a/2.7/generated/torch.cuda.get_allocator_backend.html b/2.7/generated/torch.cuda.get_allocator_backend.html
index 2e07d57bd84..03758ab733b 100644
--- a/2.7/generated/torch.cuda.get_allocator_backend.html
+++ b/2.7/generated/torch.cuda.get_allocator_backend.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_allocator_backend
                                  
 
                                  
 
                                  
-torch.cuda.get_allocator_backend()[source][source]
                                  
+torch.cuda.get_allocator_backend()[source][source]
 
                                  Return a string describing the active allocator backend as set by
 PYTORCH_CUDA_ALLOC_CONF. Currently available backends are
 native (PyTorch’s native caching allocator) and cudaMallocAsync`
diff --git a/2.7/generated/torch.cuda.get_arch_list.html b/2.7/generated/torch.cuda.get_arch_list.html
index 4e730b90cdf..992eb0f2b98 100644
--- a/2.7/generated/torch.cuda.get_arch_list.html
+++ b/2.7/generated/torch.cuda.get_arch_list.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_arch_list
                                  
 
                                  
 
                                  
-torch.cuda.get_arch_list()[source][source]
                                  
+torch.cuda.get_arch_list()[source][source]
 
                                  Return list CUDA architectures this library was compiled for.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.get_device_capability.html b/2.7/generated/torch.cuda.get_device_capability.html
index 0bbc187502c..410c2870033 100644
--- a/2.7/generated/torch.cuda.get_device_capability.html
+++ b/2.7/generated/torch.cuda.get_device_capability.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_device_capability
                                  
 
                                  
 
                                  
-torch.cuda.get_device_capability(device=None)[source][source]
                                  
+torch.cuda.get_device_capability(device=None)[source][source]
 
                                  Get the cuda capability of a device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.get_device_name.html b/2.7/generated/torch.cuda.get_device_name.html
index add1004305d..09863ce33ea 100644
--- a/2.7/generated/torch.cuda.get_device_name.html
+++ b/2.7/generated/torch.cuda.get_device_name.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_device_name
                                  
 
                                  
 
                                  
-torch.cuda.get_device_name(device=None)[source][source]
                                  
+torch.cuda.get_device_name(device=None)[source][source]
 
                                  Get the name of a device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.get_device_properties.html b/2.7/generated/torch.cuda.get_device_properties.html
index 690093306fe..5a9edb13b15 100644
--- a/2.7/generated/torch.cuda.get_device_properties.html
+++ b/2.7/generated/torch.cuda.get_device_properties.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_device_properties
                                  
 
                                  
 
                                  
-torch.cuda.get_device_properties(device=None)[source][source]
                                  
+torch.cuda.get_device_properties(device=None)[source][source]
 
                                  Get the properties of a device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.get_gencode_flags.html b/2.7/generated/torch.cuda.get_gencode_flags.html
index 90e73fd2d58..55e3e81c939 100644
--- a/2.7/generated/torch.cuda.get_gencode_flags.html
+++ b/2.7/generated/torch.cuda.get_gencode_flags.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_gencode_flags
                                  
 
                                  
 
                                  
-torch.cuda.get_gencode_flags()[source][source]
                                  
+torch.cuda.get_gencode_flags()[source][source]
 
                                  Return NVCC gencode flags this library was compiled with.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.get_per_process_memory_fraction.html b/2.7/generated/torch.cuda.get_per_process_memory_fraction.html
index b7188d6e3e7..fa89560f916 100644
--- a/2.7/generated/torch.cuda.get_per_process_memory_fraction.html
+++ b/2.7/generated/torch.cuda.get_per_process_memory_fraction.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_per_process_memory_fraction
                                  
 
                                  
 
                                  
-torch.cuda.get_per_process_memory_fraction(device=None)[source][source]
                                  
+torch.cuda.get_per_process_memory_fraction(device=None)[source][source]
 
                                  Get memory fraction for a process.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.get_rng_state.html b/2.7/generated/torch.cuda.get_rng_state.html
index 3533603984c..7a833b7f351 100644
--- a/2.7/generated/torch.cuda.get_rng_state.html
+++ b/2.7/generated/torch.cuda.get_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_rng_state
                                  
 
                                  
 
                                  
-torch.cuda.get_rng_state(device='cuda')[source][source]
                                  
+torch.cuda.get_rng_state(device='cuda')[source][source]
 
                                  Return the random number generator state of the specified GPU as a ByteTensor.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.get_rng_state_all.html b/2.7/generated/torch.cuda.get_rng_state_all.html
index e0bc4bf1375..4dea9247a98 100644
--- a/2.7/generated/torch.cuda.get_rng_state_all.html
+++ b/2.7/generated/torch.cuda.get_rng_state_all.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_rng_state_all
                                  
 
                                  
 
                                  
-torch.cuda.get_rng_state_all()[source][source]
                                  
+torch.cuda.get_rng_state_all()[source][source]
 
                                  Return a list of ByteTensor representing the random number states of all devices.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.get_stream_from_external.html b/2.7/generated/torch.cuda.get_stream_from_external.html
index 317a5022d49..681e33998be 100644
--- a/2.7/generated/torch.cuda.get_stream_from_external.html
+++ b/2.7/generated/torch.cuda.get_stream_from_external.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_stream_from_external
                                  
 
                                  
 
                                  
-torch.cuda.get_stream_from_external(data_ptr, device=None)[source][source]
                                  
+torch.cuda.get_stream_from_external(data_ptr, device=None)[source][source]
 
                                  Return a Stream from an externally allocated CUDA stream.
                                  
 
                                  This function is used to wrap streams allocated in other libraries in order
 to facilitate data exchange and multi-library interactions.
                                  
diff --git a/2.7/generated/torch.cuda.get_sync_debug_mode.html b/2.7/generated/torch.cuda.get_sync_debug_mode.html
index d3251e412ab..195f890f537 100644
--- a/2.7/generated/torch.cuda.get_sync_debug_mode.html
+++ b/2.7/generated/torch.cuda.get_sync_debug_mode.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.get_sync_debug_mode
                                  
 
                                  
 
                                  
-torch.cuda.get_sync_debug_mode()[source][source]
                                  
+torch.cuda.get_sync_debug_mode()[source][source]
 
                                  Return current value of debug mode for cuda synchronizing operations.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.graph.html b/2.7/generated/torch.cuda.graph.html
index cb580fa098f..2235666f42d 100644
--- a/2.7/generated/torch.cuda.graph.html
+++ b/2.7/generated/torch.cuda.graph.html
@@ -596,7 +596,7 @@
 
                                  graph
                                  
 
                                  
 
                                  
-class torch.cuda.graph(cuda_graph, pool=None, stream=None, capture_error_mode='global')[source][source]
                                  
+class torch.cuda.graph(cuda_graph, pool=None, stream=None, capture_error_mode='global')[source][source]
 
                                  Context-manager that captures CUDA work into a torch.cuda.CUDAGraph object for later replay.
                                  
 
                                  See CUDA Graphs for a general introduction,
 detailed use, and constraints.
                                  
diff --git a/2.7/generated/torch.cuda.graph_pool_handle.html b/2.7/generated/torch.cuda.graph_pool_handle.html
index ebc485b6d49..028abf945d5 100644
--- a/2.7/generated/torch.cuda.graph_pool_handle.html
+++ b/2.7/generated/torch.cuda.graph_pool_handle.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.graph_pool_handle
                                  
 
                                  
 
                                  
-torch.cuda.graph_pool_handle()[source][source]
                                  
+torch.cuda.graph_pool_handle()[source][source]
 
                                  Return an opaque token representing the id of a graph memory pool.
                                  
 
                                  See Graph memory management.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.host_memory_stats.html b/2.7/generated/torch.cuda.host_memory_stats.html
index 31b8f2f3bac..dd4e31c4af1 100644
--- a/2.7/generated/torch.cuda.host_memory_stats.html
+++ b/2.7/generated/torch.cuda.host_memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.host_memory_stats
                                  
 
                                  
 
                                  
-torch.cuda.host_memory_stats()[source][source]
                                  
+torch.cuda.host_memory_stats()[source][source]
 
                                  Return a dictionary of CUDA memory allocator statistics for a given device.
                                  
 
                                  
 
                                  The return value of this function is a dictionary of statistics, each of
diff --git a/2.7/generated/torch.cuda.init.html b/2.7/generated/torch.cuda.init.html
index 717a79d3e48..663dc5ddcf2 100644
--- a/2.7/generated/torch.cuda.init.html
+++ b/2.7/generated/torch.cuda.init.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.init
                                  
 
                                  
 
                                  
-torch.cuda.init()[source][source]
                                  
+torch.cuda.init()[source][source]
 
                                  Initialize PyTorch’s CUDA state.
                                  
 
                                  You may need to call this explicitly if you are interacting with
 PyTorch via its C API, as Python bindings for CUDA functionality
diff --git a/2.7/generated/torch.cuda.initial_seed.html b/2.7/generated/torch.cuda.initial_seed.html
index ce2e18f3457..90f0973d439 100644
--- a/2.7/generated/torch.cuda.initial_seed.html
+++ b/2.7/generated/torch.cuda.initial_seed.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.initial_seed
                                  
 
                                  
 
                                  
-torch.cuda.initial_seed()[source][source]
                                  
+torch.cuda.initial_seed()[source][source]
 
                                  Return the current random seed of the current GPU.
                                  
 
                                  
 
                                  Warning
                                  
diff --git a/2.7/generated/torch.cuda.ipc_collect.html b/2.7/generated/torch.cuda.ipc_collect.html
index a5a1ede1e47..ff7308bef80 100644
--- a/2.7/generated/torch.cuda.ipc_collect.html
+++ b/2.7/generated/torch.cuda.ipc_collect.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.ipc_collect
                                  
 
                                  
 
                                  
-torch.cuda.ipc_collect()[source][source]
                                  
+torch.cuda.ipc_collect()[source][source]
 
                                  Force collects GPU memory after it has been released by CUDA IPC.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.cuda.is_available.html b/2.7/generated/torch.cuda.is_available.html
index 3707d0ecd9b..b21789f3aa6 100644
--- a/2.7/generated/torch.cuda.is_available.html
+++ b/2.7/generated/torch.cuda.is_available.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.is_available
                                  
 
                                  
 
                                  
-torch.cuda.is_available()[source][source]
                                  
+torch.cuda.is_available()[source][source]
 
                                  Return a bool indicating if CUDA is currently available.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.is_current_stream_capturing.html b/2.7/generated/torch.cuda.is_current_stream_capturing.html
index 4c89ad87ef5..c7141f38df8 100644
--- a/2.7/generated/torch.cuda.is_current_stream_capturing.html
+++ b/2.7/generated/torch.cuda.is_current_stream_capturing.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.is_current_stream_capturing
                                  
 
                                  
 
                                  
-torch.cuda.is_current_stream_capturing()[source][source]
                                  
+torch.cuda.is_current_stream_capturing()[source][source]
 
                                  Return True if CUDA graph capture is underway on the current CUDA stream, False otherwise.
                                  
 
                                  If a CUDA context does not exist on the current device, returns False without initializing the context.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.is_initialized.html b/2.7/generated/torch.cuda.is_initialized.html
index da042546590..13e53946042 100644
--- a/2.7/generated/torch.cuda.is_initialized.html
+++ b/2.7/generated/torch.cuda.is_initialized.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.is_initialized
                                  
 
                                  
 
                                  
-torch.cuda.is_initialized()[source][source]
                                  
+torch.cuda.is_initialized()[source][source]
 
                                  Return whether PyTorch’s CUDA state has been initialized.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.cuda.is_tf32_supported.html b/2.7/generated/torch.cuda.is_tf32_supported.html
index 2cad7e3cb4f..f5fff9343ca 100644
--- a/2.7/generated/torch.cuda.is_tf32_supported.html
+++ b/2.7/generated/torch.cuda.is_tf32_supported.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.is_tf32_supported
                                  
 
                                  
 
                                  
-torch.cuda.is_tf32_supported()[source][source]
                                  
+torch.cuda.is_tf32_supported()[source][source]
 
                                  Return a bool indicating if the current CUDA/ROCm device supports dtype tf32.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.jiterator._create_jit_fn.html b/2.7/generated/torch.cuda.jiterator._create_jit_fn.html
index 89349823da0..d32c03bb55a 100644
--- a/2.7/generated/torch.cuda.jiterator._create_jit_fn.html
+++ b/2.7/generated/torch.cuda.jiterator._create_jit_fn.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.jiterator._create_jit_fn
                                  
 
                                  
 
                                  
-torch.cuda.jiterator._create_jit_fn(code_string, **kwargs)[source][source]
                                  
+torch.cuda.jiterator._create_jit_fn(code_string, **kwargs)[source][source]
 
                                  Create a jiterator-generated cuda kernel for an elementwise op.
                                  
 
                                  The code string has to be a valid CUDA function that describes the computation for a single element. The code
 string has to follow the c++ template pattern, as shown in the example below. This function will be inlined
diff --git a/2.7/generated/torch.cuda.jiterator._create_multi_output_jit_fn.html b/2.7/generated/torch.cuda.jiterator._create_multi_output_jit_fn.html
index 7b08b93eee5..ce0af563eb7 100644
--- a/2.7/generated/torch.cuda.jiterator._create_multi_output_jit_fn.html
+++ b/2.7/generated/torch.cuda.jiterator._create_multi_output_jit_fn.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.jiterator._create_multi_output_jit_fn
                                  
 
                                  
 
                                  
-torch.cuda.jiterator._create_multi_output_jit_fn(code_string, num_outputs, **kwargs)[source][source]
                                  
+torch.cuda.jiterator._create_multi_output_jit_fn(code_string, num_outputs, **kwargs)[source][source]
 
                                  Create a jiterator-generated cuda kernel for an elementwise op that supports returning one or more outputs.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.list_gpu_processes.html b/2.7/generated/torch.cuda.list_gpu_processes.html
index a67216d2492..89cd9d679b7 100644
--- a/2.7/generated/torch.cuda.list_gpu_processes.html
+++ b/2.7/generated/torch.cuda.list_gpu_processes.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.list_gpu_processes
                                  
 
                                  
 
                                  
-torch.cuda.list_gpu_processes(device=None)[source][source]
                                  
+torch.cuda.list_gpu_processes(device=None)[source][source]
 
                                  Return a human-readable printout of the running processes and their GPU memory use for a given device.
                                  
 
                                  This can be useful to display periodically during training, or when
 handling out-of-memory exceptions.
                                  
diff --git a/2.7/generated/torch.cuda.make_graphed_callables.html b/2.7/generated/torch.cuda.make_graphed_callables.html
index f36912d6071..f0a7a9ce083 100644
--- a/2.7/generated/torch.cuda.make_graphed_callables.html
+++ b/2.7/generated/torch.cuda.make_graphed_callables.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.make_graphed_callables
                                  
 
                                  
 
                                  
-torch.cuda.make_graphed_callables(callables, sample_args, num_warmup_iters=3, allow_unused_input=False, pool=None)[source][source]
                                  
+torch.cuda.make_graphed_callables(callables, sample_args, num_warmup_iters=3, allow_unused_input=False, pool=None)[source][source]
 
                                  Accept callables (functions or nn.Modules) and returns graphed versions.
                                  
 
                                  Each graphed callable’s forward pass runs its source callable’s
 forward CUDA work as a CUDA graph inside a single autograd node.
                                  
diff --git a/2.7/generated/torch.cuda.manual_seed.html b/2.7/generated/torch.cuda.manual_seed.html
index 720021376ce..348c6bf1e4a 100644
--- a/2.7/generated/torch.cuda.manual_seed.html
+++ b/2.7/generated/torch.cuda.manual_seed.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.manual_seed
                                  
 
                                  
 
                                  
-torch.cuda.manual_seed(seed)[source][source]
                                  
+torch.cuda.manual_seed(seed)[source][source]
 
                                  Set the seed for generating random numbers for the current GPU.
                                  
 
                                  It’s safe to call this function if CUDA is not available; in that
 case, it is silently ignored.
                                  
diff --git a/2.7/generated/torch.cuda.manual_seed_all.html b/2.7/generated/torch.cuda.manual_seed_all.html
index da3a5db5026..ac03288f2b5 100644
--- a/2.7/generated/torch.cuda.manual_seed_all.html
+++ b/2.7/generated/torch.cuda.manual_seed_all.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.manual_seed_all
                                  
 
                                  
 
                                  
-torch.cuda.manual_seed_all(seed)[source][source]
                                  
+torch.cuda.manual_seed_all(seed)[source][source]
 
                                  Set the seed for generating random numbers on all GPUs.
                                  
 
                                  It’s safe to call this function if CUDA is not available; in that
 case, it is silently ignored.
                                  
diff --git a/2.7/generated/torch.cuda.max_memory_allocated.html b/2.7/generated/torch.cuda.max_memory_allocated.html
index ce5dceff2e8..ee817c96847 100644
--- a/2.7/generated/torch.cuda.max_memory_allocated.html
+++ b/2.7/generated/torch.cuda.max_memory_allocated.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.max_memory_allocated
                                  
 
                                  
 
                                  
-torch.cuda.max_memory_allocated(device=None)[source][source]
                                  
+torch.cuda.max_memory_allocated(device=None)[source][source]
 
                                  Return the maximum GPU memory occupied by tensors in bytes for a given device.
                                  
 
                                  By default, this returns the peak allocated memory since the beginning of
 this program. reset_peak_memory_stats() can be used to
diff --git a/2.7/generated/torch.cuda.max_memory_cached.html b/2.7/generated/torch.cuda.max_memory_cached.html
index e34c87d4d53..ed43bb6eb50 100644
--- a/2.7/generated/torch.cuda.max_memory_cached.html
+++ b/2.7/generated/torch.cuda.max_memory_cached.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.max_memory_cached
                                  
 
                                  
 
                                  
-torch.cuda.max_memory_cached(device=None)[source][source]
                                  
+torch.cuda.max_memory_cached(device=None)[source][source]
 
                                  Deprecated; see max_memory_reserved().
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.max_memory_reserved.html b/2.7/generated/torch.cuda.max_memory_reserved.html
index 05f2291d6af..735610695ed 100644
--- a/2.7/generated/torch.cuda.max_memory_reserved.html
+++ b/2.7/generated/torch.cuda.max_memory_reserved.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.max_memory_reserved
                                  
 
                                  
 
                                  
-torch.cuda.max_memory_reserved(device=None)[source][source]
                                  
+torch.cuda.max_memory_reserved(device=None)[source][source]
 
                                  Return the maximum GPU memory managed by the caching allocator in bytes for a given device.
                                  
 
                                  By default, this returns the peak cached memory since the beginning of this
 program. reset_peak_memory_stats() can be used to reset
diff --git a/2.7/generated/torch.cuda.mem_get_info.html b/2.7/generated/torch.cuda.mem_get_info.html
index 54e460c52b4..78eb42ded32 100644
--- a/2.7/generated/torch.cuda.mem_get_info.html
+++ b/2.7/generated/torch.cuda.mem_get_info.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.mem_get_info
                                  
 
                                  
 
                                  
-torch.cuda.mem_get_info(device=None)[source][source]
                                  
+torch.cuda.mem_get_info(device=None)[source][source]
 
                                  Return the global free and total GPU memory for a given device using cudaMemGetInfo.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.memory.caching_allocator_enable.html b/2.7/generated/torch.cuda.memory.caching_allocator_enable.html
index 4e1c9f2e7bb..ae294e7bf6b 100644
--- a/2.7/generated/torch.cuda.memory.caching_allocator_enable.html
+++ b/2.7/generated/torch.cuda.memory.caching_allocator_enable.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory.caching_allocator_enable
                                  
 
                                  
 
                                  
-torch.cuda.memory.caching_allocator_enable(value=True)[source][source]
                                  
+torch.cuda.memory.caching_allocator_enable(value=True)[source][source]
 
                                  Enable or disable the CUDA memory allocator. On by default.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.memory_allocated.html b/2.7/generated/torch.cuda.memory_allocated.html
index 3a1a0991477..fff86a45657 100644
--- a/2.7/generated/torch.cuda.memory_allocated.html
+++ b/2.7/generated/torch.cuda.memory_allocated.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_allocated
                                  
 
                                  
 
                                  
-torch.cuda.memory_allocated(device=None)[source][source]
                                  
+torch.cuda.memory_allocated(device=None)[source][source]
 
                                  Return the current GPU memory occupied by tensors in bytes for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.memory_cached.html b/2.7/generated/torch.cuda.memory_cached.html
index d952f39b56c..2ee77ece83a 100644
--- a/2.7/generated/torch.cuda.memory_cached.html
+++ b/2.7/generated/torch.cuda.memory_cached.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_cached
                                  
 
                                  
 
                                  
-torch.cuda.memory_cached(device=None)[source][source]
                                  
+torch.cuda.memory_cached(device=None)[source][source]
 
                                  Deprecated; see memory_reserved().
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.cuda.memory_reserved.html b/2.7/generated/torch.cuda.memory_reserved.html
index aee062e1b94..09f0b535cd7 100644
--- a/2.7/generated/torch.cuda.memory_reserved.html
+++ b/2.7/generated/torch.cuda.memory_reserved.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_reserved
                                  
 
                                  
 
                                  
-torch.cuda.memory_reserved(device=None)[source][source]
                                  
+torch.cuda.memory_reserved(device=None)[source][source]
 
                                  Return the current GPU memory managed by the caching allocator in bytes for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.memory_snapshot.html b/2.7/generated/torch.cuda.memory_snapshot.html
index 56dca67e168..7d1a73af550 100644
--- a/2.7/generated/torch.cuda.memory_snapshot.html
+++ b/2.7/generated/torch.cuda.memory_snapshot.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_snapshot
                                  
 
                                  
 
                                  
-torch.cuda.memory_snapshot()[source][source]
                                  
+torch.cuda.memory_snapshot()[source][source]
 
                                  Return a snapshot of the CUDA memory allocator state across all devices.
                                  
 
                                  Interpreting the output of this function requires familiarity with the
 memory allocator internals.
                                  
diff --git a/2.7/generated/torch.cuda.memory_stats.html b/2.7/generated/torch.cuda.memory_stats.html
index 4d538456d36..9ff6f8b694e 100644
--- a/2.7/generated/torch.cuda.memory_stats.html
+++ b/2.7/generated/torch.cuda.memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_stats
                                  
 
                                  
 
                                  
-torch.cuda.memory_stats(device=None)[source][source]
                                  
+torch.cuda.memory_stats(device=None)[source][source]
 
                                  Return a dictionary of CUDA memory allocator statistics for a given device.
                                  
 
                                  The return value of this function is a dictionary of statistics, each of
 which is a non-negative integer.
                                  
diff --git a/2.7/generated/torch.cuda.memory_summary.html b/2.7/generated/torch.cuda.memory_summary.html
index 9467b3c768e..a73a6aebd81 100644
--- a/2.7/generated/torch.cuda.memory_summary.html
+++ b/2.7/generated/torch.cuda.memory_summary.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_summary
                                  
 
                                  
 
                                  
-torch.cuda.memory_summary(device=None, abbreviated=False)[source][source]
                                  
+torch.cuda.memory_summary(device=None, abbreviated=False)[source][source]
 
                                  Return a human-readable printout of the current memory allocator statistics for a given device.
                                  
 
                                  This can be useful to display periodically during training, or when
 handling out-of-memory exceptions.
                                  
diff --git a/2.7/generated/torch.cuda.memory_usage.html b/2.7/generated/torch.cuda.memory_usage.html
index 61a825c47cc..204b53ec21e 100644
--- a/2.7/generated/torch.cuda.memory_usage.html
+++ b/2.7/generated/torch.cuda.memory_usage.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.memory_usage
                                  
 
                                  
 
                                  
-torch.cuda.memory_usage(device=None)[source][source]
                                  
+torch.cuda.memory_usage(device=None)[source][source]
 
                                  Return the percent of time over the past sample period during which global (device)
 memory was being read or written as given by nvidia-smi.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.nvtx.mark.html b/2.7/generated/torch.cuda.nvtx.mark.html
index 4e42e451b12..d591f5b4c67 100644
--- a/2.7/generated/torch.cuda.nvtx.mark.html
+++ b/2.7/generated/torch.cuda.nvtx.mark.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.nvtx.mark
                                  
 
                                  
 
                                  
-torch.cuda.nvtx.mark(msg)[source][source]
                                  
+torch.cuda.nvtx.mark(msg)[source][source]
 
                                  Describe an instantaneous event that occurred at some point.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.nvtx.range.html b/2.7/generated/torch.cuda.nvtx.range.html
index 9f997bf0eaa..4e0a0ab79fa 100644
--- a/2.7/generated/torch.cuda.nvtx.range.html
+++ b/2.7/generated/torch.cuda.nvtx.range.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.nvtx.range
                                  
 
                                  
 
                                  
-torch.cuda.nvtx.range(msg, *args, **kwargs)[source][source]
                                  
+torch.cuda.nvtx.range(msg, *args, **kwargs)[source][source]
 
                                  Context manager / decorator that pushes an NVTX range at the beginning
 of its scope, and pops it at the end. If extra arguments are given,
 they are passed as arguments to msg.format().
                                  
diff --git a/2.7/generated/torch.cuda.nvtx.range_pop.html b/2.7/generated/torch.cuda.nvtx.range_pop.html
index b2a60291040..3f2784ba166 100644
--- a/2.7/generated/torch.cuda.nvtx.range_pop.html
+++ b/2.7/generated/torch.cuda.nvtx.range_pop.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.nvtx.range_pop
                                  
 
                                  
 
                                  
-torch.cuda.nvtx.range_pop()[source][source]
                                  
+torch.cuda.nvtx.range_pop()[source][source]
 
                                  Pop a range off of a stack of nested range spans.  Returns the  zero-based depth of the range that is ended.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.cuda.nvtx.range_push.html b/2.7/generated/torch.cuda.nvtx.range_push.html
index ab686474a3e..6af2bd7d322 100644
--- a/2.7/generated/torch.cuda.nvtx.range_push.html
+++ b/2.7/generated/torch.cuda.nvtx.range_push.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.nvtx.range_push
                                  
 
                                  
 
                                  
-torch.cuda.nvtx.range_push(msg)[source][source]
                                  
+torch.cuda.nvtx.range_push(msg)[source][source]
 
                                  Push a range onto a stack of nested range span.  Returns zero-based depth of the range that is started.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.power_draw.html b/2.7/generated/torch.cuda.power_draw.html
index c4999e65588..1a9ecff193b 100644
--- a/2.7/generated/torch.cuda.power_draw.html
+++ b/2.7/generated/torch.cuda.power_draw.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.power_draw
                                  
 
                                  
 
                                  
-torch.cuda.power_draw(device=None)[source][source]
                                  
+torch.cuda.power_draw(device=None)[source][source]
 
                                  
 
                                  Return the average power draw of the GPU sensor in mW (MilliWatts)
                                  over the past sample period as given by nvidia-smi for Fermi or newer fully supported devices.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.reset_max_memory_allocated.html b/2.7/generated/torch.cuda.reset_max_memory_allocated.html
index 93f0a41119e..c41310cc8a8 100644
--- a/2.7/generated/torch.cuda.reset_max_memory_allocated.html
+++ b/2.7/generated/torch.cuda.reset_max_memory_allocated.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.reset_max_memory_allocated
                                  
 
                                  
 
                                  
-torch.cuda.reset_max_memory_allocated(device=None)[source][source]
                                  
+torch.cuda.reset_max_memory_allocated(device=None)[source][source]
 
                                  Reset the starting point in tracking maximum GPU memory occupied by tensors for a given device.
                                  
 
                                  See max_memory_allocated() for details.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.reset_max_memory_cached.html b/2.7/generated/torch.cuda.reset_max_memory_cached.html
index bd4a6c169d9..622b013f503 100644
--- a/2.7/generated/torch.cuda.reset_max_memory_cached.html
+++ b/2.7/generated/torch.cuda.reset_max_memory_cached.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.reset_max_memory_cached
                                  
 
                                  
 
                                  
-torch.cuda.reset_max_memory_cached(device=None)[source][source]
                                  
+torch.cuda.reset_max_memory_cached(device=None)[source][source]
 
                                  Reset the starting point in tracking maximum GPU memory managed by the caching allocator for a given device.
                                  
 
                                  See max_memory_cached() for details.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.reset_peak_host_memory_stats.html b/2.7/generated/torch.cuda.reset_peak_host_memory_stats.html
index 2cae9600bef..82a5df4e6af 100644
--- a/2.7/generated/torch.cuda.reset_peak_host_memory_stats.html
+++ b/2.7/generated/torch.cuda.reset_peak_host_memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.reset_peak_host_memory_stats
                                  
 
                                  
 
                                  
-torch.cuda.reset_peak_host_memory_stats()[source][source]
                                  
+torch.cuda.reset_peak_host_memory_stats()[source][source]
 
                                  Reset the “peak” stats tracked by the host memory allocator.
                                  
 
                                  See host_memory_stats() for details. Peak stats correspond to the
 “peak” key in each individual stat dict.
                                  
diff --git a/2.7/generated/torch.cuda.reset_peak_memory_stats.html b/2.7/generated/torch.cuda.reset_peak_memory_stats.html
index 15699a3d84f..75012409640 100644
--- a/2.7/generated/torch.cuda.reset_peak_memory_stats.html
+++ b/2.7/generated/torch.cuda.reset_peak_memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.reset_peak_memory_stats
                                  
 
                                  
 
                                  
-torch.cuda.reset_peak_memory_stats(device=None)[source][source]
                                  
+torch.cuda.reset_peak_memory_stats(device=None)[source][source]
 
                                  Reset the “peak” stats tracked by the CUDA memory allocator.
                                  
 
                                  See memory_stats() for details. Peak stats correspond to the
 “peak” key in each individual stat dict.
                                  
diff --git a/2.7/generated/torch.cuda.seed.html b/2.7/generated/torch.cuda.seed.html
index 2ec1aa6a890..740a976cf2b 100644
--- a/2.7/generated/torch.cuda.seed.html
+++ b/2.7/generated/torch.cuda.seed.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.seed
                                  
 
                                  
 
                                  
-torch.cuda.seed()[source][source]
                                  
+torch.cuda.seed()[source][source]
 
                                  Set the seed for generating random numbers to a random number for the current GPU.
                                  
 
                                  It’s safe to call this function if CUDA is not available; in that
 case, it is silently ignored.
                                  
diff --git a/2.7/generated/torch.cuda.seed_all.html b/2.7/generated/torch.cuda.seed_all.html
index 9a1aee5f777..04f2c78021e 100644
--- a/2.7/generated/torch.cuda.seed_all.html
+++ b/2.7/generated/torch.cuda.seed_all.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.seed_all
                                  
 
                                  
 
                                  
-torch.cuda.seed_all()[source][source]
                                  
+torch.cuda.seed_all()[source][source]
 
                                  Set the seed for generating random numbers to a random number on all GPUs.
                                  
 
                                  It’s safe to call this function if CUDA is not available; in that
 case, it is silently ignored.
                                  
diff --git a/2.7/generated/torch.cuda.set_device.html b/2.7/generated/torch.cuda.set_device.html
index 270ad1bc34a..f0c0acbfd90 100644
--- a/2.7/generated/torch.cuda.set_device.html
+++ b/2.7/generated/torch.cuda.set_device.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_device
                                  
 
                                  
 
                                  
-torch.cuda.set_device(device)[source][source]
                                  
+torch.cuda.set_device(device)[source][source]
 
                                  Set the current device.
                                  
 
                                  Usage of this function is discouraged in favor of device. In most
 cases it’s better to use CUDA_VISIBLE_DEVICES environmental variable.
                                  
diff --git a/2.7/generated/torch.cuda.set_per_process_memory_fraction.html b/2.7/generated/torch.cuda.set_per_process_memory_fraction.html
index ac640fd3839..6cdd515e419 100644
--- a/2.7/generated/torch.cuda.set_per_process_memory_fraction.html
+++ b/2.7/generated/torch.cuda.set_per_process_memory_fraction.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_per_process_memory_fraction
                                  
 
                                  
 
                                  
-torch.cuda.set_per_process_memory_fraction(fraction, device=None)[source][source]
                                  
+torch.cuda.set_per_process_memory_fraction(fraction, device=None)[source][source]
 
                                  Set memory fraction for a process.
                                  
 
                                  The fraction is used to limit an caching allocator to allocated memory on a CUDA device.
 The allowed value equals the total visible memory multiplied fraction.
diff --git a/2.7/generated/torch.cuda.set_rng_state.html b/2.7/generated/torch.cuda.set_rng_state.html
index 7992aaef482..29af22bf0ad 100644
--- a/2.7/generated/torch.cuda.set_rng_state.html
+++ b/2.7/generated/torch.cuda.set_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_rng_state
                                  
 
                                  
 
                                  
-torch.cuda.set_rng_state(new_state, device='cuda')[source][source]
                                  
+torch.cuda.set_rng_state(new_state, device='cuda')[source][source]
 
                                  Set the random number generator state of the specified GPU.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.set_rng_state_all.html b/2.7/generated/torch.cuda.set_rng_state_all.html
index 1d5387a7517..00c973222ad 100644
--- a/2.7/generated/torch.cuda.set_rng_state_all.html
+++ b/2.7/generated/torch.cuda.set_rng_state_all.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_rng_state_all
                                  
 
                                  
 
                                  
-torch.cuda.set_rng_state_all(new_states)[source][source]
                                  
+torch.cuda.set_rng_state_all(new_states)[source][source]
 
                                  Set the random number generator state of all devices.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.set_stream.html b/2.7/generated/torch.cuda.set_stream.html
index 22a7b92d531..5a2f38892a4 100644
--- a/2.7/generated/torch.cuda.set_stream.html
+++ b/2.7/generated/torch.cuda.set_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_stream
                                  
 
                                  
 
                                  
-torch.cuda.set_stream(stream)[source][source]
                                  
+torch.cuda.set_stream(stream)[source][source]
 
                                  
 
                                  Set the current stream.This is a wrapper API to set the stream.
                                  Usage of this function is discouraged in favor of the stream
 context manager.
                                  
diff --git a/2.7/generated/torch.cuda.set_sync_debug_mode.html b/2.7/generated/torch.cuda.set_sync_debug_mode.html
index a0e7a5ed0ae..c582725f64c 100644
--- a/2.7/generated/torch.cuda.set_sync_debug_mode.html
+++ b/2.7/generated/torch.cuda.set_sync_debug_mode.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.set_sync_debug_mode
                                  
 
                                  
 
                                  
-torch.cuda.set_sync_debug_mode(debug_mode)[source][source]
                                  
+torch.cuda.set_sync_debug_mode(debug_mode)[source][source]
 
                                  Set the debug mode for cuda synchronizing operations.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.stream.html b/2.7/generated/torch.cuda.stream.html
index 5ec22082777..376c47801f3 100644
--- a/2.7/generated/torch.cuda.stream.html
+++ b/2.7/generated/torch.cuda.stream.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.stream
                                  
 
                                  
 
                                  
-torch.cuda.stream(stream)[source][source]
                                  
+torch.cuda.stream(stream)[source][source]
 
                                  Wrap around the Context-manager StreamContext that selects a given stream.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.synchronize.html b/2.7/generated/torch.cuda.synchronize.html
index 9a90b42f8a8..96082d9ccfb 100644
--- a/2.7/generated/torch.cuda.synchronize.html
+++ b/2.7/generated/torch.cuda.synchronize.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.synchronize
                                  
 
                                  
 
                                  
-torch.cuda.synchronize(device=None)[source][source]
                                  
+torch.cuda.synchronize(device=None)[source][source]
 
                                  Wait for all kernels in all streams on a CUDA device to complete.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.cuda.temperature.html b/2.7/generated/torch.cuda.temperature.html
index f07fa71942f..7921284a6b1 100644
--- a/2.7/generated/torch.cuda.temperature.html
+++ b/2.7/generated/torch.cuda.temperature.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.temperature
                                  
 
                                  
 
                                  
-torch.cuda.temperature(device=None)[source][source]
                                  
+torch.cuda.temperature(device=None)[source][source]
 
                                  Return the average temperature of the GPU sensor in Degrees C (Centigrades).
                                  
 
                                  The average temperature is computed based on past sample period as given by nvidia-smi.
                                  
 
                                  
diff --git a/2.7/generated/torch.cuda.utilization.html b/2.7/generated/torch.cuda.utilization.html
index 3f1b6751e6e..2ec4423e753 100644
--- a/2.7/generated/torch.cuda.utilization.html
+++ b/2.7/generated/torch.cuda.utilization.html
@@ -596,7 +596,7 @@
 
                                  torch.cuda.utilization
                                  
 
                                  
 
                                  
-torch.cuda.utilization(device=None)[source][source]
                                  
+torch.cuda.utilization(device=None)[source][source]
 
                                  Return the percent of time over the past sample period during which one or
 more kernels was executing on the GPU as given by nvidia-smi.
                                  
 
                                  
diff --git a/2.7/generated/torch.einsum.html b/2.7/generated/torch.einsum.html
index 8c172a03de8..8d1a6187b6b 100644
--- a/2.7/generated/torch.einsum.html
+++ b/2.7/generated/torch.einsum.html
@@ -596,7 +596,7 @@
 
                                  torch.einsum
                                  
 
                                  
 
                                  
-torch.einsum(equation, *operands) → Tensor[source][source]
                                  
+torch.einsum(equation, *operands) → Tensor[source][source]
 
                                  Sums the product of the elements of the input operands along dimensions specified using a notation
 based on the Einstein summation convention.
                                  
 
                                  Einsum allows computing many common multi-dimensional linear algebraic array operations by representing them
diff --git a/2.7/generated/torch.enable_grad.html b/2.7/generated/torch.enable_grad.html
index 15ff7d8ecf0..df5b0081299 100644
--- a/2.7/generated/torch.enable_grad.html
+++ b/2.7/generated/torch.enable_grad.html
@@ -596,7 +596,7 @@
 
                                  enable_grad
                                  
 
                                  
 
                                  
-class torch.enable_grad(orig_func=None)[source][source]
                                  
+class torch.enable_grad(orig_func=None)[source][source]
 
                                  Context-manager that enables gradient calculation.
                                  
 
                                  Enables gradient calculation, if it has been disabled via no_grad
 or set_grad_enabled.
                                  
diff --git a/2.7/generated/torch.from_dlpack.html b/2.7/generated/torch.from_dlpack.html
index 569e6835660..f137f8f9a01 100644
--- a/2.7/generated/torch.from_dlpack.html
+++ b/2.7/generated/torch.from_dlpack.html
@@ -596,7 +596,7 @@
 
                                  torch.from_dlpack
                                  
 
                                  
 
                                  
-torch.from_dlpack(ext_tensor) → Tensor[source][source]
                                  
+torch.from_dlpack(ext_tensor) → Tensor[source][source]
 
                                  Converts a tensor from an external library into a torch.Tensor.
                                  
 
                                  The returned PyTorch tensor will share the memory with the input tensor
 (which may have come from another library). Note that in-place operations
diff --git a/2.7/generated/torch.func.debug_unwrap.html b/2.7/generated/torch.func.debug_unwrap.html
index 87b13a2828b..78d3b102a35 100644
--- a/2.7/generated/torch.func.debug_unwrap.html
+++ b/2.7/generated/torch.func.debug_unwrap.html
@@ -598,7 +598,7 @@
 
                                  torch.func.debug_unwrap
                                  
 
                                  
 
                                  
-torch.func.debug_unwrap(tensor, *, recurse=True)[source]
                                  
+torch.func.debug_unwrap(tensor, *, recurse=True)[source]
 
                                  Unwraps a functorch tensor (e.g. BatchedTensor, GradTrackingTensor) to its underlying tensor.
                                  
 
                                  This function should only be used in a debug setting (e.g. trying to print the
 value of a Tensor in a debugger). Otherwise, using the result of function
diff --git a/2.7/generated/torch.func.functional_call.html b/2.7/generated/torch.func.functional_call.html
index 60def3a08a3..813d792bf56 100644
--- a/2.7/generated/torch.func.functional_call.html
+++ b/2.7/generated/torch.func.functional_call.html
@@ -598,7 +598,7 @@
 
                                  torch.func.functional_call
                                  
 
                                  
 
                                  
-torch.func.functional_call(module, parameter_and_buffer_dicts, args=None, kwargs=None, *, tie_weights=True, strict=False)[source]
                                  
+torch.func.functional_call(module, parameter_and_buffer_dicts, args=None, kwargs=None, *, tie_weights=True, strict=False)[source]
 
                                  Performs a functional call on the module by replacing the module parameters
 and buffers with the provided ones.
                                  
 
                                  
diff --git a/2.7/generated/torch.func.functionalize.html b/2.7/generated/torch.func.functionalize.html
index 45b626b3f49..2491478676a 100644
--- a/2.7/generated/torch.func.functionalize.html
+++ b/2.7/generated/torch.func.functionalize.html
@@ -598,7 +598,7 @@
 
                                  torch.func.functionalize
                                  
 
                                  
 
                                  
-torch.func.functionalize(func, *, remove='mutations')[source]
                                  
+torch.func.functionalize(func, *, remove='mutations')[source]
 
                                  functionalize is a transform that can be used to remove (intermediate)
 mutations and aliasing from a function, while preserving the function’s
 semantics.
                                  
diff --git a/2.7/generated/torch.func.grad.html b/2.7/generated/torch.func.grad.html
index 8490f49ff1f..122a994160f 100644
--- a/2.7/generated/torch.func.grad.html
+++ b/2.7/generated/torch.func.grad.html
@@ -598,7 +598,7 @@
 
                                  torch.func.grad
                                  
 
                                  
 
                                  
-torch.func.grad(func, argnums=0, has_aux=False)[source]
                                  
+torch.func.grad(func, argnums=0, has_aux=False)[source]
 
                                  grad operator helps computing gradients of func with respect to the
 input(s) specified by argnums. This operator can be nested to
 compute higher-order gradients.
                                  
diff --git a/2.7/generated/torch.func.grad_and_value.html b/2.7/generated/torch.func.grad_and_value.html
index 628de977cef..ff4c93ed858 100644
--- a/2.7/generated/torch.func.grad_and_value.html
+++ b/2.7/generated/torch.func.grad_and_value.html
@@ -598,7 +598,7 @@
 
                                  torch.func.grad_and_value
                                  
 
                                  
 
                                  
-torch.func.grad_and_value(func, argnums=0, has_aux=False)[source]
                                  
+torch.func.grad_and_value(func, argnums=0, has_aux=False)[source]
 
                                  Returns a function to compute a tuple of the gradient and primal, or
 forward, computation.
                                  
 
                                  
diff --git a/2.7/generated/torch.func.hessian.html b/2.7/generated/torch.func.hessian.html
index 75be688ca3e..bd1cba90855 100644
--- a/2.7/generated/torch.func.hessian.html
+++ b/2.7/generated/torch.func.hessian.html
@@ -598,7 +598,7 @@
 
                                  torch.func.hessian
                                  
 
                                  
 
                                  
-torch.func.hessian(func, argnums=0)[source]
                                  
+torch.func.hessian(func, argnums=0)[source]
 
                                  Computes the Hessian of func with respect to the arg(s) at index
 argnum via a forward-over-reverse strategy.
                                  
 
                                  The forward-over-reverse strategy (composing jacfwd(jacrev(func))) is
diff --git a/2.7/generated/torch.func.jacfwd.html b/2.7/generated/torch.func.jacfwd.html
index dc728ae882f..cce214b0801 100644
--- a/2.7/generated/torch.func.jacfwd.html
+++ b/2.7/generated/torch.func.jacfwd.html
@@ -598,7 +598,7 @@
 
                                  torch.func.jacfwd
                                  
 
                                  
 
                                  
-torch.func.jacfwd(func, argnums=0, has_aux=False, *, randomness='error')[source]
                                  
+torch.func.jacfwd(func, argnums=0, has_aux=False, *, randomness='error')[source]
 
                                  Computes the Jacobian of func with respect to the arg(s) at index
 argnum using forward-mode autodiff
                                  
 
                                  
diff --git a/2.7/generated/torch.func.jacrev.html b/2.7/generated/torch.func.jacrev.html
index 661d42f5ac6..e3424df0fd4 100644
--- a/2.7/generated/torch.func.jacrev.html
+++ b/2.7/generated/torch.func.jacrev.html
@@ -598,7 +598,7 @@
 
                                  torch.func.jacrev
                                  
 
                                  
 
                                  
-torch.func.jacrev(func, argnums=0, *, has_aux=False, chunk_size=None, _preallocate_and_copy=False)[source]
                                  
+torch.func.jacrev(func, argnums=0, *, has_aux=False, chunk_size=None, _preallocate_and_copy=False)[source]
 
                                  Computes the Jacobian of func with respect to the arg(s) at index
 argnum using reverse mode autodiff
                                  
 
                                  
diff --git a/2.7/generated/torch.func.jvp.html b/2.7/generated/torch.func.jvp.html
index efb376e07ca..367889a9242 100644
--- a/2.7/generated/torch.func.jvp.html
+++ b/2.7/generated/torch.func.jvp.html
@@ -598,7 +598,7 @@
 
                                  torch.func.jvp
                                  
 
                                  
 
                                  
-torch.func.jvp(func, primals, tangents, *, strict=False, has_aux=False)[source]
                                  
+torch.func.jvp(func, primals, tangents, *, strict=False, has_aux=False)[source]
 
                                  Standing for the Jacobian-vector product, returns a tuple containing
 the output of func(*primals) and the “Jacobian of func evaluated at
 primals” times tangents. This is also known as forward-mode autodiff.
                                  
diff --git a/2.7/generated/torch.func.linearize.html b/2.7/generated/torch.func.linearize.html
index 27c55a69252..372a88e8cc0 100644
--- a/2.7/generated/torch.func.linearize.html
+++ b/2.7/generated/torch.func.linearize.html
@@ -598,7 +598,7 @@
 
                                  torch.func.linearize
                                  
 
                                  
 
                                  
-torch.func.linearize(func, *primals)[source]
                                  
+torch.func.linearize(func, *primals)[source]
 
                                  Returns the value of func at primals and linear approximation
 at primals.
                                  
 
                                  
diff --git a/2.7/generated/torch.func.replace_all_batch_norm_modules_.html b/2.7/generated/torch.func.replace_all_batch_norm_modules_.html
index fad7d74caa0..0e520f605c2 100644
--- a/2.7/generated/torch.func.replace_all_batch_norm_modules_.html
+++ b/2.7/generated/torch.func.replace_all_batch_norm_modules_.html
@@ -598,7 +598,7 @@
 
                                  torch.func.replace_all_batch_norm_modules_
                                  
 
                                  
 
                                  
-torch.func.replace_all_batch_norm_modules_(root)[source]
                                  
+torch.func.replace_all_batch_norm_modules_(root)[source]
 
                                  In place updates root by setting the running_mean and running_var to be None and
 setting track_running_stats to be False for any nn.BatchNorm module in root
                                  
 
                                  
diff --git a/2.7/generated/torch.func.stack_module_state.html b/2.7/generated/torch.func.stack_module_state.html
index 504d4bd6d4a..7a08c91a360 100644
--- a/2.7/generated/torch.func.stack_module_state.html
+++ b/2.7/generated/torch.func.stack_module_state.html
@@ -598,7 +598,7 @@
 
                                  torch.func.stack_module_state
                                  
 
                                  
 
                                  
-torch.func.stack_module_state(models) → params, buffers[source]
                                  
+torch.func.stack_module_state(models) → params, buffers[source]
 
                                  Prepares a list of torch.nn.Modules for ensembling with vmap().
                                  
 
                                  Given a list of M nn.Modules of the same class, returns two dictionaries
 that stack all of their parameters and buffers together, indexed by name.
diff --git a/2.7/generated/torch.func.vjp.html b/2.7/generated/torch.func.vjp.html
index f3560c42a79..271e0542e4f 100644
--- a/2.7/generated/torch.func.vjp.html
+++ b/2.7/generated/torch.func.vjp.html
@@ -598,7 +598,7 @@
 
                                  torch.func.vjp
                                  
 
                                  
 
                                  
-torch.func.vjp(func, *primals, has_aux=False)[source]
                                  
+torch.func.vjp(func, *primals, has_aux=False)[source]
 
                                  Standing for the vector-Jacobian product, returns a tuple containing the
 results of func applied to primals and a function that, when
 given cotangents, computes the reverse-mode Jacobian of func with
diff --git a/2.7/generated/torch.func.vmap.html b/2.7/generated/torch.func.vmap.html
index bf3507d91a6..8140ca2b42c 100644
--- a/2.7/generated/torch.func.vmap.html
+++ b/2.7/generated/torch.func.vmap.html
@@ -598,7 +598,7 @@
 
                                  torch.func.vmap
                                  
 
                                  
 
                                  
-torch.func.vmap(func, in_dims=0, out_dims=0, randomness='error', *, chunk_size=None)[source]
                                  
+torch.func.vmap(func, in_dims=0, out_dims=0, randomness='error', *, chunk_size=None)[source]
 
                                  vmap is the vectorizing map; vmap(func) returns a new function that
 maps func over some dimension of the inputs. Semantically, vmap
 pushes the map into PyTorch operations called by func, effectively
diff --git a/2.7/generated/torch.fx.experimental.proxy_tensor.get_proxy_mode.html b/2.7/generated/torch.fx.experimental.proxy_tensor.get_proxy_mode.html
index 0702bc3d392..2b696749441 100644
--- a/2.7/generated/torch.fx.experimental.proxy_tensor.get_proxy_mode.html
+++ b/2.7/generated/torch.fx.experimental.proxy_tensor.get_proxy_mode.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.proxy_tensor.get_proxy_mode
                                  
 
                                  
 
                                  
-torch.fx.experimental.proxy_tensor.get_proxy_mode()[source][source]
                                  
+torch.fx.experimental.proxy_tensor.get_proxy_mode()[source][source]
 
                                  Current the currently active proxy tracing mode, or None if
 we are not currently tracing.  This includes pre-dispatch proxy
 tracing.
                                  
diff --git a/2.7/generated/torch.fx.experimental.proxy_tensor.handle_sym_dispatch.html b/2.7/generated/torch.fx.experimental.proxy_tensor.handle_sym_dispatch.html
index 29c16b6cd9b..9657640f812 100644
--- a/2.7/generated/torch.fx.experimental.proxy_tensor.handle_sym_dispatch.html
+++ b/2.7/generated/torch.fx.experimental.proxy_tensor.handle_sym_dispatch.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.proxy_tensor.handle_sym_dispatch
                                  
 
                                  
 
                                  
-torch.fx.experimental.proxy_tensor.handle_sym_dispatch(func, args, kwargs)[source][source]
                                  
+torch.fx.experimental.proxy_tensor.handle_sym_dispatch(func, args, kwargs)[source][source]
 
                                  Call into the currently active proxy tracing mode to do a
 SymInt/SymFloat/SymBool dispatch trace on a function that operates on
 these arguments.
                                  
diff --git a/2.7/generated/torch.fx.experimental.proxy_tensor.make_fx.html b/2.7/generated/torch.fx.experimental.proxy_tensor.make_fx.html
index e144808820f..52fb9d1350f 100644
--- a/2.7/generated/torch.fx.experimental.proxy_tensor.make_fx.html
+++ b/2.7/generated/torch.fx.experimental.proxy_tensor.make_fx.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.proxy_tensor.make_fx
                                  
 
                                  
 
                                  
-torch.fx.experimental.proxy_tensor.make_fx(f, decomposition_table=None, tracing_mode='real', _allow_non_fake_inputs=False, *, pre_dispatch=False, record_module_stack=False, _allow_fake_constant=False, _error_on_data_dependent_ops=True)[source][source]
                                  
+torch.fx.experimental.proxy_tensor.make_fx(f, decomposition_table=None, tracing_mode='real', _allow_non_fake_inputs=False, *, pre_dispatch=False, record_module_stack=False, _allow_fake_constant=False, _error_on_data_dependent_ops=True)[source][source]
 
                                  Given a function f, return a new function which when executed with valid
 arguments to f, returns an FX GraphModule representing the set of operations that
 were executed during the course of execution.
                                  
diff --git a/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_disable_thunkify.html b/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_disable_thunkify.html
index 7f11eaebac3..775ad33ce45 100644
--- a/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_disable_thunkify.html
+++ b/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_disable_thunkify.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.proxy_tensor.maybe_disable_thunkify
                                  
 
                                  
 
                                  
-torch.fx.experimental.proxy_tensor.maybe_disable_thunkify()[source][source]
                                  
+torch.fx.experimental.proxy_tensor.maybe_disable_thunkify()[source][source]
 
                                  Within a context, disable thunkification.  See maybe_enable_thunkify()
 for more details.  This is helpful if you have a wrapper function which
 you want to enable thunkification on, but in some segment on the inside (say,
diff --git a/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_enable_thunkify.html b/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_enable_thunkify.html
index 0829bdadd04..9a26118151f 100644
--- a/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_enable_thunkify.html
+++ b/2.7/generated/torch.fx.experimental.proxy_tensor.maybe_enable_thunkify.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.proxy_tensor.maybe_enable_thunkify
                                  
 
                                  
 
                                  
-torch.fx.experimental.proxy_tensor.maybe_enable_thunkify()[source][source]
                                  
+torch.fx.experimental.proxy_tensor.maybe_enable_thunkify()[source][source]
 
                                  Within this context manager, if you are doing make_fx tracing, we will thunkify
 all SymNode compute and avoid tracing it into the graph unless it is actually needed.
 You should prefer to avoid using this as much as possible, as lazy evaluation of
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.CallMethodKey.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.CallMethodKey.html
index 8e97721dc2b..16ea9752a6d 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.CallMethodKey.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.CallMethodKey.html
@@ -596,12 +596,12 @@
 
                                  CallMethodKey
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.CallMethodKey(name: 'str')[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.CallMethodKey(name: 'str')[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-get(o)[source][source]
                                  
+get(o)[source][source]
 
                                  Call the method on object
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.ConvertIntKey.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.ConvertIntKey.html
index 8dcc291925a..3d83b335db5 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.ConvertIntKey.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.ConvertIntKey.html
@@ -596,12 +596,12 @@
 
                                  ConvertIntKey
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.ConvertIntKey[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.ConvertIntKey[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-get(b)[source][source]
                                  
+get(b)[source][source]
 
                                  Get the int value from bool
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.DimConstraints.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.DimConstraints.html
index 0938df58ef9..1290768aa51 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.DimConstraints.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.DimConstraints.html
@@ -596,14 +596,14 @@
 
                                  DimConstraints
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.DimConstraints(symbol_to_source, var_to_val, marked_dynamic, source_name_to_debug_name)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.DimConstraints(symbol_to_source, var_to_val, marked_dynamic, source_name_to_debug_name)[source][source]
 
                                  Custom solver for a system of constraints on symbolic dimensions.
 Solutions are “static” values or simplified “dynamic” constraints.
                                  
 
                                  
 
                                  
 
                                  
 
                                  
-add(expr)[source][source]
                                  
+add(expr)[source][source]
 
                                  Add an expression to the set of constraints.
                                  
 
                                  Return whether the expression is a trivial constraint (i.e., an obvious tautology).
                                  
 
                                  
@@ -615,7 +615,7 @@ 
                                  DimConstraints
 
                                  
-add_equality(source, expr)[source][source]
                                  
+add_equality(source, expr)[source][source]
 
                                  Add an equality constraint
                                  
 
                                  
 
                                  
@@ -623,7 +623,7 @@ 
                                  DimConstraints
 
                                  
-forced_specializations()[source][source]
                                  
+forced_specializations()[source][source]
 
                                  Returns a dictionary of the names of symbols to their specialized value
                                  
 
                                  
 
                                  Return type
                                  
@@ -634,7 +634,7 @@ 
                                  DimConstraints
 
                                  
-prettify_results(original_signature, dynamic_shapes, constraint_violation_error, forced_specializations)[source][source]
                                  
+prettify_results(original_signature, dynamic_shapes, constraint_violation_error, forced_specializations)[source][source]
 
                                  Format a message for constraint violation erros
                                  
 
                                  
 
                                  Return type
                                  
@@ -645,7 +645,7 @@ 
                                  DimConstraints
 
                                  
-rewrite_with_congruences(s, expr)[source][source]
                                  
+rewrite_with_congruences(s, expr)[source][source]
 
                                  Eliminate expressions of the form b // d and b % d while adding congruences of the form b % d == k.
 This leaves rational operators (in particular of the form b / d) that our inequality solver can handle.
 We solve the added congruences separately (using our congruence solver, see below).
                                  
@@ -658,7 +658,7 @@ 
                                  DimConstraints
 
                                  
-solve()[source][source]
                                  
+solve()[source][source]
 
                                  Solve the system of constraint equations to find simplified constraints
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.DimDynamic.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.DimDynamic.html
index bbbd38896fd..1224f49890a 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.DimDynamic.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.DimDynamic.html
@@ -596,7 +596,7 @@
 
                                  DimDynamic
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.DimDynamic(value)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.DimDynamic(value)[source][source]
 
                                  Controls how to perform symbol allocation for a dimension.  It is always
 sound to default this to DYNAMIC, but the policies DUCK and STATIC can
 result in better trace-time and compile-time performance, as they reduce
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.DivideByKey.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.DivideByKey.html
index 01b8ae41c4e..302446a59f9 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.DivideByKey.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.DivideByKey.html
@@ -596,12 +596,12 @@
 
                                  DivideByKey
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.DivideByKey(divisor: 'Union[int, SymInt]')[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.DivideByKey(divisor: 'Union[int, SymInt]')[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-get(o)[source][source]
                                  
+get(o)[source][source]
 
                                  Divide object by divisor
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.EqualityConstraint.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.EqualityConstraint.html
index fd5a2311ff6..8c5bb122754 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.EqualityConstraint.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.EqualityConstraint.html
@@ -596,7 +596,7 @@
 
                                  EqualityConstraint
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.EqualityConstraint(warn_only, source_pairs, derived_equalities, phantom_symbols, relaxed_sources)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.EqualityConstraint(warn_only, source_pairs, derived_equalities, phantom_symbols, relaxed_sources)[source][source]
 
                                  Represent and decide various kinds of equality constraints between input sources.
                                  
 
                                  A “source pair” is a pair of input sources for dynamic dimensions that
 are specified equal. We represent source_pairs in a union-find forest
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.InnerTensorKey.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.InnerTensorKey.html
index 51b3e4cf17c..d796f24a668 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.InnerTensorKey.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.InnerTensorKey.html
@@ -596,12 +596,12 @@
 
                                  InnerTensorKey
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.InnerTensorKey(inner_name: 'str')[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.InnerTensorKey(inner_name: 'str')[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-get(o)[source][source]
                                  
+get(o)[source][source]
 
                                  Get the inner tensor attribute
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts.html
index b529b70eca0..c9f22a25941 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts.html
@@ -596,12 +596,12 @@
 
                                  PropagateUnbackedSymInts
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts(module, garbage_collect_values=True, graph=None)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.PropagateUnbackedSymInts(module, garbage_collect_values=True, graph=None)[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-boxed_run(args_list)[source]
                                  
+boxed_run(args_list)[source]
 
                                  Run module via interpretation and return the result.  This uses the “boxed”
 calling convention, where you pass a list of arguments, which will be cleared
 by the interpreter.  This ensures that input tensors are promptly deallocated.
                                  
@@ -613,7 +613,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-call_function(target, args, kwargs)[source]
                                  
+call_function(target, args, kwargs)[source]
 
                                  Execute a call_function node and return the result.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -641,7 +641,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-call_method(target, args, kwargs)[source]
                                  
+call_method(target, args, kwargs)[source]
 
                                  Execute a call_method node and return the result.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -669,7 +669,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-call_module(target, args, kwargs)[source]
                                  
+call_module(target, args, kwargs)[source]
 
                                  Execute a call_module node and return the result.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -697,7 +697,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-fetch_args_kwargs_from_env(n)[source]
                                  
+fetch_args_kwargs_from_env(n)[source]
 
                                  Fetch the concrete values of args and kwargs of node n
 from the current execution environment.
                                  
 
                                  
@@ -719,7 +719,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-fetch_attr(target)[source]
                                  
+fetch_attr(target)[source]
 
                                  Fetch an attribute from the Module hierarchy of self.module.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -740,7 +740,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-get_attr(target, args, kwargs)[source]
                                  
+get_attr(target, args, kwargs)[source]
 
                                  Execute a get_attr node. Will retrieve an attribute
 value from the Module hierarchy of self.module.
                                  
 
                                  
@@ -768,7 +768,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-map_nodes_to_values(args, n)[source]
                                  
+map_nodes_to_values(args, n)[source]
 
                                  Recursively descend through args and look up the concrete value
 for each Node in the current execution environment.
                                  
 
                                  
@@ -790,7 +790,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-output(target, args, kwargs)[source]
                                  
+output(target, args, kwargs)[source]
 
                                  Execute an output node. This really just retrieves
 the value referenced by the output node and returns it.
                                  
 
                                  
@@ -818,7 +818,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-placeholder(target, args, kwargs)[source]
                                  
+placeholder(target, args, kwargs)[source]
 
                                  Execute a placeholder node. Note that this is stateful:
 Interpreter maintains an internal iterator over
 arguments passed to run and this method returns
@@ -848,7 +848,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-run(*args, initial_env=None, enable_io_processing=True)[source]
                                  
+run(*args, initial_env=None, enable_io_processing=True)[source]
 
                                  Run module via interpretation and return the result.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -877,7 +877,7 @@ 
                                  PropagateUnbackedSymInts
 
                                  
-run_node(n)[source][source]
                                  
+run_node(n)[source][source]
 
                                  Run an FX node, propagating unbacked Symbol bindings to the new fake tensor
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint.html
index bf509fa64cf..f9de3a87f36 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint.html
@@ -596,7 +596,7 @@
 
                                  RelaxedUnspecConstraint
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint(warn_only)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.RelaxedUnspecConstraint(warn_only)[source][source]
 
                                  For clients: no explicit constraint; constraint is whatever is implicitly
 inferred by guards from tracing.
                                  
 
                                  For backends: there must exist at least TWO possible values for the
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnv.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnv.html
index 6a40f9c14fe..eec8c6a6e1e 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnv.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnv.html
@@ -596,12 +596,12 @@
 
                                  ShapeEnv
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.ShapeEnv(*, should_record_events=None, tracked_fakes=None, **kwargs)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.ShapeEnv(*, should_record_events=None, tracked_fakes=None, **kwargs)[source][source]
 
                                  
 
                                  
 
                                  
 
                                  
-add_var_to_val(expr, val)[source][source]
                                  
+add_var_to_val(expr, val)[source][source]
 
                                  Adds a new symbol to the symbolic environment.
                                  
 
                                  
 
                                  
@@ -609,7 +609,7 @@ 
                                  ShapeEnv
 
                                  
-bind_symbols(placeholders, args)[source][source]
                                  
+bind_symbols(placeholders, args)[source][source]
 
                                  Given a paired list of placeholders (fake tensors with
 symbolic sizes) and concrete arguments (regular tensors
 with real sizes), returns a dictionary mapping each
@@ -632,7 +632,7 @@ 
                                  ShapeEnv
 
                                  
-bound_sympy(expr, size_oblivious=False)[source][source]
                                  
+bound_sympy(expr, size_oblivious=False)[source][source]
 
                                  Given a sympy expression, computes a ValueRanges bound for what values it can be
                                  
 
                                  
 
                                  Return type
                                  
@@ -643,7 +643,7 @@ 
                                  ShapeEnv
 
                                  
-check_equal(other)[source][source]
                                  
+check_equal(other)[source][source]
 
                                  Compare another ShapeEnv for equivalence
                                  
 
                                  
 
                                  
@@ -651,7 +651,7 @@ 
                                  ShapeEnv
 
                                  
-cleanup()[source][source]
                                  
+cleanup()[source][source]
 
                                  Break reference cycles.
                                  
 
                                  This destroys the stacks. If you really want to keep them, we
 just need some way to break references on code objects.
                                  
@@ -661,7 +661,7 @@ 
                                  ShapeEnv
 
                                  
-create_symbol(val, source, dynamic_dim=DimDynamic.DUCK, constraint_dim=None, positive=True, do_not_specialize_zero_one=False, symbolic_context=None)[source][source]
                                  
+create_symbol(val, source, dynamic_dim=DimDynamic.DUCK, constraint_dim=None, positive=True, do_not_specialize_zero_one=False, symbolic_context=None)[source][source]
 
                                  Create a new symbol which is tracked by this ShapeEnv
                                  
 
                                  
 
                                  Return type
                                  
@@ -672,7 +672,7 @@ 
                                  ShapeEnv
 
                                  
-create_symbolic_sizes_strides_storage_offset(ex, source, *, symbolic_context=None)[source][source]
                                  
+create_symbolic_sizes_strides_storage_offset(ex, source, *, symbolic_context=None)[source][source]
 
                                  Returns a list of symbolic sizes and strides for the given tensor.
 We try our best to express stride in terms of the sizes, so as to not
 introduce new symbolic variables.
                                  
@@ -685,7 +685,7 @@ 
                                  ShapeEnv
 
                                  
-create_symboolnode(sym)[source][source]
                                  
+create_symboolnode(sym)[source][source]
 
                                  Create a SymBool object from a sympy boolean expression
                                  
 
                                  
 
                                  Return type
                                  
@@ -696,7 +696,7 @@ 
                                  ShapeEnv
 
                                  
-create_symfloatnode(sym, *, hint, source=None)[source][source]
                                  
+create_symfloatnode(sym, *, hint, source=None)[source][source]
 
                                  Create a SymFloat value from a symbolic expression
                                  
 
                                  
 
                                  Return type
                                  
@@ -707,7 +707,7 @@ 
                                  ShapeEnv
 
                                  
-create_symintnode(sym, *, hint, source=None)[source][source]
                                  
+create_symintnode(sym, *, hint, source=None)[source][source]
 
                                  Create a SymInt value from a symbolic expression
                                  
 
                                  If you know what the current hint value of the SymInt to be created
 is, pass it into hint.  Otherwise, pass None and we will make our best
@@ -721,7 +721,7 @@ 
                                  ShapeEnv
 
                                  
-create_unbacked_symbool()[source][source]
                                  
+create_unbacked_symbool()[source][source]
 
                                  Create a symbolic boolean without a hint value
                                  
 
                                  
 
                                  Return type
                                  
@@ -732,7 +732,7 @@ 
                                  ShapeEnv
 
                                  
-create_unbacked_symfloat()[source][source]
                                  
+create_unbacked_symfloat()[source][source]
 
                                  Create a symbolic float without a hint value
                                  
 
                                  
 
                                  Return type
                                  
@@ -743,7 +743,7 @@ 
                                  ShapeEnv
 
                                  
-create_unbacked_symint(source=None)[source][source]
                                  
+create_unbacked_symint(source=None)[source][source]
 
                                  Create a symbolic integer without a hint value
                                  
 
                                  
 
                                  Return type
                                  
@@ -754,7 +754,7 @@ 
                                  ShapeEnv
 
                                  
-create_unspecified_symbol(val, source, dynamic_dim=DimDynamic.DUCK, constraint_dim=None, symbolic_context=None)[source][source]
                                  
+create_unspecified_symbol(val, source, dynamic_dim=DimDynamic.DUCK, constraint_dim=None, symbolic_context=None)[source][source]
 
                                  Create a symbol with an unspecified value
                                  
 
                                  Compared to standard symbols we do not assume the value is positive,
 nor do we specialze on zero or one values.
                                  
@@ -767,7 +767,7 @@ 
                                  ShapeEnv
 
                                  
-create_unspecified_symint_and_symbol(value, source, dynamic_dim)[source][source]
                                  
+create_unspecified_symint_and_symbol(value, source, dynamic_dim)[source][source]
 
                                  Create a SymInt wrapping a new unspecified symbol
                                  
 
                                  
 
                                  Return type
                                  
@@ -778,7 +778,7 @@ 
                                  ShapeEnv
 
                                  
-defer_runtime_assert(orig_expr, msg, fx_node=None)[source][source]
                                  
+defer_runtime_assert(orig_expr, msg, fx_node=None)[source][source]
 
                                  Create an assert that is checked at runtime
                                  
 
                                  
 
                                  Parameters
                                  
@@ -797,7 +797,7 @@ 
                                  ShapeEnv
 
                                  
-deserialize_symexpr(code)[source][source]
                                  
+deserialize_symexpr(code)[source][source]
 
                                  To be used by compile_fx to deserialize symexprs
                                  
 
                                  
 
                                  Return type
                                  
@@ -808,7 +808,7 @@ 
                                  ShapeEnv
 
                                  
-evaluate_guards_expression(code, args)[source][source]
                                  
+evaluate_guards_expression(code, args)[source][source]
 
                                  Expected to be used with produce_guards_expression(). Evaluates an expression
 generated by produce_guards_expression for the given concrete args.
                                  
 
                                  
@@ -820,7 +820,7 @@ 
                                  ShapeEnv
 
                                  
-evaluate_guards_for_args(placeholders, args, *, ignore_static=True)[source][source]
                                  
+evaluate_guards_for_args(placeholders, args, *, ignore_static=True)[source][source]
 
                                  Generate guards for a graph’s placeholder values and evaluate the guards with args
                                  
 
                                  
 
                                  Return type
                                  
@@ -831,7 +831,7 @@ 
                                  ShapeEnv
 
                                  
-evaluate_sym_node(sym_node, size_oblivious=False)[source][source]
                                  
+evaluate_sym_node(sym_node, size_oblivious=False)[source][source]
 
                                  Given a a SymNode, evaluates sym_node.expr, adding guards if necessary.
                                  
 
                                  
 
                                  Return type
                                  
@@ -842,7 +842,7 @@ 
                                  ShapeEnv
 
                                  
-evaluate_symexpr(code)[source][source]
                                  
+evaluate_symexpr(code)[source][source]
 
                                  To be used by compile_fx to evaluate symexprs
                                  
 
                                  
 
                                  Return type
                                  
@@ -853,7 +853,7 @@ 
                                  ShapeEnv
 
                                  
-format_guards(verbose=False)[source][source]
                                  
+format_guards(verbose=False)[source][source]
 
                                  Format this shape env’s guard expressions with optional traceback info if verbose
                                  
 
                                  
 
                                  Return type
                                  
@@ -864,7 +864,7 @@ 
                                  ShapeEnv
 
                                  
-freeze()[source][source]
                                  
+freeze()[source][source]
 
                                  Freeze this ShapeEnv to stop accumulating guards
                                  
 
                                  A frozen ShapeEnv will ignore any further guards generated on it and
 only emit a warning which may lead to accuracy problems.
                                  
@@ -874,7 +874,7 @@ 
                                  ShapeEnv
 
                                  
-freeze_runtime_asserts()[source][source]
                                  
+freeze_runtime_asserts()[source][source]
 
                                  Freeze this ShapeEnv to stop adding deferred runtime asserts.
                                  
 
                                  We will error if you try to install a new runtime assert when it is
 frozen.  This would indicate a lowering violation, or perhaps something
@@ -886,7 +886,7 @@ 
                                  ShapeEnv
 
                                  
-get_axioms(symbols=None, compute_hint=False)[source][source]
                                  
+get_axioms(symbols=None, compute_hint=False)[source][source]
 
                                  Given the symbols in an expression, it returns all the runtime asserts that have those symbols
 concatenated with all the guards.
 If symbols is None, it returns all the runtime asserts (and all the guards)
                                  
@@ -899,7 +899,7 @@ 
                                  ShapeEnv
 
                                  
-get_implications(e)[source][source]
                                  
+get_implications(e)[source][source]
 
                                  Given a expression, it returns a list of predicates that follow from it
                                  
 
                                  
 
                                  Return type
                                  
@@ -910,7 +910,7 @@ 
                                  ShapeEnv
 
                                  
-get_nontrivial_guards()[source][source]
                                  
+get_nontrivial_guards()[source][source]
 
                                  Returns a list of guard expressions that aren’t statically known (i.e. not trivial)
                                  
 
                                  
 
                                  Return type
                                  
@@ -921,7 +921,7 @@ 
                                  ShapeEnv
 
                                  
-get_pruned_guards(symints)[source][source]
                                  
+get_pruned_guards(symints)[source][source]
 
                                  Get a list of guards, but pruned so it only provides guards that
 reference symints from the passed in input
                                  
 
                                  
@@ -933,7 +933,7 @@ 
                                  ShapeEnv
 
                                  
-ignore_fresh_unbacked_symbols()[source][source]
                                  
+ignore_fresh_unbacked_symbols()[source][source]
 
                                  Indicates that the newly allocated unbacked SymInts are being
 discarded
                                  
 
                                  
@@ -945,7 +945,7 @@ 
                                  ShapeEnv
 
                                  
-is_unbacked_symint(symbol)[source][source]
                                  
+is_unbacked_symint(symbol)[source][source]
 
                                  Check if a sympy symbol matches the naming convention for unbacked symbols
                                  
 
                                  
 
                                  Return type
                                  
@@ -956,7 +956,7 @@ 
                                  ShapeEnv
 
                                  
-produce_guards(*args, **kwargs)[source][source]
                                  
+produce_guards(*args, **kwargs)[source][source]
 
                                  Like produce_guards_verbose, but only returns the non-verbose python guard expressions
 (no verbose guards produced.)
                                  
 
                                  
@@ -968,7 +968,7 @@ 
                                  ShapeEnv
 
                                  
-produce_guards_expression(placeholders, *, guards=None, ignore_static=True)[source][source]
                                  
+produce_guards_expression(placeholders, *, guards=None, ignore_static=True)[source][source]
 
                                  Expected to be used with evaluate_guards_expression(). Produces the guards
 for the given placeholders and returns a string expression to be evaluated
 by evaluate_guards_expression given concrete values for the placeholders.
                                  
@@ -981,7 +981,7 @@ 
                                  ShapeEnv
 
                                  
-produce_guards_verbose(placeholders, sources, source_ref=<function ShapeEnv.<lambda>>, *, guards=None, input_contexts=None, equalities_inputs=None, _simplified=False, ignore_static=True, langs=('python', 'verbose_python'))[source][source]
                                  
+produce_guards_verbose(placeholders, sources, source_ref=<function ShapeEnv.<lambda>>, *, guards=None, input_contexts=None, equalities_inputs=None, _simplified=False, ignore_static=True, langs=('python', 'verbose_python'))[source][source]
 
                                  Generates a list of guards strings which, when evaluated in a context that
 defines tensors for all the sources, returns True or False depending
 on if the guards in the list evaluated to True or not.  Primarily used by Dynamo,
@@ -1006,7 +1006,7 @@ 
                                  ShapeEnv
 
                                  
-replace(expr)[source][source]
                                  
+replace(expr)[source][source]
 
                                  Apply symbol replacements to any symbols in the given expression
                                  
 
                                  
 
                                  Return type
                                  
@@ -1017,7 +1017,7 @@ 
                                  ShapeEnv
 
                                  
-set_unbacked_var_to_val(k, v)[source][source]
                                  
+set_unbacked_var_to_val(k, v)[source][source]
 
                                  Used only when propagate_real_tensors; registers a value for an
 unbacked symbol, which can be used last resort to resolve hints.
                                  
 
                                  
@@ -1026,7 +1026,7 @@ 
                                  ShapeEnv
 
                                  
-simplify(expr, size_oblivious=False)[source][source]
                                  
+simplify(expr, size_oblivious=False)[source][source]
 
                                  Use known constraints and replacements to simplify the given expr
                                  
 
                                  
 
                                  Return type
                                  
@@ -1037,7 +1037,7 @@ 
                                  ShapeEnv
 
                                  
-size_hint(expr, *, allow_none=False)[source][source]
                                  
+size_hint(expr, *, allow_none=False)[source][source]
 
                                  Gets a size hint for a given expression from the underlying shapes we had.
 Does not introduce a guard, so only use this when you can guarantee that
 your code is still valid for arbitrary shapes (such as optimization decisions)
                                  
@@ -1050,7 +1050,7 @@ 
                                  ShapeEnv
 
                                  
-suppress_guards()[source][source]
                                  
+suppress_guards()[source][source]
 
                                  Context manager to ignore all guards generated inside
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnvSettings.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnvSettings.html
index ccc14fd1a5c..f909759e677 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnvSettings.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.ShapeEnvSettings.html
@@ -596,7 +596,7 @@
 
                                  ShapeEnvSettings
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.ShapeEnvSettings(allow_scalar_outputs, allow_dynamic_output_shape_ops, assume_static_by_default, specialize_zero_one, duck_shape, prefer_deferred_runtime_asserts_over_guards, allow_complex_guards_as_runtime_asserts)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.ShapeEnvSettings(allow_scalar_outputs, allow_dynamic_output_shape_ops, assume_static_by_default, specialize_zero_one, duck_shape, prefer_deferred_runtime_asserts_over_guards, allow_complex_guards_as_runtime_asserts)[source][source]
 
                                  Encapsulates all shape env settings that could potentially affect
 FakeTensor dispatch. Used when creating dispatch cache keys.
                                  
 
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext.html
index 4297d60e450..a306c7f2f55 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext.html
@@ -596,7 +596,7 @@
 
                                  StatefulSymbolicContext
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None, tensor_source=None, shape_env_to_source_to_symbol_cache=None)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.StatefulSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None, tensor_source=None, shape_env_to_source_to_symbol_cache=None)[source][source]
 
                                  Create symbols in create_symbolic_sizes_strides_storage_offset via
 a symbolic_context determination as given by a cache of Source:Symbol. A cache hit
 will reuse a stored symbol, and a cache miss will write to this cache.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext.html
index 8f376c295b2..c8c9ffa2a04 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext.html
@@ -596,7 +596,7 @@
 
                                  StatelessSymbolicContext
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.StatelessSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None)[source][source]
 
                                  Create symbols in create_symbolic_sizes_strides_storage_offset via
 a symbolic_context determination as given by DimDynamic and DimConstraint.
 This will cause fresh symbols to be allocated
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint.html
index 534f8e0da97..4b258b32e1d 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint.html
@@ -596,7 +596,7 @@
 
                                  StrictMinMaxConstraint
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint(warn_only, vr)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.StrictMinMaxConstraint(warn_only, vr)[source][source]
 
                                  For clients: the size at this dimension must be within ‘vr’ (which
 specifies a lower and upper bound, inclusive-inclusive) AND it
 must be non-negative and should not be 0 or 1 (but see NB below).
                                  
@@ -614,7 +614,7 @@ 
                                  StrictMinMaxConstraint
 
                                  
 
                                  
-render(source)[source][source]
                                  
+render(source)[source][source]
 
                                  Format the constrain equation
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext.html
index ffdf016f7d6..675d88e9feb 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext.html
@@ -596,7 +596,7 @@
 
                                  SubclassSymbolicContext
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None, tensor_source=None, shape_env_to_source_to_symbol_cache=None, inner_contexts=None)[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.SubclassSymbolicContext(dynamic_sizes, dynamic_strides=None, constraint_sizes=None, constraint_strides=None, view_base_context=None, tensor_source=None, shape_env_to_source_to_symbol_cache=None, inner_contexts=None)[source][source]
 
                                  The correct symbolic context for a given inner tensor of a traceable tensor subclass
 may differ from that of the outer symbolic context. This structure allows for this
 flexibility, with inner symbolic contexts mapped via attr -> symbolic context.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.SymbolicContext.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.SymbolicContext.html
index fca56e5b4cd..2e4f30f4dd9 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.SymbolicContext.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.SymbolicContext.html
@@ -596,7 +596,7 @@
 
                                  SymbolicContext
                                  
 
                                  
 
                                  
-class torch.fx.experimental.symbolic_shapes.SymbolicContext[source][source]
                                  
+class torch.fx.experimental.symbolic_shapes.SymbolicContext[source][source]
 
                                  Data structure specifying how we should create symbols in
 create_symbolic_sizes_strides_storage_offset; e.g., should
 they be static or dynamic.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr.html
index 689aa44ea45..f453658d737 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr(expr)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.canonicalize_bool_expr(expr)[source][source]
 
                                  Canonicalize a boolean expression by transforming it into a lt / le
 inequality and moving all the non-constant terms to the rhs.
 We canonicalize And / Ors / Not via cnf and then canonicalize their subexpr
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.check_consistent.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.check_consistent.html
index 5e05eff2015..d67fa5be93a 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.check_consistent.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.check_consistent.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.check_consistent
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.check_consistent(new, old)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.check_consistent(new, old)[source][source]
 
                                  Test that two “meta” values (typically either Tensor or SymInt) have
 the same values, e.g., after retracing.  If we don’t understand the
 quantities in question, we’ll just skip the consistency check.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings.html
index 4ce1af4e258..11cfcc8501e 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings(shape_env, example_value, old_example_value=None, peek=False)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.compute_unbacked_bindings(shape_env, example_value, old_example_value=None, peek=False)[source][source]
 
                                  After having run fake tensor propagation and producing example_value
 result, traverse example_value looking for freshly bound unbacked
 symbols and record their paths for later.  It is an error if
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_range.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_range.html
index de2c54eebfe..e8afad73ac0 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_range.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_range.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.constrain_range
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.constrain_range(a, *, min, max=None)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.constrain_range(a, *, min, max=None)[source][source]
 
                                  Applies a constraint that the passed in SymInt must lie between min-max
 inclusive-inclusive, WITHOUT introducing a guard on the SymInt (meaning
 that it can be used on unbacked SymInts).  If min/max are None, we assume
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_unify.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_unify.html
index d614b34d495..2ef1b9a00f1 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_unify.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.constrain_unify.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.constrain_unify
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.constrain_unify(a, b)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.constrain_unify(a, b)[source][source]
 
                                  Given two SymInts, constrain them so that they must be equal.  NB:
 this will not work with SymInts that represent nontrivial expressions
 (yet!)
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_false.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_false.html
index d32f2a6785c..ed0199fb968 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_false.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_false.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.definitely_false
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.definitely_false(a)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.definitely_false(a)[source][source]
 
                                  Returns True only if we can tell that a is False, possibly introducing
 a guard in the process.  If a depends on some unbacked SymInt, we may
 return False even though there may exist a possible value of the SymInt
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_true.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_true.html
index 69b366ac1a3..e8842ee2348 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_true.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.definitely_true.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.definitely_true
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.definitely_true(a)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.definitely_true(a)[source][source]
 
                                  Returns True only if we can tell that a is True, possibly introducing
 a guard in the process.  If a depends on some unbacked SymInt, we may
 return False even though there may exist a possible value of the SymInt
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.guard_size_oblivious.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.guard_size_oblivious.html
index 1fd80b90fa2..92a659f2f63 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.guard_size_oblivious.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.guard_size_oblivious.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.guard_size_oblivious
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.guard_size_oblivious(expr)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.guard_size_oblivious(expr)[source][source]
 
                                  Perform a guard on a symbolic boolean expression in a size oblivious way.
 This is typically used when a non-oblivious test would result in a guard
 on a data dependent value of which we don’t know the value of at compile time.
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_symbols.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_symbols.html
index c5d176e77bc..1a22086a300 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_symbols.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_symbols.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.has_free_symbols
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.has_free_symbols(val)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.has_free_symbols(val)[source][source]
 
                                  Faster version of bool(free_symbols(val))
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols.html
index 044b9002b84..3b3a0a9bba4 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols(x)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.has_free_unbacked_symbols(x)[source][source]
 
                                  Faster version of bool(free_unbacked_symbols(val))
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.hint_int.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.hint_int.html
index bebf7588671..4408f0d0428 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.hint_int.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.hint_int.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.hint_int
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.hint_int(a, fallback=None)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.hint_int(a, fallback=None)[source][source]
 
                                  Retrieve the hint for an int (based on the underlying real values as observed
 at runtime).  If no hint is available (e.g., because data dependent shapes),
 if fallback is not None, use that instead (otherwise raise an error).
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_accessor_node.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_accessor_node.html
index 40c8e21d2f5..7693eaa36d9 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_accessor_node.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_accessor_node.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.is_accessor_node
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.is_accessor_node(node)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.is_accessor_node(node)[source][source]
 
                                  
 
                                  Return type
                                  
 
                                  bool
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_bool.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_bool.html
index 6b967e49408..6be32075cb9 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_bool.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_bool.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.is_concrete_bool
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.is_concrete_bool(a)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.is_concrete_bool(a)[source][source]
 
                                  Utility to check if underlying object
 in SymBool is concrete value. Also returns
 true if integer is passed in.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_float.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_float.html
index d1aca1efca1..c1f8f6839ec 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_float.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_float.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.is_concrete_float
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.is_concrete_float(a)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.is_concrete_float(a)[source][source]
 
                                  Utility to check if underlying object
 in SymInt is concrete value. Also returns
 true if integer is passed in.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_int.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_int.html
index 4edd6c79103..f20fde29de2 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_int.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.is_concrete_int.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.is_concrete_int
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.is_concrete_int(a)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.is_concrete_int(a)[source][source]
 
                                  Utility to check if underlying object
 in SymInt is concrete value. Also returns
 true if integer is passed in.
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.lru_cache.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.lru_cache.html
index 6767fe7d2ac..f47b9e10ad3 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.lru_cache.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.lru_cache.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.lru_cache
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.lru_cache(maxsize)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.lru_cache(maxsize)[source][source]
 
                                  
 
                                  Return type
                                  
 
                                  Callable[[Callable[…, _T]], functools._lru_cache_wrapper[_T]]
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.rebind_unbacked.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.rebind_unbacked.html
index afb702ff591..a42305257c3 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.rebind_unbacked.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.rebind_unbacked.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.rebind_unbacked
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.rebind_unbacked(shape_env, n, result)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.rebind_unbacked(shape_env, n, result)[source][source]
 
                                  Suppose we are retracing a pre-existing FX graph that previously had
 fake tensor propagation (and therefore unbacked SymInts).  When we retrace,
 we re-propagate fake tensors, which results in new unbacked SymInts.
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings.html
index bfb31a0b308..ddd51616930 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings(shape_env, bindings)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.resolve_unbacked_bindings(shape_env, bindings)[source][source]
 
                                  
 
                                  Return type
                                  
 
                                  Optional[dict[sympy.core.symbol.Symbol, tuple[torch.utils._pytree.KeyEntry, …]]]
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.statically_known_true.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.statically_known_true.html
index 6b8c95c5408..cbdc406faee 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.statically_known_true.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.statically_known_true.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.statically_known_true
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.statically_known_true(x)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.statically_known_true(x)[source][source]
 
                                  Returns True if x can be simplified to a constant and is true.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.fx.experimental.symbolic_shapes.sym_eq.html b/2.7/generated/torch.fx.experimental.symbolic_shapes.sym_eq.html
index 8a93dbea495..b2facf832e1 100644
--- a/2.7/generated/torch.fx.experimental.symbolic_shapes.sym_eq.html
+++ b/2.7/generated/torch.fx.experimental.symbolic_shapes.sym_eq.html
@@ -596,7 +596,7 @@
 
                                  torch.fx.experimental.symbolic_shapes.sym_eq
                                  
 
                                  
 
                                  
-torch.fx.experimental.symbolic_shapes.sym_eq(x, y)[source][source]
                                  
+torch.fx.experimental.symbolic_shapes.sym_eq(x, y)[source][source]
 
                                  Like ==, but when run on list/tuple, it will recursively test equality
 and use sym_and to join the results together, without guarding.
                                  
 
                                  
diff --git a/2.7/generated/torch.get_default_device.html b/2.7/generated/torch.get_default_device.html
index 42013022ceb..5749717da16 100644
--- a/2.7/generated/torch.get_default_device.html
+++ b/2.7/generated/torch.get_default_device.html
@@ -596,7 +596,7 @@
 
                                  torch.get_default_device
                                  
 
                                  
 
                                  
-torch.get_default_device()[source][source]
                                  
+torch.get_default_device()[source][source]
 
                                  Gets the default torch.Tensor to be allocated on device
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.get_deterministic_debug_mode.html b/2.7/generated/torch.get_deterministic_debug_mode.html
index 48f5909cf54..398ae8d98a9 100644
--- a/2.7/generated/torch.get_deterministic_debug_mode.html
+++ b/2.7/generated/torch.get_deterministic_debug_mode.html
@@ -596,7 +596,7 @@
 
                                  torch.get_deterministic_debug_mode
                                  
 
                                  
 
                                  
-torch.get_deterministic_debug_mode()[source][source]
                                  
+torch.get_deterministic_debug_mode()[source][source]
 
                                  Returns the current value of the debug mode for deterministic
 operations. Refer to torch.set_deterministic_debug_mode()
 documentation for more details.
                                  
diff --git a/2.7/generated/torch.get_device_module.html b/2.7/generated/torch.get_device_module.html
index f1d536c7477..13e4d37742a 100644
--- a/2.7/generated/torch.get_device_module.html
+++ b/2.7/generated/torch.get_device_module.html
@@ -596,7 +596,7 @@
 
                                  torch.get_device_module
                                  
 
                                  
 
                                  
-torch.get_device_module(device=None)[source][source]
                                  
+torch.get_device_module(device=None)[source][source]
 
                                  Returns the module associated with a given device(e.g., torch.device(‘cuda’), “mtia:0”, “xpu”, …).
 If no device is given, return the module for the current accelerator or CPU if none is present.
                                  
 
                                  
diff --git a/2.7/generated/torch.get_float32_matmul_precision.html b/2.7/generated/torch.get_float32_matmul_precision.html
index 1821af5d6af..888ff503c69 100644
--- a/2.7/generated/torch.get_float32_matmul_precision.html
+++ b/2.7/generated/torch.get_float32_matmul_precision.html
@@ -596,7 +596,7 @@
 
                                  torch.get_float32_matmul_precision
                                  
 
                                  
 
                                  
-torch.get_float32_matmul_precision()[source][source]
                                  
+torch.get_float32_matmul_precision()[source][source]
 
                                  Returns the current value of float32 matrix multiplication precision. Refer to
 torch.set_float32_matmul_precision() documentation for more details.
                                  
 
                                  
diff --git a/2.7/generated/torch.get_rng_state.html b/2.7/generated/torch.get_rng_state.html
index b12ecbccca1..0661097b156 100644
--- a/2.7/generated/torch.get_rng_state.html
+++ b/2.7/generated/torch.get_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.get_rng_state
                                  
 
                                  
 
                                  
-torch.get_rng_state()[source][source]
                                  
+torch.get_rng_state()[source][source]
 
                                  Returns the random number generator state as a torch.ByteTensor.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.initial_seed.html b/2.7/generated/torch.initial_seed.html
index 86357985302..d62a43238f1 100644
--- a/2.7/generated/torch.initial_seed.html
+++ b/2.7/generated/torch.initial_seed.html
@@ -596,7 +596,7 @@
 
                                  torch.initial_seed
                                  
 
                                  
 
                                  
-torch.initial_seed()[source][source]
                                  
+torch.initial_seed()[source][source]
 
                                  Returns the initial seed for generating random numbers as a
 Python long.
                                  
 
                                  
diff --git a/2.7/generated/torch.is_deterministic_algorithms_warn_only_enabled.html b/2.7/generated/torch.is_deterministic_algorithms_warn_only_enabled.html
index c045af27bb0..13ec6c901fd 100644
--- a/2.7/generated/torch.is_deterministic_algorithms_warn_only_enabled.html
+++ b/2.7/generated/torch.is_deterministic_algorithms_warn_only_enabled.html
@@ -596,7 +596,7 @@
 
                                  torch.is_deterministic_algorithms_warn_only_enabled
                                  
 
                                  
 
                                  
-torch.is_deterministic_algorithms_warn_only_enabled()[source][source]
                                  
+torch.is_deterministic_algorithms_warn_only_enabled()[source][source]
 
                                  Returns True if the global deterministic flag is set to warn only.
 Refer to torch.use_deterministic_algorithms() documentation for more
 details.
                                  
diff --git a/2.7/generated/torch.is_storage.html b/2.7/generated/torch.is_storage.html
index 76c79634dfb..ad0c85c3381 100644
--- a/2.7/generated/torch.is_storage.html
+++ b/2.7/generated/torch.is_storage.html
@@ -596,7 +596,7 @@
 
                                  torch.is_storage
                                  
 
                                  
 
                                  
-torch.is_storage(obj, /)[source][source]
                                  
+torch.is_storage(obj, /)[source][source]
 
                                  Returns True if obj is a PyTorch storage object.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.is_tensor.html b/2.7/generated/torch.is_tensor.html
index 4919ba4d543..9946c041a44 100644
--- a/2.7/generated/torch.is_tensor.html
+++ b/2.7/generated/torch.is_tensor.html
@@ -596,7 +596,7 @@
 
                                  torch.is_tensor
                                  
 
                                  
 
                                  
-torch.is_tensor(obj, /)[source][source]
                                  
+torch.is_tensor(obj, /)[source][source]
 
                                  Returns True if obj is a PyTorch tensor.
                                  
 
                                  Note that this function is simply doing isinstance(obj, Tensor).
 Using that isinstance check is better for typechecking with mypy,
diff --git a/2.7/generated/torch.is_warn_always_enabled.html b/2.7/generated/torch.is_warn_always_enabled.html
index 27c37215207..58847ce8f59 100644
--- a/2.7/generated/torch.is_warn_always_enabled.html
+++ b/2.7/generated/torch.is_warn_always_enabled.html
@@ -596,7 +596,7 @@
 
                                  torch.is_warn_always_enabled
                                  
 
                                  
 
                                  
-torch.is_warn_always_enabled()[source][source]
                                  
+torch.is_warn_always_enabled()[source][source]
 
                                  Returns True if the global warn_always flag is turned on. Refer to
 torch.set_warn_always() documentation for more details.
                                  
 
                                  
diff --git a/2.7/generated/torch.jit.ScriptModule.html b/2.7/generated/torch.jit.ScriptModule.html
index 242042f8288..22bcf07859c 100644
--- a/2.7/generated/torch.jit.ScriptModule.html
+++ b/2.7/generated/torch.jit.ScriptModule.html
@@ -596,7 +596,7 @@
 
                                  ScriptModule
                                  
 
                                  
 
                                  
-class torch.jit.ScriptModule[source][source]
                                  
+class torch.jit.ScriptModule[source][source]
 
                                  Wrapper for C++ torch::jit::Module with methods, attributes, and parameters.
                                  
 
                                  A wrapper around C++ torch::jit::Module. ScriptModules
 contain methods, attributes, parameters, and
@@ -605,7 +605,7 @@ 
                                  ScriptModule
 
                                  
-add_module(name, module)[source]
                                  
+add_module(name, module)[source]
 
                                  Add a child module to the current module.
                                  
 
                                  The module can be accessed as an attribute using the given name.
                                  
 
                                  
@@ -621,7 +621,7 @@ 
                                  ScriptModule
 
                                  
-apply(fn)[source]
                                  
+apply(fn)[source]
 
                                  Apply fn recursively to every submodule (as returned by .children()) as well as self.
                                  
 
                                  Typical use includes initializing the parameters of a model
 (see also torch.nn.init).
                                  
@@ -663,7 +663,7 @@ 
                                  ScriptModule
 
                                  
-bfloat16()[source]
                                  
+bfloat16()[source]
 
                                  Casts all floating point parameters and buffers to bfloat16 datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -681,7 +681,7 @@ 
                                  ScriptModule
 
                                  
-buffers(recurse=True)[source]
                                  
+buffers(recurse=True)[source]
 
                                  Return an iterator over module buffers.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -707,7 +707,7 @@ 
                                  ScriptModule
 
                                  
-children()[source]
                                  
+children()[source]
 
                                  Return an iterator over immediate children modules.
                                  
 
                                  
 
                                  Yields
                                  
@@ -740,7 +740,7 @@ 
                                  ScriptModule
 
                                  
-compile(*args, **kwargs)[source]
                                  
+compile(*args, **kwargs)[source]
 
                                  Compile this Module’s forward using torch.compile().
                                  
 
                                  This Module’s __call__ method is compiled and all arguments are passed as-is
 to torch.compile().
                                  
@@ -749,7 +749,7 @@ 
                                  ScriptModule
 
                                  
-cpu()[source]
                                  
+cpu()[source]
 
                                  Move all model parameters and buffers to the CPU.
                                  
 
                                  
 
                                  Note
                                  
@@ -767,7 +767,7 @@ 
                                  ScriptModule
 
                                  
-cuda(device=None)[source]
                                  
+cuda(device=None)[source]
 
                                  Move all model parameters and buffers to the GPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -792,7 +792,7 @@ 
                                  ScriptModule
 
                                  
-double()[source]
                                  
+double()[source]
 
                                  Casts all floating point parameters and buffers to double datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -810,7 +810,7 @@ 
                                  ScriptModule
 
                                  
-eval()[source]
                                  
+eval()[source]
 
                                  Set the module in evaluation mode.
                                  
 
                                  This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -831,7 +831,7 @@ 
                                  ScriptModule
 
                                  
-extra_repr()[source]
                                  
+extra_repr()[source]
 
                                  Return the extra representation of the module.
                                  
 
                                  To print customized extra information, you should re-implement
 this method in your own modules. Both single-line and multi-line
@@ -845,7 +845,7 @@ 
                                  ScriptModule
 
                                  
-float()[source]
                                  
+float()[source]
 
                                  Casts all floating point parameters and buffers to float datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -863,7 +863,7 @@ 
                                  ScriptModule
 
                                  
-get_buffer(target)[source]
                                  
+get_buffer(target)[source]
 
                                  Return the buffer given by target if it exists, otherwise throw an error.
                                  
 
                                  See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -890,7 +890,7 @@ 
                                  ScriptModule
 
                                  
-get_extra_state()[source]
                                  
+get_extra_state()[source]
 
                                  Return any extra state to include in the module’s state_dict.
                                  
 
                                  Implement this and a corresponding set_extra_state() for your module
 if you need to store extra state. This function is called when building the
@@ -911,7 +911,7 @@ 
                                  ScriptModule
 
                                  
-get_parameter(target)[source]
                                  
+get_parameter(target)[source]
 
                                  Return the parameter given by target if it exists, otherwise throw an error.
                                  
 
                                  See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -938,7 +938,7 @@ 
                                  ScriptModule
 
                                  
-get_submodule(target)[source]
                                  
+get_submodule(target)[source]
 
                                  Return the submodule given by target if it exists, otherwise throw an error.
                                  
 
                                  For example, let’s say you have an nn.Module A that
 looks like this:
                                  
@@ -994,7 +994,7 @@ 
                                  ScriptModule
 
                                  
-half()[source]
                                  
+half()[source]
 
                                  Casts all floating point parameters and buffers to half datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -1020,7 +1020,7 @@ 
                                  ScriptModule
 
                                  
-ipu(device=None)[source]
                                  
+ipu(device=None)[source]
 
                                  Move all model parameters and buffers to the IPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1045,7 +1045,7 @@ 
                                  ScriptModule
 
                                  
-load_state_dict(state_dict, strict=True, assign=False)[source]
                                  
+load_state_dict(state_dict, strict=True, assign=False)[source]
 
                                  Copy parameters and buffers from state_dict into this module and its descendants.
                                  
 
                                  If strict is True, then
 the keys of state_dict must exactly match the keys returned
@@ -1100,7 +1100,7 @@ 
                                  ScriptModule
 
                                  
-modules()[source]
                                  
+modules()[source]
 
                                  Return an iterator over all modules in the network.
                                  
 
                                  
 
                                  Yields
                                  
@@ -1132,7 +1132,7 @@ 
                                  ScriptModule
 
                                  
-mtia(device=None)[source]
                                  
+mtia(device=None)[source]
 
                                  Move all model parameters and buffers to the MTIA.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1157,7 +1157,7 @@ 
                                  ScriptModule
 
                                  
-named_buffers(prefix='', recurse=True, remove_duplicate=True)[source]
                                  
+named_buffers(prefix='', recurse=True, remove_duplicate=True)[source]
 
                                  Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1186,7 +1186,7 @@ 
                                  ScriptModule
 
                                  
-named_children()[source]
                                  
+named_children()[source]
 
                                  Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
                                  
 
                                  
 
                                  Yields
                                  
@@ -1206,7 +1206,7 @@ 
                                  ScriptModule
 
                                  
-named_modules(memo=None, prefix='', remove_duplicate=True)[source]
                                  
+named_modules(memo=None, prefix='', remove_duplicate=True)[source]
 
                                  Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1243,7 +1243,7 @@ 
                                  ScriptModule
 
                                  
-named_parameters(prefix='', recurse=True, remove_duplicate=True)[source]
                                  
+named_parameters(prefix='', recurse=True, remove_duplicate=True)[source]
 
                                  Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1273,7 +1273,7 @@ 
                                  ScriptModule
 
                                  
-parameters(recurse=True)[source]
                                  
+parameters(recurse=True)[source]
 
                                  Return an iterator over module parameters.
                                  
 
                                  This is typically passed to an optimizer.
                                  
 
                                  
@@ -1300,7 +1300,7 @@ 
                                  ScriptModule
 
                                  
-register_backward_hook(hook)[source]
                                  
+register_backward_hook(hook)[source]
 
                                  Register a backward hook on the module.
                                  
 
                                  This function is deprecated in favor of register_full_backward_hook() and
 the behavior of this function will change in future versions.
                                  
@@ -1317,7 +1317,7 @@ 
                                  ScriptModule
 
                                  
-register_buffer(name, tensor, persistent=True)[source]
                                  
+register_buffer(name, tensor, persistent=True)[source]
 
                                  Add a buffer to the module.
                                  
 
                                  This is typically used to register a buffer that should not to be
 considered a model parameter. For example, BatchNorm’s running_mean
@@ -1349,7 +1349,7 @@ 
                                  ScriptModule
 
                                  
-register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source]
                                  
+register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source]
 
                                  Register a forward hook on the module.
                                  
 
                                  The hook will be called every time after forward() has computed an output.
                                  
 
                                  If with_kwargs is False or not specified, the input contains only
@@ -1400,7 +1400,7 @@ 
                                  ScriptModule
 
                                  
-register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source]
                                  
+register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source]
 
                                  Register a forward pre-hook on the module.
                                  
 
                                  The hook will be called every time before forward() is invoked.
                                  
 
                                  If with_kwargs is false or not specified, the input contains only
@@ -1450,7 +1450,7 @@ 
                                  ScriptModule
 
                                  
-register_full_backward_hook(hook, prepend=False)[source]
                                  
+register_full_backward_hook(hook, prepend=False)[source]
 
                                  Register a backward hook on the module.
                                  
 
                                  The hook will be called every time the gradients with respect to a module
 are computed, i.e. the hook will execute if and only if the gradients with
@@ -1501,7 +1501,7 @@ 
                                  ScriptModule
 
                                  
-register_full_backward_pre_hook(hook, prepend=False)[source]
                                  
+register_full_backward_pre_hook(hook, prepend=False)[source]
 
                                  Register a backward pre-hook on the module.
                                  
 
                                  The hook will be called every time the gradients for the module are computed.
 The hook should have the following signature:
                                  
@@ -1547,7 +1547,7 @@ 
                                  ScriptModule
 
                                  
-register_load_state_dict_post_hook(hook)[source]
                                  
+register_load_state_dict_post_hook(hook)[source]
 
                                  Register a post-hook to be run after module’s load_state_dict() is called.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, incompatible_keys) -> None
                                  
@@ -1577,7 +1577,7 @@ 
                                  ScriptModule
 
                                  
-register_load_state_dict_pre_hook(hook)[source]
                                  
+register_load_state_dict_pre_hook(hook)[source]
 
                                  Register a pre-hook to be run before module’s load_state_dict() is called.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs) -> None  # noqa: B950
                                  
@@ -1593,7 +1593,7 @@ 
                                  ScriptModule
 
                                  
-register_module(name, module)[source]
                                  
+register_module(name, module)[source]
 
                                  Alias for add_module().
                                  
 
                                  
 
                                  
@@ -1601,7 +1601,7 @@ 
                                  ScriptModule
 
                                  
-register_parameter(name, param)[source]
                                  
+register_parameter(name, param)[source]
 
                                  Add a parameter to the module.
                                  
 
                                  The parameter can be accessed as an attribute using given name.
                                  
 
                                  
@@ -1620,7 +1620,7 @@ 
                                  ScriptModule
 
                                  
-register_state_dict_post_hook(hook)[source]
                                  
+register_state_dict_post_hook(hook)[source]
 
                                  Register a post-hook for the state_dict() method.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, state_dict, prefix, local_metadata) -> None
                                  
@@ -1631,7 +1631,7 @@ 
                                  ScriptModule
 
                                  
-register_state_dict_pre_hook(hook)[source]
                                  
+register_state_dict_pre_hook(hook)[source]
 
                                  Register a pre-hook for the state_dict() method.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, prefix, keep_vars) -> None
                                  
@@ -1643,7 +1643,7 @@ 
                                  ScriptModule
 
                                  
-requires_grad_(requires_grad=True)[source]
                                  
+requires_grad_(requires_grad=True)[source]
 
                                  Change if autograd should record operations on parameters in this module.
                                  
 
                                  This method sets the parameters’ requires_grad attributes
 in-place.
                                  
@@ -1667,7 +1667,7 @@ 
                                  ScriptModule
 
                                  
-save(f, **kwargs)[source]
                                  
+save(f, **kwargs)[source]
 
                                  Save with a file-like object.
                                  
 
                                  save(f, _extra_files={})
                                  
 
                                  See torch.jit.save which accepts a file-like object.
@@ -1677,7 +1677,7 @@ 
                                  ScriptModule
 
                                  
-set_extra_state(state)[source]
                                  
+set_extra_state(state)[source]
 
                                  Set extra state contained in the loaded state_dict.
                                  
 
                                  This function is called from load_state_dict() to handle any extra state
 found within the state_dict. Implement this function and a corresponding
@@ -1692,7 +1692,7 @@ 
                                  ScriptModule
 
                                  
-set_submodule(target, module, strict=False)[source]
                                  
+set_submodule(target, module, strict=False)[source]
 
                                  Set the submodule given by target if it exists, otherwise throw an error.
                                  
 
                                  
 
                                  Note
                                  
@@ -1750,7 +1750,7 @@ 
                                  ScriptModule
 
                                  
-share_memory()[source]
                                  
+share_memory()[source]
 
                                  See torch.Tensor.share_memory_().
                                  
 
                                  
 
                                  Return type
                                  
@@ -1761,7 +1761,7 @@ 
                                  ScriptModule
 
                                  
-state_dict(*args, destination=None, prefix='', keep_vars=False)[source]
                                  
+state_dict(*args, destination=None, prefix='', keep_vars=False)[source]
 
                                  Return a dictionary containing references to the whole state of the module.
                                  
 
                                  Both parameters and persistent buffers (e.g. running averages) are
 included. Keys are corresponding parameter and buffer names.
@@ -1814,27 +1814,27 @@ 
                                  ScriptModule
 
                                  
-to(*args, **kwargs)[source]
                                  
+to(*args, **kwargs)[source]
 
                                  Move and/or cast the parameters and buffers.
                                  
 
                                  This can be called as
                                  
 
                                  
 
                                  
-to(device=None, dtype=None, non_blocking=False)[source]
                                  
+to(device=None, dtype=None, non_blocking=False)[source]
 
                                  
 
 
                                  
 
                                  
-to(dtype, non_blocking=False)[source]
                                  
+to(dtype, non_blocking=False)[source]
 
                                  
 
 
                                  
 
                                  
-to(tensor, non_blocking=False)[source]
                                  
+to(tensor, non_blocking=False)[source]
 
                                  
 
 
                                  
 
                                  
-to(memory_format=torch.channels_last)[source]
                                  
+to(memory_format=torch.channels_last)[source]
 
                                  
 
 
                                  Its signature is similar to torch.Tensor.to(), but only accepts
@@ -1913,7 +1913,7 @@ 
                                  ScriptModule
 
                                  
-to_empty(*, device, recurse=True)[source]
                                  
+to_empty(*, device, recurse=True)[source]
 
                                  Move the parameters and buffers to the specified device without copying storage.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1935,7 +1935,7 @@ 
                                  ScriptModule
 
                                  
-train(mode=True)[source]
                                  
+train(mode=True)[source]
 
                                  Set the module in training mode.
                                  
 
                                  This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -1957,7 +1957,7 @@ 
                                  ScriptModule
 
                                  
-type(dst_type)[source]
                                  
+type(dst_type)[source]
 
                                  Casts all parameters and buffers to dst_type.
                                  
 
                                  
 
                                  Note
                                  
@@ -1978,7 +1978,7 @@ 
                                  ScriptModule
 
                                  
-xpu(device=None)[source]
                                  
+xpu(device=None)[source]
 
                                  Move all model parameters and buffers to the XPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing optimizer if the module will
@@ -2003,7 +2003,7 @@ 
                                  ScriptModule
 
                                  
-zero_grad(set_to_none=True)[source]
                                  
+zero_grad(set_to_none=True)[source]
 
                                  Reset gradients of all model parameters.
                                  
 
                                  See similar function under torch.optim.Optimizer for more context.
                                  
 
                                  
diff --git a/2.7/generated/torch.jit.annotate.html b/2.7/generated/torch.jit.annotate.html
index 78a89429a3a..0c787c8943a 100644
--- a/2.7/generated/torch.jit.annotate.html
+++ b/2.7/generated/torch.jit.annotate.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.annotate
                                  
 
                                  
 
                                  
-torch.jit.annotate(the_type, the_value)[source][source]
                                  
+torch.jit.annotate(the_type, the_value)[source][source]
 
                                  Use to give type of the_value in TorchScript compiler.
                                  
 
                                  This method is a pass-through function that returns the_value, used to hint TorchScript
 compiler the type of the_value. It is a no-op when running outside of TorchScript.
                                  
diff --git a/2.7/generated/torch.jit.enable_onednn_fusion.html b/2.7/generated/torch.jit.enable_onednn_fusion.html
index ba5faf138cb..895dde11830 100644
--- a/2.7/generated/torch.jit.enable_onednn_fusion.html
+++ b/2.7/generated/torch.jit.enable_onednn_fusion.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.enable_onednn_fusion
                                  
 
                                  
 
                                  
-torch.jit.enable_onednn_fusion(enabled)[source][source]
                                  
+torch.jit.enable_onednn_fusion(enabled)[source][source]
 
                                  Enable or disables onednn JIT fusion based on the parameter enabled.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.jit.fork.html b/2.7/generated/torch.jit.fork.html
index 60879189436..6e477c36df7 100644
--- a/2.7/generated/torch.jit.fork.html
+++ b/2.7/generated/torch.jit.fork.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.fork
                                  
 
                                  
 
                                  
-torch.jit.fork(func, *args, **kwargs)[source][source]
                                  
+torch.jit.fork(func, *args, **kwargs)[source][source]
 
                                  Create an asynchronous task executing func and a reference to the value of the result of this execution.
                                  
 
                                  fork will return immediately, so the return value of func may not have been computed yet. To force completion
 of the task and access the return value invoke torch.jit.wait on the Future. fork invoked
diff --git a/2.7/generated/torch.jit.freeze.html b/2.7/generated/torch.jit.freeze.html
index dab6cbac214..f2185266fe0 100644
--- a/2.7/generated/torch.jit.freeze.html
+++ b/2.7/generated/torch.jit.freeze.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.freeze
                                  
 
                                  
 
                                  
-torch.jit.freeze(mod, preserved_attrs=None, optimize_numerics=True)[source][source]
                                  
+torch.jit.freeze(mod, preserved_attrs=None, optimize_numerics=True)[source][source]
 
                                  Freeze ScriptModule, inline submodules, and attributes as constants.
                                  
 
                                  Freezing a ScriptModule will clone it and attempt to inline the cloned
 module’s submodules, parameters, and attributes as constants in the TorchScript IR Graph.
diff --git a/2.7/generated/torch.jit.ignore.html b/2.7/generated/torch.jit.ignore.html
index f96c0bdf011..430c1e27800 100644
--- a/2.7/generated/torch.jit.ignore.html
+++ b/2.7/generated/torch.jit.ignore.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.ignore
                                  
 
                                  
 
                                  
-torch.jit.ignore(drop=False, **kwargs)[source][source]
                                  
+torch.jit.ignore(drop=False, **kwargs)[source][source]
 
                                  This decorator indicates to the compiler that a function or method should
 be ignored and left as a Python function. This allows you to leave code in
 your model that is not yet TorchScript compatible. If called from TorchScript,
diff --git a/2.7/generated/torch.jit.interface.html b/2.7/generated/torch.jit.interface.html
index aba02e26ca7..981a178fbb1 100644
--- a/2.7/generated/torch.jit.interface.html
+++ b/2.7/generated/torch.jit.interface.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.interface
                                  
 
                                  
 
                                  
-torch.jit.interface(obj)[source][source]
                                  
+torch.jit.interface(obj)[source][source]
 
                                  Decorate to annotate classes or modules of different types.
                                  
 
                                  This decorator can be used to define an interface that can be used to annotate
 classes or modules of different types. This can be used for to annotate a submodule
diff --git a/2.7/generated/torch.jit.isinstance.html b/2.7/generated/torch.jit.isinstance.html
index a7a3ec68f3b..9230cfb9246 100644
--- a/2.7/generated/torch.jit.isinstance.html
+++ b/2.7/generated/torch.jit.isinstance.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.isinstance
                                  
 
                                  
 
                                  
-torch.jit.isinstance(obj, target_type)[source][source]
                                  
+torch.jit.isinstance(obj, target_type)[source][source]
 
                                  Provide container type refinement in TorchScript.
                                  
 
                                  It can refine parameterized containers of the List, Dict, Tuple, and Optional types. E.g. List[str],
 Dict[str, List[torch.Tensor]], Optional[Tuple[int,str,int]]. It can also
diff --git a/2.7/generated/torch.jit.load.html b/2.7/generated/torch.jit.load.html
index e8d6429d64d..7dfbcfe3bd1 100644
--- a/2.7/generated/torch.jit.load.html
+++ b/2.7/generated/torch.jit.load.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.load
                                  
 
                                  
 
                                  
-torch.jit.load(f, map_location=None, _extra_files=None, _restore_shapes=False)[source][source]
                                  
+torch.jit.load(f, map_location=None, _extra_files=None, _restore_shapes=False)[source][source]
 
                                  Load a ScriptModule or ScriptFunction previously saved with torch.jit.save.
                                  
 
                                  All previously saved modules, no matter their device, are first loaded onto CPU,
 and then are moved to the devices they were saved from. If this fails (e.g.
diff --git a/2.7/generated/torch.jit.onednn_fusion_enabled.html b/2.7/generated/torch.jit.onednn_fusion_enabled.html
index 3e64c159902..348339159f3 100644
--- a/2.7/generated/torch.jit.onednn_fusion_enabled.html
+++ b/2.7/generated/torch.jit.onednn_fusion_enabled.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.onednn_fusion_enabled
                                  
 
                                  
 
                                  
-torch.jit.onednn_fusion_enabled()[source][source]
                                  
+torch.jit.onednn_fusion_enabled()[source][source]
 
                                  Return whether onednn JIT fusion is enabled.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.jit.optimize_for_inference.html b/2.7/generated/torch.jit.optimize_for_inference.html
index 07a786fc900..9b19ad4a2b5 100644
--- a/2.7/generated/torch.jit.optimize_for_inference.html
+++ b/2.7/generated/torch.jit.optimize_for_inference.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.optimize_for_inference
                                  
 
                                  
 
                                  
-torch.jit.optimize_for_inference(mod, other_methods=None)[source][source]
                                  
+torch.jit.optimize_for_inference(mod, other_methods=None)[source][source]
 
                                  Perform a set of optimization passes to optimize a model for the purposes of inference.
                                  
 
                                  If the model is not already frozen, optimize_for_inference
 will invoke torch.jit.freeze automatically.
                                  
diff --git a/2.7/generated/torch.jit.save.html b/2.7/generated/torch.jit.save.html
index c12ae4c9fdc..94711a18458 100644
--- a/2.7/generated/torch.jit.save.html
+++ b/2.7/generated/torch.jit.save.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.save
                                  
 
                                  
 
                                  
-torch.jit.save(m, f, _extra_files=None)[source][source]
                                  
+torch.jit.save(m, f, _extra_files=None)[source][source]
 
                                  Save an offline version of this module for use in a separate process.
                                  
 
                                  The saved module serializes all of the methods, submodules, parameters, and
 attributes of this module. It can be loaded into the C++ API using
diff --git a/2.7/generated/torch.jit.script.html b/2.7/generated/torch.jit.script.html
index c04e6f318f8..c4c85f03ac4 100644
--- a/2.7/generated/torch.jit.script.html
+++ b/2.7/generated/torch.jit.script.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.script
                                  
 
                                  
 
                                  
-torch.jit.script(obj, optimize=None, _frames_up=0, _rcb=None, example_inputs=None)[source][source]
                                  
+torch.jit.script(obj, optimize=None, _frames_up=0, _rcb=None, example_inputs=None)[source][source]
 
                                  Script the function.
                                  
 
                                  Scripting a function or nn.Module will inspect the source code, compile
 it as TorchScript code using the TorchScript compiler, and return a ScriptModule or
diff --git a/2.7/generated/torch.jit.script_if_tracing.html b/2.7/generated/torch.jit.script_if_tracing.html
index 302ed33aa5b..446f1589d5a 100644
--- a/2.7/generated/torch.jit.script_if_tracing.html
+++ b/2.7/generated/torch.jit.script_if_tracing.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.script_if_tracing
                                  
 
                                  
 
                                  
-torch.jit.script_if_tracing(fn)[source][source]
                                  
+torch.jit.script_if_tracing(fn)[source][source]
 
                                  Compiles fn when it is first called during tracing.
                                  
 
                                  torch.jit.script has a non-negligible start up time when it is first called due to
 lazy-initializations of many compiler builtins. Therefore you should not use
diff --git a/2.7/generated/torch.jit.set_fusion_strategy.html b/2.7/generated/torch.jit.set_fusion_strategy.html
index 7ead65e307a..05110ff653a 100644
--- a/2.7/generated/torch.jit.set_fusion_strategy.html
+++ b/2.7/generated/torch.jit.set_fusion_strategy.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.set_fusion_strategy
                                  
 
                                  
 
                                  
-torch.jit.set_fusion_strategy(strategy)[source][source]
                                  
+torch.jit.set_fusion_strategy(strategy)[source][source]
 
                                  Set the type and number of specializations that can occur during fusion.
                                  
 
                                  Usage: provide a list of pairs (type, depth) where type is one of “STATIC” or “DYNAMIC”
 and depth is an integer.
                                  
diff --git a/2.7/generated/torch.jit.strict_fusion.html b/2.7/generated/torch.jit.strict_fusion.html
index 42391948c87..01b7266c1c2 100644
--- a/2.7/generated/torch.jit.strict_fusion.html
+++ b/2.7/generated/torch.jit.strict_fusion.html
@@ -596,7 +596,7 @@
 
                                  strict_fusion
                                  
 
                                  
 
                                  
-class torch.jit.strict_fusion[source][source]
                                  
+class torch.jit.strict_fusion[source][source]
 
                                  Give errors if not all nodes have been fused in inference, or symbolically differentiated in training.
                                  
 
                                  Example:
 Forcing fusion of additions.
                                  
diff --git a/2.7/generated/torch.jit.trace.html b/2.7/generated/torch.jit.trace.html
index 5d011c1a9bf..ff11bc61027 100644
--- a/2.7/generated/torch.jit.trace.html
+++ b/2.7/generated/torch.jit.trace.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.trace
                                  
 
                                  
 
                                  
-torch.jit.trace(func, example_inputs=None, optimize=None, check_trace=True, check_inputs=None, check_tolerance=1e-05, strict=True, _force_outplace=False, _module_class=None, _compilation_unit=<torch.jit.CompilationUnit object>, example_kwarg_inputs=None, _store_inputs=True)[source][source]
                                  
+torch.jit.trace(func, example_inputs=None, optimize=None, check_trace=True, check_inputs=None, check_tolerance=1e-05, strict=True, _force_outplace=False, _module_class=None, _compilation_unit=<torch.jit.CompilationUnit object>, example_kwarg_inputs=None, _store_inputs=True)[source][source]
 
                                  Trace a function and return an executable  or ScriptFunction that will be optimized using just-in-time compilation.
                                  
 
                                  Tracing is ideal for code that operates only on
 Tensor\s and lists, dictionaries, and
diff --git a/2.7/generated/torch.jit.trace_module.html b/2.7/generated/torch.jit.trace_module.html
index f9a2c73cabc..77e13a4e1dc 100644
--- a/2.7/generated/torch.jit.trace_module.html
+++ b/2.7/generated/torch.jit.trace_module.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.trace_module
                                  
 
                                  
 
                                  
-torch.jit.trace_module(mod, inputs, optimize=None, check_trace=True, check_inputs=None, check_tolerance=1e-05, strict=True, _force_outplace=False, _module_class=None, _compilation_unit=<torch.jit.CompilationUnit object>, example_inputs_is_kwarg=False, _store_inputs=True)[source][source]
                                  
+torch.jit.trace_module(mod, inputs, optimize=None, check_trace=True, check_inputs=None, check_tolerance=1e-05, strict=True, _force_outplace=False, _module_class=None, _compilation_unit=<torch.jit.CompilationUnit object>, example_inputs_is_kwarg=False, _store_inputs=True)[source][source]
 
                                  Trace a module and return an executable ScriptModule that will be optimized using just-in-time compilation.
                                  
 
                                  When a module is passed to torch.jit.trace, only
 the forward method is run and traced. With trace_module, you can specify a dictionary of
diff --git a/2.7/generated/torch.jit.unused.html b/2.7/generated/torch.jit.unused.html
index 8ba11b71d3a..10618a6935e 100644
--- a/2.7/generated/torch.jit.unused.html
+++ b/2.7/generated/torch.jit.unused.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.unused
                                  
 
                                  
 
                                  
-torch.jit.unused(fn)[source][source]
                                  
+torch.jit.unused(fn)[source][source]
 
                                  This decorator indicates to the compiler that a function or method should
 be ignored and replaced with the raising of an exception. This allows you
 to leave code in your model that is not yet TorchScript compatible and still
diff --git a/2.7/generated/torch.jit.wait.html b/2.7/generated/torch.jit.wait.html
index c4ee19fe496..8ccd6bd7fe3 100644
--- a/2.7/generated/torch.jit.wait.html
+++ b/2.7/generated/torch.jit.wait.html
@@ -596,7 +596,7 @@
 
                                  torch.jit.wait
                                  
 
                                  
 
                                  
-torch.jit.wait(future)[source][source]
                                  
+torch.jit.wait(future)[source][source]
 
                                  Force completion of a torch.jit.Future[T] asynchronous task, returning the result of the task.
                                  
 
                                  See fork() for docs and examples.
 :param future: an asynchronous task reference, created through torch.jit.fork
diff --git a/2.7/generated/torch.load.html b/2.7/generated/torch.load.html
index 1f8f3674263..0aa0df4d91b 100644
--- a/2.7/generated/torch.load.html
+++ b/2.7/generated/torch.load.html
@@ -596,7 +596,7 @@
 
                                  torch.load
                                  
 
                                  
 
                                  
-torch.load(f, map_location=None, pickle_module=pickle, *, weights_only=True, mmap=None, **pickle_load_args)[source][source]
                                  
+torch.load(f, map_location=None, pickle_module=pickle, *, weights_only=True, mmap=None, **pickle_load_args)[source][source]
 
                                  Loads an object saved with torch.save() from a file.
                                  
 
                                  torch.load() uses Python’s unpickling facilities but treats storages,
 which underlie tensors, specially. They are first deserialized on the
diff --git a/2.7/generated/torch.lobpcg.html b/2.7/generated/torch.lobpcg.html
index 1a277133f9d..cb4db3b0f3e 100644
--- a/2.7/generated/torch.lobpcg.html
+++ b/2.7/generated/torch.lobpcg.html
@@ -596,7 +596,7 @@
 
                                  torch.lobpcg
                                  
 
                                  
 
                                  
-torch.lobpcg(A, k=None, B=None, X=None, n=None, iK=None, niter=None, tol=None, largest=None, method=None, tracker=None, ortho_iparams=None, ortho_fparams=None, ortho_bparams=None)[source][source]
                                  
+torch.lobpcg(A, k=None, B=None, X=None, n=None, iK=None, niter=None, tol=None, largest=None, method=None, tracker=None, ortho_iparams=None, ortho_fparams=None, ortho_bparams=None)[source][source]
 
                                  Find the k largest (or smallest) eigenvalues and the corresponding
 eigenvectors of a symmetric positive definite generalized
 eigenvalue problem using matrix-free LOBPCG methods.
                                  
diff --git a/2.7/generated/torch.lu.html b/2.7/generated/torch.lu.html
index 5c0282f667a..e604c3b4c53 100644
--- a/2.7/generated/torch.lu.html
+++ b/2.7/generated/torch.lu.html
@@ -596,7 +596,7 @@
 
                                  torch.lu
                                  
 
                                  
 
                                  
-torch.lu(*args, **kwargs)[source]
                                  
+torch.lu(*args, **kwargs)[source]
 
                                  Computes the LU factorization of a matrix or batches of matrices
 A. Returns a tuple containing the LU factorization and
 pivots of A.  Pivoting is done if pivot is set to
diff --git a/2.7/generated/torch.manual_seed.html b/2.7/generated/torch.manual_seed.html
index 682d4cbe650..8c6b50c2613 100644
--- a/2.7/generated/torch.manual_seed.html
+++ b/2.7/generated/torch.manual_seed.html
@@ -596,7 +596,7 @@
 
                                  torch.manual_seed
                                  
 
                                  
 
                                  
-torch.manual_seed(seed)[source][source]
                                  
+torch.manual_seed(seed)[source][source]
 
                                  Sets the seed for generating random numbers on all devices. Returns a
 torch.Generator object.
                                  
 
                                  
diff --git a/2.7/generated/torch.meshgrid.html b/2.7/generated/torch.meshgrid.html
index e3a921cee55..c2f49a9eacc 100644
--- a/2.7/generated/torch.meshgrid.html
+++ b/2.7/generated/torch.meshgrid.html
@@ -596,7 +596,7 @@
 
                                  torch.meshgrid
                                  
 
                                  
 
                                  
-torch.meshgrid(*tensors, indexing=None)[source][source]
                                  
+torch.meshgrid(*tensors, indexing=None)[source][source]
 
                                  Creates grids of coordinates specified by the 1D inputs in attr:tensors.
                                  
 
                                  This is helpful when you want to visualize data over some
 range of inputs. See below for a plotting example.
                                  
diff --git a/2.7/generated/torch.mps.compile_shader.html b/2.7/generated/torch.mps.compile_shader.html
index bbb8ce1617a..c887e917182 100644
--- a/2.7/generated/torch.mps.compile_shader.html
+++ b/2.7/generated/torch.mps.compile_shader.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.compile_shader
                                  
 
                                  
 
                                  
-torch.mps.compile_shader(source)[source][source]
                                  
+torch.mps.compile_shader(source)[source][source]
 
                                  Compiles compute shader from source and allows one to invoke kernels
 defined there from the comfort of Python runtime
 Example:
                                  
diff --git a/2.7/generated/torch.mps.current_allocated_memory.html b/2.7/generated/torch.mps.current_allocated_memory.html
index b0a55406487..b82d83ded1e 100644
--- a/2.7/generated/torch.mps.current_allocated_memory.html
+++ b/2.7/generated/torch.mps.current_allocated_memory.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.current_allocated_memory
                                  
 
                                  
 
                                  
-torch.mps.current_allocated_memory()[source][source]
                                  
+torch.mps.current_allocated_memory()[source][source]
 
                                  Returns the current GPU memory occupied by tensors in bytes.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.mps.device_count.html b/2.7/generated/torch.mps.device_count.html
index f5e60b021c6..a982c061687 100644
--- a/2.7/generated/torch.mps.device_count.html
+++ b/2.7/generated/torch.mps.device_count.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.device_count
                                  
 
                                  
 
                                  
-torch.mps.device_count()[source][source]
                                  
+torch.mps.device_count()[source][source]
 
                                  Returns the number of available MPS devices.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mps.driver_allocated_memory.html b/2.7/generated/torch.mps.driver_allocated_memory.html
index c50450b8809..acf79d39618 100644
--- a/2.7/generated/torch.mps.driver_allocated_memory.html
+++ b/2.7/generated/torch.mps.driver_allocated_memory.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.driver_allocated_memory
                                  
 
                                  
 
                                  
-torch.mps.driver_allocated_memory()[source][source]
                                  
+torch.mps.driver_allocated_memory()[source][source]
 
                                  Returns total GPU memory allocated by Metal driver for the process in bytes.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.mps.empty_cache.html b/2.7/generated/torch.mps.empty_cache.html
index f6eea376d6d..0c743c2da25 100644
--- a/2.7/generated/torch.mps.empty_cache.html
+++ b/2.7/generated/torch.mps.empty_cache.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.empty_cache
                                  
 
                                  
 
                                  
-torch.mps.empty_cache()[source][source]
                                  
+torch.mps.empty_cache()[source][source]
 
                                  Releases all unoccupied cached memory currently held by the caching
 allocator so that those can be used in other GPU applications.
                                  
 
                                  
diff --git a/2.7/generated/torch.mps.event.Event.html b/2.7/generated/torch.mps.event.Event.html
index 83143f3630a..20615ad5bce 100644
--- a/2.7/generated/torch.mps.event.Event.html
+++ b/2.7/generated/torch.mps.event.Event.html
@@ -596,7 +596,7 @@
 
                                  Event
                                  
 
                                  
 
                                  
-class torch.mps.event.Event(enable_timing=False)[source][source]
                                  
+class torch.mps.event.Event(enable_timing=False)[source][source]
 
                                  Wrapper around an MPS event.
                                  
 
                                  MPS events are synchronization markers that can be used to monitor the
 device’s progress, to accurately measure timing, and to synchronize MPS streams.
                                  
@@ -608,7 +608,7 @@ 
                                  Event
 
                                  
 
                                  
 
                                  
-elapsed_time(end_event)[source][source]
                                  
+elapsed_time(end_event)[source][source]
 
                                  Returns the time elapsed in milliseconds after the event was
 recorded and before the end_event was recorded.
                                  
 
                                  
@@ -620,7 +620,7 @@ 
                                  Event
 
 
                                  
 
                                  
-query()[source][source]
                                  
+query()[source][source]
 
                                  Returns True if all work currently captured by event has completed.
                                  
 
                                  
 
                                  Return type
                                  
@@ -631,7 +631,7 @@ 
                                  Event
 
 
                                  
 
                                  
-record()[source][source]
                                  
+record()[source][source]
 
                                  Records the event in the default stream.
                                  
 
                                  
 
                                  
@@ -639,7 +639,7 @@ 
                                  Event
 
 
                                  
 
                                  
-synchronize()[source][source]
                                  
+synchronize()[source][source]
 
                                  Waits until the completion of all work currently captured in this event.
 This prevents the CPU thread from proceeding until the event completes.
                                  
 
                                  
@@ -648,7 +648,7 @@ 
                                  Event
 
 
                                  
 
                                  
-wait()[source][source]
                                  
+wait()[source][source]
 
                                  Makes all future work submitted to the default stream wait for this event.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.mps.get_rng_state.html b/2.7/generated/torch.mps.get_rng_state.html
index c3f09f74be5..4ed651822bd 100644
--- a/2.7/generated/torch.mps.get_rng_state.html
+++ b/2.7/generated/torch.mps.get_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.get_rng_state
                                  
 
                                  
 
                                  
-torch.mps.get_rng_state(device='mps')[source][source]
                                  
+torch.mps.get_rng_state(device='mps')[source][source]
 
                                  Returns the random number generator state as a ByteTensor.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mps.manual_seed.html b/2.7/generated/torch.mps.manual_seed.html
index 2a36076b6dc..a0b6bf00b91 100644
--- a/2.7/generated/torch.mps.manual_seed.html
+++ b/2.7/generated/torch.mps.manual_seed.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.manual_seed
                                  
 
                                  
 
                                  
-torch.mps.manual_seed(seed)[source][source]
                                  
+torch.mps.manual_seed(seed)[source][source]
 
                                  Sets the seed for generating random numbers.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mps.profiler.is_capturing_metal.html b/2.7/generated/torch.mps.profiler.is_capturing_metal.html
index dcfa918825b..e6f7ed82ab3 100644
--- a/2.7/generated/torch.mps.profiler.is_capturing_metal.html
+++ b/2.7/generated/torch.mps.profiler.is_capturing_metal.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.is_capturing_metal
                                  
 
                                  
 
                                  
-torch.mps.profiler.is_capturing_metal()[source][source]
                                  
+torch.mps.profiler.is_capturing_metal()[source][source]
 
                                  Cheks if metal capture is in progress
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mps.profiler.is_metal_capture_enabled.html b/2.7/generated/torch.mps.profiler.is_metal_capture_enabled.html
index a18d1755f63..2fe31f2ae31 100644
--- a/2.7/generated/torch.mps.profiler.is_metal_capture_enabled.html
+++ b/2.7/generated/torch.mps.profiler.is_metal_capture_enabled.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.is_metal_capture_enabled
                                  
 
                                  
 
                                  
-torch.mps.profiler.is_metal_capture_enabled()[source][source]
                                  
+torch.mps.profiler.is_metal_capture_enabled()[source][source]
 
                                  Checks if metal_capture context manager is usable
 To enable metal capture, set MTL_CAPTURE_ENABLED envvar
                                  
 
                                  
diff --git a/2.7/generated/torch.mps.profiler.metal_capture.html b/2.7/generated/torch.mps.profiler.metal_capture.html
index 7fe5d7751e9..4f7e3a289a4 100644
--- a/2.7/generated/torch.mps.profiler.metal_capture.html
+++ b/2.7/generated/torch.mps.profiler.metal_capture.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.metal_capture
                                  
 
                                  
 
                                  
-torch.mps.profiler.metal_capture(fname)[source][source]
                                  
+torch.mps.profiler.metal_capture(fname)[source][source]
 
                                  Conext manager that enables capturing of Metal calls into gputrace
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.mps.profiler.profile.html b/2.7/generated/torch.mps.profiler.profile.html
index 6f479854701..45aa13209a5 100644
--- a/2.7/generated/torch.mps.profiler.profile.html
+++ b/2.7/generated/torch.mps.profiler.profile.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.profile
                                  
 
                                  
 
                                  
-torch.mps.profiler.profile(mode='interval', wait_until_completed=False)[source][source]
                                  
+torch.mps.profiler.profile(mode='interval', wait_until_completed=False)[source][source]
 
                                  Context Manager to enabling generating OS Signpost tracing from MPS backend.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mps.profiler.start.html b/2.7/generated/torch.mps.profiler.start.html
index 0efa2e0c1a5..d55f1646c8f 100644
--- a/2.7/generated/torch.mps.profiler.start.html
+++ b/2.7/generated/torch.mps.profiler.start.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.start
                                  
 
                                  
 
                                  
-torch.mps.profiler.start(mode='interval', wait_until_completed=False)[source][source]
                                  
+torch.mps.profiler.start(mode='interval', wait_until_completed=False)[source][source]
 
                                  Start OS Signpost tracing from MPS backend.
                                  
 
                                  The generated OS Signposts could be recorded and viewed in
 XCode Instruments Logging tool.
                                  
diff --git a/2.7/generated/torch.mps.profiler.stop.html b/2.7/generated/torch.mps.profiler.stop.html
index 32125996236..1146438e27d 100644
--- a/2.7/generated/torch.mps.profiler.stop.html
+++ b/2.7/generated/torch.mps.profiler.stop.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.profiler.stop
                                  
 
                                  
 
                                  
-torch.mps.profiler.stop()[source][source]
                                  
+torch.mps.profiler.stop()[source][source]
 
                                  Stops generating OS Signpost tracing from MPS backend.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.mps.recommended_max_memory.html b/2.7/generated/torch.mps.recommended_max_memory.html
index 4e40c83562e..e8e884638a4 100644
--- a/2.7/generated/torch.mps.recommended_max_memory.html
+++ b/2.7/generated/torch.mps.recommended_max_memory.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.recommended_max_memory
                                  
 
                                  
 
                                  
-torch.mps.recommended_max_memory()[source][source]
                                  
+torch.mps.recommended_max_memory()[source][source]
 
                                  Returns recommended max Working set size for GPU memory in bytes.
                                  
 
                                  
 
                                  Note
                                  
diff --git a/2.7/generated/torch.mps.seed.html b/2.7/generated/torch.mps.seed.html
index 5ccf485bfbe..68975f146bf 100644
--- a/2.7/generated/torch.mps.seed.html
+++ b/2.7/generated/torch.mps.seed.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.seed
                                  
 
                                  
 
                                  
-torch.mps.seed()[source][source]
                                  
+torch.mps.seed()[source][source]
 
                                  Sets the seed for generating random numbers to a random number.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.mps.set_per_process_memory_fraction.html b/2.7/generated/torch.mps.set_per_process_memory_fraction.html
index c74ea78021a..3cc729a17e2 100644
--- a/2.7/generated/torch.mps.set_per_process_memory_fraction.html
+++ b/2.7/generated/torch.mps.set_per_process_memory_fraction.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.set_per_process_memory_fraction
                                  
 
                                  
 
                                  
-torch.mps.set_per_process_memory_fraction(fraction)[source][source]
                                  
+torch.mps.set_per_process_memory_fraction(fraction)[source][source]
 
                                  Set memory fraction for limiting process’s memory allocation on MPS device.
 The allowed value equals the fraction multiplied by recommended maximum device memory
 (obtained from Metal API device.recommendedMaxWorkingSetSize).
diff --git a/2.7/generated/torch.mps.set_rng_state.html b/2.7/generated/torch.mps.set_rng_state.html
index c5e2728b725..6290b2c0a25 100644
--- a/2.7/generated/torch.mps.set_rng_state.html
+++ b/2.7/generated/torch.mps.set_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.set_rng_state
                                  
 
                                  
 
                                  
-torch.mps.set_rng_state(new_state, device='mps')[source][source]
                                  
+torch.mps.set_rng_state(new_state, device='mps')[source][source]
 
                                  Sets the random number generator state.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mps.synchronize.html b/2.7/generated/torch.mps.synchronize.html
index 309dea7284c..5b549f84f14 100644
--- a/2.7/generated/torch.mps.synchronize.html
+++ b/2.7/generated/torch.mps.synchronize.html
@@ -596,7 +596,7 @@
 
                                  torch.mps.synchronize
                                  
 
                                  
 
                                  
-torch.mps.synchronize()[source][source]
                                  
+torch.mps.synchronize()[source][source]
 
                                  Waits for all kernels in all streams on a MPS device to complete.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.mtia.DeferredMtiaCallError.html b/2.7/generated/torch.mtia.DeferredMtiaCallError.html
index 847f4f7d024..46d6bf7addf 100644
--- a/2.7/generated/torch.mtia.DeferredMtiaCallError.html
+++ b/2.7/generated/torch.mtia.DeferredMtiaCallError.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.DeferredMtiaCallError
                                  
 
                                  
 
                                  
-exception torch.mtia.DeferredMtiaCallError[source][source]
                                  
+exception torch.mtia.DeferredMtiaCallError[source][source]
 
                                  
 
 
                                  
diff --git a/2.7/generated/torch.mtia.StreamContext.html b/2.7/generated/torch.mtia.StreamContext.html
index 8c7cc449349..41367147532 100644
--- a/2.7/generated/torch.mtia.StreamContext.html
+++ b/2.7/generated/torch.mtia.StreamContext.html
@@ -596,7 +596,7 @@
 
                                  StreamContext
                                  
 
                                  
 
                                  
-class torch.mtia.StreamContext(stream)[source][source]
                                  
+class torch.mtia.StreamContext(stream)[source][source]
 
                                  Context-manager that selects a given stream.
                                  
 
                                  All MTIA kernels queued within its context will be enqueued on a selected
 stream.
                                  
diff --git a/2.7/generated/torch.mtia.current_device.html b/2.7/generated/torch.mtia.current_device.html
index d44a6a2fd09..9c857df9a11 100644
--- a/2.7/generated/torch.mtia.current_device.html
+++ b/2.7/generated/torch.mtia.current_device.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.current_device
                                  
 
                                  
 
                                  
-torch.mtia.current_device()[source][source]
                                  
+torch.mtia.current_device()[source][source]
 
                                  Return the index of a currently selected device.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mtia.current_stream.html b/2.7/generated/torch.mtia.current_stream.html
index a68eb06ae18..b998d0cee13 100644
--- a/2.7/generated/torch.mtia.current_stream.html
+++ b/2.7/generated/torch.mtia.current_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.current_stream
                                  
 
                                  
 
                                  
-torch.mtia.current_stream(device=None)[source][source]
                                  
+torch.mtia.current_stream(device=None)[source][source]
 
                                  Return the currently selected Stream for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.default_stream.html b/2.7/generated/torch.mtia.default_stream.html
index 43de2a62118..155a7aa95df 100644
--- a/2.7/generated/torch.mtia.default_stream.html
+++ b/2.7/generated/torch.mtia.default_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.default_stream
                                  
 
                                  
 
                                  
-torch.mtia.default_stream(device=None)[source][source]
                                  
+torch.mtia.default_stream(device=None)[source][source]
 
                                  Return the default Stream for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.device.html b/2.7/generated/torch.mtia.device.html
index 81310614cb9..8b730ec9ed9 100644
--- a/2.7/generated/torch.mtia.device.html
+++ b/2.7/generated/torch.mtia.device.html
@@ -596,7 +596,7 @@
 
                                  device
                                  
 
                                  
 
                                  
-class torch.mtia.device(device)[source][source]
                                  
+class torch.mtia.device(device)[source][source]
 
                                  Context-manager that changes the selected device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.device_count.html b/2.7/generated/torch.mtia.device_count.html
index 3000e16a751..2604b228125 100644
--- a/2.7/generated/torch.mtia.device_count.html
+++ b/2.7/generated/torch.mtia.device_count.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.device_count
                                  
 
                                  
 
                                  
-torch.mtia.device_count()[source][source]
                                  
+torch.mtia.device_count()[source][source]
 
                                  Return the number of MTIA devices available.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mtia.empty_cache.html b/2.7/generated/torch.mtia.empty_cache.html
index 3808c42680e..a6b24d0f08d 100644
--- a/2.7/generated/torch.mtia.empty_cache.html
+++ b/2.7/generated/torch.mtia.empty_cache.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.empty_cache
                                  
 
                                  
 
                                  
-torch.mtia.empty_cache()[source][source]
                                  
+torch.mtia.empty_cache()[source][source]
 
                                  Empty the MTIA device cache.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.mtia.get_device_capability.html b/2.7/generated/torch.mtia.get_device_capability.html
index 43207601d3c..fb5e5cadc41 100644
--- a/2.7/generated/torch.mtia.get_device_capability.html
+++ b/2.7/generated/torch.mtia.get_device_capability.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.get_device_capability
                                  
 
                                  
 
                                  
-torch.mtia.get_device_capability(device=None)[source][source]
                                  
+torch.mtia.get_device_capability(device=None)[source][source]
 
                                  Return capability of a given device as a tuple of (major version, minor version).
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.get_rng_state.html b/2.7/generated/torch.mtia.get_rng_state.html
index 8efe130ef49..7d8f2c55059 100644
--- a/2.7/generated/torch.mtia.get_rng_state.html
+++ b/2.7/generated/torch.mtia.get_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.get_rng_state
                                  
 
                                  
 
                                  
-torch.mtia.get_rng_state(device='mtia')[source][source]
                                  
+torch.mtia.get_rng_state(device='mtia')[source][source]
 
                                  Returns the random number generator state as a ByteTensor.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.init.html b/2.7/generated/torch.mtia.init.html
index f7ae8a9e8b3..996c845c984 100644
--- a/2.7/generated/torch.mtia.init.html
+++ b/2.7/generated/torch.mtia.init.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.init
                                  
 
                                  
 
                                  
-torch.mtia.init()[source][source]
                                  
+torch.mtia.init()[source][source]
 
                                  
 
 
                                  
diff --git a/2.7/generated/torch.mtia.is_available.html b/2.7/generated/torch.mtia.is_available.html
index 01065412b39..b6ca387acdc 100644
--- a/2.7/generated/torch.mtia.is_available.html
+++ b/2.7/generated/torch.mtia.is_available.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.is_available
                                  
 
                                  
 
                                  
-torch.mtia.is_available()[source][source]
                                  
+torch.mtia.is_available()[source][source]
 
                                  Return true if MTIA device is available
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mtia.is_initialized.html b/2.7/generated/torch.mtia.is_initialized.html
index 124014490a6..6b514f32fd3 100644
--- a/2.7/generated/torch.mtia.is_initialized.html
+++ b/2.7/generated/torch.mtia.is_initialized.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.is_initialized
                                  
 
                                  
 
                                  
-torch.mtia.is_initialized()[source][source]
                                  
+torch.mtia.is_initialized()[source][source]
 
                                  Return whether PyTorch’s MTIA state has been initialized.
                                  
 
                                  
 
diff --git a/2.7/generated/torch.mtia.memory.memory_stats.html b/2.7/generated/torch.mtia.memory.memory_stats.html
index 9cbff37b991..0388ec878bf 100644
--- a/2.7/generated/torch.mtia.memory.memory_stats.html
+++ b/2.7/generated/torch.mtia.memory.memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.memory.memory_stats
                                  
 
                                  
 
                                  
-torch.mtia.memory.memory_stats(device=None)[source][source]
                                  
+torch.mtia.memory.memory_stats(device=None)[source][source]
 
                                  Return a dictionary of MTIA memory allocator statistics for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.memory_stats.html b/2.7/generated/torch.mtia.memory_stats.html
index 65e6404a483..511581294ea 100644
--- a/2.7/generated/torch.mtia.memory_stats.html
+++ b/2.7/generated/torch.mtia.memory_stats.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.memory_stats
                                  
 
                                  
 
                                  
-torch.mtia.memory_stats(device=None)[source][source]
                                  
+torch.mtia.memory_stats(device=None)[source][source]
 
                                  Return a dictionary of MTIA memory allocator statistics for a given device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.record_memory_history.html b/2.7/generated/torch.mtia.record_memory_history.html
index 501b3b8926b..6efe9104e86 100644
--- a/2.7/generated/torch.mtia.record_memory_history.html
+++ b/2.7/generated/torch.mtia.record_memory_history.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.record_memory_history
                                  
 
                                  
 
                                  
-torch.mtia.record_memory_history(enabled='all', stacks='python', max_entries=0)[source][source]
                                  
+torch.mtia.record_memory_history(enabled='all', stacks='python', max_entries=0)[source][source]
 
                                  Enable/Disable the memory profiler on MTIA allocator
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.set_device.html b/2.7/generated/torch.mtia.set_device.html
index 11872079f85..ec96a051bd1 100644
--- a/2.7/generated/torch.mtia.set_device.html
+++ b/2.7/generated/torch.mtia.set_device.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.set_device
                                  
 
                                  
 
                                  
-torch.mtia.set_device(device)[source][source]
                                  
+torch.mtia.set_device(device)[source][source]
 
                                  Set the current device.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.set_rng_state.html b/2.7/generated/torch.mtia.set_rng_state.html
index 35fec321251..dc4f78b4d10 100644
--- a/2.7/generated/torch.mtia.set_rng_state.html
+++ b/2.7/generated/torch.mtia.set_rng_state.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.set_rng_state
                                  
 
                                  
 
                                  
-torch.mtia.set_rng_state(new_state, device='mtia')[source][source]
                                  
+torch.mtia.set_rng_state(new_state, device='mtia')[source][source]
 
                                  Sets the random number generator state.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.set_stream.html b/2.7/generated/torch.mtia.set_stream.html
index 2298a1d970d..ea16e5a43ff 100644
--- a/2.7/generated/torch.mtia.set_stream.html
+++ b/2.7/generated/torch.mtia.set_stream.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.set_stream
                                  
 
                                  
 
                                  
-torch.mtia.set_stream(stream)[source][source]
                                  
+torch.mtia.set_stream(stream)[source][source]
 
                                  
 
                                  Set the current stream.This is a wrapper API to set the stream.
                                  Usage of this function is discouraged in favor of the stream
 context manager.
                                  
diff --git a/2.7/generated/torch.mtia.snapshot.html b/2.7/generated/torch.mtia.snapshot.html
index 35a468ee9be..7d6bf54fd68 100644
--- a/2.7/generated/torch.mtia.snapshot.html
+++ b/2.7/generated/torch.mtia.snapshot.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.snapshot
                                  
 
                                  
 
                                  
-torch.mtia.snapshot()[source][source]
                                  
+torch.mtia.snapshot()[source][source]
 
                                  Return a dictionary of MTIA memory allocator history
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.mtia.stream.html b/2.7/generated/torch.mtia.stream.html
index e167b828f3a..31e8a34ac93 100644
--- a/2.7/generated/torch.mtia.stream.html
+++ b/2.7/generated/torch.mtia.stream.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.stream
                                  
 
                                  
 
                                  
-torch.mtia.stream(stream)[source][source]
                                  
+torch.mtia.stream(stream)[source][source]
 
                                  Wrap around the Context-manager StreamContext that selects a given stream.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.mtia.synchronize.html b/2.7/generated/torch.mtia.synchronize.html
index e2b14d79f1a..2626aec2dc9 100644
--- a/2.7/generated/torch.mtia.synchronize.html
+++ b/2.7/generated/torch.mtia.synchronize.html
@@ -596,7 +596,7 @@
 
                                  torch.mtia.synchronize
                                  
 
                                  
 
                                  
-torch.mtia.synchronize(device=None)[source][source]
                                  
+torch.mtia.synchronize(device=None)[source][source]
 
                                  Waits for all jobs in all streams on a MTIA device to complete.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveAvgPool1d.html b/2.7/generated/torch.nn.AdaptiveAvgPool1d.html
index deb6337df0e..0cd602c6e29 100644
--- a/2.7/generated/torch.nn.AdaptiveAvgPool1d.html
+++ b/2.7/generated/torch.nn.AdaptiveAvgPool1d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveAvgPool1d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveAvgPool1d(output_size)[source][source]
                                  
+class torch.nn.AdaptiveAvgPool1d(output_size)[source][source]
 
                                  Applies a 1D adaptive average pooling over an input signal composed of several input planes.
                                  
 
                                  The output size is LoutL_{out}, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveAvgPool2d.html b/2.7/generated/torch.nn.AdaptiveAvgPool2d.html
index 885addaa3a3..4bdb5f38aac 100644
--- a/2.7/generated/torch.nn.AdaptiveAvgPool2d.html
+++ b/2.7/generated/torch.nn.AdaptiveAvgPool2d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveAvgPool2d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveAvgPool2d(output_size)[source][source]
                                  
+class torch.nn.AdaptiveAvgPool2d(output_size)[source][source]
 
                                  Applies a 2D adaptive average pooling over an input signal composed of several input planes.
                                  
 
                                  The output is of size H x W, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveAvgPool3d.html b/2.7/generated/torch.nn.AdaptiveAvgPool3d.html
index fb9c8e360f6..8344c89e7a8 100644
--- a/2.7/generated/torch.nn.AdaptiveAvgPool3d.html
+++ b/2.7/generated/torch.nn.AdaptiveAvgPool3d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveAvgPool3d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveAvgPool3d(output_size)[source][source]
                                  
+class torch.nn.AdaptiveAvgPool3d(output_size)[source][source]
 
                                  Applies a 3D adaptive average pooling over an input signal composed of several input planes.
                                  
 
                                  The output is of size D x H x W, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveLogSoftmaxWithLoss.html b/2.7/generated/torch.nn.AdaptiveLogSoftmaxWithLoss.html
index ecabae345d3..f3434893ce1 100644
--- a/2.7/generated/torch.nn.AdaptiveLogSoftmaxWithLoss.html
+++ b/2.7/generated/torch.nn.AdaptiveLogSoftmaxWithLoss.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveLogSoftmaxWithLoss
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveLogSoftmaxWithLoss(in_features, n_classes, cutoffs, div_value=4.0, head_bias=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.AdaptiveLogSoftmaxWithLoss(in_features, n_classes, cutoffs, div_value=4.0, head_bias=False, device=None, dtype=None)[source][source]
 
                                  Efficient softmax approximation.
                                  
 
                                  As described in
 Efficient softmax approximation for GPUs by Edouard Grave, Armand Joulin,
@@ -690,7 +690,7 @@ 
                                  AdaptiveLogSoftmaxWithLoss
 
                                  
-log_prob(input)[source][source]
                                  
+log_prob(input)[source][source]
 
                                  Compute log probabilities for all n_classes\texttt{n\_classes}.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -716,7 +716,7 @@ 
                                  AdaptiveLogSoftmaxWithLoss
 
                                  
-predict(input)[source][source]
                                  
+predict(input)[source][source]
 
                                  Return the class with the highest probability for each example in the input minibatch.
                                  
 
                                  This is equivalent to self.log_prob(input).argmax(dim=1), but is more efficient in some cases.
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveMaxPool1d.html b/2.7/generated/torch.nn.AdaptiveMaxPool1d.html
index 3c6c00627e6..2adc310cc21 100644
--- a/2.7/generated/torch.nn.AdaptiveMaxPool1d.html
+++ b/2.7/generated/torch.nn.AdaptiveMaxPool1d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveMaxPool1d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveMaxPool1d(output_size, return_indices=False)[source][source]
                                  
+class torch.nn.AdaptiveMaxPool1d(output_size, return_indices=False)[source][source]
 
                                  Applies a 1D adaptive max pooling over an input signal composed of several input planes.
                                  
 
                                  The output size is LoutL_{out}, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveMaxPool2d.html b/2.7/generated/torch.nn.AdaptiveMaxPool2d.html
index 24fbd9a4de0..5709df8e529 100644
--- a/2.7/generated/torch.nn.AdaptiveMaxPool2d.html
+++ b/2.7/generated/torch.nn.AdaptiveMaxPool2d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveMaxPool2d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveMaxPool2d(output_size, return_indices=False)[source][source]
                                  
+class torch.nn.AdaptiveMaxPool2d(output_size, return_indices=False)[source][source]
 
                                  Applies a 2D adaptive max pooling over an input signal composed of several input planes.
                                  
 
                                  The output is of size Hout×WoutH_{out} \times W_{out}, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AdaptiveMaxPool3d.html b/2.7/generated/torch.nn.AdaptiveMaxPool3d.html
index 4c284e789e8..fb37423480a 100644
--- a/2.7/generated/torch.nn.AdaptiveMaxPool3d.html
+++ b/2.7/generated/torch.nn.AdaptiveMaxPool3d.html
@@ -596,7 +596,7 @@
 
                                  AdaptiveMaxPool3d
                                  
 
                                  
 
                                  
-class torch.nn.AdaptiveMaxPool3d(output_size, return_indices=False)[source][source]
                                  
+class torch.nn.AdaptiveMaxPool3d(output_size, return_indices=False)[source][source]
 
                                  Applies a 3D adaptive max pooling over an input signal composed of several input planes.
                                  
 
                                  The output is of size Dout×Hout×WoutD_{out} \times H_{out} \times W_{out}, for any input size.
 The number of output features is equal to the number of input planes.
                                  
diff --git a/2.7/generated/torch.nn.AlphaDropout.html b/2.7/generated/torch.nn.AlphaDropout.html
index f5a040efe18..bd819c691c0 100644
--- a/2.7/generated/torch.nn.AlphaDropout.html
+++ b/2.7/generated/torch.nn.AlphaDropout.html
@@ -596,7 +596,7 @@
 
                                  AlphaDropout
                                  
 
                                  
 
                                  
-class torch.nn.AlphaDropout(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.AlphaDropout(p=0.5, inplace=False)[source][source]
 
                                  Applies Alpha Dropout over the input.
                                  
 
                                  Alpha Dropout is a type of Dropout that maintains the self-normalizing
 property.
diff --git a/2.7/generated/torch.nn.AvgPool1d.html b/2.7/generated/torch.nn.AvgPool1d.html
index c7796760fda..b110826594c 100644
--- a/2.7/generated/torch.nn.AvgPool1d.html
+++ b/2.7/generated/torch.nn.AvgPool1d.html
@@ -596,7 +596,7 @@
 
                                  AvgPool1d
                                  
 
                                  
 
                                  
-class torch.nn.AvgPool1d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True)[source][source]
                                  
+class torch.nn.AvgPool1d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True)[source][source]
 
                                  Applies a 1D average pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,L)(N, C, L),
 output (N,C,Lout)(N, C, L_{out}) and kernel_size kk
diff --git a/2.7/generated/torch.nn.AvgPool2d.html b/2.7/generated/torch.nn.AvgPool2d.html
index 02b7cc6d5f3..93886db22a1 100644
--- a/2.7/generated/torch.nn.AvgPool2d.html
+++ b/2.7/generated/torch.nn.AvgPool2d.html
@@ -596,7 +596,7 @@
 
                                  AvgPool2d
                                  
 
                                  
 
                                  
-class torch.nn.AvgPool2d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
                                  
+class torch.nn.AvgPool2d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
 
                                  Applies a 2D average pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,H,W)(N, C, H, W),
 output (N,C,Hout,Wout)(N, C, H_{out}, W_{out}) and kernel_size (kH,kW)(kH, kW)
diff --git a/2.7/generated/torch.nn.AvgPool3d.html b/2.7/generated/torch.nn.AvgPool3d.html
index fe982d07032..0e80be0ba2e 100644
--- a/2.7/generated/torch.nn.AvgPool3d.html
+++ b/2.7/generated/torch.nn.AvgPool3d.html
@@ -596,7 +596,7 @@
 
                                  AvgPool3d
                                  
 
                                  
 
                                  
-class torch.nn.AvgPool3d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
                                  
+class torch.nn.AvgPool3d(kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source][source]
 
                                  Applies a 3D average pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,D,H,W)(N, C, D, H, W),
 output (N,C,Dout,Hout,Wout)(N, C, D_{out}, H_{out}, W_{out}) and kernel_size (kD,kH,kW)(kD, kH, kW)
diff --git a/2.7/generated/torch.nn.BCELoss.html b/2.7/generated/torch.nn.BCELoss.html
index 84e74779271..9a5962edc37 100644
--- a/2.7/generated/torch.nn.BCELoss.html
+++ b/2.7/generated/torch.nn.BCELoss.html
@@ -596,7 +596,7 @@
 
                                  BCELoss
                                  
 
                                  
 
                                  
-class torch.nn.BCELoss(weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.BCELoss(weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the Binary Cross Entropy between the target and
 the input probabilities:
                                  
 
                                  The unreduced (i.e. with reduction set to 'none') loss can be described as:
                                  
diff --git a/2.7/generated/torch.nn.BCEWithLogitsLoss.html b/2.7/generated/torch.nn.BCEWithLogitsLoss.html
index e3476738d89..697424f6019 100644
--- a/2.7/generated/torch.nn.BCEWithLogitsLoss.html
+++ b/2.7/generated/torch.nn.BCEWithLogitsLoss.html
@@ -596,7 +596,7 @@
 
                                  BCEWithLogitsLoss
                                  
 
                                  
 
                                  
-class torch.nn.BCEWithLogitsLoss(weight=None, size_average=None, reduce=None, reduction='mean', pos_weight=None)[source][source]
                                  
+class torch.nn.BCEWithLogitsLoss(weight=None, size_average=None, reduce=None, reduction='mean', pos_weight=None)[source][source]
 
                                  This loss combines a Sigmoid layer and the BCELoss in one single
 class. This version is more numerically stable than using a plain Sigmoid
 followed by a BCELoss as, by combining the operations into one layer,
diff --git a/2.7/generated/torch.nn.BatchNorm1d.html b/2.7/generated/torch.nn.BatchNorm1d.html
index cf0332e38db..08efb897cea 100644
--- a/2.7/generated/torch.nn.BatchNorm1d.html
+++ b/2.7/generated/torch.nn.BatchNorm1d.html
@@ -596,7 +596,7 @@
 
                                  BatchNorm1d
                                  
 
                                  
 
                                  
-class torch.nn.BatchNorm1d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.BatchNorm1d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  Applies Batch Normalization over a 2D or 3D input.
                                  
 
                                  Method described in the paper
 Batch Normalization: Accelerating Deep Network Training by Reducing
diff --git a/2.7/generated/torch.nn.BatchNorm2d.html b/2.7/generated/torch.nn.BatchNorm2d.html
index 51b422049cd..426f8620188 100644
--- a/2.7/generated/torch.nn.BatchNorm2d.html
+++ b/2.7/generated/torch.nn.BatchNorm2d.html
@@ -596,7 +596,7 @@
 
                                  BatchNorm2d
                                  
 
                                  
 
                                  
-class torch.nn.BatchNorm2d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.BatchNorm2d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  Applies Batch Normalization over a 4D input.
                                  
 
                                  4D is a mini-batch of 2D inputs
 with additional channel dimension. Method described in the paper
diff --git a/2.7/generated/torch.nn.BatchNorm3d.html b/2.7/generated/torch.nn.BatchNorm3d.html
index f52cb1feca8..3cbe02c7095 100644
--- a/2.7/generated/torch.nn.BatchNorm3d.html
+++ b/2.7/generated/torch.nn.BatchNorm3d.html
@@ -596,7 +596,7 @@
 
                                  BatchNorm3d
                                  
 
                                  
 
                                  
-class torch.nn.BatchNorm3d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.BatchNorm3d(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  Applies Batch Normalization over a 5D input.
                                  
 
                                  5D is a mini-batch of 3D inputs with additional channel dimension as described in the paper
 Batch Normalization: Accelerating Deep Network Training by Reducing
diff --git a/2.7/generated/torch.nn.Bilinear.html b/2.7/generated/torch.nn.Bilinear.html
index 9241100f312..647daaeefcd 100644
--- a/2.7/generated/torch.nn.Bilinear.html
+++ b/2.7/generated/torch.nn.Bilinear.html
@@ -596,7 +596,7 @@
 
                                  Bilinear
                                  
 
                                  
 
                                  
-class torch.nn.Bilinear(in1_features, in2_features, out_features, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.Bilinear(in1_features, in2_features, out_features, bias=True, device=None, dtype=None)[source][source]
 
                                  Applies a bilinear transformation to the incoming data: y=x1TAx2+by = x_1^T A x_2 + b.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.CELU.html b/2.7/generated/torch.nn.CELU.html
index ee2dad0fe69..ec7a18021f2 100644
--- a/2.7/generated/torch.nn.CELU.html
+++ b/2.7/generated/torch.nn.CELU.html
@@ -596,7 +596,7 @@
 
                                  CELU
                                  
 
                                  
 
                                  
-class torch.nn.CELU(alpha=1.0, inplace=False)[source][source]
                                  
+class torch.nn.CELU(alpha=1.0, inplace=False)[source][source]
 
                                  Applies the CELU function element-wise.
                                  
 
                                  
 CELU(x)=max⁡(0,x)+min⁡(0,α∗(exp⁡(x/α)−1))\text{CELU}(x) = \max(0,x) + \min(0, \alpha * (\exp(x/\alpha) - 1))
diff --git a/2.7/generated/torch.nn.CTCLoss.html b/2.7/generated/torch.nn.CTCLoss.html
index b37d4431ac4..565bce99519 100644
--- a/2.7/generated/torch.nn.CTCLoss.html
+++ b/2.7/generated/torch.nn.CTCLoss.html
@@ -596,7 +596,7 @@
 
                                  CTCLoss
                                  
 
                                  
 
                                  
-class torch.nn.CTCLoss(blank=0, reduction='mean', zero_infinity=False)[source][source]
                                  
+class torch.nn.CTCLoss(blank=0, reduction='mean', zero_infinity=False)[source][source]
 
                                  The Connectionist Temporal Classification loss.
                                  
 
                                  Calculates loss between a continuous (unsegmented) time series and a target sequence. CTCLoss sums over the
 probability of possible alignments of input to target, producing a loss value which is differentiable
diff --git a/2.7/generated/torch.nn.ChannelShuffle.html b/2.7/generated/torch.nn.ChannelShuffle.html
index d5b0bb5786f..b3971667441 100644
--- a/2.7/generated/torch.nn.ChannelShuffle.html
+++ b/2.7/generated/torch.nn.ChannelShuffle.html
@@ -596,7 +596,7 @@
 
                                  ChannelShuffle
                                  
 
                                  
 
                                  
-class torch.nn.ChannelShuffle(groups)[source][source]
                                  
+class torch.nn.ChannelShuffle(groups)[source][source]
 
                                  Divides and rearranges the channels in a tensor.
                                  
 
                                  This operation divides the channels in a tensor of shape (N,C,∗)(N, C, *)
 into g groups as (N,Cg,g,∗)(N, \frac{C}{g}, g, *) and shuffles them,
diff --git a/2.7/generated/torch.nn.CircularPad1d.html b/2.7/generated/torch.nn.CircularPad1d.html
index 9e86fd4c837..4f6c4b9b13c 100644
--- a/2.7/generated/torch.nn.CircularPad1d.html
+++ b/2.7/generated/torch.nn.CircularPad1d.html
@@ -596,7 +596,7 @@
 
                                  CircularPad1d
                                  
 
                                  
 
                                  
-class torch.nn.CircularPad1d(padding)[source][source]
                                  
+class torch.nn.CircularPad1d(padding)[source][source]
 
                                  Pads the input tensor using circular padding of the input boundary.
                                  
 
                                  Tensor values at the beginning of the dimension are used to pad the end,
 and values at the end are used to pad the beginning. If negative padding is
diff --git a/2.7/generated/torch.nn.CircularPad2d.html b/2.7/generated/torch.nn.CircularPad2d.html
index ec1cc4466fb..a16b3c9b3b6 100644
--- a/2.7/generated/torch.nn.CircularPad2d.html
+++ b/2.7/generated/torch.nn.CircularPad2d.html
@@ -596,7 +596,7 @@
 
                                  CircularPad2d
                                  
 
                                  
 
                                  
-class torch.nn.CircularPad2d(padding)[source][source]
                                  
+class torch.nn.CircularPad2d(padding)[source][source]
 
                                  Pads the input tensor using circular padding of the input boundary.
                                  
 
                                  Tensor values at the beginning of the dimension are used to pad the end,
 and values at the end are used to pad the beginning. If negative padding is
diff --git a/2.7/generated/torch.nn.CircularPad3d.html b/2.7/generated/torch.nn.CircularPad3d.html
index fce8e0e64cb..e5d51871526 100644
--- a/2.7/generated/torch.nn.CircularPad3d.html
+++ b/2.7/generated/torch.nn.CircularPad3d.html
@@ -596,7 +596,7 @@
 
                                  CircularPad3d
                                  
 
                                  
 
                                  
-class torch.nn.CircularPad3d(padding)[source][source]
                                  
+class torch.nn.CircularPad3d(padding)[source][source]
 
                                  Pads the input tensor using circular padding of the input boundary.
                                  
 
                                  Tensor values at the beginning of the dimension are used to pad the end,
 and values at the end are used to pad the beginning. If negative padding is
diff --git a/2.7/generated/torch.nn.ConstantPad1d.html b/2.7/generated/torch.nn.ConstantPad1d.html
index 470cfabe0fa..2f71d5b9d16 100644
--- a/2.7/generated/torch.nn.ConstantPad1d.html
+++ b/2.7/generated/torch.nn.ConstantPad1d.html
@@ -596,7 +596,7 @@
 
                                  ConstantPad1d
                                  
 
                                  
 
                                  
-class torch.nn.ConstantPad1d(padding, value)[source][source]
                                  
+class torch.nn.ConstantPad1d(padding, value)[source][source]
 
                                  Pads the input tensor boundaries with a constant value.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ConstantPad2d.html b/2.7/generated/torch.nn.ConstantPad2d.html
index 6d5c05acd01..e84796210d3 100644
--- a/2.7/generated/torch.nn.ConstantPad2d.html
+++ b/2.7/generated/torch.nn.ConstantPad2d.html
@@ -596,7 +596,7 @@
 
                                  ConstantPad2d
                                  
 
                                  
 
                                  
-class torch.nn.ConstantPad2d(padding, value)[source][source]
                                  
+class torch.nn.ConstantPad2d(padding, value)[source][source]
 
                                  Pads the input tensor boundaries with a constant value.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ConstantPad3d.html b/2.7/generated/torch.nn.ConstantPad3d.html
index 2d081ada4ba..3cb950aa162 100644
--- a/2.7/generated/torch.nn.ConstantPad3d.html
+++ b/2.7/generated/torch.nn.ConstantPad3d.html
@@ -596,7 +596,7 @@
 
                                  ConstantPad3d
                                  
 
                                  
 
                                  
-class torch.nn.ConstantPad3d(padding, value)[source][source]
                                  
+class torch.nn.ConstantPad3d(padding, value)[source][source]
 
                                  Pads the input tensor boundaries with a constant value.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Conv1d.html b/2.7/generated/torch.nn.Conv1d.html
index 173f899f853..e957aa47d1e 100644
--- a/2.7/generated/torch.nn.Conv1d.html
+++ b/2.7/generated/torch.nn.Conv1d.html
@@ -596,7 +596,7 @@
 
                                  Conv1d
                                  
 
                                  
 
                                  
-class torch.nn.Conv1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.Conv1d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 1D convolution over an input signal composed of several input
 planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size
diff --git a/2.7/generated/torch.nn.Conv2d.html b/2.7/generated/torch.nn.Conv2d.html
index 98f5e7cbb95..2e8173b52b6 100644
--- a/2.7/generated/torch.nn.Conv2d.html
+++ b/2.7/generated/torch.nn.Conv2d.html
@@ -596,7 +596,7 @@
 
                                  Conv2d
                                  
 
                                  
 
                                  
-class torch.nn.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 2D convolution over an input signal composed of several input
 planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size
diff --git a/2.7/generated/torch.nn.Conv3d.html b/2.7/generated/torch.nn.Conv3d.html
index 7e242df499a..84c6b474c88 100644
--- a/2.7/generated/torch.nn.Conv3d.html
+++ b/2.7/generated/torch.nn.Conv3d.html
@@ -596,7 +596,7 @@
 
                                  Conv3d
                                  
 
                                  
 
                                  
-class torch.nn.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.Conv3d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 3D convolution over an input signal composed of several input
 planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,Cin,D,H,W)(N, C_{in}, D, H, W)
diff --git a/2.7/generated/torch.nn.ConvTranspose1d.html b/2.7/generated/torch.nn.ConvTranspose1d.html
index d0fb66eb953..7bcb1eafb5d 100644
--- a/2.7/generated/torch.nn.ConvTranspose1d.html
+++ b/2.7/generated/torch.nn.ConvTranspose1d.html
@@ -596,7 +596,7 @@
 
                                  ConvTranspose1d
                                  
 
                                  
 
                                  
-class torch.nn.ConvTranspose1d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.ConvTranspose1d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 1D transposed convolution operator over an input image
 composed of several input planes.
                                  
 
                                  This module can be seen as the gradient of Conv1d with respect to its input.
diff --git a/2.7/generated/torch.nn.ConvTranspose2d.html b/2.7/generated/torch.nn.ConvTranspose2d.html
index a5b172fd1b6..9ae8e688475 100644
--- a/2.7/generated/torch.nn.ConvTranspose2d.html
+++ b/2.7/generated/torch.nn.ConvTranspose2d.html
@@ -596,7 +596,7 @@
 
                                  ConvTranspose2d
                                  
 
                                  
 
                                  
-class torch.nn.ConvTranspose2d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.ConvTranspose2d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 2D transposed convolution operator over an input image
 composed of several input planes.
                                  
 
                                  This module can be seen as the gradient of Conv2d with respect to its input.
diff --git a/2.7/generated/torch.nn.ConvTranspose3d.html b/2.7/generated/torch.nn.ConvTranspose3d.html
index fde5515c0cf..6955e60026e 100644
--- a/2.7/generated/torch.nn.ConvTranspose3d.html
+++ b/2.7/generated/torch.nn.ConvTranspose3d.html
@@ -596,7 +596,7 @@
 
                                  ConvTranspose3d
                                  
 
                                  
 
                                  
-class torch.nn.ConvTranspose3d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.ConvTranspose3d(in_channels, out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  Applies a 3D transposed convolution operator over an input image composed of several input
 planes.
 The transposed convolution operator multiplies each input value element-wise by a learnable kernel,
diff --git a/2.7/generated/torch.nn.CosineEmbeddingLoss.html b/2.7/generated/torch.nn.CosineEmbeddingLoss.html
index b5c93551c63..e9002ff9331 100644
--- a/2.7/generated/torch.nn.CosineEmbeddingLoss.html
+++ b/2.7/generated/torch.nn.CosineEmbeddingLoss.html
@@ -596,7 +596,7 @@
 
                                  CosineEmbeddingLoss
                                  
 
                                  
 
                                  
-class torch.nn.CosineEmbeddingLoss(margin=0.0, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.CosineEmbeddingLoss(margin=0.0, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the loss given input tensors
 x1x_1, x2x_2 and a Tensor label yy with values 1 or -1.
 Use (y=1y=1) to maximize the cosine similarity of two inputs, and (y=−1y=-1) otherwise.
diff --git a/2.7/generated/torch.nn.CosineSimilarity.html b/2.7/generated/torch.nn.CosineSimilarity.html
index 8d85c519ec9..053c4706ebb 100644
--- a/2.7/generated/torch.nn.CosineSimilarity.html
+++ b/2.7/generated/torch.nn.CosineSimilarity.html
@@ -596,7 +596,7 @@
 
                                  CosineSimilarity
                                  
 
                                  
 
                                  
-class torch.nn.CosineSimilarity(dim=1, eps=1e-08)[source][source]
                                  
+class torch.nn.CosineSimilarity(dim=1, eps=1e-08)[source][source]
 
                                  Returns cosine similarity between x1x_1 and x2x_2, computed along dim.
                                  
 
                                  
 similarity=x1⋅x2max⁡(∥x1∥2⋅∥x2∥2,ϵ).\text{similarity} = \dfrac{x_1 \cdot x_2}{\max(\Vert x_1 \Vert _2 \cdot \Vert x_2 \Vert _2, \epsilon)}.
diff --git a/2.7/generated/torch.nn.CrossEntropyLoss.html b/2.7/generated/torch.nn.CrossEntropyLoss.html
index 4fb95f632f8..35f67bc8108 100644
--- a/2.7/generated/torch.nn.CrossEntropyLoss.html
+++ b/2.7/generated/torch.nn.CrossEntropyLoss.html
@@ -596,7 +596,7 @@
 
                                  CrossEntropyLoss
                                  
 
                                  
 
                                  
-class torch.nn.CrossEntropyLoss(weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean', label_smoothing=0.0)[source][source]
                                  
+class torch.nn.CrossEntropyLoss(weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean', label_smoothing=0.0)[source][source]
 
                                  This criterion computes the cross entropy loss between input logits
 and target.
                                  
 
                                  It is useful when training a classification problem with C classes.
diff --git a/2.7/generated/torch.nn.DataParallel.html b/2.7/generated/torch.nn.DataParallel.html
index 09ca36fdea5..10a8fb37646 100644
--- a/2.7/generated/torch.nn.DataParallel.html
+++ b/2.7/generated/torch.nn.DataParallel.html
@@ -596,7 +596,7 @@
 
                                  DataParallel
                                  
 
                                  
 
                                  
-class torch.nn.DataParallel(module, device_ids=None, output_device=None, dim=0)[source][source]
                                  
+class torch.nn.DataParallel(module, device_ids=None, output_device=None, dim=0)[source][source]
 
                                  Implements data parallelism at the module level.
                                  
 
                                  This container parallelizes the application of the given module by
 splitting the input across the specified devices by chunking in the batch
diff --git a/2.7/generated/torch.nn.Dropout.html b/2.7/generated/torch.nn.Dropout.html
index f3dd4019d8f..85b83b10e58 100644
--- a/2.7/generated/torch.nn.Dropout.html
+++ b/2.7/generated/torch.nn.Dropout.html
@@ -596,7 +596,7 @@
 
                                  Dropout
                                  
 
                                  
 
                                  
-class torch.nn.Dropout(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.Dropout(p=0.5, inplace=False)[source][source]
 
                                  During training, randomly zeroes some of the elements of the input tensor with probability p.
                                  
 
                                  The zeroed elements are chosen independently for each forward call and are sampled from a Bernoulli distribution.
                                  
 
                                  Each channel will be zeroed out independently on every forward call.
                                  
diff --git a/2.7/generated/torch.nn.Dropout1d.html b/2.7/generated/torch.nn.Dropout1d.html
index 34911f7c27c..9506fbf6380 100644
--- a/2.7/generated/torch.nn.Dropout1d.html
+++ b/2.7/generated/torch.nn.Dropout1d.html
@@ -596,7 +596,7 @@
 
                                  Dropout1d
                                  
 
                                  
 
                                  
-class torch.nn.Dropout1d(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.Dropout1d(p=0.5, inplace=False)[source][source]
 
                                  Randomly zero out entire channels.
                                  
 
                                  A channel is a 1D feature map,
 e.g., the jj-th channel of the ii-th sample in the
diff --git a/2.7/generated/torch.nn.Dropout2d.html b/2.7/generated/torch.nn.Dropout2d.html
index d7fd5e516a1..f0a9698efd9 100644
--- a/2.7/generated/torch.nn.Dropout2d.html
+++ b/2.7/generated/torch.nn.Dropout2d.html
@@ -596,7 +596,7 @@
 
                                  Dropout2d
                                  
 
                                  
 
                                  
-class torch.nn.Dropout2d(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.Dropout2d(p=0.5, inplace=False)[source][source]
 
                                  Randomly zero out entire channels.
                                  
 
                                  A channel is a 2D feature map,
 e.g., the jj-th channel of the ii-th sample in the
diff --git a/2.7/generated/torch.nn.Dropout3d.html b/2.7/generated/torch.nn.Dropout3d.html
index 228c33a7e3a..d5f8d9a3140 100644
--- a/2.7/generated/torch.nn.Dropout3d.html
+++ b/2.7/generated/torch.nn.Dropout3d.html
@@ -596,7 +596,7 @@
 
                                  Dropout3d
                                  
 
                                  
 
                                  
-class torch.nn.Dropout3d(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.Dropout3d(p=0.5, inplace=False)[source][source]
 
                                  Randomly zero out entire channels.
                                  
 
                                  A channel is a 3D feature map,
 e.g., the jj-th channel of the ii-th sample in the
diff --git a/2.7/generated/torch.nn.ELU.html b/2.7/generated/torch.nn.ELU.html
index 11c4e9df44f..36631f1cdd6 100644
--- a/2.7/generated/torch.nn.ELU.html
+++ b/2.7/generated/torch.nn.ELU.html
@@ -596,7 +596,7 @@
 
                                  ELU
                                  
 
                                  
 
                                  
-class torch.nn.ELU(alpha=1.0, inplace=False)[source][source]
                                  
+class torch.nn.ELU(alpha=1.0, inplace=False)[source][source]
 
                                  Applies the Exponential Linear Unit (ELU) function, element-wise.
                                  
 
                                  Method described in the paper: Fast and Accurate Deep Network Learning by Exponential Linear
 Units (ELUs).
                                  
diff --git a/2.7/generated/torch.nn.Embedding.html b/2.7/generated/torch.nn.Embedding.html
index c848a0060e9..65b84980a7f 100644
--- a/2.7/generated/torch.nn.Embedding.html
+++ b/2.7/generated/torch.nn.Embedding.html
@@ -596,7 +596,7 @@
 
                                  Embedding
                                  
 
                                  
 
                                  
-class torch.nn.Embedding(num_embeddings, embedding_dim, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False, _weight=None, _freeze=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.Embedding(num_embeddings, embedding_dim, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False, _weight=None, _freeze=False, device=None, dtype=None)[source][source]
 
                                  A simple lookup table that stores embeddings of a fixed dictionary and size.
                                  
 
                                  This module is often used to store word embeddings and retrieve them using indices.
 The input to the module is a list of indices, and the output is the corresponding
@@ -702,7 +702,7 @@ 
                                  Embedding
 
                                  
-classmethod from_pretrained(embeddings, freeze=True, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False)[source][source]
                                  
+classmethod from_pretrained(embeddings, freeze=True, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False)[source][source]
 
                                  Create Embedding instance from given 2-dimensional FloatTensor.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.EmbeddingBag.html b/2.7/generated/torch.nn.EmbeddingBag.html
index af2890a4703..0a3206f3a31 100644
--- a/2.7/generated/torch.nn.EmbeddingBag.html
+++ b/2.7/generated/torch.nn.EmbeddingBag.html
@@ -596,7 +596,7 @@
 
                                  EmbeddingBag
                                  
 
                                  
 
                                  
-class torch.nn.EmbeddingBag(num_embeddings, embedding_dim, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='mean', sparse=False, _weight=None, include_last_offset=False, padding_idx=None, device=None, dtype=None)[source][source]
                                  
+class torch.nn.EmbeddingBag(num_embeddings, embedding_dim, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='mean', sparse=False, _weight=None, include_last_offset=False, padding_idx=None, device=None, dtype=None)[source][source]
 
                                  Compute sums or means of ‘bags’ of embeddings, without instantiating the intermediate embeddings.
                                  
 
                                  For bags of constant length, no per_sample_weights, no indices equal to padding_idx,
 and with 2D inputs, this class
                                  
@@ -677,7 +677,7 @@ 
                                  EmbeddingBag
 
                                  
-forward(input, offsets=None, per_sample_weights=None)[source][source]
                                  
+forward(input, offsets=None, per_sample_weights=None)[source][source]
 
                                  Forward pass of EmbeddingBag.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -717,7 +717,7 @@ 
                                  EmbeddingBag
 
                                  
-classmethod from_pretrained(embeddings, freeze=True, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='mean', sparse=False, include_last_offset=False, padding_idx=None)[source][source]
                                  
+classmethod from_pretrained(embeddings, freeze=True, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, mode='mean', sparse=False, include_last_offset=False, padding_idx=None)[source][source]
 
                                  Create EmbeddingBag instance from given 2-dimensional FloatTensor.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.FeatureAlphaDropout.html b/2.7/generated/torch.nn.FeatureAlphaDropout.html
index f20a7a482d0..13c4d25a86b 100644
--- a/2.7/generated/torch.nn.FeatureAlphaDropout.html
+++ b/2.7/generated/torch.nn.FeatureAlphaDropout.html
@@ -596,7 +596,7 @@
 
                                  FeatureAlphaDropout
                                  
 
                                  
 
                                  
-class torch.nn.FeatureAlphaDropout(p=0.5, inplace=False)[source][source]
                                  
+class torch.nn.FeatureAlphaDropout(p=0.5, inplace=False)[source][source]
 
                                  Randomly masks out entire channels.
                                  
 
                                  A channel is a feature map,
 e.g. the jj-th channel of the ii-th sample in the batch input
diff --git a/2.7/generated/torch.nn.Flatten.html b/2.7/generated/torch.nn.Flatten.html
index cf9ac5ec2f4..3f39594ad3d 100644
--- a/2.7/generated/torch.nn.Flatten.html
+++ b/2.7/generated/torch.nn.Flatten.html
@@ -596,7 +596,7 @@
 
                                  Flatten
                                  
 
                                  
 
                                  
-class torch.nn.Flatten(start_dim=1, end_dim=-1)[source][source]
                                  
+class torch.nn.Flatten(start_dim=1, end_dim=-1)[source][source]
 
                                  Flattens a contiguous range of dims into a tensor.
                                  
 
                                  For use with Sequential, see torch.flatten() for details.
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Fold.html b/2.7/generated/torch.nn.Fold.html
index 92a965db4fa..bfb782e5a9c 100644
--- a/2.7/generated/torch.nn.Fold.html
+++ b/2.7/generated/torch.nn.Fold.html
@@ -596,7 +596,7 @@
 
                                  Fold
                                  
 
                                  
 
                                  
-class torch.nn.Fold(output_size, kernel_size, dilation=1, padding=0, stride=1)[source][source]
                                  
+class torch.nn.Fold(output_size, kernel_size, dilation=1, padding=0, stride=1)[source][source]
 
                                  Combines an array of sliding local blocks into a large containing tensor.
                                  
 
                                  Consider a batched input tensor containing sliding local blocks,
 e.g., patches of images, of shape (N,C×∏(kernel_size),L)(N, C \times  \prod(\text{kernel\_size}), L),
diff --git a/2.7/generated/torch.nn.FractionalMaxPool2d.html b/2.7/generated/torch.nn.FractionalMaxPool2d.html
index 2db87c6afc9..d4c4433932b 100644
--- a/2.7/generated/torch.nn.FractionalMaxPool2d.html
+++ b/2.7/generated/torch.nn.FractionalMaxPool2d.html
@@ -596,7 +596,7 @@
 
                                  FractionalMaxPool2d
                                  
 
                                  
 
                                  
-class torch.nn.FractionalMaxPool2d(kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source][source]
                                  
+class torch.nn.FractionalMaxPool2d(kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source][source]
 
                                  Applies a 2D fractional max pooling over an input signal composed of several input planes.
                                  
 
                                  Fractional MaxPooling is described in detail in the paper Fractional MaxPooling by Ben Graham
                                  
 
                                  The max-pooling operation is applied in kH×kWkH \times kW regions by a stochastic
diff --git a/2.7/generated/torch.nn.FractionalMaxPool3d.html b/2.7/generated/torch.nn.FractionalMaxPool3d.html
index 9ca82dd7cee..87fac498c09 100644
--- a/2.7/generated/torch.nn.FractionalMaxPool3d.html
+++ b/2.7/generated/torch.nn.FractionalMaxPool3d.html
@@ -596,7 +596,7 @@
 
                                  FractionalMaxPool3d
                                  
 
                                  
 
                                  
-class torch.nn.FractionalMaxPool3d(kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source][source]
                                  
+class torch.nn.FractionalMaxPool3d(kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source][source]
 
                                  Applies a 3D fractional max pooling over an input signal composed of several input planes.
                                  
 
                                  Fractional MaxPooling is described in detail in the paper Fractional MaxPooling by Ben Graham
                                  
 
                                  The max-pooling operation is applied in kT×kH×kWkT \times kH \times kW regions by a stochastic
diff --git a/2.7/generated/torch.nn.GELU.html b/2.7/generated/torch.nn.GELU.html
index 448e08802e4..739a6a8961f 100644
--- a/2.7/generated/torch.nn.GELU.html
+++ b/2.7/generated/torch.nn.GELU.html
@@ -596,7 +596,7 @@
 
                                  GELU
                                  
 
                                  
 
                                  
-class torch.nn.GELU(approximate='none')[source][source]
                                  
+class torch.nn.GELU(approximate='none')[source][source]
 
                                  Applies the Gaussian Error Linear Units function.
                                  
 
                                  
 GELU(x)=x∗Φ(x)\text{GELU}(x) = x * \Phi(x)
diff --git a/2.7/generated/torch.nn.GLU.html b/2.7/generated/torch.nn.GLU.html
index 459594e1563..4d00be4b0ce 100644
--- a/2.7/generated/torch.nn.GLU.html
+++ b/2.7/generated/torch.nn.GLU.html
@@ -596,7 +596,7 @@
 
                                  GLU
                                  
 
                                  
 
                                  
-class torch.nn.GLU(dim=-1)[source][source]
                                  
+class torch.nn.GLU(dim=-1)[source][source]
 
                                  Applies the gated linear unit function.
                                  
 
                                  GLU(a,b)=a⊗σ(b){GLU}(a, b)= a \otimes \sigma(b) where aa is the first half
 of the input matrices and bb is the second half.
                                  
diff --git a/2.7/generated/torch.nn.GRU.html b/2.7/generated/torch.nn.GRU.html
index a2cc61f8740..470b7cb83db 100644
--- a/2.7/generated/torch.nn.GRU.html
+++ b/2.7/generated/torch.nn.GRU.html
@@ -596,7 +596,7 @@
 
                                  GRU
                                  
 
                                  
 
                                  
-class torch.nn.GRU(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.GRU(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None)[source][source]
 
                                  Apply a multi-layer gated recurrent unit (GRU) RNN to an input sequence.
 For each element in the input sequence, each layer computes the following
 function:
                                  
diff --git a/2.7/generated/torch.nn.GRUCell.html b/2.7/generated/torch.nn.GRUCell.html
index b4305adcaec..f1c8448c102 100644
--- a/2.7/generated/torch.nn.GRUCell.html
+++ b/2.7/generated/torch.nn.GRUCell.html
@@ -596,7 +596,7 @@
 
                                  GRUCell
                                  
 
                                  
 
                                  
-class torch.nn.GRUCell(input_size, hidden_size, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.GRUCell(input_size, hidden_size, bias=True, device=None, dtype=None)[source][source]
 
                                  A gated recurrent unit (GRU) cell.
                                  
 
                                  
 r=σ(Wirx+bir+Whrh+bhr)z=σ(Wizx+biz+Whzh+bhz)n=tanh⁡(Winx+bin+r⊙(Whnh+bhn))h′=(1−z)⊙n+z⊙h\begin{array}{ll}
diff --git a/2.7/generated/torch.nn.GaussianNLLLoss.html b/2.7/generated/torch.nn.GaussianNLLLoss.html
index 18806ba7798..1c0c66ea612 100644
--- a/2.7/generated/torch.nn.GaussianNLLLoss.html
+++ b/2.7/generated/torch.nn.GaussianNLLLoss.html
@@ -596,7 +596,7 @@
 
                                  GaussianNLLLoss
                                  
 
                                  
 
                                  
-class torch.nn.GaussianNLLLoss(*, full=False, eps=1e-06, reduction='mean')[source][source]
                                  
+class torch.nn.GaussianNLLLoss(*, full=False, eps=1e-06, reduction='mean')[source][source]
 
                                  Gaussian negative log likelihood loss.
                                  
 
                                  The targets are treated as samples from Gaussian distributions with
 expectations and variances predicted by the neural network. For a
diff --git a/2.7/generated/torch.nn.GroupNorm.html b/2.7/generated/torch.nn.GroupNorm.html
index 80b25416804..a88d24f5139 100644
--- a/2.7/generated/torch.nn.GroupNorm.html
+++ b/2.7/generated/torch.nn.GroupNorm.html
@@ -596,7 +596,7 @@
 
                                  GroupNorm
                                  
 
                                  
 
                                  
-class torch.nn.GroupNorm(num_groups, num_channels, eps=1e-05, affine=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.GroupNorm(num_groups, num_channels, eps=1e-05, affine=True, device=None, dtype=None)[source][source]
 
                                  Applies Group Normalization over a mini-batch of inputs.
                                  
 
                                  This layer implements the operation as described in
 the paper Group Normalization
                                  
diff --git a/2.7/generated/torch.nn.Hardshrink.html b/2.7/generated/torch.nn.Hardshrink.html
index 32389ada69e..13529129fdd 100644
--- a/2.7/generated/torch.nn.Hardshrink.html
+++ b/2.7/generated/torch.nn.Hardshrink.html
@@ -596,7 +596,7 @@
 
                                  Hardshrink
                                  
 
                                  
 
                                  
-class torch.nn.Hardshrink(lambd=0.5)[source][source]
                                  
+class torch.nn.Hardshrink(lambd=0.5)[source][source]
 
                                  Applies the Hard Shrinkage (Hardshrink) function element-wise.
                                  
 
                                  Hardshrink is defined as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Hardsigmoid.html b/2.7/generated/torch.nn.Hardsigmoid.html
index e82c34c6d17..0142a1b98a4 100644
--- a/2.7/generated/torch.nn.Hardsigmoid.html
+++ b/2.7/generated/torch.nn.Hardsigmoid.html
@@ -596,7 +596,7 @@
 
                                  Hardsigmoid
                                  
 
                                  
 
                                  
-class torch.nn.Hardsigmoid(inplace=False)[source][source]
                                  
+class torch.nn.Hardsigmoid(inplace=False)[source][source]
 
                                  Applies the Hardsigmoid function element-wise.
                                  
 
                                  Hardsigmoid is defined as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Hardswish.html b/2.7/generated/torch.nn.Hardswish.html
index f5d8160d5ec..82e054593f9 100644
--- a/2.7/generated/torch.nn.Hardswish.html
+++ b/2.7/generated/torch.nn.Hardswish.html
@@ -596,7 +596,7 @@
 
                                  Hardswish
                                  
 
                                  
 
                                  
-class torch.nn.Hardswish(inplace=False)[source][source]
                                  
+class torch.nn.Hardswish(inplace=False)[source][source]
 
                                  Applies the Hardswish function, element-wise.
                                  
 
                                  Method described in the paper: Searching for MobileNetV3.
                                  
 
                                  Hardswish is defined as:
                                  
diff --git a/2.7/generated/torch.nn.Hardtanh.html b/2.7/generated/torch.nn.Hardtanh.html
index ef9b2f94e14..c3006d4ab41 100644
--- a/2.7/generated/torch.nn.Hardtanh.html
+++ b/2.7/generated/torch.nn.Hardtanh.html
@@ -596,7 +596,7 @@
 
                                  Hardtanh
                                  
 
                                  
 
                                  
-class torch.nn.Hardtanh(min_val=-1.0, max_val=1.0, inplace=False, min_value=None, max_value=None)[source][source]
                                  
+class torch.nn.Hardtanh(min_val=-1.0, max_val=1.0, inplace=False, min_value=None, max_value=None)[source][source]
 
                                  Applies the HardTanh function element-wise.
                                  
 
                                  HardTanh is defined as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.HingeEmbeddingLoss.html b/2.7/generated/torch.nn.HingeEmbeddingLoss.html
index eedd98f6802..f72fa12fde3 100644
--- a/2.7/generated/torch.nn.HingeEmbeddingLoss.html
+++ b/2.7/generated/torch.nn.HingeEmbeddingLoss.html
@@ -596,7 +596,7 @@
 
                                  HingeEmbeddingLoss
                                  
 
                                  
 
                                  
-class torch.nn.HingeEmbeddingLoss(margin=1.0, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.HingeEmbeddingLoss(margin=1.0, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Measures the loss given an input tensor xx and a labels tensor yy
 (containing 1 or -1).
 This is usually used for measuring whether two inputs are similar or
diff --git a/2.7/generated/torch.nn.HuberLoss.html b/2.7/generated/torch.nn.HuberLoss.html
index ff09b7e60dc..c13142f07ca 100644
--- a/2.7/generated/torch.nn.HuberLoss.html
+++ b/2.7/generated/torch.nn.HuberLoss.html
@@ -596,7 +596,7 @@
 
                                  HuberLoss
                                  
 
                                  
 
                                  
-class torch.nn.HuberLoss(reduction='mean', delta=1.0)[source][source]
                                  
+class torch.nn.HuberLoss(reduction='mean', delta=1.0)[source][source]
 
                                  Creates a criterion that uses a squared term if the absolute
 element-wise error falls below delta and a delta-scaled L1 term otherwise.
 This loss combines advantages of both L1Loss and MSELoss; the
diff --git a/2.7/generated/torch.nn.Identity.html b/2.7/generated/torch.nn.Identity.html
index f29534a43dc..8708e83de6f 100644
--- a/2.7/generated/torch.nn.Identity.html
+++ b/2.7/generated/torch.nn.Identity.html
@@ -596,7 +596,7 @@
 
                                  Identity
                                  
 
                                  
 
                                  
-class torch.nn.Identity(*args, **kwargs)[source][source]
                                  
+class torch.nn.Identity(*args, **kwargs)[source][source]
 
                                  A placeholder identity operator that is argument-insensitive.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.InstanceNorm1d.html b/2.7/generated/torch.nn.InstanceNorm1d.html
index a0aa93409e7..63820b52640 100644
--- a/2.7/generated/torch.nn.InstanceNorm1d.html
+++ b/2.7/generated/torch.nn.InstanceNorm1d.html
@@ -596,7 +596,7 @@
 
                                  InstanceNorm1d
                                  
 
                                  
 
                                  
-class torch.nn.InstanceNorm1d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.InstanceNorm1d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                                  Applies Instance Normalization.
                                  
 
                                  This operation applies Instance Normalization
 over a 2D (unbatched) or 3D (batched) input as described in the paper
diff --git a/2.7/generated/torch.nn.InstanceNorm2d.html b/2.7/generated/torch.nn.InstanceNorm2d.html
index 42a20d1079e..4f8a6bed2c7 100644
--- a/2.7/generated/torch.nn.InstanceNorm2d.html
+++ b/2.7/generated/torch.nn.InstanceNorm2d.html
@@ -596,7 +596,7 @@
 
                                  InstanceNorm2d
                                  
 
                                  
 
                                  
-class torch.nn.InstanceNorm2d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.InstanceNorm2d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                                  Applies Instance Normalization.
                                  
 
                                  This operation applies Instance Normalization
 over a 4D input (a mini-batch of 2D inputs
diff --git a/2.7/generated/torch.nn.InstanceNorm3d.html b/2.7/generated/torch.nn.InstanceNorm3d.html
index 58b4661d72a..6e3c7fec08d 100644
--- a/2.7/generated/torch.nn.InstanceNorm3d.html
+++ b/2.7/generated/torch.nn.InstanceNorm3d.html
@@ -596,7 +596,7 @@
 
                                  InstanceNorm3d
                                  
 
                                  
 
                                  
-class torch.nn.InstanceNorm3d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.InstanceNorm3d(num_features, eps=1e-05, momentum=0.1, affine=False, track_running_stats=False, device=None, dtype=None)[source][source]
 
                                  Applies Instance Normalization.
                                  
 
                                  This operation applies Instance Normalization
 over a 5D input (a mini-batch of 3D inputs with additional channel dimension) as described in the paper
diff --git a/2.7/generated/torch.nn.KLDivLoss.html b/2.7/generated/torch.nn.KLDivLoss.html
index 3988ece4036..b232c212172 100644
--- a/2.7/generated/torch.nn.KLDivLoss.html
+++ b/2.7/generated/torch.nn.KLDivLoss.html
@@ -596,7 +596,7 @@
 
                                  KLDivLoss
                                  
 
                                  
 
                                  
-class torch.nn.KLDivLoss(size_average=None, reduce=None, reduction='mean', log_target=False)[source][source]
                                  
+class torch.nn.KLDivLoss(size_average=None, reduce=None, reduction='mean', log_target=False)[source][source]
 
                                  The Kullback-Leibler divergence loss.
                                  
 
                                  For tensors of the same shape ypred, ytruey_{\text{pred}},\ y_{\text{true}},
 where ypredy_{\text{pred}} is the input and ytruey_{\text{true}} is the
diff --git a/2.7/generated/torch.nn.L1Loss.html b/2.7/generated/torch.nn.L1Loss.html
index 43ce91b07ef..0b0c5826517 100644
--- a/2.7/generated/torch.nn.L1Loss.html
+++ b/2.7/generated/torch.nn.L1Loss.html
@@ -596,7 +596,7 @@
 
                                  L1Loss
                                  
 
                                  
 
                                  
-class torch.nn.L1Loss(size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.L1Loss(size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the mean absolute error (MAE) between each element in
 the input xx and target yy.
                                  
 
                                  The unreduced (i.e. with reduction set to 'none') loss can be described as:
                                  
diff --git a/2.7/generated/torch.nn.LPPool1d.html b/2.7/generated/torch.nn.LPPool1d.html
index e352e1f318e..ac9286d13a8 100644
--- a/2.7/generated/torch.nn.LPPool1d.html
+++ b/2.7/generated/torch.nn.LPPool1d.html
@@ -596,7 +596,7 @@
 
                                  LPPool1d
                                  
 
                                  
 
                                  
-class torch.nn.LPPool1d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                  
+class torch.nn.LPPool1d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                  Applies a 1D power-average pooling over an input signal composed of several input planes.
                                  
 
                                  On each window, the function computed is:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.LPPool2d.html b/2.7/generated/torch.nn.LPPool2d.html
index 86f2002b872..28504f6a906 100644
--- a/2.7/generated/torch.nn.LPPool2d.html
+++ b/2.7/generated/torch.nn.LPPool2d.html
@@ -596,7 +596,7 @@
 
                                  LPPool2d
                                  
 
                                  
 
                                  
-class torch.nn.LPPool2d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                  
+class torch.nn.LPPool2d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                  Applies a 2D power-average pooling over an input signal composed of several input planes.
                                  
 
                                  On each window, the function computed is:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.LPPool3d.html b/2.7/generated/torch.nn.LPPool3d.html
index 45e5964277f..876b3a001ff 100644
--- a/2.7/generated/torch.nn.LPPool3d.html
+++ b/2.7/generated/torch.nn.LPPool3d.html
@@ -596,7 +596,7 @@
 
                                  LPPool3d
                                  
 
                                  
 
                                  
-class torch.nn.LPPool3d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                  
+class torch.nn.LPPool3d(norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                  Applies a 3D power-average pooling over an input signal composed of several input planes.
                                  
 
                                  On each window, the function computed is:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.LSTM.html b/2.7/generated/torch.nn.LSTM.html
index 3001adefeed..ff1968ef0c9 100644
--- a/2.7/generated/torch.nn.LSTM.html
+++ b/2.7/generated/torch.nn.LSTM.html
@@ -596,7 +596,7 @@
 
                                  LSTM
                                  
 
                                  
 
                                  
-class torch.nn.LSTM(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, proj_size=0, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LSTM(input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, proj_size=0, device=None, dtype=None)[source][source]
 
                                  Apply a multi-layer long short-term memory (LSTM) RNN to an input sequence.
 For each element in the input sequence, each layer computes the following
 function:
                                  
diff --git a/2.7/generated/torch.nn.LSTMCell.html b/2.7/generated/torch.nn.LSTMCell.html
index d978adda8c7..46dd67086c2 100644
--- a/2.7/generated/torch.nn.LSTMCell.html
+++ b/2.7/generated/torch.nn.LSTMCell.html
@@ -596,7 +596,7 @@
 
                                  LSTMCell
                                  
 
                                  
 
                                  
-class torch.nn.LSTMCell(input_size, hidden_size, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LSTMCell(input_size, hidden_size, bias=True, device=None, dtype=None)[source][source]
 
                                  A long short-term memory (LSTM) cell.
                                  
 
                                  
 i=σ(Wiix+bii+Whih+bhi)f=σ(Wifx+bif+Whfh+bhf)g=tanh⁡(Wigx+big+Whgh+bhg)o=σ(Wiox+bio+Whoh+bho)c′=f⊙c+i⊙gh′=o⊙tanh⁡(c′)\begin{array}{ll}
diff --git a/2.7/generated/torch.nn.LayerNorm.html b/2.7/generated/torch.nn.LayerNorm.html
index fc325982648..3da3a11f6c8 100644
--- a/2.7/generated/torch.nn.LayerNorm.html
+++ b/2.7/generated/torch.nn.LayerNorm.html
@@ -596,7 +596,7 @@
 
                                  LayerNorm
                                  
 
                                  
 
                                  
-class torch.nn.LayerNorm(normalized_shape, eps=1e-05, elementwise_affine=True, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LayerNorm(normalized_shape, eps=1e-05, elementwise_affine=True, bias=True, device=None, dtype=None)[source][source]
 
                                  Applies Layer Normalization over a mini-batch of inputs.
                                  
 
                                  This layer implements the operation as described in
 the paper Layer Normalization
                                  
diff --git a/2.7/generated/torch.nn.LazyBatchNorm1d.html b/2.7/generated/torch.nn.LazyBatchNorm1d.html
index 0b96d6d5561..9a8a3acd101 100644
--- a/2.7/generated/torch.nn.LazyBatchNorm1d.html
+++ b/2.7/generated/torch.nn.LazyBatchNorm1d.html
@@ -596,7 +596,7 @@
 
                                  LazyBatchNorm1d
                                  
 
                                  
 
                                  
-class torch.nn.LazyBatchNorm1d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyBatchNorm1d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.BatchNorm1d module with lazy initialization.
                                  
 
                                  Lazy initialization based on the num_features argument of the BatchNorm1d that is inferred
 from the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyBatchNorm1d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of BatchNorm1d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyBatchNorm2d.html b/2.7/generated/torch.nn.LazyBatchNorm2d.html
index 6ede3e31da0..48a0c4a3536 100644
--- a/2.7/generated/torch.nn.LazyBatchNorm2d.html
+++ b/2.7/generated/torch.nn.LazyBatchNorm2d.html
@@ -596,7 +596,7 @@
 
                                  LazyBatchNorm2d
                                  
 
                                  
 
                                  
-class torch.nn.LazyBatchNorm2d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyBatchNorm2d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.BatchNorm2d module with lazy initialization.
                                  
 
                                  Lazy initialization is done for the num_features argument of the BatchNorm2d that is inferred
 from the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyBatchNorm2d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of BatchNorm2d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyBatchNorm3d.html b/2.7/generated/torch.nn.LazyBatchNorm3d.html
index ab9e2bbb401..1d467944704 100644
--- a/2.7/generated/torch.nn.LazyBatchNorm3d.html
+++ b/2.7/generated/torch.nn.LazyBatchNorm3d.html
@@ -596,7 +596,7 @@
 
                                  LazyBatchNorm3d
                                  
 
                                  
 
                                  
-class torch.nn.LazyBatchNorm3d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyBatchNorm3d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.BatchNorm3d module with lazy initialization.
                                  
 
                                  Lazy initialization is done for the num_features argument of the BatchNorm3d that is inferred
 from the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyBatchNorm3d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of BatchNorm3d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConv1d.html b/2.7/generated/torch.nn.LazyConv1d.html
index ac875830aa9..39fbd6bf9f1 100644
--- a/2.7/generated/torch.nn.LazyConv1d.html
+++ b/2.7/generated/torch.nn.LazyConv1d.html
@@ -596,7 +596,7 @@
 
                                  LazyConv1d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConv1d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConv1d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.Conv1d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the Conv1d is inferred from the input.size(1).
 The attributes that will be lazily initialized are weight and bias.
                                  
@@ -627,7 +627,7 @@ 
                                  LazyConv1d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of Conv1d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConv2d.html b/2.7/generated/torch.nn.LazyConv2d.html
index a06ea8831e4..fb737699c02 100644
--- a/2.7/generated/torch.nn.LazyConv2d.html
+++ b/2.7/generated/torch.nn.LazyConv2d.html
@@ -596,7 +596,7 @@
 
                                  LazyConv2d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConv2d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConv2d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.Conv2d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the Conv2d that is inferred from the input.size(1).
 The attributes that will be lazily initialized are weight and bias.
                                  
@@ -627,7 +627,7 @@ 
                                  LazyConv2d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of Conv2d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConv3d.html b/2.7/generated/torch.nn.LazyConv3d.html
index f0db2bfd9e4..e5054895a72 100644
--- a/2.7/generated/torch.nn.LazyConv3d.html
+++ b/2.7/generated/torch.nn.LazyConv3d.html
@@ -596,7 +596,7 @@
 
                                  LazyConv3d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConv3d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConv3d(out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.Conv3d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the Conv3d that is inferred from
 the input.size(1).
@@ -628,7 +628,7 @@ 
                                  LazyConv3d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of Conv3d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConvTranspose1d.html b/2.7/generated/torch.nn.LazyConvTranspose1d.html
index 102c124a2e4..6fa6e572593 100644
--- a/2.7/generated/torch.nn.LazyConvTranspose1d.html
+++ b/2.7/generated/torch.nn.LazyConvTranspose1d.html
@@ -596,7 +596,7 @@
 
                                  LazyConvTranspose1d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConvTranspose1d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConvTranspose1d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.ConvTranspose1d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the ConvTranspose1d that is inferred from
 the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyConvTranspose1d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of ConvTranspose1d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConvTranspose2d.html b/2.7/generated/torch.nn.LazyConvTranspose2d.html
index 1b66a6da1fb..8766a86ffac 100644
--- a/2.7/generated/torch.nn.LazyConvTranspose2d.html
+++ b/2.7/generated/torch.nn.LazyConvTranspose2d.html
@@ -596,7 +596,7 @@
 
                                  LazyConvTranspose2d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConvTranspose2d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConvTranspose2d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.ConvTranspose2d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the ConvTranspose2d is inferred from
 the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyConvTranspose2d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of ConvTranspose2d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyConvTranspose3d.html b/2.7/generated/torch.nn.LazyConvTranspose3d.html
index 31150e666cd..8a89e18fb75 100644
--- a/2.7/generated/torch.nn.LazyConvTranspose3d.html
+++ b/2.7/generated/torch.nn.LazyConvTranspose3d.html
@@ -596,7 +596,7 @@
 
                                  LazyConvTranspose3d
                                  
 
                                  
 
                                  
-class torch.nn.LazyConvTranspose3d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyConvTranspose3d(out_channels, kernel_size, stride=1, padding=0, output_padding=0, groups=1, bias=True, dilation=1, padding_mode='zeros', device=None, dtype=None)[source][source]
 
                                  A torch.nn.ConvTranspose3d module with lazy initialization of the in_channels argument.
                                  
 
                                  The in_channels argument of the ConvTranspose3d is inferred from
 the input.size(1).
@@ -625,7 +625,7 @@ 
                                  LazyConvTranspose3d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of ConvTranspose3d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyInstanceNorm1d.html b/2.7/generated/torch.nn.LazyInstanceNorm1d.html
index 5da9526b03d..02c103984b8 100644
--- a/2.7/generated/torch.nn.LazyInstanceNorm1d.html
+++ b/2.7/generated/torch.nn.LazyInstanceNorm1d.html
@@ -596,7 +596,7 @@
 
                                  LazyInstanceNorm1d
                                  
 
                                  
 
                                  
-class torch.nn.LazyInstanceNorm1d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyInstanceNorm1d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.InstanceNorm1d module with lazy initialization of the num_features argument.
                                  
 
                                  The num_features argument of the InstanceNorm1d is inferred from the input.size(1).
 The attributes that will be lazily initialized are weight, bias, running_mean and running_var.
                                  
@@ -628,7 +628,7 @@ 
                                  LazyInstanceNorm1d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of InstanceNorm1d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyInstanceNorm2d.html b/2.7/generated/torch.nn.LazyInstanceNorm2d.html
index bd27465eaff..884cc461c59 100644
--- a/2.7/generated/torch.nn.LazyInstanceNorm2d.html
+++ b/2.7/generated/torch.nn.LazyInstanceNorm2d.html
@@ -596,7 +596,7 @@
 
                                  LazyInstanceNorm2d
                                  
 
                                  
 
                                  
-class torch.nn.LazyInstanceNorm2d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyInstanceNorm2d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.InstanceNorm2d module with lazy initialization of the num_features argument.
                                  
 
                                  The num_features argument of the InstanceNorm2d is inferred from the input.size(1).
 The attributes that will be lazily initialized are weight, bias,
@@ -629,7 +629,7 @@ 
                                  LazyInstanceNorm2d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of InstanceNorm2d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyInstanceNorm3d.html b/2.7/generated/torch.nn.LazyInstanceNorm3d.html
index 8238e5490d2..bb5fecae649 100644
--- a/2.7/generated/torch.nn.LazyInstanceNorm3d.html
+++ b/2.7/generated/torch.nn.LazyInstanceNorm3d.html
@@ -596,7 +596,7 @@
 
                                  LazyInstanceNorm3d
                                  
 
                                  
 
                                  
-class torch.nn.LazyInstanceNorm3d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyInstanceNorm3d(eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.InstanceNorm3d module with lazy initialization of the num_features argument.
                                  
 
                                  The num_features argument of the InstanceNorm3d is inferred from the input.size(1).
 The attributes that will be lazily initialized are weight, bias,
@@ -629,7 +629,7 @@ 
                                  LazyInstanceNorm3d
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of InstanceNorm3d
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LazyLinear.html b/2.7/generated/torch.nn.LazyLinear.html
index 9971f8f6b33..ea76a97869e 100644
--- a/2.7/generated/torch.nn.LazyLinear.html
+++ b/2.7/generated/torch.nn.LazyLinear.html
@@ -596,7 +596,7 @@
 
                                  LazyLinear
                                  
 
                                  
 
                                  
-class torch.nn.LazyLinear(out_features, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.LazyLinear(out_features, bias=True, device=None, dtype=None)[source][source]
 
                                  A torch.nn.Linear module where in_features is inferred.
                                  
 
                                  In this module, the weight and bias are of torch.nn.UninitializedParameter
 class. They will be initialized after the first call to forward is done and the
@@ -671,7 +671,7 @@ 
                                  LazyLinear
 
                                  
-cls_to_become[source]
                                  
+cls_to_become[source]
 
                                  alias of Linear
                                  
 
                                  
 
diff --git a/2.7/generated/torch.nn.LeakyReLU.html b/2.7/generated/torch.nn.LeakyReLU.html
index 047d7d1adb0..d6aa5521371 100644
--- a/2.7/generated/torch.nn.LeakyReLU.html
+++ b/2.7/generated/torch.nn.LeakyReLU.html
@@ -596,7 +596,7 @@
 
                                  LeakyReLU
                                  
 
                                  
 
                                  
-class torch.nn.LeakyReLU(negative_slope=0.01, inplace=False)[source][source]
                                  
+class torch.nn.LeakyReLU(negative_slope=0.01, inplace=False)[source][source]
 
                                  Applies the LeakyReLU function element-wise.
                                  
 
                                  
 LeakyReLU(x)=max⁡(0,x)+negative_slope∗min⁡(0,x)\text{LeakyReLU}(x) = \max(0, x) + \text{negative\_slope} * \min(0, x)
diff --git a/2.7/generated/torch.nn.Linear.html b/2.7/generated/torch.nn.Linear.html
index 7b5d10dad4c..54f024cb785 100644
--- a/2.7/generated/torch.nn.Linear.html
+++ b/2.7/generated/torch.nn.Linear.html
@@ -596,7 +596,7 @@
 
                                  Linear
                                  
 
                                  
 
                                  
-class torch.nn.Linear(in_features, out_features, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.Linear(in_features, out_features, bias=True, device=None, dtype=None)[source][source]
 
                                  Applies an affine linear transformation to the incoming data: y=xAT+by = xA^T + b.
                                  
 
                                  This module supports TensorFloat32.
                                  
 
                                  On certain ROCm devices, when using float16 inputs this module will use different precision for backward.
                                  
diff --git a/2.7/generated/torch.nn.LocalResponseNorm.html b/2.7/generated/torch.nn.LocalResponseNorm.html
index db0f2403adf..229bce8b8be 100644
--- a/2.7/generated/torch.nn.LocalResponseNorm.html
+++ b/2.7/generated/torch.nn.LocalResponseNorm.html
@@ -596,7 +596,7 @@
 
                                  LocalResponseNorm
                                  
 
                                  
 
                                  
-class torch.nn.LocalResponseNorm(size, alpha=0.0001, beta=0.75, k=1.0)[source][source]
                                  
+class torch.nn.LocalResponseNorm(size, alpha=0.0001, beta=0.75, k=1.0)[source][source]
 
                                  Applies local response normalization over an input signal.
                                  
 
                                  The input signal is composed of several input planes, where channels occupy the second dimension.
 Applies normalization across channels.
                                  
diff --git a/2.7/generated/torch.nn.LogSigmoid.html b/2.7/generated/torch.nn.LogSigmoid.html
index 01dd25ff6a2..d91a1301f4a 100644
--- a/2.7/generated/torch.nn.LogSigmoid.html
+++ b/2.7/generated/torch.nn.LogSigmoid.html
@@ -596,7 +596,7 @@
 
                                  LogSigmoid
                                  
 
                                  
 
                                  
-class torch.nn.LogSigmoid(*args, **kwargs)[source][source]
                                  
+class torch.nn.LogSigmoid(*args, **kwargs)[source][source]
 
                                  Applies the Logsigmoid function element-wise.
                                  
 
                                  
 LogSigmoid(x)=log⁡(11+exp⁡(−x))\text{LogSigmoid}(x) = \log\left(\frac{ 1 }{ 1 + \exp(-x)}\right)
diff --git a/2.7/generated/torch.nn.LogSoftmax.html b/2.7/generated/torch.nn.LogSoftmax.html
index bfa330a1f6f..76c931e629b 100644
--- a/2.7/generated/torch.nn.LogSoftmax.html
+++ b/2.7/generated/torch.nn.LogSoftmax.html
@@ -596,7 +596,7 @@
 
                                  LogSoftmax
                                  
 
                                  
 
                                  
-class torch.nn.LogSoftmax(dim=None)[source][source]
                                  
+class torch.nn.LogSoftmax(dim=None)[source][source]
 
                                  Applies the log⁡(Softmax(x))\log(\text{Softmax}(x)) function to an n-dimensional input Tensor.
                                  
 
                                  The LogSoftmax formulation can be simplified as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.MSELoss.html b/2.7/generated/torch.nn.MSELoss.html
index 3f7b5f0ff4e..9e79275ca21 100644
--- a/2.7/generated/torch.nn.MSELoss.html
+++ b/2.7/generated/torch.nn.MSELoss.html
@@ -596,7 +596,7 @@
 
                                  MSELoss
                                  
 
                                  
 
                                  
-class torch.nn.MSELoss(size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.MSELoss(size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the mean squared error (squared L2 norm) between
 each element in the input xx and target yy.
                                  
 
                                  The unreduced (i.e. with reduction set to 'none') loss can be described as:
                                  
diff --git a/2.7/generated/torch.nn.MarginRankingLoss.html b/2.7/generated/torch.nn.MarginRankingLoss.html
index 846dd25f36f..4e0f3998c6d 100644
--- a/2.7/generated/torch.nn.MarginRankingLoss.html
+++ b/2.7/generated/torch.nn.MarginRankingLoss.html
@@ -596,7 +596,7 @@
 
                                  MarginRankingLoss
                                  
 
                                  
 
                                  
-class torch.nn.MarginRankingLoss(margin=0.0, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.MarginRankingLoss(margin=0.0, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the loss given
 inputs x1x1, x2x2, two 1D mini-batch or 0D Tensors,
 and a label 1D mini-batch or 0D Tensor yy (containing 1 or -1).
                                  
diff --git a/2.7/generated/torch.nn.MaxPool1d.html b/2.7/generated/torch.nn.MaxPool1d.html
index 1b0c2430591..261d5cf9b29 100644
--- a/2.7/generated/torch.nn.MaxPool1d.html
+++ b/2.7/generated/torch.nn.MaxPool1d.html
@@ -596,7 +596,7 @@
 
                                  MaxPool1d
                                  
 
                                  
 
                                  
-class torch.nn.MaxPool1d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
                                  
+class torch.nn.MaxPool1d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
 
                                  Applies a 1D max pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,L)(N, C, L)
 and output (N,C,Lout)(N, C, L_{out}) can be precisely described as:
                                  
diff --git a/2.7/generated/torch.nn.MaxPool2d.html b/2.7/generated/torch.nn.MaxPool2d.html
index ba176b236a9..9e3abeb7e4f 100644
--- a/2.7/generated/torch.nn.MaxPool2d.html
+++ b/2.7/generated/torch.nn.MaxPool2d.html
@@ -596,7 +596,7 @@
 
                                  MaxPool2d
                                  
 
                                  
 
                                  
-class torch.nn.MaxPool2d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
                                  
+class torch.nn.MaxPool2d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
 
                                  Applies a 2D max pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,H,W)(N, C, H, W),
 output (N,C,Hout,Wout)(N, C, H_{out}, W_{out}) and kernel_size (kH,kW)(kH, kW)
diff --git a/2.7/generated/torch.nn.MaxPool3d.html b/2.7/generated/torch.nn.MaxPool3d.html
index 384f4dfc61b..d17a06571f8 100644
--- a/2.7/generated/torch.nn.MaxPool3d.html
+++ b/2.7/generated/torch.nn.MaxPool3d.html
@@ -596,7 +596,7 @@
 
                                  MaxPool3d
                                  
 
                                  
 
                                  
-class torch.nn.MaxPool3d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
                                  
+class torch.nn.MaxPool3d(kernel_size, stride=None, padding=0, dilation=1, return_indices=False, ceil_mode=False)[source][source]
 
                                  Applies a 3D max pooling over an input signal composed of several input planes.
                                  
 
                                  In the simplest case, the output value of the layer with input size (N,C,D,H,W)(N, C, D, H, W),
 output (N,C,Dout,Hout,Wout)(N, C, D_{out}, H_{out}, W_{out}) and kernel_size (kD,kH,kW)(kD, kH, kW)
diff --git a/2.7/generated/torch.nn.MaxUnpool1d.html b/2.7/generated/torch.nn.MaxUnpool1d.html
index 1b83b42581f..d3537312e04 100644
--- a/2.7/generated/torch.nn.MaxUnpool1d.html
+++ b/2.7/generated/torch.nn.MaxUnpool1d.html
@@ -596,7 +596,7 @@
 
                                  MaxUnpool1d
                                  
 
                                  
 
                                  
-class torch.nn.MaxUnpool1d(kernel_size, stride=None, padding=0)[source][source]
                                  
+class torch.nn.MaxUnpool1d(kernel_size, stride=None, padding=0)[source][source]
 
                                  Computes a partial inverse of MaxPool1d.
                                  
 
                                  MaxPool1d is not fully invertible, since the non-maximal values are lost.
                                  
 
                                  MaxUnpool1d takes in as input the output of MaxPool1d
diff --git a/2.7/generated/torch.nn.MaxUnpool2d.html b/2.7/generated/torch.nn.MaxUnpool2d.html
index 4144f1c4440..b5c4478cbf5 100644
--- a/2.7/generated/torch.nn.MaxUnpool2d.html
+++ b/2.7/generated/torch.nn.MaxUnpool2d.html
@@ -596,7 +596,7 @@
 
                                  MaxUnpool2d
                                  
 
                                  
 
                                  
-class torch.nn.MaxUnpool2d(kernel_size, stride=None, padding=0)[source][source]
                                  
+class torch.nn.MaxUnpool2d(kernel_size, stride=None, padding=0)[source][source]
 
                                  Computes a partial inverse of MaxPool2d.
                                  
 
                                  MaxPool2d is not fully invertible, since the non-maximal values are lost.
                                  
 
                                  MaxUnpool2d takes in as input the output of MaxPool2d
diff --git a/2.7/generated/torch.nn.MaxUnpool3d.html b/2.7/generated/torch.nn.MaxUnpool3d.html
index 635aecbc96c..2fd1a4af337 100644
--- a/2.7/generated/torch.nn.MaxUnpool3d.html
+++ b/2.7/generated/torch.nn.MaxUnpool3d.html
@@ -596,7 +596,7 @@
 
                                  MaxUnpool3d
                                  
 
                                  
 
                                  
-class torch.nn.MaxUnpool3d(kernel_size, stride=None, padding=0)[source][source]
                                  
+class torch.nn.MaxUnpool3d(kernel_size, stride=None, padding=0)[source][source]
 
                                  Computes a partial inverse of MaxPool3d.
                                  
 
                                  MaxPool3d is not fully invertible, since the non-maximal values are lost.
 MaxUnpool3d takes in as input the output of MaxPool3d
diff --git a/2.7/generated/torch.nn.Mish.html b/2.7/generated/torch.nn.Mish.html
index d48a858dff1..f20d9c8df37 100644
--- a/2.7/generated/torch.nn.Mish.html
+++ b/2.7/generated/torch.nn.Mish.html
@@ -596,7 +596,7 @@
 
                                  Mish
                                  
 
                                  
 
                                  
-class torch.nn.Mish(inplace=False)[source][source]
                                  
+class torch.nn.Mish(inplace=False)[source][source]
 
                                  Applies the Mish function, element-wise.
                                  
 
                                  Mish: A Self Regularized Non-Monotonic Neural Activation Function.
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Module.html b/2.7/generated/torch.nn.Module.html
index 3d5889673ac..d188117d349 100644
--- a/2.7/generated/torch.nn.Module.html
+++ b/2.7/generated/torch.nn.Module.html
@@ -596,7 +596,7 @@
 
                                  Module
                                  
 
                                  
 
                                  
-class torch.nn.Module(*args, **kwargs)[source][source]
                                  
+class torch.nn.Module(*args, **kwargs)[source][source]
 
                                  Base class for all neural network modules.
                                  
 
                                  Your models should also subclass this class.
                                  
 
                                  Modules can also contain other Modules, allowing them to be nested in
@@ -630,7 +630,7 @@ 
                                  Module
 
                                  
 
                                  
-add_module(name, module)[source][source]
                                  
+add_module(name, module)[source][source]
 
                                  Add a child module to the current module.
                                  
 
                                  The module can be accessed as an attribute using the given name.
                                  
 
                                  
@@ -646,7 +646,7 @@ 
                                  Module
 
                                  
-apply(fn)[source][source]
                                  
+apply(fn)[source][source]
 
                                  Apply fn recursively to every submodule (as returned by .children()) as well as self.
                                  
 
                                  Typical use includes initializing the parameters of a model
 (see also torch.nn.init).
                                  
@@ -688,7 +688,7 @@ 
                                  Module
 
                                  
-bfloat16()[source][source]
                                  
+bfloat16()[source][source]
 
                                  Casts all floating point parameters and buffers to bfloat16 datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -706,7 +706,7 @@ 
                                  Module
 
                                  
-buffers(recurse=True)[source][source]
                                  
+buffers(recurse=True)[source][source]
 
                                  Return an iterator over module buffers.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -732,7 +732,7 @@ 
                                  Module
 
                                  
-children()[source][source]
                                  
+children()[source][source]
 
                                  Return an iterator over immediate children modules.
                                  
 
                                  
 
                                  Yields
                                  
@@ -746,7 +746,7 @@ 
                                  Module
 
                                  
-compile(*args, **kwargs)[source][source]
                                  
+compile(*args, **kwargs)[source][source]
 
                                  Compile this Module’s forward using torch.compile().
                                  
 
                                  This Module’s __call__ method is compiled and all arguments are passed as-is
 to torch.compile().
                                  
@@ -755,7 +755,7 @@ 
                                  Module
 
                                  
-cpu()[source][source]
                                  
+cpu()[source][source]
 
                                  Move all model parameters and buffers to the CPU.
                                  
 
                                  
 
                                  Note
                                  
@@ -773,7 +773,7 @@ 
                                  Module
 
                                  
-cuda(device=None)[source][source]
                                  
+cuda(device=None)[source][source]
 
                                  Move all model parameters and buffers to the GPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -798,7 +798,7 @@ 
                                  Module
 
                                  
-double()[source][source]
                                  
+double()[source][source]
 
                                  Casts all floating point parameters and buffers to double datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -816,7 +816,7 @@ 
                                  Module
 
                                  
-eval()[source][source]
                                  
+eval()[source][source]
 
                                  Set the module in evaluation mode.
                                  
 
                                  This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -837,7 +837,7 @@ 
                                  Module
 
                                  
-extra_repr()[source][source]
                                  
+extra_repr()[source][source]
 
                                  Return the extra representation of the module.
                                  
 
                                  To print customized extra information, you should re-implement
 this method in your own modules. Both single-line and multi-line
@@ -851,7 +851,7 @@ 
                                  Module
 
                                  
-float()[source][source]
                                  
+float()[source][source]
 
                                  Casts all floating point parameters and buffers to float datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -869,7 +869,7 @@ 
                                  Module
 
                                  
-forward(*input)[source]
                                  
+forward(*input)[source]
 
                                  Define the computation performed at every call.
                                  
 
                                  Should be overridden by all subclasses.
                                  
 
                                  
@@ -885,7 +885,7 @@ 
                                  Module
 
                                  
-get_buffer(target)[source][source]
                                  
+get_buffer(target)[source][source]
 
                                  Return the buffer given by target if it exists, otherwise throw an error.
                                  
 
                                  See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -912,7 +912,7 @@ 
                                  Module
 
                                  
-get_extra_state()[source][source]
                                  
+get_extra_state()[source][source]
 
                                  Return any extra state to include in the module’s state_dict.
                                  
 
                                  Implement this and a corresponding set_extra_state() for your module
 if you need to store extra state. This function is called when building the
@@ -933,7 +933,7 @@ 
                                  Module
 
                                  
-get_parameter(target)[source][source]
                                  
+get_parameter(target)[source][source]
 
                                  Return the parameter given by target if it exists, otherwise throw an error.
                                  
 
                                  See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -960,7 +960,7 @@ 
                                  Module
 
                                  
-get_submodule(target)[source][source]
                                  
+get_submodule(target)[source][source]
 
                                  Return the submodule given by target if it exists, otherwise throw an error.
                                  
 
                                  For example, let’s say you have an nn.Module A that
 looks like this:
                                  
@@ -1009,7 +1009,7 @@ 
                                  Module
 
                                  
-half()[source][source]
                                  
+half()[source][source]
 
                                  Casts all floating point parameters and buffers to half datatype.
                                  
 
                                  
 
                                  Note
                                  
@@ -1027,7 +1027,7 @@ 
                                  Module
 
                                  
-ipu(device=None)[source][source]
                                  
+ipu(device=None)[source][source]
 
                                  Move all model parameters and buffers to the IPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1052,7 +1052,7 @@ 
                                  Module
 
                                  
-load_state_dict(state_dict, strict=True, assign=False)[source][source]
                                  
+load_state_dict(state_dict, strict=True, assign=False)[source][source]
 
                                  Copy parameters and buffers from state_dict into this module and its descendants.
                                  
 
                                  If strict is True, then
 the keys of state_dict must exactly match the keys returned
@@ -1107,7 +1107,7 @@ 
                                  Module
 
                                  
-modules()[source][source]
                                  
+modules()[source][source]
 
                                  Return an iterator over all modules in the network.
                                  
 
                                  
 
                                  Yields
                                  
@@ -1139,7 +1139,7 @@ 
                                  Module
 
                                  
-mtia(device=None)[source][source]
                                  
+mtia(device=None)[source][source]
 
                                  Move all model parameters and buffers to the MTIA.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1164,7 +1164,7 @@ 
                                  Module
 
                                  
-named_buffers(prefix='', recurse=True, remove_duplicate=True)[source][source]
                                  
+named_buffers(prefix='', recurse=True, remove_duplicate=True)[source][source]
 
                                  Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1193,7 +1193,7 @@ 
                                  Module
 
                                  
-named_children()[source][source]
                                  
+named_children()[source][source]
 
                                  Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
                                  
 
                                  
 
                                  Yields
                                  
@@ -1213,7 +1213,7 @@ 
                                  Module
 
                                  
-named_modules(memo=None, prefix='', remove_duplicate=True)[source][source]
                                  
+named_modules(memo=None, prefix='', remove_duplicate=True)[source][source]
 
                                  Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1250,7 +1250,7 @@ 
                                  Module
 
                                  
-named_parameters(prefix='', recurse=True, remove_duplicate=True)[source][source]
                                  
+named_parameters(prefix='', recurse=True, remove_duplicate=True)[source][source]
 
                                  Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1280,7 +1280,7 @@ 
                                  Module
 
                                  
-parameters(recurse=True)[source][source]
                                  
+parameters(recurse=True)[source][source]
 
                                  Return an iterator over module parameters.
                                  
 
                                  This is typically passed to an optimizer.
                                  
 
                                  
@@ -1307,7 +1307,7 @@ 
                                  Module
 
                                  
-register_backward_hook(hook)[source][source]
                                  
+register_backward_hook(hook)[source][source]
 
                                  Register a backward hook on the module.
                                  
 
                                  This function is deprecated in favor of register_full_backward_hook() and
 the behavior of this function will change in future versions.
                                  
@@ -1324,7 +1324,7 @@ 
                                  Module
 
                                  
-register_buffer(name, tensor, persistent=True)[source][source]
                                  
+register_buffer(name, tensor, persistent=True)[source][source]
 
                                  Add a buffer to the module.
                                  
 
                                  This is typically used to register a buffer that should not to be
 considered a model parameter. For example, BatchNorm’s running_mean
@@ -1356,7 +1356,7 @@ 
                                  Module
 
                                  
-register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source][source]
                                  
+register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source][source]
 
                                  Register a forward hook on the module.
                                  
 
                                  The hook will be called every time after forward() has computed an output.
                                  
 
                                  If with_kwargs is False or not specified, the input contains only
@@ -1407,7 +1407,7 @@ 
                                  Module
 
                                  
-register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source][source]
                                  
+register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source][source]
 
                                  Register a forward pre-hook on the module.
                                  
 
                                  The hook will be called every time before forward() is invoked.
                                  
 
                                  If with_kwargs is false or not specified, the input contains only
@@ -1457,7 +1457,7 @@ 
                                  Module
 
                                  
-register_full_backward_hook(hook, prepend=False)[source][source]
                                  
+register_full_backward_hook(hook, prepend=False)[source][source]
 
                                  Register a backward hook on the module.
                                  
 
                                  The hook will be called every time the gradients with respect to a module
 are computed, i.e. the hook will execute if and only if the gradients with
@@ -1508,7 +1508,7 @@ 
                                  Module
 
                                  
-register_full_backward_pre_hook(hook, prepend=False)[source][source]
                                  
+register_full_backward_pre_hook(hook, prepend=False)[source][source]
 
                                  Register a backward pre-hook on the module.
                                  
 
                                  The hook will be called every time the gradients for the module are computed.
 The hook should have the following signature:
                                  
@@ -1554,7 +1554,7 @@ 
                                  Module
 
                                  
-register_load_state_dict_post_hook(hook)[source][source]
                                  
+register_load_state_dict_post_hook(hook)[source][source]
 
                                  Register a post-hook to be run after module’s load_state_dict() is called.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, incompatible_keys) -> None
                                  
@@ -1584,7 +1584,7 @@ 
                                  Module
 
                                  
-register_load_state_dict_pre_hook(hook)[source][source]
                                  
+register_load_state_dict_pre_hook(hook)[source][source]
 
                                  Register a pre-hook to be run before module’s load_state_dict() is called.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs) -> None  # noqa: B950
                                  
@@ -1600,7 +1600,7 @@ 
                                  Module
 
                                  
-register_module(name, module)[source][source]
                                  
+register_module(name, module)[source][source]
 
                                  Alias for add_module().
                                  
 
                                  
 
                                  
@@ -1608,7 +1608,7 @@ 
                                  Module
 
                                  
-register_parameter(name, param)[source][source]
                                  
+register_parameter(name, param)[source][source]
 
                                  Add a parameter to the module.
                                  
 
                                  The parameter can be accessed as an attribute using given name.
                                  
 
                                  
@@ -1627,7 +1627,7 @@ 
                                  Module
 
                                  
-register_state_dict_post_hook(hook)[source][source]
                                  
+register_state_dict_post_hook(hook)[source][source]
 
                                  Register a post-hook for the state_dict() method.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, state_dict, prefix, local_metadata) -> None
                                  
@@ -1638,7 +1638,7 @@ 
                                  Module
 
                                  
-register_state_dict_pre_hook(hook)[source][source]
                                  
+register_state_dict_pre_hook(hook)[source][source]
 
                                  Register a pre-hook for the state_dict() method.
                                  
 
                                  
 
                                  It should have the following signature::
                                  hook(module, prefix, keep_vars) -> None
                                  
@@ -1650,7 +1650,7 @@ 
                                  Module
 
                                  
-requires_grad_(requires_grad=True)[source][source]
                                  
+requires_grad_(requires_grad=True)[source][source]
 
                                  Change if autograd should record operations on parameters in this module.
                                  
 
                                  This method sets the parameters’ requires_grad attributes
 in-place.
                                  
@@ -1674,7 +1674,7 @@ 
                                  Module
 
                                  
-set_extra_state(state)[source][source]
                                  
+set_extra_state(state)[source][source]
 
                                  Set extra state contained in the loaded state_dict.
                                  
 
                                  This function is called from load_state_dict() to handle any extra state
 found within the state_dict. Implement this function and a corresponding
@@ -1689,7 +1689,7 @@ 
                                  Module
 
                                  
-set_submodule(target, module, strict=False)[source][source]
                                  
+set_submodule(target, module, strict=False)[source][source]
 
                                  Set the submodule given by target if it exists, otherwise throw an error.
                                  
 
                                  
 
                                  Note
                                  
@@ -1747,7 +1747,7 @@ 
                                  Module
 
                                  
-share_memory()[source][source]
                                  
+share_memory()[source][source]
 
                                  See torch.Tensor.share_memory_().
                                  
 
                                  
 
                                  Return type
                                  
@@ -1758,7 +1758,7 @@ 
                                  Module
 
                                  
-state_dict(*, destination: T_destination, prefix: str = '', keep_vars: bool = False) → T_destination[source][source]
                                  
+state_dict(*, destination: T_destination, prefix: str = '', keep_vars: bool = False) → T_destination[source][source]
 
                                  
 state_dict(*, prefix: str = '', keep_vars: bool = False) → dict[str, Any]
                                  
 
                                  Return a dictionary containing references to the whole state of the module.
                                  
@@ -1813,7 +1813,7 @@ 
                                  Module
 
                                  
-to(device: Optional[Union[str, device, int]] = ..., dtype: Optional[dtype] = ..., non_blocking: bool = ...) → Self[source][source]
                                  
+to(device: Optional[Union[str, device, int]] = ..., dtype: Optional[dtype] = ..., non_blocking: bool = ...) → Self[source][source]
 
                                  
 to(dtype: dtype, non_blocking: bool = ...) → Self
                                  
 
                                  
@@ -1822,22 +1822,22 @@ 
                                  ModuleThis can be called as
                                  
 
                                  
 
                                  
-to(device=None, dtype=None, non_blocking=False)[source][source]
                                  
+to(device=None, dtype=None, non_blocking=False)[source][source]
                                  
 
                                  
 
 
                                  
 
                                  
-to(dtype, non_blocking=False)[source][source]
                                  
+to(dtype, non_blocking=False)[source][source]
 
                                  
 
 
                                  
 
                                  
-to(tensor, non_blocking=False)[source][source]
                                  
+to(tensor, non_blocking=False)[source][source]
 
                                  
 
 
                                  
 
                                  
-to(memory_format=torch.channels_last)[source][source]
                                  
+to(memory_format=torch.channels_last)[source][source]
 
                                  
 
 
                                  Its signature is similar to torch.Tensor.to(), but only accepts
@@ -1916,7 +1916,7 @@ 
                                  Module
 
                                  
-to_empty(*, device, recurse=True)[source][source]
                                  
+to_empty(*, device, recurse=True)[source][source]
 
                                  Move the parameters and buffers to the specified device without copying storage.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -1938,7 +1938,7 @@ 
                                  Module
 
                                  
-train(mode=True)[source][source]
                                  
+train(mode=True)[source][source]
 
                                  Set the module in training mode.
                                  
 
                                  This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -1960,7 +1960,7 @@ 
                                  Module
 
                                  
-type(dst_type)[source][source]
                                  
+type(dst_type)[source][source]
 
                                  Casts all parameters and buffers to dst_type.
                                  
 
                                  
 
                                  Note
                                  
@@ -1981,7 +1981,7 @@ 
                                  Module
 
                                  
-xpu(device=None)[source][source]
                                  
+xpu(device=None)[source][source]
 
                                  Move all model parameters and buffers to the XPU.
                                  
 
                                  This also makes associated parameters and buffers different objects. So
 it should be called before constructing optimizer if the module will
@@ -2006,7 +2006,7 @@ 
                                  Module
 
                                  
-zero_grad(set_to_none=True)[source][source]
                                  
+zero_grad(set_to_none=True)[source][source]
 
                                  Reset gradients of all model parameters.
                                  
 
                                  See similar function under torch.optim.Optimizer for more context.
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ModuleDict.html b/2.7/generated/torch.nn.ModuleDict.html
index 000fb3f7f09..73c3846363f 100644
--- a/2.7/generated/torch.nn.ModuleDict.html
+++ b/2.7/generated/torch.nn.ModuleDict.html
@@ -596,7 +596,7 @@
 
                                  ModuleDict
                                  
 
                                  
 
                                  
-class torch.nn.ModuleDict(modules=None)[source][source]
                                  
+class torch.nn.ModuleDict(modules=None)[source][source]
 
                                  Holds submodules in a dictionary.
                                  
 
                                  ModuleDict can be indexed like a regular Python dictionary,
 but modules it contains are properly registered, and will be visible by all
@@ -639,7 +639,7 @@ 
                                  ModuleDict
 
                                  
-clear()[source][source]
                                  
+clear()[source][source]
 
                                  Remove all items from the ModuleDict.
                                  
 
                                  
 
                                  
@@ -647,7 +647,7 @@ 
                                  ModuleDict
 
                                  
-items()[source][source]
                                  
+items()[source][source]
 
                                  Return an iterable of the ModuleDict key/value pairs.
                                  
 
                                  
 
                                  Return type
                                  
@@ -658,7 +658,7 @@ 
                                  ModuleDict
 
                                  
-keys()[source][source]
                                  
+keys()[source][source]
 
                                  Return an iterable of the ModuleDict keys.
                                  
 
                                  
 
                                  Return type
                                  
@@ -669,7 +669,7 @@ 
                                  ModuleDict
 
                                  
-pop(key)[source][source]
                                  
+pop(key)[source][source]
 
                                  Remove key from the ModuleDict and return its module.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -683,7 +683,7 @@ 
                                  ModuleDict
 
                                  
-update(modules)[source][source]
                                  
+update(modules)[source][source]
 
                                  Update the ModuleDict with key-value pairs from a mapping, overwriting existing keys.
                                  
 
                                  
 
                                  Note
                                  
@@ -700,7 +700,7 @@ 
                                  ModuleDict
 
                                  
-values()[source][source]
                                  
+values()[source][source]
 
                                  Return an iterable of the ModuleDict values.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.nn.ModuleList.html b/2.7/generated/torch.nn.ModuleList.html
index 1edac189453..f9a063e794c 100644
--- a/2.7/generated/torch.nn.ModuleList.html
+++ b/2.7/generated/torch.nn.ModuleList.html
@@ -596,7 +596,7 @@
 
                                  ModuleList
                                  
 
                                  
 
                                  
-class torch.nn.ModuleList(modules=None)[source][source]
                                  
+class torch.nn.ModuleList(modules=None)[source][source]
 
                                  Holds submodules in a list.
                                  
 
                                  ModuleList can be indexed like a regular Python list, but
 modules it contains are properly registered, and will be visible by all
@@ -621,7 +621,7 @@ 
                                  ModuleList
 
                                  
-append(module)[source][source]
                                  
+append(module)[source][source]
 
                                  Append a given module to the end of the list.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -635,7 +635,7 @@ 
                                  ModuleList
 
                                  
-extend(modules)[source][source]
                                  
+extend(modules)[source][source]
 
                                  Append modules from a Python iterable to the end of the list.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -649,7 +649,7 @@ 
                                  ModuleList
 
                                  
-insert(index, module)[source][source]
                                  
+insert(index, module)[source][source]
 
                                  Insert a given module before a given index in the list.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.MultiLabelMarginLoss.html b/2.7/generated/torch.nn.MultiLabelMarginLoss.html
index 93796429ad2..f71f5213eb1 100644
--- a/2.7/generated/torch.nn.MultiLabelMarginLoss.html
+++ b/2.7/generated/torch.nn.MultiLabelMarginLoss.html
@@ -596,7 +596,7 @@
 
                                  MultiLabelMarginLoss
                                  
 
                                  
 
                                  
-class torch.nn.MultiLabelMarginLoss(size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.MultiLabelMarginLoss(size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that optimizes a multi-class multi-classification
 hinge loss (margin-based loss) between input xx (a 2D mini-batch Tensor)
 and output yy (which is a 2D Tensor of target class indices).
diff --git a/2.7/generated/torch.nn.MultiLabelSoftMarginLoss.html b/2.7/generated/torch.nn.MultiLabelSoftMarginLoss.html
index 525a60b2ec6..bd4cbb0a7e0 100644
--- a/2.7/generated/torch.nn.MultiLabelSoftMarginLoss.html
+++ b/2.7/generated/torch.nn.MultiLabelSoftMarginLoss.html
@@ -596,7 +596,7 @@
 
                                  MultiLabelSoftMarginLoss
                                  
 
                                  
 
                                  
-class torch.nn.MultiLabelSoftMarginLoss(weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.MultiLabelSoftMarginLoss(weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that optimizes a multi-label one-versus-all
 loss based on max-entropy, between input xx and target yy of size
 (N,C)(N, C).
diff --git a/2.7/generated/torch.nn.MultiMarginLoss.html b/2.7/generated/torch.nn.MultiMarginLoss.html
index 9b9a4b10a83..fae5d6a0e60 100644
--- a/2.7/generated/torch.nn.MultiMarginLoss.html
+++ b/2.7/generated/torch.nn.MultiMarginLoss.html
@@ -596,7 +596,7 @@
 
                                  MultiMarginLoss
                                  
 
                                  
 
                                  
-class torch.nn.MultiMarginLoss(p=1, margin=1.0, weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.MultiMarginLoss(p=1, margin=1.0, weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that optimizes a multi-class classification hinge
 loss (margin-based loss) between input xx (a 2D mini-batch Tensor) and
 output yy (which is a 1D tensor of target class indices,
diff --git a/2.7/generated/torch.nn.MultiheadAttention.html b/2.7/generated/torch.nn.MultiheadAttention.html
index 77804b0e2f3..f059eaa4ebd 100644
--- a/2.7/generated/torch.nn.MultiheadAttention.html
+++ b/2.7/generated/torch.nn.MultiheadAttention.html
@@ -596,7 +596,7 @@
 
                                  MultiheadAttention
                                  
 
                                  
 
                                  
-class torch.nn.MultiheadAttention(embed_dim, num_heads, dropout=0.0, bias=True, add_bias_kv=False, add_zero_attn=False, kdim=None, vdim=None, batch_first=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.MultiheadAttention(embed_dim, num_heads, dropout=0.0, bias=True, add_bias_kv=False, add_zero_attn=False, kdim=None, vdim=None, batch_first=False, device=None, dtype=None)[source][source]
 
                                  Allows the model to jointly attend to information from different representation subspaces.
                                  
 
                                  
 
                                  Note
                                  
@@ -659,7 +659,7 @@ 
                                  MultiheadAttention
 
                                  
-forward(query, key, value, key_padding_mask=None, need_weights=True, attn_mask=None, average_attn_weights=True, is_causal=False)[source][source]
                                  
+forward(query, key, value, key_padding_mask=None, need_weights=True, attn_mask=None, average_attn_weights=True, is_causal=False)[source][source]
 
                                  Compute attention outputs using query, key, and value embeddings.
                                  
 
                                  
 
                                  Supports optional parameters for padding, masks and attention weights.
                                  
@@ -735,7 +735,7 @@ 
                                  MultiheadAttention
 
                                  
-merge_masks(attn_mask, key_padding_mask, query)[source][source]
                                  
+merge_masks(attn_mask, key_padding_mask, query)[source][source]
 
                                  Determine mask type and combine masks if necessary.
                                  
 
                                  If only one mask is provided, that mask
 and the corresponding mask type will be returned. If both masks are provided, they will be both
diff --git a/2.7/generated/torch.nn.NLLLoss.html b/2.7/generated/torch.nn.NLLLoss.html
index 35c0fd3ef40..60c5e191408 100644
--- a/2.7/generated/torch.nn.NLLLoss.html
+++ b/2.7/generated/torch.nn.NLLLoss.html
@@ -596,7 +596,7 @@
 
                                  NLLLoss
                                  
 
                                  
 
                                  
-class torch.nn.NLLLoss(weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.NLLLoss(weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean')[source][source]
 
                                  The negative log likelihood loss. It is useful to train a classification
 problem with C classes.
                                  
 
                                  If provided, the optional argument weight should be a 1D Tensor assigning
diff --git a/2.7/generated/torch.nn.PReLU.html b/2.7/generated/torch.nn.PReLU.html
index 08db33d67b9..027c06b8791 100644
--- a/2.7/generated/torch.nn.PReLU.html
+++ b/2.7/generated/torch.nn.PReLU.html
@@ -596,7 +596,7 @@
 
                                  PReLU
                                  
 
                                  
 
                                  
-class torch.nn.PReLU(num_parameters=1, init=0.25, device=None, dtype=None)[source][source]
                                  
+class torch.nn.PReLU(num_parameters=1, init=0.25, device=None, dtype=None)[source][source]
 
                                  Applies the element-wise PReLU function.
                                  
 
                                  
 PReLU(x)=max⁡(0,x)+a∗min⁡(0,x)\text{PReLU}(x) = \max(0,x) + a * \min(0,x)
diff --git a/2.7/generated/torch.nn.PairwiseDistance.html b/2.7/generated/torch.nn.PairwiseDistance.html
index e618926ab8f..bd9c81c1bdf 100644
--- a/2.7/generated/torch.nn.PairwiseDistance.html
+++ b/2.7/generated/torch.nn.PairwiseDistance.html
@@ -596,7 +596,7 @@
 
                                  PairwiseDistance
                                  
 
                                  
 
                                  
-class torch.nn.PairwiseDistance(p=2.0, eps=1e-06, keepdim=False)[source][source]
                                  
+class torch.nn.PairwiseDistance(p=2.0, eps=1e-06, keepdim=False)[source][source]
 
                                  Computes the pairwise distance between input vectors, or between columns of input matrices.
                                  
 
                                  Distances are computed using p-norm, with constant eps added to avoid division by zero
 if p is negative, i.e.:
                                  
diff --git a/2.7/generated/torch.nn.ParameterDict.html b/2.7/generated/torch.nn.ParameterDict.html
index 00ddfc6ba6e..836d82bf20f 100644
--- a/2.7/generated/torch.nn.ParameterDict.html
+++ b/2.7/generated/torch.nn.ParameterDict.html
@@ -596,7 +596,7 @@
 
                                  ParameterDict
                                  
 
                                  
 
                                  
-class torch.nn.ParameterDict(parameters=None)[source][source]
                                  
+class torch.nn.ParameterDict(parameters=None)[source][source]
 
                                  Holds parameters in a dictionary.
                                  
 
                                  ParameterDict can be indexed like a regular Python dictionary, but Parameters it
 contains are properly registered, and will be visible by all Module methods.
@@ -632,7 +632,7 @@ 
                                  ParameterDict
 
                                  
-clear()[source][source]
                                  
+clear()[source][source]
 
                                  Remove all items from the ParameterDict.
                                  
 
                                  
 
                                  
@@ -640,7 +640,7 @@ 
                                  ParameterDict
 
                                  
-copy()[source][source]
                                  
+copy()[source][source]
 
                                  Return a copy of this ParameterDict instance.
                                  
 
                                  
 
                                  Return type
                                  
@@ -651,7 +651,7 @@ 
                                  ParameterDict
 
                                  
-fromkeys(keys, default=None)[source][source]
                                  
+fromkeys(keys, default=None)[source][source]
 
                                  Return a new ParameterDict with the keys provided.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -668,7 +668,7 @@ 
                                  ParameterDict
 
                                  
-get(key, default=None)[source][source]
                                  
+get(key, default=None)[source][source]
 
                                  Return the parameter associated with key if present. Otherwise return default if provided, None if not.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -685,7 +685,7 @@ 
                                  ParameterDict
 
                                  
-items()[source][source]
                                  
+items()[source][source]
 
                                  Return an iterable of the ParameterDict key/value pairs.
                                  
 
                                  
 
                                  Return type
                                  
@@ -696,7 +696,7 @@ 
                                  ParameterDict
 
                                  
-keys()[source][source]
                                  
+keys()[source][source]
 
                                  Return an iterable of the ParameterDict keys.
                                  
 
                                  
 
                                  Return type
                                  
@@ -707,7 +707,7 @@ 
                                  ParameterDict
 
                                  
-pop(key)[source][source]
                                  
+pop(key)[source][source]
 
                                  Remove key from the ParameterDict and return its parameter.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -721,7 +721,7 @@ 
                                  ParameterDict
 
                                  
-popitem()[source][source]
                                  
+popitem()[source][source]
 
                                  Remove and return the last inserted (key, parameter) pair from the ParameterDict.
                                  
 
                                  
 
                                  Return type
                                  
@@ -732,7 +732,7 @@ 
                                  ParameterDict
 
                                  
-setdefault(key, default=None)[source][source]
                                  
+setdefault(key, default=None)[source][source]
 
                                  Set the default for a key in the Parameterdict.
                                  
 
                                  If key is in the ParameterDict, return its value.
 If not, insert key with a parameter default and return default.
@@ -752,7 +752,7 @@ 
                                  ParameterDict
 
                                  
-update(parameters)[source][source]
                                  
+update(parameters)[source][source]
 
                                  Update the ParameterDict with key-value pairs from parameters, overwriting existing keys.
                                  
 
                                  
 
                                  Note
                                  
@@ -770,7 +770,7 @@ 
                                  ParameterDict
 
                                  
-values()[source][source]
                                  
+values()[source][source]
 
                                  Return an iterable of the ParameterDict values.
                                  
 
                                  
 
                                  Return type
                                  
diff --git a/2.7/generated/torch.nn.ParameterList.html b/2.7/generated/torch.nn.ParameterList.html
index 5332fa26a00..9f0bf4363b7 100644
--- a/2.7/generated/torch.nn.ParameterList.html
+++ b/2.7/generated/torch.nn.ParameterList.html
@@ -596,7 +596,7 @@
 
                                  ParameterList
                                  
 
                                  
 
                                  
-class torch.nn.ParameterList(values=None)[source][source]
                                  
+class torch.nn.ParameterList(values=None)[source][source]
 
                                  Holds parameters in a list.
                                  
 
                                  ParameterList can be used like a regular Python
 list, but Tensors that are Parameter are properly registered,
@@ -624,7 +624,7 @@ 
                                  ParameterList
 
                                  
-append(value)[source][source]
                                  
+append(value)[source][source]
 
                                  Append a given value at the end of the list.
                                  
 
                                  
 
                                  Parameters
                                  
@@ -638,7 +638,7 @@ 
                                  ParameterList
 
                                  
-extend(values)[source][source]
                                  
+extend(values)[source][source]
 
                                  Append values from a Python iterable to the end of the list.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.PixelShuffle.html b/2.7/generated/torch.nn.PixelShuffle.html
index 064217f4270..a8300708426 100644
--- a/2.7/generated/torch.nn.PixelShuffle.html
+++ b/2.7/generated/torch.nn.PixelShuffle.html
@@ -596,7 +596,7 @@
 
                                  PixelShuffle
                                  
 
                                  
 
                                  
-class torch.nn.PixelShuffle(upscale_factor)[source][source]
                                  
+class torch.nn.PixelShuffle(upscale_factor)[source][source]
 
                                  Rearrange elements in a tensor according to an upscaling factor.
                                  
 
                                  Rearranges elements in a tensor of shape (∗,C×r2,H,W)(*, C \times r^2, H, W)
 to a tensor of shape (∗,C,H×r,W×r)(*, C, H \times r, W \times r), where r is an upscale factor.
                                  
diff --git a/2.7/generated/torch.nn.PixelUnshuffle.html b/2.7/generated/torch.nn.PixelUnshuffle.html
index 450d0a7b67c..81b784ea4cc 100644
--- a/2.7/generated/torch.nn.PixelUnshuffle.html
+++ b/2.7/generated/torch.nn.PixelUnshuffle.html
@@ -596,7 +596,7 @@
 
                                  PixelUnshuffle
                                  
 
                                  
 
                                  
-class torch.nn.PixelUnshuffle(downscale_factor)[source][source]
                                  
+class torch.nn.PixelUnshuffle(downscale_factor)[source][source]
 
                                  Reverse the PixelShuffle operation.
                                  
 
                                  Reverses the PixelShuffle operation by rearranging elements
 in a tensor of shape (∗,C,H×r,W×r)(*, C, H \times r, W \times r) to a tensor of shape
diff --git a/2.7/generated/torch.nn.PoissonNLLLoss.html b/2.7/generated/torch.nn.PoissonNLLLoss.html
index 15aa24e635e..f03f7bc04b5 100644
--- a/2.7/generated/torch.nn.PoissonNLLLoss.html
+++ b/2.7/generated/torch.nn.PoissonNLLLoss.html
@@ -596,7 +596,7 @@
 
                                  PoissonNLLLoss
                                  
 
                                  
 
                                  
-class torch.nn.PoissonNLLLoss(log_input=True, full=False, size_average=None, eps=1e-08, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.PoissonNLLLoss(log_input=True, full=False, size_average=None, eps=1e-08, reduce=None, reduction='mean')[source][source]
 
                                  Negative log likelihood loss with Poisson distribution of target.
                                  
 
                                  The loss can be described as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.RMSNorm.html b/2.7/generated/torch.nn.RMSNorm.html
index 2a195963bc3..7a8bdc058c2 100644
--- a/2.7/generated/torch.nn.RMSNorm.html
+++ b/2.7/generated/torch.nn.RMSNorm.html
@@ -596,7 +596,7 @@
 
                                  RMSNorm
                                  
 
                                  
 
                                  
-class torch.nn.RMSNorm(normalized_shape, eps=None, elementwise_affine=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.RMSNorm(normalized_shape, eps=None, elementwise_affine=True, device=None, dtype=None)[source][source]
 
                                  Applies Root Mean Square Layer Normalization over a mini-batch of inputs.
                                  
 
                                  This layer implements the operation as described in
 the paper Root Mean Square Layer Normalization
                                  
@@ -645,7 +645,7 @@ 
                                  RMSNorm
 
                                  
-extra_repr()[source][source]
                                  
+extra_repr()[source][source]
 
                                  Extra information about the module.
                                  
 
                                  
 
                                  Return type
                                  
@@ -656,7 +656,7 @@ 
                                  RMSNorm
 
                                  
-forward(x)[source][source]
                                  
+forward(x)[source][source]
 
                                  Runs forward pass.
                                  
 
                                  
 
                                  Return type
                                  
@@ -667,7 +667,7 @@ 
                                  RMSNorm
 
                                  
-reset_parameters()[source][source]
                                  
+reset_parameters()[source][source]
 
                                  Resets parameters based on their initialization used in __init__.
                                  
 
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.RNN.html b/2.7/generated/torch.nn.RNN.html
index 49fa7ca1c87..b234b091118 100644
--- a/2.7/generated/torch.nn.RNN.html
+++ b/2.7/generated/torch.nn.RNN.html
@@ -596,7 +596,7 @@
 
                                  RNN
                                  
 
                                  
 
                                  
-class torch.nn.RNN(input_size, hidden_size, num_layers=1, nonlinearity='tanh', bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None)[source][source]
                                  
+class torch.nn.RNN(input_size, hidden_size, num_layers=1, nonlinearity='tanh', bias=True, batch_first=False, dropout=0.0, bidirectional=False, device=None, dtype=None)[source][source]
 
                                  Apply a multi-layer Elman RNN with tanh⁡\tanh or ReLU\text{ReLU}
 non-linearity to an input sequence. For each element in the input sequence,
 each layer computes the following function:
                                  
diff --git a/2.7/generated/torch.nn.RNNBase.html b/2.7/generated/torch.nn.RNNBase.html
index 1e4dc1b5bd2..12e13bd4c6b 100644
--- a/2.7/generated/torch.nn.RNNBase.html
+++ b/2.7/generated/torch.nn.RNNBase.html
@@ -596,7 +596,7 @@
 
                                  RNNBase
                                  
 
                                  
 
                                  
-class torch.nn.RNNBase(mode, input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, proj_size=0, device=None, dtype=None)[source][source]
                                  
+class torch.nn.RNNBase(mode, input_size, hidden_size, num_layers=1, bias=True, batch_first=False, dropout=0.0, bidirectional=False, proj_size=0, device=None, dtype=None)[source][source]
 
                                  Base class for RNN modules (RNN, LSTM, GRU).
                                  
 
                                  Implements aspects of RNNs shared by the RNN, LSTM, and GRU classes, such as module initialization
 and utility methods for parameter storage management.
                                  
@@ -612,7 +612,7 @@ 
                                  RNNBase
 
                                  
-flatten_parameters()[source][source]
                                  
+flatten_parameters()[source][source]
 
                                  Reset parameter data pointer so that they can use faster code paths.
                                  
 
                                  Right now, this works only if the module is on the GPU and cuDNN is enabled.
 Otherwise, it’s a no-op.
                                  
diff --git a/2.7/generated/torch.nn.RNNCell.html b/2.7/generated/torch.nn.RNNCell.html
index 3e384421249..a1608c5735e 100644
--- a/2.7/generated/torch.nn.RNNCell.html
+++ b/2.7/generated/torch.nn.RNNCell.html
@@ -596,7 +596,7 @@
 
                                  RNNCell
                                  
 
                                  
 
                                  
-class torch.nn.RNNCell(input_size, hidden_size, bias=True, nonlinearity='tanh', device=None, dtype=None)[source][source]
                                  
+class torch.nn.RNNCell(input_size, hidden_size, bias=True, nonlinearity='tanh', device=None, dtype=None)[source][source]
 
                                  An Elman RNN cell with tanh or ReLU non-linearity.
                                  
 
                                  
 h′=tanh⁡(Wihx+bih+Whhh+bhh)h' = \tanh(W_{ih} x + b_{ih}  +  W_{hh} h + b_{hh})
                                  If nonlinearity is ‘relu’, then ReLU is used in place of tanh.
                                  
diff --git a/2.7/generated/torch.nn.RReLU.html b/2.7/generated/torch.nn.RReLU.html
index 1b3f5ea0eb4..653904b2537 100644
--- a/2.7/generated/torch.nn.RReLU.html
+++ b/2.7/generated/torch.nn.RReLU.html
@@ -596,7 +596,7 @@
 
                                  RReLU
                                  
 
                                  
 
                                  
-class torch.nn.RReLU(lower=0.125, upper=0.3333333333333333, inplace=False)[source][source]
                                  
+class torch.nn.RReLU(lower=0.125, upper=0.3333333333333333, inplace=False)[source][source]
 
                                  Applies the randomized leaky rectified linear unit function, element-wise.
                                  
 
                                  Method described in the paper:
 Empirical Evaluation of Rectified Activations in Convolutional Network.
                                  
diff --git a/2.7/generated/torch.nn.ReLU.html b/2.7/generated/torch.nn.ReLU.html
index 18c6bdcb304..4d2481aaf47 100644
--- a/2.7/generated/torch.nn.ReLU.html
+++ b/2.7/generated/torch.nn.ReLU.html
@@ -596,7 +596,7 @@
 
                                  ReLU
                                  
 
                                  
 
                                  
-class torch.nn.ReLU(inplace=False)[source][source]
                                  
+class torch.nn.ReLU(inplace=False)[source][source]
 
                                  Applies the rectified linear unit function element-wise.
                                  
 
                                  ReLU(x)=(x)+=max⁡(0,x)\text{ReLU}(x) = (x)^+ = \max(0, x)
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReLU6.html b/2.7/generated/torch.nn.ReLU6.html
index 6dc89b57a84..bb1e722c2d8 100644
--- a/2.7/generated/torch.nn.ReLU6.html
+++ b/2.7/generated/torch.nn.ReLU6.html
@@ -596,7 +596,7 @@
 
                                  ReLU6
                                  
 
                                  
 
                                  
-class torch.nn.ReLU6(inplace=False)[source][source]
                                  
+class torch.nn.ReLU6(inplace=False)[source][source]
 
                                  Applies the ReLU6 function element-wise.
                                  
 
                                  
 ReLU6(x)=min⁡(max⁡(0,x),6)\text{ReLU6}(x) = \min(\max(0,x), 6)
diff --git a/2.7/generated/torch.nn.ReflectionPad1d.html b/2.7/generated/torch.nn.ReflectionPad1d.html
index 7f5d8731db4..4919e22ac88 100644
--- a/2.7/generated/torch.nn.ReflectionPad1d.html
+++ b/2.7/generated/torch.nn.ReflectionPad1d.html
@@ -596,7 +596,7 @@
 
                                  ReflectionPad1d
                                  
 
                                  
 
                                  
-class torch.nn.ReflectionPad1d(padding)[source][source]
                                  
+class torch.nn.ReflectionPad1d(padding)[source][source]
 
                                  Pads the input tensor using the reflection of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReflectionPad2d.html b/2.7/generated/torch.nn.ReflectionPad2d.html
index 59b2308d12a..767f3c36551 100644
--- a/2.7/generated/torch.nn.ReflectionPad2d.html
+++ b/2.7/generated/torch.nn.ReflectionPad2d.html
@@ -596,7 +596,7 @@
 
                                  ReflectionPad2d
                                  
 
                                  
 
                                  
-class torch.nn.ReflectionPad2d(padding)[source][source]
                                  
+class torch.nn.ReflectionPad2d(padding)[source][source]
 
                                  Pads the input tensor using the reflection of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReflectionPad3d.html b/2.7/generated/torch.nn.ReflectionPad3d.html
index aab9fd7aafd..f7b4f842921 100644
--- a/2.7/generated/torch.nn.ReflectionPad3d.html
+++ b/2.7/generated/torch.nn.ReflectionPad3d.html
@@ -596,7 +596,7 @@
 
                                  ReflectionPad3d
                                  
 
                                  
 
                                  
-class torch.nn.ReflectionPad3d(padding)[source][source]
                                  
+class torch.nn.ReflectionPad3d(padding)[source][source]
 
                                  Pads the input tensor using the reflection of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReplicationPad1d.html b/2.7/generated/torch.nn.ReplicationPad1d.html
index 7c7d4a9425e..3d0149a3f79 100644
--- a/2.7/generated/torch.nn.ReplicationPad1d.html
+++ b/2.7/generated/torch.nn.ReplicationPad1d.html
@@ -596,7 +596,7 @@
 
                                  ReplicationPad1d
                                  
 
                                  
 
                                  
-class torch.nn.ReplicationPad1d(padding)[source][source]
                                  
+class torch.nn.ReplicationPad1d(padding)[source][source]
 
                                  Pads the input tensor using replication of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReplicationPad2d.html b/2.7/generated/torch.nn.ReplicationPad2d.html
index ce0446c1bd3..a3858ae5816 100644
--- a/2.7/generated/torch.nn.ReplicationPad2d.html
+++ b/2.7/generated/torch.nn.ReplicationPad2d.html
@@ -596,7 +596,7 @@
 
                                  ReplicationPad2d
                                  
 
                                  
 
                                  
-class torch.nn.ReplicationPad2d(padding)[source][source]
                                  
+class torch.nn.ReplicationPad2d(padding)[source][source]
 
                                  Pads the input tensor using replication of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.ReplicationPad3d.html b/2.7/generated/torch.nn.ReplicationPad3d.html
index f15f3c26325..fd559704c83 100644
--- a/2.7/generated/torch.nn.ReplicationPad3d.html
+++ b/2.7/generated/torch.nn.ReplicationPad3d.html
@@ -596,7 +596,7 @@
 
                                  ReplicationPad3d
                                  
 
                                  
 
                                  
-class torch.nn.ReplicationPad3d(padding)[source][source]
                                  
+class torch.nn.ReplicationPad3d(padding)[source][source]
 
                                  Pads the input tensor using replication of the input boundary.
                                  
 
                                  For N-dimensional padding, use torch.nn.functional.pad().
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.SELU.html b/2.7/generated/torch.nn.SELU.html
index eb1cb832cdb..d25a1056040 100644
--- a/2.7/generated/torch.nn.SELU.html
+++ b/2.7/generated/torch.nn.SELU.html
@@ -596,7 +596,7 @@
 
                                  SELU
                                  
 
                                  
 
                                  
-class torch.nn.SELU(inplace=False)[source][source]
                                  
+class torch.nn.SELU(inplace=False)[source][source]
 
                                  Applies the SELU function element-wise.
                                  
 
                                  
 SELU(x)=scale∗(max⁡(0,x)+min⁡(0,α∗(exp⁡(x)−1)))\text{SELU}(x) = \text{scale} * (\max(0,x) + \min(0, \alpha * (\exp(x) - 1)))
diff --git a/2.7/generated/torch.nn.Sequential.html b/2.7/generated/torch.nn.Sequential.html
index fa520cd9f42..e9bb43ef2c4 100644
--- a/2.7/generated/torch.nn.Sequential.html
+++ b/2.7/generated/torch.nn.Sequential.html
@@ -596,7 +596,7 @@
 
                                  Sequential
                                  
 
                                  
 
                                  
-class torch.nn.Sequential(*args: Module)[source][source]
                                  
+class torch.nn.Sequential(*args: Module)[source][source]
 
                                  
 class torch.nn.Sequential(arg: OrderedDict[str, Module])
                                  
 
                                  A sequential container.
                                  
@@ -641,7 +641,7 @@ 
                                  Sequential
 
                                  
-append(module)[source][source]
                                  
+append(module)[source][source]
 
                                  Append a given module to the end.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.SiLU.html b/2.7/generated/torch.nn.SiLU.html
index 24123430a75..9d73f48fb7e 100644
--- a/2.7/generated/torch.nn.SiLU.html
+++ b/2.7/generated/torch.nn.SiLU.html
@@ -596,7 +596,7 @@
 
                                  SiLU
                                  
 
                                  
 
                                  
-class torch.nn.SiLU(inplace=False)[source][source]
                                  
+class torch.nn.SiLU(inplace=False)[source][source]
 
                                  Applies the Sigmoid Linear Unit (SiLU) function, element-wise.
                                  
 
                                  The SiLU function is also known as the swish function.
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Sigmoid.html b/2.7/generated/torch.nn.Sigmoid.html
index d0d89b55458..6b1b72532b7 100644
--- a/2.7/generated/torch.nn.Sigmoid.html
+++ b/2.7/generated/torch.nn.Sigmoid.html
@@ -596,7 +596,7 @@
 
                                  Sigmoid
                                  
 
                                  
 
                                  
-class torch.nn.Sigmoid(*args, **kwargs)[source][source]
                                  
+class torch.nn.Sigmoid(*args, **kwargs)[source][source]
 
                                  Applies the Sigmoid function element-wise.
                                  
 
                                  
 Sigmoid(x)=σ(x)=11+exp⁡(−x)\text{Sigmoid}(x) = \sigma(x) = \frac{1}{1 + \exp(-x)}
diff --git a/2.7/generated/torch.nn.SmoothL1Loss.html b/2.7/generated/torch.nn.SmoothL1Loss.html
index 9a7586779f2..909816ba6f6 100644
--- a/2.7/generated/torch.nn.SmoothL1Loss.html
+++ b/2.7/generated/torch.nn.SmoothL1Loss.html
@@ -596,7 +596,7 @@
 
                                  SmoothL1Loss
                                  
 
                                  
 
                                  
-class torch.nn.SmoothL1Loss(size_average=None, reduce=None, reduction='mean', beta=1.0)[source][source]
                                  
+class torch.nn.SmoothL1Loss(size_average=None, reduce=None, reduction='mean', beta=1.0)[source][source]
 
                                  Creates a criterion that uses a squared term if the absolute
 element-wise error falls below beta and an L1 term otherwise.
 It is less sensitive to outliers than torch.nn.MSELoss and in some cases
diff --git a/2.7/generated/torch.nn.SoftMarginLoss.html b/2.7/generated/torch.nn.SoftMarginLoss.html
index 8734b5d5098..7f17ece1d85 100644
--- a/2.7/generated/torch.nn.SoftMarginLoss.html
+++ b/2.7/generated/torch.nn.SoftMarginLoss.html
@@ -596,7 +596,7 @@
 
                                  SoftMarginLoss
                                  
 
                                  
 
                                  
-class torch.nn.SoftMarginLoss(size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.SoftMarginLoss(size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that optimizes a two-class classification
 logistic loss between input tensor xx and target tensor yy
 (containing 1 or -1).
                                  
diff --git a/2.7/generated/torch.nn.Softmax.html b/2.7/generated/torch.nn.Softmax.html
index 1f9235922b0..1ad9fb60fd9 100644
--- a/2.7/generated/torch.nn.Softmax.html
+++ b/2.7/generated/torch.nn.Softmax.html
@@ -596,7 +596,7 @@
 
                                  Softmax
                                  
 
                                  
 
                                  
-class torch.nn.Softmax(dim=None)[source][source]
                                  
+class torch.nn.Softmax(dim=None)[source][source]
 
                                  Applies the Softmax function to an n-dimensional input Tensor.
                                  
 
                                  Rescales them so that the elements of the n-dimensional output Tensor
 lie in the range [0,1] and sum to 1.
                                  
diff --git a/2.7/generated/torch.nn.Softmax2d.html b/2.7/generated/torch.nn.Softmax2d.html
index 25a46fe5928..8fb21474b2c 100644
--- a/2.7/generated/torch.nn.Softmax2d.html
+++ b/2.7/generated/torch.nn.Softmax2d.html
@@ -596,7 +596,7 @@
 
                                  Softmax2d
                                  
 
                                  
 
                                  
-class torch.nn.Softmax2d(*args, **kwargs)[source][source]
                                  
+class torch.nn.Softmax2d(*args, **kwargs)[source][source]
 
                                  Applies SoftMax over features to each spatial location.
                                  
 
                                  When given an image of Channels x Height x Width, it will
 apply Softmax to each location (Channels,hi,wj)(Channels, h_i, w_j)
                                  
diff --git a/2.7/generated/torch.nn.Softmin.html b/2.7/generated/torch.nn.Softmin.html
index 3f47a8aa771..af91be3b889 100644
--- a/2.7/generated/torch.nn.Softmin.html
+++ b/2.7/generated/torch.nn.Softmin.html
@@ -596,7 +596,7 @@
 
                                  Softmin
                                  
 
                                  
 
                                  
-class torch.nn.Softmin(dim=None)[source][source]
                                  
+class torch.nn.Softmin(dim=None)[source][source]
 
                                  Applies the Softmin function to an n-dimensional input Tensor.
                                  
 
                                  Rescales them so that the elements of the n-dimensional output Tensor
 lie in the range [0, 1] and sum to 1.
                                  
diff --git a/2.7/generated/torch.nn.Softplus.html b/2.7/generated/torch.nn.Softplus.html
index 0585585c9c1..d7a9018c45d 100644
--- a/2.7/generated/torch.nn.Softplus.html
+++ b/2.7/generated/torch.nn.Softplus.html
@@ -596,7 +596,7 @@
 
                                  Softplus
                                  
 
                                  
 
                                  
-class torch.nn.Softplus(beta=1.0, threshold=20.0)[source][source]
                                  
+class torch.nn.Softplus(beta=1.0, threshold=20.0)[source][source]
 
                                  Applies the Softplus function element-wise.
                                  
 
                                  
 Softplus(x)=1β∗log⁡(1+exp⁡(β∗x))\text{Softplus}(x) = \frac{1}{\beta} * \log(1 + \exp(\beta * x))
diff --git a/2.7/generated/torch.nn.Softshrink.html b/2.7/generated/torch.nn.Softshrink.html
index 5442a06c9a7..c4879fb973a 100644
--- a/2.7/generated/torch.nn.Softshrink.html
+++ b/2.7/generated/torch.nn.Softshrink.html
@@ -596,7 +596,7 @@
 
                                  Softshrink
                                  
 
                                  
 
                                  
-class torch.nn.Softshrink(lambd=0.5)[source][source]
                                  
+class torch.nn.Softshrink(lambd=0.5)[source][source]
 
                                  Applies the soft shrinkage function element-wise.
                                  
 
                                  
 SoftShrinkage(x)={x−λ, if x>λx+λ, if x<−λ0, otherwise \text{SoftShrinkage}(x) =
diff --git a/2.7/generated/torch.nn.Softsign.html b/2.7/generated/torch.nn.Softsign.html
index 4e371c2f936..4ea15a9e7b5 100644
--- a/2.7/generated/torch.nn.Softsign.html
+++ b/2.7/generated/torch.nn.Softsign.html
@@ -596,7 +596,7 @@
 
                                  Softsign
                                  
 
                                  
 
                                  
-class torch.nn.Softsign(*args, **kwargs)[source][source]
                                  
+class torch.nn.Softsign(*args, **kwargs)[source][source]
 
                                  Applies the element-wise Softsign function.
                                  
 
                                  
 SoftSign(x)=x1+∣x∣\text{SoftSign}(x) = \frac{x}{ 1 + |x|}
diff --git a/2.7/generated/torch.nn.SyncBatchNorm.html b/2.7/generated/torch.nn.SyncBatchNorm.html
index 4e9c998bac9..21f50c3336e 100644
--- a/2.7/generated/torch.nn.SyncBatchNorm.html
+++ b/2.7/generated/torch.nn.SyncBatchNorm.html
@@ -596,7 +596,7 @@
 
                                  SyncBatchNorm
                                  
 
                                  
 
                                  
-class torch.nn.SyncBatchNorm(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, process_group=None, device=None, dtype=None)[source][source]
                                  
+class torch.nn.SyncBatchNorm(num_features, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True, process_group=None, device=None, dtype=None)[source][source]
 
                                  Applies Batch Normalization over a N-Dimensional input.
                                  
 
                                  The N-D input is a mini-batch of [N-2]D inputs with additional channel dimension) as described in the paper
 Batch Normalization: Accelerating Deep Network Training by Reducing
@@ -708,7 +708,7 @@ 
                                  SyncBatchNorm
 
                                  
-classmethod convert_sync_batchnorm(module, process_group=None)[source][source]
                                  
+classmethod convert_sync_batchnorm(module, process_group=None)[source][source]
 
                                  Converts all BatchNorm*D layers in the model to torch.nn.SyncBatchNorm layers.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.Tanh.html b/2.7/generated/torch.nn.Tanh.html
index 0578c39234c..c9298c2cec4 100644
--- a/2.7/generated/torch.nn.Tanh.html
+++ b/2.7/generated/torch.nn.Tanh.html
@@ -596,7 +596,7 @@
 
                                  Tanh
                                  
 
                                  
 
                                  
-class torch.nn.Tanh(*args, **kwargs)[source][source]
                                  
+class torch.nn.Tanh(*args, **kwargs)[source][source]
 
                                  Applies the Hyperbolic Tangent (Tanh) function element-wise.
                                  
 
                                  Tanh is defined as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Tanhshrink.html b/2.7/generated/torch.nn.Tanhshrink.html
index df22573a3d5..34143a39153 100644
--- a/2.7/generated/torch.nn.Tanhshrink.html
+++ b/2.7/generated/torch.nn.Tanhshrink.html
@@ -596,7 +596,7 @@
 
                                  Tanhshrink
                                  
 
                                  
 
                                  
-class torch.nn.Tanhshrink(*args, **kwargs)[source][source]
                                  
+class torch.nn.Tanhshrink(*args, **kwargs)[source][source]
 
                                  Applies the element-wise Tanhshrink function.
                                  
 
                                  
 Tanhshrink(x)=x−tanh⁡(x)\text{Tanhshrink}(x) = x - \tanh(x)
diff --git a/2.7/generated/torch.nn.Threshold.html b/2.7/generated/torch.nn.Threshold.html
index 0f71fec1282..59b2507e33e 100644
--- a/2.7/generated/torch.nn.Threshold.html
+++ b/2.7/generated/torch.nn.Threshold.html
@@ -596,7 +596,7 @@
 
                                  Threshold
                                  
 
                                  
 
                                  
-class torch.nn.Threshold(threshold, value, inplace=False)[source][source]
                                  
+class torch.nn.Threshold(threshold, value, inplace=False)[source][source]
 
                                  Thresholds each element of the input Tensor.
                                  
 
                                  Threshold is defined as:
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.Transformer.html b/2.7/generated/torch.nn.Transformer.html
index 3f51fa523ee..f2d5f646e0d 100644
--- a/2.7/generated/torch.nn.Transformer.html
+++ b/2.7/generated/torch.nn.Transformer.html
@@ -596,7 +596,7 @@
 
                                  Transformer
                                  
 
                                  
 
                                  
-class torch.nn.Transformer(d_model=512, nhead=8, num_encoder_layers=6, num_decoder_layers=6, dim_feedforward=2048, dropout=0.1, activation=<function relu>, custom_encoder=None, custom_decoder=None, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.Transformer(d_model=512, nhead=8, num_encoder_layers=6, num_decoder_layers=6, dim_feedforward=2048, dropout=0.1, activation=<function relu>, custom_encoder=None, custom_decoder=None, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
 
                                  A transformer model.
                                  
 
                                  
 
                                  Note
                                  
@@ -642,7 +642,7 @@ 
                                  Transformerhttps://github.com/pytorch/examples/tree/master/word_language_model
                                  
 
                                  
 
                                  
-forward(src, tgt, src_mask=None, tgt_mask=None, memory_mask=None, src_key_padding_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, src_is_causal=None, tgt_is_causal=None, memory_is_causal=False)[source][source]
                                  
+forward(src, tgt, src_mask=None, tgt_mask=None, memory_mask=None, src_key_padding_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, src_is_causal=None, tgt_is_causal=None, memory_is_causal=False)[source][source]
 
                                  Take in and process masked source/target sequences.
                                  
 
                                  
 
                                  Note
                                  
@@ -729,7 +729,7 @@ 
                                  Transformer
 
                                  
-static generate_square_subsequent_mask(sz, device=None, dtype=None)[source][source]
                                  
+static generate_square_subsequent_mask(sz, device=None, dtype=None)[source][source]
 
                                  Generate a square causal mask for the sequence.
                                  
 
                                  The masked positions are filled with float(‘-inf’). Unmasked positions are filled with float(0.0).
                                  
 
                                  
diff --git a/2.7/generated/torch.nn.TransformerDecoder.html b/2.7/generated/torch.nn.TransformerDecoder.html
index 715bd95d080..ffd961e1fab 100644
--- a/2.7/generated/torch.nn.TransformerDecoder.html
+++ b/2.7/generated/torch.nn.TransformerDecoder.html
@@ -596,7 +596,7 @@
 
                                  TransformerDecoder
                                  
 
                                  
 
                                  
-class torch.nn.TransformerDecoder(decoder_layer, num_layers, norm=None)[source][source]
                                  
+class torch.nn.TransformerDecoder(decoder_layer, num_layers, norm=None)[source][source]
 
                                  TransformerDecoder is a stack of N decoder layers.
                                  
 
                                  
 
                                  Note
                                  
@@ -625,7 +625,7 @@ 
                                  TransformerDecoder
 
                                  
-forward(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, tgt_is_causal=None, memory_is_causal=False)[source][source]
                                  
+forward(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, tgt_is_causal=None, memory_is_causal=False)[source][source]
 
                                  Pass the inputs (and mask) through the decoder layer in turn.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.TransformerDecoderLayer.html b/2.7/generated/torch.nn.TransformerDecoderLayer.html
index 8bf98413a15..ea9a37e042e 100644
--- a/2.7/generated/torch.nn.TransformerDecoderLayer.html
+++ b/2.7/generated/torch.nn.TransformerDecoderLayer.html
@@ -596,7 +596,7 @@
 
                                  TransformerDecoderLayer
                                  
 
                                  
 
                                  
-class torch.nn.TransformerDecoderLayer(d_model, nhead, dim_feedforward=2048, dropout=0.1, activation=<function relu>, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.TransformerDecoderLayer(d_model, nhead, dim_feedforward=2048, dropout=0.1, activation=<function relu>, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
 
                                  TransformerDecoderLayer is made up of self-attn, multi-head-attn and feedforward network.
                                  
 
                                  
 
                                  Note
                                  
@@ -644,7 +644,7 @@ 
                                  TransformerDecoderLayer
 
                                  
 
                                  
-forward(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, tgt_is_causal=False, memory_is_causal=False)[source][source]
                                  
+forward(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None, tgt_is_causal=False, memory_is_causal=False)[source][source]
 
                                  Pass the inputs (and mask) through the decoder layer.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.TransformerEncoder.html b/2.7/generated/torch.nn.TransformerEncoder.html
index 536730f0f88..cfd4bac62e5 100644
--- a/2.7/generated/torch.nn.TransformerEncoder.html
+++ b/2.7/generated/torch.nn.TransformerEncoder.html
@@ -596,7 +596,7 @@
 
                                  TransformerEncoder
                                  
 
                                  
 
                                  
-class torch.nn.TransformerEncoder(encoder_layer, num_layers, norm=None, enable_nested_tensor=True, mask_check=True)[source][source]
                                  
+class torch.nn.TransformerEncoder(encoder_layer, num_layers, norm=None, enable_nested_tensor=True, mask_check=True)[source][source]
 
                                  TransformerEncoder is a stack of N encoder layers.
                                  
 
                                  
 
                                  Note
                                  
@@ -628,7 +628,7 @@ 
                                  TransformerEncoder
 
                                  
-forward(src, mask=None, src_key_padding_mask=None, is_causal=None)[source][source]
                                  
+forward(src, mask=None, src_key_padding_mask=None, is_causal=None)[source][source]
 
                                  Pass the input through the encoder layers in turn.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.TransformerEncoderLayer.html b/2.7/generated/torch.nn.TransformerEncoderLayer.html
index 50c0cb6809c..1af20c7d7ff 100644
--- a/2.7/generated/torch.nn.TransformerEncoderLayer.html
+++ b/2.7/generated/torch.nn.TransformerEncoderLayer.html
@@ -596,7 +596,7 @@
 
                                  TransformerEncoderLayer
                                  
 
                                  
 
                                  
-class torch.nn.TransformerEncoderLayer(d_model, nhead, dim_feedforward=2048, dropout=0.1, activation=<function relu>, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
                                  
+class torch.nn.TransformerEncoderLayer(d_model, nhead, dim_feedforward=2048, dropout=0.1, activation=<function relu>, layer_norm_eps=1e-05, batch_first=False, norm_first=False, bias=True, device=None, dtype=None)[source][source]
 
                                  TransformerEncoderLayer is made up of self-attn and feedforward network.
                                  
 
                                  
 
                                  Note
                                  
@@ -674,7 +674,7 @@ 
                                  TransformerEncoderLayer
 
                                  
 
                                  
-forward(src, src_mask=None, src_key_padding_mask=None, is_causal=False)[source][source]
                                  
+forward(src, src_mask=None, src_key_padding_mask=None, is_causal=False)[source][source]
 
                                  Pass the input through the encoder layer.
                                  
 
                                  
 
                                  Parameters
                                  
diff --git a/2.7/generated/torch.nn.TripletMarginLoss.html b/2.7/generated/torch.nn.TripletMarginLoss.html
index 8505a550477..72519b34de5 100644
--- a/2.7/generated/torch.nn.TripletMarginLoss.html
+++ b/2.7/generated/torch.nn.TripletMarginLoss.html
@@ -596,7 +596,7 @@
 
                                  TripletMarginLoss
                                  
 
                                  
 
                                  
-class torch.nn.TripletMarginLoss(margin=1.0, p=2.0, eps=1e-06, swap=False, size_average=None, reduce=None, reduction='mean')[source][source]
                                  
+class torch.nn.TripletMarginLoss(margin=1.0, p=2.0, eps=1e-06, swap=False, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the triplet loss given an input
 tensors x1x1, x2x2, x3x3 and a margin with a value greater than 00.
 This is used for measuring a relative similarity between samples. A triplet
diff --git a/2.7/generated/torch.nn.TripletMarginWithDistanceLoss.html b/2.7/generated/torch.nn.TripletMarginWithDistanceLoss.html
index e702c2118d5..12fce0f3eeb 100644
--- a/2.7/generated/torch.nn.TripletMarginWithDistanceLoss.html
+++ b/2.7/generated/torch.nn.TripletMarginWithDistanceLoss.html
@@ -596,7 +596,7 @@
 
                                  TripletMarginWithDistanceLoss
                                  
 
                                  
 
                                  
-class torch.nn.TripletMarginWithDistanceLoss(*, distance_function=None, margin=1.0, swap=False, reduction='mean')[source][source]
                                  
+class torch.nn.TripletMarginWithDistanceLoss(*, distance_function=None, margin=1.0, swap=False, reduction='mean')[source][source]
 
                                  Creates a criterion that measures the triplet loss given input
 tensors aa, pp, and nn (representing anchor,
 positive, and negative examples, respectively), and a nonnegative,
diff --git a/2.7/generated/torch.nn.Unflatten.html b/2.7/generated/torch.nn.Unflatten.html
index 2b80b3f242e..a0f91561dd2 100644
--- a/2.7/generated/torch.nn.Unflatten.html
+++ b/2.7/generated/torch.nn.Unflatten.html
@@ -596,7 +596,7 @@
 
                                  Unflatten
                                  
 
                                  
 
                                  
-class torch.nn.Unflatten(dim, unflattened_size)[source][source]
                                  
+class torch.nn.Unflatten(dim, unflattened_size)[source][source]
 
                                  Unflattens a tensor dim expanding it to a desired shape. For use with Sequential.
                                  
 
                                    
 
                                  • dim specifies the dimension of the input tensor to be unflattened, and it can
diff --git a/2.7/generated/torch.nn.Unfold.html b/2.7/generated/torch.nn.Unfold.html
index c2c87c4b731..42f5f8c01ed 100644
--- a/2.7/generated/torch.nn.Unfold.html
+++ b/2.7/generated/torch.nn.Unfold.html
@@ -596,7 +596,7 @@
 
                                    Unfold
                                    
 
                                    
 
                                    
-class torch.nn.Unfold(kernel_size, dilation=1, padding=0, stride=1)[source][source]
                                    
+class torch.nn.Unfold(kernel_size, dilation=1, padding=0, stride=1)[source][source]
 
                                    Extracts sliding local blocks from a batched input tensor.
                                    
 
                                    Consider a batched input tensor of shape (N,C,∗)(N, C, *),
 where NN is the batch dimension, CC is the channel dimension,
diff --git a/2.7/generated/torch.nn.Upsample.html b/2.7/generated/torch.nn.Upsample.html
index 2eae7039a79..30e3bbdcc4e 100644
--- a/2.7/generated/torch.nn.Upsample.html
+++ b/2.7/generated/torch.nn.Upsample.html
@@ -596,7 +596,7 @@
 
                                    Upsample
                                    
 
                                    
 
                                    
-class torch.nn.Upsample(size=None, scale_factor=None, mode='nearest', align_corners=None, recompute_scale_factor=None)[source][source]
                                    
+class torch.nn.Upsample(size=None, scale_factor=None, mode='nearest', align_corners=None, recompute_scale_factor=None)[source][source]
 
                                    Upsamples a given multi-channel 1D (temporal), 2D (spatial) or 3D (volumetric) data.
                                    
 
                                    The input data is assumed to be of the form
 minibatch x channels x [optional depth] x [optional height] x width.
diff --git a/2.7/generated/torch.nn.UpsamplingBilinear2d.html b/2.7/generated/torch.nn.UpsamplingBilinear2d.html
index 09be7e27faf..343962ba9ea 100644
--- a/2.7/generated/torch.nn.UpsamplingBilinear2d.html
+++ b/2.7/generated/torch.nn.UpsamplingBilinear2d.html
@@ -596,7 +596,7 @@
 
                                    UpsamplingBilinear2d
                                    
 
                                    
 
                                    
-class torch.nn.UpsamplingBilinear2d(size=None, scale_factor=None)[source][source]
                                    
+class torch.nn.UpsamplingBilinear2d(size=None, scale_factor=None)[source][source]
 
                                    Applies a 2D bilinear upsampling to an input signal composed of several input channels.
                                    
 
                                    To specify the scale, it takes either the size or the scale_factor
 as it’s constructor argument.
                                    
diff --git a/2.7/generated/torch.nn.UpsamplingNearest2d.html b/2.7/generated/torch.nn.UpsamplingNearest2d.html
index bff9b893a16..6cda6bedcff 100644
--- a/2.7/generated/torch.nn.UpsamplingNearest2d.html
+++ b/2.7/generated/torch.nn.UpsamplingNearest2d.html
@@ -596,7 +596,7 @@
 
                                    UpsamplingNearest2d
                                    
 
                                    
 
                                    
-class torch.nn.UpsamplingNearest2d(size=None, scale_factor=None)[source][source]
                                    
+class torch.nn.UpsamplingNearest2d(size=None, scale_factor=None)[source][source]
 
                                    Applies a 2D nearest neighbor upsampling to an input signal composed of several input channels.
                                    
 
                                    To specify the scale, it takes either the size or the scale_factor
 as it’s constructor argument.
                                    
diff --git a/2.7/generated/torch.nn.ZeroPad1d.html b/2.7/generated/torch.nn.ZeroPad1d.html
index 9a19f05400d..7c8e448998e 100644
--- a/2.7/generated/torch.nn.ZeroPad1d.html
+++ b/2.7/generated/torch.nn.ZeroPad1d.html
@@ -596,7 +596,7 @@
 
                                    ZeroPad1d
                                    
 
                                    
 
                                    
-class torch.nn.ZeroPad1d(padding)[source][source]
                                    
+class torch.nn.ZeroPad1d(padding)[source][source]
 
                                    Pads the input tensor boundaries with zero.
                                    
 
                                    For N-dimensional padding, use torch.nn.functional.pad().
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.ZeroPad2d.html b/2.7/generated/torch.nn.ZeroPad2d.html
index 5aea86614a9..771cb008fe6 100644
--- a/2.7/generated/torch.nn.ZeroPad2d.html
+++ b/2.7/generated/torch.nn.ZeroPad2d.html
@@ -596,7 +596,7 @@
 
                                    ZeroPad2d
                                    
 
                                    
 
                                    
-class torch.nn.ZeroPad2d(padding)[source][source]
                                    
+class torch.nn.ZeroPad2d(padding)[source][source]
 
                                    Pads the input tensor boundaries with zero.
                                    
 
                                    For N-dimensional padding, use torch.nn.functional.pad().
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.ZeroPad3d.html b/2.7/generated/torch.nn.ZeroPad3d.html
index 51a9483b31c..dee44ec6c8b 100644
--- a/2.7/generated/torch.nn.ZeroPad3d.html
+++ b/2.7/generated/torch.nn.ZeroPad3d.html
@@ -596,7 +596,7 @@
 
                                    ZeroPad3d
                                    
 
                                    
 
                                    
-class torch.nn.ZeroPad3d(padding)[source][source]
                                    
+class torch.nn.ZeroPad3d(padding)[source][source]
 
                                    Pads the input tensor boundaries with zero.
                                    
 
                                    For N-dimensional padding, use torch.nn.functional.pad().
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.attention.bias.CausalBias.html b/2.7/generated/torch.nn.attention.bias.CausalBias.html
index 6662c6d8454..1de9cfb273b 100644
--- a/2.7/generated/torch.nn.attention.bias.CausalBias.html
+++ b/2.7/generated/torch.nn.attention.bias.CausalBias.html
@@ -598,7 +598,7 @@
 
                                    torch.nn.attention.bias.CausalBias
                                    
 
                                    
 
                                    
-class torch.nn.attention.bias.CausalBias(variant, seq_len_q, seq_len_kv)[source][source]
                                    
+class torch.nn.attention.bias.CausalBias(variant, seq_len_q, seq_len_kv)[source][source]
 
                                    A bias representing causal attention patterns. For an overview of the bias structure, see the CausalVariant enum.
                                    
 
                                    This class is used for defining causal (triangular) attention biases. For construing the bias, there exist
 two factory functions: causal_upper_left() and causal_lower_right().
                                    
diff --git a/2.7/generated/torch.nn.attention.bias.CausalVariant.html b/2.7/generated/torch.nn.attention.bias.CausalVariant.html
index c5f9e27d909..f6e55ba9bc5 100644
--- a/2.7/generated/torch.nn.attention.bias.CausalVariant.html
+++ b/2.7/generated/torch.nn.attention.bias.CausalVariant.html
@@ -598,7 +598,7 @@
 
                                    CausalVariant
                                    
 
                                    
 
                                    
-class torch.nn.attention.bias.CausalVariant(value)[source][source]
                                    
+class torch.nn.attention.bias.CausalVariant(value)[source][source]
 
                                    Enum for causal variants used in attention mechanisms.
                                    
 
                                    Defines two types of causal biases:
                                    
 
                                    UPPER_LEFT: Represents upper-left triangular bias for standard causal attention.
diff --git a/2.7/generated/torch.nn.attention.bias.causal_lower_right.html b/2.7/generated/torch.nn.attention.bias.causal_lower_right.html
index 779160e5403..c56637ad920 100644
--- a/2.7/generated/torch.nn.attention.bias.causal_lower_right.html
+++ b/2.7/generated/torch.nn.attention.bias.causal_lower_right.html
@@ -598,7 +598,7 @@
 
                                    torch.nn.attention.bias.causal_lower_right
                                    
 
                                    
 
                                    
-torch.nn.attention.bias.causal_lower_right(*size)[source][source]
                                    
+torch.nn.attention.bias.causal_lower_right(*size)[source][source]
 
                                    Creates a lower-right triangular causal bias.
                                    
 
                                    This function generates a lower-right triangular matrix to represent causal attention bias with a
 diagonal offset set so that the inclusive values are aligned to the lower right corner of the matrix.
                                    
diff --git a/2.7/generated/torch.nn.attention.bias.causal_upper_left.html b/2.7/generated/torch.nn.attention.bias.causal_upper_left.html
index f33885a217f..f3725474617 100644
--- a/2.7/generated/torch.nn.attention.bias.causal_upper_left.html
+++ b/2.7/generated/torch.nn.attention.bias.causal_upper_left.html
@@ -598,7 +598,7 @@
 
                                    torch.nn.attention.bias.causal_upper_left
                                    
 
                                    
 
                                    
-torch.nn.attention.bias.causal_upper_left(*size)[source][source]
                                    
+torch.nn.attention.bias.causal_upper_left(*size)[source][source]
 
                                    Creates an upper-left triangular causal bias.
                                    
 
                                    This function generates a upper-left triangular matrix to represent causal attention bias with a
 diagonal offset set so that the inclusive values are aligned to the upper left corner of the matrix.
diff --git a/2.7/generated/torch.nn.attention.sdpa_kernel.html b/2.7/generated/torch.nn.attention.sdpa_kernel.html
index 2b0d8266e4e..79bfbeae260 100644
--- a/2.7/generated/torch.nn.attention.sdpa_kernel.html
+++ b/2.7/generated/torch.nn.attention.sdpa_kernel.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.attention.sdpa_kernel
                                    
 
                                    
 
                                    
-torch.nn.attention.sdpa_kernel(backends, set_priority=False)[source][source]
                                    
+torch.nn.attention.sdpa_kernel(backends, set_priority=False)[source][source]
 
                                    Context manager to select which backend to use for scaled dot product attention.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.functional.adaptive_avg_pool2d.html b/2.7/generated/torch.nn.functional.adaptive_avg_pool2d.html
index 19b7a378e35..13f325120a7 100644
--- a/2.7/generated/torch.nn.functional.adaptive_avg_pool2d.html
+++ b/2.7/generated/torch.nn.functional.adaptive_avg_pool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.adaptive_avg_pool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.adaptive_avg_pool2d(input, output_size)[source][source]
                                    
+torch.nn.functional.adaptive_avg_pool2d(input, output_size)[source][source]
 
                                    Apply a 2D adaptive average pooling over an input signal composed of several input planes.
                                    
 
                                    See AdaptiveAvgPool2d for details and output shape.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.adaptive_avg_pool3d.html b/2.7/generated/torch.nn.functional.adaptive_avg_pool3d.html
index 0d4c8532fbd..d454a239520 100644
--- a/2.7/generated/torch.nn.functional.adaptive_avg_pool3d.html
+++ b/2.7/generated/torch.nn.functional.adaptive_avg_pool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.adaptive_avg_pool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.adaptive_avg_pool3d(input, output_size)[source][source]
                                    
+torch.nn.functional.adaptive_avg_pool3d(input, output_size)[source][source]
 
                                    Apply a 3D adaptive average pooling over an input signal composed of several input planes.
                                    
 
                                    See AdaptiveAvgPool3d for details and output shape.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.adaptive_max_pool1d.html b/2.7/generated/torch.nn.functional.adaptive_max_pool1d.html
index 69cdfd144ed..f4289a05768 100644
--- a/2.7/generated/torch.nn.functional.adaptive_max_pool1d.html
+++ b/2.7/generated/torch.nn.functional.adaptive_max_pool1d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.adaptive_max_pool1d
                                    
 
                                    
 
                                    
-torch.nn.functional.adaptive_max_pool1d(input, output_size, return_indices=False)[source]
                                    
+torch.nn.functional.adaptive_max_pool1d(input, output_size, return_indices=False)[source]
 
                                    Applies a 1D adaptive max pooling over an input signal composed of
 several input planes.
                                    
 
                                    See AdaptiveMaxPool1d for details and output shape.
                                    
diff --git a/2.7/generated/torch.nn.functional.adaptive_max_pool2d.html b/2.7/generated/torch.nn.functional.adaptive_max_pool2d.html
index 2b5a0aa7599..358a158fd9c 100644
--- a/2.7/generated/torch.nn.functional.adaptive_max_pool2d.html
+++ b/2.7/generated/torch.nn.functional.adaptive_max_pool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.adaptive_max_pool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.adaptive_max_pool2d(input, output_size, return_indices=False)[source]
                                    
+torch.nn.functional.adaptive_max_pool2d(input, output_size, return_indices=False)[source]
 
                                    Applies a 2D adaptive max pooling over an input signal composed of
 several input planes.
                                    
 
                                    See AdaptiveMaxPool2d for details and output shape.
                                    
diff --git a/2.7/generated/torch.nn.functional.adaptive_max_pool3d.html b/2.7/generated/torch.nn.functional.adaptive_max_pool3d.html
index 555fd51b2d2..efbe9b2e04c 100644
--- a/2.7/generated/torch.nn.functional.adaptive_max_pool3d.html
+++ b/2.7/generated/torch.nn.functional.adaptive_max_pool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.adaptive_max_pool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.adaptive_max_pool3d(input, output_size, return_indices=False)[source]
                                    
+torch.nn.functional.adaptive_max_pool3d(input, output_size, return_indices=False)[source]
 
                                    Applies a 3D adaptive max pooling over an input signal composed of
 several input planes.
                                    
 
                                    See AdaptiveMaxPool3d for details and output shape.
                                    
diff --git a/2.7/generated/torch.nn.functional.affine_grid.html b/2.7/generated/torch.nn.functional.affine_grid.html
index 3ec64c18848..87ec7aa930d 100644
--- a/2.7/generated/torch.nn.functional.affine_grid.html
+++ b/2.7/generated/torch.nn.functional.affine_grid.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.affine_grid
                                    
 
                                    
 
                                    
-torch.nn.functional.affine_grid(theta, size, align_corners=None)[source][source]
                                    
+torch.nn.functional.affine_grid(theta, size, align_corners=None)[source][source]
 
                                    Generate 2D or 3D flow field (sampling grid), given a batch of affine matrices theta.
                                    
 
                                    
 
                                    Note
                                    
diff --git a/2.7/generated/torch.nn.functional.alpha_dropout.html b/2.7/generated/torch.nn.functional.alpha_dropout.html
index 97e24f4e7db..4477e0c1b06 100644
--- a/2.7/generated/torch.nn.functional.alpha_dropout.html
+++ b/2.7/generated/torch.nn.functional.alpha_dropout.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.alpha_dropout
                                    
 
                                    
 
                                    
-torch.nn.functional.alpha_dropout(input, p=0.5, training=False, inplace=False)[source][source]
                                    
+torch.nn.functional.alpha_dropout(input, p=0.5, training=False, inplace=False)[source][source]
 
                                    Apply alpha dropout to the input.
                                    
 
                                    See AlphaDropout for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.batch_norm.html b/2.7/generated/torch.nn.functional.batch_norm.html
index 0b9fb479b12..09991a4a4c0 100644
--- a/2.7/generated/torch.nn.functional.batch_norm.html
+++ b/2.7/generated/torch.nn.functional.batch_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.batch_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.batch_norm(input, running_mean, running_var, weight=None, bias=None, training=False, momentum=0.1, eps=1e-05)[source][source]
                                    
+torch.nn.functional.batch_norm(input, running_mean, running_var, weight=None, bias=None, training=False, momentum=0.1, eps=1e-05)[source][source]
 
                                    Apply Batch Normalization for each channel across a batch of data.
                                    
 
                                    See BatchNorm1d, BatchNorm2d,
 BatchNorm3d for details.
                                    
diff --git a/2.7/generated/torch.nn.functional.binary_cross_entropy.html b/2.7/generated/torch.nn.functional.binary_cross_entropy.html
index 13252e0d3f5..19e6e9e16c8 100644
--- a/2.7/generated/torch.nn.functional.binary_cross_entropy.html
+++ b/2.7/generated/torch.nn.functional.binary_cross_entropy.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.binary_cross_entropy
                                    
 
                                    
 
                                    
-torch.nn.functional.binary_cross_entropy(input, target, weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
                                    
+torch.nn.functional.binary_cross_entropy(input, target, weight=None, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                    Measure Binary Cross Entropy between the target and input probabilities.
                                    
 
                                    See BCELoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.binary_cross_entropy_with_logits.html b/2.7/generated/torch.nn.functional.binary_cross_entropy_with_logits.html
index 29b36cc518f..249d37559e0 100644
--- a/2.7/generated/torch.nn.functional.binary_cross_entropy_with_logits.html
+++ b/2.7/generated/torch.nn.functional.binary_cross_entropy_with_logits.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.binary_cross_entropy_with_logits
                                    
 
                                    
 
                                    
-torch.nn.functional.binary_cross_entropy_with_logits(input, target, weight=None, size_average=None, reduce=None, reduction='mean', pos_weight=None)[source][source]
                                    
+torch.nn.functional.binary_cross_entropy_with_logits(input, target, weight=None, size_average=None, reduce=None, reduction='mean', pos_weight=None)[source][source]
 
                                    Calculate Binary Cross Entropy between target and input logits.
                                    
 
                                    See BCEWithLogitsLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.celu.html b/2.7/generated/torch.nn.functional.celu.html
index 1335b3dae64..771873bce9a 100644
--- a/2.7/generated/torch.nn.functional.celu.html
+++ b/2.7/generated/torch.nn.functional.celu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.celu
                                    
 
                                    
 
                                    
-torch.nn.functional.celu(input, alpha=1., inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.celu(input, alpha=1., inplace=False) → Tensor[source][source]
 
                                    Applies element-wise,
 CELU(x)=max⁡(0,x)+min⁡(0,α∗(exp⁡(x/α)−1))\text{CELU}(x) = \max(0,x) + \min(0, \alpha * (\exp(x/\alpha) - 1)).
                                    
 
                                    See CELU for more details.
                                    
diff --git a/2.7/generated/torch.nn.functional.cosine_embedding_loss.html b/2.7/generated/torch.nn.functional.cosine_embedding_loss.html
index 513a4ebdbd7..539bd5b6c98 100644
--- a/2.7/generated/torch.nn.functional.cosine_embedding_loss.html
+++ b/2.7/generated/torch.nn.functional.cosine_embedding_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.cosine_embedding_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.cosine_embedding_loss(input1, input2, target, margin=0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.cosine_embedding_loss(input1, input2, target, margin=0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See CosineEmbeddingLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.cross_entropy.html b/2.7/generated/torch.nn.functional.cross_entropy.html
index ecd3069677f..272304a10b1 100644
--- a/2.7/generated/torch.nn.functional.cross_entropy.html
+++ b/2.7/generated/torch.nn.functional.cross_entropy.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.cross_entropy
                                    
 
                                    
 
                                    
-torch.nn.functional.cross_entropy(input, target, weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean', label_smoothing=0.0)[source][source]
                                    
+torch.nn.functional.cross_entropy(input, target, weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean', label_smoothing=0.0)[source][source]
 
                                    Compute the cross entropy loss between input logits and target.
                                    
 
                                    See CrossEntropyLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.ctc_loss.html b/2.7/generated/torch.nn.functional.ctc_loss.html
index c393855c3ca..6b535d8d348 100644
--- a/2.7/generated/torch.nn.functional.ctc_loss.html
+++ b/2.7/generated/torch.nn.functional.ctc_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.ctc_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.ctc_loss(log_probs, targets, input_lengths, target_lengths, blank=0, reduction='mean', zero_infinity=False)[source][source]
                                    
+torch.nn.functional.ctc_loss(log_probs, targets, input_lengths, target_lengths, blank=0, reduction='mean', zero_infinity=False)[source][source]
 
                                    Apply the Connectionist Temporal Classification loss.
                                    
 
                                    See CTCLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.dropout.html b/2.7/generated/torch.nn.functional.dropout.html
index a6bdea1e2a1..ee065fcbf43 100644
--- a/2.7/generated/torch.nn.functional.dropout.html
+++ b/2.7/generated/torch.nn.functional.dropout.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.dropout
                                    
 
                                    
 
                                    
-torch.nn.functional.dropout(input, p=0.5, training=True, inplace=False)[source][source]
                                    
+torch.nn.functional.dropout(input, p=0.5, training=True, inplace=False)[source][source]
 
                                    During training, randomly zeroes some elements of the input tensor with probability p.
                                    
 
                                    Uses samples from a Bernoulli distribution.
                                    
 
                                    See Dropout for details.
                                    
diff --git a/2.7/generated/torch.nn.functional.dropout1d.html b/2.7/generated/torch.nn.functional.dropout1d.html
index a4ceb6d940c..1307d9619cf 100644
--- a/2.7/generated/torch.nn.functional.dropout1d.html
+++ b/2.7/generated/torch.nn.functional.dropout1d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.dropout1d
                                    
 
                                    
 
                                    
-torch.nn.functional.dropout1d(input, p=0.5, training=True, inplace=False)[source][source]
                                    
+torch.nn.functional.dropout1d(input, p=0.5, training=True, inplace=False)[source][source]
 
                                    Randomly zero out entire channels (a channel is a 1D feature map).
                                    
 
                                    For example, the jj-th channel of the ii-th sample in the
 batched input is a 1D tensor input[i,j]\text{input}[i, j] of the input tensor.
diff --git a/2.7/generated/torch.nn.functional.dropout2d.html b/2.7/generated/torch.nn.functional.dropout2d.html
index 023fcf0b6f0..c53db1ffa11 100644
--- a/2.7/generated/torch.nn.functional.dropout2d.html
+++ b/2.7/generated/torch.nn.functional.dropout2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.dropout2d
                                    
 
                                    
 
                                    
-torch.nn.functional.dropout2d(input, p=0.5, training=True, inplace=False)[source][source]
                                    
+torch.nn.functional.dropout2d(input, p=0.5, training=True, inplace=False)[source][source]
 
                                    Randomly zero out entire channels (a channel is a 2D feature map).
                                    
 
                                    For example, the jj-th channel of the ii-th sample in the
 batched input is a 2D tensor input[i,j]\text{input}[i, j] of the input tensor.
diff --git a/2.7/generated/torch.nn.functional.dropout3d.html b/2.7/generated/torch.nn.functional.dropout3d.html
index 12d2ef32d76..19de6c9ac07 100644
--- a/2.7/generated/torch.nn.functional.dropout3d.html
+++ b/2.7/generated/torch.nn.functional.dropout3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.dropout3d
                                    
 
                                    
 
                                    
-torch.nn.functional.dropout3d(input, p=0.5, training=True, inplace=False)[source][source]
                                    
+torch.nn.functional.dropout3d(input, p=0.5, training=True, inplace=False)[source][source]
 
                                    Randomly zero out entire channels (a channel is a 3D feature map).
                                    
 
                                    For example, the jj-th channel of the ii-th sample in the
 batched input is a 3D tensor input[i,j]\text{input}[i, j] of the input tensor.
diff --git a/2.7/generated/torch.nn.functional.elu.html b/2.7/generated/torch.nn.functional.elu.html
index 378208ff55b..5a964b7a050 100644
--- a/2.7/generated/torch.nn.functional.elu.html
+++ b/2.7/generated/torch.nn.functional.elu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.elu
                                    
 
                                    
 
                                    
-torch.nn.functional.elu(input, alpha=1.0, inplace=False)[source][source]
                                    
+torch.nn.functional.elu(input, alpha=1.0, inplace=False)[source][source]
 
                                    Apply the Exponential Linear Unit (ELU) function element-wise.
                                    
 
                                    See ELU for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.embedding.html b/2.7/generated/torch.nn.functional.embedding.html
index 73c5f439b15..d505f68c2be 100644
--- a/2.7/generated/torch.nn.functional.embedding.html
+++ b/2.7/generated/torch.nn.functional.embedding.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.embedding
                                    
 
                                    
 
                                    
-torch.nn.functional.embedding(input, weight, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False)[source][source]
                                    
+torch.nn.functional.embedding(input, weight, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False)[source][source]
 
                                    Generate a simple lookup table that looks up embeddings in a fixed dictionary and size.
                                    
 
                                    This module is often used to retrieve word embeddings using indices.
 The input to the module is a list of indices, and the embedding matrix,
diff --git a/2.7/generated/torch.nn.functional.embedding_bag.html b/2.7/generated/torch.nn.functional.embedding_bag.html
index fb41ace894e..cb68b3e19c1 100644
--- a/2.7/generated/torch.nn.functional.embedding_bag.html
+++ b/2.7/generated/torch.nn.functional.embedding_bag.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.embedding_bag
                                    
 
                                    
 
                                    
-torch.nn.functional.embedding_bag(input, weight, offsets=None, max_norm=None, norm_type=2, scale_grad_by_freq=False, mode='mean', sparse=False, per_sample_weights=None, include_last_offset=False, padding_idx=None)[source][source]
                                    
+torch.nn.functional.embedding_bag(input, weight, offsets=None, max_norm=None, norm_type=2, scale_grad_by_freq=False, mode='mean', sparse=False, per_sample_weights=None, include_last_offset=False, padding_idx=None)[source][source]
 
                                    Compute sums, means or maxes of bags of embeddings.
                                    
 
                                    Calculation is done without instantiating the intermediate embeddings.
 See torch.nn.EmbeddingBag for more details.
                                    
diff --git a/2.7/generated/torch.nn.functional.feature_alpha_dropout.html b/2.7/generated/torch.nn.functional.feature_alpha_dropout.html
index 861e4a3f453..f4d650f9740 100644
--- a/2.7/generated/torch.nn.functional.feature_alpha_dropout.html
+++ b/2.7/generated/torch.nn.functional.feature_alpha_dropout.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.feature_alpha_dropout
                                    
 
                                    
 
                                    
-torch.nn.functional.feature_alpha_dropout(input, p=0.5, training=False, inplace=False)[source][source]
                                    
+torch.nn.functional.feature_alpha_dropout(input, p=0.5, training=False, inplace=False)[source][source]
 
                                    Randomly masks out entire channels (a channel is a feature map).
                                    
 
                                    For example, the jj-th channel of the ii-th sample in the batch input
 is a tensor input[i,j]\text{input}[i, j] of the input tensor. Instead of
diff --git a/2.7/generated/torch.nn.functional.fold.html b/2.7/generated/torch.nn.functional.fold.html
index 52cebe865d8..1300efd8f08 100644
--- a/2.7/generated/torch.nn.functional.fold.html
+++ b/2.7/generated/torch.nn.functional.fold.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.fold
                                    
 
                                    
 
                                    
-torch.nn.functional.fold(input, output_size, kernel_size, dilation=1, padding=0, stride=1)[source][source]
                                    
+torch.nn.functional.fold(input, output_size, kernel_size, dilation=1, padding=0, stride=1)[source][source]
 
                                    Combine an array of sliding local blocks into a large containing tensor.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.functional.fractional_max_pool2d.html b/2.7/generated/torch.nn.functional.fractional_max_pool2d.html
index b96aebe1375..faac1a50fca 100644
--- a/2.7/generated/torch.nn.functional.fractional_max_pool2d.html
+++ b/2.7/generated/torch.nn.functional.fractional_max_pool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.fractional_max_pool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.fractional_max_pool2d(input, kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source]
                                    
+torch.nn.functional.fractional_max_pool2d(input, kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source]
 
                                    Applies 2D fractional max pooling over an input signal composed of several input planes.
                                    
 
                                    Fractional MaxPooling is described in detail in the paper Fractional MaxPooling by Ben Graham
                                    
 
                                    The max-pooling operation is applied in kH×kWkH \times kW regions by a stochastic
diff --git a/2.7/generated/torch.nn.functional.fractional_max_pool3d.html b/2.7/generated/torch.nn.functional.fractional_max_pool3d.html
index fa8190a76df..058d6db78bd 100644
--- a/2.7/generated/torch.nn.functional.fractional_max_pool3d.html
+++ b/2.7/generated/torch.nn.functional.fractional_max_pool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.fractional_max_pool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.fractional_max_pool3d(input, kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source]
                                    
+torch.nn.functional.fractional_max_pool3d(input, kernel_size, output_size=None, output_ratio=None, return_indices=False, _random_samples=None)[source]
 
                                    Applies 3D fractional max pooling over an input signal composed of several input planes.
                                    
 
                                    Fractional MaxPooling is described in detail in the paper Fractional MaxPooling by Ben Graham
                                    
 
                                    The max-pooling operation is applied in kT×kH×kWkT \times kH \times kW regions by a stochastic
diff --git a/2.7/generated/torch.nn.functional.gaussian_nll_loss.html b/2.7/generated/torch.nn.functional.gaussian_nll_loss.html
index 91a6a8cd2fc..9fe88012fb0 100644
--- a/2.7/generated/torch.nn.functional.gaussian_nll_loss.html
+++ b/2.7/generated/torch.nn.functional.gaussian_nll_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.gaussian_nll_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.gaussian_nll_loss(input, target, var, full=False, eps=1e-06, reduction='mean')[source][source]
                                    
+torch.nn.functional.gaussian_nll_loss(input, target, var, full=False, eps=1e-06, reduction='mean')[source][source]
 
                                    Gaussian negative log likelihood loss.
                                    
 
                                    See GaussianNLLLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.glu.html b/2.7/generated/torch.nn.functional.glu.html
index 055cf6b2004..26b55c9e6e4 100644
--- a/2.7/generated/torch.nn.functional.glu.html
+++ b/2.7/generated/torch.nn.functional.glu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.glu
                                    
 
                                    
 
                                    
-torch.nn.functional.glu(input, dim=-1) → Tensor[source][source]
                                    
+torch.nn.functional.glu(input, dim=-1) → Tensor[source][source]
 
                                    The gated linear unit. Computes:
                                    
 
                                    
 GLU(a,b)=a⊗σ(b)\text{GLU}(a, b) = a \otimes \sigma(b)
diff --git a/2.7/generated/torch.nn.functional.grid_sample.html b/2.7/generated/torch.nn.functional.grid_sample.html
index 790828dcf88..b8a386760f5 100644
--- a/2.7/generated/torch.nn.functional.grid_sample.html
+++ b/2.7/generated/torch.nn.functional.grid_sample.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.grid_sample
                                    
 
                                    
 
                                    
-torch.nn.functional.grid_sample(input, grid, mode='bilinear', padding_mode='zeros', align_corners=None)[source][source]
                                    
+torch.nn.functional.grid_sample(input, grid, mode='bilinear', padding_mode='zeros', align_corners=None)[source][source]
 
                                    Compute grid sample.
                                    
 
                                    Given an input and a flow-field grid, computes the
 output using input values and pixel locations from grid.
                                    
diff --git a/2.7/generated/torch.nn.functional.group_norm.html b/2.7/generated/torch.nn.functional.group_norm.html
index 73523d0f728..1c3590393c3 100644
--- a/2.7/generated/torch.nn.functional.group_norm.html
+++ b/2.7/generated/torch.nn.functional.group_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.group_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.group_norm(input, num_groups, weight=None, bias=None, eps=1e-05)[source][source]
                                    
+torch.nn.functional.group_norm(input, num_groups, weight=None, bias=None, eps=1e-05)[source][source]
 
                                    Apply Group Normalization for last certain number of dimensions.
                                    
 
                                    See GroupNorm for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.gumbel_softmax.html b/2.7/generated/torch.nn.functional.gumbel_softmax.html
index 7a2d1e2516a..e1eed2ecf46 100644
--- a/2.7/generated/torch.nn.functional.gumbel_softmax.html
+++ b/2.7/generated/torch.nn.functional.gumbel_softmax.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.gumbel_softmax
                                    
 
                                    
 
                                    
-torch.nn.functional.gumbel_softmax(logits, tau=1, hard=False, eps=1e-10, dim=-1)[source][source]
                                    
+torch.nn.functional.gumbel_softmax(logits, tau=1, hard=False, eps=1e-10, dim=-1)[source][source]
 
                                    Sample from the Gumbel-Softmax distribution (Link 1  Link 2) and optionally discretize.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.functional.hardsigmoid.html b/2.7/generated/torch.nn.functional.hardsigmoid.html
index 7c1728375e1..8fd35b5280b 100644
--- a/2.7/generated/torch.nn.functional.hardsigmoid.html
+++ b/2.7/generated/torch.nn.functional.hardsigmoid.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.hardsigmoid
                                    
 
                                    
 
                                    
-torch.nn.functional.hardsigmoid(input, inplace=False)[source][source]
                                    
+torch.nn.functional.hardsigmoid(input, inplace=False)[source][source]
 
                                    Apply the Hardsigmoid function element-wise.
                                    
 
                                    
 Hardsigmoid(x)={0if x≤−3,1if x≥+3,x/6+1/2otherwise\text{Hardsigmoid}(x) = \begin{cases}
diff --git a/2.7/generated/torch.nn.functional.hardswish.html b/2.7/generated/torch.nn.functional.hardswish.html
index 2cc67eb372d..b162ad1e0f4 100644
--- a/2.7/generated/torch.nn.functional.hardswish.html
+++ b/2.7/generated/torch.nn.functional.hardswish.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.hardswish
                                    
 
                                    
 
                                    
-torch.nn.functional.hardswish(input, inplace=False)[source][source]
                                    
+torch.nn.functional.hardswish(input, inplace=False)[source][source]
 
                                    Apply hardswish function, element-wise.
                                    
 
                                    Follows implementation as described in the paper:
 Searching for MobileNetV3.
                                    
diff --git a/2.7/generated/torch.nn.functional.hardtanh.html b/2.7/generated/torch.nn.functional.hardtanh.html
index 1be6f81cfe4..66921207570 100644
--- a/2.7/generated/torch.nn.functional.hardtanh.html
+++ b/2.7/generated/torch.nn.functional.hardtanh.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.hardtanh
                                    
 
                                    
 
                                    
-torch.nn.functional.hardtanh(input, min_val=-1., max_val=1., inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.hardtanh(input, min_val=-1., max_val=1., inplace=False) → Tensor[source][source]
 
                                    Applies the HardTanh function element-wise. See Hardtanh for more
 details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.hinge_embedding_loss.html b/2.7/generated/torch.nn.functional.hinge_embedding_loss.html
index a51a7f20e4c..a759d313d96 100644
--- a/2.7/generated/torch.nn.functional.hinge_embedding_loss.html
+++ b/2.7/generated/torch.nn.functional.hinge_embedding_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.hinge_embedding_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.hinge_embedding_loss(input, target, margin=1.0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.hinge_embedding_loss(input, target, margin=1.0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See HingeEmbeddingLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.huber_loss.html b/2.7/generated/torch.nn.functional.huber_loss.html
index 24903003072..1d992baf339 100644
--- a/2.7/generated/torch.nn.functional.huber_loss.html
+++ b/2.7/generated/torch.nn.functional.huber_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.huber_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.huber_loss(input, target, reduction='mean', delta=1.0, weight=None) → Tensor[source][source]
                                    
+torch.nn.functional.huber_loss(input, target, reduction='mean', delta=1.0, weight=None) → Tensor[source][source]
 
                                    Computes the Huber loss, with optional weighting.
                                    
 
                                    Function uses a squared term if the absolute
 element-wise error falls below delta and a delta-scaled L1 term otherwise.
                                    
diff --git a/2.7/generated/torch.nn.functional.instance_norm.html b/2.7/generated/torch.nn.functional.instance_norm.html
index 575aff3d9e9..057c4b4cbbf 100644
--- a/2.7/generated/torch.nn.functional.instance_norm.html
+++ b/2.7/generated/torch.nn.functional.instance_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.instance_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.instance_norm(input, running_mean=None, running_var=None, weight=None, bias=None, use_input_stats=True, momentum=0.1, eps=1e-05)[source][source]
                                    
+torch.nn.functional.instance_norm(input, running_mean=None, running_var=None, weight=None, bias=None, use_input_stats=True, momentum=0.1, eps=1e-05)[source][source]
 
                                    Apply Instance Normalization independently for each channel in every data sample within a batch.
                                    
 
                                    See InstanceNorm1d, InstanceNorm2d,
 InstanceNorm3d for details.
                                    
diff --git a/2.7/generated/torch.nn.functional.interpolate.html b/2.7/generated/torch.nn.functional.interpolate.html
index 5ba621119c6..b349e7019a7 100644
--- a/2.7/generated/torch.nn.functional.interpolate.html
+++ b/2.7/generated/torch.nn.functional.interpolate.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.interpolate
                                    
 
                                    
 
                                    
-torch.nn.functional.interpolate(input, size=None, scale_factor=None, mode='nearest', align_corners=None, recompute_scale_factor=None, antialias=False)[source][source]
                                    
+torch.nn.functional.interpolate(input, size=None, scale_factor=None, mode='nearest', align_corners=None, recompute_scale_factor=None, antialias=False)[source][source]
 
                                    Down/up samples the input.
                                    
 
                                    Tensor interpolated to either the given size or the given
 scale_factor
                                    
diff --git a/2.7/generated/torch.nn.functional.kl_div.html b/2.7/generated/torch.nn.functional.kl_div.html
index 27954b3fd9b..a1b79fa4aa2 100644
--- a/2.7/generated/torch.nn.functional.kl_div.html
+++ b/2.7/generated/torch.nn.functional.kl_div.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.kl_div
                                    
 
                                    
 
                                    
-torch.nn.functional.kl_div(input, target, size_average=None, reduce=None, reduction='mean', log_target=False)[source][source]
                                    
+torch.nn.functional.kl_div(input, target, size_average=None, reduce=None, reduction='mean', log_target=False)[source][source]
 
                                    Compute the KL Divergence loss.
                                    
 
                                    Refer - The Kullback-Leibler divergence Loss
                                    
 
                                    See KLDivLoss for details.
                                    
diff --git a/2.7/generated/torch.nn.functional.l1_loss.html b/2.7/generated/torch.nn.functional.l1_loss.html
index 7cd78c9e714..3ba2b03bb35 100644
--- a/2.7/generated/torch.nn.functional.l1_loss.html
+++ b/2.7/generated/torch.nn.functional.l1_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.l1_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.l1_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.l1_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    Function that takes the mean element-wise absolute value difference.
                                    
 
                                    See L1Loss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.layer_norm.html b/2.7/generated/torch.nn.functional.layer_norm.html
index 01a21c73d96..7e3f4d47dc4 100644
--- a/2.7/generated/torch.nn.functional.layer_norm.html
+++ b/2.7/generated/torch.nn.functional.layer_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.layer_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.layer_norm(input, normalized_shape, weight=None, bias=None, eps=1e-05)[source][source]
                                    
+torch.nn.functional.layer_norm(input, normalized_shape, weight=None, bias=None, eps=1e-05)[source][source]
 
                                    Apply Layer Normalization for last certain number of dimensions.
                                    
 
                                    See LayerNorm for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.leaky_relu.html b/2.7/generated/torch.nn.functional.leaky_relu.html
index f7b8c9e8467..8fe467a659c 100644
--- a/2.7/generated/torch.nn.functional.leaky_relu.html
+++ b/2.7/generated/torch.nn.functional.leaky_relu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.leaky_relu
                                    
 
                                    
 
                                    
-torch.nn.functional.leaky_relu(input, negative_slope=0.01, inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.leaky_relu(input, negative_slope=0.01, inplace=False) → Tensor[source][source]
 
                                    Applies element-wise,
 LeakyReLU(x)=max⁡(0,x)+negative_slope∗min⁡(0,x)\text{LeakyReLU}(x) = \max(0, x) + \text{negative\_slope} * \min(0, x)
                                    
 
                                    See LeakyReLU for more details.
                                    
diff --git a/2.7/generated/torch.nn.functional.local_response_norm.html b/2.7/generated/torch.nn.functional.local_response_norm.html
index 57569a7ae39..60b8b35a6b0 100644
--- a/2.7/generated/torch.nn.functional.local_response_norm.html
+++ b/2.7/generated/torch.nn.functional.local_response_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.local_response_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.local_response_norm(input, size, alpha=0.0001, beta=0.75, k=1.0)[source][source]
                                    
+torch.nn.functional.local_response_norm(input, size, alpha=0.0001, beta=0.75, k=1.0)[source][source]
 
                                    Apply local response normalization over an input signal.
                                    
 
                                    The input signal is composed of several input planes, where channels occupy the second dimension.
 Normalization is applied across channels.
                                    
diff --git a/2.7/generated/torch.nn.functional.log_softmax.html b/2.7/generated/torch.nn.functional.log_softmax.html
index e52bb58376b..a65b2765014 100644
--- a/2.7/generated/torch.nn.functional.log_softmax.html
+++ b/2.7/generated/torch.nn.functional.log_softmax.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.log_softmax
                                    
 
                                    
 
                                    
-torch.nn.functional.log_softmax(input, dim=None, _stacklevel=3, dtype=None)[source][source]
                                    
+torch.nn.functional.log_softmax(input, dim=None, _stacklevel=3, dtype=None)[source][source]
 
                                    Apply a softmax followed by a logarithm.
                                    
 
                                    While mathematically equivalent to log(softmax(x)), doing these two
 operations separately is slower and numerically unstable. This function
diff --git a/2.7/generated/torch.nn.functional.lp_pool1d.html b/2.7/generated/torch.nn.functional.lp_pool1d.html
index 6fe91c81ce8..e3d9cdd563f 100644
--- a/2.7/generated/torch.nn.functional.lp_pool1d.html
+++ b/2.7/generated/torch.nn.functional.lp_pool1d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.lp_pool1d
                                    
 
                                    
 
                                    
-torch.nn.functional.lp_pool1d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                    
+torch.nn.functional.lp_pool1d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                    Apply a 1D power-average pooling over an input signal composed of several input planes.
                                    
 
                                    If the sum of all inputs to the power of p is
 zero, the gradient is set to zero as well.
                                    
diff --git a/2.7/generated/torch.nn.functional.lp_pool2d.html b/2.7/generated/torch.nn.functional.lp_pool2d.html
index ed43f0b8c0f..f5257147b37 100644
--- a/2.7/generated/torch.nn.functional.lp_pool2d.html
+++ b/2.7/generated/torch.nn.functional.lp_pool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.lp_pool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.lp_pool2d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                    
+torch.nn.functional.lp_pool2d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                    Apply a 2D power-average pooling over an input signal composed of several input planes.
                                    
 
                                    If the sum of all inputs to the power of p is
 zero, the gradient is set to zero as well.
                                    
diff --git a/2.7/generated/torch.nn.functional.lp_pool3d.html b/2.7/generated/torch.nn.functional.lp_pool3d.html
index 3f16d3c2ba0..7647580ade9 100644
--- a/2.7/generated/torch.nn.functional.lp_pool3d.html
+++ b/2.7/generated/torch.nn.functional.lp_pool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.lp_pool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.lp_pool3d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
                                    
+torch.nn.functional.lp_pool3d(input, norm_type, kernel_size, stride=None, ceil_mode=False)[source][source]
 
                                    Apply a 3D power-average pooling over an input signal composed of several input planes.
                                    
 
                                    If the sum of all inputs to the power of p is
 zero, the gradient is set to zero as well.
                                    
diff --git a/2.7/generated/torch.nn.functional.margin_ranking_loss.html b/2.7/generated/torch.nn.functional.margin_ranking_loss.html
index 0fd3c2e2038..e34958dee90 100644
--- a/2.7/generated/torch.nn.functional.margin_ranking_loss.html
+++ b/2.7/generated/torch.nn.functional.margin_ranking_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.margin_ranking_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.margin_ranking_loss(input1, input2, target, margin=0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.margin_ranking_loss(input1, input2, target, margin=0, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See MarginRankingLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.max_pool1d.html b/2.7/generated/torch.nn.functional.max_pool1d.html
index 8e04b7427ed..b442c633738 100644
--- a/2.7/generated/torch.nn.functional.max_pool1d.html
+++ b/2.7/generated/torch.nn.functional.max_pool1d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_pool1d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_pool1d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
                                    
+torch.nn.functional.max_pool1d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
 
                                    Applies a 1D max pooling over an input signal composed of several input
 planes.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.max_pool2d.html b/2.7/generated/torch.nn.functional.max_pool2d.html
index 2b6ca0ed2ac..c3fdb5e8b9c 100644
--- a/2.7/generated/torch.nn.functional.max_pool2d.html
+++ b/2.7/generated/torch.nn.functional.max_pool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_pool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_pool2d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
                                    
+torch.nn.functional.max_pool2d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
 
                                    Applies a 2D max pooling over an input signal composed of several input
 planes.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.max_pool3d.html b/2.7/generated/torch.nn.functional.max_pool3d.html
index 083f416d96b..2af1918396c 100644
--- a/2.7/generated/torch.nn.functional.max_pool3d.html
+++ b/2.7/generated/torch.nn.functional.max_pool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_pool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_pool3d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
                                    
+torch.nn.functional.max_pool3d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]
 
                                    Applies a 3D max pooling over an input signal composed of several input
 planes.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.max_unpool1d.html b/2.7/generated/torch.nn.functional.max_unpool1d.html
index 9598fb7430d..67706930705 100644
--- a/2.7/generated/torch.nn.functional.max_unpool1d.html
+++ b/2.7/generated/torch.nn.functional.max_unpool1d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_unpool1d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_unpool1d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
                                    
+torch.nn.functional.max_unpool1d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
 
                                    Compute a partial inverse of MaxPool1d.
                                    
 
                                    See MaxUnpool1d for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.max_unpool2d.html b/2.7/generated/torch.nn.functional.max_unpool2d.html
index c029985b758..b2feb54a5eb 100644
--- a/2.7/generated/torch.nn.functional.max_unpool2d.html
+++ b/2.7/generated/torch.nn.functional.max_unpool2d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_unpool2d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_unpool2d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
                                    
+torch.nn.functional.max_unpool2d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
 
                                    Compute a partial inverse of MaxPool2d.
                                    
 
                                    See MaxUnpool2d for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.max_unpool3d.html b/2.7/generated/torch.nn.functional.max_unpool3d.html
index 96048d3265d..007d7ed6ed9 100644
--- a/2.7/generated/torch.nn.functional.max_unpool3d.html
+++ b/2.7/generated/torch.nn.functional.max_unpool3d.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.max_unpool3d
                                    
 
                                    
 
                                    
-torch.nn.functional.max_unpool3d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
                                    
+torch.nn.functional.max_unpool3d(input, indices, kernel_size, stride=None, padding=0, output_size=None)[source][source]
 
                                    Compute a partial inverse of MaxPool3d.
                                    
 
                                    See MaxUnpool3d for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.mish.html b/2.7/generated/torch.nn.functional.mish.html
index abdb61ae9d4..8e63e9ef7c2 100644
--- a/2.7/generated/torch.nn.functional.mish.html
+++ b/2.7/generated/torch.nn.functional.mish.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.mish
                                    
 
                                    
 
                                    
-torch.nn.functional.mish(input, inplace=False)[source][source]
                                    
+torch.nn.functional.mish(input, inplace=False)[source][source]
 
                                    Apply the Mish function, element-wise.
                                    
 
                                    Mish: A Self Regularized Non-Monotonic Neural Activation Function.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.mse_loss.html b/2.7/generated/torch.nn.functional.mse_loss.html
index d8c0e14003e..8c29074126f 100644
--- a/2.7/generated/torch.nn.functional.mse_loss.html
+++ b/2.7/generated/torch.nn.functional.mse_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.mse_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.mse_loss(input, target, size_average=None, reduce=None, reduction='mean', weight=None) → Tensor[source][source]
                                    
+torch.nn.functional.mse_loss(input, target, size_average=None, reduce=None, reduction='mean', weight=None) → Tensor[source][source]
 
                                    Measures the element-wise mean squared error, with optional weighting.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.functional.multi_margin_loss.html b/2.7/generated/torch.nn.functional.multi_margin_loss.html
index bb6c5151900..2552093202a 100644
--- a/2.7/generated/torch.nn.functional.multi_margin_loss.html
+++ b/2.7/generated/torch.nn.functional.multi_margin_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.multi_margin_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.multi_margin_loss(input, target, p=1, margin=1, weight=None, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.multi_margin_loss(input, target, p=1, margin=1, weight=None, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See MultiMarginLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.multilabel_margin_loss.html b/2.7/generated/torch.nn.functional.multilabel_margin_loss.html
index 47773105ec6..9fba4cd83ad 100644
--- a/2.7/generated/torch.nn.functional.multilabel_margin_loss.html
+++ b/2.7/generated/torch.nn.functional.multilabel_margin_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.multilabel_margin_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.multilabel_margin_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.multilabel_margin_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See MultiLabelMarginLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.multilabel_soft_margin_loss.html b/2.7/generated/torch.nn.functional.multilabel_soft_margin_loss.html
index 78d1d015759..a65636b382b 100644
--- a/2.7/generated/torch.nn.functional.multilabel_soft_margin_loss.html
+++ b/2.7/generated/torch.nn.functional.multilabel_soft_margin_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.multilabel_soft_margin_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.multilabel_soft_margin_loss(input, target, weight=None, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.multilabel_soft_margin_loss(input, target, weight=None, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See MultiLabelSoftMarginLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.nll_loss.html b/2.7/generated/torch.nn.functional.nll_loss.html
index 0f2e247c3e3..f119fd7ad80 100644
--- a/2.7/generated/torch.nn.functional.nll_loss.html
+++ b/2.7/generated/torch.nn.functional.nll_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.nll_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.nll_loss(input, target, weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean')[source][source]
                                    
+torch.nn.functional.nll_loss(input, target, weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean')[source][source]
 
                                    Compute the negative log likelihood loss.
                                    
 
                                    See NLLLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.normalize.html b/2.7/generated/torch.nn.functional.normalize.html
index 613390491c1..4e9ac646408 100644
--- a/2.7/generated/torch.nn.functional.normalize.html
+++ b/2.7/generated/torch.nn.functional.normalize.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.normalize
                                    
 
                                    
 
                                    
-torch.nn.functional.normalize(input, p=2.0, dim=1, eps=1e-12, out=None)[source][source]
                                    
+torch.nn.functional.normalize(input, p=2.0, dim=1, eps=1e-12, out=None)[source][source]
 
                                    Perform LpL_p normalization of inputs over specified dimension.
                                    
 
                                    For a tensor input of sizes (n0,...,ndim,...,nk)(n_0, ..., n_{dim}, ..., n_k), each
 ndimn_{dim} -element vector vv along dimension dim is transformed as
                                    
diff --git a/2.7/generated/torch.nn.functional.pad.html b/2.7/generated/torch.nn.functional.pad.html
index 48f43da76ed..1e292fca40b 100644
--- a/2.7/generated/torch.nn.functional.pad.html
+++ b/2.7/generated/torch.nn.functional.pad.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.pad
                                    
 
                                    
 
                                    
-torch.nn.functional.pad(input, pad, mode='constant', value=None) → Tensor[source][source]
                                    
+torch.nn.functional.pad(input, pad, mode='constant', value=None) → Tensor[source][source]
 
                                    Pads tensor.
                                    
 
                                    
 
                                    Padding size:
                                    The padding size by which to pad some dimensions of input
diff --git a/2.7/generated/torch.nn.functional.poisson_nll_loss.html b/2.7/generated/torch.nn.functional.poisson_nll_loss.html
index 0417b15d9c0..850af72b153 100644
--- a/2.7/generated/torch.nn.functional.poisson_nll_loss.html
+++ b/2.7/generated/torch.nn.functional.poisson_nll_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.poisson_nll_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.poisson_nll_loss(input, target, log_input=True, full=False, size_average=None, eps=1e-08, reduce=None, reduction='mean')[source][source]
                                    
+torch.nn.functional.poisson_nll_loss(input, target, log_input=True, full=False, size_average=None, eps=1e-08, reduce=None, reduction='mean')[source][source]
 
                                    Poisson negative log likelihood loss.
                                    
 
                                    See PoissonNLLLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.relu.html b/2.7/generated/torch.nn.functional.relu.html
index 8e3149687fa..64d7ff57233 100644
--- a/2.7/generated/torch.nn.functional.relu.html
+++ b/2.7/generated/torch.nn.functional.relu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.relu
                                    
 
                                    
 
                                    
-torch.nn.functional.relu(input, inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.relu(input, inplace=False) → Tensor[source][source]
 
                                    Applies the rectified linear unit function element-wise. See
 ReLU for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.relu6.html b/2.7/generated/torch.nn.functional.relu6.html
index f92f98bc945..8b31b052943 100644
--- a/2.7/generated/torch.nn.functional.relu6.html
+++ b/2.7/generated/torch.nn.functional.relu6.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.relu6
                                    
 
                                    
 
                                    
-torch.nn.functional.relu6(input, inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.relu6(input, inplace=False) → Tensor[source][source]
 
                                    Applies the element-wise function ReLU6(x)=min⁡(max⁡(0,x),6)\text{ReLU6}(x) = \min(\max(0,x), 6).
                                    
 
                                    See ReLU6 for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.rms_norm.html b/2.7/generated/torch.nn.functional.rms_norm.html
index c7b144b1230..e356ab27df3 100644
--- a/2.7/generated/torch.nn.functional.rms_norm.html
+++ b/2.7/generated/torch.nn.functional.rms_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.rms_norm
                                    
 
                                    
 
                                    
-torch.nn.functional.rms_norm(input, normalized_shape, weight=None, eps=None)[source][source]
                                    
+torch.nn.functional.rms_norm(input, normalized_shape, weight=None, eps=None)[source][source]
 
                                    Apply Root Mean Square Layer Normalization.
                                    
 
                                    See RMSNorm for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.rrelu.html b/2.7/generated/torch.nn.functional.rrelu.html
index a48e6caae3c..440311baac7 100644
--- a/2.7/generated/torch.nn.functional.rrelu.html
+++ b/2.7/generated/torch.nn.functional.rrelu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.rrelu
                                    
 
                                    
 
                                    
-torch.nn.functional.rrelu(input, lower=1. / 8, upper=1. / 3, training=False, inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.rrelu(input, lower=1. / 8, upper=1. / 3, training=False, inplace=False) → Tensor[source][source]
 
                                    Randomized leaky ReLU.
                                    
 
                                    See RReLU for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.selu.html b/2.7/generated/torch.nn.functional.selu.html
index b7411cb9b12..683d2468b12 100644
--- a/2.7/generated/torch.nn.functional.selu.html
+++ b/2.7/generated/torch.nn.functional.selu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.selu
                                    
 
                                    
 
                                    
-torch.nn.functional.selu(input, inplace=False) → Tensor[source][source]
                                    
+torch.nn.functional.selu(input, inplace=False) → Tensor[source][source]
 
                                    Applies element-wise,
 SELU(x)=scale∗(max⁡(0,x)+min⁡(0,α∗(exp⁡(x)−1)))\text{SELU}(x) = scale * (\max(0,x) + \min(0, \alpha * (\exp(x) - 1))),
 with α=1.6732632423543772848170429916717\alpha=1.6732632423543772848170429916717 and
diff --git a/2.7/generated/torch.nn.functional.sigmoid.html b/2.7/generated/torch.nn.functional.sigmoid.html
index 152cacce417..7053e90d842 100644
--- a/2.7/generated/torch.nn.functional.sigmoid.html
+++ b/2.7/generated/torch.nn.functional.sigmoid.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.sigmoid
                                    
 
                                    
 
                                    
-torch.nn.functional.sigmoid(input) → Tensor[source][source]
                                    
+torch.nn.functional.sigmoid(input) → Tensor[source][source]
 
                                    Applies the element-wise function Sigmoid(x)=11+exp⁡(−x)\text{Sigmoid}(x) = \frac{1}{1 + \exp(-x)}
                                    
 
                                    See Sigmoid for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.silu.html b/2.7/generated/torch.nn.functional.silu.html
index 9a05687d20f..8f212979b85 100644
--- a/2.7/generated/torch.nn.functional.silu.html
+++ b/2.7/generated/torch.nn.functional.silu.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.silu
                                    
 
                                    
 
                                    
-torch.nn.functional.silu(input, inplace=False)[source][source]
                                    
+torch.nn.functional.silu(input, inplace=False)[source][source]
 
                                    Apply the Sigmoid Linear Unit (SiLU) function, element-wise.
                                    
 
                                    The SiLU function is also known as the swish function.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.smooth_l1_loss.html b/2.7/generated/torch.nn.functional.smooth_l1_loss.html
index 8c85449fa76..f98470e346a 100644
--- a/2.7/generated/torch.nn.functional.smooth_l1_loss.html
+++ b/2.7/generated/torch.nn.functional.smooth_l1_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.smooth_l1_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.smooth_l1_loss(input, target, size_average=None, reduce=None, reduction='mean', beta=1.0)[source][source]
                                    
+torch.nn.functional.smooth_l1_loss(input, target, size_average=None, reduce=None, reduction='mean', beta=1.0)[source][source]
 
                                    Compute the Smooth L1 loss.
                                    
 
                                    Function uses a squared term if the absolute
 element-wise error falls below beta and an L1 term otherwise.
                                    
diff --git a/2.7/generated/torch.nn.functional.soft_margin_loss.html b/2.7/generated/torch.nn.functional.soft_margin_loss.html
index 577f31af629..77f3251a30c 100644
--- a/2.7/generated/torch.nn.functional.soft_margin_loss.html
+++ b/2.7/generated/torch.nn.functional.soft_margin_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.soft_margin_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.soft_margin_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
                                    
+torch.nn.functional.soft_margin_loss(input, target, size_average=None, reduce=None, reduction='mean') → Tensor[source][source]
 
                                    See SoftMarginLoss for details.
                                    
 
                                    
 
                                    Return type
                                    
diff --git a/2.7/generated/torch.nn.functional.softmax.html b/2.7/generated/torch.nn.functional.softmax.html
index fbd582e9af8..3806f1ac0e0 100644
--- a/2.7/generated/torch.nn.functional.softmax.html
+++ b/2.7/generated/torch.nn.functional.softmax.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.softmax
                                    
 
                                    
 
                                    
-torch.nn.functional.softmax(input, dim=None, _stacklevel=3, dtype=None)[source][source]
                                    
+torch.nn.functional.softmax(input, dim=None, _stacklevel=3, dtype=None)[source][source]
 
                                    Apply a softmax function.
                                    
 
                                    Softmax is defined as:
                                    
 
                                    Softmax(xi)=exp⁡(xi)∑jexp⁡(xj)\text{Softmax}(x_{i}) = \frac{\exp(x_i)}{\sum_j \exp(x_j)}
                                    
diff --git a/2.7/generated/torch.nn.functional.softmin.html b/2.7/generated/torch.nn.functional.softmin.html
index e27e42ed5bb..73775fbcdd5 100644
--- a/2.7/generated/torch.nn.functional.softmin.html
+++ b/2.7/generated/torch.nn.functional.softmin.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.softmin
                                    
 
                                    
 
                                    
-torch.nn.functional.softmin(input, dim=None, _stacklevel=3, dtype=None)[source][source]
                                    
+torch.nn.functional.softmin(input, dim=None, _stacklevel=3, dtype=None)[source][source]
 
                                    Apply a softmin function.
                                    
 
                                    Note that Softmin(x)=Softmax(−x)\text{Softmin}(x) = \text{Softmax}(-x). See softmax definition for mathematical formula.
                                    
 
                                    See Softmin for more details.
                                    
diff --git a/2.7/generated/torch.nn.functional.softsign.html b/2.7/generated/torch.nn.functional.softsign.html
index d9713414b9b..00209d7bb19 100644
--- a/2.7/generated/torch.nn.functional.softsign.html
+++ b/2.7/generated/torch.nn.functional.softsign.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.softsign
                                    
 
                                    
 
                                    
-torch.nn.functional.softsign(input) → Tensor[source][source]
                                    
+torch.nn.functional.softsign(input) → Tensor[source][source]
 
                                    Applies element-wise, the function SoftSign(x)=x1+∣x∣\text{SoftSign}(x) = \frac{x}{1 + |x|}
                                    
 
                                    See Softsign for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.tanh.html b/2.7/generated/torch.nn.functional.tanh.html
index fbe5f64e4e6..0864ca82c72 100644
--- a/2.7/generated/torch.nn.functional.tanh.html
+++ b/2.7/generated/torch.nn.functional.tanh.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.tanh
                                    
 
                                    
 
                                    
-torch.nn.functional.tanh(input) → Tensor[source][source]
                                    
+torch.nn.functional.tanh(input) → Tensor[source][source]
 
                                    Applies element-wise,
 Tanh(x)=tanh⁡(x)=exp⁡(x)−exp⁡(−x)exp⁡(x)+exp⁡(−x)\text{Tanh}(x) = \tanh(x) = \frac{\exp(x) - \exp(-x)}{\exp(x) + \exp(-x)}
                                    
 
                                    See Tanh for more details.
                                    
diff --git a/2.7/generated/torch.nn.functional.tanhshrink.html b/2.7/generated/torch.nn.functional.tanhshrink.html
index a10826a849f..2a46cdb358e 100644
--- a/2.7/generated/torch.nn.functional.tanhshrink.html
+++ b/2.7/generated/torch.nn.functional.tanhshrink.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.tanhshrink
                                    
 
                                    
 
                                    
-torch.nn.functional.tanhshrink(input) → Tensor[source][source]
                                    
+torch.nn.functional.tanhshrink(input) → Tensor[source][source]
 
                                    Applies element-wise, Tanhshrink(x)=x−Tanh(x)\text{Tanhshrink}(x) = x - \text{Tanh}(x)
                                    
 
                                    See Tanhshrink for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.threshold.html b/2.7/generated/torch.nn.functional.threshold.html
index 88066bd979d..5bbf06ac6aa 100644
--- a/2.7/generated/torch.nn.functional.threshold.html
+++ b/2.7/generated/torch.nn.functional.threshold.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.threshold
                                    
 
                                    
 
                                    
-torch.nn.functional.threshold(input, threshold, value, inplace=False)[source]
                                    
+torch.nn.functional.threshold(input, threshold, value, inplace=False)[source]
 
                                    Apply a threshold to each element of the input Tensor.
                                    
 
                                    See Threshold for more details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.torch.nn.parallel.data_parallel.html b/2.7/generated/torch.nn.functional.torch.nn.parallel.data_parallel.html
index cdaff8dbf71..a682e233ce7 100644
--- a/2.7/generated/torch.nn.functional.torch.nn.parallel.data_parallel.html
+++ b/2.7/generated/torch.nn.functional.torch.nn.parallel.data_parallel.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.torch.nn.parallel.data_parallel
                                    
 
                                    
 
                                    
-torch.nn.parallel.data_parallel(module, inputs, device_ids=None, output_device=None, dim=0, module_kwargs=None)[source][source]
                                    
+torch.nn.parallel.data_parallel(module, inputs, device_ids=None, output_device=None, dim=0, module_kwargs=None)[source][source]
 
                                    Evaluate module(input) in parallel across the GPUs given in device_ids.
                                    
 
                                    This is the functional version of the DataParallel module.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.triplet_margin_loss.html b/2.7/generated/torch.nn.functional.triplet_margin_loss.html
index 0ca636d3869..d6d8ded453e 100644
--- a/2.7/generated/torch.nn.functional.triplet_margin_loss.html
+++ b/2.7/generated/torch.nn.functional.triplet_margin_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.triplet_margin_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.triplet_margin_loss(anchor, positive, negative, margin=1.0, p=2, eps=1e-06, swap=False, size_average=None, reduce=None, reduction='mean')[source][source]
                                    
+torch.nn.functional.triplet_margin_loss(anchor, positive, negative, margin=1.0, p=2, eps=1e-06, swap=False, size_average=None, reduce=None, reduction='mean')[source][source]
 
                                    Compute the triplet loss between given input tensors and a margin greater than 0.
                                    
 
                                    See TripletMarginLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.triplet_margin_with_distance_loss.html b/2.7/generated/torch.nn.functional.triplet_margin_with_distance_loss.html
index b47bd98ac25..09e792c139f 100644
--- a/2.7/generated/torch.nn.functional.triplet_margin_with_distance_loss.html
+++ b/2.7/generated/torch.nn.functional.triplet_margin_with_distance_loss.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.triplet_margin_with_distance_loss
                                    
 
                                    
 
                                    
-torch.nn.functional.triplet_margin_with_distance_loss(anchor, positive, negative, *, distance_function=None, margin=1.0, swap=False, reduction='mean')[source][source]
                                    
+torch.nn.functional.triplet_margin_with_distance_loss(anchor, positive, negative, *, distance_function=None, margin=1.0, swap=False, reduction='mean')[source][source]
 
                                    Compute the triplet margin loss for input tensors using a custom distance function.
                                    
 
                                    See TripletMarginWithDistanceLoss for details.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.functional.unfold.html b/2.7/generated/torch.nn.functional.unfold.html
index 14b7f13650f..f1c4804e336 100644
--- a/2.7/generated/torch.nn.functional.unfold.html
+++ b/2.7/generated/torch.nn.functional.unfold.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.unfold
                                    
 
                                    
 
                                    
-torch.nn.functional.unfold(input, kernel_size, dilation=1, padding=0, stride=1)[source][source]
                                    
+torch.nn.functional.unfold(input, kernel_size, dilation=1, padding=0, stride=1)[source][source]
 
                                    Extract sliding local blocks from a batched input tensor.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.functional.upsample.html b/2.7/generated/torch.nn.functional.upsample.html
index b4f3acfefc9..26c5049b461 100644
--- a/2.7/generated/torch.nn.functional.upsample.html
+++ b/2.7/generated/torch.nn.functional.upsample.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.upsample
                                    
 
                                    
 
                                    
-torch.nn.functional.upsample(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
                                    
+torch.nn.functional.upsample(input, size=None, scale_factor=None, mode='nearest', align_corners=None)[source][source]
 
                                    Upsample input.
                                    
 
                                    Provided tensor is upsampled to either the given size or the given
 scale_factor
                                    
diff --git a/2.7/generated/torch.nn.functional.upsample_bilinear.html b/2.7/generated/torch.nn.functional.upsample_bilinear.html
index b478a638422..215ff2f774c 100644
--- a/2.7/generated/torch.nn.functional.upsample_bilinear.html
+++ b/2.7/generated/torch.nn.functional.upsample_bilinear.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.upsample_bilinear
                                    
 
                                    
 
                                    
-torch.nn.functional.upsample_bilinear(input, size=None, scale_factor=None)[source][source]
                                    
+torch.nn.functional.upsample_bilinear(input, size=None, scale_factor=None)[source][source]
 
                                    Upsamples the input, using bilinear upsampling.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.functional.upsample_nearest.html b/2.7/generated/torch.nn.functional.upsample_nearest.html
index b2f237ebffe..2c121bd897b 100644
--- a/2.7/generated/torch.nn.functional.upsample_nearest.html
+++ b/2.7/generated/torch.nn.functional.upsample_nearest.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.functional.upsample_nearest
                                    
 
                                    
 
                                    
-torch.nn.functional.upsample_nearest(input, size=None, scale_factor=None)[source][source]
                                    
+torch.nn.functional.upsample_nearest(input, size=None, scale_factor=None)[source][source]
 
                                    Upsamples the input, using nearest neighbours’ pixel values.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.lazy.LazyModuleMixin.html b/2.7/generated/torch.nn.modules.lazy.LazyModuleMixin.html
index 39ae210579c..87bdf0ac894 100644
--- a/2.7/generated/torch.nn.modules.lazy.LazyModuleMixin.html
+++ b/2.7/generated/torch.nn.modules.lazy.LazyModuleMixin.html
@@ -596,7 +596,7 @@
 
                                    LazyModuleMixin
                                    
 
                                    
 
                                    
-class torch.nn.modules.lazy.LazyModuleMixin(*args, **kwargs)[source][source]
                                    
+class torch.nn.modules.lazy.LazyModuleMixin(*args, **kwargs)[source][source]
 
                                    A mixin for modules that lazily initialize parameters, also known as “lazy modules”.
                                    
 
                                    Modules that lazily initialize parameters, or “lazy modules”,
 derive the shapes of their parameters from the first input(s)
@@ -704,7 +704,7 @@ 
                                    LazyModuleMixin
 
                                    
-has_uninitialized_params()[source][source]
                                    
+has_uninitialized_params()[source][source]
 
                                    Check if a module has parameters that are not initialized.
                                    
 
                                    
 
                                    
@@ -712,7 +712,7 @@ 
                                    LazyModuleMixin
 
                                    
-initialize_parameters(*args, **kwargs)[source][source]
                                    
+initialize_parameters(*args, **kwargs)[source][source]
 
                                    Initialize parameters according to the input batch properties.
                                    
 
                                    This adds an interface to isolate parameter initialization from the
 forward pass when doing parameter shape inference.
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_backward_hook.html b/2.7/generated/torch.nn.modules.module.register_module_backward_hook.html
index 591098a4bc7..5fa0f96b08b 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_backward_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_backward_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_backward_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_backward_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_backward_hook(hook)[source][source]
 
                                    Register a backward hook common to all the modules.
                                    
 
                                    This function is deprecated in favor of
 torch.nn.modules.module.register_module_full_backward_hook()
diff --git a/2.7/generated/torch.nn.modules.module.register_module_buffer_registration_hook.html b/2.7/generated/torch.nn.modules.module.register_module_buffer_registration_hook.html
index afc960f481d..33cb5eadf3a 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_buffer_registration_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_buffer_registration_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_buffer_registration_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_buffer_registration_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_buffer_registration_hook(hook)[source][source]
 
                                    Register a buffer registration hook common to all modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_forward_hook.html b/2.7/generated/torch.nn.modules.module.register_module_forward_hook.html
index 6439d7b8d10..b8209e37811 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_forward_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_forward_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_forward_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_forward_hook(hook, *, with_kwargs=False, always_call=False)[source][source]
                                    
+torch.nn.modules.module.register_module_forward_hook(hook, *, with_kwargs=False, always_call=False)[source][source]
 
                                    Register a global forward hook for all the modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_forward_pre_hook.html b/2.7/generated/torch.nn.modules.module.register_module_forward_pre_hook.html
index fa2dcbd803c..75712795546 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_forward_pre_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_forward_pre_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_forward_pre_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_forward_pre_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_forward_pre_hook(hook)[source][source]
 
                                    Register a forward pre-hook common to all modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_full_backward_hook.html b/2.7/generated/torch.nn.modules.module.register_module_full_backward_hook.html
index 49e5c914b4e..af94834715c 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_full_backward_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_full_backward_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_full_backward_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_full_backward_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_full_backward_hook(hook)[source][source]
 
                                    Register a backward hook common to all the modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_full_backward_pre_hook.html b/2.7/generated/torch.nn.modules.module.register_module_full_backward_pre_hook.html
index a797c7e03c8..eff01c23bfd 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_full_backward_pre_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_full_backward_pre_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_full_backward_pre_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_full_backward_pre_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_full_backward_pre_hook(hook)[source][source]
 
                                    Register a backward pre-hook common to all the modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_module_registration_hook.html b/2.7/generated/torch.nn.modules.module.register_module_module_registration_hook.html
index d5fa29005c8..7bfa54126e2 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_module_registration_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_module_registration_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_module_registration_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_module_registration_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_module_registration_hook(hook)[source][source]
 
                                    Register a module registration hook common to all modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.module.register_module_parameter_registration_hook.html b/2.7/generated/torch.nn.modules.module.register_module_parameter_registration_hook.html
index fc145adb3e1..bb3b91bb591 100644
--- a/2.7/generated/torch.nn.modules.module.register_module_parameter_registration_hook.html
+++ b/2.7/generated/torch.nn.modules.module.register_module_parameter_registration_hook.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.modules.module.register_module_parameter_registration_hook
                                    
 
                                    
 
                                    
-torch.nn.modules.module.register_module_parameter_registration_hook(hook)[source][source]
                                    
+torch.nn.modules.module.register_module_parameter_registration_hook(hook)[source][source]
 
                                    Register a parameter registration hook common to all modules.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.modules.normalization.RMSNorm.html b/2.7/generated/torch.nn.modules.normalization.RMSNorm.html
index 94b519fd4e5..06dda21e781 100644
--- a/2.7/generated/torch.nn.modules.normalization.RMSNorm.html
+++ b/2.7/generated/torch.nn.modules.normalization.RMSNorm.html
@@ -596,7 +596,7 @@
 
                                    RMSNorm
                                    
 
                                    
 
                                    
-class torch.nn.modules.normalization.RMSNorm(normalized_shape, eps=None, elementwise_affine=True, device=None, dtype=None)[source][source]
                                    
+class torch.nn.modules.normalization.RMSNorm(normalized_shape, eps=None, elementwise_affine=True, device=None, dtype=None)[source][source]
 
                                    Applies Root Mean Square Layer Normalization over a mini-batch of inputs.
                                    
 
                                    This layer implements the operation as described in
 the paper Root Mean Square Layer Normalization
                                    
@@ -645,7 +645,7 @@ 
                                    RMSNorm
 
                                    
-extra_repr()[source][source]
                                    
+extra_repr()[source][source]
 
                                    Extra information about the module.
                                    
 
                                    
 
                                    Return type
                                    
@@ -656,7 +656,7 @@ 
                                    RMSNorm
 
                                    
-forward(x)[source][source]
                                    
+forward(x)[source][source]
 
                                    Runs forward pass.
                                    
 
                                    
 
                                    Return type
                                    
@@ -667,7 +667,7 @@ 
                                    RMSNorm
 
                                    
-reset_parameters()[source][source]
                                    
+reset_parameters()[source][source]
 
                                    Resets parameters based on their initialization used in __init__.
                                    
 
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.parallel.DistributedDataParallel.html b/2.7/generated/torch.nn.parallel.DistributedDataParallel.html
index 8f6d4e3e1d5..b85b7b02175 100644
--- a/2.7/generated/torch.nn.parallel.DistributedDataParallel.html
+++ b/2.7/generated/torch.nn.parallel.DistributedDataParallel.html
@@ -596,7 +596,7 @@
 
                                    DistributedDataParallel
                                    
 
                                    
 
                                    
-class torch.nn.parallel.DistributedDataParallel(module, device_ids=None, output_device=None, dim=0, broadcast_buffers=True, init_sync=True, process_group=None, bucket_cap_mb=None, find_unused_parameters=False, check_reduction=False, gradient_as_bucket_view=False, static_graph=False, delay_all_reduce_named_params=None, param_to_hook_all_reduce=None, mixed_precision=None, device_mesh=None)[source][source]
                                    
+class torch.nn.parallel.DistributedDataParallel(module, device_ids=None, output_device=None, dim=0, broadcast_buffers=True, init_sync=True, process_group=None, bucket_cap_mb=None, find_unused_parameters=False, check_reduction=False, gradient_as_bucket_view=False, static_graph=False, delay_all_reduce_named_params=None, param_to_hook_all_reduce=None, mixed_precision=None, device_mesh=None)[source][source]
 
                                    Implement distributed data parallelism based on torch.distributed at module level.
                                    
 
                                    This container provides data parallelism by synchronizing gradients
 across each model replica. The devices to synchronize across are
@@ -920,7 +920,7 @@ 
                                    DistributedDataParallel
 
                                    
 
                                    
-join(divide_by_initial_world_size=True, enable=True, throw_on_early_termination=False)[source][source]
                                    
+join(divide_by_initial_world_size=True, enable=True, throw_on_early_termination=False)[source][source]
 
                                    Context manager for training with uneven inputs across processes in DDP.
                                    
 
                                    This context manager will keep track of already-joined DDP processes,
 and “shadow” the forward and backward passes by inserting collective
@@ -1016,7 +1016,7 @@ 
                                    DistributedDataParallel
 
                                    
-join_hook(**kwargs)[source][source]
                                    
+join_hook(**kwargs)[source][source]
 
                                    DDP join hook enables training on uneven inputs by mirroring communications in forward and backward passes.
                                    
 
                                    
 
                                    Parameters
                                    
@@ -1045,7 +1045,7 @@ 
                                    DistributedDataParallel
 
                                    
-no_sync()[source][source]
                                    
+no_sync()[source][source]
 
                                    Context manager to disable gradient synchronizations across DDP processes.
                                    
 
                                    Within this context, gradients will be accumulated on module
 variables, which will later be synchronized in the first
@@ -1067,7 +1067,7 @@ 
                                    DistributedDataParallel
 
                                    
-register_comm_hook(state, hook)[source][source]
                                    
+register_comm_hook(state, hook)[source][source]
 
                                    Register communication hook for user-defined DDP aggregation of gradients across multiple workers.
                                    
 
                                    This hook would be very useful for researchers to try out new ideas. For
 example, this hook can be used to implement several algorithms like GossipGrad
diff --git a/2.7/generated/torch.nn.parameter.Buffer.html b/2.7/generated/torch.nn.parameter.Buffer.html
index 8f4a879652c..ae6d44a3a40 100644
--- a/2.7/generated/torch.nn.parameter.Buffer.html
+++ b/2.7/generated/torch.nn.parameter.Buffer.html
@@ -596,7 +596,7 @@
 
                                    Buffer
                                    
 
                                    
 
                                    
-class torch.nn.parameter.Buffer(data=None, *, persistent=True)[source][source]
                                    
+class torch.nn.parameter.Buffer(data=None, *, persistent=True)[source][source]
 
                                    A kind of Tensor that should not be considered a model
 parameter. For example, BatchNorm’s running_mean is not a parameter, but is part of the module’s state.
                                    
 
                                    Buffers are Tensor subclasses, that have a
diff --git a/2.7/generated/torch.nn.parameter.Parameter.html b/2.7/generated/torch.nn.parameter.Parameter.html
index 9ffd5c8c3a9..9e5f1876f50 100644
--- a/2.7/generated/torch.nn.parameter.Parameter.html
+++ b/2.7/generated/torch.nn.parameter.Parameter.html
@@ -596,7 +596,7 @@
 
                                    Parameter
                                    
 
                                    
 
                                    
-class torch.nn.parameter.Parameter(data=None, requires_grad=True)[source][source]
                                    
+class torch.nn.parameter.Parameter(data=None, requires_grad=True)[source][source]
 
                                    A kind of Tensor that is to be considered a module parameter.
                                    
 
                                    Parameters are Tensor subclasses, that have a
 very special property when used with Module s - when they’re
diff --git a/2.7/generated/torch.nn.parameter.UninitializedBuffer.html b/2.7/generated/torch.nn.parameter.UninitializedBuffer.html
index 29bfd55e684..114f0538fa5 100644
--- a/2.7/generated/torch.nn.parameter.UninitializedBuffer.html
+++ b/2.7/generated/torch.nn.parameter.UninitializedBuffer.html
@@ -596,7 +596,7 @@
 
                                    UninitializedBuffer
                                    
 
                                    
 
                                    
-class torch.nn.parameter.UninitializedBuffer(requires_grad=False, device=None, dtype=None, persistent=True)[source][source]
                                    
+class torch.nn.parameter.UninitializedBuffer(requires_grad=False, device=None, dtype=None, persistent=True)[source][source]
 
                                    A buffer that is not initialized.
                                    
 
                                    Uninitialized Buffer is a a special case of torch.Tensor
 where the shape of the data is still unknown.
                                    
diff --git a/2.7/generated/torch.nn.parameter.UninitializedParameter.html b/2.7/generated/torch.nn.parameter.UninitializedParameter.html
index e9756a0f0b8..159b112ef3e 100644
--- a/2.7/generated/torch.nn.parameter.UninitializedParameter.html
+++ b/2.7/generated/torch.nn.parameter.UninitializedParameter.html
@@ -596,7 +596,7 @@
 
                                    UninitializedParameter
                                    
 
                                    
 
                                    
-class torch.nn.parameter.UninitializedParameter(requires_grad=True, device=None, dtype=None)[source][source]
                                    
+class torch.nn.parameter.UninitializedParameter(requires_grad=True, device=None, dtype=None)[source][source]
 
                                    A parameter that is not initialized.
                                    
 
                                    Uninitialized Parameters are a special case of torch.nn.Parameter
 where the shape of the data is still unknown.
                                    
@@ -611,7 +611,7 @@ 
                                    UninitializedParameter
 
                                    
 
                                    
-cls_to_become[source]
                                    
+cls_to_become[source]
 
                                    alias of Parameter
                                    
 
                                    
 
diff --git a/2.7/generated/torch.nn.utils.clip_grad_norm.html b/2.7/generated/torch.nn.utils.clip_grad_norm.html
index 771d2542fd4..2c4d8ffc9a8 100644
--- a/2.7/generated/torch.nn.utils.clip_grad_norm.html
+++ b/2.7/generated/torch.nn.utils.clip_grad_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.clip_grad_norm
                                    
 
                                    
 
                                    
-torch.nn.utils.clip_grad_norm(parameters, max_norm, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source][source]
                                    
+torch.nn.utils.clip_grad_norm(parameters, max_norm, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source][source]
 
                                    Clip the gradient norm of an iterable of parameters.
                                    
 
                                    
 
                                    Warning
                                    
diff --git a/2.7/generated/torch.nn.utils.clip_grad_norm_.html b/2.7/generated/torch.nn.utils.clip_grad_norm_.html
index 93a9d63d453..0bda8965e54 100644
--- a/2.7/generated/torch.nn.utils.clip_grad_norm_.html
+++ b/2.7/generated/torch.nn.utils.clip_grad_norm_.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.clip_grad_norm_
                                    
 
                                    
 
                                    
-torch.nn.utils.clip_grad_norm_(parameters, max_norm, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source][source]
                                    
+torch.nn.utils.clip_grad_norm_(parameters, max_norm, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source][source]
 
                                    Clip the gradient norm of an iterable of parameters.
                                    
 
                                    The norm is computed over the norms of the individual gradients of all parameters,
 as if the norms of the individual gradients were concatenated into a single vector.
diff --git a/2.7/generated/torch.nn.utils.clip_grad_value_.html b/2.7/generated/torch.nn.utils.clip_grad_value_.html
index de3b9f4b21a..33da240b18d 100644
--- a/2.7/generated/torch.nn.utils.clip_grad_value_.html
+++ b/2.7/generated/torch.nn.utils.clip_grad_value_.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.clip_grad_value_
                                    
 
                                    
 
                                    
-torch.nn.utils.clip_grad_value_(parameters, clip_value, foreach=None)[source][source]
                                    
+torch.nn.utils.clip_grad_value_(parameters, clip_value, foreach=None)[source][source]
 
                                    Clip the gradients of an iterable of parameters at specified value.
                                    
 
                                    Gradients are modified in-place.
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.utils.clip_grads_with_norm_.html b/2.7/generated/torch.nn.utils.clip_grads_with_norm_.html
index c1ca0365647..cfd9d4d8c54 100644
--- a/2.7/generated/torch.nn.utils.clip_grads_with_norm_.html
+++ b/2.7/generated/torch.nn.utils.clip_grads_with_norm_.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.clip_grads_with_norm_
                                    
 
                                    
 
                                    
-torch.nn.utils.clip_grads_with_norm_(parameters, max_norm, total_norm, foreach=None)[source]
                                    
+torch.nn.utils.clip_grads_with_norm_(parameters, max_norm, total_norm, foreach=None)[source]
 
                                    Scale the gradients of an iterable of parameters given a pre-calculated total norm and desired max norm.
                                    
 
                                    The gradients will be scaled by the following calculation
                                    
 
                                    
diff --git a/2.7/generated/torch.nn.utils.convert_conv2d_weight_memory_format.html b/2.7/generated/torch.nn.utils.convert_conv2d_weight_memory_format.html
index 75d1a9cb874..e71135bf8a4 100644
--- a/2.7/generated/torch.nn.utils.convert_conv2d_weight_memory_format.html
+++ b/2.7/generated/torch.nn.utils.convert_conv2d_weight_memory_format.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.convert_conv2d_weight_memory_format
                                    
 
                                    
 
                                    
-torch.nn.utils.convert_conv2d_weight_memory_format(module, memory_format)[source][source]
                                    
+torch.nn.utils.convert_conv2d_weight_memory_format(module, memory_format)[source][source]
 
                                    Convert memory_format of nn.Conv2d.weight to memory_format.
                                    
 
                                    The conversion recursively applies to nested nn.Module, including module.
 Note that it only changes the memory_format, but not the semantics of each dimensions.
diff --git a/2.7/generated/torch.nn.utils.convert_conv3d_weight_memory_format.html b/2.7/generated/torch.nn.utils.convert_conv3d_weight_memory_format.html
index e05e17699a6..117189b6d15 100644
--- a/2.7/generated/torch.nn.utils.convert_conv3d_weight_memory_format.html
+++ b/2.7/generated/torch.nn.utils.convert_conv3d_weight_memory_format.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.convert_conv3d_weight_memory_format
                                    
 
                                    
 
                                    
-torch.nn.utils.convert_conv3d_weight_memory_format(module, memory_format)[source][source]
                                    
+torch.nn.utils.convert_conv3d_weight_memory_format(module, memory_format)[source][source]
 
                                    Convert memory_format of nn.Conv3d.weight to memory_format
 The conversion recursively applies to nested nn.Module, including module.
 Note that it only changes the memory_format, but not the semantics of each dimensions.
diff --git a/2.7/generated/torch.nn.utils.fuse_conv_bn_eval.html b/2.7/generated/torch.nn.utils.fuse_conv_bn_eval.html
index 3ab2c8aead7..460638c0b22 100644
--- a/2.7/generated/torch.nn.utils.fuse_conv_bn_eval.html
+++ b/2.7/generated/torch.nn.utils.fuse_conv_bn_eval.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.fuse_conv_bn_eval
                                    
 
                                    
 
                                    
-torch.nn.utils.fuse_conv_bn_eval(conv, bn, transpose=False)[source][source]
                                    
+torch.nn.utils.fuse_conv_bn_eval(conv, bn, transpose=False)[source][source]
 
                                    Fuse a convolutional module and a BatchNorm module into a single, new convolutional module.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.fuse_conv_bn_weights.html b/2.7/generated/torch.nn.utils.fuse_conv_bn_weights.html
index 7fb1d32f594..9049c196607 100644
--- a/2.7/generated/torch.nn.utils.fuse_conv_bn_weights.html
+++ b/2.7/generated/torch.nn.utils.fuse_conv_bn_weights.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.fuse_conv_bn_weights
                                    
 
                                    
 
                                    
-torch.nn.utils.fuse_conv_bn_weights(conv_w, conv_b, bn_rm, bn_rv, bn_eps, bn_w, bn_b, transpose=False)[source][source]
                                    
+torch.nn.utils.fuse_conv_bn_weights(conv_w, conv_b, bn_rm, bn_rv, bn_eps, bn_w, bn_b, transpose=False)[source][source]
 
                                    Fuse convolutional module parameters and BatchNorm module parameters into new convolutional module parameters.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.fuse_linear_bn_eval.html b/2.7/generated/torch.nn.utils.fuse_linear_bn_eval.html
index 164050fe41a..76cc579e60b 100644
--- a/2.7/generated/torch.nn.utils.fuse_linear_bn_eval.html
+++ b/2.7/generated/torch.nn.utils.fuse_linear_bn_eval.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.fuse_linear_bn_eval
                                    
 
                                    
 
                                    
-torch.nn.utils.fuse_linear_bn_eval(linear, bn)[source][source]
                                    
+torch.nn.utils.fuse_linear_bn_eval(linear, bn)[source][source]
 
                                    Fuse a linear module and a BatchNorm module into a single, new linear module.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.fuse_linear_bn_weights.html b/2.7/generated/torch.nn.utils.fuse_linear_bn_weights.html
index 570209db36f..1cdc7b16dfc 100644
--- a/2.7/generated/torch.nn.utils.fuse_linear_bn_weights.html
+++ b/2.7/generated/torch.nn.utils.fuse_linear_bn_weights.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.fuse_linear_bn_weights
                                    
 
                                    
 
                                    
-torch.nn.utils.fuse_linear_bn_weights(linear_w, linear_b, bn_rm, bn_rv, bn_eps, bn_w, bn_b)[source][source]
                                    
+torch.nn.utils.fuse_linear_bn_weights(linear_w, linear_b, bn_rm, bn_rv, bn_eps, bn_w, bn_b)[source][source]
 
                                    Fuse linear module parameters and BatchNorm module parameters into new linear module parameters.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.get_total_norm.html b/2.7/generated/torch.nn.utils.get_total_norm.html
index cb4d2234763..7673a76e8d5 100644
--- a/2.7/generated/torch.nn.utils.get_total_norm.html
+++ b/2.7/generated/torch.nn.utils.get_total_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.get_total_norm
                                    
 
                                    
 
                                    
-torch.nn.utils.get_total_norm(tensors, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source]
                                    
+torch.nn.utils.get_total_norm(tensors, norm_type=2.0, error_if_nonfinite=False, foreach=None)[source]
 
                                    Compute the norm of an iterable of tensors.
                                    
 
                                    The norm is computed over the norms of the individual tensors, as if the norms of
 the individual tensors were concatenated into a single vector.
                                    
diff --git a/2.7/generated/torch.nn.utils.parameters_to_vector.html b/2.7/generated/torch.nn.utils.parameters_to_vector.html
index 1f03136ec2a..8ec58c184ff 100644
--- a/2.7/generated/torch.nn.utils.parameters_to_vector.html
+++ b/2.7/generated/torch.nn.utils.parameters_to_vector.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parameters_to_vector
                                    
 
                                    
 
                                    
-torch.nn.utils.parameters_to_vector(parameters)[source][source]
                                    
+torch.nn.utils.parameters_to_vector(parameters)[source][source]
 
                                    Flatten an iterable of parameters into a single vector.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.parametrizations.orthogonal.html b/2.7/generated/torch.nn.utils.parametrizations.orthogonal.html
index ddf4a3cfd6c..e469ef18fe2 100644
--- a/2.7/generated/torch.nn.utils.parametrizations.orthogonal.html
+++ b/2.7/generated/torch.nn.utils.parametrizations.orthogonal.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrizations.orthogonal
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrizations.orthogonal(module, name='weight', orthogonal_map=None, *, use_trivialization=True)[source][source]
                                    
+torch.nn.utils.parametrizations.orthogonal(module, name='weight', orthogonal_map=None, *, use_trivialization=True)[source][source]
 
                                    Apply an orthogonal or unitary parametrization to a matrix or a batch of matrices.
                                    
 
                                    Letting K\mathbb{K} be R\mathbb{R} or C\mathbb{C}, the parametrized
 matrix Q∈Km×nQ \in \mathbb{K}^{m \times n} is orthogonal as
                                    
diff --git a/2.7/generated/torch.nn.utils.parametrizations.spectral_norm.html b/2.7/generated/torch.nn.utils.parametrizations.spectral_norm.html
index fc17c25d036..24e84142cb6 100644
--- a/2.7/generated/torch.nn.utils.parametrizations.spectral_norm.html
+++ b/2.7/generated/torch.nn.utils.parametrizations.spectral_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrizations.spectral_norm
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrizations.spectral_norm(module, name='weight', n_power_iterations=1, eps=1e-12, dim=None)[source][source]
                                    
+torch.nn.utils.parametrizations.spectral_norm(module, name='weight', n_power_iterations=1, eps=1e-12, dim=None)[source][source]
 
                                    Apply spectral normalization to a parameter in the given module.
                                    
 
                                    
 WSN=Wσ(W),σ(W)=max⁡h:h≠0∥Wh∥2∥h∥2\mathbf{W}_{SN} = \dfrac{\mathbf{W}}{\sigma(\mathbf{W})},
diff --git a/2.7/generated/torch.nn.utils.parametrizations.weight_norm.html b/2.7/generated/torch.nn.utils.parametrizations.weight_norm.html
index 2c91ece41fa..764de7e2882 100644
--- a/2.7/generated/torch.nn.utils.parametrizations.weight_norm.html
+++ b/2.7/generated/torch.nn.utils.parametrizations.weight_norm.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrizations.weight_norm
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrizations.weight_norm(module, name='weight', dim=0)[source][source]
                                    
+torch.nn.utils.parametrizations.weight_norm(module, name='weight', dim=0)[source][source]
 
                                    Apply weight normalization to a parameter in the given module.
                                    
 
                                    
 w=gv∥v∥\mathbf{w} = g \dfrac{\mathbf{v}}{\|\mathbf{v}\|}
diff --git a/2.7/generated/torch.nn.utils.parametrize.ParametrizationList.html b/2.7/generated/torch.nn.utils.parametrize.ParametrizationList.html
index c33a741fc07..ee097ec8433 100644
--- a/2.7/generated/torch.nn.utils.parametrize.ParametrizationList.html
+++ b/2.7/generated/torch.nn.utils.parametrize.ParametrizationList.html
@@ -596,7 +596,7 @@
 
                                    ParametrizationList
                                    
 
                                    
 
                                    
-class torch.nn.utils.parametrize.ParametrizationList(modules, original, unsafe=False)[source][source]
                                    
+class torch.nn.utils.parametrize.ParametrizationList(modules, original, unsafe=False)[source][source]
 
                                    A sequential container that holds and manages the original parameters or buffers of a parametrized torch.nn.Module.
                                    
 
                                    It is the type of module.parametrizations[tensor_name] when module[tensor_name]
 has been parametrized with register_parametrization().
                                    
@@ -624,7 +624,7 @@ 
                                    ParametrizationList
 
                                    
-right_inverse(value)[source][source]
                                    
+right_inverse(value)[source][source]
 
                                    Call the right_inverse methods of the parametrizations in the inverse registration order.
                                    
 
                                    Then, it stores the result in self.original if right_inverse outputs one tensor
 or in self.original0, self.original1, … if it outputs several.
                                    
diff --git a/2.7/generated/torch.nn.utils.parametrize.cached.html b/2.7/generated/torch.nn.utils.parametrize.cached.html
index fa659c790af..1955db91984 100644
--- a/2.7/generated/torch.nn.utils.parametrize.cached.html
+++ b/2.7/generated/torch.nn.utils.parametrize.cached.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrize.cached
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrize.cached()[source][source]
                                    
+torch.nn.utils.parametrize.cached()[source][source]
 
                                    Context manager that enables the caching system within parametrizations registered with register_parametrization().
                                    
 
                                    The value of the parametrized objects is computed and cached the first time
 they are required when this context manager is active. The cached values are
diff --git a/2.7/generated/torch.nn.utils.parametrize.is_parametrized.html b/2.7/generated/torch.nn.utils.parametrize.is_parametrized.html
index 1b444e0e6e7..4cab096fa92 100644
--- a/2.7/generated/torch.nn.utils.parametrize.is_parametrized.html
+++ b/2.7/generated/torch.nn.utils.parametrize.is_parametrized.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrize.is_parametrized
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrize.is_parametrized(module, tensor_name=None)[source][source]
                                    
+torch.nn.utils.parametrize.is_parametrized(module, tensor_name=None)[source][source]
 
                                    Determine if a module has a parametrization.
                                    
 
                                    
 
                                    Parameters
                                    
diff --git a/2.7/generated/torch.nn.utils.parametrize.register_parametrization.html b/2.7/generated/torch.nn.utils.parametrize.register_parametrization.html
index 3695e67c4eb..79d6f0e8126 100644
--- a/2.7/generated/torch.nn.utils.parametrize.register_parametrization.html
+++ b/2.7/generated/torch.nn.utils.parametrize.register_parametrization.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrize.register_parametrization
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrize.register_parametrization(module, tensor_name, parametrization, *, unsafe=False)[source][source]
                                    
+torch.nn.utils.parametrize.register_parametrization(module, tensor_name, parametrization, *, unsafe=False)[source][source]
 
                                    Register a parametrization to a tensor in a module.
                                    
 
                                    Assume that tensor_name="weight" for simplicity. When accessing module.weight,
 the module will return the parametrized version parametrization(module.weight).
diff --git a/2.7/generated/torch.nn.utils.parametrize.remove_parametrizations.html b/2.7/generated/torch.nn.utils.parametrize.remove_parametrizations.html
index 9775b860783..22058aa3e24 100644
--- a/2.7/generated/torch.nn.utils.parametrize.remove_parametrizations.html
+++ b/2.7/generated/torch.nn.utils.parametrize.remove_parametrizations.html
@@ -596,7 +596,7 @@
 
                                    torch.nn.utils.parametrize.remove_parametrizations
                                    
 
                                    
 
                                    
-torch.nn.utils.parametrize.remove_parametrizations(module, tensor_name, leave_parametrized=True)[source][source]
                                    
+torch.nn.utils.parametrize.remove_parametrizations(module, tensor_name, leave_parametrized=True)[source][source]
 
                                    Remove the parametrizations on a tensor in a module.
                                    
 
                                      
 
                                    • If leave_parametrized=True, module[tensor_name] will be set to
diff --git a/2.7/generated/torch.nn.utils.prune.BasePruningMethod.html b/2.7/generated/torch.nn.utils.prune.BasePruningMethod.html
index 0635d71bc37..94390df89dd 100644
--- a/2.7/generated/torch.nn.utils.prune.BasePruningMethod.html
+++ b/2.7/generated/torch.nn.utils.prune.BasePruningMethod.html
@@ -596,13 +596,13 @@
 
                                      BasePruningMethod
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.BasePruningMethod[source][source]
                                      
+class torch.nn.utils.prune.BasePruningMethod[source][source]
 
                                      Abstract base class for creation of new pruning techniques.
                                      
 
                                      Provides a skeleton for customization requiring the overriding of methods
 such as compute_mask() and apply().
                                      
 
                                      
 
                                      
-classmethod apply(module, name, *args, importance_scores=None, **kwargs)[source][source]
                                      
+classmethod apply(module, name, *args, importance_scores=None, **kwargs)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -629,7 +629,7 @@ 
                                      BasePruningMethod
 
                                      
-apply_mask(module)[source][source]
                                      
+apply_mask(module)[source][source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -648,7 +648,7 @@ 
                                      BasePruningMethod
 
                                      
-abstract compute_mask(t, default_mask)[source][source]
                                      
+abstract compute_mask(t, default_mask)[source][source]
 
                                      Compute and returns a mask for the input tensor t.
                                      
 
                                      Starting from a base default_mask (which should be a mask of ones
 if the tensor has not been pruned yet), generate a random mask to
@@ -676,7 +676,7 @@ 
                                      BasePruningMethod
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source][source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source][source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -703,7 +703,7 @@ 
                                      BasePruningMethod
 
                                      
-remove(module)[source][source]
                                      
+remove(module)[source][source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.CustomFromMask.html b/2.7/generated/torch.nn.utils.prune.CustomFromMask.html
index 5d8900f8300..f6da677c18a 100644
--- a/2.7/generated/torch.nn.utils.prune.CustomFromMask.html
+++ b/2.7/generated/torch.nn.utils.prune.CustomFromMask.html
@@ -596,10 +596,10 @@
 
                                      CustomFromMask
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.CustomFromMask(mask)[source][source]
                                      
+class torch.nn.utils.prune.CustomFromMask(mask)[source][source]
 
                                      
 
                                      
-classmethod apply(module, name, mask)[source][source]
                                      
+classmethod apply(module, name, mask)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -617,7 +617,7 @@ 
                                      CustomFromMask
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -636,7 +636,7 @@ 
                                      CustomFromMask
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -663,7 +663,7 @@ 
                                      CustomFromMask
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.Identity.html b/2.7/generated/torch.nn.utils.prune.Identity.html
index 49b33e9beea..25dcb7158c3 100644
--- a/2.7/generated/torch.nn.utils.prune.Identity.html
+++ b/2.7/generated/torch.nn.utils.prune.Identity.html
@@ -596,11 +596,11 @@
 
                                      Identity
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.Identity[source][source]
                                      
+class torch.nn.utils.prune.Identity[source][source]
 
                                      Utility pruning method that does not prune any units but generates the pruning parametrization with a mask of ones.
                                      
 
                                      
 
                                      
-classmethod apply(module, name)[source][source]
                                      
+classmethod apply(module, name)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -618,7 +618,7 @@ 
                                      Identity
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -637,7 +637,7 @@ 
                                      Identity
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -664,7 +664,7 @@ 
                                      Identity
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.L1Unstructured.html b/2.7/generated/torch.nn.utils.prune.L1Unstructured.html
index 91a2e94605d..784f9ec69be 100644
--- a/2.7/generated/torch.nn.utils.prune.L1Unstructured.html
+++ b/2.7/generated/torch.nn.utils.prune.L1Unstructured.html
@@ -596,7 +596,7 @@
 
                                      L1Unstructured
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.L1Unstructured(amount)[source][source]
                                      
+class torch.nn.utils.prune.L1Unstructured(amount)[source][source]
 
                                      Prune (currently unpruned) units in a tensor by zeroing out the ones with the lowest L1-norm.
                                      
 
                                      
 
                                      Parameters
                                      
@@ -608,7 +608,7 @@ 
                                      L1Unstructured
 
                                      
-classmethod apply(module, name, amount, importance_scores=None)[source][source]
                                      
+classmethod apply(module, name, amount, importance_scores=None)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -635,7 +635,7 @@ 
                                      L1Unstructured
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -654,7 +654,7 @@ 
                                      L1Unstructured
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -681,7 +681,7 @@ 
                                      L1Unstructured
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.LnStructured.html b/2.7/generated/torch.nn.utils.prune.LnStructured.html
index 3a5f2cf898b..a80c3c0a974 100644
--- a/2.7/generated/torch.nn.utils.prune.LnStructured.html
+++ b/2.7/generated/torch.nn.utils.prune.LnStructured.html
@@ -596,7 +596,7 @@
 
                                      LnStructured
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.LnStructured(amount, n, dim=-1)[source][source]
                                      
+class torch.nn.utils.prune.LnStructured(amount, n, dim=-1)[source][source]
 
                                      Prune entire (currently unpruned) channels in a tensor based on their Ln-norm.
                                      
 
                                      
 
                                      Parameters
                                      
@@ -614,7 +614,7 @@ 
                                      LnStructured
 
                                      
-classmethod apply(module, name, amount, n, dim, importance_scores=None)[source][source]
                                      
+classmethod apply(module, name, amount, n, dim, importance_scores=None)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -645,7 +645,7 @@ 
                                      LnStructured
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -664,7 +664,7 @@ 
                                      LnStructured
 
                                      
-compute_mask(t, default_mask)[source][source]
                                      
+compute_mask(t, default_mask)[source][source]
 
                                      Compute and returns a mask for the input tensor t.
                                      
 
                                      Starting from a base default_mask (which should be a mask of ones
 if the tensor has not been pruned yet), generate a mask to apply on
@@ -693,7 +693,7 @@ 
                                      LnStructured
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -720,7 +720,7 @@ 
                                      LnStructured
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.PruningContainer.html b/2.7/generated/torch.nn.utils.prune.PruningContainer.html
index 75a8c9b261f..f70d7e97220 100644
--- a/2.7/generated/torch.nn.utils.prune.PruningContainer.html
+++ b/2.7/generated/torch.nn.utils.prune.PruningContainer.html
@@ -596,7 +596,7 @@
 
                                      PruningContainer
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.PruningContainer(*args)[source][source]
                                      
+class torch.nn.utils.prune.PruningContainer(*args)[source][source]
 
                                      Container holding a sequence of pruning methods for iterative pruning.
                                      
 
                                      Keeps track of the order in which pruning methods are applied and handles
 combining successive pruning calls.
                                      
@@ -604,7 +604,7 @@ 
                                      PruningContainer
 
                                      
-add_pruning_method(method)[source][source]
                                      
+add_pruning_method(method)[source][source]
 
                                      Add a child pruning method to the container.
                                      
 
                                      
 
                                      Parameters
                                      
@@ -616,7 +616,7 @@ 
                                      PruningContainer
 
                                      
-classmethod apply(module, name, *args, importance_scores=None, **kwargs)[source]
                                      
+classmethod apply(module, name, *args, importance_scores=None, **kwargs)[source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -643,7 +643,7 @@ 
                                      PruningContainer
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -662,7 +662,7 @@ 
                                      PruningContainer
 
                                      
-compute_mask(t, default_mask)[source][source]
                                      
+compute_mask(t, default_mask)[source][source]
 
                                      Apply the latest method by computing the new partial masks and returning its combination with the default_mask.
                                      
 
                                      The new partial mask should be computed on the entries or channels
 that were not zeroed out by the default_mask.
@@ -697,7 +697,7 @@ 
                                      PruningContainer
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -724,7 +724,7 @@ 
                                      PruningContainer
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.RandomStructured.html b/2.7/generated/torch.nn.utils.prune.RandomStructured.html
index e6e3f9a5ea0..ed31fbd32c5 100644
--- a/2.7/generated/torch.nn.utils.prune.RandomStructured.html
+++ b/2.7/generated/torch.nn.utils.prune.RandomStructured.html
@@ -596,7 +596,7 @@
 
                                      RandomStructured
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.RandomStructured(amount, dim=-1)[source][source]
                                      
+class torch.nn.utils.prune.RandomStructured(amount, dim=-1)[source][source]
 
                                      Prune entire (currently unpruned) channels in a tensor at random.
                                      
 
                                      
 
                                      Parameters
                                      
@@ -612,7 +612,7 @@ 
                                      RandomStructured
 
                                      
-classmethod apply(module, name, amount, dim=-1)[source][source]
                                      
+classmethod apply(module, name, amount, dim=-1)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -636,7 +636,7 @@ 
                                      RandomStructured
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -655,7 +655,7 @@ 
                                      RandomStructured
 
                                      
-compute_mask(t, default_mask)[source][source]
                                      
+compute_mask(t, default_mask)[source][source]
 
                                      Compute and returns a mask for the input tensor t.
                                      
 
                                      Starting from a base default_mask (which should be a mask of ones
 if the tensor has not been pruned yet), generate a random mask to
@@ -684,7 +684,7 @@ 
                                      RandomStructured
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -711,7 +711,7 @@ 
                                      RandomStructured
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.RandomUnstructured.html b/2.7/generated/torch.nn.utils.prune.RandomUnstructured.html
index 544ba75c343..48e900b6d44 100644
--- a/2.7/generated/torch.nn.utils.prune.RandomUnstructured.html
+++ b/2.7/generated/torch.nn.utils.prune.RandomUnstructured.html
@@ -596,7 +596,7 @@
 
                                      RandomUnstructured
                                      
 
                                      
 
                                      
-class torch.nn.utils.prune.RandomUnstructured(amount)[source][source]
                                      
+class torch.nn.utils.prune.RandomUnstructured(amount)[source][source]
 
                                      Prune (currently unpruned) units in a tensor at random.
                                      
 
                                      
 
                                      Parameters
                                      
@@ -612,7 +612,7 @@ 
                                      RandomUnstructured
 
                                      
-classmethod apply(module, name, amount)[source][source]
                                      
+classmethod apply(module, name, amount)[source][source]
 
                                      Add pruning on the fly and reparametrization of a tensor.
                                      
 
                                      Adds the forward pre-hook that enables pruning on the fly and
 the reparametrization of a tensor in terms of the original tensor
@@ -634,7 +634,7 @@ 
                                      RandomUnstructured
 
                                      
-apply_mask(module)[source]
                                      
+apply_mask(module)[source]
 
                                      Simply handles the multiplication between the parameter being pruned and the generated mask.
                                      
 
                                      Fetches the mask and the original tensor from the module
 and returns the pruned version of the tensor.
                                      
@@ -653,7 +653,7 @@ 
                                      RandomUnstructured
 
                                      
-prune(t, default_mask=None, importance_scores=None)[source]
                                      
+prune(t, default_mask=None, importance_scores=None)[source]
 
                                      Compute and returns a pruned version of input tensor t.
                                      
 
                                      According to the pruning rule specified in compute_mask().
                                      
 
                                      
@@ -680,7 +680,7 @@ 
                                      RandomUnstructured
 
                                      
-remove(module)[source]
                                      
+remove(module)[source]
 
                                      Remove the pruning reparameterization from a module.
                                      
 
                                      The pruned parameter named name remains permanently pruned,
 and the parameter named name+'_orig' is removed from the parameter list.
diff --git a/2.7/generated/torch.nn.utils.prune.custom_from_mask.html b/2.7/generated/torch.nn.utils.prune.custom_from_mask.html
index 89b5d273c9f..aab9328ff58 100644
--- a/2.7/generated/torch.nn.utils.prune.custom_from_mask.html
+++ b/2.7/generated/torch.nn.utils.prune.custom_from_mask.html
@@ -596,7 +596,7 @@
 
                                      torch.nn.utils.prune.custom_from_mask
                                      
 
                                      
 
                                      
-torch.nn.utils.prune.custom_from_mask(module, name, mask)[source][source]
                                      
+torch.nn.utils.prune.custom_from_mask(module, name, mask)[source][source]
 
                                      Prune tensor corresponding to parameter called name in module by applying the pre-computed mask in mask.
                                      
 
                                      Modifies module in place (and also return the modified module) by:
                                      
 
                                        
diff --git a/2.7/generated/torch.nn.utils.prune.global_unstructured.html b/2.7/generated/torch.nn.utils.prune.global_unstructured.html
index c58b16c2a74..bdc3f8dce16 100644
--- a/2.7/generated/torch.nn.utils.prune.global_unstructured.html
+++ b/2.7/generated/torch.nn.utils.prune.global_unstructured.html
@@ -596,7 +596,7 @@
 
                                        torch.nn.utils.prune.global_unstructured
                                        
 
                                        
 
                                        
-torch.nn.utils.prune.global_unstructured(parameters, pruning_method, importance_scores=None, **kwargs)[source][source]
                                        
+torch.nn.utils.prune.global_unstructured(parameters, pruning_method, importance_scores=None, **kwargs)[source][source]
 
                                        Globally prunes tensors corresponding to all parameters in parameters by applying the specified pruning_method.
                                        
 
                                        Modifies modules in place by:
                                        
 
                                          
diff --git a/2.7/generated/torch.nn.utils.prune.identity.html b/2.7/generated/torch.nn.utils.prune.identity.html
index 65e1920005b..bafc0cc4f87 100644
--- a/2.7/generated/torch.nn.utils.prune.identity.html
+++ b/2.7/generated/torch.nn.utils.prune.identity.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.identity
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.identity(module, name)[source][source]
                                          
+torch.nn.utils.prune.identity(module, name)[source][source]
 
                                          Apply pruning reparametrization without pruning any units.
                                          
 
                                          Applies pruning reparametrization to the tensor corresponding to the
 parameter called name in module without actually pruning any
diff --git a/2.7/generated/torch.nn.utils.prune.is_pruned.html b/2.7/generated/torch.nn.utils.prune.is_pruned.html
index f1f639e5e02..dd87f1b7cf0 100644
--- a/2.7/generated/torch.nn.utils.prune.is_pruned.html
+++ b/2.7/generated/torch.nn.utils.prune.is_pruned.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.is_pruned
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.is_pruned(module)[source][source]
                                          
+torch.nn.utils.prune.is_pruned(module)[source][source]
 
                                          Check if a module is pruned by looking for pruning pre-hooks.
                                          
 
                                          Check whether module is pruned by looking for
 forward_pre_hooks in its modules that inherit from the
diff --git a/2.7/generated/torch.nn.utils.prune.l1_unstructured.html b/2.7/generated/torch.nn.utils.prune.l1_unstructured.html
index 84adb813da2..a510d35e97e 100644
--- a/2.7/generated/torch.nn.utils.prune.l1_unstructured.html
+++ b/2.7/generated/torch.nn.utils.prune.l1_unstructured.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.l1_unstructured
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.l1_unstructured(module, name, amount, importance_scores=None)[source][source]
                                          
+torch.nn.utils.prune.l1_unstructured(module, name, amount, importance_scores=None)[source][source]
 
                                          Prune tensor by removing units with the lowest L1-norm.
                                          
 
                                          Prunes tensor corresponding to parameter called name in module
 by removing the specified amount of (currently unpruned) units with the
diff --git a/2.7/generated/torch.nn.utils.prune.ln_structured.html b/2.7/generated/torch.nn.utils.prune.ln_structured.html
index eab966e9bb2..dd155c0b4ff 100644
--- a/2.7/generated/torch.nn.utils.prune.ln_structured.html
+++ b/2.7/generated/torch.nn.utils.prune.ln_structured.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.ln_structured
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.ln_structured(module, name, amount, n, dim, importance_scores=None)[source][source]
                                          
+torch.nn.utils.prune.ln_structured(module, name, amount, n, dim, importance_scores=None)[source][source]
 
                                          Prune tensor by removing channels with the lowest Ln-norm along the specified dimension.
                                          
 
                                          Prunes tensor corresponding to parameter called name in module
 by removing the specified amount of (currently unpruned) channels
diff --git a/2.7/generated/torch.nn.utils.prune.random_structured.html b/2.7/generated/torch.nn.utils.prune.random_structured.html
index b93fe08b2e5..150eaa69a89 100644
--- a/2.7/generated/torch.nn.utils.prune.random_structured.html
+++ b/2.7/generated/torch.nn.utils.prune.random_structured.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.random_structured
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.random_structured(module, name, amount, dim)[source][source]
                                          
+torch.nn.utils.prune.random_structured(module, name, amount, dim)[source][source]
 
                                          Prune tensor by removing random channels along the specified dimension.
                                          
 
                                          Prunes tensor corresponding to parameter called name in module
 by removing the specified amount of (currently unpruned) channels
diff --git a/2.7/generated/torch.nn.utils.prune.random_unstructured.html b/2.7/generated/torch.nn.utils.prune.random_unstructured.html
index a19dc50dfb3..aeee0ece553 100644
--- a/2.7/generated/torch.nn.utils.prune.random_unstructured.html
+++ b/2.7/generated/torch.nn.utils.prune.random_unstructured.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.random_unstructured
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.random_unstructured(module, name, amount)[source][source]
                                          
+torch.nn.utils.prune.random_unstructured(module, name, amount)[source][source]
 
                                          Prune tensor by removing random (currently unpruned) units.
                                          
 
                                          Prunes tensor corresponding to parameter called name in module
 by removing the specified amount of (currently unpruned) units
diff --git a/2.7/generated/torch.nn.utils.prune.remove.html b/2.7/generated/torch.nn.utils.prune.remove.html
index 61f20de3bef..d035f06c36a 100644
--- a/2.7/generated/torch.nn.utils.prune.remove.html
+++ b/2.7/generated/torch.nn.utils.prune.remove.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.prune.remove
                                          
 
                                          
 
                                          
-torch.nn.utils.prune.remove(module, name)[source][source]
                                          
+torch.nn.utils.prune.remove(module, name)[source][source]
 
                                          Remove the pruning reparameterization from a module and the pruning method from the forward hook.
                                          
 
                                          The pruned parameter named name remains permanently pruned, and the parameter
 named name+'_orig' is removed from the parameter list. Similarly,
diff --git a/2.7/generated/torch.nn.utils.remove_spectral_norm.html b/2.7/generated/torch.nn.utils.remove_spectral_norm.html
index 14e3c30cd27..af051c57a31 100644
--- a/2.7/generated/torch.nn.utils.remove_spectral_norm.html
+++ b/2.7/generated/torch.nn.utils.remove_spectral_norm.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.remove_spectral_norm
                                          
 
                                          
 
                                          
-torch.nn.utils.remove_spectral_norm(module, name='weight')[source][source]
                                          
+torch.nn.utils.remove_spectral_norm(module, name='weight')[source][source]
 
                                          Remove the spectral normalization reparameterization from a module.
                                          
 
                                          
 
                                          Parameters
                                          
diff --git a/2.7/generated/torch.nn.utils.remove_weight_norm.html b/2.7/generated/torch.nn.utils.remove_weight_norm.html
index 1cb012c138e..b1d0598c27d 100644
--- a/2.7/generated/torch.nn.utils.remove_weight_norm.html
+++ b/2.7/generated/torch.nn.utils.remove_weight_norm.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.remove_weight_norm
                                          
 
                                          
 
                                          
-torch.nn.utils.remove_weight_norm(module, name='weight')[source][source]
                                          
+torch.nn.utils.remove_weight_norm(module, name='weight')[source][source]
 
                                          Remove the weight normalization reparameterization from a module.
                                          
 
                                          
 
                                          Parameters
                                          
diff --git a/2.7/generated/torch.nn.utils.rnn.PackedSequence.html b/2.7/generated/torch.nn.utils.rnn.PackedSequence.html
index 8c60a5b057c..54c37295a37 100644
--- a/2.7/generated/torch.nn.utils.rnn.PackedSequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.PackedSequence.html
@@ -596,7 +596,7 @@
 
                                          PackedSequence
                                          
 
                                          
 
                                          
-class torch.nn.utils.rnn.PackedSequence(data, batch_sizes=None, sorted_indices=None, unsorted_indices=None)[source][source]
                                          
+class torch.nn.utils.rnn.PackedSequence(data, batch_sizes=None, sorted_indices=None, unsorted_indices=None)[source][source]
 
                                          Holds the data and list of batch_sizes of a packed sequence.
                                          
 
                                          All RNN modules accept packed sequences as inputs.
                                          
 
                                          
@@ -668,7 +668,7 @@ 
                                          PackedSequence
 
                                          
-is_pinned()[source][source]
                                          
+is_pinned()[source][source]
 
                                          Return true if self.data stored on in pinned memory.
                                          
 
                                          
 
                                          Return type
                                          
@@ -685,7 +685,7 @@ 
                                          PackedSequence
 
                                          
-to(dtype: dtype, non_blocking: bool = ..., copy: bool = ...) → Self[source][source]
                                          
+to(dtype: dtype, non_blocking: bool = ..., copy: bool = ...) → Self[source][source]
 
                                          
 to(device: Optional[Union[str, device, int]] = ..., dtype: Optional[dtype] = ..., non_blocking: bool = ..., copy: bool = ...) → Self
                                          
 
                                          
diff --git a/2.7/generated/torch.nn.utils.rnn.pack_padded_sequence.html b/2.7/generated/torch.nn.utils.rnn.pack_padded_sequence.html
index 1edff0832dd..54c72997c26 100644
--- a/2.7/generated/torch.nn.utils.rnn.pack_padded_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.pack_padded_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.pack_padded_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.pack_padded_sequence(input, lengths, batch_first=False, enforce_sorted=True)[source][source]
                                          
+torch.nn.utils.rnn.pack_padded_sequence(input, lengths, batch_first=False, enforce_sorted=True)[source][source]
                                          
 
                                          Packs a Tensor containing padded sequences of variable length.
                                          
 
                                          input can be of size T x B x * (if batch_first is False)
 or B x T x * (if batch_first is True) where T is the length
diff --git a/2.7/generated/torch.nn.utils.rnn.pack_sequence.html b/2.7/generated/torch.nn.utils.rnn.pack_sequence.html
index 9341b3fe715..60fab98f7e2 100644
--- a/2.7/generated/torch.nn.utils.rnn.pack_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.pack_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.pack_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.pack_sequence(sequences, enforce_sorted=True)[source][source]
                                          
+torch.nn.utils.rnn.pack_sequence(sequences, enforce_sorted=True)[source][source]
 
                                          Packs a list of variable length Tensors.
                                          
 
                                          Consecutive call of the next functions: pad_sequence, pack_padded_sequence.
                                          
 
                                          sequences should be a list of Tensors of size L x *, where L is
diff --git a/2.7/generated/torch.nn.utils.rnn.pad_packed_sequence.html b/2.7/generated/torch.nn.utils.rnn.pad_packed_sequence.html
index f604ddfd892..b9faec3f33d 100644
--- a/2.7/generated/torch.nn.utils.rnn.pad_packed_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.pad_packed_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.pad_packed_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.pad_packed_sequence(sequence, batch_first=False, padding_value=0.0, total_length=None)[source][source]
                                          
+torch.nn.utils.rnn.pad_packed_sequence(sequence, batch_first=False, padding_value=0.0, total_length=None)[source][source]
 
                                          Pad a packed batch of variable length sequences.
                                          
 
                                          It is an inverse operation to pack_padded_sequence().
                                          
 
                                          The returned Tensor’s data will be of size T x B x * (if batch_first is False)
diff --git a/2.7/generated/torch.nn.utils.rnn.pad_sequence.html b/2.7/generated/torch.nn.utils.rnn.pad_sequence.html
index 5460ed90409..056cf0d15a4 100644
--- a/2.7/generated/torch.nn.utils.rnn.pad_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.pad_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.pad_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.pad_sequence(sequences, batch_first=False, padding_value=0.0, padding_side='right')[source][source]
                                          
+torch.nn.utils.rnn.pad_sequence(sequences, batch_first=False, padding_value=0.0, padding_side='right')[source][source]
 
                                          Pad a list of variable length Tensors with padding_value.
                                          
 
                                          pad_sequence stacks a list of Tensors along a new dimension, and pads them
 to equal length. sequences can be list of sequences with size L x *,
diff --git a/2.7/generated/torch.nn.utils.rnn.unpack_sequence.html b/2.7/generated/torch.nn.utils.rnn.unpack_sequence.html
index 1119b3bf9d7..866527f00e9 100644
--- a/2.7/generated/torch.nn.utils.rnn.unpack_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.unpack_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.unpack_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.unpack_sequence(packed_sequences)[source][source]
                                          
+torch.nn.utils.rnn.unpack_sequence(packed_sequences)[source][source]
 
                                          Unpack PackedSequence into a list of variable length Tensors.
                                          
 
                                          packed_sequences should be a PackedSequence object.
                                          
 
                                          Example
                                          
diff --git a/2.7/generated/torch.nn.utils.rnn.unpad_sequence.html b/2.7/generated/torch.nn.utils.rnn.unpad_sequence.html
index 166f90e52d3..a852b3667f5 100644
--- a/2.7/generated/torch.nn.utils.rnn.unpad_sequence.html
+++ b/2.7/generated/torch.nn.utils.rnn.unpad_sequence.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.rnn.unpad_sequence
                                          
 
                                          
 
                                          
-torch.nn.utils.rnn.unpad_sequence(padded_sequences, lengths, batch_first=False)[source][source]
                                          
+torch.nn.utils.rnn.unpad_sequence(padded_sequences, lengths, batch_first=False)[source][source]
 
                                          Unpad padded Tensor into a list of variable length Tensors.
                                          
 
                                          unpad_sequence unstacks padded Tensor into a list of variable length Tensors.
                                          
 
                                          Example
                                          
diff --git a/2.7/generated/torch.nn.utils.skip_init.html b/2.7/generated/torch.nn.utils.skip_init.html
index 22c9d9d1a4c..6cb2078db70 100644
--- a/2.7/generated/torch.nn.utils.skip_init.html
+++ b/2.7/generated/torch.nn.utils.skip_init.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.skip_init
                                          
 
                                          
 
                                          
-torch.nn.utils.skip_init(module_cls, *args, **kwargs)[source][source]
                                          
+torch.nn.utils.skip_init(module_cls, *args, **kwargs)[source][source]
 
                                          Given a module class object and args / kwargs, instantiate the module without initializing parameters / buffers.
                                          
 
                                          This can be useful if initialization is slow or if custom initialization will
 be performed, making the default initialization unnecessary. There are some caveats to this, due to
diff --git a/2.7/generated/torch.nn.utils.spectral_norm.html b/2.7/generated/torch.nn.utils.spectral_norm.html
index db718e6c7f5..8887335169b 100644
--- a/2.7/generated/torch.nn.utils.spectral_norm.html
+++ b/2.7/generated/torch.nn.utils.spectral_norm.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.spectral_norm
                                          
 
                                          
 
                                          
-torch.nn.utils.spectral_norm(module, name='weight', n_power_iterations=1, eps=1e-12, dim=None)[source][source]
                                          
+torch.nn.utils.spectral_norm(module, name='weight', n_power_iterations=1, eps=1e-12, dim=None)[source][source]
 
                                          Apply spectral normalization to a parameter in the given module.
                                          
 
                                          
 WSN=Wσ(W),σ(W)=max⁡h:h≠0∥Wh∥2∥h∥2\mathbf{W}_{SN} = \dfrac{\mathbf{W}}{\sigma(\mathbf{W})},
diff --git a/2.7/generated/torch.nn.utils.stateless.functional_call.html b/2.7/generated/torch.nn.utils.stateless.functional_call.html
index 2f65f59c1ab..eb458c2d2ec 100644
--- a/2.7/generated/torch.nn.utils.stateless.functional_call.html
+++ b/2.7/generated/torch.nn.utils.stateless.functional_call.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.stateless.functional_call
                                          
 
                                          
 
                                          
-torch.nn.utils.stateless.functional_call(module, parameters_and_buffers, args=None, kwargs=None, *, tie_weights=True, strict=False)[source][source]
                                          
+torch.nn.utils.stateless.functional_call(module, parameters_and_buffers, args=None, kwargs=None, *, tie_weights=True, strict=False)[source][source]
 
                                          Perform a functional call on the module by replacing the module parameters and buffers with the provided ones.
                                          
 
                                          
 
                                          Warning
                                          
diff --git a/2.7/generated/torch.nn.utils.vector_to_parameters.html b/2.7/generated/torch.nn.utils.vector_to_parameters.html
index ad9f53cceab..1ca048b4827 100644
--- a/2.7/generated/torch.nn.utils.vector_to_parameters.html
+++ b/2.7/generated/torch.nn.utils.vector_to_parameters.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.vector_to_parameters
                                          
 
                                          
 
                                          
-torch.nn.utils.vector_to_parameters(vec, parameters)[source][source]
                                          
+torch.nn.utils.vector_to_parameters(vec, parameters)[source][source]
 
                                          Copy slices of a vector into an iterable of parameters.
                                          
 
                                          
 
                                          Parameters
                                          
diff --git a/2.7/generated/torch.nn.utils.weight_norm.html b/2.7/generated/torch.nn.utils.weight_norm.html
index 7ca777cfc9f..3889b4baae4 100644
--- a/2.7/generated/torch.nn.utils.weight_norm.html
+++ b/2.7/generated/torch.nn.utils.weight_norm.html
@@ -596,7 +596,7 @@
 
                                          torch.nn.utils.weight_norm
                                          
 
                                          
 
                                          
-torch.nn.utils.weight_norm(module, name='weight', dim=0)[source][source]
                                          
+torch.nn.utils.weight_norm(module, name='weight', dim=0)[source][source]
 
                                          Apply weight normalization to a parameter in the given module.
                                          
 
                                          
 w=gv∥v∥\mathbf{w} = g \dfrac{\mathbf{v}}{\|\mathbf{v}\|}
diff --git a/2.7/generated/torch.no_grad.html b/2.7/generated/torch.no_grad.html
index 4676c977c64..0b23e6887b9 100644
--- a/2.7/generated/torch.no_grad.html
+++ b/2.7/generated/torch.no_grad.html
@@ -596,7 +596,7 @@
 
                                          no_grad
                                          
 
                                          
 
                                          
-class torch.no_grad(orig_func=None)[source][source]
                                          
+class torch.no_grad(orig_func=None)[source][source]
 
                                          Context-manager that disables gradient calculation.
                                          
 
                                          Disabling gradient calculation is useful for inference, when you are sure
 that you will not call Tensor.backward(). It will reduce memory
diff --git a/2.7/generated/torch.norm.html b/2.7/generated/torch.norm.html
index a982b296bc3..730c2cc4e69 100644
--- a/2.7/generated/torch.norm.html
+++ b/2.7/generated/torch.norm.html
@@ -596,7 +596,7 @@
 
                                          torch.norm
                                          
 
                                          
 
                                          
-torch.norm(input, p='fro', dim=None, keepdim=False, out=None, dtype=None)[source][source]
                                          
+torch.norm(input, p='fro', dim=None, keepdim=False, out=None, dtype=None)[source][source]
 
                                          Returns the matrix norm or vector norm of a given tensor.
                                          
 
                                          
 
                                          Warning
                                          
diff --git a/2.7/generated/torch.onnx.JitScalarType.html b/2.7/generated/torch.onnx.JitScalarType.html
index b64647ef350..f2768048a17 100644
--- a/2.7/generated/torch.onnx.JitScalarType.html
+++ b/2.7/generated/torch.onnx.JitScalarType.html
@@ -616,7 +616,7 @@ 
                                          JitScalarType
 
                                          
-dtype()[source][source]
                                          
+dtype()[source][source]
 
                                          Convert a JitScalarType to a torch dtype.
                                          
 
                                          
 
                                          Return type
                                          
@@ -627,7 +627,7 @@ 
                                          JitScalarType
 
                                          
-classmethod from_dtype(dtype)[source][source]
                                          
+classmethod from_dtype(dtype)[source][source]
 
                                          Convert a torch dtype to JitScalarType.
                                          
 
                                          
 
                                          Note: DO NOT USE this API when dtype comes from a torch._C.Value.type() calls.
                                          A “RuntimeError: INTERNAL ASSERT FAILED at “../aten/src/ATen/core/jit_type_base.h” can
@@ -653,7 +653,7 @@ 
                                          JitScalarType
 
                                          
-classmethod from_onnx_type(onnx_type)[source][source]
                                          
+classmethod from_onnx_type(onnx_type)[source][source]
 
                                          Convert a ONNX data type to JitScalarType.
                                          
 
                                          
 
                                          Parameters
                                          
@@ -673,7 +673,7 @@ 
                                          JitScalarType
 
                                          
-classmethod from_value(value, default=None)[source][source]
                                          
+classmethod from_value(value, default=None)[source][source]
 
                                          Create a JitScalarType from an value’s scalar type.
                                          
 
                                          
 
                                          Parameters
                                          
@@ -699,7 +699,7 @@ 
                                          JitScalarType
 
                                          
-onnx_compatible()[source][source]
                                          
+onnx_compatible()[source][source]
 
                                          Return whether this JitScalarType is compatible with ONNX.
                                          
 
                                          
 
                                          Return type
                                          
@@ -710,7 +710,7 @@ 
                                          JitScalarType
 
                                          
-onnx_type()[source][source]
                                          
+onnx_type()[source][source]
 
                                          Convert a JitScalarType to an ONNX data type.
                                          
 
                                          
 
                                          Return type
                                          
@@ -721,7 +721,7 @@ 
                                          JitScalarType
 
                                          
-scalar_name()[source][source]
                                          
+scalar_name()[source][source]
 
                                          Convert a JitScalarType to a JIT scalar type name.
                                          
 
                                          
 
                                          Return type
                                          
@@ -732,7 +732,7 @@ 
                                          JitScalarType
 
                                          
-torch_name()[source][source]
                                          
+torch_name()[source][source]
 
                                          Convert a JitScalarType to a torch type name.
                                          
 
                                          
 
                                          Return type
                                          
diff --git a/2.7/generated/torch.optim.ASGD.html b/2.7/generated/torch.optim.ASGD.html
index e8c16bc98d5..fa4b9cd1595 100644
--- a/2.7/generated/torch.optim.ASGD.html
+++ b/2.7/generated/torch.optim.ASGD.html
@@ -596,7 +596,7 @@
 
                                          ASGD
                                          
 
                                          
 
                                          
-class torch.optim.ASGD(params, lr=0.01, lambd=0.0001, alpha=0.75, t0=1000000.0, weight_decay=0, foreach=None, maximize=False, differentiable=False, capturable=False)[source][source]
                                          
+class torch.optim.ASGD(params, lr=0.01, lambd=0.0001, alpha=0.75, t0=1000000.0, weight_decay=0, foreach=None, maximize=False, differentiable=False, capturable=False)[source][source]
 
                                          Implements Averaged Stochastic Gradient Descent.
                                          
 
                                          It has been proposed in Acceleration of stochastic approximation by
 averaging.
                                          
@@ -634,7 +634,7 @@ 
                                          ASGD¶<
 
                                          
 
                                          
 
                                          
-add_param_group(param_group)[source]
                                          
+add_param_group(param_group)[source]
 
                                          Add a param group to the Optimizer s param_groups.
                                          
 
                                          This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                          
@@ -648,7 +648,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-load_state_dict(state_dict)[source]
                                          
+load_state_dict(state_dict)[source]
 
                                          Load the optimizer state.
                                          
 
                                          
 
                                          Parameters
                                          
@@ -672,7 +672,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                          
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                          Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                          
@@ -706,7 +706,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                          
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                          Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                          
@@ -743,7 +743,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_state_dict_post_hook(hook, prepend=False)[source]
                                          
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                          Register a state dict post-hook which will be called after state_dict() is called.
                                          
 
                                          It should have the following signature:
                                          
 
                                          hook(optimizer, state_dict) -> state_dict or None
@@ -775,7 +775,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                          
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                          Register a state dict pre-hook which will be called before state_dict() is called.
                                          
 
                                          It should have the following signature:
                                          
 
                                          hook(optimizer) -> None
@@ -807,7 +807,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_step_post_hook(hook)[source]
                                          
+register_step_post_hook(hook)[source]
 
                                          Register an optimizer step post hook which will be called after optimizer step.
                                          
 
                                          It should have the following signature:
                                          
 
                                          hook(optimizer, args, kwargs) -> None
@@ -830,7 +830,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-register_step_pre_hook(hook)[source]
                                          
+register_step_pre_hook(hook)[source]
 
                                          Register an optimizer step pre hook which will be called before optimizer step.
                                          
 
                                          It should have the following signature:
                                          
 
                                          hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -855,7 +855,7 @@ 
                                          ASGD¶<
 
 
                                          
 
                                          
-state_dict()[source]
                                          
+state_dict()[source]
 
                                          Return the state of the optimizer as a dict.
                                          
 
                                          It contains two entries:
                                          
 
                                            
@@ -918,7 +918,7 @@ 
                                            ASGD¶<
 
 
                                            
 
                                            
-step(closure=None)[source][source]
                                            
+step(closure=None)[source][source]
 
                                            Perform a single optimization step.
                                            
 
                                            
 
                                            Parameters
                                            
@@ -930,7 +930,7 @@ 
                                            ASGD¶<
 
 
                                            
 
                                            
-zero_grad(set_to_none=True)[source]
                                            
+zero_grad(set_to_none=True)[source]
 
                                            Reset the gradients of all optimized torch.Tensor s.
                                            
 
                                            
 
                                            Parameters
                                            
diff --git a/2.7/generated/torch.optim.Adadelta.html b/2.7/generated/torch.optim.Adadelta.html
index 9c298cb00e5..a4300d7d29b 100644
--- a/2.7/generated/torch.optim.Adadelta.html
+++ b/2.7/generated/torch.optim.Adadelta.html
@@ -596,7 +596,7 @@
 
                                            Adadelta
                                            
 
                                            
 
                                            
-class torch.optim.Adadelta(params, lr=1.0, rho=0.9, eps=1e-06, weight_decay=0, foreach=None, *, capturable=False, maximize=False, differentiable=False)[source][source]
                                            
+class torch.optim.Adadelta(params, lr=1.0, rho=0.9, eps=1e-06, weight_decay=0, foreach=None, *, capturable=False, maximize=False, differentiable=False)[source][source]
 
                                            Implements Adadelta algorithm.
                                            
 
                                            
 input:γ (lr), θ0 (params), f(θ) (objective), ρ (decay), λ (weight decay)initialize:v0←0  (square avg), u0←0  (accumulate variables)for t=1 to … dogt←∇θft(θt−1)if λ≠0gt←gt+λθt−1vt←vt−1ρ+gt2(1−ρ)Δxt←ut−1+ϵvt+ϵgtut←ut−1ρ+Δxt2(1−ρ)θt←θt−1−γΔxtreturn θt\begin{aligned}
@@ -683,7 +683,7 @@ 
                                            Adadelta
 
                                            
-add_param_group(param_group)[source]
                                            
+add_param_group(param_group)[source]
 
                                            Add a param group to the Optimizer s param_groups.
                                            
 
                                            This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                            
@@ -697,7 +697,7 @@ 
                                            Adadelta
 
                                            
-load_state_dict(state_dict)[source]
                                            
+load_state_dict(state_dict)[source]
 
                                            Load the optimizer state.
                                            
 
                                            
 
                                            Parameters
                                            
@@ -721,7 +721,7 @@ 
                                            Adadelta
 
                                            
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                            
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                            Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                            
@@ -755,7 +755,7 @@ 
                                            Adadelta
 
                                            
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                            
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                            Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                            
@@ -792,7 +792,7 @@ 
                                            Adadelta
 
                                            
-register_state_dict_post_hook(hook, prepend=False)[source]
                                            
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                            Register a state dict post-hook which will be called after state_dict() is called.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, state_dict) -> state_dict or None
@@ -824,7 +824,7 @@ 
                                            Adadelta
 
                                            
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                            
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                            Register a state dict pre-hook which will be called before state_dict() is called.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer) -> None
@@ -856,7 +856,7 @@ 
                                            Adadelta
 
                                            
-register_step_post_hook(hook)[source]
                                            
+register_step_post_hook(hook)[source]
 
                                            Register an optimizer step post hook which will be called after optimizer step.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, args, kwargs) -> None
@@ -879,7 +879,7 @@ 
                                            Adadelta
 
                                            
-register_step_pre_hook(hook)[source]
                                            
+register_step_pre_hook(hook)[source]
 
                                            Register an optimizer step pre hook which will be called before optimizer step.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -904,7 +904,7 @@ 
                                            Adadelta
 
                                            
-state_dict()[source]
                                            
+state_dict()[source]
 
                                            Return the state of the optimizer as a dict.
                                            
 
                                            It contains two entries:
                                            
 
                                            
 
                                            
 
                                            
-add_param_group(param_group)[source]
                                            
+add_param_group(param_group)[source]
 
                                            Add a param group to the Optimizer s param_groups.
                                            
 
                                            This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                            
@@ -837,7 +837,7 @@ 
                                            Adafactor
 
                                            
-load_state_dict(state_dict)[source]
                                            
+load_state_dict(state_dict)[source]
 
                                            Load the optimizer state.
                                            
 
                                            
 
                                            Parameters
                                            
@@ -861,7 +861,7 @@ 
                                            Adafactor
 
                                            
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                            
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                            Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                            
@@ -895,7 +895,7 @@ 
                                            Adafactor
 
                                            
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                            
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                            Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                            
@@ -932,7 +932,7 @@ 
                                            Adafactor
 
                                            
-register_state_dict_post_hook(hook, prepend=False)[source]
                                            
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                            Register a state dict post-hook which will be called after state_dict() is called.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, state_dict) -> state_dict or None
@@ -964,7 +964,7 @@ 
                                            Adafactor
 
                                            
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                            
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                            Register a state dict pre-hook which will be called before state_dict() is called.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer) -> None
@@ -996,7 +996,7 @@ 
                                            Adafactor
 
                                            
-register_step_post_hook(hook)[source]
                                            
+register_step_post_hook(hook)[source]
 
                                            Register an optimizer step post hook which will be called after optimizer step.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, args, kwargs) -> None
@@ -1019,7 +1019,7 @@ 
                                            Adafactor
 
                                            
-register_step_pre_hook(hook)[source]
                                            
+register_step_pre_hook(hook)[source]
 
                                            Register an optimizer step pre hook which will be called before optimizer step.
                                            
 
                                            It should have the following signature:
                                            
 
                                            hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -1044,7 +1044,7 @@ 
                                            Adafactor
 
                                            
-state_dict()[source]
                                            
+state_dict()[source]
 
                                            Return the state of the optimizer as a dict.
                                            
 
                                            It contains two entries:
                                            
 
                                              
@@ -1107,7 +1107,7 @@ 
                                              Adafactor
 
                                              
-step(closure=None)[source][source]
                                              
+step(closure=None)[source][source]
 
                                              Perform a single optimization step.
                                              
 
                                              
 
                                              Parameters
                                              
@@ -1119,7 +1119,7 @@ 
                                              Adafactor
 
                                              
-zero_grad(set_to_none=True)[source]
                                              
+zero_grad(set_to_none=True)[source]
 
                                              Reset the gradients of all optimized torch.Tensor s.
                                              
 
                                              
 
                                              Parameters
                                              
diff --git a/2.7/generated/torch.optim.Adagrad.html b/2.7/generated/torch.optim.Adagrad.html
index 2fe204db094..ba488e43b92 100644
--- a/2.7/generated/torch.optim.Adagrad.html
+++ b/2.7/generated/torch.optim.Adagrad.html
@@ -596,7 +596,7 @@
 
                                              Adagrad
                                              
 
                                              
 
                                              
-class torch.optim.Adagrad(params, lr=0.01, lr_decay=0, weight_decay=0, initial_accumulator_value=0, eps=1e-10, foreach=None, *, maximize=False, differentiable=False, fused=None)[source][source]
                                              
+class torch.optim.Adagrad(params, lr=0.01, lr_decay=0, weight_decay=0, initial_accumulator_value=0, eps=1e-10, foreach=None, *, maximize=False, differentiable=False, fused=None)[source][source]
 
                                              Implements Adagrad algorithm.
                                              
 
                                              
 input:γ (lr), θ0 (params), f(θ) (objective), λ (weight decay),τ (initial accumulator value), η (lr decay)initialize:state_sum0←τfor t=1 to … dogt←∇θft(θt−1)γ~←γ/(1+(t−1)η)if λ≠0gt←gt+λθt−1state_sumt←state_sumt−1+gt2θt←θt−1−γ~gtstate_sumt+ϵreturn θt\begin{aligned}
@@ -667,7 +667,7 @@ 
                                              Adagrad
 
                                              
-add_param_group(param_group)[source]
                                              
+add_param_group(param_group)[source]
 
                                              Add a param group to the Optimizer s param_groups.
                                              
 
                                              This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                              
@@ -681,7 +681,7 @@ 
                                              Adagrad
 
                                              
-load_state_dict(state_dict)[source]
                                              
+load_state_dict(state_dict)[source]
 
                                              Load the optimizer state.
                                              
 
                                              
 
                                              Parameters
                                              
@@ -705,7 +705,7 @@ 
                                              Adagrad
 
                                              
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                              
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                              Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                              
@@ -739,7 +739,7 @@ 
                                              Adagrad
 
                                              
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                              
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                              Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                              
@@ -776,7 +776,7 @@ 
                                              Adagrad
 
                                              
-register_state_dict_post_hook(hook, prepend=False)[source]
                                              
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                              Register a state dict post-hook which will be called after state_dict() is called.
                                              
 
                                              It should have the following signature:
                                              
 
                                              hook(optimizer, state_dict) -> state_dict or None
@@ -808,7 +808,7 @@ 
                                              Adagrad
 
                                              
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                              
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                              Register a state dict pre-hook which will be called before state_dict() is called.
                                              
 
                                              It should have the following signature:
                                              
 
                                              hook(optimizer) -> None
@@ -840,7 +840,7 @@ 
                                              Adagrad
 
                                              
-register_step_post_hook(hook)[source]
                                              
+register_step_post_hook(hook)[source]
 
                                              Register an optimizer step post hook which will be called after optimizer step.
                                              
 
                                              It should have the following signature:
                                              
 
                                              hook(optimizer, args, kwargs) -> None
@@ -863,7 +863,7 @@ 
                                              Adagrad
 
                                              
-register_step_pre_hook(hook)[source]
                                              
+register_step_pre_hook(hook)[source]
 
                                              Register an optimizer step pre hook which will be called before optimizer step.
                                              
 
                                              It should have the following signature:
                                              
 
                                              hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -888,7 +888,7 @@ 
                                              Adagrad
 
                                              
-state_dict()[source]
                                              
+state_dict()[source]
 
                                              Return the state of the optimizer as a dict.
                                              
 
                                              It contains two entries:
                                              
 
                                                
@@ -951,7 +951,7 @@ 
                                                Adagrad
 
                                                
-step(closure=None)[source][source]
                                                
+step(closure=None)[source][source]
 
                                                Perform a single optimization step.
                                                
 
                                                
 
                                                Parameters
                                                
@@ -963,7 +963,7 @@ 
                                                Adagrad
 
                                                
-zero_grad(set_to_none=True)[source]
                                                
+zero_grad(set_to_none=True)[source]
 
                                                Reset the gradients of all optimized torch.Tensor s.
                                                
 
                                                
 
                                                Parameters
                                                
diff --git a/2.7/generated/torch.optim.Adam.html b/2.7/generated/torch.optim.Adam.html
index 0e3cf8c34aa..ad9150f8541 100644
--- a/2.7/generated/torch.optim.Adam.html
+++ b/2.7/generated/torch.optim.Adam.html
@@ -596,7 +596,7 @@
 
                                                Adam
                                                
 
                                                
 
                                                
-class torch.optim.Adam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, amsgrad=False, *, foreach=None, maximize=False, capturable=False, differentiable=False, fused=None, decoupled_weight_decay=False)[source][source]
                                                
+class torch.optim.Adam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, amsgrad=False, *, foreach=None, maximize=False, capturable=False, differentiable=False, fused=None, decoupled_weight_decay=False)[source][source]
 
                                                Implements Adam algorithm.
                                                
 
                                                
 input:γ (lr),β1,β2 (betas),θ0 (params),f(θ) (objective)λ (weight decay), amsgrad, maximize, ϵ (epsilon)initialize:m0←0 ( first moment),v0←0 (second moment), v0max←0for t=1 to … doif maximize:gt←−∇θft(θt−1)elsegt←∇θft(θt−1)if λ≠0gt←gt+λθt−1mt←β1mt−1+(1−β1)gtvt←β2vt−1+(1−β2)gt2mt^←mt/(1−β1t)if amsgradvtmax←max(vt−1max,vt)vt^←vtmax/(1−β2t)elsevt^←vt/(1−β2t)θt←θt−1−γmt^/(vt^+ϵ)return θt\begin{aligned}
@@ -706,7 +706,7 @@ 
                                                Adam¶<
 
                                                
 
                                                
 
                                                
-add_param_group(param_group)[source]
                                                
+add_param_group(param_group)[source]
 
                                                Add a param group to the Optimizer s param_groups.
                                                
 
                                                This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                
@@ -720,7 +720,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-load_state_dict(state_dict)[source]
                                                
+load_state_dict(state_dict)[source]
 
                                                Load the optimizer state.
                                                
 
                                                
 
                                                Parameters
                                                
@@ -744,7 +744,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                
@@ -778,7 +778,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                
@@ -815,7 +815,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                Register a state dict post-hook which will be called after state_dict() is called.
                                                
 
                                                It should have the following signature:
                                                
 
                                                hook(optimizer, state_dict) -> state_dict or None
@@ -847,7 +847,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                Register a state dict pre-hook which will be called before state_dict() is called.
                                                
 
                                                It should have the following signature:
                                                
 
                                                hook(optimizer) -> None
@@ -879,7 +879,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_step_post_hook(hook)[source]
                                                
+register_step_post_hook(hook)[source]
 
                                                Register an optimizer step post hook which will be called after optimizer step.
                                                
 
                                                It should have the following signature:
                                                
 
                                                hook(optimizer, args, kwargs) -> None
@@ -902,7 +902,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-register_step_pre_hook(hook)[source]
                                                
+register_step_pre_hook(hook)[source]
 
                                                Register an optimizer step pre hook which will be called before optimizer step.
                                                
 
                                                It should have the following signature:
                                                
 
                                                hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -927,7 +927,7 @@ 
                                                Adam¶<
 
 
                                                
 
                                                
-state_dict()[source]
                                                
+state_dict()[source]
 
                                                Return the state of the optimizer as a dict.
                                                
 
                                                It contains two entries:
                                                
 
                                                  
@@ -990,7 +990,7 @@ 
                                                  Adam¶<
 
 
                                                  
 
                                                  
-step(closure=None)[source][source]
                                                  
+step(closure=None)[source][source]
 
                                                  Perform a single optimization step.
                                                  
 
                                                  
 
                                                  Parameters
                                                  
@@ -1002,7 +1002,7 @@ 
                                                  Adam¶<
 
 
                                                  
 
                                                  
-zero_grad(set_to_none=True)[source]
                                                  
+zero_grad(set_to_none=True)[source]
 
                                                  Reset the gradients of all optimized torch.Tensor s.
                                                  
 
                                                  
 
                                                  Parameters
                                                  
diff --git a/2.7/generated/torch.optim.AdamW.html b/2.7/generated/torch.optim.AdamW.html
index 1a07509f863..e97867500ee 100644
--- a/2.7/generated/torch.optim.AdamW.html
+++ b/2.7/generated/torch.optim.AdamW.html
@@ -596,7 +596,7 @@
 
                                                  AdamW
                                                  
 
                                                  
 
                                                  
-class torch.optim.AdamW(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0.01, amsgrad=False, *, maximize=False, foreach=None, capturable=False, differentiable=False, fused=None)[source][source]
                                                  
+class torch.optim.AdamW(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0.01, amsgrad=False, *, maximize=False, foreach=None, capturable=False, differentiable=False, fused=None)[source][source]
 
                                                  Implements AdamW algorithm, where weight decay does not accumulate in the momentum nor variance.
                                                  
 
                                                  
 input:γ(lr), β1,β2(betas), θ0(params), f(θ)(objective), ϵ (epsilon)λ(weight decay), amsgrad, maximizeinitialize:m0←0 (first moment),v0←0 ( second moment), v0max←0for t=1 to … doif maximize:gt←−∇θft(θt−1)elsegt←∇θft(θt−1)θt←θt−1−γλθt−1mt←β1mt−1+(1−β1)gtvt←β2vt−1+(1−β2)gt2mt^←mt/(1−β1t)if amsgradvtmax←max(vt−1max,vt)vt^←vtmax/(1−β2t)elsevt^←vt/(1−β2t)θt←θt−γmt^/(vt^+ϵ)return θt\begin{aligned}
@@ -703,7 +703,7 @@ 
                                                  AdamW
 
                                                  
 
                                                  
 
                                                  
-add_param_group(param_group)[source]
                                                  
+add_param_group(param_group)[source]
 
                                                  Add a param group to the Optimizer s param_groups.
                                                  
 
                                                  This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                  
@@ -717,7 +717,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-load_state_dict(state_dict)[source]
                                                  
+load_state_dict(state_dict)[source]
 
                                                  Load the optimizer state.
                                                  
 
                                                  
 
                                                  Parameters
                                                  
@@ -741,7 +741,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                  
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                  Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                  
@@ -775,7 +775,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                  
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                  Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                  
@@ -812,7 +812,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                  
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                  Register a state dict post-hook which will be called after state_dict() is called.
                                                  
 
                                                  It should have the following signature:
                                                  
 
                                                  hook(optimizer, state_dict) -> state_dict or None
@@ -844,7 +844,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                  
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                  Register a state dict pre-hook which will be called before state_dict() is called.
                                                  
 
                                                  It should have the following signature:
                                                  
 
                                                  hook(optimizer) -> None
@@ -876,7 +876,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_step_post_hook(hook)[source]
                                                  
+register_step_post_hook(hook)[source]
 
                                                  Register an optimizer step post hook which will be called after optimizer step.
                                                  
 
                                                  It should have the following signature:
                                                  
 
                                                  hook(optimizer, args, kwargs) -> None
@@ -899,7 +899,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-register_step_pre_hook(hook)[source]
                                                  
+register_step_pre_hook(hook)[source]
 
                                                  Register an optimizer step pre hook which will be called before optimizer step.
                                                  
 
                                                  It should have the following signature:
                                                  
 
                                                  hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -924,7 +924,7 @@ 
                                                  AdamW
 
 
                                                  
 
                                                  
-state_dict()[source]
                                                  
+state_dict()[source]
 
                                                  Return the state of the optimizer as a dict.
                                                  
 
                                                  It contains two entries:
                                                  
 
                                                    
@@ -987,7 +987,7 @@ 
                                                    AdamW
 
 
                                                    
 
                                                    
-step(closure=None)[source]
                                                    
+step(closure=None)[source]
 
                                                    Perform a single optimization step.
                                                    
 
                                                    
 
                                                    Parameters
                                                    
@@ -999,7 +999,7 @@ 
                                                    AdamW
 
 
                                                    
 
                                                    
-zero_grad(set_to_none=True)[source]
                                                    
+zero_grad(set_to_none=True)[source]
 
                                                    Reset the gradients of all optimized torch.Tensor s.
                                                    
 
                                                    
 
                                                    Parameters
                                                    
diff --git a/2.7/generated/torch.optim.Adamax.html b/2.7/generated/torch.optim.Adamax.html
index 794f026cac9..e6209c5050d 100644
--- a/2.7/generated/torch.optim.Adamax.html
+++ b/2.7/generated/torch.optim.Adamax.html
@@ -596,7 +596,7 @@
 
                                                    Adamax
                                                    
 
                                                    
 
                                                    
-class torch.optim.Adamax(params, lr=0.002, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, foreach=None, *, maximize=False, differentiable=False, capturable=False)[source][source]
                                                    
+class torch.optim.Adamax(params, lr=0.002, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, foreach=None, *, maximize=False, differentiable=False, capturable=False)[source][source]
 
                                                    Implements Adamax algorithm (a variant of Adam based on infinity norm).
                                                    
 
                                                    
 input:γ (lr),β1,β2 (betas),θ0 (params),f(θ) (objective), λ (weight decay),ϵ (epsilon)initialize:m0←0 ( first moment),u0←0 ( infinity norm)for t=1 to … dogt←∇θft(θt−1)if λ≠0gt←gt+λθt−1mt←β1mt−1+(1−β1)gtut←max(β2ut−1,∣gt∣+ϵ)θt←θt−1−γmt(1−β1t)utreturn θt\begin{aligned}
@@ -656,7 +656,7 @@ 
                                                    Adamax
 
                                                    
 
                                                    
-add_param_group(param_group)[source]
                                                    
+add_param_group(param_group)[source]
 
                                                    Add a param group to the Optimizer s param_groups.
                                                    
 
                                                    This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                    
@@ -670,7 +670,7 @@ 
                                                    Adamax
 
                                                    
-load_state_dict(state_dict)[source]
                                                    
+load_state_dict(state_dict)[source]
 
                                                    Load the optimizer state.
                                                    
 
                                                    
 
                                                    Parameters
                                                    
@@ -694,7 +694,7 @@ 
                                                    Adamax
 
                                                    
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                    
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                    Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                    
@@ -728,7 +728,7 @@ 
                                                    Adamax
 
                                                    
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                    
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                    Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                    
@@ -765,7 +765,7 @@ 
                                                    Adamax
 
                                                    
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                    
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                    Register a state dict post-hook which will be called after state_dict() is called.
                                                    
 
                                                    It should have the following signature:
                                                    
 
                                                    hook(optimizer, state_dict) -> state_dict or None
@@ -797,7 +797,7 @@ 
                                                    Adamax
 
                                                    
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                    
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                    Register a state dict pre-hook which will be called before state_dict() is called.
                                                    
 
                                                    It should have the following signature:
                                                    
 
                                                    hook(optimizer) -> None
@@ -829,7 +829,7 @@ 
                                                    Adamax
 
                                                    
-register_step_post_hook(hook)[source]
                                                    
+register_step_post_hook(hook)[source]
 
                                                    Register an optimizer step post hook which will be called after optimizer step.
                                                    
 
                                                    It should have the following signature:
                                                    
 
                                                    hook(optimizer, args, kwargs) -> None
@@ -852,7 +852,7 @@ 
                                                    Adamax
 
                                                    
-register_step_pre_hook(hook)[source]
                                                    
+register_step_pre_hook(hook)[source]
 
                                                    Register an optimizer step pre hook which will be called before optimizer step.
                                                    
 
                                                    It should have the following signature:
                                                    
 
                                                    hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -877,7 +877,7 @@ 
                                                    Adamax
 
                                                    
-state_dict()[source]
                                                    
+state_dict()[source]
 
                                                    Return the state of the optimizer as a dict.
                                                    
 
                                                    It contains two entries:
                                                    
 
                                                      
@@ -940,7 +940,7 @@ 
                                                      Adamax
 
                                                      
-step(closure=None)[source][source]
                                                      
+step(closure=None)[source][source]
 
                                                      Performs a single optimization step.
                                                      
 
                                                      
 
                                                      Parameters
                                                      
@@ -952,7 +952,7 @@ 
                                                      Adamax
 
                                                      
-zero_grad(set_to_none=True)[source]
                                                      
+zero_grad(set_to_none=True)[source]
 
                                                      Reset the gradients of all optimized torch.Tensor s.
                                                      
 
                                                      
 
                                                      Parameters
                                                      
diff --git a/2.7/generated/torch.optim.LBFGS.html b/2.7/generated/torch.optim.LBFGS.html
index f1839d36938..25d80585381 100644
--- a/2.7/generated/torch.optim.LBFGS.html
+++ b/2.7/generated/torch.optim.LBFGS.html
@@ -596,7 +596,7 @@
 
                                                      LBFGS
                                                      
 
                                                      
 
                                                      
-class torch.optim.LBFGS(params, lr=1, max_iter=20, max_eval=None, tolerance_grad=1e-07, tolerance_change=1e-09, history_size=100, line_search_fn=None)[source][source]
                                                      
+class torch.optim.LBFGS(params, lr=1, max_iter=20, max_eval=None, tolerance_grad=1e-07, tolerance_change=1e-09, history_size=100, line_search_fn=None)[source][source]
 
                                                      Implements L-BFGS algorithm.
                                                      
 
                                                      Heavily inspired by minFunc.
                                                      
 
                                                      
@@ -635,7 +635,7 @@ 
                                                      LBFGS
 
                                                      
 
                                                      
 
                                                      
-add_param_group(param_group)[source]
                                                      
+add_param_group(param_group)[source]
 
                                                      Add a param group to the Optimizer s param_groups.
                                                      
 
                                                      This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                      
@@ -649,7 +649,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-load_state_dict(state_dict)[source]
                                                      
+load_state_dict(state_dict)[source]
 
                                                      Load the optimizer state.
                                                      
 
                                                      
 
                                                      Parameters
                                                      
@@ -673,7 +673,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                      
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                      Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                      
@@ -707,7 +707,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                      
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                      Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                      
@@ -744,7 +744,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                      
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                      Register a state dict post-hook which will be called after state_dict() is called.
                                                      
 
                                                      It should have the following signature:
                                                      
 
                                                      hook(optimizer, state_dict) -> state_dict or None
@@ -776,7 +776,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                      
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                      Register a state dict pre-hook which will be called before state_dict() is called.
                                                      
 
                                                      It should have the following signature:
                                                      
 
                                                      hook(optimizer) -> None
@@ -808,7 +808,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_step_post_hook(hook)[source]
                                                      
+register_step_post_hook(hook)[source]
 
                                                      Register an optimizer step post hook which will be called after optimizer step.
                                                      
 
                                                      It should have the following signature:
                                                      
 
                                                      hook(optimizer, args, kwargs) -> None
@@ -831,7 +831,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-register_step_pre_hook(hook)[source]
                                                      
+register_step_pre_hook(hook)[source]
 
                                                      Register an optimizer step pre hook which will be called before optimizer step.
                                                      
 
                                                      It should have the following signature:
                                                      
 
                                                      hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -856,7 +856,7 @@ 
                                                      LBFGS
 
 
                                                      
 
                                                      
-state_dict()[source]
                                                      
+state_dict()[source]
 
                                                      Return the state of the optimizer as a dict.
                                                      
 
                                                      It contains two entries:
                                                      
 
                                                        
@@ -919,7 +919,7 @@ 
                                                        LBFGS
 
 
                                                        
 
                                                        
-step(closure)[source][source]
                                                        
+step(closure)[source][source]
 
                                                        Perform a single optimization step.
                                                        
 
                                                        
 
                                                        Parameters
                                                        
@@ -931,7 +931,7 @@ 
                                                        LBFGS
 
 
                                                        
 
                                                        
-zero_grad(set_to_none=True)[source]
                                                        
+zero_grad(set_to_none=True)[source]
 
                                                        Reset the gradients of all optimized torch.Tensor s.
                                                        
 
                                                        
 
                                                        Parameters
                                                        
diff --git a/2.7/generated/torch.optim.NAdam.html b/2.7/generated/torch.optim.NAdam.html
index 74856c408a7..35bdce7e672 100644
--- a/2.7/generated/torch.optim.NAdam.html
+++ b/2.7/generated/torch.optim.NAdam.html
@@ -596,7 +596,7 @@
 
                                                        NAdam
                                                        
 
                                                        
 
                                                        
-class torch.optim.NAdam(params, lr=0.002, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, momentum_decay=0.004, decoupled_weight_decay=False, *, foreach=None, maximize=False, capturable=False, differentiable=False)[source][source]
                                                        
+class torch.optim.NAdam(params, lr=0.002, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, momentum_decay=0.004, decoupled_weight_decay=False, *, foreach=None, maximize=False, capturable=False, differentiable=False)[source][source]
 
                                                        Implements NAdam algorithm.
                                                        
 
                                                        
 input:γt (lr), β1,β2 (betas), θ0 (params), f(θ) (objective) λ (weight decay), ψ (momentum decay) decoupled_weight_decay, maximizeinitialize:m0←0 ( first moment),v0←0 ( second moment)for t=1 to … doif maximize:gt←−∇θft(θt−1)elsegt←∇θft(θt−1)θt←θt−1if λ≠0if decoupled_weight_decayθt←θt−1−γλθt−1elsegt←gt+λθt−1μt←β1(1−120.96tψ)μt+1←β1(1−120.96(t+1)ψ)mt←β1mt−1+(1−β1)gtvt←β2vt−1+(1−β2)gt2mt^←μt+1mt/(1−∏i=1t+1μi)+(1−μt)gt/(1−∏i=1tμi)vt^←vt/(1−β2t)θt←θt−γmt^/(vt^+ϵ)return θt\begin{aligned}
@@ -687,7 +687,7 @@ 
                                                        NAdam
 
                                                        
 
                                                        
 
                                                        
-add_param_group(param_group)[source]
                                                        
+add_param_group(param_group)[source]
 
                                                        Add a param group to the Optimizer s param_groups.
                                                        
 
                                                        This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                        
@@ -701,7 +701,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-load_state_dict(state_dict)[source]
                                                        
+load_state_dict(state_dict)[source]
 
                                                        Load the optimizer state.
                                                        
 
                                                        
 
                                                        Parameters
                                                        
@@ -725,7 +725,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                        
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                        Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                        
@@ -759,7 +759,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                        
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                        Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                        
@@ -796,7 +796,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                        
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                        Register a state dict post-hook which will be called after state_dict() is called.
                                                        
 
                                                        It should have the following signature:
                                                        
 
                                                        hook(optimizer, state_dict) -> state_dict or None
@@ -828,7 +828,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                        
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                        Register a state dict pre-hook which will be called before state_dict() is called.
                                                        
 
                                                        It should have the following signature:
                                                        
 
                                                        hook(optimizer) -> None
@@ -860,7 +860,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_step_post_hook(hook)[source]
                                                        
+register_step_post_hook(hook)[source]
 
                                                        Register an optimizer step post hook which will be called after optimizer step.
                                                        
 
                                                        It should have the following signature:
                                                        
 
                                                        hook(optimizer, args, kwargs) -> None
@@ -883,7 +883,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-register_step_pre_hook(hook)[source]
                                                        
+register_step_pre_hook(hook)[source]
 
                                                        Register an optimizer step pre hook which will be called before optimizer step.
                                                        
 
                                                        It should have the following signature:
                                                        
 
                                                        hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -908,7 +908,7 @@ 
                                                        NAdam
 
 
                                                        
 
                                                        
-state_dict()[source]
                                                        
+state_dict()[source]
 
                                                        Return the state of the optimizer as a dict.
                                                        
 
                                                        It contains two entries:
                                                        
 
                                                          
@@ -971,7 +971,7 @@ 
                                                          NAdam
 
 
                                                          
 
                                                          
-step(closure=None)[source][source]
                                                          
+step(closure=None)[source][source]
 
                                                          Perform a single optimization step.
                                                          
 
                                                          
 
                                                          Parameters
                                                          
@@ -983,7 +983,7 @@ 
                                                          NAdam
 
 
                                                          
 
                                                          
-zero_grad(set_to_none=True)[source]
                                                          
+zero_grad(set_to_none=True)[source]
 
                                                          Reset the gradients of all optimized torch.Tensor s.
                                                          
 
                                                          
 
                                                          Parameters
                                                          
diff --git a/2.7/generated/torch.optim.Optimizer.add_param_group.html b/2.7/generated/torch.optim.Optimizer.add_param_group.html
index 8f7bd58ddc0..7cd238f55cf 100644
--- a/2.7/generated/torch.optim.Optimizer.add_param_group.html
+++ b/2.7/generated/torch.optim.Optimizer.add_param_group.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.add_param_group
                                                          
 
                                                          
 
                                                          
-Optimizer.add_param_group(param_group)[source][source]
                                                          
+Optimizer.add_param_group(param_group)[source][source]
 
                                                          Add a param group to the Optimizer s param_groups.
                                                          
 
                                                          This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                          
diff --git a/2.7/generated/torch.optim.Optimizer.load_state_dict.html b/2.7/generated/torch.optim.Optimizer.load_state_dict.html
index cf8aed341a9..f1be8453a48 100644
--- a/2.7/generated/torch.optim.Optimizer.load_state_dict.html
+++ b/2.7/generated/torch.optim.Optimizer.load_state_dict.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.load_state_dict
                                                          
 
                                                          
 
                                                          
-Optimizer.load_state_dict(state_dict)[source][source]
                                                          
+Optimizer.load_state_dict(state_dict)[source][source]
 
                                                          Load the optimizer state.
                                                          
 
                                                          
 
                                                          Parameters
                                                          
diff --git a/2.7/generated/torch.optim.Optimizer.register_load_state_dict_post_hook.html b/2.7/generated/torch.optim.Optimizer.register_load_state_dict_post_hook.html
index a8bd7afba6d..bcbea46b4d3 100644
--- a/2.7/generated/torch.optim.Optimizer.register_load_state_dict_post_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_load_state_dict_post_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_load_state_dict_post_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_load_state_dict_post_hook(hook, prepend=False)[source][source]
                                                          
+Optimizer.register_load_state_dict_post_hook(hook, prepend=False)[source][source]
 
                                                          Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                          
diff --git a/2.7/generated/torch.optim.Optimizer.register_load_state_dict_pre_hook.html b/2.7/generated/torch.optim.Optimizer.register_load_state_dict_pre_hook.html
index b3c5d30bc4a..55992cce234 100644
--- a/2.7/generated/torch.optim.Optimizer.register_load_state_dict_pre_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_load_state_dict_pre_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_load_state_dict_pre_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_load_state_dict_pre_hook(hook, prepend=False)[source][source]
                                                          
+Optimizer.register_load_state_dict_pre_hook(hook, prepend=False)[source][source]
 
                                                          Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                          
diff --git a/2.7/generated/torch.optim.Optimizer.register_state_dict_post_hook.html b/2.7/generated/torch.optim.Optimizer.register_state_dict_post_hook.html
index 590af7fde30..288e3a63dec 100644
--- a/2.7/generated/torch.optim.Optimizer.register_state_dict_post_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_state_dict_post_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_state_dict_post_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_state_dict_post_hook(hook, prepend=False)[source][source]
                                                          
+Optimizer.register_state_dict_post_hook(hook, prepend=False)[source][source]
 
                                                          Register a state dict post-hook which will be called after state_dict() is called.
                                                          
 
                                                          It should have the following signature:
                                                          
 
                                                          hook(optimizer, state_dict) -> state_dict or None
diff --git a/2.7/generated/torch.optim.Optimizer.register_state_dict_pre_hook.html b/2.7/generated/torch.optim.Optimizer.register_state_dict_pre_hook.html
index 2c27c34eadb..f13041b9869 100644
--- a/2.7/generated/torch.optim.Optimizer.register_state_dict_pre_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_state_dict_pre_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_state_dict_pre_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_state_dict_pre_hook(hook, prepend=False)[source][source]
                                                          
+Optimizer.register_state_dict_pre_hook(hook, prepend=False)[source][source]
 
                                                          Register a state dict pre-hook which will be called before state_dict() is called.
                                                          
 
                                                          It should have the following signature:
                                                          
 
                                                          hook(optimizer) -> None
diff --git a/2.7/generated/torch.optim.Optimizer.register_step_post_hook.html b/2.7/generated/torch.optim.Optimizer.register_step_post_hook.html
index 1a7f4c895b3..b787e4071ed 100644
--- a/2.7/generated/torch.optim.Optimizer.register_step_post_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_step_post_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_step_post_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_step_post_hook(hook)[source][source]
                                                          
+Optimizer.register_step_post_hook(hook)[source][source]
 
                                                          Register an optimizer step post hook which will be called after optimizer step.
                                                          
 
                                                          It should have the following signature:
                                                          
 
                                                          hook(optimizer, args, kwargs) -> None
diff --git a/2.7/generated/torch.optim.Optimizer.register_step_pre_hook.html b/2.7/generated/torch.optim.Optimizer.register_step_pre_hook.html
index c577c79ed89..4bda4463ee0 100644
--- a/2.7/generated/torch.optim.Optimizer.register_step_pre_hook.html
+++ b/2.7/generated/torch.optim.Optimizer.register_step_pre_hook.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.register_step_pre_hook
                                                          
 
                                                          
 
                                                          
-Optimizer.register_step_pre_hook(hook)[source][source]
                                                          
+Optimizer.register_step_pre_hook(hook)[source][source]
 
                                                          Register an optimizer step pre hook which will be called before optimizer step.
                                                          
 
                                                          It should have the following signature:
                                                          
 
                                                          hook(optimizer, args, kwargs) -> None or modified args and kwargs
diff --git a/2.7/generated/torch.optim.Optimizer.state_dict.html b/2.7/generated/torch.optim.Optimizer.state_dict.html
index 38a4db95246..adbd8b7e323 100644
--- a/2.7/generated/torch.optim.Optimizer.state_dict.html
+++ b/2.7/generated/torch.optim.Optimizer.state_dict.html
@@ -596,7 +596,7 @@
 
                                                          torch.optim.Optimizer.state_dict
                                                          
 
                                                          
 
                                                          
-Optimizer.state_dict()[source][source]
                                                          
+Optimizer.state_dict()[source][source]
 
                                                          Return the state of the optimizer as a dict.
                                                          
 
                                                          It contains two entries:
                                                          
 
                                                            
diff --git a/2.7/generated/torch.optim.Optimizer.step.html b/2.7/generated/torch.optim.Optimizer.step.html
index ef357fb60b0..cc6f3455eb4 100644
--- a/2.7/generated/torch.optim.Optimizer.step.html
+++ b/2.7/generated/torch.optim.Optimizer.step.html
@@ -596,7 +596,7 @@
 
                                                            torch.optim.Optimizer.step
                                                            
 
                                                            
 
                                                            
-Optimizer.step(closure: None = None) → None[source][source]
                                                            
+Optimizer.step(closure: None = None) → None[source][source]
 
                                                            
 Optimizer.step(closure: Callable[[], float]) → float
                                                            
 
                                                            Perform a single optimization step to update parameter.
                                                            
diff --git a/2.7/generated/torch.optim.Optimizer.zero_grad.html b/2.7/generated/torch.optim.Optimizer.zero_grad.html
index d400490f5a3..34f3a84bf27 100644
--- a/2.7/generated/torch.optim.Optimizer.zero_grad.html
+++ b/2.7/generated/torch.optim.Optimizer.zero_grad.html
@@ -596,7 +596,7 @@
 
                                                            torch.optim.Optimizer.zero_grad
                                                            
 
                                                            
 
                                                            
-Optimizer.zero_grad(set_to_none=True)[source][source]
                                                            
+Optimizer.zero_grad(set_to_none=True)[source][source]
 
                                                            Reset the gradients of all optimized torch.Tensor s.
                                                            
 
                                                            
 
                                                            Parameters
                                                            
diff --git a/2.7/generated/torch.optim.RAdam.html b/2.7/generated/torch.optim.RAdam.html
index 3661d561a77..f19be9a8b7f 100644
--- a/2.7/generated/torch.optim.RAdam.html
+++ b/2.7/generated/torch.optim.RAdam.html
@@ -596,7 +596,7 @@
 
                                                            RAdam
                                                            
 
                                                            
 
                                                            
-class torch.optim.RAdam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, decoupled_weight_decay=False, *, foreach=None, maximize=False, capturable=False, differentiable=False)[source][source]
                                                            
+class torch.optim.RAdam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, decoupled_weight_decay=False, *, foreach=None, maximize=False, capturable=False, differentiable=False)[source][source]
 
                                                            Implements RAdam algorithm.
                                                            
 
                                                            
 input:γ (lr), β1,β2 (betas), θ0 (params), f(θ) (objective), λ (weightdecay), maximizeϵ (epsilon),decoupled_weight_decayinitialize:m0←0 ( first moment),v0←0 ( second moment),ρ∞←2/(1−β2)−1for t=1 to … doif maximize:gt←−∇θft(θt−1)elsegt←∇θft(θt−1)θt←θt−1if λ≠0if decoupled_weight_decayθt←θt−γλθtelsegt←gt+λθtmt←β1mt−1+(1−β1)gtvt←β2vt−1+(1−β2)gt2mt^←mt/(1−β1t)ρt←ρ∞−2tβ2t/(1−β2t)if ρt>5lt←(1−β2t)vt+ϵrt←(ρt−4)(ρt−2)ρ∞(ρ∞−4)(ρ∞−2)ρtθt←θt−γmt^rtltelseθt←θt−γmt^return θt\begin{aligned}
@@ -713,7 +713,7 @@ 
                                                            RAdam
 
                                                            
 
                                                            
 
                                                            
-add_param_group(param_group)[source]
                                                            
+add_param_group(param_group)[source]
 
                                                            Add a param group to the Optimizer s param_groups.
                                                            
 
                                                            This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                            
@@ -727,7 +727,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-load_state_dict(state_dict)[source]
                                                            
+load_state_dict(state_dict)[source]
 
                                                            Load the optimizer state.
                                                            
 
                                                            
 
                                                            Parameters
                                                            
@@ -751,7 +751,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                            
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                            Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                            
@@ -785,7 +785,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                            
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                            Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                            
@@ -822,7 +822,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                            
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                            Register a state dict post-hook which will be called after state_dict() is called.
                                                            
 
                                                            It should have the following signature:
                                                            
 
                                                            hook(optimizer, state_dict) -> state_dict or None
@@ -854,7 +854,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                            
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                            Register a state dict pre-hook which will be called before state_dict() is called.
                                                            
 
                                                            It should have the following signature:
                                                            
 
                                                            hook(optimizer) -> None
@@ -886,7 +886,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_step_post_hook(hook)[source]
                                                            
+register_step_post_hook(hook)[source]
 
                                                            Register an optimizer step post hook which will be called after optimizer step.
                                                            
 
                                                            It should have the following signature:
                                                            
 
                                                            hook(optimizer, args, kwargs) -> None
@@ -909,7 +909,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-register_step_pre_hook(hook)[source]
                                                            
+register_step_pre_hook(hook)[source]
 
                                                            Register an optimizer step pre hook which will be called before optimizer step.
                                                            
 
                                                            It should have the following signature:
                                                            
 
                                                            hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -934,7 +934,7 @@ 
                                                            RAdam
 
 
                                                            
 
                                                            
-state_dict()[source]
                                                            
+state_dict()[source]
 
                                                            Return the state of the optimizer as a dict.
                                                            
 
                                                            It contains two entries:
                                                            
 
                                                              
@@ -997,7 +997,7 @@ 
                                                              RAdam
 
 
                                                              
 
                                                              
-step(closure=None)[source][source]
                                                              
+step(closure=None)[source][source]
 
                                                              Perform a single optimization step.
                                                              
 
                                                              
 
                                                              Parameters
                                                              
@@ -1009,7 +1009,7 @@ 
                                                              RAdam
 
 
                                                              
 
                                                              
-zero_grad(set_to_none=True)[source]
                                                              
+zero_grad(set_to_none=True)[source]
 
                                                              Reset the gradients of all optimized torch.Tensor s.
                                                              
 
                                                              
 
                                                              Parameters
                                                              
diff --git a/2.7/generated/torch.optim.RMSprop.html b/2.7/generated/torch.optim.RMSprop.html
index b96202399c7..05fb8dc9905 100644
--- a/2.7/generated/torch.optim.RMSprop.html
+++ b/2.7/generated/torch.optim.RMSprop.html
@@ -596,7 +596,7 @@
 
                                                              RMSprop
                                                              
 
                                                              
 
                                                              
-class torch.optim.RMSprop(params, lr=0.01, alpha=0.99, eps=1e-08, weight_decay=0, momentum=0, centered=False, capturable=False, foreach=None, maximize=False, differentiable=False)[source][source]
                                                              
+class torch.optim.RMSprop(params, lr=0.01, alpha=0.99, eps=1e-08, weight_decay=0, momentum=0, centered=False, capturable=False, foreach=None, maximize=False, differentiable=False)[source][source]
 
                                                              Implements RMSprop algorithm.
                                                              
 
                                                              
 input:α (alpha), γ (lr), θ0 (params), f(θ) (objective)λ (weight decay), μ (momentum), centered, ϵ (epsilon)initialize:v0←0 (square average), b0←0 (buffer), g0ave←0for t=1 to … dogt←∇θft(θt−1)if λ≠0gt←gt+λθt−1vt←αvt−1+(1−α)gt2vt~←vtif centeredgtave←gt−1aveα+(1−α)gtvt~←vt~−(gtave)2if μ>0bt←μbt−1+gt/(vt~+ϵ)θt←θt−1−γbtelseθt←θt−1−γgt/(vt~+ϵ)return θt\begin{aligned}
@@ -707,7 +707,7 @@ 
                                                              RMSprop
 
                                                              
-add_param_group(param_group)[source]
                                                              
+add_param_group(param_group)[source]
 
                                                              Add a param group to the Optimizer s param_groups.
                                                              
 
                                                              This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                              
@@ -721,7 +721,7 @@ 
                                                              RMSprop
 
                                                              
-load_state_dict(state_dict)[source]
                                                              
+load_state_dict(state_dict)[source]
 
                                                              Load the optimizer state.
                                                              
 
                                                              
 
                                                              Parameters
                                                              
@@ -745,7 +745,7 @@ 
                                                              RMSprop
 
                                                              
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                              
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                              Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                              
@@ -779,7 +779,7 @@ 
                                                              RMSprop
 
                                                              
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                              
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                              Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                              
@@ -816,7 +816,7 @@ 
                                                              RMSprop
 
                                                              
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                              
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                              Register a state dict post-hook which will be called after state_dict() is called.
                                                              
 
                                                              It should have the following signature:
                                                              
 
                                                              hook(optimizer, state_dict) -> state_dict or None
@@ -848,7 +848,7 @@ 
                                                              RMSprop
 
                                                              
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                              
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                              Register a state dict pre-hook which will be called before state_dict() is called.
                                                              
 
                                                              It should have the following signature:
                                                              
 
                                                              hook(optimizer) -> None
@@ -880,7 +880,7 @@ 
                                                              RMSprop
 
                                                              
-register_step_post_hook(hook)[source]
                                                              
+register_step_post_hook(hook)[source]
 
                                                              Register an optimizer step post hook which will be called after optimizer step.
                                                              
 
                                                              It should have the following signature:
                                                              
 
                                                              hook(optimizer, args, kwargs) -> None
@@ -903,7 +903,7 @@ 
                                                              RMSprop
 
                                                              
-register_step_pre_hook(hook)[source]
                                                              
+register_step_pre_hook(hook)[source]
 
                                                              Register an optimizer step pre hook which will be called before optimizer step.
                                                              
 
                                                              It should have the following signature:
                                                              
 
                                                              hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -928,7 +928,7 @@ 
                                                              RMSprop
 
                                                              
-state_dict()[source]
                                                              
+state_dict()[source]
 
                                                              Return the state of the optimizer as a dict.
                                                              
 
                                                              It contains two entries:
                                                              
 
                                                                
@@ -991,7 +991,7 @@ 
                                                                RMSprop
 
                                                                
-step(closure=None)[source][source]
                                                                
+step(closure=None)[source][source]
 
                                                                Perform a single optimization step.
                                                                
 
                                                                
 
                                                                Parameters
                                                                
@@ -1003,7 +1003,7 @@ 
                                                                RMSprop
 
                                                                
-zero_grad(set_to_none=True)[source]
                                                                
+zero_grad(set_to_none=True)[source]
 
                                                                Reset the gradients of all optimized torch.Tensor s.
                                                                
 
                                                                
 
                                                                Parameters
                                                                
diff --git a/2.7/generated/torch.optim.Rprop.html b/2.7/generated/torch.optim.Rprop.html
index e5d6f5a15b9..9ef7f1e71a0 100644
--- a/2.7/generated/torch.optim.Rprop.html
+++ b/2.7/generated/torch.optim.Rprop.html
@@ -596,7 +596,7 @@
 
                                                                Rprop
                                                                
 
                                                                
 
                                                                
-class torch.optim.Rprop(params, lr=0.01, etas=(0.5, 1.2), step_sizes=(1e-06, 50), *, capturable=False, foreach=None, maximize=False, differentiable=False)[source][source]
                                                                
+class torch.optim.Rprop(params, lr=0.01, etas=(0.5, 1.2), step_sizes=(1e-06, 50), *, capturable=False, foreach=None, maximize=False, differentiable=False)[source][source]
 
                                                                Implements the resilient backpropagation algorithm.
                                                                
 
                                                                
 input:θ0∈Rd (params),f(θ) (objective),η+/− (etaplus, etaminus),Γmax/min (step sizes)initialize:gprev0←0, η0←lr (learning rate)for t=1 to … dogt←∇θft(θt−1)for i=0,1,…,d−1 doif gprevigti>0ηti←min(ηt−1iη+,Γmax)else if gprevigti<0ηti←max(ηt−1iη−,Γmin)gti←0else ηti←ηt−1iθt←θt−1−ηtsign(gt)gprev←gtreturn θt\begin{aligned}
@@ -664,7 +664,7 @@ 
                                                                Rprop
 
                                                                
 
                                                                
 
                                                                
-add_param_group(param_group)[source]
                                                                
+add_param_group(param_group)[source]
 
                                                                Add a param group to the Optimizer s param_groups.
                                                                
 
                                                                This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                                
@@ -678,7 +678,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-load_state_dict(state_dict)[source]
                                                                
+load_state_dict(state_dict)[source]
 
                                                                Load the optimizer state.
                                                                
 
                                                                
 
                                                                Parameters
                                                                
@@ -702,7 +702,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                                
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                                
@@ -736,7 +736,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                                
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                                
@@ -773,7 +773,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                                
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                Register a state dict post-hook which will be called after state_dict() is called.
                                                                
 
                                                                It should have the following signature:
                                                                
 
                                                                hook(optimizer, state_dict) -> state_dict or None
@@ -805,7 +805,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                                
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                Register a state dict pre-hook which will be called before state_dict() is called.
                                                                
 
                                                                It should have the following signature:
                                                                
 
                                                                hook(optimizer) -> None
@@ -837,7 +837,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_step_post_hook(hook)[source]
                                                                
+register_step_post_hook(hook)[source]
 
                                                                Register an optimizer step post hook which will be called after optimizer step.
                                                                
 
                                                                It should have the following signature:
                                                                
 
                                                                hook(optimizer, args, kwargs) -> None
@@ -860,7 +860,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-register_step_pre_hook(hook)[source]
                                                                
+register_step_pre_hook(hook)[source]
 
                                                                Register an optimizer step pre hook which will be called before optimizer step.
                                                                
 
                                                                It should have the following signature:
                                                                
 
                                                                hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -885,7 +885,7 @@ 
                                                                Rprop
 
 
                                                                
 
                                                                
-state_dict()[source]
                                                                
+state_dict()[source]
 
                                                                Return the state of the optimizer as a dict.
                                                                
 
                                                                It contains two entries:
                                                                
 
                                                                  
@@ -948,7 +948,7 @@ 
                                                                  Rprop
 
 
                                                                  
 
                                                                  
-step(closure=None)[source][source]
                                                                  
+step(closure=None)[source][source]
 
                                                                  Perform a single optimization step.
                                                                  
 
                                                                  
 
                                                                  Parameters
                                                                  
@@ -960,7 +960,7 @@ 
                                                                  Rprop
 
 
                                                                  
 
                                                                  
-zero_grad(set_to_none=True)[source]
                                                                  
+zero_grad(set_to_none=True)[source]
 
                                                                  Reset the gradients of all optimized torch.Tensor s.
                                                                  
 
                                                                  
 
                                                                  Parameters
                                                                  
diff --git a/2.7/generated/torch.optim.SGD.html b/2.7/generated/torch.optim.SGD.html
index 1f21ef03eeb..76c1f089579 100644
--- a/2.7/generated/torch.optim.SGD.html
+++ b/2.7/generated/torch.optim.SGD.html
@@ -596,7 +596,7 @@
 
                                                                  SGD
                                                                  
 
                                                                  
 
                                                                  
-class torch.optim.SGD(params, lr=0.001, momentum=0, dampening=0, weight_decay=0, nesterov=False, *, maximize=False, foreach=None, differentiable=False, fused=None)[source][source]
                                                                  
+class torch.optim.SGD(params, lr=0.001, momentum=0, dampening=0, weight_decay=0, nesterov=False, *, maximize=False, foreach=None, differentiable=False, fused=None)[source][source]
 
                                                                  Implements stochastic gradient descent (optionally with momentum).
                                                                  
 
                                                                  
 input:γ (lr), θ0 (params), f(θ) (objective), λ (weight decay), μ (momentum), τ (dampening),  nesterov,  maximizefor t=1 to … dogt←∇θft(θt−1)if λ≠0gt←gt+λθt−1if μ≠0if t>1bt←μbt−1+(1−τ)gtelsebt←gtif nesterovgt←gt+μbtelsegt←btif maximizeθt←θt−1+γgtelseθt←θt−1−γgtreturn θt\begin{aligned}
@@ -708,7 +708,7 @@ 
                                                                  SGD
 
                                                                  
 
                                                                  
-add_param_group(param_group)[source]
                                                                  
+add_param_group(param_group)[source]
 
                                                                  Add a param group to the Optimizer s param_groups.
                                                                  
 
                                                                  This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                                  
@@ -722,7 +722,7 @@ 
                                                                  SGD
 
                                                                  
-load_state_dict(state_dict)[source]
                                                                  
+load_state_dict(state_dict)[source]
 
                                                                  Load the optimizer state.
                                                                  
 
                                                                  
 
                                                                  Parameters
                                                                  
@@ -746,7 +746,7 @@ 
                                                                  SGD
 
                                                                  
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                                  
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                  Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                                  
@@ -780,7 +780,7 @@ 
                                                                  SGD
 
                                                                  
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                                  
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                  Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                                  
@@ -817,7 +817,7 @@ 
                                                                  SGD
 
                                                                  
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                                  
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                  Register a state dict post-hook which will be called after state_dict() is called.
                                                                  
 
                                                                  It should have the following signature:
                                                                  
 
                                                                  hook(optimizer, state_dict) -> state_dict or None
@@ -849,7 +849,7 @@ 
                                                                  SGD
 
                                                                  
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                                  
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                  Register a state dict pre-hook which will be called before state_dict() is called.
                                                                  
 
                                                                  It should have the following signature:
                                                                  
 
                                                                  hook(optimizer) -> None
@@ -881,7 +881,7 @@ 
                                                                  SGD
 
                                                                  
-register_step_post_hook(hook)[source]
                                                                  
+register_step_post_hook(hook)[source]
 
                                                                  Register an optimizer step post hook which will be called after optimizer step.
                                                                  
 
                                                                  It should have the following signature:
                                                                  
 
                                                                  hook(optimizer, args, kwargs) -> None
@@ -904,7 +904,7 @@ 
                                                                  SGD
 
                                                                  
-register_step_pre_hook(hook)[source]
                                                                  
+register_step_pre_hook(hook)[source]
 
                                                                  Register an optimizer step pre hook which will be called before optimizer step.
                                                                  
 
                                                                  It should have the following signature:
                                                                  
 
                                                                  hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -929,7 +929,7 @@ 
                                                                  SGD
 
                                                                  
-state_dict()[source]
                                                                  
+state_dict()[source]
 
                                                                  Return the state of the optimizer as a dict.
                                                                  
 
                                                                  It contains two entries:
                                                                  
 
                                                                    
@@ -992,7 +992,7 @@ 
                                                                    SGD
 
                                                                    
-step(closure=None)[source][source]
                                                                    
+step(closure=None)[source][source]
 
                                                                    Perform a single optimization step.
                                                                    
 
                                                                    
 
                                                                    Parameters
                                                                    
@@ -1004,7 +1004,7 @@ 
                                                                    SGD
 
                                                                    
-zero_grad(set_to_none=True)[source]
                                                                    
+zero_grad(set_to_none=True)[source]
 
                                                                    Reset the gradients of all optimized torch.Tensor s.
                                                                    
 
                                                                    
 
                                                                    Parameters
                                                                    
diff --git a/2.7/generated/torch.optim.SparseAdam.html b/2.7/generated/torch.optim.SparseAdam.html
index 411c4b93f72..292f91ebb29 100644
--- a/2.7/generated/torch.optim.SparseAdam.html
+++ b/2.7/generated/torch.optim.SparseAdam.html
@@ -596,7 +596,7 @@
 
                                                                    SparseAdam
                                                                    
 
                                                                    
 
                                                                    
-class torch.optim.SparseAdam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, maximize=False)[source][source]
                                                                    
+class torch.optim.SparseAdam(params, lr=0.001, betas=(0.9, 0.999), eps=1e-08, maximize=False)[source][source]
 
                                                                    SparseAdam implements a masked version of the Adam algorithm
 suitable for sparse gradients. Currently, due to implementation constraints (explained
 below), SparseAdam is only intended for a narrow subset of use cases, specifically
@@ -655,7 +655,7 @@ 
                                                                    SparseAdam
 
                                                                    
-add_param_group(param_group)[source]
                                                                    
+add_param_group(param_group)[source]
 
                                                                    Add a param group to the Optimizer s param_groups.
                                                                    
 
                                                                    This can be useful when fine tuning a pre-trained network as frozen layers can be made
 trainable and added to the Optimizer as training progresses.
                                                                    
@@ -669,7 +669,7 @@ 
                                                                    SparseAdam
 
                                                                    
-load_state_dict(state_dict)[source]
                                                                    
+load_state_dict(state_dict)[source]
 
                                                                    Load the optimizer state.
                                                                    
 
                                                                    
 
                                                                    Parameters
                                                                    
@@ -693,7 +693,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_load_state_dict_post_hook(hook, prepend=False)[source]
                                                                    
+register_load_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                    Register a load_state_dict post-hook which will be called after
 load_state_dict() is called. It should have the
 following signature:
                                                                    
@@ -727,7 +727,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_load_state_dict_pre_hook(hook, prepend=False)[source]
                                                                    
+register_load_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                    Register a load_state_dict pre-hook which will be called before
 load_state_dict() is called. It should have the
 following signature:
                                                                    
@@ -764,7 +764,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_state_dict_post_hook(hook, prepend=False)[source]
                                                                    
+register_state_dict_post_hook(hook, prepend=False)[source]
 
                                                                    Register a state dict post-hook which will be called after state_dict() is called.
                                                                    
 
                                                                    It should have the following signature:
                                                                    
 
                                                                    hook(optimizer, state_dict) -> state_dict or None
@@ -796,7 +796,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_state_dict_pre_hook(hook, prepend=False)[source]
                                                                    
+register_state_dict_pre_hook(hook, prepend=False)[source]
 
                                                                    Register a state dict pre-hook which will be called before state_dict() is called.
                                                                    
 
                                                                    It should have the following signature:
                                                                    
 
                                                                    hook(optimizer) -> None
@@ -828,7 +828,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_step_post_hook(hook)[source]
                                                                    
+register_step_post_hook(hook)[source]
 
                                                                    Register an optimizer step post hook which will be called after optimizer step.
                                                                    
 
                                                                    It should have the following signature:
                                                                    
 
                                                                    hook(optimizer, args, kwargs) -> None
@@ -851,7 +851,7 @@ 
                                                                    SparseAdam
 
                                                                    
-register_step_pre_hook(hook)[source]
                                                                    
+register_step_pre_hook(hook)[source]
 
                                                                    Register an optimizer step pre hook which will be called before optimizer step.
                                                                    
 
                                                                    It should have the following signature:
                                                                    
 
                                                                    hook(optimizer, args, kwargs) -> None or modified args and kwargs
@@ -876,7 +876,7 @@ 
                                                                    SparseAdam
 
                                                                    
-state_dict()[source]
                                                                    
+state_dict()[source]
 
                                                                    Return the state of the optimizer as a dict.
                                                                    
 
                                                                    It contains two entries:
                                                                    
 
                                                                      
@@ -939,7 +939,7 @@ 
                                                                      SparseAdam
 
                                                                      
-step(closure=None)[source][source]
                                                                      
+step(closure=None)[source][source]
 
                                                                      Perform a single optimization step.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -951,7 +951,7 @@ 
                                                                      SparseAdam
 
                                                                      
-zero_grad(set_to_none=True)[source]
                                                                      
+zero_grad(set_to_none=True)[source]
 
                                                                      Reset the gradients of all optimized torch.Tensor s.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.ChainedScheduler.html b/2.7/generated/torch.optim.lr_scheduler.ChainedScheduler.html
index 3ffd0643c6b..1f08b889338 100644
--- a/2.7/generated/torch.optim.lr_scheduler.ChainedScheduler.html
+++ b/2.7/generated/torch.optim.lr_scheduler.ChainedScheduler.html
@@ -596,7 +596,7 @@
 
                                                                      ChainedScheduler
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.ChainedScheduler(schedulers, optimizer=None)[source][source]
                                                                      
+class torch.optim.lr_scheduler.ChainedScheduler(schedulers, optimizer=None)[source][source]
 
                                                                      Chains a list of learning rate schedulers.
                                                                      
 
                                                                      Takes in a sequence of chainable learning rate schedulers and calls their
 step() functions consecutively in just one call to step().
                                                                      
@@ -626,7 +626,7 @@ 
                                                                      ChainedScheduler
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -637,7 +637,7 @@ 
                                                                      ChainedScheduler
 
                                                                      
-get_lr()[source]
                                                                      
+get_lr()[source]
 
                                                                      Compute learning rate using chainable form of the scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -648,7 +648,7 @@ 
                                                                      ChainedScheduler
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -660,7 +660,7 @@ 
                                                                      ChainedScheduler
 
                                                                      
-state_dict()[source][source]
                                                                      
+state_dict()[source][source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
@@ -669,7 +669,7 @@ 
                                                                      ChainedScheduler
 
                                                                      
-step()[source][source]
                                                                      
+step()[source][source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
diff --git a/2.7/generated/torch.optim.lr_scheduler.ConstantLR.html b/2.7/generated/torch.optim.lr_scheduler.ConstantLR.html
index 546776ba527..32dbfc7d8ad 100644
--- a/2.7/generated/torch.optim.lr_scheduler.ConstantLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.ConstantLR.html
@@ -596,7 +596,7 @@
 
                                                                      ConstantLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.ConstantLR(optimizer, factor=0.3333333333333333, total_iters=5, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.ConstantLR(optimizer, factor=0.3333333333333333, total_iters=5, last_epoch=-1)[source][source]
 
                                                                      Multiply the learning rate of each parameter group by a small constant factor.
                                                                      
 
                                                                      The multiplication is done until the number of epoch reaches a pre-defined milestone: total_iters.
 Notice that such multiplication of the small constant factor can
@@ -629,7 +629,7 @@ 
                                                                      ConstantLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -640,13 +640,13 @@ 
                                                                      ConstantLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -658,7 +658,7 @@ 
                                                                      ConstantLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -666,7 +666,7 @@ 
                                                                      ConstantLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingLR.html b/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingLR.html
index 662ac4d1cde..9e5d038654c 100644
--- a/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingLR.html
@@ -596,7 +596,7 @@
 
                                                                      CosineAnnealingLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max, eta_min=0.0, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max, eta_min=0.0, last_epoch=-1)[source][source]
 
                                                                      Set the learning rate of each parameter group using a cosine annealing schedule.
                                                                      
 
                                                                      The ηmax\eta_{max} is set to the initial lr and
 TcurT_{cur} is the number of epochs since the last restart in SGDR:
                                                                      
@@ -633,7 +633,7 @@ 
                                                                      CosineAnnealingLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -644,13 +644,13 @@ 
                                                                      CosineAnnealingLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Retrieve the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -662,7 +662,7 @@ 
                                                                      CosineAnnealingLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -670,7 +670,7 @@ 
                                                                      CosineAnnealingLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingWarmRestarts.html b/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingWarmRestarts.html
index 01886e653f3..aeb78c12a8b 100644
--- a/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingWarmRestarts.html
+++ b/2.7/generated/torch.optim.lr_scheduler.CosineAnnealingWarmRestarts.html
@@ -596,7 +596,7 @@
 
                                                                      CosineAnnealingWarmRestarts
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, T_0, T_mult=1, eta_min=0.0, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, T_0, T_mult=1, eta_min=0.0, last_epoch=-1)[source][source]
 
                                                                      Set the learning rate of each parameter group using a cosine annealing schedule.
                                                                      
 
                                                                      The ηmax\eta_{max} is set to the initial lr, TcurT_{cur}
 is the number of epochs since the last restart and TiT_{i} is the number
@@ -622,7 +622,7 @@ 
                                                                      CosineAnnealingWarmRestarts
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -633,13 +633,13 @@ 
                                                                      CosineAnnealingWarmRestarts
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the initial learning rate.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -651,7 +651,7 @@ 
                                                                      CosineAnnealingWarmRestarts
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -659,7 +659,7 @@ 
                                                                      CosineAnnealingWarmRestarts
 
                                                                      
-step(epoch=None)[source][source]
                                                                      
+step(epoch=None)[source][source]
 
                                                                      Step could be called after every batch update.
                                                                      
 
                                                                      Example
                                                                      
 
                                                                      >>> scheduler = CosineAnnealingWarmRestarts(optimizer, T_0, T_mult)
diff --git a/2.7/generated/torch.optim.lr_scheduler.CyclicLR.html b/2.7/generated/torch.optim.lr_scheduler.CyclicLR.html
index dbb3e72f37d..127928a31c6 100644
--- a/2.7/generated/torch.optim.lr_scheduler.CyclicLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.CyclicLR.html
@@ -596,7 +596,7 @@
 
                                                                      CyclicLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.CyclicLR(optimizer, base_lr, max_lr, step_size_up=2000, step_size_down=None, mode='triangular', gamma=1.0, scale_fn=None, scale_mode='cycle', cycle_momentum=True, base_momentum=0.8, max_momentum=0.9, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.CyclicLR(optimizer, base_lr, max_lr, step_size_up=2000, step_size_down=None, mode='triangular', gamma=1.0, scale_fn=None, scale_mode='cycle', cycle_momentum=True, base_momentum=0.8, max_momentum=0.9, last_epoch=-1)[source][source]
 
                                                                      Sets the learning rate of each parameter group according to cyclical learning rate policy (CLR).
                                                                      
 
                                                                      The policy cycles the learning rate between two boundaries with a constant frequency,
 as detailed in the paper Cyclical Learning Rates for Training Neural Networks.
@@ -686,7 +686,7 @@ 
                                                                      CyclicLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -697,7 +697,7 @@ 
                                                                      CyclicLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Calculate the learning rate at batch index.
                                                                      
 
                                                                      This function treats self.last_epoch as the last batch index.
                                                                      
 
                                                                      If self.cycle_momentum is True, this function has a side effect of
@@ -706,13 +706,13 @@ 
                                                                      CyclicLR
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-scale_fn(x)[source][source]
                                                                      
+scale_fn(x)[source][source]
 
                                                                      Get the scaling policy.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -723,7 +723,7 @@ 
                                                                      CyclicLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.ExponentialLR.html b/2.7/generated/torch.optim.lr_scheduler.ExponentialLR.html
index c3578ffb387..103d0818bd7 100644
--- a/2.7/generated/torch.optim.lr_scheduler.ExponentialLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.ExponentialLR.html
@@ -596,7 +596,7 @@
 
                                                                      ExponentialLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma, last_epoch=-1)[source][source]
 
                                                                      Decays the learning rate of each parameter group by gamma every epoch.
                                                                      
 
                                                                      When last_epoch=-1, sets initial lr as lr.
                                                                      
 
                                                                      
@@ -610,7 +610,7 @@ 
                                                                      ExponentialLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -621,13 +621,13 @@ 
                                                                      ExponentialLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -639,7 +639,7 @@ 
                                                                      ExponentialLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -647,7 +647,7 @@ 
                                                                      ExponentialLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.LRScheduler.html b/2.7/generated/torch.optim.lr_scheduler.LRScheduler.html
index 0f4fbd82ea7..ae69a0af474 100644
--- a/2.7/generated/torch.optim.lr_scheduler.LRScheduler.html
+++ b/2.7/generated/torch.optim.lr_scheduler.LRScheduler.html
@@ -596,13 +596,13 @@
 
                                                                      LRScheduler
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.LRScheduler(optimizer, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.LRScheduler(optimizer, last_epoch=-1)[source][source]
 
                                                                      Adjusts the learning rate during optimization.
                                                                      
 
                                                                      
 
                                                                      
 
                                                                      
 
                                                                      
-get_last_lr()[source][source]
                                                                      
+get_last_lr()[source][source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -613,7 +613,7 @@ 
                                                                      LRScheduler
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute learning rate using chainable form of the scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -624,7 +624,7 @@ 
                                                                      LRScheduler
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -636,7 +636,7 @@ 
                                                                      LRScheduler
 
                                                                      
-state_dict()[source][source]
                                                                      
+state_dict()[source][source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -644,7 +644,7 @@ 
                                                                      LRScheduler
 
                                                                      
-step(epoch=None)[source][source]
                                                                      
+step(epoch=None)[source][source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.LambdaLR.html b/2.7/generated/torch.optim.lr_scheduler.LambdaLR.html
index 795c374df34..c9a8b31c6e1 100644
--- a/2.7/generated/torch.optim.lr_scheduler.LambdaLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.LambdaLR.html
@@ -596,7 +596,7 @@
 
                                                                      LambdaLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.LambdaLR(optimizer, lr_lambda, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.LambdaLR(optimizer, lr_lambda, last_epoch=-1)[source][source]
 
                                                                      Sets the initial learning rate.
                                                                      
 
                                                                      The learning rate of each parameter group is set to the initial lr
 times a given function. When last_epoch=-1, sets initial lr as lr.
                                                                      
@@ -624,7 +624,7 @@ 
                                                                      LambdaLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -635,13 +635,13 @@ 
                                                                      LambdaLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute learning rate.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      When saving or loading the scheduler, please make sure to also save or load the state of the optimizer.
                                                                      
 
                                                                      
@@ -654,7 +654,7 @@ 
                                                                      LambdaLR
 
                                                                      
-state_dict()[source][source]
                                                                      
+state_dict()[source][source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
@@ -665,7 +665,7 @@ 
                                                                      LambdaLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.LinearLR.html b/2.7/generated/torch.optim.lr_scheduler.LinearLR.html
index fb8a9e9d616..83b60cddf88 100644
--- a/2.7/generated/torch.optim.lr_scheduler.LinearLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.LinearLR.html
@@ -596,7 +596,7 @@
 
                                                                      LinearLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.LinearLR(optimizer, start_factor=0.3333333333333333, end_factor=1.0, total_iters=5, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.LinearLR(optimizer, start_factor=0.3333333333333333, end_factor=1.0, total_iters=5, last_epoch=-1)[source][source]
 
                                                                      Decays the learning rate of each parameter group by linearly changing small multiplicative factor.
                                                                      
 
                                                                      The multiplication is done until the number of epoch reaches a pre-defined milestone: total_iters.
 Notice that such decay can happen simultaneously with other changes to the learning rate
@@ -632,7 +632,7 @@ 
                                                                      LinearLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -643,13 +643,13 @@ 
                                                                      LinearLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -661,7 +661,7 @@ 
                                                                      LinearLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -669,7 +669,7 @@ 
                                                                      LinearLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.MultiStepLR.html b/2.7/generated/torch.optim.lr_scheduler.MultiStepLR.html
index 8cf2d2f368e..977937fc7d5 100644
--- a/2.7/generated/torch.optim.lr_scheduler.MultiStepLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.MultiStepLR.html
@@ -596,7 +596,7 @@
 
                                                                      MultiStepLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones, gamma=0.1, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones, gamma=0.1, last_epoch=-1)[source][source]
 
                                                                      Decays the learning rate of each parameter group by gamma once the number of epoch reaches one of the milestones.
                                                                      
 
                                                                      Notice that such decay can happen simultaneously with other changes to the learning rate
 from outside this scheduler. When last_epoch=-1, sets initial lr as lr.
                                                                      
@@ -625,7 +625,7 @@ 
                                                                      MultiStepLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -636,13 +636,13 @@ 
                                                                      MultiStepLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -654,7 +654,7 @@ 
                                                                      MultiStepLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -662,7 +662,7 @@ 
                                                                      MultiStepLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.MultiplicativeLR.html b/2.7/generated/torch.optim.lr_scheduler.MultiplicativeLR.html
index a0d0517fd6a..6b8c8a7af0b 100644
--- a/2.7/generated/torch.optim.lr_scheduler.MultiplicativeLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.MultiplicativeLR.html
@@ -596,7 +596,7 @@
 
                                                                      MultiplicativeLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.MultiplicativeLR(optimizer, lr_lambda, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.MultiplicativeLR(optimizer, lr_lambda, last_epoch=-1)[source][source]
 
                                                                      Multiply the learning rate of each parameter group by the factor given in the specified function.
                                                                      
 
                                                                      When last_epoch=-1, set initial lr as lr.
                                                                      
 
                                                                      
@@ -621,7 +621,7 @@ 
                                                                      MultiplicativeLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -632,13 +632,13 @@ 
                                                                      MultiplicativeLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -650,7 +650,7 @@ 
                                                                      MultiplicativeLR
 
                                                                      
-state_dict()[source][source]
                                                                      
+state_dict()[source][source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
@@ -660,7 +660,7 @@ 
                                                                      MultiplicativeLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.OneCycleLR.html b/2.7/generated/torch.optim.lr_scheduler.OneCycleLR.html
index fd8a955c6f4..a4149e855ec 100644
--- a/2.7/generated/torch.optim.lr_scheduler.OneCycleLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.OneCycleLR.html
@@ -596,7 +596,7 @@
 
                                                                      OneCycleLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.OneCycleLR(optimizer, max_lr, total_steps=None, epochs=None, steps_per_epoch=None, pct_start=0.3, anneal_strategy='cos', cycle_momentum=True, base_momentum=0.85, max_momentum=0.95, div_factor=25.0, final_div_factor=10000.0, three_phase=False, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.OneCycleLR(optimizer, max_lr, total_steps=None, epochs=None, steps_per_epoch=None, pct_start=0.3, anneal_strategy='cos', cycle_momentum=True, base_momentum=0.85, max_momentum=0.95, div_factor=25.0, final_div_factor=10000.0, three_phase=False, last_epoch=-1)[source][source]
 
                                                                      Sets the learning rate of each parameter group according to the 1cycle learning rate policy.
                                                                      
 
                                                                      The 1cycle policy anneals the learning rate from an initial learning rate to some maximum
 learning rate and then from that maximum learning rate to some minimum learning rate much
@@ -692,7 +692,7 @@ 
                                                                      OneCycleLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -703,13 +703,13 @@ 
                                                                      OneCycleLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -721,7 +721,7 @@ 
                                                                      OneCycleLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -729,7 +729,7 @@ 
                                                                      OneCycleLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.PolynomialLR.html b/2.7/generated/torch.optim.lr_scheduler.PolynomialLR.html
index 25c1d41876b..4a26731856e 100644
--- a/2.7/generated/torch.optim.lr_scheduler.PolynomialLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.PolynomialLR.html
@@ -596,7 +596,7 @@
 
                                                                      PolynomialLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.PolynomialLR(optimizer, total_iters=5, power=1.0, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.PolynomialLR(optimizer, total_iters=5, power=1.0, last_epoch=-1)[source][source]
 
                                                                      Decays the learning rate of each parameter group using a polynomial function in the given total_iters.
                                                                      
 
                                                                      When last_epoch=-1, sets initial lr as lr.
                                                                      
 
                                                                      
@@ -624,7 +624,7 @@ 
                                                                      PolynomialLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -635,13 +635,13 @@ 
                                                                      PolynomialLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -653,7 +653,7 @@ 
                                                                      PolynomialLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -661,7 +661,7 @@ 
                                                                      PolynomialLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html b/2.7/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html
index fd9352751ca..a8c212b2101 100644
--- a/2.7/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html
+++ b/2.7/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html
@@ -596,7 +596,7 @@
 
                                                                      ReduceLROnPlateau
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', factor=0.1, patience=10, threshold=0.0001, threshold_mode='rel', cooldown=0, min_lr=0, eps=1e-08)[source][source]
                                                                      
+class torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', factor=0.1, patience=10, threshold=0.0001, threshold_mode='rel', cooldown=0, min_lr=0, eps=1e-08)[source][source]
 
                                                                      Reduce learning rate when a metric has stopped improving.
                                                                      
 
                                                                      Models often benefit from reducing the learning rate by a factor
 of 2-10 once learning stagnates. This scheduler reads a metrics
@@ -654,7 +654,7 @@ 
                                                                      ReduceLROnPlateau
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -665,7 +665,7 @@ 
                                                                      ReduceLROnPlateau
 
                                                                      
-get_lr()[source]
                                                                      
+get_lr()[source]
 
                                                                      Compute learning rate using chainable form of the scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -676,13 +676,13 @@ 
                                                                      ReduceLROnPlateau
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-step(metrics, epoch=None)[source][source]
                                                                      
+step(metrics, epoch=None)[source][source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.lr_scheduler.SequentialLR.html b/2.7/generated/torch.optim.lr_scheduler.SequentialLR.html
index 83273e4e62f..a6f022c7f8c 100644
--- a/2.7/generated/torch.optim.lr_scheduler.SequentialLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.SequentialLR.html
@@ -596,7 +596,7 @@
 
                                                                      SequentialLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.SequentialLR(optimizer, schedulers, milestones, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.SequentialLR(optimizer, schedulers, milestones, last_epoch=-1)[source][source]
 
                                                                      Contains a list of schedulers expected to be called sequentially during the optimization process.
                                                                      
 
                                                                      Specifically, the schedulers will be called according to the milestone points, which should provide exact
 intervals by which each scheduler should be called at a given epoch.
                                                                      
@@ -628,7 +628,7 @@ 
                                                                      SequentialLR
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -639,7 +639,7 @@ 
                                                                      SequentialLR
 
                                                                      
-get_lr()[source]
                                                                      
+get_lr()[source]
 
                                                                      Compute learning rate using chainable form of the scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -650,7 +650,7 @@ 
                                                                      SequentialLR
 
                                                                      
-load_state_dict(state_dict)[source][source]
                                                                      
+load_state_dict(state_dict)[source][source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -662,14 +662,14 @@ 
                                                                      SequentialLR
 
                                                                      
-recursive_undo(sched=None)[source][source]
                                                                      
+recursive_undo(sched=None)[source][source]
 
                                                                      Recursively undo any step performed by the initialisation of
 schedulers.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-state_dict()[source][source]
                                                                      
+state_dict()[source][source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
@@ -678,7 +678,7 @@ 
                                                                      SequentialLR
 
                                                                      
-step()[source][source]
                                                                      
+step()[source][source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
diff --git a/2.7/generated/torch.optim.lr_scheduler.StepLR.html b/2.7/generated/torch.optim.lr_scheduler.StepLR.html
index e5255d6f50b..54cbad078c7 100644
--- a/2.7/generated/torch.optim.lr_scheduler.StepLR.html
+++ b/2.7/generated/torch.optim.lr_scheduler.StepLR.html
@@ -596,7 +596,7 @@
 
                                                                      StepLR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.lr_scheduler.StepLR(optimizer, step_size, gamma=0.1, last_epoch=-1)[source][source]
                                                                      
+class torch.optim.lr_scheduler.StepLR(optimizer, step_size, gamma=0.1, last_epoch=-1)[source][source]
 
                                                                      Decays the learning rate of each parameter group by gamma every step_size epochs.
                                                                      
 
                                                                      Notice that such decay can happen simultaneously with other changes to the learning rate
 from outside this scheduler. When last_epoch=-1, sets initial lr as lr.
                                                                      
@@ -626,7 +626,7 @@ 
                                                                      StepLR
 
                                                                      
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -637,13 +637,13 @@ 
                                                                      StepLR
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Compute the learning rate of each parameter group.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -655,7 +655,7 @@ 
                                                                      StepLR
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -663,7 +663,7 @@ 
                                                                      StepLR
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.swa_utils.AveragedModel.html b/2.7/generated/torch.optim.swa_utils.AveragedModel.html
index 446e9a45e69..617fd36732e 100644
--- a/2.7/generated/torch.optim.swa_utils.AveragedModel.html
+++ b/2.7/generated/torch.optim.swa_utils.AveragedModel.html
@@ -596,7 +596,7 @@
 
                                                                      AveragedModel
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.swa_utils.AveragedModel(model, device=None, avg_fn=None, multi_avg_fn=None, use_buffers=False)[source][source]
                                                                      
+class torch.optim.swa_utils.AveragedModel(model, device=None, avg_fn=None, multi_avg_fn=None, use_buffers=False)[source][source]
 
                                                                      Implements averaged model for Stochastic Weight Averaging (SWA) and Exponential Moving Average (EMA).
                                                                      
 
                                                                      Stochastic Weight Averaging was proposed in Averaging Weights Leads to
 Wider Optima and Better Generalization by Pavel Izmailov, Dmitrii
@@ -685,7 +685,7 @@ 
                                                                      AveragedModel
 
                                                                      
-add_module(name, module)[source]
                                                                      
+add_module(name, module)[source]
 
                                                                      Add a child module to the current module.
                                                                      
 
                                                                      The module can be accessed as an attribute using the given name.
                                                                      
 
                                                                      
@@ -701,7 +701,7 @@ 
                                                                      AveragedModel
 
                                                                      
-apply(fn)[source]
                                                                      
+apply(fn)[source]
 
                                                                      Apply fn recursively to every submodule (as returned by .children()) as well as self.
                                                                      
 
                                                                      Typical use includes initializing the parameters of a model
 (see also torch.nn.init).
                                                                      
@@ -743,7 +743,7 @@ 
                                                                      AveragedModel
 
                                                                      
-bfloat16()[source]
                                                                      
+bfloat16()[source]
 
                                                                      Casts all floating point parameters and buffers to bfloat16 datatype.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -761,7 +761,7 @@ 
                                                                      AveragedModel
 
                                                                      
-buffers(recurse=True)[source]
                                                                      
+buffers(recurse=True)[source]
 
                                                                      Return an iterator over module buffers.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -787,7 +787,7 @@ 
                                                                      AveragedModel
 
                                                                      
-children()[source]
                                                                      
+children()[source]
 
                                                                      Return an iterator over immediate children modules.
                                                                      
 
                                                                      
 
                                                                      Yields
                                                                      
@@ -801,7 +801,7 @@ 
                                                                      AveragedModel
 
                                                                      
-compile(*args, **kwargs)[source]
                                                                      
+compile(*args, **kwargs)[source]
 
                                                                      Compile this Module’s forward using torch.compile().
                                                                      
 
                                                                      This Module’s __call__ method is compiled and all arguments are passed as-is
 to torch.compile().
                                                                      
@@ -810,7 +810,7 @@ 
                                                                      AveragedModel
 
                                                                      
-cpu()[source]
                                                                      
+cpu()[source]
 
                                                                      Move all model parameters and buffers to the CPU.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -828,7 +828,7 @@ 
                                                                      AveragedModel
 
                                                                      
-cuda(device=None)[source]
                                                                      
+cuda(device=None)[source]
 
                                                                      Move all model parameters and buffers to the GPU.
                                                                      
 
                                                                      This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -853,7 +853,7 @@ 
                                                                      AveragedModel
 
                                                                      
-double()[source]
                                                                      
+double()[source]
 
                                                                      Casts all floating point parameters and buffers to double datatype.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -871,7 +871,7 @@ 
                                                                      AveragedModel
 
                                                                      
-eval()[source]
                                                                      
+eval()[source]
 
                                                                      Set the module in evaluation mode.
                                                                      
 
                                                                      This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -892,7 +892,7 @@ 
                                                                      AveragedModel
 
                                                                      
-extra_repr()[source]
                                                                      
+extra_repr()[source]
 
                                                                      Return the extra representation of the module.
                                                                      
 
                                                                      To print customized extra information, you should re-implement
 this method in your own modules. Both single-line and multi-line
@@ -906,7 +906,7 @@ 
                                                                      AveragedModel
 
                                                                      
-float()[source]
                                                                      
+float()[source]
 
                                                                      Casts all floating point parameters and buffers to float datatype.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -924,13 +924,13 @@ 
                                                                      AveragedModel
 
                                                                      
-forward(*args, **kwargs)[source][source]
                                                                      
+forward(*args, **kwargs)[source][source]
 
                                                                      Forward pass.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-get_buffer(target)[source]
                                                                      
+get_buffer(target)[source]
 
                                                                      Return the buffer given by target if it exists, otherwise throw an error.
                                                                      
 
                                                                      See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -957,7 +957,7 @@ 
                                                                      AveragedModel
 
                                                                      
-get_extra_state()[source]
                                                                      
+get_extra_state()[source]
 
                                                                      Return any extra state to include in the module’s state_dict.
                                                                      
 
                                                                      Implement this and a corresponding set_extra_state() for your module
 if you need to store extra state. This function is called when building the
@@ -978,7 +978,7 @@ 
                                                                      AveragedModel
 
                                                                      
-get_parameter(target)[source]
                                                                      
+get_parameter(target)[source]
 
                                                                      Return the parameter given by target if it exists, otherwise throw an error.
                                                                      
 
                                                                      See the docstring for get_submodule for a more detailed
 explanation of this method’s functionality as well as how to
@@ -1005,7 +1005,7 @@ 
                                                                      AveragedModel
 
                                                                      
-get_submodule(target)[source]
                                                                      
+get_submodule(target)[source]
 
                                                                      Return the submodule given by target if it exists, otherwise throw an error.
                                                                      
 
                                                                      For example, let’s say you have an nn.Module A that
 looks like this:
                                                                      
@@ -1054,7 +1054,7 @@ 
                                                                      AveragedModel
 
                                                                      
-half()[source]
                                                                      
+half()[source]
 
                                                                      Casts all floating point parameters and buffers to half datatype.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -1072,7 +1072,7 @@ 
                                                                      AveragedModel
 
                                                                      
-ipu(device=None)[source]
                                                                      
+ipu(device=None)[source]
 
                                                                      Move all model parameters and buffers to the IPU.
                                                                      
 
                                                                      This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1097,7 +1097,7 @@ 
                                                                      AveragedModel
 
                                                                      
-load_state_dict(state_dict, strict=True, assign=False)[source]
                                                                      
+load_state_dict(state_dict, strict=True, assign=False)[source]
 
                                                                      Copy parameters and buffers from state_dict into this module and its descendants.
                                                                      
 
                                                                      If strict is True, then
 the keys of state_dict must exactly match the keys returned
@@ -1152,7 +1152,7 @@ 
                                                                      AveragedModel
 
                                                                      
-modules()[source]
                                                                      
+modules()[source]
 
                                                                      Return an iterator over all modules in the network.
                                                                      
 
                                                                      
 
                                                                      Yields
                                                                      
@@ -1184,7 +1184,7 @@ 
                                                                      AveragedModel
 
                                                                      
-mtia(device=None)[source]
                                                                      
+mtia(device=None)[source]
 
                                                                      Move all model parameters and buffers to the MTIA.
                                                                      
 
                                                                      This also makes associated parameters and buffers different objects. So
 it should be called before constructing the optimizer if the module will
@@ -1209,7 +1209,7 @@ 
                                                                      AveragedModel
 
                                                                      
-named_buffers(prefix='', recurse=True, remove_duplicate=True)[source]
                                                                      
+named_buffers(prefix='', recurse=True, remove_duplicate=True)[source]
 
                                                                      Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -1238,7 +1238,7 @@ 
                                                                      AveragedModel
 
                                                                      
-named_children()[source]
                                                                      
+named_children()[source]
 
                                                                      Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
                                                                      
 
                                                                      
 
                                                                      Yields
                                                                      
@@ -1258,7 +1258,7 @@ 
                                                                      AveragedModel
 
                                                                      
-named_modules(memo=None, prefix='', remove_duplicate=True)[source]
                                                                      
+named_modules(memo=None, prefix='', remove_duplicate=True)[source]
 
                                                                      Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -1295,7 +1295,7 @@ 
                                                                      AveragedModel
 
                                                                      
-named_parameters(prefix='', recurse=True, remove_duplicate=True)[source]
                                                                      
+named_parameters(prefix='', recurse=True, remove_duplicate=True)[source]
 
                                                                      Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -1325,7 +1325,7 @@ 
                                                                      AveragedModel
 
                                                                      
-parameters(recurse=True)[source]
                                                                      
+parameters(recurse=True)[source]
 
                                                                      Return an iterator over module parameters.
                                                                      
 
                                                                      This is typically passed to an optimizer.
                                                                      
 
                                                                      
@@ -1352,7 +1352,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_backward_hook(hook)[source]
                                                                      
+register_backward_hook(hook)[source]
 
                                                                      Register a backward hook on the module.
                                                                      
 
                                                                      This function is deprecated in favor of register_full_backward_hook() and
 the behavior of this function will change in future versions.
                                                                      
@@ -1369,7 +1369,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_buffer(name, tensor, persistent=True)[source]
                                                                      
+register_buffer(name, tensor, persistent=True)[source]
 
                                                                      Add a buffer to the module.
                                                                      
 
                                                                      This is typically used to register a buffer that should not to be
 considered a model parameter. For example, BatchNorm’s running_mean
@@ -1401,7 +1401,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source]
                                                                      
+register_forward_hook(hook, *, prepend=False, with_kwargs=False, always_call=False)[source]
 
                                                                      Register a forward hook on the module.
                                                                      
 
                                                                      The hook will be called every time after forward() has computed an output.
                                                                      
 
                                                                      If with_kwargs is False or not specified, the input contains only
@@ -1452,7 +1452,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source]
                                                                      
+register_forward_pre_hook(hook, *, prepend=False, with_kwargs=False)[source]
 
                                                                      Register a forward pre-hook on the module.
                                                                      
 
                                                                      The hook will be called every time before forward() is invoked.
                                                                      
 
                                                                      If with_kwargs is false or not specified, the input contains only
@@ -1502,7 +1502,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_full_backward_hook(hook, prepend=False)[source]
                                                                      
+register_full_backward_hook(hook, prepend=False)[source]
 
                                                                      Register a backward hook on the module.
                                                                      
 
                                                                      The hook will be called every time the gradients with respect to a module
 are computed, i.e. the hook will execute if and only if the gradients with
@@ -1553,7 +1553,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_full_backward_pre_hook(hook, prepend=False)[source]
                                                                      
+register_full_backward_pre_hook(hook, prepend=False)[source]
 
                                                                      Register a backward pre-hook on the module.
                                                                      
 
                                                                      The hook will be called every time the gradients for the module are computed.
 The hook should have the following signature:
                                                                      
@@ -1599,7 +1599,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_load_state_dict_post_hook(hook)[source]
                                                                      
+register_load_state_dict_post_hook(hook)[source]
 
                                                                      Register a post-hook to be run after module’s load_state_dict() is called.
                                                                      
 
                                                                      
 
                                                                      It should have the following signature::
                                                                      hook(module, incompatible_keys) -> None
                                                                      
@@ -1629,7 +1629,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_load_state_dict_pre_hook(hook)[source]
                                                                      
+register_load_state_dict_pre_hook(hook)[source]
 
                                                                      Register a pre-hook to be run before module’s load_state_dict() is called.
                                                                      
 
                                                                      
 
                                                                      It should have the following signature::
                                                                      hook(module, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs) -> None  # noqa: B950
                                                                      
@@ -1645,7 +1645,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_module(name, module)[source]
                                                                      
+register_module(name, module)[source]
 
                                                                      Alias for add_module().
                                                                      
 
                                                                      
 
                                                                      
@@ -1653,7 +1653,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_parameter(name, param)[source]
                                                                      
+register_parameter(name, param)[source]
 
                                                                      Add a parameter to the module.
                                                                      
 
                                                                      The parameter can be accessed as an attribute using given name.
                                                                      
 
                                                                      
@@ -1672,7 +1672,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_state_dict_post_hook(hook)[source]
                                                                      
+register_state_dict_post_hook(hook)[source]
 
                                                                      Register a post-hook for the state_dict() method.
                                                                      
 
                                                                      
 
                                                                      It should have the following signature::
                                                                      hook(module, state_dict, prefix, local_metadata) -> None
                                                                      
@@ -1683,7 +1683,7 @@ 
                                                                      AveragedModel
 
                                                                      
-register_state_dict_pre_hook(hook)[source]
                                                                      
+register_state_dict_pre_hook(hook)[source]
 
                                                                      Register a pre-hook for the state_dict() method.
                                                                      
 
                                                                      
 
                                                                      It should have the following signature::
                                                                      hook(module, prefix, keep_vars) -> None
                                                                      
@@ -1695,7 +1695,7 @@ 
                                                                      AveragedModel
 
                                                                      
-requires_grad_(requires_grad=True)[source]
                                                                      
+requires_grad_(requires_grad=True)[source]
 
                                                                      Change if autograd should record operations on parameters in this module.
                                                                      
 
                                                                      This method sets the parameters’ requires_grad attributes
 in-place.
                                                                      
@@ -1719,7 +1719,7 @@ 
                                                                      AveragedModel
 
                                                                      
-set_extra_state(state)[source]
                                                                      
+set_extra_state(state)[source]
 
                                                                      Set extra state contained in the loaded state_dict.
                                                                      
 
                                                                      This function is called from load_state_dict() to handle any extra state
 found within the state_dict. Implement this function and a corresponding
@@ -1734,7 +1734,7 @@ 
                                                                      AveragedModel
 
                                                                      
-set_submodule(target, module, strict=False)[source]
                                                                      
+set_submodule(target, module, strict=False)[source]
 
                                                                      Set the submodule given by target if it exists, otherwise throw an error.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -1792,7 +1792,7 @@ 
                                                                      AveragedModel
 
                                                                      
-share_memory()[source]
                                                                      
+share_memory()[source]
 
                                                                      See torch.Tensor.share_memory_().
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -1803,7 +1803,7 @@ 
                                                                      AveragedModel
 
                                                                      
-state_dict(*args, destination=None, prefix='', keep_vars=False)[source]
                                                                      
+state_dict(*args, destination=None, prefix='', keep_vars=False)[source]
 
                                                                      Return a dictionary containing references to the whole state of the module.
                                                                      
 
                                                                      Both parameters and persistent buffers (e.g. running averages) are
 included. Keys are corresponding parameter and buffer names.
@@ -1856,27 +1856,27 @@ 
                                                                      AveragedModel
 
                                                                      
-to(*args, **kwargs)[source]
                                                                      
+to(*args, **kwargs)[source]
 
                                                                      Move and/or cast the parameters and buffers.
                                                                      
 
                                                                      This can be called as
                                                                      
 
                                                                      
 
                                                                      
-to(device=None, dtype=None, non_blocking=False)[source]
                                                                      
+to(device=None, dtype=None, non_blocking=False)[source]
 
                                                                      
 
 
                                                                      
 
                                                                      
-to(dtype, non_blocking=False)[source]
                                                                      
+to(dtype, non_blocking=False)[source]
 
                                                                      
 
 
                                                                      
 
                                                                      
-to(tensor, non_blocking=False)[source]
                                                                      
+to(tensor, non_blocking=False)[source]
 
                                                                      
 
 
                                                                      
 
                                                                      
-to(memory_format=torch.channels_last)[source]
                                                                      
+to(memory_format=torch.channels_last)[source]
 
                                                                      
 
 
                                                                      Its signature is similar to torch.Tensor.to(), but only accepts
@@ -1955,7 +1955,7 @@ 
                                                                      AveragedModel
 
                                                                      
-to_empty(*, device, recurse=True)[source]
                                                                      
+to_empty(*, device, recurse=True)[source]
 
                                                                      Move the parameters and buffers to the specified device without copying storage.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -1977,7 +1977,7 @@ 
                                                                      AveragedModel
 
                                                                      
-train(mode=True)[source]
                                                                      
+train(mode=True)[source]
 
                                                                      Set the module in training mode.
                                                                      
 
                                                                      This has an effect only on certain modules. See the documentation of
 particular modules for details of their behaviors in training/evaluation
@@ -1999,7 +1999,7 @@ 
                                                                      AveragedModel
 
                                                                      
-type(dst_type)[source]
                                                                      
+type(dst_type)[source]
 
                                                                      Casts all parameters and buffers to dst_type.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -2020,7 +2020,7 @@ 
                                                                      AveragedModel
 
                                                                      
-update_parameters(model)[source][source]
                                                                      
+update_parameters(model)[source][source]
 
                                                                      Update model parameters.
                                                                      
 
                                                                      
 
                                                                      
@@ -2028,7 +2028,7 @@ 
                                                                      AveragedModel
 
                                                                      
-xpu(device=None)[source]
                                                                      
+xpu(device=None)[source]
 
                                                                      Move all model parameters and buffers to the XPU.
                                                                      
 
                                                                      This also makes associated parameters and buffers different objects. So
 it should be called before constructing optimizer if the module will
@@ -2053,7 +2053,7 @@ 
                                                                      AveragedModel
 
                                                                      
-zero_grad(set_to_none=True)[source]
                                                                      
+zero_grad(set_to_none=True)[source]
 
                                                                      Reset gradients of all model parameters.
                                                                      
 
                                                                      See similar function under torch.optim.Optimizer for more context.
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.optim.swa_utils.SWALR.html b/2.7/generated/torch.optim.swa_utils.SWALR.html
index dc50ad131c5..e1e34104ad0 100644
--- a/2.7/generated/torch.optim.swa_utils.SWALR.html
+++ b/2.7/generated/torch.optim.swa_utils.SWALR.html
@@ -596,7 +596,7 @@
 
                                                                      SWALR
                                                                      
 
                                                                      
 
                                                                      
-class torch.optim.swa_utils.SWALR(optimizer, swa_lr, anneal_epochs=10, anneal_strategy='cos', last_epoch=-1)[source][source]
                                                                      
+class torch.optim.swa_utils.SWALR(optimizer, swa_lr, anneal_epochs=10, anneal_strategy='cos', last_epoch=-1)[source][source]
 
                                                                      Anneals the learning rate in each parameter group to a fixed value.
                                                                      
 
                                                                      This learning rate scheduler is meant to be used with Stochastic Weight
 Averaging (SWA) method (see torch.optim.swa_utils.AveragedModel).
                                                                      
@@ -639,7 +639,7 @@ 
                                                                      SWALR
 
                                                                      
 
                                                                      
 
                                                                      
-get_last_lr()[source]
                                                                      
+get_last_lr()[source]
 
                                                                      Return last computed learning rate by current scheduler.
                                                                      
 
                                                                      
 
                                                                      Return type
                                                                      
@@ -650,13 +650,13 @@ 
                                                                      SWALR
 
 
                                                                      
 
                                                                      
-get_lr()[source][source]
                                                                      
+get_lr()[source][source]
 
                                                                      Get learning rate.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-load_state_dict(state_dict)[source]
                                                                      
+load_state_dict(state_dict)[source]
 
                                                                      Load the scheduler’s state.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
@@ -668,7 +668,7 @@ 
                                                                      SWALR
 
 
                                                                      
 
                                                                      
-state_dict()[source]
                                                                      
+state_dict()[source]
 
                                                                      Return the state of the scheduler as a dict.
                                                                      
 
                                                                      It contains an entry for every variable in self.__dict__ which
 is not the optimizer.
                                                                      
@@ -676,7 +676,7 @@ 
                                                                      SWALR
 
 
                                                                      
 
                                                                      
-step(epoch=None)[source]
                                                                      
+step(epoch=None)[source]
 
                                                                      Perform a step.
                                                                      
 
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.pca_lowrank.html b/2.7/generated/torch.pca_lowrank.html
index ad8851527c4..a1f5f65abaa 100644
--- a/2.7/generated/torch.pca_lowrank.html
+++ b/2.7/generated/torch.pca_lowrank.html
@@ -596,7 +596,7 @@
 
                                                                      torch.pca_lowrank
                                                                      
 
                                                                      
 
                                                                      
-torch.pca_lowrank(A, q=None, center=True, niter=2)[source][source]
                                                                      
+torch.pca_lowrank(A, q=None, center=True, niter=2)[source][source]
 
                                                                      Performs linear Principal Component Analysis (PCA) on a low-rank
 matrix, batches of such matrices, or sparse matrix.
                                                                      
 
                                                                      This function returns a namedtuple (U, S, V) which is the
diff --git a/2.7/generated/torch.quasirandom.SobolEngine.html b/2.7/generated/torch.quasirandom.SobolEngine.html
index 89d559151f8..bc6763fa4ff 100644
--- a/2.7/generated/torch.quasirandom.SobolEngine.html
+++ b/2.7/generated/torch.quasirandom.SobolEngine.html
@@ -596,7 +596,7 @@
 
                                                                      SobolEngine
                                                                      
 
                                                                      
 
                                                                      
-class torch.quasirandom.SobolEngine(dimension, scramble=False, seed=None)[source][source]
                                                                      
+class torch.quasirandom.SobolEngine(dimension, scramble=False, seed=None)[source][source]
 
                                                                      The torch.quasirandom.SobolEngine is an engine for generating
 (scrambled) Sobol sequences. Sobol sequences are an example of low
 discrepancy quasi-random sequences.
                                                                      
@@ -638,7 +638,7 @@ 
                                                                      SobolEngine
 
                                                                      
-draw(n=1, out=None, dtype=None)[source][source]
                                                                      
+draw(n=1, out=None, dtype=None)[source][source]
 
                                                                      Function to draw a sequence of n points from a Sobol sequence.
 Note that the samples are dependent on the previous samples. The size
 of the result is (n,dimension)(n, dimension).
                                                                      
@@ -661,7 +661,7 @@ 
                                                                      SobolEngine
 
                                                                      
-draw_base2(m, out=None, dtype=None)[source][source]
                                                                      
+draw_base2(m, out=None, dtype=None)[source][source]
 
                                                                      Function to draw a sequence of 2**m points from a Sobol sequence.
 Note that the samples are dependent on the previous samples. The size
 of the result is (2∗∗m,dimension)(2**m, dimension).
                                                                      
@@ -683,7 +683,7 @@ 
                                                                      SobolEngine
 
                                                                      
-fast_forward(n)[source][source]
                                                                      
+fast_forward(n)[source][source]
 
                                                                      Function to fast-forward the state of the SobolEngine by
 n steps. This is equivalent to drawing n samples
 without using the samples.
                                                                      
@@ -696,7 +696,7 @@ 
                                                                      SobolEngine
 
                                                                      
-reset()[source][source]
                                                                      
+reset()[source][source]
 
                                                                      Function to reset the SobolEngine to base state.
                                                                      
 
                                                                      
 
diff --git a/2.7/generated/torch.save.html b/2.7/generated/torch.save.html
index 865fefd24ff..0c545970252 100644
--- a/2.7/generated/torch.save.html
+++ b/2.7/generated/torch.save.html
@@ -596,7 +596,7 @@
 
                                                                      torch.save
                                                                      
 
                                                                      
 
                                                                      
-torch.save(obj, f, pickle_module=pickle, pickle_protocol=2, _use_new_zipfile_serialization=True)[source][source]
                                                                      
+torch.save(obj, f, pickle_module=pickle, pickle_protocol=2, _use_new_zipfile_serialization=True)[source][source]
 
                                                                      Saves an object to a disk file.
                                                                      
 
                                                                      See also: Saving and loading tensors
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.seed.html b/2.7/generated/torch.seed.html
index 6f95e06816e..cd0fb285ea5 100644
--- a/2.7/generated/torch.seed.html
+++ b/2.7/generated/torch.seed.html
@@ -596,7 +596,7 @@
 
                                                                      torch.seed
                                                                      
 
                                                                      
 
                                                                      
-torch.seed()[source][source]
                                                                      
+torch.seed()[source][source]
 
                                                                      Sets the seed for generating random numbers to a non-deterministic
 random number on all devices. Returns a 64 bit number used to seed the RNG.
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.set_default_device.html b/2.7/generated/torch.set_default_device.html
index 5ea44a66c9a..0a814ff1990 100644
--- a/2.7/generated/torch.set_default_device.html
+++ b/2.7/generated/torch.set_default_device.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_default_device
                                                                      
 
                                                                      
 
                                                                      
-torch.set_default_device(device)[source][source]
                                                                      
+torch.set_default_device(device)[source][source]
 
                                                                      Sets the default torch.Tensor to be allocated on device.  This
 does not affect factory function calls which are called with an explicit
 device argument.  Factory calls will be performed as if they
diff --git a/2.7/generated/torch.set_default_dtype.html b/2.7/generated/torch.set_default_dtype.html
index ac90acb172d..320f5a8e483 100644
--- a/2.7/generated/torch.set_default_dtype.html
+++ b/2.7/generated/torch.set_default_dtype.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_default_dtype
                                                                      
 
                                                                      
 
                                                                      
-torch.set_default_dtype(d, /)[source][source]
                                                                      
+torch.set_default_dtype(d, /)[source][source]
 
                                                                      Sets the default floating point dtype to d. Supports floating point dtype
 as inputs. Other dtypes will cause torch to raise an exception.
                                                                      
 
                                                                      When PyTorch is initialized its default floating point dtype is torch.float32,
diff --git a/2.7/generated/torch.set_default_tensor_type.html b/2.7/generated/torch.set_default_tensor_type.html
index eedc3488144..41b3c0c2ee3 100644
--- a/2.7/generated/torch.set_default_tensor_type.html
+++ b/2.7/generated/torch.set_default_tensor_type.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_default_tensor_type
                                                                      
 
                                                                      
 
                                                                      
-torch.set_default_tensor_type(t, /)[source][source]
                                                                      
+torch.set_default_tensor_type(t, /)[source][source]
 
                                                                      
 
                                                                      Warning
                                                                      
 
                                                                      This function is deprecated as of PyTorch 2.1, please use torch.set_default_dtype() and
diff --git a/2.7/generated/torch.set_deterministic_debug_mode.html b/2.7/generated/torch.set_deterministic_debug_mode.html
index 82278ed6005..790b63cf966 100644
--- a/2.7/generated/torch.set_deterministic_debug_mode.html
+++ b/2.7/generated/torch.set_deterministic_debug_mode.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_deterministic_debug_mode
                                                                      
 
                                                                      
 
                                                                      
-torch.set_deterministic_debug_mode(debug_mode)[source][source]
                                                                      
+torch.set_deterministic_debug_mode(debug_mode)[source][source]
 
                                                                      Sets the debug mode for deterministic operations.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
diff --git a/2.7/generated/torch.set_float32_matmul_precision.html b/2.7/generated/torch.set_float32_matmul_precision.html
index fe9276dfa03..50cc75a112b 100644
--- a/2.7/generated/torch.set_float32_matmul_precision.html
+++ b/2.7/generated/torch.set_float32_matmul_precision.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_float32_matmul_precision
                                                                      
 
                                                                      
 
                                                                      
-torch.set_float32_matmul_precision(precision)[source][source]
                                                                      
+torch.set_float32_matmul_precision(precision)[source][source]
 
                                                                      Sets the internal precision of float32 matrix multiplications.
                                                                      
 
                                                                      Running float32 matrix multiplications in lower precision may significantly increase
 performance, and in some programs the loss of precision has a negligible impact.
                                                                      
diff --git a/2.7/generated/torch.set_printoptions.html b/2.7/generated/torch.set_printoptions.html
index 058af428aac..a00048b3d70 100644
--- a/2.7/generated/torch.set_printoptions.html
+++ b/2.7/generated/torch.set_printoptions.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_printoptions
                                                                      
 
                                                                      
 
                                                                      
-torch.set_printoptions(precision=None, threshold=None, edgeitems=None, linewidth=None, profile=None, sci_mode=None)[source][source]
                                                                      
+torch.set_printoptions(precision=None, threshold=None, edgeitems=None, linewidth=None, profile=None, sci_mode=None)[source][source]
 
                                                                      Set options for printing. Items shamelessly taken from NumPy
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.set_rng_state.html b/2.7/generated/torch.set_rng_state.html
index 23c2a2d0c1b..e28fad9f4eb 100644
--- a/2.7/generated/torch.set_rng_state.html
+++ b/2.7/generated/torch.set_rng_state.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_rng_state
                                                                      
 
                                                                      
 
                                                                      
-torch.set_rng_state(new_state)[source][source]
                                                                      
+torch.set_rng_state(new_state)[source][source]
 
                                                                      Sets the random number generator state.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
diff --git a/2.7/generated/torch.set_warn_always.html b/2.7/generated/torch.set_warn_always.html
index 8b625975cb1..ef14101eb45 100644
--- a/2.7/generated/torch.set_warn_always.html
+++ b/2.7/generated/torch.set_warn_always.html
@@ -596,7 +596,7 @@
 
                                                                      torch.set_warn_always
                                                                      
 
                                                                      
 
                                                                      
-torch.set_warn_always(b, /)[source][source]
                                                                      
+torch.set_warn_always(b, /)[source][source]
 
                                                                      When this flag is False (default) then some PyTorch warnings may only
 appear once per process. This helps avoid excessive warning information.
 Setting it to True causes these warnings to always appear, which may be
diff --git a/2.7/generated/torch.signal.windows.bartlett.html b/2.7/generated/torch.signal.windows.bartlett.html
index 0261221b6bd..bd4bffbd6c4 100644
--- a/2.7/generated/torch.signal.windows.bartlett.html
+++ b/2.7/generated/torch.signal.windows.bartlett.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.bartlett
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.bartlett(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.bartlett(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the Bartlett window.
                                                                      
 
                                                                      The Bartlett window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.blackman.html b/2.7/generated/torch.signal.windows.blackman.html
index f294df7d378..697ebdd6fbc 100644
--- a/2.7/generated/torch.signal.windows.blackman.html
+++ b/2.7/generated/torch.signal.windows.blackman.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.blackman
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.blackman(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.blackman(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the Blackman window.
                                                                      
 
                                                                      The Blackman window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.cosine.html b/2.7/generated/torch.signal.windows.cosine.html
index a11b2ce3d4d..7c4ae76b792 100644
--- a/2.7/generated/torch.signal.windows.cosine.html
+++ b/2.7/generated/torch.signal.windows.cosine.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.cosine
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.cosine(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.cosine(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes a window with a simple cosine waveform, following the same implementation as SciPy.
 This window is also known as the sine window.
                                                                      
 
                                                                      The cosine window is defined as follows:
                                                                      
diff --git a/2.7/generated/torch.signal.windows.exponential.html b/2.7/generated/torch.signal.windows.exponential.html
index 8b0c761c717..a9e86ed287f 100644
--- a/2.7/generated/torch.signal.windows.exponential.html
+++ b/2.7/generated/torch.signal.windows.exponential.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.exponential
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.exponential(M, *, center=None, tau=1.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.exponential(M, *, center=None, tau=1.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes a window with an exponential waveform.
 Also known as Poisson window.
                                                                      
 
                                                                      The exponential window is defined as follows:
                                                                      
diff --git a/2.7/generated/torch.signal.windows.gaussian.html b/2.7/generated/torch.signal.windows.gaussian.html
index 2a25e2ab10c..877343946be 100644
--- a/2.7/generated/torch.signal.windows.gaussian.html
+++ b/2.7/generated/torch.signal.windows.gaussian.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.gaussian
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.gaussian(M, *, std=1.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.gaussian(M, *, std=1.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes a window with a gaussian waveform.
                                                                      
 
                                                                      The gaussian window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.general_cosine.html b/2.7/generated/torch.signal.windows.general_cosine.html
index 9eb43c79fbf..feed330b686 100644
--- a/2.7/generated/torch.signal.windows.general_cosine.html
+++ b/2.7/generated/torch.signal.windows.general_cosine.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.general_cosine
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.general_cosine(M, *, a, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.general_cosine(M, *, a, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the general cosine window.
                                                                      
 
                                                                      The general cosine window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.general_hamming.html b/2.7/generated/torch.signal.windows.general_hamming.html
index c6098f1f513..468cd3f0ece 100644
--- a/2.7/generated/torch.signal.windows.general_hamming.html
+++ b/2.7/generated/torch.signal.windows.general_hamming.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.general_hamming
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.general_hamming(M, *, alpha=0.54, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.general_hamming(M, *, alpha=0.54, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the general Hamming window.
                                                                      
 
                                                                      The general Hamming window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.hamming.html b/2.7/generated/torch.signal.windows.hamming.html
index 5fa04fcc3bd..0bd3def52a9 100644
--- a/2.7/generated/torch.signal.windows.hamming.html
+++ b/2.7/generated/torch.signal.windows.hamming.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.hamming
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.hamming(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.hamming(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the Hamming window.
                                                                      
 
                                                                      The Hamming window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.hann.html b/2.7/generated/torch.signal.windows.hann.html
index 2dcd913abab..4f33124b8df 100644
--- a/2.7/generated/torch.signal.windows.hann.html
+++ b/2.7/generated/torch.signal.windows.hann.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.hann
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.hann(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.hann(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the Hann window.
                                                                      
 
                                                                      The Hann window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.kaiser.html b/2.7/generated/torch.signal.windows.kaiser.html
index a1ccf3d7131..5450876333b 100644
--- a/2.7/generated/torch.signal.windows.kaiser.html
+++ b/2.7/generated/torch.signal.windows.kaiser.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.kaiser
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.kaiser(M, *, beta=12.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.kaiser(M, *, beta=12.0, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the Kaiser window.
                                                                      
 
                                                                      The Kaiser window is defined as follows:
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.signal.windows.nuttall.html b/2.7/generated/torch.signal.windows.nuttall.html
index 6a4630560ec..cb3bc122cf7 100644
--- a/2.7/generated/torch.signal.windows.nuttall.html
+++ b/2.7/generated/torch.signal.windows.nuttall.html
@@ -596,7 +596,7 @@
 
                                                                      torch.signal.windows.nuttall
                                                                      
 
                                                                      
 
                                                                      
-torch.signal.windows.nuttall(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
                                                                      
+torch.signal.windows.nuttall(M, *, sym=True, dtype=None, layout=torch.strided, device=None, requires_grad=False)[source][source]
 
                                                                      Computes the minimum 4-term Blackman-Harris window according to Nuttall.
                                                                      
 
                                                                      
 wn=1−0.36358cos⁡(zn)+0.48917cos⁡(2zn)−0.13659cos⁡(3zn)+0.01064cos⁡(4zn)w_n = 1 - 0.36358 \cos{(z_n)} + 0.48917 \cos{(2z_n)} - 0.13659 \cos{(3z_n)} + 0.01064 \cos{(4z_n)}
diff --git a/2.7/generated/torch.sparse.as_sparse_gradcheck.html b/2.7/generated/torch.sparse.as_sparse_gradcheck.html
index 83f18ea2f63..87fe03c2bce 100644
--- a/2.7/generated/torch.sparse.as_sparse_gradcheck.html
+++ b/2.7/generated/torch.sparse.as_sparse_gradcheck.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sparse.as_sparse_gradcheck
                                                                      
 
                                                                      
 
                                                                      
-torch.sparse.as_sparse_gradcheck(gradcheck)[source][source]
                                                                      
+torch.sparse.as_sparse_gradcheck(gradcheck)[source][source]
 
                                                                      Decorate function, to extend gradcheck for sparse tensors.
                                                                      
 
                                                                      Decorator for torch.autograd.gradcheck or its functools.partial
 variants that extends the gradcheck function with support to input
diff --git a/2.7/generated/torch.sparse.check_sparse_tensor_invariants.html b/2.7/generated/torch.sparse.check_sparse_tensor_invariants.html
index da9e59113b1..5eabdfdfb73 100644
--- a/2.7/generated/torch.sparse.check_sparse_tensor_invariants.html
+++ b/2.7/generated/torch.sparse.check_sparse_tensor_invariants.html
@@ -596,7 +596,7 @@
 
                                                                      check_sparse_tensor_invariants
                                                                      
 
                                                                      
 
                                                                      
-class torch.sparse.check_sparse_tensor_invariants(enable=True)[source][source]
                                                                      
+class torch.sparse.check_sparse_tensor_invariants(enable=True)[source][source]
 
                                                                      A tool to control checking sparse tensor invariants.
                                                                      
 
                                                                      The following options exists to manage sparsr tensor invariants
 checking in sparse tensor construction:
                                                                      
@@ -639,14 +639,14 @@ 
                                                                      check_sparse_tensor_invariants
 
                                                                      
-static disable()[source][source]
                                                                      
+static disable()[source][source]
 
                                                                      Disable sparse tensor invariants checking in sparse tensor constructors.
                                                                      
 
                                                                      See torch.sparse.check_sparse_tensor_invariants.enable() for more information.
                                                                      
 
                                                                      
 
 
                                                                      
 
                                                                      
-static enable()[source][source]
                                                                      
+static enable()[source][source]
 
                                                                      Enable sparse tensor invariants checking in sparse tensor constructors.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
@@ -665,7 +665,7 @@ 
                                                                      check_sparse_tensor_invariants
 
                                                                      
-static is_enabled()[source][source]
                                                                      
+static is_enabled()[source][source]
 
                                                                      Return True if the sparse tensor invariants checking is enabled.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
diff --git a/2.7/generated/torch.sparse.sum.html b/2.7/generated/torch.sparse.sum.html
index a0177104602..d629ab8f65b 100644
--- a/2.7/generated/torch.sparse.sum.html
+++ b/2.7/generated/torch.sparse.sum.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sparse.sum
                                                                      
 
                                                                      
 
                                                                      
-torch.sparse.sum(input, dim=None, dtype=None)[source][source]
                                                                      
+torch.sparse.sum(input, dim=None, dtype=None)[source][source]
 
                                                                      Return the sum of each row of the given sparse tensor.
                                                                      
 
                                                                      Returns the sum of each row of the sparse tensor input in the given
 dimensions dim. If dim is a list of dimensions,
diff --git a/2.7/generated/torch.split.html b/2.7/generated/torch.split.html
index df3d6944e71..7d4ba39645a 100644
--- a/2.7/generated/torch.split.html
+++ b/2.7/generated/torch.split.html
@@ -596,7 +596,7 @@
 
                                                                      torch.split
                                                                      
 
                                                                      
 
                                                                      
-torch.split(tensor, split_size_or_sections, dim=0)[source][source]
                                                                      
+torch.split(tensor, split_size_or_sections, dim=0)[source][source]
 
                                                                      Splits the tensor into chunks. Each chunk is a view of the original tensor.
                                                                      
 
                                                                      If split_size_or_sections is an integer type, then tensor will
 be split into equally sized chunks (if possible). Last chunk will be smaller if
diff --git a/2.7/generated/torch.stft.html b/2.7/generated/torch.stft.html
index df18b576f5a..353bb6cbd76 100644
--- a/2.7/generated/torch.stft.html
+++ b/2.7/generated/torch.stft.html
@@ -596,7 +596,7 @@
 
                                                                      torch.stft
                                                                      
 
                                                                      
 
                                                                      
-torch.stft(input, n_fft, hop_length=None, win_length=None, window=None, center=True, pad_mode='reflect', normalized=False, onesided=None, return_complex=None, align_to_window=None)[source][source]
                                                                      
+torch.stft(input, n_fft, hop_length=None, win_length=None, window=None, center=True, pad_mode='reflect', normalized=False, onesided=None, return_complex=None, align_to_window=None)[source][source]
 
                                                                      Short-time Fourier transform (STFT).
                                                                      
 
                                                                      
 
                                                                      Warning
                                                                      
diff --git a/2.7/generated/torch.svd_lowrank.html b/2.7/generated/torch.svd_lowrank.html
index 5b3cd69c18a..f124438746e 100644
--- a/2.7/generated/torch.svd_lowrank.html
+++ b/2.7/generated/torch.svd_lowrank.html
@@ -596,7 +596,7 @@
 
                                                                      torch.svd_lowrank
                                                                      
 
                                                                      
 
                                                                      
-torch.svd_lowrank(A, q=6, niter=2, M=None)[source][source]
                                                                      
+torch.svd_lowrank(A, q=6, niter=2, M=None)[source][source]
 
                                                                      Return the singular value decomposition (U, S, V) of a matrix,
 batches of matrices, or a sparse matrix AA such that
 A≈Udiag⁡(S)VHA \approx U \operatorname{diag}(S) V^{\text{H}}. In case MM is given, then
diff --git a/2.7/generated/torch.sym_float.html b/2.7/generated/torch.sym_float.html
index 7bcc40b5687..f3bc18526ce 100644
--- a/2.7/generated/torch.sym_float.html
+++ b/2.7/generated/torch.sym_float.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_float
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_float(a)[source][source]
                                                                      
+torch.sym_float(a)[source][source]
 
                                                                      SymInt-aware utility for float casting.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.sym_fresh_size.html b/2.7/generated/torch.sym_fresh_size.html
index 6ff34d6196b..ab302ffd7cc 100644
--- a/2.7/generated/torch.sym_fresh_size.html
+++ b/2.7/generated/torch.sym_fresh_size.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_fresh_size
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_fresh_size(expr)[source][source]
                                                                      
+torch.sym_fresh_size(expr)[source][source]
 
                                                                      
 
 
                                                                      
diff --git a/2.7/generated/torch.sym_int.html b/2.7/generated/torch.sym_int.html
index 07612bc9495..abd49b8cead 100644
--- a/2.7/generated/torch.sym_int.html
+++ b/2.7/generated/torch.sym_int.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_int
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_int(a)[source][source]
                                                                      
+torch.sym_int(a)[source][source]
 
                                                                      SymInt-aware utility for int casting.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.sym_ite.html b/2.7/generated/torch.sym_ite.html
index 32920ef9e64..38a65a4f1e3 100644
--- a/2.7/generated/torch.sym_ite.html
+++ b/2.7/generated/torch.sym_ite.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_ite
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_ite(b, t, f)[source][source]
                                                                      
+torch.sym_ite(b, t, f)[source][source]
 
                                                                      
 
 
                                                                      
diff --git a/2.7/generated/torch.sym_max.html b/2.7/generated/torch.sym_max.html
index 95ace0ece79..f3d68b94401 100644
--- a/2.7/generated/torch.sym_max.html
+++ b/2.7/generated/torch.sym_max.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_max
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_max(a, b)[source][source]
                                                                      
+torch.sym_max(a, b)[source][source]
 
                                                                      SymInt-aware utility for max which avoids branching on a < b.
 Unlike builtins.max(), this only works for int/float, and it always
 promotes to float if any argument is float (unlike builtins.max, which
diff --git a/2.7/generated/torch.sym_min.html b/2.7/generated/torch.sym_min.html
index 479b3b377d6..202f9f52e42 100644
--- a/2.7/generated/torch.sym_min.html
+++ b/2.7/generated/torch.sym_min.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_min
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_min(a, b)[source][source]
                                                                      
+torch.sym_min(a, b)[source][source]
 
                                                                      SymInt-aware utility for min().
                                                                      
 
                                                                      
 
diff --git a/2.7/generated/torch.sym_not.html b/2.7/generated/torch.sym_not.html
index 5837c671eb9..b05bed84afd 100644
--- a/2.7/generated/torch.sym_not.html
+++ b/2.7/generated/torch.sym_not.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_not
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_not(a)[source][source]
                                                                      
+torch.sym_not(a)[source][source]
 
                                                                      SymInt-aware utility for logical negation.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.sym_sum.html b/2.7/generated/torch.sym_sum.html
index 15648fa697d..58761bfe0fb 100644
--- a/2.7/generated/torch.sym_sum.html
+++ b/2.7/generated/torch.sym_sum.html
@@ -596,7 +596,7 @@
 
                                                                      torch.sym_sum
                                                                      
 
                                                                      
 
                                                                      
-torch.sym_sum(args)[source][source]
                                                                      
+torch.sym_sum(args)[source][source]
 
                                                                      N-ary add which is faster to compute for long lists than iterated binary
 addition.  Only does something special for integers.
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.tensordot.html b/2.7/generated/torch.tensordot.html
index 2d822f3244b..2f71af296aa 100644
--- a/2.7/generated/torch.tensordot.html
+++ b/2.7/generated/torch.tensordot.html
@@ -596,7 +596,7 @@
 
                                                                      torch.tensordot
                                                                      
 
                                                                      
 
                                                                      
-torch.tensordot(a, b, dims=2, out=None)[source][source]
                                                                      
+torch.tensordot(a, b, dims=2, out=None)[source][source]
 
                                                                      Returns a contraction of a and b over multiple dimensions.
                                                                      
 
                                                                      tensordot implements a generalized matrix product.
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.unique.html b/2.7/generated/torch.unique.html
index a5de4556245..e1e1fdc959c 100644
--- a/2.7/generated/torch.unique.html
+++ b/2.7/generated/torch.unique.html
@@ -596,7 +596,7 @@
 
                                                                      torch.unique
                                                                      
 
                                                                      
 
                                                                      
-torch.unique(input, sorted=True, return_inverse=False, return_counts=False, dim=None) → tuple[Tensor, Tensor, Tensor][source]
                                                                      
+torch.unique(input, sorted=True, return_inverse=False, return_counts=False, dim=None) → tuple[Tensor, Tensor, Tensor][source]
 
                                                                      Returns the unique elements of the input tensor.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
diff --git a/2.7/generated/torch.unique_consecutive.html b/2.7/generated/torch.unique_consecutive.html
index 10a2384080d..d1ec03a78fa 100644
--- a/2.7/generated/torch.unique_consecutive.html
+++ b/2.7/generated/torch.unique_consecutive.html
@@ -596,7 +596,7 @@
 
                                                                      torch.unique_consecutive
                                                                      
 
                                                                      
 
                                                                      
-torch.unique_consecutive(*args, **kwargs)[source]
                                                                      
+torch.unique_consecutive(*args, **kwargs)[source]
 
                                                                      Eliminates all but the first element from every consecutive group of equivalent elements.
                                                                      
 
                                                                      
 
                                                                      Note
                                                                      
diff --git a/2.7/generated/torch.unravel_index.html b/2.7/generated/torch.unravel_index.html
index 62e079b1154..eb9132f59e4 100644
--- a/2.7/generated/torch.unravel_index.html
+++ b/2.7/generated/torch.unravel_index.html
@@ -596,7 +596,7 @@
 
                                                                      torch.unravel_index
                                                                      
 
                                                                      
 
                                                                      
-torch.unravel_index(indices, shape)[source][source]
                                                                      
+torch.unravel_index(indices, shape)[source][source]
 
                                                                      Converts a tensor of flat indices into a tuple of coordinate tensors that
 index into an arbitrary tensor of the specified shape.
                                                                      
 
                                                                      
diff --git a/2.7/generated/torch.use_deterministic_algorithms.html b/2.7/generated/torch.use_deterministic_algorithms.html
index b6abbd171c2..f650b43f683 100644
--- a/2.7/generated/torch.use_deterministic_algorithms.html
+++ b/2.7/generated/torch.use_deterministic_algorithms.html
@@ -596,7 +596,7 @@
 
                                                                      torch.use_deterministic_algorithms
                                                                      
 
                                                                      
 
                                                                      
-torch.use_deterministic_algorithms(mode, *, warn_only=False)[source][source]
                                                                      
+torch.use_deterministic_algorithms(mode, *, warn_only=False)[source][source]
 
                                                                      Sets whether PyTorch operations must use “deterministic”
 algorithms. That is, algorithms which, given the same input, and when
 run on the same software and hardware, always produce the same output.
diff --git a/2.7/generated/torch.utils.generate_methods_for_privateuse1_backend.html b/2.7/generated/torch.utils.generate_methods_for_privateuse1_backend.html
index df0dbdfcef0..00e9c8d5340 100644
--- a/2.7/generated/torch.utils.generate_methods_for_privateuse1_backend.html
+++ b/2.7/generated/torch.utils.generate_methods_for_privateuse1_backend.html
@@ -596,7 +596,7 @@
 
                                                                      torch.utils.generate_methods_for_privateuse1_backend
                                                                      
 
                                                                      
 
                                                                      
-torch.utils.generate_methods_for_privateuse1_backend(for_tensor=True, for_module=True, for_packed_sequence=True, for_storage=False, unsupported_dtype=None)[source][source]
                                                                      
+torch.utils.generate_methods_for_privateuse1_backend(for_tensor=True, for_module=True, for_packed_sequence=True, for_storage=False, unsupported_dtype=None)[source][source]
 
                                                                      Automatically generate attributes and methods for the custom backend after rename privateuse1 backend.
                                                                      
 
                                                                      In the default scenario, storage-related methods will not be generated automatically.
                                                                      
 
                                                                      When you implement kernels for various torch operations, and register them to the PrivateUse1 dispatch key.
diff --git a/2.7/generated/torch.utils.get_cpp_backtrace.html b/2.7/generated/torch.utils.get_cpp_backtrace.html
index 6f6b0fcdb05..c7afe1ce060 100644
--- a/2.7/generated/torch.utils.get_cpp_backtrace.html
+++ b/2.7/generated/torch.utils.get_cpp_backtrace.html
@@ -596,7 +596,7 @@
 
                                                                      torch.utils.get_cpp_backtrace
                                                                      
 
                                                                      
 
                                                                      
-torch.utils.get_cpp_backtrace(frames_to_skip=0, maximum_number_of_frames=64)[source][source]
                                                                      
+torch.utils.get_cpp_backtrace(frames_to_skip=0, maximum_number_of_frames=64)[source][source]
 
                                                                      Return a string containing the C++ stack trace of the current thread.
                                                                      
 
                                                                      
 
                                                                      Parameters
                                                                      
diff --git a/2.7/generated/torch.utils.rename_privateuse1_backend.html b/2.7/generated/torch.utils.rename_privateuse1_backend.html
index ce9a6ecf260..6f710b0cc3e 100644
--- a/2.7/generated/torch.utils.rename_privateuse1_backend.html
+++ b/2.7/generated/torch.utils.rename_privateuse1_backend.html
@@ -596,7 +596,7 @@
 
                                                                      torch.utils.rename_privateuse1_backend
                                                                      
 
                                                                      
 
                                                                      
-torch.utils.rename_privateuse1_backend(backend_name)[source][source]
                                                                      
+torch.utils.rename_privateuse1_backend(backend_name)[source][source]
 
                                                                      Rename the privateuse1 backend device to make it more convenient to use as a device name within PyTorch APIs.
                                                                      
 
                                                                      The steps are:
                                                                      
 
                                                                        
diff --git a/2.7/generated/torch.utils.set_module.html b/2.7/generated/torch.utils.set_module.html
index 6286ec644c7..c86ef611cdd 100644
--- a/2.7/generated/torch.utils.set_module.html
+++ b/2.7/generated/torch.utils.set_module.html
@@ -596,7 +596,7 @@
 
                                                                        torch.utils.set_module
                                                                        
 
                                                                        
 
                                                                        
-torch.utils.set_module(obj, mod)[source][source]
                                                                        
+torch.utils.set_module(obj, mod)[source][source]
 
                                                                        Set the module attribute on a python object for a given object for nicer printing
                                                                        
 
                                                                        
 
diff --git a/2.7/generated/torch.utils.swap_tensors.html b/2.7/generated/torch.utils.swap_tensors.html
index 0f2223ffaad..a85e2bcabf8 100644
--- a/2.7/generated/torch.utils.swap_tensors.html
+++ b/2.7/generated/torch.utils.swap_tensors.html
@@ -596,7 +596,7 @@
 
                                                                        torch.utils.swap_tensors
                                                                        
 
                                                                        
 
                                                                        
-torch.utils.swap_tensors(t1, t2)[source][source]
                                                                        
+torch.utils.swap_tensors(t1, t2)[source][source]
 
                                                                        This function swaps the content of the two Tensor objects.
 At a high level, this will make t1 have the content of t2 while preserving
 its identity.
                                                                        
diff --git a/2.7/generated/torch.vmap.html b/2.7/generated/torch.vmap.html
index 7c6573124dd..a3d767924fa 100644
--- a/2.7/generated/torch.vmap.html
+++ b/2.7/generated/torch.vmap.html
@@ -596,7 +596,7 @@
 
                                                                        torch.vmap
                                                                        
 
                                                                        
 
                                                                        
-torch.vmap(func, in_dims=0, out_dims=0, randomness='error', *, chunk_size=None)[source]
                                                                        
+torch.vmap(func, in_dims=0, out_dims=0, randomness='error', *, chunk_size=None)[source]
 
                                                                        vmap is the vectorizing map; vmap(func) returns a new function that
 maps func over some dimension of the inputs. Semantically, vmap
 pushes the map into PyTorch operations called by func, effectively
diff --git a/2.7/generated/torch.xpu.Event.html b/2.7/generated/torch.xpu.Event.html
index f54a220249d..61a4391a6de 100644
--- a/2.7/generated/torch.xpu.Event.html
+++ b/2.7/generated/torch.xpu.Event.html
@@ -596,7 +596,7 @@
 
                                                                        Event
                                                                        
 
                                                                        
 
                                                                        
-class torch.xpu.Event(enable_timing=False)[source][source]
                                                                        
+class torch.xpu.Event(enable_timing=False)[source][source]
 
                                                                        Wrapper around a XPU event.
                                                                        
 
                                                                        XPU events are synchronization markers that can be used to monitor the
 device’s progress, and to synchronize XPU streams.
                                                                        
@@ -611,7 +611,7 @@ 
                                                                        Event
 
                                                                        
 
                                                                        
 
                                                                        
-elapsed_time(end_event)[source][source]
                                                                        
+elapsed_time(end_event)[source][source]
 
                                                                        Return the time elapsed.
                                                                        
 
                                                                        Time reported in milliseconds after the event was recorded and
 before the end_event was recorded.
                                                                        
@@ -619,7 +619,7 @@ 
                                                                        Event
 
 
                                                                        
 
                                                                        
-query()[source][source]
                                                                        
+query()[source][source]
 
                                                                        Check if all work currently captured by event has completed.
                                                                        
 
                                                                        
 
                                                                        Returns
                                                                        
@@ -634,7 +634,7 @@ 
                                                                        Event
 
 
                                                                        
 
                                                                        
-record(stream=None)[source][source]
                                                                        
+record(stream=None)[source][source]
 
                                                                        Record the event in a given stream.
                                                                        
 
                                                                        Uses torch.xpu.current_stream() if no stream is specified. The
 stream’s device must match the event’s device.
                                                                        
@@ -644,7 +644,7 @@ 
                                                                        Event
 
 
                                                                        
 
                                                                        
-synchronize()[source][source]
                                                                        
+synchronize()[source][source]
 
                                                                        Wait for the event to complete.
                                                                        
 
                                                                        Waits until the completion of all work currently captured in this event.
 This prevents the CPU thread from proceeding until the event completes.
                                                                        
@@ -654,7 +654,7 @@ 
                                                                        Event
 
 
                                                                        
 
                                                                        
-wait(stream=None)[source][source]
                                                                        
+wait(stream=None)[source][source]
 
                                                                        Make all future work submitted to the given stream wait for this event.
                                                                        
 
                                                                        Use torch.xpu.current_stream() if no stream is specified.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.Stream.html b/2.7/generated/torch.xpu.Stream.html
index 7a5d3699be6..bf8cd4082bf 100644
--- a/2.7/generated/torch.xpu.Stream.html
+++ b/2.7/generated/torch.xpu.Stream.html
@@ -596,7 +596,7 @@
 
                                                                        Stream
                                                                        
 
                                                                        
 
                                                                        
-class torch.xpu.Stream(device=None, priority=0, **kwargs)[source][source]
                                                                        
+class torch.xpu.Stream(device=None, priority=0, **kwargs)[source][source]
 
                                                                        Wrapper around a XPU stream.
                                                                        
 
                                                                        A XPU stream is a linear sequence of execution that belongs to a specific
 device, independent from other streams. It supports with statement as a
@@ -618,7 +618,7 @@ 
                                                                        Stream
 
                                                                        
 
                                                                        
-query()[source][source]
                                                                        
+query()[source][source]
 
                                                                        Check if all the work submitted has been completed.
                                                                        
 
                                                                        
 
                                                                        Returns
                                                                        
@@ -632,7 +632,7 @@ 
                                                                        Stream
 
                                                                        
-record_event(event=None)[source][source]
                                                                        
+record_event(event=None)[source][source]
 
                                                                        Record an event.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -647,7 +647,7 @@ 
                                                                        Stream
 
                                                                        
-synchronize()[source][source]
                                                                        
+synchronize()[source][source]
 
                                                                        Wait for all the kernels in this stream to complete.
                                                                        
 
                                                                        
 
                                                                        
@@ -655,7 +655,7 @@ 
                                                                        Stream
 
                                                                        
-wait_event(event)[source][source]
                                                                        
+wait_event(event)[source][source]
 
                                                                        Make all future work submitted to the stream wait for an event.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -666,7 +666,7 @@ 
                                                                        Stream
 
                                                                        
-wait_stream(stream)[source][source]
                                                                        
+wait_stream(stream)[source][source]
 
                                                                        Synchronize with another stream.
                                                                        
 
                                                                        All future work submitted to this stream will wait until all kernels
 submitted to a given stream at the time of call complete.
                                                                        
diff --git a/2.7/generated/torch.xpu.StreamContext.html b/2.7/generated/torch.xpu.StreamContext.html
index 29e054ad75a..34662608e49 100644
--- a/2.7/generated/torch.xpu.StreamContext.html
+++ b/2.7/generated/torch.xpu.StreamContext.html
@@ -596,7 +596,7 @@
 
                                                                        StreamContext
                                                                        
 
                                                                        
 
                                                                        
-class torch.xpu.StreamContext(stream)[source][source]
                                                                        
+class torch.xpu.StreamContext(stream)[source][source]
 
                                                                        Context-manager that selects a given stream.
                                                                        
 
                                                                        All XPU kernels queued within its context will be enqueued on a selected
 stream.
                                                                        
diff --git a/2.7/generated/torch.xpu.current_device.html b/2.7/generated/torch.xpu.current_device.html
index 8951517fd6b..0c7a797c9a0 100644
--- a/2.7/generated/torch.xpu.current_device.html
+++ b/2.7/generated/torch.xpu.current_device.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.current_device
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.current_device()[source][source]
                                                                        
+torch.xpu.current_device()[source][source]
 
                                                                        Return the index of a currently selected device.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.current_stream.html b/2.7/generated/torch.xpu.current_stream.html
index 08eb78fcb86..b8d8f3c06e2 100644
--- a/2.7/generated/torch.xpu.current_stream.html
+++ b/2.7/generated/torch.xpu.current_stream.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.current_stream
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.current_stream(device=None)[source][source]
                                                                        
+torch.xpu.current_stream(device=None)[source][source]
 
                                                                        Return the currently selected Stream for a given device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.device.html b/2.7/generated/torch.xpu.device.html
index 880d54a302d..375bcd62e22 100644
--- a/2.7/generated/torch.xpu.device.html
+++ b/2.7/generated/torch.xpu.device.html
@@ -596,7 +596,7 @@
 
                                                                        device
                                                                        
 
                                                                        
 
                                                                        
-class torch.xpu.device(device)[source][source]
                                                                        
+class torch.xpu.device(device)[source][source]
 
                                                                        Context-manager that changes the selected device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.device_count.html b/2.7/generated/torch.xpu.device_count.html
index dd90db93857..aa42f44c676 100644
--- a/2.7/generated/torch.xpu.device_count.html
+++ b/2.7/generated/torch.xpu.device_count.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.device_count
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.device_count()[source][source]
                                                                        
+torch.xpu.device_count()[source][source]
 
                                                                        Return the number of XPU device available.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.device_of.html b/2.7/generated/torch.xpu.device_of.html
index 94f3c83fcd8..a8867826a47 100644
--- a/2.7/generated/torch.xpu.device_of.html
+++ b/2.7/generated/torch.xpu.device_of.html
@@ -596,7 +596,7 @@
 
                                                                        device_of
                                                                        
 
                                                                        
 
                                                                        
-class torch.xpu.device_of(obj)[source][source]
                                                                        
+class torch.xpu.device_of(obj)[source][source]
 
                                                                        Context-manager that changes the current device to that of given object.
                                                                        
 
                                                                        You can use both tensors and storages as arguments. If a given object is
 not allocated on a XPU, this is a no-op.
                                                                        
diff --git a/2.7/generated/torch.xpu.empty_cache.html b/2.7/generated/torch.xpu.empty_cache.html
index 82674b06364..6aa50d3cf6d 100644
--- a/2.7/generated/torch.xpu.empty_cache.html
+++ b/2.7/generated/torch.xpu.empty_cache.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.empty_cache
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.empty_cache()[source][source]
                                                                        
+torch.xpu.empty_cache()[source][source]
 
                                                                        Release all unoccupied cached memory currently held by the caching
 allocator so that those can be used in other XPU application.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.get_arch_list.html b/2.7/generated/torch.xpu.get_arch_list.html
index f96b9980f9f..68f2e07a725 100644
--- a/2.7/generated/torch.xpu.get_arch_list.html
+++ b/2.7/generated/torch.xpu.get_arch_list.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_arch_list
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_arch_list()[source][source]
                                                                        
+torch.xpu.get_arch_list()[source][source]
 
                                                                        Return list XPU architectures this library was compiled for.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.get_device_capability.html b/2.7/generated/torch.xpu.get_device_capability.html
index a5583f3ded1..d4d800b147b 100644
--- a/2.7/generated/torch.xpu.get_device_capability.html
+++ b/2.7/generated/torch.xpu.get_device_capability.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_device_capability
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_device_capability(device=None)[source][source]
                                                                        
+torch.xpu.get_device_capability(device=None)[source][source]
 
                                                                        Get the xpu capability of a device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.get_device_name.html b/2.7/generated/torch.xpu.get_device_name.html
index 2c062874561..b323acc38c9 100644
--- a/2.7/generated/torch.xpu.get_device_name.html
+++ b/2.7/generated/torch.xpu.get_device_name.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_device_name
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_device_name(device=None)[source][source]
                                                                        
+torch.xpu.get_device_name(device=None)[source][source]
 
                                                                        Get the name of a device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.get_device_properties.html b/2.7/generated/torch.xpu.get_device_properties.html
index e8fc2992dad..746678c6f75 100644
--- a/2.7/generated/torch.xpu.get_device_properties.html
+++ b/2.7/generated/torch.xpu.get_device_properties.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_device_properties
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_device_properties(device=None)[source][source]
                                                                        
+torch.xpu.get_device_properties(device=None)[source][source]
 
                                                                        Get the properties of a device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.get_gencode_flags.html b/2.7/generated/torch.xpu.get_gencode_flags.html
index 9a608e831e7..152c5ae53c1 100644
--- a/2.7/generated/torch.xpu.get_gencode_flags.html
+++ b/2.7/generated/torch.xpu.get_gencode_flags.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_gencode_flags
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_gencode_flags()[source][source]
                                                                        
+torch.xpu.get_gencode_flags()[source][source]
 
                                                                        Return XPU AOT(ahead-of-time) build flags this library was compiled with.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.get_rng_state.html b/2.7/generated/torch.xpu.get_rng_state.html
index a7af15f9397..a9ba882541e 100644
--- a/2.7/generated/torch.xpu.get_rng_state.html
+++ b/2.7/generated/torch.xpu.get_rng_state.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_rng_state
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_rng_state(device='xpu')[source][source]
                                                                        
+torch.xpu.get_rng_state(device='xpu')[source][source]
 
                                                                        Return the random number generator state of the specified GPU as a ByteTensor.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.get_rng_state_all.html b/2.7/generated/torch.xpu.get_rng_state_all.html
index 69394f61de1..fc09335bee1 100644
--- a/2.7/generated/torch.xpu.get_rng_state_all.html
+++ b/2.7/generated/torch.xpu.get_rng_state_all.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_rng_state_all
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_rng_state_all()[source][source]
                                                                        
+torch.xpu.get_rng_state_all()[source][source]
 
                                                                        Return a list of ByteTensor representing the random number states of all devices.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.get_stream_from_external.html b/2.7/generated/torch.xpu.get_stream_from_external.html
index a1017ecfd1e..f6adba3891f 100644
--- a/2.7/generated/torch.xpu.get_stream_from_external.html
+++ b/2.7/generated/torch.xpu.get_stream_from_external.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.get_stream_from_external
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.get_stream_from_external(data_ptr, device=None)[source][source]
                                                                        
+torch.xpu.get_stream_from_external(data_ptr, device=None)[source][source]
 
                                                                        Return a Stream from an external SYCL queue.
                                                                        
 
                                                                        This function is used to wrap SYCL queue created in other libraries in order
 to facilitate data exchange and multi-library interactions.
                                                                        
diff --git a/2.7/generated/torch.xpu.init.html b/2.7/generated/torch.xpu.init.html
index dfb8b24766f..91ac29f6a9b 100644
--- a/2.7/generated/torch.xpu.init.html
+++ b/2.7/generated/torch.xpu.init.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.init
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.init()[source][source]
                                                                        
+torch.xpu.init()[source][source]
 
                                                                        Initialize PyTorch’s XPU state.
 This is a Python API about lazy initialization that avoids initializing
 XPU until the first time it is accessed. Does nothing if the XPU state is
diff --git a/2.7/generated/torch.xpu.initial_seed.html b/2.7/generated/torch.xpu.initial_seed.html
index 45bc8f50b39..5eb3029db38 100644
--- a/2.7/generated/torch.xpu.initial_seed.html
+++ b/2.7/generated/torch.xpu.initial_seed.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.initial_seed
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.initial_seed()[source][source]
                                                                        
+torch.xpu.initial_seed()[source][source]
 
                                                                        Return the current random seed of the current GPU.
                                                                        
 
                                                                        
 
                                                                        Warning
                                                                        
diff --git a/2.7/generated/torch.xpu.is_available.html b/2.7/generated/torch.xpu.is_available.html
index 81c07c1e2ae..9b1d693600c 100644
--- a/2.7/generated/torch.xpu.is_available.html
+++ b/2.7/generated/torch.xpu.is_available.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.is_available
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.is_available()[source][source]
                                                                        
+torch.xpu.is_available()[source][source]
 
                                                                        Return a bool indicating if XPU is currently available.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.is_initialized.html b/2.7/generated/torch.xpu.is_initialized.html
index 041b3f2a36d..cd065b2e06e 100644
--- a/2.7/generated/torch.xpu.is_initialized.html
+++ b/2.7/generated/torch.xpu.is_initialized.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.is_initialized
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.is_initialized()[source][source]
                                                                        
+torch.xpu.is_initialized()[source][source]
 
                                                                        Return whether PyTorch’s XPU state has been initialized.
                                                                        
 
                                                                        
 
diff --git a/2.7/generated/torch.xpu.manual_seed.html b/2.7/generated/torch.xpu.manual_seed.html
index 99ff2069faa..f23af8731f6 100644
--- a/2.7/generated/torch.xpu.manual_seed.html
+++ b/2.7/generated/torch.xpu.manual_seed.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.manual_seed
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.manual_seed(seed)[source][source]
                                                                        
+torch.xpu.manual_seed(seed)[source][source]
 
                                                                        Set the seed for generating random numbers for the current GPU.
                                                                        
 
                                                                        It’s safe to call this function if XPU is not available; in that case, it is silently ignored.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.manual_seed_all.html b/2.7/generated/torch.xpu.manual_seed_all.html
index d4cc3f361cc..ba80bacef9c 100644
--- a/2.7/generated/torch.xpu.manual_seed_all.html
+++ b/2.7/generated/torch.xpu.manual_seed_all.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.manual_seed_all
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.manual_seed_all(seed)[source][source]
                                                                        
+torch.xpu.manual_seed_all(seed)[source][source]
 
                                                                        Set the seed for generating random numbers on all GPUs.
                                                                        
 
                                                                        It’s safe to call this function if XPU is not available; in that case, it is silently ignored.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.max_memory_allocated.html b/2.7/generated/torch.xpu.max_memory_allocated.html
index 866bfdf8f44..3324268c96b 100644
--- a/2.7/generated/torch.xpu.max_memory_allocated.html
+++ b/2.7/generated/torch.xpu.max_memory_allocated.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.max_memory_allocated
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.max_memory_allocated(device=None)[source][source]
                                                                        
+torch.xpu.max_memory_allocated(device=None)[source][source]
 
                                                                        Return the maximum GPU memory occupied by tensors in bytes for a given device.
                                                                        
 
                                                                        By default, this returns the peak allocated memory since the beginning of
 this program. reset_peak_memory_stats() can be used to
diff --git a/2.7/generated/torch.xpu.max_memory_reserved.html b/2.7/generated/torch.xpu.max_memory_reserved.html
index 2496554c222..9c38970ea72 100644
--- a/2.7/generated/torch.xpu.max_memory_reserved.html
+++ b/2.7/generated/torch.xpu.max_memory_reserved.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.max_memory_reserved
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.max_memory_reserved(device=None)[source][source]
                                                                        
+torch.xpu.max_memory_reserved(device=None)[source][source]
 
                                                                        Return the maximum GPU memory managed by the caching allocator in bytes for a given device.
                                                                        
 
                                                                        By default, this returns the peak cached memory since the beginning of this
 program. reset_peak_memory_stats() can be used to reset
diff --git a/2.7/generated/torch.xpu.mem_get_info.html b/2.7/generated/torch.xpu.mem_get_info.html
index 894d65b96d9..7971ac4f8be 100644
--- a/2.7/generated/torch.xpu.mem_get_info.html
+++ b/2.7/generated/torch.xpu.mem_get_info.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.mem_get_info
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.mem_get_info(device=None)[source][source]
                                                                        
+torch.xpu.mem_get_info(device=None)[source][source]
 
                                                                        Return the global free and total GPU memory for a given device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.memory_allocated.html b/2.7/generated/torch.xpu.memory_allocated.html
index a718d409a4b..442a50ad4f5 100644
--- a/2.7/generated/torch.xpu.memory_allocated.html
+++ b/2.7/generated/torch.xpu.memory_allocated.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.memory_allocated
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.memory_allocated(device=None)[source][source]
                                                                        
+torch.xpu.memory_allocated(device=None)[source][source]
 
                                                                        Return the current GPU memory occupied by tensors in bytes for a given device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.memory_reserved.html b/2.7/generated/torch.xpu.memory_reserved.html
index 15b1a1654f0..41f57ba3cf0 100644
--- a/2.7/generated/torch.xpu.memory_reserved.html
+++ b/2.7/generated/torch.xpu.memory_reserved.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.memory_reserved
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.memory_reserved(device=None)[source][source]
                                                                        
+torch.xpu.memory_reserved(device=None)[source][source]
 
                                                                        Return the current GPU memory managed by the caching allocator in bytes for a given device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.memory_stats.html b/2.7/generated/torch.xpu.memory_stats.html
index 55238728139..63fa6ea665f 100644
--- a/2.7/generated/torch.xpu.memory_stats.html
+++ b/2.7/generated/torch.xpu.memory_stats.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.memory_stats
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.memory_stats(device=None)[source][source]
                                                                        
+torch.xpu.memory_stats(device=None)[source][source]
 
                                                                        Return a dictionary of XPU memory allocator statistics for a given device.
                                                                        
 
                                                                        The return value of this function is a dictionary of statistics, each of
 which is a non-negative integer.
                                                                        
diff --git a/2.7/generated/torch.xpu.memory_stats_as_nested_dict.html b/2.7/generated/torch.xpu.memory_stats_as_nested_dict.html
index 10c555d7fd0..dbdd2bbedc5 100644
--- a/2.7/generated/torch.xpu.memory_stats_as_nested_dict.html
+++ b/2.7/generated/torch.xpu.memory_stats_as_nested_dict.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.memory_stats_as_nested_dict
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.memory_stats_as_nested_dict(device=None)[source][source]
                                                                        
+torch.xpu.memory_stats_as_nested_dict(device=None)[source][source]
 
                                                                        Return the result of memory_stats() as a nested dictionary.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
diff --git a/2.7/generated/torch.xpu.reset_accumulated_memory_stats.html b/2.7/generated/torch.xpu.reset_accumulated_memory_stats.html
index 7ef83ff777d..4c736a99392 100644
--- a/2.7/generated/torch.xpu.reset_accumulated_memory_stats.html
+++ b/2.7/generated/torch.xpu.reset_accumulated_memory_stats.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.reset_accumulated_memory_stats
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.reset_accumulated_memory_stats(device=None)[source][source]
                                                                        
+torch.xpu.reset_accumulated_memory_stats(device=None)[source][source]
 
                                                                        Reset the “accumulated” (historical) stats tracked by the XPU memory allocator.
                                                                        
 
                                                                        See memory_stats() for details. Accumulated stats correspond to
 the “allocated” and “freed” keys in each individual stat dict.
                                                                        
diff --git a/2.7/generated/torch.xpu.reset_peak_memory_stats.html b/2.7/generated/torch.xpu.reset_peak_memory_stats.html
index a8ce0ae819c..f24a0008efd 100644
--- a/2.7/generated/torch.xpu.reset_peak_memory_stats.html
+++ b/2.7/generated/torch.xpu.reset_peak_memory_stats.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.reset_peak_memory_stats
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.reset_peak_memory_stats(device=None)[source][source]
                                                                        
+torch.xpu.reset_peak_memory_stats(device=None)[source][source]
 
                                                                        Reset the “peak” stats tracked by the XPU memory allocator.
                                                                        
 
                                                                        See memory_stats() for details. Peak stats correspond to the
 “peak” key in each individual stat dict.
                                                                        
diff --git a/2.7/generated/torch.xpu.seed.html b/2.7/generated/torch.xpu.seed.html
index f283ff9e22c..bc6a96fd562 100644
--- a/2.7/generated/torch.xpu.seed.html
+++ b/2.7/generated/torch.xpu.seed.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.seed
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.seed()[source][source]
                                                                        
+torch.xpu.seed()[source][source]
 
                                                                        Set the seed for generating random numbers to a random number for the current GPU.
                                                                        
 
                                                                        It’s safe to call this function if XPU is not available; in that case, it is silently ignored.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.seed_all.html b/2.7/generated/torch.xpu.seed_all.html
index fcd16760eab..fe4beb31ef5 100644
--- a/2.7/generated/torch.xpu.seed_all.html
+++ b/2.7/generated/torch.xpu.seed_all.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.seed_all
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.seed_all()[source][source]
                                                                        
+torch.xpu.seed_all()[source][source]
 
                                                                        Set the seed for generating random numbers to a random number on all GPUs.
                                                                        
 
                                                                        It’s safe to call this function if XPU is not available; in that case, it is silently ignored.
                                                                        
 
                                                                        
diff --git a/2.7/generated/torch.xpu.set_device.html b/2.7/generated/torch.xpu.set_device.html
index 41c727b7ccd..704c28e59c1 100644
--- a/2.7/generated/torch.xpu.set_device.html
+++ b/2.7/generated/torch.xpu.set_device.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.set_device
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.set_device(device)[source][source]
                                                                        
+torch.xpu.set_device(device)[source][source]
 
                                                                        Set the current device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.set_rng_state.html b/2.7/generated/torch.xpu.set_rng_state.html
index 0698c46da9d..9c8002311a4 100644
--- a/2.7/generated/torch.xpu.set_rng_state.html
+++ b/2.7/generated/torch.xpu.set_rng_state.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.set_rng_state
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.set_rng_state(new_state, device='xpu')[source][source]
                                                                        
+torch.xpu.set_rng_state(new_state, device='xpu')[source][source]
 
                                                                        Set the random number generator state of the specified GPU.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.set_rng_state_all.html b/2.7/generated/torch.xpu.set_rng_state_all.html
index a5c3dc7d67f..cd6daf285c2 100644
--- a/2.7/generated/torch.xpu.set_rng_state_all.html
+++ b/2.7/generated/torch.xpu.set_rng_state_all.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.set_rng_state_all
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.set_rng_state_all(new_states)[source][source]
                                                                        
+torch.xpu.set_rng_state_all(new_states)[source][source]
 
                                                                        Set the random number generator state of all devices.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.set_stream.html b/2.7/generated/torch.xpu.set_stream.html
index 045808ba59d..00422e32639 100644
--- a/2.7/generated/torch.xpu.set_stream.html
+++ b/2.7/generated/torch.xpu.set_stream.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.set_stream
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.set_stream(stream)[source][source]
                                                                        
+torch.xpu.set_stream(stream)[source][source]
 
                                                                        
 
                                                                        Set the current stream.This is a wrapper API to set the stream.
                                                                        Usage of this function is discouraged in favor of the stream
 context manager.
                                                                        
diff --git a/2.7/generated/torch.xpu.stream.html b/2.7/generated/torch.xpu.stream.html
index 6ddb1364a95..588e3e3ada1 100644
--- a/2.7/generated/torch.xpu.stream.html
+++ b/2.7/generated/torch.xpu.stream.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.stream
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.stream(stream)[source][source]
                                                                        
+torch.xpu.stream(stream)[source][source]
 
                                                                        Wrap around the Context-manager StreamContext that selects a given stream.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/generated/torch.xpu.synchronize.html b/2.7/generated/torch.xpu.synchronize.html
index 33efa0376ec..968fa2b36e3 100644
--- a/2.7/generated/torch.xpu.synchronize.html
+++ b/2.7/generated/torch.xpu.synchronize.html
@@ -596,7 +596,7 @@
 
                                                                        torch.xpu.synchronize
                                                                        
 
                                                                        
 
                                                                        
-torch.xpu.synchronize(device=None)[source][source]
                                                                        
+torch.xpu.synchronize(device=None)[source][source]
 
                                                                        Wait for all kernels in all streams on a XPU device to complete.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/hub.html b/2.7/hub.html
index 75a9281128b..9abc9cb3e5f 100644
--- a/2.7/hub.html
+++ b/2.7/hub.html
@@ -667,7 +667,7 @@ 
                                                                        Loading models from Hubtorch.hub.load().
                                                                        
 
                                                                        
 
                                                                        
-torch.hub.list(github, force_reload=False, skip_validation=False, trust_repo=None, verbose=True)[source][source]
                                                                        
+torch.hub.list(github, force_reload=False, skip_validation=False, trust_repo=None, verbose=True)[source][source]
 
                                                                        List all callable entrypoints available in the repo specified by github.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -720,7 +720,7 @@ 
                                                                        Loading models from Hub
 
                                                                        
-torch.hub.help(github, model, force_reload=False, skip_validation=False, trust_repo=None)[source][source]
                                                                        
+torch.hub.help(github, model, force_reload=False, skip_validation=False, trust_repo=None)[source][source]
 
                                                                        Show the docstring of entrypoint model.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -765,7 +765,7 @@ 
                                                                        Loading models from Hub
 
                                                                        
-torch.hub.load(repo_or_dir, model, *args, source='github', trust_repo=None, force_reload=False, verbose=True, skip_validation=False, **kwargs)[source][source]
                                                                        
+torch.hub.load(repo_or_dir, model, *args, source='github', trust_repo=None, force_reload=False, verbose=True, skip_validation=False, **kwargs)[source][source]
 
                                                                        Load a model from a github repo or a local directory.
                                                                        
 
                                                                        Note: Loading a model is the typical use case, but this can also be used to
 for loading other objects such as tokenizers, loss functions, etc.
                                                                        
@@ -839,7 +839,7 @@ 
                                                                        Loading models from Hub
 
                                                                        
-torch.hub.download_url_to_file(url, dst, hash_prefix=None, progress=True)[source][source]
                                                                        
+torch.hub.download_url_to_file(url, dst, hash_prefix=None, progress=True)[source][source]
 
                                                                        Download object at the given URL to a local path.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -864,7 +864,7 @@ 
                                                                        Loading models from Hub
 
                                                                        
-torch.hub.load_state_dict_from_url(url, model_dir=None, map_location=None, progress=True, check_hash=False, file_name=None, weights_only=False)[source][source]
                                                                        
+torch.hub.load_state_dict_from_url(url, model_dir=None, map_location=None, progress=True, check_hash=False, file_name=None, weights_only=False)[source][source]
 
                                                                        Loads the Torch serialized object at the given URL.
                                                                        
 
                                                                        If downloaded file is a zip file, it will be automatically
 decompressed.
                                                                        
@@ -927,7 +927,7 @@ 
                                                                        Where are my downloaded models saved?
 
                                                                        
-torch.hub.get_dir()[source][source]
                                                                        
+torch.hub.get_dir()[source][source]
 
                                                                        Get the Torch Hub cache directory used for storing downloaded models & weights.
                                                                        
 
                                                                        If set_dir() is not called, default path is $TORCH_HOME/hub where
 environment variable $TORCH_HOME defaults to $XDG_CACHE_HOME/torch.
@@ -943,7 +943,7 @@ 
                                                                        Where are my downloaded models saved?
 
                                                                        
-torch.hub.set_dir(d)[source][source]
                                                                        
+torch.hub.set_dir(d)[source][source]
 
                                                                        Optionally set the Torch Hub directory used to save downloaded models & weights.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
diff --git a/2.7/jit.html b/2.7/jit.html
index a2d8a96cab2..b5df1651d56 100644
--- a/2.7/jit.html
+++ b/2.7/jit.html
@@ -1255,7 +1255,7 @@ 
                                                                        Modules@torch.jit.export methods.
                                                                        
 
                                                                        
 
                                                                        
-torch.jit.export(fn)[source][source]
                                                                        
+torch.jit.export(fn)[source][source]
 
                                                                        This decorator indicates that a method on an nn.Module is used as an entry point into a
 ScriptModule and should be compiled.
                                                                        
 
                                                                        forward implicitly is assumed to be an entry point, so it does not need this decorator.
diff --git a/2.7/jit_language_reference.html b/2.7/jit_language_reference.html
index d67e68432c4..71683acd468 100644
--- a/2.7/jit_language_reference.html
+++ b/2.7/jit_language_reference.html
@@ -1347,7 +1347,7 @@ 
                                                                        Functions
 
                                                                        
-torch.jit.is_scripting()[source][source]
                                                                        
+torch.jit.is_scripting()[source][source]
 
                                                                        Function that returns True when in compilation and False otherwise. This
 is useful especially with the @unused decorator to leave code in your
 model that is not yet TorchScript compatible.
@@ -1374,7 +1374,7 @@ 
                                                                        Functions
 
                                                                        
-torch.jit.is_tracing()[source][source]
                                                                        
+torch.jit.is_tracing()[source][source]
 
                                                                        Return a boolean value.
                                                                        
 
                                                                        Returns True in tracing (if a function is called during the
 tracing of code with torch.jit.trace) and False otherwise.
                                                                        
diff --git a/2.7/library.html b/2.7/library.html
index 4938c362dbc..a53f96e4bc6 100644
--- a/2.7/library.html
+++ b/2.7/library.html
@@ -607,7 +607,7 @@ 
                                                                        Testing custom ops
 
                                                                        
-torch.library.opcheck(op, args, kwargs=None, *, test_utils=('test_schema', 'test_autograd_registration', 'test_faketensor', 'test_aot_dispatch_dynamic'), raise_exception=True, atol=None, rtol=None)[source][source]
                                                                        
+torch.library.opcheck(op, args, kwargs=None, *, test_utils=('test_schema', 'test_autograd_registration', 'test_faketensor', 'test_aot_dispatch_dynamic'), raise_exception=True, atol=None, rtol=None)[source][source]
 
                                                                        Given an operator and some sample arguments, tests if the operator is
 registered correctly.
                                                                        
 
                                                                        That is, when you use the torch.library/TORCH_LIBRARY APIs to create a
@@ -725,7 +725,7 @@ 
                                                                        Creating new custom ops in Pythontorch.library.custom_op() to create new custom ops.
                                                                        
 
                                                                        
 
                                                                        
-torch.library.custom_op(name, fn=None, /, *, mutates_args, device_types=None, schema=None)[source]
                                                                        
+torch.library.custom_op(name, fn=None, /, *, mutates_args, device_types=None, schema=None)[source]
 
                                                                        Wraps a function into custom operator.
                                                                        
 
                                                                        Reasons why you may want to create a custom op include:
 - Wrapping a third-party library or custom kernel to work with PyTorch
@@ -823,7 +823,7 @@ 
                                                                        Creating new custom ops in Python
 
                                                                        
-torch.library.triton_op(name, fn=None, /, *, mutates_args, schema=None)[source]
                                                                        
+torch.library.triton_op(name, fn=None, /, *, mutates_args, schema=None)[source]
 
                                                                        Create a custom operator whose implementation is backed by 1+ triton kernels.
                                                                        
 
                                                                        This is a more structured way of using triton kernels with PyTorch.
 Prefer using triton kernels with no torch.library custom operator wrappers
@@ -917,7 +917,7 @@ 
                                                                        Creating new custom ops in Python
 
                                                                        
-torch.library.wrap_triton(triton_kernel, /)[source]
                                                                        
+torch.library.wrap_triton(triton_kernel, /)[source]
 
                                                                        Allows capture of a triton kernel into a graph via make_fx or
 non-strict torch.export.
                                                                        
 
                                                                        These technologies perform Dispatcher-based tracing (via
@@ -990,7 +990,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
 
                                                                        
-torch.library.register_kernel(op, device_types, func=None, /, *, lib=None)[source][source]
                                                                        
+torch.library.register_kernel(op, device_types, func=None, /, *, lib=None)[source][source]
 
                                                                        Register an implementation for a device type for this operator.
                                                                        
 
                                                                        Some valid device_types are: “cpu”, “cuda”, “xla”, “mps”, “ipu”, “xpu”.
 This API may be used as a decorator.
                                                                        
@@ -1039,7 +1039,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.register_autocast(op, device_type, cast_inputs, /, *, lib=None)[source][source]
                                                                        
+torch.library.register_autocast(op, device_type, cast_inputs, /, *, lib=None)[source][source]
 
                                                                        Register an autocast dispatch rule for this custom op.
                                                                        
 
                                                                        Valid device_type include: “cpu” and “cuda”.
                                                                        
 
                                                                        
@@ -1081,7 +1081,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.register_autograd(op, backward, /, *, setup_context=None, lib=None)[source][source]
                                                                        
+torch.library.register_autograd(op, backward, /, *, setup_context=None, lib=None)[source][source]
 
                                                                        Register a backward formula for this custom op.
                                                                        
 
                                                                        In order for an operator to work with autograd, you need to register
 a backward formula:
@@ -1165,7 +1165,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.register_fake(op, func=None, /, *, lib=None, _stacklevel=1)[source][source]
                                                                        
+torch.library.register_fake(op, func=None, /, *, lib=None, _stacklevel=1)[source][source]
 
                                                                        Register a FakeTensor implementation (“fake impl”) for this operator.
                                                                        
 
                                                                        Also sometimes known as a “meta kernel”, “abstract impl”.
                                                                        
 
                                                                        An “FakeTensor implementation” specifies the behavior of this operator on
@@ -1246,7 +1246,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.register_vmap(op, func=None, /, *, lib=None)[source][source]
                                                                        
+torch.library.register_vmap(op, func=None, /, *, lib=None)[source][source]
 
                                                                        Register a vmap implementation to support torch.vmap() for this custom op.
                                                                        
 
                                                                        This API may be used as a decorator (see examples).
                                                                        
 
                                                                        In order for an operator to work with torch.vmap(), you may need to register a
@@ -1324,14 +1324,14 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.impl_abstract(qualname, func=None, *, lib=None, _stacklevel=1)[source][source]
                                                                        
+torch.library.impl_abstract(qualname, func=None, *, lib=None, _stacklevel=1)[source][source]
 
                                                                        This API was renamed to torch.library.register_fake() in PyTorch 2.4.
 Please use that instead.
                                                                        
 
                                                                        
 
 
                                                                        
 
                                                                        
-torch.library.get_ctx()[source][source]
                                                                        
+torch.library.get_ctx()[source][source]
 
                                                                        get_ctx() returns the current AbstractImplCtx object.
                                                                        
 
                                                                        Calling get_ctx() is only valid inside of an fake impl
 (see torch.library.register_fake() for more usage details.
                                                                        
@@ -1344,7 +1344,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.register_torch_dispatch(op, torch_dispatch_class, func=None, /, *, lib=None)[source][source]
                                                                        
+torch.library.register_torch_dispatch(op, torch_dispatch_class, func=None, /, *, lib=None)[source][source]
 
                                                                        Registers a torch_dispatch rule for the given operator and torch_dispatch_class.
                                                                        
 
                                                                        This allows for open registration to specify the behavior between the operator
 and the torch_dispatch_class without needing to modify the torch_dispatch_class
@@ -1388,7 +1388,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-torch.library.infer_schema(prototype_function, /, *, mutates_args, op_name=None)[source]
                                                                        
+torch.library.infer_schema(prototype_function, /, *, mutates_args, op_name=None)[source]
 
                                                                        Parses the schema of a given function with type hints. The schema is inferred from the
 function’s type hints, and can be used to define a new operator.
                                                                        
 
                                                                        We make the following assumptions:
                                                                        
@@ -1439,7 +1439,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
-class torch._library.custom_ops.CustomOpDef(namespace, name, schema, fn)[source][source]
                                                                        
+class torch._library.custom_ops.CustomOpDef(namespace, name, schema, fn)[source][source]
 
                                                                        CustomOpDef is a wrapper around a function that turns it into a custom op.
                                                                        
 
                                                                        It has various methods for registering additional behavior for this
 custom op.
                                                                        
@@ -1449,7 +1449,7 @@ 
                                                                        Extending custom ops (created from Python or C++)
 
                                                                        
 
                                                                        
-set_kernel_enabled(device_type, enabled=True)[source][source]
                                                                        
+set_kernel_enabled(device_type, enabled=True)[source][source]
 
                                                                        Disable or re-enable an already registered kernel for this custom operator.
                                                                        
 
                                                                        If the kernel is already disabled/enabled, this is a no-op.
                                                                        
 
                                                                        
@@ -1505,7 +1505,7 @@ 
                                                                        Low-level APIsGoogle Colab.
                                                                        
 
                                                                        
 
                                                                        
-class torch.library.Library(ns, kind, dispatch_key='')[source][source]
                                                                        
+class torch.library.Library(ns, kind, dispatch_key='')[source][source]
 
                                                                        A class to create libraries that can be used to register new operators or
 override operators in existing libraries from Python.
 A user can optionally pass in a dispatch keyname if they only want to register
@@ -1526,7 +1526,7 @@ 
                                                                        Low-level APIs
 
                                                                        
-define(schema, alias_analysis='', *, tags=())[source][source]
                                                                        
+define(schema, alias_analysis='', *, tags=())[source][source]
 
                                                                        Defines a new operator and its semantics in the ns namespace.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -1555,7 +1555,7 @@ 
                                                                        Low-level APIs
 
                                                                        
-fallback(fn, dispatch_key='', *, with_keyset=False)[source][source]
                                                                        
+fallback(fn, dispatch_key='', *, with_keyset=False)[source][source]
 
                                                                        Registers the function implementation as the fallback for the given key.
                                                                        
 
                                                                        This function only works for a library with global namespace (“_”).
                                                                        
 
                                                                        
@@ -1584,7 +1584,7 @@ 
                                                                        Low-level APIs
 
                                                                        
-impl(op_name, fn, dispatch_key='', *, with_keyset=False)[source][source]
                                                                        
+impl(op_name, fn, dispatch_key='', *, with_keyset=False)[source][source]
 
                                                                        Registers the function implementation for an operator defined in the library.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -1614,13 +1614,13 @@ 
                                                                        Low-level APIs
 
                                                                        
-torch.library.fallthrough_kernel()[source][source]
                                                                        
+torch.library.fallthrough_kernel()[source][source]
 
                                                                        A dummy function to pass to Library.impl in order to register a fallthrough.
                                                                        
 
                                                                        
 
 
                                                                        
 
                                                                        
-torch.library.define(qualname, schema, *, lib=None, tags=())[source][source]
                                                                        
+torch.library.define(qualname, schema, *, lib=None, tags=())[source][source]
 
                                                                        
 torch.library.define(lib, schema, alias_analysis='')
                                                                        
 
                                                                        Defines a new operator.
                                                                        
@@ -1677,7 +1677,7 @@ 
                                                                        Low-level APIs
 
                                                                        
-torch.library.impl(lib, name, dispatch_key='')[source][source]
                                                                        
+torch.library.impl(lib, name, dispatch_key='')[source][source]
 
                                                                        
 torch.library.impl(qualname: str, types: Union[str, Sequence[str]], func: Literal[None] = None, *, lib: Optional[Library] = None) → Callable[[Callable[..., object]], None]
                                                                        
 
                                                                        
diff --git a/2.7/model_zoo.html b/2.7/model_zoo.html
index f1d7b2cb82c..de1fc19a129 100644
--- a/2.7/model_zoo.html
+++ b/2.7/model_zoo.html
@@ -595,7 +595,7 @@ 
                                                                        torch.utils.model_zooMoved to torch.hub.
                                                                        
 
                                                                        
 
                                                                        
-torch.utils.model_zoo.load_url(url, model_dir=None, map_location=None, progress=True, check_hash=False, file_name=None, weights_only=False)[source]
                                                                        
+torch.utils.model_zoo.load_url(url, model_dir=None, map_location=None, progress=True, check_hash=False, file_name=None, weights_only=False)[source]
                                                                        
 
                                                                        Loads the Torch serialized object at the given URL.
                                                                        
 
                                                                        If downloaded file is a zip file, it will be automatically
 decompressed.
                                                                        
diff --git a/2.7/module_tracker.html b/2.7/module_tracker.html
index 88c5f0d8e37..00ffb971d3a 100644
--- a/2.7/module_tracker.html
+++ b/2.7/module_tracker.html
@@ -596,7 +596,7 @@
 It can be used within other tracking tools to be able to easily associate measured quantities to user-friendly names. This is used in particular in the FlopCounterMode today.
                                                                        
 
                                                                        
 
                                                                        
-class torch.utils.module_tracker.ModuleTracker[source][source]
                                                                        
+class torch.utils.module_tracker.ModuleTracker[source][source]
 
                                                                        ModuleTracker is a context manager that tracks the nn.Module hierarchy during execution
 so that other system can query which Module is currently being executed (or its backward is being
 executed).
                                                                        
diff --git a/2.7/monitor.html b/2.7/monitor.html
index f3cd460748a..70cf4d0c794 100644
--- a/2.7/monitor.html
+++ b/2.7/monitor.html
@@ -769,7 +769,7 @@ 
                                                                        torch.monitor
 
                                                                        
-class torch.monitor.TensorboardEventHandler(writer)[source][source]
                                                                        
+class torch.monitor.TensorboardEventHandler(writer)[source][source]
 
                                                                        TensorboardEventHandler is an event handler that will write known events to
 the provided SummaryWriter.
                                                                        
 
                                                                        This currently only supports torch.monitor.Stat events which are logged
@@ -785,7 +785,7 @@ 
                                                                        torch.monitor
 
                                                                        
-__init__(writer)[source][source]
                                                                        
+__init__(writer)[source][source]
 
                                                                        Constructs the TensorboardEventHandler.
                                                                        
 
                                                                        
 
                                                                        
diff --git a/2.7/multiprocessing.html b/2.7/multiprocessing.html
index 27e36c8ef96..be78bf2e60d 100644
--- a/2.7/multiprocessing.html
+++ b/2.7/multiprocessing.html
@@ -613,19 +613,19 @@
 
                                                                        Strategy management
                                                                        
 
                                                                        
 
                                                                        
-torch.multiprocessing.get_all_sharing_strategies()[source][source]
                                                                        
+torch.multiprocessing.get_all_sharing_strategies()[source][source]
 
                                                                        Return a set of sharing strategies supported on a current system.
                                                                        
 
                                                                        
 
 
                                                                        
 
                                                                        
-torch.multiprocessing.get_sharing_strategy()[source][source]
                                                                        
+torch.multiprocessing.get_sharing_strategy()[source][source]
 
                                                                        Return the current strategy for sharing CPU tensors.
                                                                        
 
                                                                        
 
 
                                                                        
 
                                                                        
-torch.multiprocessing.set_sharing_strategy(new_strategy)[source][source]
                                                                        
+torch.multiprocessing.set_sharing_strategy(new_strategy)[source][source]
 
                                                                        Set the strategy for sharing CPU tensors.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -762,7 +762,7 @@ 
                                                                        Spawning subprocesses
 
                                                                        
 
                                                                        
-torch.multiprocessing.spawn.spawn(fn, args=(), nprocs=1, join=True, daemon=False, start_method='spawn')[source][source]
                                                                        
+torch.multiprocessing.spawn.spawn(fn, args=(), nprocs=1, join=True, daemon=False, start_method='spawn')[source][source]
 
                                                                        Spawns nprocs processes that run fn with args.
                                                                        
 
                                                                        If one of the processes exits with a non-zero exit status, the
 remaining processes are killed and an exception is raised with the
@@ -799,11 +799,11 @@ 
                                                                        Spawning subprocesses
 
                                                                        
-class torch.multiprocessing.SpawnContext[source][source]
                                                                        
+class torch.multiprocessing.SpawnContext[source][source]
 
                                                                        Returned by spawn() when called with join=False.
                                                                        
 
                                                                        
 
                                                                        
-join(timeout=None, grace_period=None)[source]
                                                                        
+join(timeout=None, grace_period=None)[source]
 
                                                                        Join one or more processes within spawn context.
                                                                        
 
                                                                        Attempt to join one or more processes in this spawn context.
 If one of them exited with a non-zero exit status, this function
diff --git a/2.7/named_tensor.html b/2.7/named_tensor.html
index d5b5f7b2de6..89c49eb3f15 100644
--- a/2.7/named_tensor.html
+++ b/2.7/named_tensor.html
@@ -859,7 +859,7 @@ 
                                                                        Named tensor API reference
 
                                                                        
-rename(*names, **rename_map)[source][source]
                                                                        
+rename(*names, **rename_map)[source][source]
 
                                                                        Renames dimension names of self.
                                                                        
 
                                                                        There are two main usages:
                                                                        
 
                                                                        self.rename(**rename_map) returns a view on tensor that has dims
@@ -892,13 +892,13 @@ 
                                                                        Named tensor API reference
 
                                                                        
-rename_(*names, **rename_map)[source][source]
                                                                        
+rename_(*names, **rename_map)[source][source]
 
                                                                        In-place version of rename().
                                                                        
 
                                                                        
 
 
                                                                        
 
                                                                        
-refine_names(*names)[source][source]
                                                                        
+refine_names(*names)[source][source]
 
                                                                        Refines the dimension names of self according to names.
                                                                        
 
                                                                        Refining is a special case of renaming that “lifts” unnamed dimensions.
 A None dim can be refined to have any name; a named dim can only be
@@ -980,7 +980,7 @@ 
                                                                        Named tensor API reference
 
                                                                        
-align_to(*names)[source][source]
                                                                        
+align_to(*names)[source][source]
 
                                                                        Permutes the dimensions of the self tensor to match the order
 specified in names, adding size-one dims for any new names.
                                                                        
 
                                                                        All of the dims of self must be named in order to use this method.
diff --git a/2.7/nested.html b/2.7/nested.html
index 7dbcd46ff30..2fcbba59fe6 100644
--- a/2.7/nested.html
+++ b/2.7/nested.html
@@ -1029,7 +1029,7 @@ 
                                                                        Data dependent operation within torch.compile
                                                                        Detailed Docs for Construction and Conversion Functions
                                                                        
 
                                                                        
 
                                                                        
-torch.nested.nested_tensor(tensor_list, *, dtype=None, layout=None, device=None, requires_grad=False, pin_memory=False)[source][source]
                                                                        
+torch.nested.nested_tensor(tensor_list, *, dtype=None, layout=None, device=None, requires_grad=False, pin_memory=False)[source][source]
 
                                                                        Constructs a nested tensor with no autograd history (also known as a “leaf tensor”, see
 Autograd mechanics) from tensor_list a list of tensors.
                                                                        
 
                                                                        
@@ -1069,7 +1069,7 @@ 
                                                                        Data dependent operation within torch.compile
 
                                                                        
-torch.nested.nested_tensor_from_jagged(values, offsets=None, lengths=None, jagged_dim=None, min_seqlen=None, max_seqlen=None)[source][source]
                                                                        
+torch.nested.nested_tensor_from_jagged(values, offsets=None, lengths=None, jagged_dim=None, min_seqlen=None, max_seqlen=None)[source][source]
 
                                                                        Constructs a jagged layout nested tensor from the given jagged components. The jagged layout
 consists of a required values buffer with the jagged dimension packed into a single dimension.
 The offsets / lengths metadata determines how this dimension is split into batch elements
@@ -1145,7 +1145,7 @@ 
                                                                        Data dependent operation within torch.compile
 
                                                                        
-torch.nested.as_nested_tensor(ts, dtype=None, device=None, layout=None)[source][source]
                                                                        
+torch.nested.as_nested_tensor(ts, dtype=None, device=None, layout=None)[source][source]
 
                                                                        Constructs a nested tensor preserving autograd history from a tensor or a list / tuple of
 tensors.
                                                                        
 
                                                                        If a nested tensor is passed, it will be returned directly unless the device / dtype / layout
@@ -1250,7 +1250,7 @@ 
                                                                        Data dependent operation within torch.compile
 
                                                                        
-torch.nested.masked_select(tensor, mask)[source][source]
                                                                        
+torch.nested.masked_select(tensor, mask)[source][source]
 
                                                                        Constructs a nested tensor given a strided tensor input and a strided mask, the resulting jagged layout nested tensor
 will have values retain values where the mask is equal to True. The dimensionality of the mask is preserved and is
 represented with the offsets, this is unlike masked_select() where the output is collapsed to a 1D tensor.
                                                                        
@@ -1286,7 +1286,7 @@ 
                                                                        Data dependent operation within torch.compile
 
                                                                        
-torch.nested.narrow(tensor, dim, start, length, layout=torch.strided)[source][source]
                                                                        
+torch.nested.narrow(tensor, dim, start, length, layout=torch.strided)[source][source]
 
                                                                        Constructs a nested tensor (which might be a view) from tensor, a strided tensor. This follows
 similar semantics to torch.Tensor.narrow, where in the dim-th dimension the new nested tensor
 shows only the elements in the interval [start, start+length). As nested representations
diff --git a/2.7/nn.attention.flex_attention.html b/2.7/nn.attention.flex_attention.html
index 6c4741cedff..fc57d25a220 100644
--- a/2.7/nn.attention.flex_attention.html
+++ b/2.7/nn.attention.flex_attention.html
@@ -596,7 +596,7 @@
 
                                                                        torch.nn.attention.flex_attention
                                                                        
 
                                                                        
 
                                                                        
-torch.nn.attention.flex_attention.flex_attention(query, key, value, score_mod=None, block_mask=None, scale=None, enable_gqa=False, return_lse=False, kernel_options=None)[source][source]
                                                                        
+torch.nn.attention.flex_attention.flex_attention(query, key, value, score_mod=None, block_mask=None, scale=None, enable_gqa=False, return_lse=False, kernel_options=None)[source][source]
 
                                                                        This function implements scaled dot product attention with an arbitrary attention score modification function.
                                                                        
 
                                                                        This function computes the scaled dot product attention between query, key, and value tensors with a user-defined
 attention score modification function. The attention score modification function will be applied after the attention
@@ -675,7 +675,7 @@
 
                                                                        BlockMask Utilities
                                                                        
 
                                                                        
 
                                                                        
-torch.nn.attention.flex_attention.create_block_mask(mask_mod, B, H, Q_LEN, KV_LEN, device='cuda', BLOCK_SIZE=128, _compile=False)[source][source]
                                                                        
+torch.nn.attention.flex_attention.create_block_mask(mask_mod, B, H, Q_LEN, KV_LEN, device='cuda', BLOCK_SIZE=128, _compile=False)[source][source]
 
                                                                        This function creates a block mask tuple from a mask_mod function.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -717,7 +717,7 @@ 
                                                                        BlockMask Utilities
 
                                                                        
-torch.nn.attention.flex_attention.create_mask(mod_fn, B, H, Q_LEN, KV_LEN, device='cuda')[source][source]
                                                                        
+torch.nn.attention.flex_attention.create_mask(mod_fn, B, H, Q_LEN, KV_LEN, device='cuda')[source][source]
 
                                                                        This function creates a mask tensor from a mod_fn function.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -741,7 +741,7 @@ 
                                                                        BlockMask Utilities
 
                                                                        
-torch.nn.attention.flex_attention.create_nested_block_mask(mask_mod, B, H, q_nt, kv_nt=None, BLOCK_SIZE=128, _compile=False)[source][source]
                                                                        
+torch.nn.attention.flex_attention.create_nested_block_mask(mask_mod, B, H, q_nt, kv_nt=None, BLOCK_SIZE=128, _compile=False)[source][source]
 
                                                                        This function creates a nested tensor compatible block mask tuple from a mask_mod
 function. The returned BlockMask will be on the device specified by the input nested tensor.
                                                                        
 
                                                                        
@@ -805,7 +805,7 @@ 
                                                                        BlockMask Utilities
 
                                                                        
-torch.nn.attention.flex_attention.and_masks(*mask_mods)[source][source]
                                                                        
+torch.nn.attention.flex_attention.and_masks(*mask_mods)[source][source]
 
                                                                        Returns a mask_mod that’s the intersection of provided mask_mods
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
@@ -816,7 +816,7 @@ 
                                                                        BlockMask Utilities
 
                                                                        
-torch.nn.attention.flex_attention.or_masks(*mask_mods)[source][source]
                                                                        
+torch.nn.attention.flex_attention.or_masks(*mask_mods)[source][source]
 
                                                                        Returns a mask_mod that’s the union of provided mask_mods
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
@@ -827,7 +827,7 @@ 
                                                                        BlockMask Utilities
 
                                                                        
-torch.nn.attention.flex_attention.noop_mask(batch, head, token_q, token_kv)[source][source]
                                                                        
+torch.nn.attention.flex_attention.noop_mask(batch, head, token_q, token_kv)[source][source]
 
                                                                        Returns a noop mask_mod
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
@@ -841,7 +841,7 @@ 
                                                                        BlockMask Utilities
                                                                        
 
                                                                        
 
                                                                        
-class torch.nn.attention.flex_attention.BlockMask(seq_lengths, kv_num_blocks, kv_indices, full_kv_num_blocks, full_kv_indices, q_num_blocks, q_indices, full_q_num_blocks, full_q_indices, BLOCK_SIZE, mask_mod)[source][source]
                                                                        
+class torch.nn.attention.flex_attention.BlockMask(seq_lengths, kv_num_blocks, kv_indices, full_kv_num_blocks, full_kv_indices, q_num_blocks, q_indices, full_q_num_blocks, full_q_indices, BLOCK_SIZE, mask_mod)[source][source]
 
                                                                        BlockMask is our format for representing a block-sparse attention mask.
 It is somewhat of a cross in-between BCSR and a non-sparse format.
                                                                        
 
                                                                        
@@ -894,7 +894,7 @@ 
                                                                        Details
 
                                                                        
-as_tuple(flatten=True)[source][source]
                                                                        
+as_tuple(flatten=True)[source][source]
 
                                                                        Returns a tuple of the attributes of the BlockMask.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -905,7 +905,7 @@ 
                                                                        Details
 
                                                                        
-classmethod from_kv_blocks(kv_num_blocks, kv_indices, full_kv_num_blocks=None, full_kv_indices=None, BLOCK_SIZE=128, mask_mod=None, seq_lengths=None)[source][source]
                                                                        
+classmethod from_kv_blocks(kv_num_blocks, kv_indices, full_kv_num_blocks=None, full_kv_indices=None, BLOCK_SIZE=128, mask_mod=None, seq_lengths=None)[source][source]
 
                                                                        Creates a BlockMask instance from key-value block information.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -970,7 +970,7 @@ 
                                                                        Details
 
                                                                        
-numel()[source][source]
                                                                        
+numel()[source][source]
 
                                                                        Returns the number of elements (not accounting for sparsity) in the mask.
                                                                        
 
                                                                        
 
@@ -996,7 +996,7 @@ 
                                                                        Details
 
                                                                        
-sparsity()[source][source]
                                                                        
+sparsity()[source][source]
 
                                                                        Computes the percentage of blocks that are sparse (i.e. not computed)
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
@@ -1007,7 +1007,7 @@ 
                                                                        Details
 
                                                                        
-to(device)[source][source]
                                                                        
+to(device)[source][source]
 
                                                                        Moves the BlockMask to the specified device.
                                                                        
 
                                                                        
 
                                                                        Parameters
                                                                        
@@ -1033,7 +1033,7 @@ 
                                                                        Details
 
                                                                        
-to_dense()[source][source]
                                                                        
+to_dense()[source][source]
 
                                                                        Returns a dense block that is equivalent to the block mask.
                                                                        
 
                                                                        
 
                                                                        Return type
                                                                        
@@ -1044,7 +1044,7 @@ 
                                                                        Details
 
                                                                        
-to_string(grid_size=(20, 20), limit=4)[source][source]
                                                                        
+to_string(grid_size=(20, 20), limit=4)[source][source]
 
                                                                        Returns a string representation of the block mask. Quite nifty.
                                                                        
 
                                                                        If grid_size is None, prints out an uncompressed version. Warning, it can be quite big!
                                                                        
 
                                                                        
diff --git a/2.7/nn.init.html b/2.7/nn.init.html
index be9d2126ae9..1f135904d52 100644
--- a/2.7/nn.init.html
+++ b/2.7/nn.init.html
@@ -600,7 +600,7 @@
 
                                                                        
 
                                                                        
 
                                                                        
-torch.nn.init.calculate_gain(nonlinearity, param=None)[source][source]
                                                                        
+torch.nn.init.calculate_gain(nonlinearity, param=None)[source][source]
 
                                                                        Return the recommended gain value for the given nonlinearity function.
                                                                        
 
                                                                        The values are as follows:
                                                                        
 
                    
 
                    
@@ -678,7 +678,7 @@
 
 
                    
 
                    
-torch.nn.init.uniform_(tensor, a=0.0, b=1.0, generator=None)[source][source]
                    
+torch.nn.init.uniform_(tensor, a=0.0, b=1.0, generator=None)[source][source]
 
                    Fill the input Tensor with values drawn from the uniform distribution.
                    
 
                    U(a,b)\mathcal{U}(a, b).
                    
 
                    
@@ -703,7 +703,7 @@
 
 
                    
 
                    
-torch.nn.init.normal_(tensor, mean=0.0, std=1.0, generator=None)[source][source]
                    
+torch.nn.init.normal_(tensor, mean=0.0, std=1.0, generator=None)[source][source]
 
                    Fill the input Tensor with values drawn from the normal distribution.
                    
 
                    N(mean,std2)\mathcal{N}(\text{mean}, \text{std}^2).
                    
 
                    
@@ -728,7 +728,7 @@
 
 
                    
 
                    
-torch.nn.init.constant_(tensor, val)[source][source]
                    
+torch.nn.init.constant_(tensor, val)[source][source]
 
                    Fill the input Tensor with the value val\text{val}.
                    
 
                    
 
                    Parameters
                    
@@ -750,7 +750,7 @@
 
 
                    
 
                    
-torch.nn.init.ones_(tensor)[source][source]
                    
+torch.nn.init.ones_(tensor)[source][source]
 
                    Fill the input Tensor with the scalar value 1.
                    
 
                    
 
                    Parameters
                    
@@ -769,7 +769,7 @@
 
 
                    
 
                    
-torch.nn.init.zeros_(tensor)[source][source]
                    
+torch.nn.init.zeros_(tensor)[source][source]
 
                    Fill the input Tensor with the scalar value 0.
                    
 
                    
 
                    Parameters
                    
@@ -788,7 +788,7 @@
 
 
                    
 
                    
-torch.nn.init.eye_(tensor)[source][source]
                    
+torch.nn.init.eye_(tensor)[source][source]
 
                    Fill the 2-dimensional input Tensor with the identity matrix.
                    
 
                    Preserves the identity of the inputs in Linear layers, where as
 many inputs are preserved as possible.
                    
@@ -806,7 +806,7 @@
 
 
                    
 
                    
-torch.nn.init.dirac_(tensor, groups=1)[source][source]
                    
+torch.nn.init.dirac_(tensor, groups=1)[source][source]
 
                    Fill the {3, 4, 5}-dimensional input Tensor with the Dirac delta function.
                    
 
                    Preserves the identity of the inputs in Convolutional
 layers, where as many input channels are preserved as possible. In case
@@ -830,7 +830,7 @@
 
 
                    
 
                    
-torch.nn.init.xavier_uniform_(tensor, gain=1.0, generator=None)[source][source]
                    
+torch.nn.init.xavier_uniform_(tensor, gain=1.0, generator=None)[source][source]
 
                    Fill the input Tensor with values using a Xavier uniform distribution.
                    
 
                    The method is described in Understanding the difficulty of training
 deep feedforward neural networks - Glorot, X. & Bengio, Y. (2010).
@@ -877,7 +877,7 @@
 
 
                    
 
                    
-torch.nn.init.xavier_normal_(tensor, gain=1.0, generator=None)[source][source]
                    
+torch.nn.init.xavier_normal_(tensor, gain=1.0, generator=None)[source][source]
 
                    Fill the input Tensor with values using a Xavier normal distribution.
                    
 
                    The method is described in Understanding the difficulty of training deep feedforward
 neural networks - Glorot, X. & Bengio, Y. (2010). The resulting tensor
@@ -923,7 +923,7 @@
 
 
                    
 
                    
-torch.nn.init.kaiming_uniform_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu', generator=None)[source][source]
                    
+torch.nn.init.kaiming_uniform_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu', generator=None)[source][source]
 
                    Fill the input Tensor with values using a Kaiming uniform distribution.
                    
 
                    The method is described in Delving deep into rectifiers: Surpassing
 human-level performance on ImageNet classification - He, K. et al. (2015).
@@ -974,7 +974,7 @@
 
 
                    
 
                    
-torch.nn.init.kaiming_normal_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu', generator=None)[source][source]
                    
+torch.nn.init.kaiming_normal_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu', generator=None)[source][source]
 
                    Fill the input Tensor with values using a Kaiming normal distribution.
                    
 
                    The method is described in Delving deep into rectifiers: Surpassing
 human-level performance on ImageNet classification - He, K. et al. (2015).
@@ -1028,7 +1028,7 @@
 
 
                    
 
                    
-torch.nn.init.trunc_normal_(tensor, mean=0.0, std=1.0, a=-2.0, b=2.0, generator=None)[source][source]
                    
+torch.nn.init.trunc_normal_(tensor, mean=0.0, std=1.0, a=-2.0, b=2.0, generator=None)[source][source]
 
                    Fill the input Tensor with values drawn from a truncated normal distribution.
                    
 
                    The values are effectively drawn from the
 normal distribution N(mean,std2)\mathcal{N}(\text{mean}, \text{std}^2)
@@ -1059,7 +1059,7 @@
 
 
                    
 
                    
-torch.nn.init.orthogonal_(tensor, gain=1, generator=None)[source][source]
                    
+torch.nn.init.orthogonal_(tensor, gain=1, generator=None)[source][source]
 
                    Fill the input Tensor with a (semi) orthogonal matrix.
                    
 
                    Described in Exact solutions to the nonlinear dynamics of learning in deep
 linear neural networks - Saxe, A. et al. (2013). The input tensor must have
@@ -1083,7 +1083,7 @@
 
 
                    
 
                    
-torch.nn.init.sparse_(tensor, sparsity, std=0.01, generator=None)[source][source]
                    
+torch.nn.init.sparse_(tensor, sparsity, std=0.01, generator=None)[source][source]
 
                    Fill the 2D input Tensor as a sparse matrix.
                    
 
                    The non-zero elements will be drawn from the normal distribution
 N(0,0.01)\mathcal{N}(0, 0.01), as described in Deep learning via
diff --git a/2.7/notes/serialization.html b/2.7/notes/serialization.html
index d7449edde02..7685004943a 100644
--- a/2.7/notes/serialization.html
+++ b/2.7/notes/serialization.html
@@ -1012,7 +1012,7 @@ 
                    torch.full always inferring a float dtyp
 
                    The following utility functions are related to serialization:
                    
 
                    
 
                    
-torch.serialization.register_package(priority, tagger, deserializer)[source][source]
                    
+torch.serialization.register_package(priority, tagger, deserializer)[source][source]
 
                    Registers callables for tagging and deserializing storage objects with an associated priority.
 Tagging associates a device with a storage object at save time while deserializing moves a
 storage object to an appropriate device at load time. tagger and deserializer
@@ -1053,7 +1053,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.get_crc32_options()[source][source]
                    
+torch.serialization.get_crc32_options()[source][source]
 
                    Get whether torch.save() computes and writes crc32 for each record.
                    
 
                    Defaults to True.
                    
 
                    
@@ -1065,7 +1065,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.set_crc32_options(compute_crc32)[source][source]
                    
+torch.serialization.set_crc32_options(compute_crc32)[source][source]
 
                    Set whether torch.save() computes and writes crc32 for each record.
                    
 
                    
 
                    Note
                    
@@ -1082,7 +1082,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.get_default_load_endianness()[source][source]
                    
+torch.serialization.get_default_load_endianness()[source][source]
 
                    Get fallback byte order for loading files
                    
 
                    If byteorder mark is not present in saved checkpoint,
 this byte order is used as fallback.
@@ -1099,7 +1099,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.set_default_load_endianness(endianness)[source][source]
                    
+torch.serialization.set_default_load_endianness(endianness)[source][source]
 
                    Set fallback byte order for loading files
                    
 
                    If byteorder mark is not present in saved checkpoint,
 this byte order is used as fallback.
@@ -1113,7 +1113,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.get_default_mmap_options()[source][source]
                    
+torch.serialization.get_default_mmap_options()[source][source]
 
                    Get default mmap options for torch.load() with mmap=True.
                    
 
                    Defaults to mmap.MAP_PRIVATE.
                    
 
                    
@@ -1128,7 +1128,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.set_default_mmap_options(flags)[source][source]
                    
+torch.serialization.set_default_mmap_options(flags)[source][source]
 
                    Context manager or function to set default mmap options for torch.load() with mmap=True to flags.
                    
 
                    For now, only either mmap.MAP_PRIVATE or mmap.MAP_SHARED are supported.
 Please open an issue if you need any other option to be added here.
                    
@@ -1145,7 +1145,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.add_safe_globals(safe_globals)[source][source]
                    
+torch.serialization.add_safe_globals(safe_globals)[source][source]
 
                    Marks the given globals as safe for weights_only load. For example, functions
 added to this list can be called during unpickling, classes could be instantiated
 and have state set.
                    
@@ -1180,7 +1180,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.clear_safe_globals()[source][source]
                    
+torch.serialization.clear_safe_globals()[source][source]
 
                    Clears the list of globals that are safe for weights_only load.
                    
 
                    
 
                    
@@ -1188,7 +1188,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.get_safe_globals()[source][source]
                    
+torch.serialization.get_safe_globals()[source][source]
 
                    Returns the list of user-added globals that are safe for weights_only load.
                    
 
                    
 
                    Return type
                    
@@ -1199,7 +1199,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-torch.serialization.get_unsafe_globals_in_checkpoint(f)[source][source]
                    
+torch.serialization.get_unsafe_globals_in_checkpoint(f)[source][source]
 
                    Returns a list of strings of functions/classes in a torch.save object that are not safe for weights_only.
                    
 
                    For a given function or class f, the corresponding string will be of the form
 {f.__module__}.{f.__name__}.
                    
@@ -1227,7 +1227,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-class torch.serialization.safe_globals(safe_globals)[source][source]
                    
+class torch.serialization.safe_globals(safe_globals)[source][source]
 
                    Context-manager that adds certain globals as safe for weights_only load.
                    
 
                    
 
                    Parameters
                    
@@ -1255,7 +1255,7 @@ 
                    torch.full always inferring a float dtyp
 
 
                    
 
                    
-class torch.serialization.skip_data(materialize_fake_tensors=False)[source][source]
                    
+class torch.serialization.skip_data(materialize_fake_tensors=False)[source][source]
 
                    Context-manager that skips writing/reading storage bytes for torch.save / torch.load calls.
                    
 
                    For the save path, storages will still be saved, but the space that their bytes would usually be written to
 will be empty space. The storage bytes can then be populated in a separate pass.
                    
diff --git a/2.7/onnx_dynamo.html b/2.7/onnx_dynamo.html
index 204b9c20c7f..bb590109e8b 100644
--- a/2.7/onnx_dynamo.html
+++ b/2.7/onnx_dynamo.html
@@ -705,7 +705,7 @@ 
                    When the conversion failsAPI Reference
                    
 
                    
 
                    
-torch.onnx.dynamo_export(model, /, *model_args, export_options=None, **model_kwargs)[source][source]
                    
+torch.onnx.dynamo_export(model, /, *model_args, export_options=None, **model_kwargs)[source][source]
 
                    Export a torch.nn.Module to an ONNX graph.
                    
 
                    
 
                    Deprecated since version 2.7: Please use torch.onnx.export(..., dynamo=True) instead.
                    
@@ -736,7 +736,7 @@ 
                    API Reference
 
                    
 
                    
-apply_weights(state_dict)[source][source]
                    
+apply_weights(state_dict)[source][source]
 
                    Apply the weights from the specified state dict to the ONNX model.
                    
 
                    Use this method to replace FakeTensors or other weights.
                    
 
                    
@@ -748,7 +748,7 @@ 
                    API Reference
 
                    
-compute_values(value_names, args=(), kwargs=None)[source][source]
                    
+compute_values(value_names, args=(), kwargs=None)[source][source]
 
                    Compute the values of the specified names in the ONNX model.
                    
 
                    This method is used to compute the values of the specified names in the ONNX model.
 The values are returned as a dictionary mapping names to tensors.
                    
@@ -767,7 +767,7 @@ 
                    API Reference
 
                    
-initialize_inference_session(initializer=<function _ort_session_initializer>)[source][source]
                    
+initialize_inference_session(initializer=<function _ort_session_initializer>)[source][source]
 
                    Initialize the ONNX Runtime inference session.
                    
 
                    
 
                    Parameters
                    
@@ -786,7 +786,7 @@ 
                    API Reference
 
                    
-optimize()[source][source]
                    
+optimize()[source][source]
 
                    Optimize the ONNX model.
                    
 
                    This method optimizes the ONNX model by performing constant folding and
 eliminating redundancies in the graph. The optimization is done in-place.
                    
@@ -796,7 +796,7 @@ 
                    API Reference
 
                    
-release()[source][source]
                    
+release()[source][source]
 
                    Release the inference session.
                    
 
                    You may call this method to release the resources used by the inference session.
                    
 
                    
@@ -805,7 +805,7 @@ 
                    API Reference
 
                    
-save(destination, *, include_initializers=True, keep_initializers_as_inputs=False, external_data=None)[source][source]
                    
+save(destination, *, include_initializers=True, keep_initializers_as_inputs=False, external_data=None)[source][source]
 
                    Save the ONNX model to the specified destination.
                    
 
                    When external_data is True or the model is larger than 2GB,
 the weights are saved as external data in a separate file.
                    
@@ -867,7 +867,7 @@ 
                    API Reference
 
                    
-torch.onnx.enable_fake_mode()[source]
                    
+torch.onnx.enable_fake_mode()[source]
 
                    Enable fake mode for the duration of the context.
                    
 
                    Internally it instantiates a torch._subclasses.fake_tensor.FakeTensorMode context manager
 that converts user input and model parameters into torch._subclasses.fake_tensor.FakeTensor.
                    
@@ -951,7 +951,7 @@ 
                    API Reference
 
                    
 
                    
-get_op_functions(namespace, op_name, overload=None)[source][source]
                    
+get_op_functions(namespace, op_name, overload=None)[source][source]
 
                    Returns a list of ONNXFunctions for the given op: torch.ops.<namespace>.<op_name>.<overload>.
                    
 
                    The list is ordered by the time of registration. The custom operators should be
 in the second half of the list.
                    
@@ -976,7 +976,7 @@ 
                    API Reference
 
                    
-is_registered_op(namespace, op_name, overload=None)[source][source]
                    
+is_registered_op(namespace, op_name, overload=None)[source][source]
 
                    Returns whether the given op is registered: torch.ops.<namespace>.<op_name>.<overload>.
                    
 
                    
 
                    Parameters
                    
@@ -1004,7 +1004,7 @@ 
                    API Reference
 
                    
-register_op(function, namespace, op_name, overload=None, is_complex=False)[source][source]
                    
+register_op(function, namespace, op_name, overload=None, is_complex=False)[source][source]
 
                    Registers a custom operator: torch.ops.<namespace>.<op_name>.<overload>.
                    
 
                    
 
                    Parameters
                    
diff --git a/2.7/onnx_dynamo_onnxruntime_backend.html b/2.7/onnx_dynamo_onnxruntime_backend.html
index 11a98b1e25f..7899ddbda0a 100644
--- a/2.7/onnx_dynamo_onnxruntime_backend.html
+++ b/2.7/onnx_dynamo_onnxruntime_backend.html
@@ -601,7 +601,7 @@ 
                    ONNX Backend for TorchDynamo
 
                    
-torch.onnx.is_onnxrt_backend_supported()[source]
                    
+torch.onnx.is_onnxrt_backend_supported()[source]
 
                    Returns True if ONNX Runtime dependencies are installed and usable
 to support TorchDynamo backend integration; False otherwise.
                    
 
                    Example:
                    
diff --git a/2.7/onnx_torchscript.html b/2.7/onnx_torchscript.html
index 19845fa17f5..4313c489f75 100644
--- a/2.7/onnx_torchscript.html
+++ b/2.7/onnx_torchscript.html
@@ -1242,7 +1242,7 @@ 
                    Frequently Asked QuestionsFunctions
                    
 
                    
 
                    
-torch.onnx.export(model, args=(), f=None, *, kwargs=None, export_params=True, verbose=None, input_names=None, output_names=None, opset_version=None, dynamic_axes=None, keep_initializers_as_inputs=False, dynamo=False, external_data=True, dynamic_shapes=None, custom_translation_table=None, report=False, optimize=True, verify=False, profile=False, dump_exported_program=False, artifacts_dir='.', fallback=False, training=<TrainingMode.EVAL: 0>, operator_export_type=<OperatorExportTypes.ONNX: 0>, do_constant_folding=True, custom_opsets=None, export_modules_as_functions=False, autograd_inlining=True)[source][source]
                    
+torch.onnx.export(model, args=(), f=None, *, kwargs=None, export_params=True, verbose=None, input_names=None, output_names=None, opset_version=None, dynamic_axes=None, keep_initializers_as_inputs=False, dynamo=False, external_data=True, dynamic_shapes=None, custom_translation_table=None, report=False, optimize=True, verify=False, profile=False, dump_exported_program=False, artifacts_dir='.', fallback=False, training=<TrainingMode.EVAL: 0>, operator_export_type=<OperatorExportTypes.ONNX: 0>, do_constant_folding=True, custom_opsets=None, export_modules_as_functions=False, autograd_inlining=True)[source][source]
 
                    Exports a model into ONNX format.
                    
 
                    
 
                    Parameters
                    
@@ -1449,7 +1449,7 @@ 
                    Functions
 
                    
-torch.onnx.register_custom_op_symbolic(symbolic_name, symbolic_fn, opset_version)[source][source]
                    
+torch.onnx.register_custom_op_symbolic(symbolic_name, symbolic_fn, opset_version)[source][source]
 
                    Registers a symbolic function for a custom operator.
                    
 
                    When the user registers symbolic for custom/contrib ops,
 it is highly recommended to add shape inference for that operator via setType API,
@@ -1472,7 +1472,7 @@ 
                    Functions
 
                    
-torch.onnx.unregister_custom_op_symbolic(symbolic_name, opset_version)[source][source]
                    
+torch.onnx.unregister_custom_op_symbolic(symbolic_name, opset_version)[source][source]
 
                    Unregisters symbolic_name.
                    
 
                    See “Custom Operators” in the module documentation for an example usage.
                    
 
                    
@@ -1488,7 +1488,7 @@ 
                    Functions
 
                    
-torch.onnx.select_model_mode_for_export(model, mode)[source][source]
                    
+torch.onnx.select_model_mode_for_export(model, mode)[source][source]
 
                    A context manager to temporarily set the training mode of model
 to mode, resetting it when we exit the with-block.
                    
 
                    
@@ -1506,7 +1506,7 @@ 
                    Functions
 
                    
-torch.onnx.is_in_onnx_export()[source][source]
                    
+torch.onnx.is_in_onnx_export()[source][source]
 
                    Returns whether it is in the middle of ONNX export.
                    
 
                    
 
                    Return type
                    
diff --git a/2.7/onnx_verification.html b/2.7/onnx_verification.html
index a227d04598d..209e27253e9 100644
--- a/2.7/onnx_verification.html
+++ b/2.7/onnx_verification.html
@@ -597,7 +597,7 @@
 
                    The ONNX verification module provides a set of tools to verify the correctness of ONNX models.
                    
 
                    
 
                    
-torch.onnx.verification.verify_onnx_program(onnx_program, args=None, kwargs=None, compare_intermediates=False)[source]
                    
+torch.onnx.verification.verify_onnx_program(onnx_program, args=None, kwargs=None, compare_intermediates=False)[source]
 
                    Verify the ONNX model by comparing the values with the expected values from ExportedProgram.
                    
 
                    
 
                    Parameters
                    
@@ -644,7 +644,7 @@
 
                    
 
                    
 
                    
-classmethod from_tensors(name, expected, actual)[source][source]
                    
+classmethod from_tensors(name, expected, actual)[source][source]
 
                    Create a VerificationInfo object from two tensors.
                    
 
                    
 
                    Parameters
                    
@@ -667,7 +667,7 @@
 
 
                    
 
                    
-torch.onnx.verification.verify(model, input_args, input_kwargs=None, do_constant_folding=True, dynamic_axes=None, input_names=None, output_names=None, training=<TrainingMode.EVAL: 0>, opset_version=None, keep_initializers_as_inputs=True, verbose=False, fixed_batch_size=False, use_external_data=False, additional_test_inputs=None, options=None)[source][source]
                    
+torch.onnx.verification.verify(model, input_args, input_kwargs=None, do_constant_folding=True, dynamic_axes=None, input_names=None, output_names=None, training=<TrainingMode.EVAL: 0>, opset_version=None, keep_initializers_as_inputs=True, verbose=False, fixed_batch_size=False, use_external_data=False, additional_test_inputs=None, options=None)[source][source]
 
                    Verify model export to ONNX against original PyTorch model.
                    
 
                    
 
                    Deprecated since version 2.7: Consider using torch.onnx.export(..., dynamo=True) and use the returned
@@ -709,42 +709,42 @@ 
                    Deprecated
 
                    
-class torch.onnx.verification.check_export_model_diff[source][source]
                    
+class torch.onnx.verification.check_export_model_diff[source][source]
 
                    
 
 
                    
 
                    
-class torch.onnx.verification.GraphInfo[source][source]
                    
+class torch.onnx.verification.GraphInfo[source][source]
 
                    
 
 
                    
 
                    
-class torch.onnx.verification.GraphInfoPrettyPrinter[source][source]
                    
+class torch.onnx.verification.GraphInfoPrettyPrinter[source][source]
 
                    
 
 
                    
 
                    
-class torch.onnx.verification.OnnxBackend[source][source]
                    
+class torch.onnx.verification.OnnxBackend[source][source]
 
                    
 
 
                    
 
                    
-class torch.onnx.verification.OnnxTestCaseRepro[source][source]
                    
+class torch.onnx.verification.OnnxTestCaseRepro[source][source]
 
                    
 
 
                    
 
                    
-class torch.onnx.verification.VerificationOptions[source][source]
                    
+class torch.onnx.verification.VerificationOptions[source][source]
 
                    
 
 
                    
 
                    
-torch.onnx.verification.find_mismatch()[source][source]
                    
+torch.onnx.verification.find_mismatch()[source][source]
 
                    
 
 
                    
 
                    
-torch.onnx.verification.verify_aten_graph()[source][source]
                    
+torch.onnx.verification.verify_aten_graph()[source][source]
 
                    
 
 
                    
diff --git a/2.7/optim.html b/2.7/optim.html
index db9e63f15e2..90d49dcf2a8 100644
--- a/2.7/optim.html
+++ b/2.7/optim.html
@@ -708,7 +708,7 @@ 
                    optimizer.step(
 
                    Base class
                    
 
                    
 
                    
-class torch.optim.Optimizer(params, defaults)[source][source]
                    
+class torch.optim.Optimizer(params, defaults)[source][source]
 
                    Base class for all optimizers.
                    
 
                    
 
                    Warning
                    
@@ -1436,13 +1436,13 @@ 
                    Putting it all together: EMA
 
                    
-torch.optim.swa_utils.get_ema_multi_avg_fn(decay=0.999)[source][source]
                    
+torch.optim.swa_utils.get_ema_multi_avg_fn(decay=0.999)[source][source]
 
                    Get the function applying exponential moving average (EMA) across multiple params.
                    
 
                    
 
 
                    
 
                    
-torch.optim.swa_utils.update_bn(loader, model, device=None)[source][source]
                    
+torch.optim.swa_utils.update_bn(loader, model, device=None)[source][source]
 
                    Update BatchNorm running_mean, running_var buffers in the model.
                    
 
                    It performs one pass over data in loader to estimate the activation
 statistics for BatchNorm layers in the model.
                    
diff --git a/2.7/package.html b/2.7/package.html
index 94774602e27..fc1fc2c3fe1 100644
--- a/2.7/package.html
+++ b/2.7/package.html
@@ -1283,7 +1283,7 @@ 
                    ManglingAPI Reference
                    
 
                    
 
                    
-class torch.package.PackagingError(dependency_graph, debug=False)[source][source]
                    
+class torch.package.PackagingError(dependency_graph, debug=False)[source][source]
 
                    This exception is raised when there is an issue with exporting a package.
 PackageExporter will attempt to gather up all the errors and present
 them to you at once.
                    
@@ -1293,14 +1293,14 @@ 
                    API Reference
 
                    
-class torch.package.EmptyMatchError[source][source]
                    
+class torch.package.EmptyMatchError[source][source]
 
                    This is an exception that is thrown when a mock or extern is marked as
 allow_empty=False, and is not matched with any module during packaging.
                    
 
                    
 
 
                    
 
                    
-class torch.package.PackageExporter(f, importer=<torch.package.importer._SysImporter object>, debug=False)[source][source]
                    
+class torch.package.PackageExporter(f, importer=<torch.package.importer._SysImporter object>, debug=False)[source][source]
 
                    Exporters allow you to write packages of code, pickled Python data, and
 arbitrary binary and text resources into a self-contained package.
                    
 
                    Imports can load this code in a hermetic way, such that code is loaded
@@ -1324,7 +1324,7 @@ 
                    API Reference
 
                    
 
                    
-__init__(f, importer=<torch.package.importer._SysImporter object>, debug=False)[source][source]
                    
+__init__(f, importer=<torch.package.importer._SysImporter object>, debug=False)[source][source]
 
                    Create an exporter.
                    
 
                    
 
                    Parameters
                    
@@ -1341,7 +1341,7 @@ 
                    API Reference
 
                    
-add_dependency(module_name, dependencies=True)[source][source]
                    
+add_dependency(module_name, dependencies=True)[source][source]
 
                    Given a module, add it to the dependency graph according to patterns
 specified by the user.
                    
 
                    
@@ -1350,7 +1350,7 @@ 
                    API Reference
 
                    
-all_paths(src, dst)[source][source]
                    
+all_paths(src, dst)[source][source]
 
                    
 
                    Return a dot representation of the subgraph
                    that has all paths from src to dst.
                    
 
                    
@@ -1368,7 +1368,7 @@ 
                    API Reference
 
                    
-close()[source][source]
                    
+close()[source][source]
 
                    Write the package to the filesystem. Any calls after close() are now invalid.
 It is preferable to use resource guard syntax instead:
                    
 
                    with PackageExporter("file.zip") as e:
@@ -1379,7 +1379,7 @@ 
                    API Reference
 
                    
-denied_modules()[source][source]
                    
+denied_modules()[source][source]
 
                    Return all modules that are currently denied.
                    
 
                    
 
                    Returns
                    
@@ -1394,7 +1394,7 @@ 
                    API Reference
 
                    
-deny(include, *, exclude=())[source][source]
                    
+deny(include, *, exclude=())[source][source]
 
                    Blocklist modules who names match the given glob patterns from the list of modules the package can import.
 If a dependency on any matching packages is found, a PackagingError is raised.
                    
 
                    
@@ -1410,7 +1410,7 @@ 
                    API Reference
 
                    
-dependency_graph_string()[source][source]
                    
+dependency_graph_string()[source][source]
 
                    Returns digraph string representation of dependencies in package.
                    
 
                    
 
                    Returns
                    
@@ -1424,7 +1424,7 @@ 
                    API Reference
 
                    
-extern(include, *, exclude=(), allow_empty=True)[source][source]
                    
+extern(include, *, exclude=(), allow_empty=True)[source][source]
 
                    Include module in the list of external modules the package can import.
 This will prevent dependency discovery from saving
 it in the package. The importer will load an external module directly from the standard import system.
@@ -1449,7 +1449,7 @@ 
                    API Reference
 
                    
-externed_modules()[source][source]
                    
+externed_modules()[source][source]
 
                    Return all modules that are currently externed.
                    
 
                    
 
                    Returns
                    
@@ -1464,7 +1464,7 @@ 
                    API Reference
 
                    
-get_rdeps(module_name)[source][source]
                    
+get_rdeps(module_name)[source][source]
 
                    Return a list of all modules which depend on the module module_name.
                    
 
                    
 
                    Returns
                    
@@ -1478,7 +1478,7 @@ 
                    API Reference
 
                    
-get_unique_id()[source][source]
                    
+get_unique_id()[source][source]
 
                    Get an id. This id is guaranteed to only be handed out once for this package.
                    
 
                    
 
                    Return type
                    
@@ -1489,7 +1489,7 @@ 
                    API Reference
 
                    
-intern(include, *, exclude=(), allow_empty=True)[source][source]
                    
+intern(include, *, exclude=(), allow_empty=True)[source][source]
 
                    Specify modules that should be packaged. A module must match some intern pattern in order to be
 included in the package and have its dependencies processed recursively.
                    
 
                    
@@ -1509,7 +1509,7 @@ 
                    API Reference
 
                    
-interned_modules()[source][source]
                    
+interned_modules()[source][source]
 
                    Return all modules that are currently interned.
                    
 
                    
 
                    Returns
                    
@@ -1524,7 +1524,7 @@ 
                    API Reference
 
                    
-mock(include, *, exclude=(), allow_empty=True)[source][source]
                    
+mock(include, *, exclude=(), allow_empty=True)[source][source]
 
                    Replace some required modules with a mock implementation.  Mocked modules will return a fake
 object for any attribute accessed from it. Because we copy file-by-file, the dependency resolution will sometimes
 find files that are imported by model files but whose functionality is never used
@@ -1560,7 +1560,7 @@ 
                    API Reference
 
                    
-mocked_modules()[source][source]
                    
+mocked_modules()[source][source]
 
                    Return all modules that are currently mocked.
                    
 
                    
 
                    Returns
                    
@@ -1575,7 +1575,7 @@ 
                    API Reference
 
                    
-register_extern_hook(hook)[source][source]
                    
+register_extern_hook(hook)[source][source]
 
                    Registers an extern hook on the exporter.
                    
 
                    The hook will be called each time a module matches against an extern() pattern.
 It should have the following signature:
                    
@@ -1596,7 +1596,7 @@ 
                    API Reference
 
                    
-register_intern_hook(hook)[source][source]
                    
+register_intern_hook(hook)[source][source]
 
                    Registers an intern hook on the exporter.
                    
 
                    The hook will be called each time a module matches against an intern() pattern.
 It should have the following signature:
                    
@@ -1617,7 +1617,7 @@ 
                    API Reference
 
                    
-register_mock_hook(hook)[source][source]
                    
+register_mock_hook(hook)[source][source]
 
                    Registers a mock hook on the exporter.
                    
 
                    The hook will be called each time a module matches against a mock() pattern.
 It should have the following signature:
                    
@@ -1638,7 +1638,7 @@ 
                    API Reference
 
                    
-save_binary(package, resource, binary)[source][source]
                    
+save_binary(package, resource, binary)[source][source]
 
                    Save raw bytes to the package.
                    
 
                    
 
                    Parameters
                    
@@ -1653,7 +1653,7 @@ 
                    API Reference
 
                    
-save_module(module_name, dependencies=True)[source][source]
                    
+save_module(module_name, dependencies=True)[source][source]
 
                    Save the code for module into the package. Code for the module is resolved using the importers path to find the
 module object, and then using its __file__ attribute to find the source code.
                    
 
                    
@@ -1669,7 +1669,7 @@ 
                    API Reference
 
                    
-save_pickle(package, resource, obj, dependencies=True, pickle_protocol=3)[source][source]
                    
+save_pickle(package, resource, obj, dependencies=True, pickle_protocol=3)[source][source]
 
                    Save a python object to the archive using pickle. Equivalent to torch.save() but saving into
 the archive rather than a stand-alone file. Standard pickle does not save the code, only the objects.
 If dependencies is true, this method will also scan the pickled objects for which modules are required
@@ -1692,7 +1692,7 @@ 
                    API Reference
 
                    
-save_source_file(module_name, file_or_directory, dependencies=True)[source][source]
                    
+save_source_file(module_name, file_or_directory, dependencies=True)[source][source]
 
                    Adds the local file system file_or_directory to the source package to provide the code
 for module_name.
                    
 
                    
@@ -1710,7 +1710,7 @@ 
                    API Reference
 
                    
-save_source_string(module_name, src, is_package=False, dependencies=True)[source][source]
                    
+save_source_string(module_name, src, is_package=False, dependencies=True)[source][source]
 
                    Adds src as the source code for module_name in the exported package.
                    
 
                    
 
                    Parameters
                    
@@ -1727,7 +1727,7 @@ 
                    API Reference
 
                    
-save_text(package, resource, text)[source][source]
                    
+save_text(package, resource, text)[source][source]
 
                    Save text data to the package.
                    
 
                    
 
                    Parameters
                    
@@ -1744,7 +1744,7 @@ 
                    API Reference
 
                    
-class torch.package.PackageImporter(file_or_buffer, module_allowed=<function PackageImporter.<lambda>>)[source][source]
                    
+class torch.package.PackageImporter(file_or_buffer, module_allowed=<function PackageImporter.<lambda>>)[source][source]
 
                    Importers allow you to load code written to packages by PackageExporter.
 Code is loaded in a hermetic way, using files from the package
 rather than the normal python import system. This allows
@@ -1759,7 +1759,7 @@ 
                    API Reference
 
                    
 
                    
-__init__(file_or_buffer, module_allowed=<function PackageImporter.<lambda>>)[source][source]
                    
+__init__(file_or_buffer, module_allowed=<function PackageImporter.<lambda>>)[source][source]
 
                    Open file_or_buffer for importing. This checks that the imported package only requires modules
 allowed by module_allowed
                    
 
                    
@@ -1780,7 +1780,7 @@ 
                    API Reference
 
                    
-file_structure(*, include='**', exclude=())[source][source]
                    
+file_structure(*, include='**', exclude=())[source][source]
 
                    Returns a file structure representation of package’s zipfile.
                    
 
                    
 
                    Parameters
                    
@@ -1802,7 +1802,7 @@ 
                    API Reference
 
                    
-id()[source][source]
                    
+id()[source][source]
 
                    Returns internal identifier that torch.package uses to distinguish PackageImporter instances.
 Looks like:
                    
 
                    <torch_package_0>
@@ -1812,7 +1812,7 @@ 
                    API Reference
 
                    
-import_module(name, package=None)[source][source]
                    
+import_module(name, package=None)[source][source]
 
                    Load a module from the package if it hasn’t already been loaded, and then return
 the module. Modules are loaded locally
 to the importer and will appear in self.modules rather than sys.modules.
                    
@@ -1834,7 +1834,7 @@ 
                    API Reference
 
                    
-load_binary(package, resource)[source][source]
                    
+load_binary(package, resource)[source][source]
 
                    Load raw bytes.
                    
 
                    
 
                    Parameters
                    
@@ -1854,7 +1854,7 @@ 
                    API Reference
 
                    
-load_pickle(package, resource, map_location=None)[source][source]
                    
+load_pickle(package, resource, map_location=None)[source][source]
 
                    Unpickles the resource from the package, loading any modules that are needed to construct the objects
 using import_module().
                    
 
                    
@@ -1876,7 +1876,7 @@ 
                    API Reference
 
                    
-load_text(package, resource, encoding='utf-8', errors='strict')[source][source]
                    
+load_text(package, resource, encoding='utf-8', errors='strict')[source][source]
 
                    Load a string.
                    
 
                    
 
                    Parameters
                    
@@ -1898,7 +1898,7 @@ 
                    API Reference
 
                    
-python_version()[source][source]
                    
+python_version()[source][source]
 
                    Returns the version of python that was used to create this package.
                    
 
                    Note: this function is experimental and not Forward Compatible. The plan is to move this into a lock
 file later on.
                    
@@ -1913,7 +1913,7 @@ 
                    API Reference
 
                    
-class torch.package.Directory(name, is_dir)[source][source]
                    
+class torch.package.Directory(name, is_dir)[source][source]
 
                    A file structure representation. Organized as Directory nodes that have lists of
 their Directory children. Directories for a package are created by calling
 PackageImporter.file_structure().
                    
@@ -1921,7 +1921,7 @@ 
                    API Reference
 
                    
 
                    
-has_file(filename)[source][source]
                    
+has_file(filename)[source][source]
 
                    Checks if a file is present in a Directory.
                    
 
                    
 
                    Parameters
                    
diff --git a/2.7/profiler.html b/2.7/profiler.html
index 78157cacb21..b68c186b3b9 100644
--- a/2.7/profiler.html
+++ b/2.7/profiler.html
@@ -606,7 +606,7 @@ 
                    torch.profiler
                    
 
                    
 
                    
-class torch.profiler._KinetoProfile(*, activities=None, record_shapes=False, profile_memory=False, with_stack=False, with_flops=False, with_modules=False, experimental_config=None, execution_trace_observer=None, acc_events=False, custom_trace_id_callback=None)[source][source]
                    
+class torch.profiler._KinetoProfile(*, activities=None, record_shapes=False, profile_memory=False, with_stack=False, with_flops=False, with_modules=False, experimental_config=None, execution_trace_observer=None, acc_events=False, custom_trace_id_callback=None)[source][source]
 
                    Low-level profiler wrap the autograd profile
                    
 
                    
 
                    Parameters
                    
@@ -649,7 +649,7 @@ 
                    API Reference
 
                    
-add_metadata(key, value)[source][source]
                    
+add_metadata(key, value)[source][source]
 
                    Adds a user defined metadata with a string key and a string value
 into the trace file
                    
 
                    
@@ -658,7 +658,7 @@ 
                    API Reference
 
                    
-add_metadata_json(key, value)[source][source]
                    
+add_metadata_json(key, value)[source][source]
 
                    Adds a user defined metadata with a string key and a valid json value
 into the trace file
                    
 
                    
@@ -667,14 +667,14 @@ 
                    API Reference
 
                    
-events()[source][source]
                    
+events()[source][source]
 
                    Returns the list of unaggregated profiler events,
 to be used in the trace callback or after the profiling is finished
                    
 
                    
 
 
                    
 
                    
-export_chrome_trace(path)[source][source]
                    
+export_chrome_trace(path)[source][source]
 
                    Exports the collected trace in Chrome JSON format. If kineto is enabled, only
 last cycle in schedule is exported.
                    
 
                    
@@ -683,7 +683,7 @@ 
                    API Reference
 
                    
-export_memory_timeline(path, device=None)[source][source]
                    
+export_memory_timeline(path, device=None)[source][source]
 
                    Export memory event information from the profiler collected
 tree for a given device, and export a timeline plot. There are 3
 exportable files using export_memory_timeline, each controlled by the
@@ -708,7 +708,7 @@ 
                    API Reference
 
                    
-export_stacks(path, metric='self_cpu_time_total')[source][source]
                    
+export_stacks(path, metric='self_cpu_time_total')[source][source]
 
                    Save stack traces to a file
                    
 
                    
 
                    Parameters
                    
@@ -722,7 +722,7 @@ 
                    API Reference
 
                    
-key_averages(group_by_input_shape=False, group_by_stack_n=0, group_by_overload_name=False)[source][source]
                    
+key_averages(group_by_input_shape=False, group_by_stack_n=0, group_by_overload_name=False)[source][source]
 
                    Averages events, grouping them by operator name and (optionally) input shapes, stack
 and overload name.
                    
 
                    
@@ -736,7 +736,7 @@ 
                    API Reference
 
                    
-preset_metadata_json(key, value)[source][source]
                    
+preset_metadata_json(key, value)[source][source]
 
                    Preset a user defined metadata when the profiler is not started
 and added into the trace file later.
 Metadata is in the format of a string key and a valid json value
                    
@@ -746,7 +746,7 @@ 
                    API Reference
 
                    
-toggle_collection_dynamic(enable, activities)[source][source]
                    
+toggle_collection_dynamic(enable, activities)[source][source]
 
                    Toggle collection of activities on/off at any point of collection. Currently supports toggling Torch Ops
 (CPU) and CUDA activity supported in Kineto
                    
 
                    
@@ -779,7 +779,7 @@ 
                    API Reference
 
                    
-class torch.profiler.profile(*, activities=None, schedule=None, on_trace_ready=None, record_shapes=False, profile_memory=False, with_stack=False, with_flops=False, with_modules=False, experimental_config=None, execution_trace_observer=None, acc_events=False, use_cuda=None, custom_trace_id_callback=None)[source][source]
                    
+class torch.profiler.profile(*, activities=None, schedule=None, on_trace_ready=None, record_shapes=False, profile_memory=False, with_stack=False, with_flops=False, with_modules=False, experimental_config=None, execution_trace_observer=None, acc_events=False, use_cuda=None, custom_trace_id_callback=None)[source][source]
 
                    Profiler context manager.
                    
 
                    
 
                    Parameters
                    
@@ -911,19 +911,19 @@ 
                    API Reference
 
                    
-get_trace_id()[source][source]
                    
+get_trace_id()[source][source]
 
                    Returns the current trace ID.
                    
 
                    
 
 
                    
 
                    
-set_custom_trace_id_callback(callback)[source][source]
                    
+set_custom_trace_id_callback(callback)[source][source]
 
                    Sets a callback to be called when a new trace ID is generated.
                    
 
                    
 
 
                    
 
                    
-step()[source][source]
                    
+step()[source][source]
 
                    Signals the profiler that the next profiling step has started.
                    
 
                    
 
@@ -931,7 +931,7 @@ 
                    API Reference
 
                    
-class torch.profiler.ProfilerAction(value)[source][source]
                    
+class torch.profiler.ProfilerAction(value)[source][source]
 
                    Profiler actions that can be taken at the specified intervals
                    
 
                    
 
@@ -954,7 +954,7 @@ 
                    API Reference
 
                    
-torch.profiler.schedule(*, wait, warmup, active, repeat=0, skip_first=0, skip_first_wait=0)[source][source]
                    
+torch.profiler.schedule(*, wait, warmup, active, repeat=0, skip_first=0, skip_first_wait=0)[source][source]
 
                    Returns a callable that can be used as profiler schedule argument. The profiler will skip
 the first skip_first steps, then wait for wait steps, then do the warmup for the next warmup steps,
 then do the active recording for the next active steps and then repeat the cycle starting with wait steps.
@@ -975,7 +975,7 @@ 
                    API Reference
 
                    
-torch.profiler.tensorboard_trace_handler(dir_name, worker_name=None, use_gzip=False)[source][source]
                    
+torch.profiler.tensorboard_trace_handler(dir_name, worker_name=None, use_gzip=False)[source][source]
 
                    Outputs tracing files to directory of dir_name, then that directory can be
 directly delivered to tensorboard as logdir.
 worker_name should be unique for each worker in distributed scenario,
@@ -989,13 +989,13 @@ 
                    API Reference
                    
 
                    
 
                    
-torch.profiler.itt.is_available()[source][source]
                    
+torch.profiler.itt.is_available()[source][source]
 
                    Check if ITT feature is available or not
                    
 
                    
 
 
                    
 
                    
-torch.profiler.itt.mark(msg)[source][source]
                    
+torch.profiler.itt.mark(msg)[source][source]
 
                    Describe an instantaneous event that occurred at some point.
                    
 
                    
 
                    Parameters
                    
@@ -1006,7 +1006,7 @@ 
                    Intel Instrumentation and Tracing Technology APIs
 
                    
-torch.profiler.itt.range_push(msg)[source][source]
                    
+torch.profiler.itt.range_push(msg)[source][source]
 
                    Pushes a range onto a stack of nested range span.  Returns zero-based
 depth of the range that is started.
                    
 
                    
@@ -1018,7 +1018,7 @@ 
                    Intel Instrumentation and Tracing Technology APIs
 
                    
-torch.profiler.itt.range_pop()[source][source]
                    
+torch.profiler.itt.range_pop()[source][source]
 
                    Pops a range off of a stack of nested range spans. Returns the
 zero-based depth of the range that is ended.
                    
 
                    
diff --git a/2.7/random.html b/2.7/random.html
index 01458832087..fd9b1f8bbef 100644
--- a/2.7/random.html
+++ b/2.7/random.html
@@ -594,7 +594,7 @@
 
                    torch.random
                    
 
                    
 
                    
-torch.random.fork_rng(devices=None, enabled=True, _caller='fork_rng', _devices_kw='devices', device_type='cuda')[source][source]
                    
+torch.random.fork_rng(devices=None, enabled=True, _caller='fork_rng', _devices_kw='devices', device_type='cuda')[source][source]
 
                    Forks the RNG, so that when you return, the RNG is reset
 to the state that it was previously in.
                    
 
                    
@@ -620,7 +620,7 @@
 
 
                    
 
                    
-torch.random.get_rng_state()[source][source]
                    
+torch.random.get_rng_state()[source][source]
 
                    Returns the random number generator state as a torch.ByteTensor.
                    
 
                    
 
                    Note
                    
@@ -636,7 +636,7 @@
 
 
                    
 
                    
-torch.random.initial_seed()[source][source]
                    
+torch.random.initial_seed()[source][source]
 
                    Returns the initial seed for generating random numbers as a
 Python long.
                    
 
                    
@@ -652,7 +652,7 @@
 
 
                    
 
                    
-torch.random.manual_seed(seed)[source][source]
                    
+torch.random.manual_seed(seed)[source][source]
 
                    Sets the seed for generating random numbers on all devices. Returns a
 torch.Generator object.
                    
 
                    
@@ -670,7 +670,7 @@
 
 
                    
 
                    
-torch.random.seed()[source][source]
                    
+torch.random.seed()[source][source]
 
                    Sets the seed for generating random numbers to a non-deterministic
 random number on all devices. Returns a 64 bit number used to seed the RNG.
                    
 
                    
@@ -682,7 +682,7 @@
 
 
                    
 
                    
-torch.random.set_rng_state(new_state)[source][source]
                    
+torch.random.set_rng_state(new_state)[source][source]
 
                    Sets the random number generator state.
                    
 
                    
 
                    Note
                    
diff --git a/2.7/rpc.html b/2.7/rpc.html
index 5445d9e6524..c2d786491b0 100644
--- a/2.7/rpc.html
+++ b/2.7/rpc.html
@@ -677,7 +677,7 @@ 
                    Basics
 
                    
 
                    
-torch.distributed.rpc.init_rpc(name, backend=None, rank=-1, world_size=None, rpc_backend_options=None)[source][source]
                    
+torch.distributed.rpc.init_rpc(name, backend=None, rank=-1, world_size=None, rpc_backend_options=None)[source][source]
 
                    Initializes RPC primitives such as the local RPC agent
 and distributed autograd, which immediately makes the current
 process ready to send and receive RPCs.
                    
@@ -727,7 +727,7 @@ 
                    Basics
 
                    
 
                    
-torch.distributed.rpc.rpc_sync(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
                    
+torch.distributed.rpc.rpc_sync(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
 
                    Make a blocking RPC call to run function func on worker to. RPC
 messages are sent and received in parallel to execution of Python code. This
 method is thread-safe.
                    
@@ -802,7 +802,7 @@ 
                    Basics
 
                    
-torch.distributed.rpc.rpc_async(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
                    
+torch.distributed.rpc.rpc_async(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
 
                    Make a non-blocking RPC call to run function func on worker to. RPC
 messages are sent and received in parallel to execution of Python code. This
 method is thread-safe. This method will immediately return a
@@ -899,7 +899,7 @@ 
                    Basics
 
                    
-torch.distributed.rpc.remote(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
                    
+torch.distributed.rpc.remote(to, func, args=None, kwargs=None, timeout=-1.0)[source][source]
 
                    Make a remote call to run func on worker to and return an
 RRef to the result value immediately.
 Worker to will be the owner of the returned
@@ -1008,7 +1008,7 @@ 
                    Basics
 
                    
-torch.distributed.rpc.get_worker_info(worker_name=None)[source][source]
                    
+torch.distributed.rpc.get_worker_info(worker_name=None)[source][source]
 
                    Get WorkerInfo of a given worker name.
 Use this WorkerInfo to avoid passing an
 expensive string on every invocation.
                    
@@ -1027,7 +1027,7 @@ 
                    Basics
 
                    
-torch.distributed.rpc.shutdown(graceful=True, timeout=0)[source][source]
                    
+torch.distributed.rpc.shutdown(graceful=True, timeout=0)[source][source]
 
                    Perform a shutdown of the RPC agent, and then destroy the RPC agent. This
 stops the local agent from accepting outstanding requests, and shuts
 down the RPC framework by terminating all RPC threads. If graceful=True,
@@ -1108,7 +1108,7 @@ 
                    Basics
 
                    
 
                    
-torch.distributed.rpc.functions.async_execution(fn)[source][source]
                    
+torch.distributed.rpc.functions.async_execution(fn)[source][source]
 
                    A decorator for a function indicating that the return value of the function
 is guaranteed to be a Future object and this
 function can run asynchronously on the RPC callee. More specifically, the
@@ -1351,7 +1351,7 @@ 
                    TensorPipe Backend
 
                    
-class torch.distributed.rpc.TensorPipeRpcBackendOptions(*, num_worker_threads=16, rpc_timeout=60.0, init_method='env://', device_maps=None, devices=None, _transports=None, _channels=None)[source][source]
                    
+class torch.distributed.rpc.TensorPipeRpcBackendOptions(*, num_worker_threads=16, rpc_timeout=60.0, init_method='env://', device_maps=None, devices=None, _transports=None, _channels=None)[source][source]
 
                    The backend options for
 TensorPipeAgent, derived from
 RpcBackendOptions.
                    
@@ -1423,7 +1423,7 @@ 
                    TensorPipe Backend
 
                    
-set_device_map(to, device_map)[source][source]
                    
+set_device_map(to, device_map)[source][source]
 
                    Set device mapping between each RPC caller and callee pair. This
 function can be called multiple times to incrementally add
 device placement configurations.
                    
@@ -1474,7 +1474,7 @@ 
                    TensorPipe Backend
 
                    
-set_devices(devices)[source][source]
                    
+set_devices(devices)[source][source]
 
                    Set local devices used by the TensorPipe RPC agent. When processing
 CUDA RPC requests, the TensorPipe RPC agent will properly synchronize
 CUDA streams for all devices in this List.
                    
@@ -1821,7 +1821,7 @@ 
                    TensorPipe Backend
 
                    
-class torch.distributed.nn.api.remote_module.RemoteModule(*args, **kwargs)[source][source]
                    
+class torch.distributed.nn.api.remote_module.RemoteModule(*args, **kwargs)[source][source]
 
                    
 
                    A RemoteModule instance can only be created after RPC initialization.
                    
 
                    It creates a user-specified module on a specified remote node.
@@ -1902,7 +1902,7 @@ 
                    TensorPipe Backend
 
                    
-get_module_rref()[source]
                    
+get_module_rref()[source]
 
                    Return an RRef (RRef[nn.Module]) pointing to the remote module.
                    
 
                    
 
                    Return type
                    
@@ -1913,7 +1913,7 @@ 
                    TensorPipe Backend
 
                    
-remote_parameters(recurse=True)[source]
                    
+remote_parameters(recurse=True)[source]
 
                    Return a list of RRef pointing to the remote module’s parameters.
                    
 
                    This can typically be used in conjunction
 with DistributedOptimizer.
                    
@@ -1994,7 +1994,7 @@ 
                    Distributed Autograd Framework
 
                    
-class torch.distributed.autograd.context[source][source]
                    
+class torch.distributed.autograd.context[source][source]
 
                    Context object to wrap forward and backward passes when using
 distributed autograd. The context_id generated in the with
 statement  is required to uniquely identify a distributed backward pass
diff --git a/2.7/storage.html b/2.7/storage.html
index 77a8eab9817..d7694979c00 100644
--- a/2.7/storage.html
+++ b/2.7/storage.html
@@ -702,52 +702,52 @@ 
                    Special cases
 
                    
-class torch.UntypedStorage(*args, **kwargs)[source][source]
                    
+class torch.UntypedStorage(*args, **kwargs)[source][source]
 
                    
 
                    
-bfloat16()[source]
                    
+bfloat16()[source]
 
                    Casts this storage to bfloat16 type.
                    
 
                    
 
 
                    
 
                    
-bool()[source]
                    
+bool()[source]
 
                    Casts this storage to bool type.
                    
 
                    
 
 
                    
 
                    
-byte()[source]
                    
+byte()[source]
 
                    Casts this storage to byte type.
                    
 
                    
 
 
                    
 
                    
-byteswap(dtype)[source]
                    
+byteswap(dtype)[source]
 
                    Swap bytes in underlying data.
                    
 
                    
 
 
                    
 
                    
-char()[source]
                    
+char()[source]
 
                    Casts this storage to char type.
                    
 
                    
 
 
                    
 
                    
-clone()[source]
                    
+clone()[source]
 
                    Return a copy of this storage.
                    
 
                    
 
 
                    
 
                    
-complex_double()[source]
                    
+complex_double()[source]
 
                    Casts this storage to complex double type.
                    
 
                    
 
 
                    
 
                    
-complex_float()[source]
                    
+complex_float()[source]
 
                    Casts this storage to complex float type.
                    
 
                    
 
@@ -758,13 +758,13 @@ 
                    Special cases
 
                    
-cpu()[source]
                    
+cpu()[source]
 
                    Return a CPU copy of this storage if it’s not already on the CPU.
                    
 
                    
 
 
                    
 
                    
-cuda(device=None, non_blocking=False)[source]
                    
+cuda(device=None, non_blocking=False)[source]
 
                    Returns a copy of this object in CUDA memory.
                    
 
                    If this object is already in CUDA memory and on the correct device, then
 no copy is performed and the original object is returned.
                    
@@ -795,7 +795,7 @@ 
                    Special cases
 
                    
-double()[source]
                    
+double()[source]
 
                    Casts this storage to double type.
                    
 
                    
 
@@ -819,31 +819,31 @@ 
                    Special cases
 
                    
-float()[source]
                    
+float()[source]
 
                    Casts this storage to float type.
                    
 
                    
 
 
                    
 
                    
-float8_e4m3fn()[source]
                    
+float8_e4m3fn()[source]
 
                    Casts this storage to float8_e4m3fn type
                    
 
                    
 
 
                    
 
                    
-float8_e4m3fnuz()[source]
                    
+float8_e4m3fnuz()[source]
 
                    Casts this storage to float8_e4m3fnuz type
                    
 
                    
 
 
                    
 
                    
-float8_e5m2()[source]
                    
+float8_e5m2()[source]
 
                    Casts this storage to float8_e5m2 type
                    
 
                    
 
 
                    
 
                    
-float8_e5m2fnuz()[source]
                    
+float8_e5m2fnuz()[source]
 
                    Casts this storage to float8_e5m2fnuz type
                    
 
                    
 
@@ -877,7 +877,7 @@ 
                    Special cases
 
                    
-get_device()[source]
                    
+get_device()[source]
 
                    
 
                    Return type
                    
 
                    int
                    
@@ -887,13 +887,13 @@ 
                    Special cases
 
                    
-half()[source]
                    
+half()[source]
 
                    Casts this storage to half type.
                    
 
                    
 
 
                    
 
                    
-hpu(device=None, non_blocking=False)[source]
                    
+hpu(device=None, non_blocking=False)[source]
 
                    Returns a copy of this object in HPU memory.
                    
 
                    If this object is already in HPU memory and on the correct device, then
 no copy is performed and the original object is returned.
                    
@@ -914,7 +914,7 @@ 
                    Special cases
 
                    
-int()[source]
                    
+int()[source]
 
                    Casts this storage to int type.
                    
 
                    
 
@@ -930,7 +930,7 @@ 
                    Special cases
 
                    
-is_pinned(device='cuda')[source]
                    
+is_pinned(device='cuda')[source]
 
                    Determine whether the CPU storage is already pinned on device.
                    
 
                    
 
                    Parameters
                    
@@ -960,13 +960,13 @@ 
                    Special cases
 
                    
-long()[source]
                    
+long()[source]
 
                    Casts this storage to long type.
                    
 
                    
 
 
                    
 
                    
-mps()[source]
                    
+mps()[source]
 
                    Return a MPS copy of this storage if it’s not already on the MPS.
                    
 
                    
 
@@ -982,7 +982,7 @@ 
                    Special cases
 
                    
-pin_memory(device='cuda')[source]
                    
+pin_memory(device='cuda')[source]
 
                    Copy the CPU storage to pinned memory, if it’s not already pinned.
                    
 
                    
 
                    Parameters
                    
@@ -1007,7 +1007,7 @@ 
                    Special cases
 
                    
-share_memory_(*args, **kwargs)[source][source]
                    
+share_memory_(*args, **kwargs)[source][source]
 
                    Moves the storage to shared memory.
                    
 
                    This is a no-op for storages already in shared memory and for CUDA
 storages, which do not need to be moved for sharing across processes.
@@ -1042,13 +1042,13 @@ 
                    Special cases
 
                    
-short()[source]
                    
+short()[source]
 
                    Casts this storage to short type.
                    
 
                    
 
 
                    
 
                    
-size()[source]
                    
+size()[source]
 
                    
 
                    Return type
                    
 
                    int
                    
@@ -1058,20 +1058,20 @@ 
                    Special cases
 
                    
-to(*, device, non_blocking=False)[source]
                    
+to(*, device, non_blocking=False)[source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-tolist()[source]
                    
+tolist()[source]
 
                    Return a list containing the elements of this storage.
                    
 
                    
 
 
                    
 
                    
-type(dtype=None, non_blocking=False)[source]
                    
+type(dtype=None, non_blocking=False)[source]
 
                    
 
                    Return type
                    
 
                    Union[_StorageBase, TypedStorage]
                    
@@ -1081,7 +1081,7 @@ 
                    Special cases
 
                    
-untyped()[source]
                    
+untyped()[source]
 
                    
 
 
                    
@@ -1116,67 +1116,67 @@ 
                    Legacy Typed Storagetorch.Tensor views.
                    
 
                    
 
                    
-class torch.TypedStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.TypedStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
 
                    
-bfloat16()[source][source]
                    
+bfloat16()[source][source]
 
                    Casts this storage to bfloat16 type.
                    
 
                    
 
 
                    
 
                    
-bool()[source][source]
                    
+bool()[source][source]
 
                    Casts this storage to bool type.
                    
 
                    
 
 
                    
 
                    
-byte()[source][source]
                    
+byte()[source][source]
 
                    Casts this storage to byte type.
                    
 
                    
 
 
                    
 
                    
-char()[source][source]
                    
+char()[source][source]
 
                    Casts this storage to char type.
                    
 
                    
 
 
                    
 
                    
-clone()[source][source]
                    
+clone()[source][source]
 
                    Return a copy of this storage.
                    
 
                    
 
 
                    
 
                    
-complex_double()[source][source]
                    
+complex_double()[source][source]
 
                    Casts this storage to complex double type.
                    
 
                    
 
 
                    
 
                    
-complex_float()[source][source]
                    
+complex_float()[source][source]
 
                    Casts this storage to complex float type.
                    
 
                    
 
 
                    
 
                    
-copy_(source, non_blocking=None)[source][source]
                    
+copy_(source, non_blocking=None)[source][source]
 
                    
 
                    
 
                    
 
 
                    
 
                    
-cpu()[source][source]
                    
+cpu()[source][source]
 
                    Return a CPU copy of this storage if it’s not already on the CPU.
                    
 
                    
 
 
                    
 
                    
-cuda(device=None, non_blocking=False)[source][source]
                    
+cuda(device=None, non_blocking=False)[source][source]
 
                    Returns a copy of this object in CUDA memory.
                    
 
                    If this object is already in CUDA memory and on the correct device, then
 no copy is performed and the original object is returned.
                    
@@ -1197,7 +1197,7 @@ 
                    Legacy Typed Storage
 
                    
-data_ptr()[source][source]
                    
+data_ptr()[source][source]
 
                    
 
 
                    
@@ -1207,7 +1207,7 @@ 
                    Legacy Typed Storage
 
                    
-double()[source][source]
                    
+double()[source][source]
 
                    Casts this storage to double type.
                    
 
                    
 
@@ -1218,7 +1218,7 @@ 
                    Legacy Typed Storage
 
                    
-element_size()[source][source]
                    
+element_size()[source][source]
 
                    
 
 
                    
@@ -1230,47 +1230,47 @@ 
                    Legacy Typed Storage
 
                    
-fill_(value)[source][source]
                    
+fill_(value)[source][source]
 
                    
 
 
                    
 
                    
-float()[source][source]
                    
+float()[source][source]
 
                    Casts this storage to float type.
                    
 
                    
 
 
                    
 
                    
-float8_e4m3fn()[source][source]
                    
+float8_e4m3fn()[source][source]
 
                    Casts this storage to float8_e4m3fn type
                    
 
                    
 
 
                    
 
                    
-float8_e4m3fnuz()[source][source]
                    
+float8_e4m3fnuz()[source][source]
 
                    Casts this storage to float8_e4m3fnuz type
                    
 
                    
 
 
                    
 
                    
-float8_e5m2()[source][source]
                    
+float8_e5m2()[source][source]
 
                    Casts this storage to float8_e5m2 type
                    
 
                    
 
 
                    
 
                    
-float8_e5m2fnuz()[source][source]
                    
+float8_e5m2fnuz()[source][source]
 
                    Casts this storage to float8_e5m2fnuz type
                    
 
                    
 
 
                    
 
                    
-classmethod from_buffer(*args, **kwargs)[source][source]
                    
+classmethod from_buffer(*args, **kwargs)[source][source]
 
                    
 
 
                    
 
                    
-classmethod from_file(filename, shared=False, size=0) → Storage[source][source]
                    
+classmethod from_file(filename, shared=False, size=0) → Storage[source][source]
 
                    Creates a CPU storage backed by a memory-mapped file.
                    
 
                    If shared is True, then memory is shared between all processes.
 All changes are written to the file. If shared is False, then the changes on
@@ -1293,7 +1293,7 @@ 
                    Legacy Typed Storage
 
                    
-get_device()[source][source]
                    
+get_device()[source][source]
 
                    
 
                    Return type
                    
 
                    int
                    
@@ -1303,13 +1303,13 @@ 
                    Legacy Typed Storage
 
                    
-half()[source][source]
                    
+half()[source][source]
 
                    Casts this storage to half type.
                    
 
                    
 
 
                    
 
                    
-hpu(device=None, non_blocking=False)[source][source]
                    
+hpu(device=None, non_blocking=False)[source][source]
 
                    Returns a copy of this object in HPU memory.
                    
 
                    If this object is already in HPU memory and on the correct device, then
 no copy is performed and the original object is returned.
                    
@@ -1330,7 +1330,7 @@ 
                    Legacy Typed Storage
 
                    
-int()[source][source]
                    
+int()[source][source]
 
                    Casts this storage to int type.
                    
 
                    
 
@@ -1346,7 +1346,7 @@ 
                    Legacy Typed Storage
 
                    
-is_pinned(device='cuda')[source][source]
                    
+is_pinned(device='cuda')[source][source]
 
                    Determine whether the CPU TypedStorage is already pinned on device.
                    
 
                    
 
                    Parameters
                    
@@ -1361,7 +1361,7 @@ 
                    Legacy Typed Storage
 
                    
-is_shared()[source][source]
                    
+is_shared()[source][source]
 
                    
 
 
                    
@@ -1371,23 +1371,23 @@ 
                    Legacy Typed Storage
 
                    
-long()[source][source]
                    
+long()[source][source]
 
                    Casts this storage to long type.
                    
 
                    
 
 
                    
 
                    
-nbytes()[source][source]
                    
+nbytes()[source][source]
 
                    
 
 
                    
 
                    
-pickle_storage_type()[source][source]
                    
+pickle_storage_type()[source][source]
 
                    
 
 
                    
 
                    
-pin_memory(device='cuda')[source][source]
                    
+pin_memory(device='cuda')[source][source]
 
                    Copy the CPU TypedStorage to pinned memory, if it’s not already pinned.
                    
 
                    
 
                    Parameters
                    
@@ -1402,34 +1402,34 @@ 
                    Legacy Typed Storage
 
                    
-resizable()[source][source]
                    
+resizable()[source][source]
 
                    
 
 
                    
 
                    
-resize_(size)[source][source]
                    
+resize_(size)[source][source]
 
                    
 
 
                    
 
                    
-share_memory_()[source][source]
                    
+share_memory_()[source][source]
 
                    See torch.UntypedStorage.share_memory_()
                    
 
                    
 
 
                    
 
                    
-short()[source][source]
                    
+short()[source][source]
 
                    Casts this storage to short type.
                    
 
                    
 
 
                    
 
                    
-size()[source][source]
                    
+size()[source][source]
 
                    
 
 
                    
 
                    
-to(*, device, non_blocking=False)[source][source]
                    
+to(*, device, non_blocking=False)[source][source]
 
                    Returns a copy of this object in device memory.
                    
 
                    If this object is already on the correct device, then no copy is performed
 and the original object is returned.
                    
@@ -1450,13 +1450,13 @@ 
                    Legacy Typed Storage
 
                    
-tolist()[source][source]
                    
+tolist()[source][source]
 
                    Return a list containing the elements of this storage.
                    
 
                    
 
 
                    
 
                    
-type(dtype=None, non_blocking=False)[source][source]
                    
+type(dtype=None, non_blocking=False)[source][source]
 
                    Returns the type if dtype is not provided, else casts this object to
 the specified type.
                    
 
                    If this is already of the correct type, no copy is performed and the
@@ -1481,7 +1481,7 @@ 
                    Legacy Typed Storage
 
                    
-untyped()[source][source]
                    
+untyped()[source][source]
 
                    Return the internal torch.UntypedStorage.
                    
 
                    
 
@@ -1489,7 +1489,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.DoubleStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.DoubleStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1501,7 +1501,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.FloatStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.FloatStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1513,7 +1513,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.HalfStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.HalfStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1525,7 +1525,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.LongStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.LongStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1537,7 +1537,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.IntStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.IntStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1549,7 +1549,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.ShortStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.ShortStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1561,7 +1561,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.CharStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.CharStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1573,7 +1573,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.ByteStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.ByteStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1585,7 +1585,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.BoolStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.BoolStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1597,7 +1597,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.BFloat16Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.BFloat16Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1609,7 +1609,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.ComplexDoubleStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.ComplexDoubleStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1621,7 +1621,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.ComplexFloatStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.ComplexFloatStorage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1633,7 +1633,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.QUInt8Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.QUInt8Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1645,7 +1645,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.QInt8Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.QInt8Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1657,7 +1657,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.QInt32Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.QInt32Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1669,7 +1669,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.QUInt4x2Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.QUInt4x2Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
@@ -1681,7 +1681,7 @@ 
                    Legacy Typed Storage
 
                    
-class torch.QUInt2x4Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
                    
+class torch.QUInt2x4Storage(*args, wrap_storage=None, dtype=None, device=None, _internal=False)[source][source]
 
                    
 
                    
 
                    
diff --git a/2.7/tensorboard.html b/2.7/tensorboard.html
index 247d689e3ff..21d2cda1eb1 100644
--- a/2.7/tensorboard.html
+++ b/2.7/tensorboard.html
@@ -653,7 +653,7 @@
 
                    
 
                    
 
                    
-class torch.utils.tensorboard.writer.SummaryWriter(log_dir=None, comment='', purge_step=None, max_queue=10, flush_secs=120, filename_suffix='')[source][source]
                    
+class torch.utils.tensorboard.writer.SummaryWriter(log_dir=None, comment='', purge_step=None, max_queue=10, flush_secs=120, filename_suffix='')[source][source]
 
                    Writes entries directly to event files in the log_dir to be consumed by TensorBoard.
                    
 
                    The SummaryWriter class provides a high-level API to create an event file
 in a given directory and add summaries and events to it. The class updates the
@@ -662,7 +662,7 @@
 training.
                    
 
                    
 
                    
-__init__(log_dir=None, comment='', purge_step=None, max_queue=10, flush_secs=120, filename_suffix='')[source][source]
                    
+__init__(log_dir=None, comment='', purge_step=None, max_queue=10, flush_secs=120, filename_suffix='')[source][source]
 
                    Create a SummaryWriter that will write out events and summaries to the event file.
                    
 
                    
 
                    Parameters
                    
@@ -709,7 +709,7 @@
 
 
                    
 
                    
-add_scalar(tag, scalar_value, global_step=None, walltime=None, new_style=False, double_precision=False)[source][source]
                    
+add_scalar(tag, scalar_value, global_step=None, walltime=None, new_style=False, double_precision=False)[source][source]
 
                    Add scalar data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -739,7 +739,7 @@
 
 
                    
 
                    
-add_scalars(main_tag, tag_scalar_dict, global_step=None, walltime=None)[source][source]
                    
+add_scalars(main_tag, tag_scalar_dict, global_step=None, walltime=None)[source][source]
 
                    Add many scalar data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -771,7 +771,7 @@
 
 
                    
 
                    
-add_histogram(tag, values, global_step=None, bins='tensorflow', walltime=None, max_bins=None)[source][source]
                    
+add_histogram(tag, values, global_step=None, bins='tensorflow', walltime=None, max_bins=None)[source][source]
 
                    Add histogram to summary.
                    
 
                    
 
                    Parameters
                    
@@ -802,7 +802,7 @@
 
 
                    
 
                    
-add_image(tag, img_tensor, global_step=None, walltime=None, dataformats='CHW')[source][source]
                    
+add_image(tag, img_tensor, global_step=None, walltime=None, dataformats='CHW')[source][source]
 
                    Add image data to summary.
                    
 
                    Note that this requires the pillow package.
                    
 
                    
@@ -850,7 +850,7 @@
 
 
                    
 
                    
-add_images(tag, img_tensor, global_step=None, walltime=None, dataformats='NCHW')[source][source]
                    
+add_images(tag, img_tensor, global_step=None, walltime=None, dataformats='NCHW')[source][source]
 
                    Add batched image data to summary.
                    
 
                    Note that this requires the pillow package.
                    
 
                    
@@ -891,7 +891,7 @@
 
 
                    
 
                    
-add_figure(tag, figure, global_step=None, close=True, walltime=None)[source][source]
                    
+add_figure(tag, figure, global_step=None, close=True, walltime=None)[source][source]
 
                    Render matplotlib figure into an image and add it to summary.
                    
 
                    Note that this requires the matplotlib package.
                    
 
                    
@@ -910,7 +910,7 @@
 
 
                    
 
                    
-add_video(tag, vid_tensor, global_step=None, fps=4, walltime=None)[source][source]
                    
+add_video(tag, vid_tensor, global_step=None, fps=4, walltime=None)[source][source]
 
                    Add video data to summary.
                    
 
                    Note that this requires the moviepy package.
                    
 
                    
@@ -933,7 +933,7 @@
 
 
                    
 
                    
-add_audio(tag, snd_tensor, global_step=None, sample_rate=44100, walltime=None)[source][source]
                    
+add_audio(tag, snd_tensor, global_step=None, sample_rate=44100, walltime=None)[source][source]
 
                    Add audio data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -955,7 +955,7 @@
 
 
                    
 
                    
-add_text(tag, text_string, global_step=None, walltime=None)[source][source]
                    
+add_text(tag, text_string, global_step=None, walltime=None)[source][source]
 
                    Add text data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -977,7 +977,7 @@
 
 
                    
 
                    
-add_graph(model, input_to_model=None, verbose=False, use_strict_trace=True)[source][source]
                    
+add_graph(model, input_to_model=None, verbose=False, use_strict_trace=True)[source][source]
 
                    Add graph data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -996,7 +996,7 @@
 
 
                    
 
                    
-add_embedding(mat, metadata=None, label_img=None, global_step=None, tag='default', metadata_header=None)[source][source]
                    
+add_embedding(mat, metadata=None, label_img=None, global_step=None, tag='default', metadata_header=None)[source][source]
 
                    Add embedding projector data to summary.
                    
 
                    
 
                    Parameters
                    
@@ -1045,7 +1045,7 @@
 
 
                    
 
                    
-add_pr_curve(tag, labels, predictions, global_step=None, num_thresholds=127, weights=None, walltime=None)[source][source]
                    
+add_pr_curve(tag, labels, predictions, global_step=None, num_thresholds=127, weights=None, walltime=None)[source][source]
 
                    Add precision recall curve.
                    
 
                    Plotting a precision-recall curve lets you understand your model’s
 performance under different threshold settings. With this function,
@@ -1080,7 +1080,7 @@
 
 
                    
 
                    
-add_custom_scalars(layout)[source][source]
                    
+add_custom_scalars(layout)[source][source]
 
                    Create special chart by collecting charts tags in ‘scalars’.
                    
 
                    NOTE: This function can only be called once for each SummaryWriter() object.
                    
 
                    Because it only provides metadata to tensorboard, the function can be called before or after the training loop.
                    
@@ -1104,7 +1104,7 @@
 
 
                    
 
                    
-add_mesh(tag, vertices, colors=None, faces=None, config_dict=None, global_step=None, walltime=None)[source][source]
                    
+add_mesh(tag, vertices, colors=None, faces=None, config_dict=None, global_step=None, walltime=None)[source][source]
 
                    Add meshes or 3D point clouds to TensorBoard.
                    
 
                    The visualization is based on Three.js,
 so it allows users to interact with the rendered object. Besides the basic definitions
@@ -1162,7 +1162,7 @@
 
 
                    
 
                    
-add_hparams(hparam_dict, metric_dict, hparam_domain_discrete=None, run_name=None, global_step=None)[source][source]
                    
+add_hparams(hparam_dict, metric_dict, hparam_domain_discrete=None, run_name=None, global_step=None)[source][source]
 
                    Add a set of hyperparameters to be compared in TensorBoard.
                    
 
                    
 
                    Parameters
                    
@@ -1198,7 +1198,7 @@
 
 
                    
 
                    
-flush()[source][source]
                    
+flush()[source][source]
 
                    Flushes the event file to disk.
                    
 
                    Call this method to make sure that all pending events have been written to
 disk.
                    
@@ -1206,7 +1206,7 @@
 
 
                    
 
                    
-close()[source][source]
                    
+close()[source][source]
 
                    
 
 
                    
diff --git a/2.7/testing.html b/2.7/testing.html
index 1c6aec5554b..3282e5eba42 100644
--- a/2.7/testing.html
+++ b/2.7/testing.html
@@ -594,7 +594,7 @@
 
                    torch.testing
                    
 
                    
 
                    
-torch.testing.assert_close(actual, expected, *, allow_subclasses=True, rtol=None, atol=None, equal_nan=False, check_device=True, check_dtype=True, check_layout=True, check_stride=False, msg=None)[source][source]
                    
+torch.testing.assert_close(actual, expected, *, allow_subclasses=True, rtol=None, atol=None, equal_nan=False, check_device=True, check_dtype=True, check_layout=True, check_stride=False, msg=None)[source][source]
 
                    Asserts that actual and expected are close.
                    
 
                    If actual and expected are strided, non-quantized, real-valued, and finite, they are considered close if
                    
 
                    
@@ -866,7 +866,7 @@
 
 
                    
 
                    
-torch.testing.make_tensor(*shape, dtype, device, low=None, high=None, requires_grad=False, noncontiguous=False, exclude_zero=False, memory_format=None)[source][source]
                    
+torch.testing.make_tensor(*shape, dtype, device, low=None, high=None, requires_grad=False, noncontiguous=False, exclude_zero=False, memory_format=None)[source][source]
 
                    Creates a tensor with the given shape, device, and dtype, and filled with
 values uniformly drawn from [low, high).
                    
 
                    If low or high are specified and are outside the range of the dtype’s representable
@@ -961,7 +961,7 @@
 
 
                    
 
                    
-torch.testing.assert_allclose(actual, expected, rtol=None, atol=None, equal_nan=True, msg='')[source][source]
                    
+torch.testing.assert_allclose(actual, expected, rtol=None, atol=None, equal_nan=True, msg='')[source][source]
 
                    
 
                    Warning
                    
 
                    torch.testing.assert_allclose() is deprecated since 1.12 and will be removed in a future release.
diff --git a/2.7/torch.ao.ns._numeric_suite.html b/2.7/torch.ao.ns._numeric_suite.html
index 88f285da59e..d59d2517286 100644
--- a/2.7/torch.ao.ns._numeric_suite.html
+++ b/2.7/torch.ao.ns._numeric_suite.html
@@ -602,7 +602,7 @@
 
                    
 
                    
 
                    
-torch.ao.ns._numeric_suite.compare_weights(float_dict, quantized_dict)[source][source]
                    
+torch.ao.ns._numeric_suite.compare_weights(float_dict, quantized_dict)[source][source]
 
                    Compare the weights of the float module with its corresponding quantized
 module. Return a dict with key corresponding to module names and each entry being
 a dictionary with two keys ‘float’ and ‘quantized’, containing the float and
@@ -641,7 +641,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.get_logger_dict(mod, prefix='')[source][source]
                    
+torch.ao.ns._numeric_suite.get_logger_dict(mod, prefix='')[source][source]
 
                    Traverse the modules and save all logger stats into target dict.
 This is mainly used for quantization accuracy debug.
                    
 
                    
@@ -667,41 +667,41 @@
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite.Logger[source][source]
                    
+class torch.ao.ns._numeric_suite.Logger[source][source]
 
                    Base class for stats logging
                    
 
                    
 
                    
-forward(x)[source][source]
                    
+forward(x)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite.ShadowLogger[source][source]
                    
+class torch.ao.ns._numeric_suite.ShadowLogger[source][source]
 
                    Class used in Shadow module to record the outputs of the original and
 shadow modules.
                    
 
                    
 
                    
-forward(x, y)[source][source]
                    
+forward(x, y)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite.OutputLogger[source][source]
                    
+class torch.ao.ns._numeric_suite.OutputLogger[source][source]
 
                    Class used to log the outputs of the module
                    
 
                    
 
                    
-forward(x)[source][source]
                    
+forward(x)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite.Shadow(q_module, float_module, logger_cls)[source][source]
                    
+class torch.ao.ns._numeric_suite.Shadow(q_module, float_module, logger_cls)[source][source]
 
                    Shadow module attaches the float module to its matching quantized module
 as the shadow. Then it uses Logger module to process the outputs of both
 modules.
                    
@@ -717,7 +717,7 @@
 
                    
 
                    
 
                    
-forward(*x)[source][source]
                    
+forward(*x)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -727,7 +727,7 @@
 
 
                    
 
                    
-add(x, y)[source][source]
                    
+add(x, y)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -737,7 +737,7 @@
 
 
                    
 
                    
-add_scalar(x, y)[source][source]
                    
+add_scalar(x, y)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -747,7 +747,7 @@
 
 
                    
 
                    
-mul(x, y)[source][source]
                    
+mul(x, y)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -757,7 +757,7 @@
 
 
                    
 
                    
-mul_scalar(x, y)[source][source]
                    
+mul_scalar(x, y)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -767,7 +767,7 @@
 
 
                    
 
                    
-cat(x, dim=0)[source][source]
                    
+cat(x, dim=0)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -777,7 +777,7 @@
 
 
                    
 
                    
-add_relu(x, y)[source][source]
                    
+add_relu(x, y)[source][source]
 
                    
 
                    Return type
                    
 
                    Tensor
                    
@@ -789,7 +789,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.prepare_model_with_stubs(float_module, q_module, module_swap_list, logger_cls)[source][source]
                    
+torch.ao.ns._numeric_suite.prepare_model_with_stubs(float_module, q_module, module_swap_list, logger_cls)[source][source]
 
                    Prepare the model by attaching the float module to its matching quantized
 module as the shadow if the float module type is in module_swap_list.
                    
 
                    Example usage:
                    
@@ -813,7 +813,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.compare_model_stub(float_model, q_model, module_swap_list, *data, logger_cls=<class 'torch.ao.ns._numeric_suite.ShadowLogger'>)[source][source]
                    
+torch.ao.ns._numeric_suite.compare_model_stub(float_model, q_model, module_swap_list, *data, logger_cls=<class 'torch.ao.ns._numeric_suite.ShadowLogger'>)[source][source]
 
                    Compare quantized module in a model with its floating point counterpart,
 feeding both of them the same input. Return a dict with key corresponding to
 module names and each entry being a dictionary with two keys ‘float’ and
@@ -854,7 +854,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.get_matching_activations(float_module, q_module)[source][source]
                    
+torch.ao.ns._numeric_suite.get_matching_activations(float_module, q_module)[source][source]
 
                    Find the matching activation between float and quantized modules.
                    
 
                    
 
                    Parameters
                    
@@ -876,7 +876,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.prepare_model_outputs(float_module, q_module, logger_cls=<class 'torch.ao.ns._numeric_suite.OutputLogger'>, allow_list=None)[source][source]
                    
+torch.ao.ns._numeric_suite.prepare_model_outputs(float_module, q_module, logger_cls=<class 'torch.ao.ns._numeric_suite.OutputLogger'>, allow_list=None)[source][source]
 
                    Prepare the model by attaching the logger to both float module
 and quantized module if they are in the allow_list.
                    
 
                    
@@ -893,7 +893,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite.compare_model_outputs(float_model, q_model, *data, logger_cls=<class 'torch.ao.ns._numeric_suite.OutputLogger'>, allow_list=None)[source][source]
                    
+torch.ao.ns._numeric_suite.compare_model_outputs(float_model, q_model, *data, logger_cls=<class 'torch.ao.ns._numeric_suite.OutputLogger'>, allow_list=None)[source][source]
 
                    Compare output activations between float and quantized models at
 corresponding locations for the same input. Return a dict with key corresponding
 to quantized module names and each entry being a dictionary with two keys
diff --git a/2.7/torch.ao.ns._numeric_suite_fx.html b/2.7/torch.ao.ns._numeric_suite_fx.html
index ca36357be48..96d75576965 100644
--- a/2.7/torch.ao.ns._numeric_suite_fx.html
+++ b/2.7/torch.ao.ns._numeric_suite_fx.html
@@ -685,39 +685,39 @@
 
                    
 
                    
 
                    
-class torch.ao.ns._numeric_suite_fx.OutputLogger(ref_node_name, prev_node_name, model_name, ref_name, prev_node_target_type, ref_node_target_type, results_type, index_within_arg, index_of_arg, fqn, qconfig_str='')[source][source]
                    
+class torch.ao.ns._numeric_suite_fx.OutputLogger(ref_node_name, prev_node_name, model_name, ref_name, prev_node_target_type, ref_node_target_type, results_type, index_within_arg, index_of_arg, fqn, qconfig_str='')[source][source]
 
                    Base class for capturing intermediate values.
                    
 
                    
 
                    
 
                    
 
                    
-forward(x)[source][source]
                    
+forward(x)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite_fx.OutputComparisonLogger(*args, **kwargs)[source][source]
                    
+class torch.ao.ns._numeric_suite_fx.OutputComparisonLogger(*args, **kwargs)[source][source]
 
                    Same as OutputLogger, but also requires the original activation
 in order to calculate the comparison at calibration time
                    
 
                    
 
                    
-forward(x, x_ref)[source][source]
                    
+forward(x, x_ref)[source][source]
 
                    
 
 
                    
 
 
                    
 
                    
-class torch.ao.ns._numeric_suite_fx.NSTracer(skipped_module_names, skipped_module_classes)[source][source]
                    
+class torch.ao.ns._numeric_suite_fx.NSTracer(skipped_module_names, skipped_module_classes)[source][source]
 
                    Just like a regular FX quantization tracer, but treats observers and fake_quantize
 modules as leaf modules.
                    
 
                    
 
                    
 
                    
 
                    
-is_leaf_module(m, module_qualified_name)[source][source]
                    
+is_leaf_module(m, module_qualified_name)[source][source]
 
                    
 
                    Return type
                    
 
                    bool
                    
@@ -729,7 +729,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.extract_weights(model_name_a, model_a, model_name_b, model_b, base_name_to_sets_of_related_ops=None, unmatchable_types_map=None, op_to_type_to_weight_extraction_fn=None)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.extract_weights(model_name_a, model_a, model_name_b, model_b, base_name_to_sets_of_related_ops=None, unmatchable_types_map=None, op_to_type_to_weight_extraction_fn=None)[source][source]
 
                    Extract weights from model A and model B, and return a comparison.
                    
 
                    
 
                    Parameters
                    
@@ -755,7 +755,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.add_loggers(name_a, model_a, name_b, model_b, logger_cls, should_log_inputs=False, base_name_to_sets_of_related_ops=None, unmatchable_types_map=None)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.add_loggers(name_a, model_a, name_b, model_b, logger_cls, should_log_inputs=False, base_name_to_sets_of_related_ops=None, unmatchable_types_map=None)[source][source]
 
                    Instrument model A and model B with loggers.
                    
 
                    
 
                    Parameters
                    
@@ -780,7 +780,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.extract_logger_info(model_a, model_b, logger_cls, model_name_to_use_for_layer_names)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.extract_logger_info(model_a, model_b, logger_cls, model_name_to_use_for_layer_names)[source][source]
 
                    Traverse all loggers in model_a and model_b, and extract the logged
 information.
                    
 
                    
@@ -804,7 +804,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.add_shadow_loggers(name_a, model_a, name_b, model_b, logger_cls, should_log_inputs=False, base_name_to_sets_of_related_ops=None, node_type_to_io_type_map=None, unmatchable_types_map=None)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.add_shadow_loggers(name_a, model_a, name_b, model_b, logger_cls, should_log_inputs=False, base_name_to_sets_of_related_ops=None, node_type_to_io_type_map=None, unmatchable_types_map=None)[source][source]
 
                    Instrument model A and model B with shadow loggers.
                    
 
                    
 
                    Parameters
                    
@@ -827,7 +827,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.extract_shadow_logger_info(model_a_shadows_b, logger_cls, model_name_to_use_for_layer_names)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.extract_shadow_logger_info(model_a_shadows_b, logger_cls, model_name_to_use_for_layer_names)[source][source]
 
                    Traverse all loggers in a shadow model, and extract the logged
 information.
                    
 
                    
@@ -850,7 +850,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.extend_logger_results_with_comparison(results, model_name_1, model_name_2, comparison_fn, comparison_name)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.extend_logger_results_with_comparison(results, model_name_1, model_name_2, comparison_fn, comparison_name)[source][source]
 
                    Compares the logged values from model_name_2 against the corresponding
 values in model_name_1, using comparison_fn. Records the result
 in model_name_2’s results under comparison_name. Modifies results inplace.
                    
@@ -871,7 +871,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.prepare_n_shadows_model(model, example_inputs, qconfig_multi_mapping, backend_config, custom_prepare_fn=None, custom_prepare_kwargs=None, custom_tracer=None)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.prepare_n_shadows_model(model, example_inputs, qconfig_multi_mapping, backend_config, custom_prepare_fn=None, custom_prepare_kwargs=None, custom_tracer=None)[source][source]
 
                    Given a model with a graph with M ops such as
                    
 
                    
 
                    args_kwargs_m -> op_m -> output_m
                    
@@ -905,7 +905,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.loggers_set_enabled(model, enabled)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.loggers_set_enabled(model, enabled)[source][source]
 
                    Sets the enabled setting on a model’s loggers
                    
 
                    
 
                    
@@ -913,7 +913,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.loggers_set_save_activations(model, save_activations)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.loggers_set_save_activations(model, save_activations)[source][source]
 
                    Sets the save_activations setting on a model’s loggers
                    
 
                    
 
                    
@@ -921,7 +921,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.convert_n_shadows_model(model, custom_convert_fn=None, custom_convert_kwargs=None)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.convert_n_shadows_model(model, custom_convert_fn=None, custom_convert_kwargs=None)[source][source]
 
                    Given a model from prepare_n_shadows_model, runs convert_fx
 on each shadow submodule.
                    
 
                    
@@ -933,7 +933,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.extract_results_n_shadows_model(model)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.extract_results_n_shadows_model(model)[source][source]
 
                    Extracts logger results from model.
                    
 
                    
 
                    Return type
                    
@@ -944,7 +944,7 @@
 
 
                    
 
                    
-torch.ao.ns._numeric_suite_fx.print_comparisons_n_shadows_model(results)[source][source]
                    
+torch.ao.ns._numeric_suite_fx.print_comparisons_n_shadows_model(results)[source][source]
 
                    Prints a summary of extracted results.
                    
 
                    
 
                    
@@ -959,17 +959,17 @@ 
                    torch.ao.ns.fx.utils
 
                    
 
                    
-torch.ao.ns.fx.utils.compute_sqnr(x, y)[source][source]
                    
+torch.ao.ns.fx.utils.compute_sqnr(x, y)[source][source]
 
                    
 
 
                    
 
                    
-torch.ao.ns.fx.utils.compute_normalized_l2_error(x, y)[source][source]
                    
+torch.ao.ns.fx.utils.compute_normalized_l2_error(x, y)[source][source]
 
                    
 
 
                    
 
                    
-torch.ao.ns.fx.utils.compute_cosine_similarity(x, y)[source][source]
                    
+torch.ao.ns.fx.utils.compute_cosine_similarity(x, y)[source][source]
 
                    
 
 
                    
diff --git a/2.7/torch.compiler_aot_inductor.html b/2.7/torch.compiler_aot_inductor.html
index e409f644790..ecd419e606a 100644
--- a/2.7/torch.compiler_aot_inductor.html
+++ b/2.7/torch.compiler_aot_inductor.html
@@ -785,7 +785,7 @@ 
                    Troubleshooting
                    
 
                    
 
                    
-torch._inductor.aoti_compile_and_package(exported_program, _deprecated_unused_args=None, _deprecated_unused_kwargs=None, *, package_path=None, inductor_configs=None)[source][source]
                    
+torch._inductor.aoti_compile_and_package(exported_program, _deprecated_unused_args=None, _deprecated_unused_kwargs=None, *, package_path=None, inductor_configs=None)[source][source]
 
                    Compiles the exported program with AOTInductor, and packages it into a .pt2
 artifact specified by the input package_path. To load the package, you can
 call torch._inductor.aoti_load_package(package_path).
                    
@@ -835,7 +835,7 @@ 
                    API Reference
 
                    
-torch._inductor.aoti_load_package(path, run_single_threaded=False)[source][source]
                    
+torch._inductor.aoti_load_package(path, run_single_threaded=False)[source][source]
 
                    Loads the model from the PT2 package.
                    
 
                    If multiple models were packaged into the PT2, this will load the default
 model. To load a specific model, you can directly call the load API
                    
diff --git a/2.7/torch.html b/2.7/torch.html
index 074dd6bef22..0221b65ce7c 100644
--- a/2.7/torch.html
+++ b/2.7/torch.html
@@ -2251,13 +2251,13 @@ 
                    Utilities
                    
 
                    
 
                    
-class torch.SymInt(node)[source][source]
                    
+class torch.SymInt(node)[source][source]
 
                    Like an int (including magic methods), but redirects all operations on the
 wrapped node. This is used in particular to symbolically record operations
 in the symbolic shape workflow.
                    
 
                    
 
                    
-as_integer_ratio()[source][source]
                    
+as_integer_ratio()[source][source]
 
                    Represent this int as an exact integer ratio
                    
 
                    
 
                    Return type
                    
@@ -2270,13 +2270,13 @@ 
                    Symbolic Numbers
 
                    
-class torch.SymFloat(node)[source][source]
                    
+class torch.SymFloat(node)[source][source]
 
                    Like an float (including magic methods), but redirects all operations on the
 wrapped node. This is used in particular to symbolically record operations
 in the symbolic shape workflow.
                    
 
                    
 
                    
-as_integer_ratio()[source][source]
                    
+as_integer_ratio()[source][source]
 
                    Represent this float as an exact integer ratio
                    
 
                    
 
                    Return type
                    
@@ -2287,7 +2287,7 @@ 
                    Symbolic Numbers
 
                    
-conjugate()[source][source]
                    
+conjugate()[source][source]
 
                    Returns the complex conjugate of the float.
                    
 
                    
 
                    Return type
                    
@@ -2298,7 +2298,7 @@ 
                    Symbolic Numbers
 
                    
-hex()[source][source]
                    
+hex()[source][source]
 
                    Returns the hexadecimal representation of the float.
                    
 
                    
 
                    Return type
                    
@@ -2309,7 +2309,7 @@ 
                    Symbolic Numbers
 
                    
-is_integer()[source][source]
                    
+is_integer()[source][source]
 
                    Return True if the float is an integer.
                    
 
                    
 
@@ -2317,7 +2317,7 @@ 
                    Symbolic Numbers
 
                    
-class torch.SymBool(node)[source][source]
                    
+class torch.SymBool(node)[source][source]
 
                    Like an bool (including magic methods), but redirects all operations on the
 wrapped node. This is used in particular to symbolically record operations
 in the symbolic shape workflow.
                    
diff --git a/2.7/torch.overrides.html b/2.7/torch.overrides.html
index 812259ab6e5..d5f4c69be21 100644
--- a/2.7/torch.overrides.html
+++ b/2.7/torch.overrides.html
@@ -599,7 +599,7 @@
 
                    Functions
                    
 
                    
 
                    
-torch.overrides.get_ignored_functions()[source][source]
                    
+torch.overrides.get_ignored_functions()[source][source]
 
                    Return public functions that cannot be overridden by __torch_function__.
                    
 
                    
 
                    Returns
                    
@@ -622,7 +622,7 @@ 
                    Functions
 
                    
-torch.overrides.get_overridable_functions()[source][source]
                    
+torch.overrides.get_overridable_functions()[source][source]
 
                    List functions that are overridable via __torch_function__
                    
 
                    
 
                    Returns
                    
@@ -637,7 +637,7 @@ 
                    Functions
 
                    
-torch.overrides.resolve_name(f)[source][source]
                    
+torch.overrides.resolve_name(f)[source][source]
 
                    Get a human readable string name for a function passed to
 __torch_function__
                    
 
                    
@@ -656,7 +656,7 @@ 
                    Functions
 
                    
-torch.overrides.get_testing_overrides()[source][source]
                    
+torch.overrides.get_testing_overrides()[source][source]
 
                    Return a dict containing dummy overrides for all overridable functions
                    
 
                    
 
                    Returns
                    
@@ -680,7 +680,7 @@ 
                    Functions
 
                    
-torch.overrides.handle_torch_function(public_api, relevant_args, *args, **kwargs)[source][source]
                    
+torch.overrides.handle_torch_function(public_api, relevant_args, *args, **kwargs)[source][source]
 
                    Implement a function with checks for __torch_function__ overrides.
                    
 
                    See torch::autograd::handle_torch_function for the equivalent of this
 function in the C++ implementation.
                    
@@ -743,7 +743,7 @@ 
                    Functions
 
                    
-torch.overrides.is_tensor_like(inp)[source][source]
                    
+torch.overrides.is_tensor_like(inp)[source][source]
 
                    Returns True if the passed-in input is a Tensor-like.
                    
 
                    Currently, this occurs whenever there’s a __torch_function__
 attribute on the type of the input.
                    
@@ -777,7 +777,7 @@ 
                    Functions
 
                    
-torch.overrides.is_tensor_method_or_property(func)[source][source]
                    
+torch.overrides.is_tensor_method_or_property(func)[source][source]
 
                    Returns True if the function passed in is a handler for a
 method or property belonging to torch.Tensor, as passed
 into __torch_function__.
                    
@@ -807,7 +807,7 @@ 
                    Functions
 
                    
-torch.overrides.wrap_torch_function(dispatcher)[source][source]
                    
+torch.overrides.wrap_torch_function(dispatcher)[source][source]
 
                    Wraps a given function with __torch_function__ -related functionality.
                    
 
                    
 
                    Parameters
                    
diff --git a/2.7/torch_cuda_memory.html b/2.7/torch_cuda_memory.html
index ed63fec9f86..d726286df10 100644
--- a/2.7/torch_cuda_memory.html
+++ b/2.7/torch_cuda_memory.html
@@ -638,7 +638,7 @@ 
                    Allocator State HistorySnapshot API Reference
                    
 
                    
 
                    
-torch.cuda.memory._record_memory_history(enabled='all', context='all', stacks='all', max_entries=9223372036854775807, device=None)[source][source]
                    
+torch.cuda.memory._record_memory_history(enabled='all', context='all', stacks='all', max_entries=9223372036854775807, device=None)[source][source]
 
                    Enable recording of stack traces associated with memory
 allocations, so you can tell what allocated any piece of memory in
 torch.cuda.memory._snapshot().
                    
@@ -675,7 +675,7 @@ 
                    Snapshot API Reference
 
                    
-torch.cuda.memory._snapshot(device=None)[source][source]
                    
+torch.cuda.memory._snapshot(device=None)[source][source]
 
                    Save a snapshot of CUDA memory state at the time it was called.
                    
 
                    The state is represented as a dictionary with the following structure.
                    
 
                    class Snapshot(TypedDict):
@@ -752,7 +752,7 @@ 
                    Snapshot API Reference
 
                    
-torch.cuda.memory._dump_snapshot(filename='dump_snapshot.pickle')[source][source]
                    
+torch.cuda.memory._dump_snapshot(filename='dump_snapshot.pickle')[source][source]
 
                    Save a pickled version of the torch.memory._snapshot() dictionary to a file.
                    
 
                    This file can be opened by the interactive snapshot viewer at pytorch.org/memory_viz
                    
 
                    
diff --git a/replace_github_links.sh b/replace_github_links.sh
new file mode 100755
index 00000000000..7ca5e217f84
--- /dev/null
+++ b/replace_github_links.sh
@@ -0,0 +1,18 @@
+#!/bin/bash
+# Replaces GitHub links from v2.7.1 to v2.7.0 in all html files in a directory
+#
+# Usage:
+# ./replace_github_links.sh directory
+#
+# Example (from the root directory)
+# ./replace_github_links.sh 2.7
+
+if [ "$1" == "" ]; then
+  echo "Incorrect usage. Correct Usage: replace_github_links.sh "
+  exit 1
+fi
+
+find $1 -name "*.html" -print0 | xargs -0 sed -i 's|github.com/pytorch/pytorch/blob/v2.7.1|github.com/pytorch/pytorch/blob/v2.7.0|g'
+
+echo "Replaced v2.7.1 with v2.7.0 in GitHub links in $1 directory"
+