8000 Cannot Convert Pytorch model with fft_rfftn layers to ONNX using latest torch.onnx.dynamo_export · Issue #133785 · pytorch/pytorch · GitHub
[go: up one dir, main page]
Skip to content

Cannot Convert Pytorch model with fft_rfftn layers to ONNX using latest torch.onnx.dynamo_export #133785

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
K-prog opened this issue Aug 17, 2024 · 2 comments
Open

Cannot Convert Pytorch model with fft_rfftn layers to ONNX using latest torch.onnx.dynamo_export #133785

K-prog opened this issue Aug 17, 2024 · 2 comments
Assignees
@justinchuby
Labels
module: onnx Related to torch.onnx triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Milestone
2.8.0

Comments

@K-prog
Copy link
K-prog commented Aug 17, 2024 •
edited by justinchuby
Loading

🐛 Describe the bug

I am trying to convert the well renowned LAMA Inpainting model to ONNX via the new dynamo_export, as earlier these fft_rfftn were not supported by onnx there wasn't a way to export this model to ONNX, although there was a workaround by adding a custom FourierUnitJIT class which made it compatible with older torch.onnx.export itself but has accuracy losses.
Conversion Notebook of older approach.

As now we do have the support of these layers in onnx and dynamo_export allows us to export these, there is still some issue

Minimum Reproducible Code:

!curl -LJO "https://github.com/Sanster/models/releases/download/add_big_lama/big-lama.pt"
import torch
model = torch.jit.load("big-lama.pt", map_location="cpu")

dummy_image = torch.randn(1, 3, 512, 512)
dummy_mask = torch.randint(0, 2, (1, 1, 512, 512)).float()

onnx_program = torch.onnx.dynamo_export(model, (dummy_image, dummy_mask))

Torch ONNX Dynamo gives an error stating
OnnxExporterError: Failed to export the model to ONNX. Generating SARIF report at 'report_dynamo_export.sarif'. SARIF is a standard format for the output of static analysis tools
SARIF REPORT

torch.onnx.export(model, (dummy_image, dummy_mask), "lama_inpainting.onnx", 
                    opset_version = 18,
                    input_names=["image", "mask"], 
                    output_names=["output"],
                    dynamic_axes={'image': {0: 'batch_size'},
                                'mask': {0: 'batch_size'},
                                'output': {0: 'batch_size'}})

Tried using the old onnx exporter but obviously it wont work as stated in #107588 and the code errors out stating 'aten::fft_rfftn' is not supported as expected

Versions

Collecting environment information...
PyTorch version: 2.4.0+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.4 LTS (x86_64)
GCC version: Could not collect
Clang version: Could not collect
CMake version: Could not collect
Libc version: glibc-2.35

Python version: 3.11.9 | packaged by conda-forge | (main, Apr 19 2024, 18:36:13) [GCC 12.3.0] (64-bit runtime)
Python platform: Linux-6.8.0-40-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration: GPU 0: NVIDIA L4
Nvidia driver version: 535.183.01
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 43 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 128
On-line CPU(s) list: 0-127
Vendor ID: AuthenticAMD
Model name: AMD EPYC 7742 64-Core Processor
CPU family: 23
Model: 49
Thread(s) per core: 2
Core(s) per socket: 64
Socket(s): 1
Stepping: 0
Frequency boost: enabled
CPU max MHz: 2250.0000
CPU min MHz: 1500.0000
BogoMIPS: 4491.48
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sev sev_es
Virtualization: AMD-V
L1d cache: 2 MiB (64 instances)
L1i cache: 2 MiB (64 instances)
L2 cache: 32 MiB (64 instances)
L3 cache: 256 MiB (16 instances)
NUMA node(s): 1
NUMA node0 CPU(s): 0-127
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Mitigation; untrained return thunk; SMT enabled with STIBP protection
Vulnerability Spec rstack overflow: Mitigation; Safe RET
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Retpolines; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected

Versions of relevant libraries:
[pip3] numpy==1.26.4
[pip3] onnx==1.16.2
[pip3] onnxscript==0.1.0.dev20240817
[pip3] torch==2.4.0
[pip3] torchvision==0.19.0
[pip3] triton==3.0.0
[conda] numpy 1.26.4 pypi_0 pypi
[conda] torchvision 0.19.0 pypi_0 pypi
@K-prog K-prog mentioned this issue Aug 17, 2024
Open
@malfet malfet added module: onnx Related to torch.onnx triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module labels Aug 19, 2024
@justinchuby justinchuby self-assigned this Aug 31, 2024
@justinchuby
Copy link
Collaborator

Thank you. I will be working on this in the coming week or so.

@justinchuby justinchuby added this to the 2.5.0 milestone Sep 1, 2024
@kit1980
Copy link
Contributor
kit1980 commented Sep 26, 2024

@justinchuby is there a fix already?
If not, move to the next release milestone?

@justinchuby justinchuby modified the milestones: 2.5.0, 2.6.0 Sep 26, 2024
@justinchuby justinchuby mentioned this issue Sep 27, 2024
Closed
@justinchuby justinchuby modified the milestones: 2.6.0, 2.7.0 Nov 24, 2024
@justinchuby justinchuby modified the milestones: 2.7.0, 2.8.0 Mar 13, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@justinchuby justinchuby

Labels
module: onnx Related to torch.onnx triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Projects
None yet
Milestone
2.8.0
Development

No branches or pull requests
4 participants
@kit1980 @malfet @justinchuby @K-prog
0