Update llama.cpp

abetlen · abetlen · commit 3720c739d4bf · 2023-09-29T19:58:21.000-04:00
diff --git a/llama_cpp/llama_cpp.py b/llama_cpp/llama_cpp.py
@@ -301,19 +301,19 @@ class llama_model_params(Structure):
 
 # struct llama_context_params {
 #     uint32_t seed;            // RNG seed, -1 for random
-#     uint32_t n_ctx;           // text context
-#     uint32_t n_batch;         // prompt processing batch size
+#     uint32_t n_ctx;           // text context, 0 = from model
+#     uint32_t n_batch;         // prompt processing maximum batch size
 #     uint32_t n_threads;       // number of threads to use for generation
 #     uint32_t n_threads_batch; // number of threads to use for batch processing
 
 #     // ref: https://github.com/ggerganov/llama.cpp/pull/2054
-#     float rope_freq_base;  // RoPE base frequency
-#     float rope_freq_scale; // RoPE frequency scaling factor
+#     float rope_freq_base;  // RoPE base frequency, 0 = from model
+#     float rope_freq_scale; // RoPE frequency scaling factor, 0 = from model
 
 
 #     // Keep the booleans together to avoid misalignment during copy-by-value.
 #     bool mul_mat_q;  // if true, use experimental mul_mat_q kernels
-#     bool f16_kv;     // use fp16 for KV cache
+#     bool f16_kv;     // use fp16 for KV cache, fp32 otherwise
 #     bool logits_all; // the llama_eval() call computes all logits, not just the last one
 #     bool embedding;  // embedding mode only
 # };
diff --git a/vendor/llama.cpp b/vendor/llama.cpp
@@ -1 +1 @@
-Subproject commit bc39553c901a91cfcb757863586250838c83eeab
+Subproject commit 40e07a60f9ce06e79f3ccd4c903eba300fb31b5e
