Jinyang Yuan 6b583f6f83 perf: Enable CUDA graphs when attention DP is used and active requests on different GPUs are uneven (#3010) ... Signed-off-by: Jinyang Yuan <154768711+jinyangyuan-nvidia@users.noreply.github.com> Co-authored-by: raccoonliukai <raccoonliu@tencent.com>		2025-03-26 21:09:25 +08:00
..
__init__.py	Update TensorRT-LLM (#2755)	2025-02-11 03:01:00 +00:00
_util.py	Refactor imports inside tensorrt_llm._torch. (#3015)	2025-03-26 11:01:07 +08:00
config.py	Update (#2978)	2025-03-23 16:39:35 +08:00
cuda_graph_runner.py	Refactor imports inside tensorrt_llm._torch. (#3015)	2025-03-26 11:01:07 +08:00
decoder.py	Refactor imports inside tensorrt_llm._torch. (#3015)	2025-03-26 11:01:07 +08:00
distributed.py	Only gather responses on rank 0 (#3040)	2025-03-24 21:54:51 -07:00
guided_decoder.py	Update (#2978)	2025-03-23 16:39:35 +08:00
kv_cache_transceiver.py	Update TensorRT-LLM (#2849)	2025-03-04 18:44:00 +08:00
layerwise_nvtx_marker.py	Update TensorRT-LLM (#2849)	2025-03-04 18:44:00 +08:00
llm_request.py	Update (#2978)	2025-03-23 16:39:35 +08:00
model_engine.py	fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP (#3053)	2025-03-26 11:32:57 +08:00
py_executor_creator.py	perf: Enable CUDA graphs when attention DP is used and active requests on different GPUs are uneven (#3010)	2025-03-26 21:09:25 +08:00
py_executor.py	perf: Enable CUDA graphs when attention DP is used and active requests on different GPUs are uneven (#3010)	2025-03-26 21:09:25 +08:00
resource_manager.py	perf: Enable CUDA graphs when attention DP is used and active requests on different GPUs are uneven (#3010)	2025-03-26 21:09:25 +08:00
scheduler.py	Update TensorRT-LLM (#2936)	2025-03-18 21:25:19 +08:00