kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
4ef60d5fbb
TensorRT-LLMs/cpp/tensorrt_llm/kernels/trtllmGenKernels/batchedGemm
History
Dom Brown 44fb3c1673
[TRTLLM-5770] feat: Integrate TRT-LLM Gen FP8 block scale MoE with Pytorch workflow kernel autotuner (#5207) 
- Adds a new Python custom op (fp8_block_scale_moe_runner) and a FP8BlockScaleMoERunner class for autotuning.
- Updates C++ MoE and batched GEMM kernels to accept a configIndex for workspace sizing and execution.
- Extends the unit test to run both autotuned and non-autotuned code paths.

Signed-off-by: Dom Brown <3886319+DomBrown@users.noreply.github.com>
2025-06-17 21:01:56 +08:00
..
trtllmGen_bmm_export feat: MoE trtllm backend kernel update (#5183) 2025-06-16 14:46:13 +08:00
CMakeLists.txt feat: TRT-LLM Gen integration for BMM and MoE refactoring (#4280) 2025-05-16 13:31:53 +02:00
KernelRunner.cpp [TRTLLM-5770] feat: Integrate TRT-LLM Gen FP8 block scale MoE with Pytorch workflow kernel autotuner (#5207) 2025-06-17 21:01:56 +08:00
KernelRunner.h [TRTLLM-5770] feat: Integrate TRT-LLM Gen FP8 block scale MoE with Pytorch workflow kernel autotuner (#5207) 2025-06-17 21:01:56 +08:00