Xiaoyu Zhang 44e86480e8 fuse allreduce and residual_rmsnorm (#8731)

2025-08-11 13:50:53 -07:00

..

configs

[perf] add kimi-k2 b200 fused moe config (#9010)

2025-08-09 12:40:49 -07:00

__init__.py

Refactor: move all quantization-related code to srt/layer/quantization (#7989)

2025-07-17 00:47:07 -07:00

fused_moe.py

Expert Parallelism for GPT-OSS (#8944)

2025-08-08 00:46:42 -07:00

layer.py

fuse allreduce and residual_rmsnorm (#8731)

2025-08-11 13:50:53 -07:00

triton_kernels_moe.py

Support mxfp4 for GPT-OSS (#8843)

2025-08-06 00:05:25 -07:00