biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Kernel] Add NVFP4 MoE CUTLASS support for SM120 (#29242)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
This commit is contained in:
Michael Goin
2025-11-25 09:59:07 -05:00
committed by GitHub
parent dbc3d9991a
commit e502098643
8 changed files with 264 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/design/moe_kernel_features.md
View File

@@ -60,7 +60,7 @@ Modular kernels are supported by the following `FusedMoEMethodBase` classes.
- [`ModelOptFp8MoEMethod`][vllm.model_executor.layers.quantization.modelopt.ModelOptFp8MoEMethod]
- [`Fp8MoEMethod`][vllm.model_executor.layers.quantization.fp8.Fp8MoEMethod]
- [`CompressedTensorsW4A4MoeMethod`][vllm.model_executor.layers.quantization.compressed_tensors.compressed_tensors_moe.CompressedTensorsW4A4MoeMethod]
- [`CompressedTensorsW4A4Nvfp4MoeMethod`][vllm.model_executor.layers.quantization.compressed_tensors.compressed_tensors_moe.CompressedTensorsW4A4Nvfp4MoeMethod]
- [`CompressedTensorsW8A8Fp8MoEMethod`][vllm.model_executor.layers.quantization.compressed_tensors.compressed_tensors_moe.CompressedTensorsW8A8Fp8MoEMethod]
- [`Mxfp4MoEMethod`][vllm.model_executor.layers.quantization.mxfp4.Mxfp4MoEMethod]
- [`UnquantizedFusedMoEMethod`][vllm.model_executor.layers.fused_moe.layer.UnquantizedFusedMoEMethod]