A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[Docs] Fix broken table in moe_kernel_features doc (#26314)
Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com> Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
Varun Sundar Rabindranath committed
Oct 6, 2025 at 15:58 UTC
93540958b8fd1c5bbe3382727e494fbac57c9832
1 file changed
-1
docs/design/moe_kernel_features.md
-1
@@ -93,7 +93,6 @@ To be used with a particular `FusedMoEPrepareAndFinalize` sub-class, MoE kernels
93
| gpt oss triton | standard | N/A | N/A | <sup>5</sup> | Y | Y | [`triton_kernel_fused_experts`][vllm.model_executor.layers.fused_moe.gpt_oss_triton_kernels_moe.triton_kernel_fused_experts],</br>[`OAITritonExperts`][vllm.model_executor.layers.fused_moe.gpt_oss_triton_kernels_moe.OAITritonExperts] |
94
| deep gemm+triton<sup>2</sup> | standard,</br>batched | all<sup>1</sup> | G(128),A,T | silu, gelu | <sup>6</sup> | Y | [`TritonOrDeepGemmExperts`][vllm.model_executor.layers.fused_moe.triton_deep_gemm_moe.TritonOrDeepGemmExperts],</br>[`BatchedTritonOrDeepGemmExperts`][vllm.model_executor.layers.fused_moe.batched_triton_or_deep_gemm_moe.BatchedTritonOrDeepGemmExperts] |
95
| marlin | standard | <sup>3</sup> | <sup>3</sup> | silu,</br>swigluoai | Y | N | [`fused_marlin_moe`][vllm.model_executor.layers.fused_moe.fused_marlin_moe.fused_marlin_moe] |
96
-
96
| marlin experts | standard | N/A | N/A | silu,</br>swigluoai | Y | Y | [`MarlinExperts`][vllm.model_executor.layers.fused_moe.fused_marlin_moe.MarlinExperts] |
97
| trtllm | standard | mxfp4,</br>nvfp4 | G(16),G(32) | <sup>5</sup> | N | Y | [`TrtLlmGenExperts`][vllm.model_executor.layers.fused_moe.trtllm_moe.TrtLlmGenExperts] |
98
| pallas | standard | N/A | N/A | silu | N | N | [`fused_moe`][vllm.model_executor.layers.fused_moe.moe_pallas.fused_moe] |