vllm.model_executor.layers.fused_moe.experts ¶

Modules:

• aiter_mxfp4_w4a8_moe –
• batched_deep_gemm_moe –
• cpu_int4_moe –

  CPU INT4 W4A8 dynamic quantized fused MoE experts.

• cpu_moe –

  CPU FP8 W8A16 and MXFP4 W4A16 fused MoE experts.

• cutlass_moe –

  CUTLASS based Fused MoE kernels.

• deep_gemm_moe –
• fallback –
• flashinfer_b12x_moe –
• flashinfer_cutedsl_batched_moe –
• flashinfer_cutedsl_moe –
• flashinfer_cutlass_moe –
• fused_batched_moe –

  Fused batched MoE kernel.

• fused_humming_moe –

  Fused MoE utilities for Humming.

• gpt_oss_triton_kernels_moe –
• lora_context –
• lora_experts_mixin –
• marlin_moe –

  Fused MoE utilities for GPTQ.

• nvfp4_emulation_moe –

  NVFP4 quantization emulation for MoE.

• ocp_mx_emulation_moe –

  OCP MX quantization emulation for MoE.

• rocm_aiter_moe –
• triton_cutlass_moe –
• triton_deep_gemm_moe –
• triton_moe –

  Triton-based MoE expert implementations.

• trtllm_bf16_moe –
• trtllm_fp8_moe –
• trtllm_mxfp4_moe –
• trtllm_mxint4_moe –
• trtllm_nvfp4_moe –
• xpu_moe –