[Doc] Fix outdated --fp4-gemm-backend documentation (#18350)

This commit is contained in:
Mohammad Miadh Angkad
2026-02-07 20:42:47 +08:00
committed by GitHub
parent d792aa7618
commit fddef76619
3 changed files with 5 additions and 9 deletions

View File

@@ -133,11 +133,7 @@ def fp4_gemm(
fp4_backend = get_fp4_gemm_runner_backend()
if enable_flashinfer_fp4_gemm:
# Use the remapping logic to convert SGLang backend names to FlashInfer API names
backend = (
fp4_backend.get_flashinfer_backend()
if not fp4_backend.is_auto()
else "cutlass"
)
backend = fp4_backend.get_flashinfer_backend()
return flashinfer_fp4_gemm(
input, weight, input_sf, weight_sf, alpha, out_dtype, backend=backend
)

View File

@@ -448,7 +448,7 @@ class ServerArgs:
grammar_backend: Optional[str] = None
mm_attention_backend: Optional[str] = None
fp8_gemm_runner_backend: str = "auto"
fp4_gemm_runner_backend: str = "auto"
fp4_gemm_runner_backend: str = "flashinfer_cutlass"
nsa_prefill_backend: Optional[str] = (
None # None = auto-detect based on hardware/kv_cache_dtype
)
@@ -3789,9 +3789,9 @@ class ServerArgs:
default=ServerArgs.fp4_gemm_runner_backend,
dest="fp4_gemm_runner_backend",
help="Choose the runner backend for NVFP4 GEMM operations. "
"Options: 'auto' (default, selects between flashinfer_cudnn/flashinfer_cutlass based on CUDA/cuDNN version), "
"Options: 'flashinfer_cutlass' (default), "
"'auto' (auto-selects between flashinfer_cudnn/flashinfer_cutlass based on CUDA/cuDNN version), "
"'flashinfer_cudnn' (FlashInfer cuDNN backend, optimal on CUDA 13+ with cuDNN 9.15+), "
"'flashinfer_cutlass' (FlashInfer CUTLASS backend, optimal on CUDA 12), "
"'flashinfer_trtllm' (FlashInfer TensorRT-LLM backend, requires different weight preparation with shuffling). "
"NOTE: This replaces the deprecated environment variable "
"SGLANG_FLASHINFER_FP4_GEMM_BACKEND.",