[AMD] Honor an explicit triton moe_runner_backend for mxfp8 on ROCm (#41377)
This commit is contained in:
@@ -1412,8 +1412,12 @@ def _moe_runner_backend_quant_constraints(view: Any) -> dict:
|
||||
|
||||
is_gfx95_mxfp8 = get_platform().is_hip and is_gfx95_supported()
|
||||
allowed = list(MXFP8_MOE_RUNNER_BACKEND_CHOICES)
|
||||
if is_gfx95_mxfp8:
|
||||
# Every other entry is CUDA-only. Honor an explicit triton request on ROCm
|
||||
# instead of sending it back to flashinfer_trtllm, whose MoE apply path
|
||||
# imports flashinfer.
|
||||
if is_gfx95_mxfp8 or (get_platform().is_hip and moe_runner_backend == "triton"):
|
||||
allowed.append("triton")
|
||||
if is_gfx95_mxfp8:
|
||||
# the aiter MXFP8 MoE quant info is built only when aiter is enabled
|
||||
if envs.SGLANG_USE_AITER.get():
|
||||
allowed.append("aiter")
|
||||
|
||||
Reference in New Issue
Block a user