[Bugfix] Fix DeepGemm E8M0 accuracy degradation for Qwen3.5 FP8 on Blackwell (#38083)
Signed-off-by: Vadim Gimpelson <vadim.gimpelson@gmail.com>
This commit is contained in:
@@ -359,10 +359,14 @@ class W8A8BlockFp8LinearOp:
|
||||
act_quant_group_shape: GroupShape,
|
||||
cutlass_block_fp8_supported: bool = CUTLASS_BLOCK_FP8_SUPPORTED,
|
||||
use_aiter_and_is_supported: bool = False,
|
||||
use_deep_gemm: bool | None = None,
|
||||
):
|
||||
self.weight_group_shape = weight_group_shape
|
||||
self.act_quant_group_shape = act_quant_group_shape
|
||||
self.is_deep_gemm_supported = is_deep_gemm_supported()
|
||||
if use_deep_gemm is not None:
|
||||
self.is_deep_gemm_supported = use_deep_gemm
|
||||
else:
|
||||
self.is_deep_gemm_supported = is_deep_gemm_supported()
|
||||
self.is_hopper = current_platform.is_device_capability(90)
|
||||
self.use_deep_gemm_e8m0 = is_deep_gemm_e8m0_used()
|
||||
self.is_flashinfer_supported = is_flashinfer_fp8_blockscale_gemm_supported()
|
||||
|
||||
Reference in New Issue
Block a user