xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-21 04:55:01 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Fix fp8 DeepGemm compilation issues (#30336)

Browse Source
This commit is contained in:
ElizaWszola 2025-12-10 02:17:25 +01:00 committed by GitHub
parent 4c2e10ea19
commit 2e7035dd8c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/model_executor/layers/quantization/utils/fp8_utils.py
View File

@ -31,7 +31,6 @@ from vllm.model_executor.utils import replace_parameter
from vllm.platforms import current_platform
from vllm.triton_utils import tl, triton
from vllm.utils.deep_gemm import (
DeepGemmQuantScaleFMT,
fp8_gemm_nt,
is_deep_gemm_e8m0_used,
is_deep_gemm_supported,
@ -248,6 +247,7 @@ class W8A8BlockFp8LinearOp:
self.act_quant_group_shape = act_quant_group_shape
self.is_deep_gemm_supported = is_deep_gemm_supported()
self.is_hopper = current_platform.is_device_capability(90)
self.is_blackwell = current_platform.is_device_capability(100)
self.use_deep_gemm_e8m0 = is_deep_gemm_e8m0_used()
# Get the correct blockscale mul and input quant operations.
@ -303,7 +303,7 @@ class W8A8BlockFp8LinearOp:
weight: torch.Tensor,
weight_scale: torch.Tensor,
) -> torch.Tensor:
if DeepGemmQuantScaleFMT.from_oracle() == DeepGemmQuantScaleFMT.UE8M0:
if self.use_deep_gemm_e8m0 and self.is_blackwell:
q_input, input_scale = per_token_group_quant_fp8_packed_for_deepgemm(
input_2d,
group_size=self.act_quant_group_shape.col,