xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-09 07:47:09 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bug] Fix AttributeError: 'ColumnParallelLinear' object has no attribute weight_scale_inv (#30823)

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
Wentao Ye 2025-12-17 05:00:35 -05:00 committed by GitHub
parent 53cd7f868b
commit f284d7bd0c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/model_executor/layers/quantization/utils/fp8_utils.py
View File

@ -1437,14 +1437,17 @@ def maybe_post_process_fp8_weight_block(layer: torch.nn.Module):
layer.orig_dtype, layer.weight layer.orig_dtype, layer.weight
) )
if should_use_deepgemm: if should_use_deepgemm:
scale_attr = (
"weight_scale_inv" if hasattr(layer, "weight_scale_inv") else "weight_scale"
)
dg_weight, dg_weight_scale = deepgemm_post_process_fp8_weight_block( dg_weight, dg_weight_scale = deepgemm_post_process_fp8_weight_block(
wq=layer.weight.data, wq=layer.weight.data,
ws=layer.weight_scale_inv.data, ws=getattr(layer, scale_attr).data,
quant_block_shape=tuple(layer.weight_block_size), quant_block_shape=tuple(layer.weight_block_size),
use_e8m0=is_deep_gemm_e8m0_used(), use_e8m0=is_deep_gemm_e8m0_used(),
) )
replace_parameter(layer, "weight", dg_weight) replace_parameter(layer, "weight", dg_weight)
replace_parameter(layer, "weight_scale_inv", dg_weight_scale) replace_parameter(layer, scale_attr, dg_weight_scale)
def expert_weight_is_col_major(x: torch.Tensor) -> bool: def expert_weight_is_col_major(x: torch.Tensor) -> bool: