frozenleaves/vllm
1
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2025-10-20 14:53:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] update should_ignore_layer (#11354)

Browse Source 
Signed-off-by: George Ohashi <george@neuralmagic.com>
This commit is contained in:
George
2024-12-21 01:36:23 -05:00
committed by GitHub
parent dd2b5633dd
commit 51ff216d85
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/model_executor/layers/quantization/compressed_tensors/utils.py
View File

@ -30,7 +30,7 @@ def should_ignore_layer(layer_name: Optional[str],
# in the safetensors checkpoint. So, we convert the name
# from the fused version to unfused + check to make sure that
# each shard of the fused layer has the same scheme.
if proj_name in FUSED_LAYER_NAME_MAPPING:
if proj_name in FUSED_LAYER_NAME_MAPPING and layer_name not in ignore:
shard_proj_names = FUSED_LAYER_NAME_MAPPING[proj_name]
# Convert fused_name --> [shard_names]