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

Allow torchao quantization in SiglipMLP (#15575)

Browse Source
This commit is contained in:
Jerry Zhang
2025-03-26 22:45:51 -07:00
committed by GitHub
parent 54aa619459
commit dcf2a590f5
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/model_executor/models/siglip.py
View File

@ -208,8 +208,10 @@ class SiglipMLP(nn.Module):
self.config = config
self.activation_fn = get_act_fn(config.hidden_act)
# Special handling for BNB quantization
if quant_config and quant_config.get_name() == "bitsandbytes":
# Special handling for BNB and torchao quantization
if quant_config and quant_config.get_name() in [
"bitsandbytes", "torchao"
]:
quantizable = True
else:
# For other quantization, we require the hidden size to be a