xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-05-24 00:04:32 +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
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 4 additions and 2 deletions
Show Stats 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.config = config
self.activation_fn = get_act_fn(config.hidden_act) self.activation_fn = get_act_fn(config.hidden_act)
# Special handling for BNB quantization # Special handling for BNB and torchao quantization
if quant_config and quant_config.get_name() == "bitsandbytes": if quant_config and quant_config.get_name() in [
"bitsandbytes", "torchao"
]:
quantizable = True quantizable = True
else: else:
# For other quantization, we require the hidden size to be a # For other quantization, we require the hidden size to be a