LLM: Fix vLLM CPU convert error (#11926)
This commit is contained in:
parent
5a8fc1baa2
commit
7ca557aada
1 changed files with 2 additions and 2 deletions
|
|
@ -254,8 +254,8 @@ def get_load_function(low_bit):
|
|||
scheduler_config=self.scheduler_config)
|
||||
return
|
||||
|
||||
_model_mlp_convert()
|
||||
_model_attention_convert()
|
||||
# _model_mlp_convert()
|
||||
# _model_attention_convert()
|
||||
|
||||
self.model = get_model(
|
||||
model_config=self.model_config,
|
||||
|
|
|
|||
Loading…
Reference in a new issue