simcop2387/AutoGPTQ
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity

fix bug

Browse source
This commit is contained in:
qwopqwop200 2023-05-02 19:13:13 +09:00 • committed by GitHub
parent 6c23e5b3a5
commit 1388acac94
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
auto_gptq/modeling/llama.py
View file

@ -109,7 +109,7 @@ class LlamaGPTQForCausalLM(BaseGPTQForCausalLM):
)
if fused_attn:
make_quant_attn(model, use_triton=use_triton, desc_act=quantize_config.desc_act)
make_quant_attn(model, use_triton=use_triton, group_size = quantize_config.group_size, desc_act=quantize_config.desc_act,)
if use_triton and fused_mlp:
make_fused_mlp(model)
model_config = model.config.to_dict()