ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update llama.py

Browse source
This commit is contained in:
Daniel Han 2024-07-23 12:27:46 -07:00
commit 92b3752cad
1 changed files with 5 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

9
unsloth/models/llama.py
View file

@ -1178,10 +1178,11 @@ def _wrap_fast_inference(generate, device_type, dtype, model):
kwargs.pop("token_type_ids", None)
# Check pad_token
kwargs["pad_token_id"] = kwargs.pop(
"pad_token_id",
getattr(model.config, "eos_token_id", None),
)
model_eos_token_id = getattr(model.config, "eos_token_id", None)
if hasattr(model_eos_token_id, "__iter__"):
model_eos_token_id = model_eos_token_id[0]
kwargs["pad_token_id"] = kwargs.pop("pad_token_id", model_eos_token_id)
# Set pad token
# old_pad_token_id = getattr(model.config, "pad_token_id", None)