423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-22 22:02:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update vllm_engine.py

Browse Source 
Former-commit-id: 17d0005b8cb9cf75b8247bcdf4ce022e1a5afd0b
This commit is contained in:
hoshi-hiyouga 2024-05-07 00:37:05 +08:00 committed by GitHub
parent 4c91104471
commit df66b288a2
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/llmtuner/chat/vllm_engine.py
View File

@ -106,7 +106,6 @@ class VllmEngine(BaseEngine):
top_k=top_k or generating_args["top_k"],
num_return_sequences=num_return_sequences or 1,
repetition_penalty=repetition_penalty or generating_args["repetition_penalty"],
stop=stop or generating_args["stop"]
)
)
@ -124,10 +123,10 @@ class VllmEngine(BaseEngine):
top_k=generating_args["top_k"],
use_beam_search=generating_args["num_beams"] > 1,
length_penalty=generating_args["length_penalty"],
stop=stop,
stop_token_ids=[self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids,
max_tokens=generating_args["max_new_tokens"],
skip_special_tokens=True,
stop=generating_args["stop"],
)
if self.processor is not None and image is not None: