ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

LLM: fix inconsistency between output token number and max_new_token (#8479)

Browse source
This commit is contained in:
binbin Deng 2023-07-07 17:31:05 +08:00 committed by GitHub
parent bcc1eae322
commit d489775d2c
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
python/llm/src/bigdl/llm/ggml/model/generation/utils.py
View file

@ -164,7 +164,7 @@ class GenerationMixin:
res_list = [] res_list = []
word_count = 0 word_count = 0
for token in tokens: for token in tokens:
if word_count > max_new_tokens: if word_count >= max_new_tokens:
break break
res_list.append(token) res_list.append(token)
word_count += 1 word_count += 1