baalajimaestro/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Now the last token sent when stream=True

Browse source
This commit is contained in:
Carlos Tejada 2023-07-19 22:47:14 -04:00
parent 36872620d0
commit 0756a2d3fb
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
llama_cpp/llama.py
View file

@ -953,7 +953,7 @@ class Llama:
token_end_position += len(self.detokenize([token])) token_end_position += len(self.detokenize([token]))
# Check if stop sequence is in the token # Check if stop sequence is in the token
if token_end_position >= ( if token_end_position >= (
remaining_length - first_stop_position - 1 remaining_length - first_stop_position
): ):
break break
logprobs_or_none: Optional[CompletionLogprobs] = None logprobs_or_none: Optional[CompletionLogprobs] = None