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

Fix: default repeat_penalty

Browse source
This commit is contained in:
Andrei Betlen 2023-05-08 18:49:11 -04:00
parent a3cc7bf5b2
commit 82d138fe54
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/server/app.py
View file

@ -146,7 +146,7 @@ top_k_field = Field(
) )
repeat_penalty_field = Field( repeat_penalty_field = Field(
default=0.0, default=1.1,
ge=0.0, ge=0.0,
description="A penalty applied to each token that is already generated. This helps prevent the model from repeating itself.\n\n" description="A penalty applied to each token that is already generated. This helps prevent the model from repeating itself.\n\n"
+ "Repeat penalty is a hyperparameter used to penalize the repetition of token sequences during text generation. It helps prevent the model from generating repetitive or monotonous text. A higher value (e.g., 1.5) will penalize repetitions more strongly, while a lower value (e.g., 0.9) will be more lenient.", + "Repeat penalty is a hyperparameter used to penalize the repetition of token sequences during text generation. It helps prevent the model from generating repetitive or monotonous text. A higher value (e.g., 1.5) will penalize repetitions more strongly, while a lower value (e.g., 0.9) will be more lenient.",