From 338fbd546bfe0c605a608fa4f6fb0235287175af Mon Sep 17 00:00:00 2001 From: Parag Ekbote Date: Tue, 4 Feb 2025 12:16:29 -0800 Subject: [PATCH] =?UTF-8?q?=F0=9F=93=96=20Clarification=20max=20len=20in?= =?UTF-8?q?=20Reward=20documentation=20(#2740)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * Nit fix about max_lenth argument. * copy to docstring * typo * consistency --------- Co-authored-by: Quentin Gallouédec Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com> --- trl/trainer/reward_config.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/trl/trainer/reward_config.py b/trl/trainer/reward_config.py index ba40e3f80a..f324637e52 100644 --- a/trl/trainer/reward_config.py +++ b/trl/trainer/reward_config.py @@ -29,8 +29,8 @@ class RewardConfig(TrainingArguments): Parameters: max_length (`int` or `None`, *optional*, defaults to `1024`): - Maximum length of the sequences (prompt + completion) in the batch. This argument is required if you want - to use the default data collator. + Maximum length of the sequences (prompt + completion) in the batch, filters out entries that exceed the + limit. This argument is required if you want to use the default data collator. disable_dropout (`bool`, *optional*, defaults to `True`): Whether to disable dropout in the model. dataset_num_proc (`int`, *optional*, defaults to `None`): @@ -46,8 +46,8 @@ class RewardConfig(TrainingArguments): max_length: Optional[int] = field( default=1024, metadata={ - "help": "Maximum length of the sequences (prompt + completion) in the batch. This argument is required if " - "you want to use the default data collator." + "help": "Maximum length of the sequences (prompt + completion) in the batch, filters out entries that " + "exceed the limit. This argument is required if you want to use the default data collator." }, ) disable_dropout: bool = field(