ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update rl.py

Browse source
This commit is contained in:
Daniel Han 2025-05-28 12:27:55 -07:00
commit b087bfaf55
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
unsloth/models/rl.py
View file

@ -374,7 +374,7 @@ def _patch_trl_rl_trainers(trainer_file = "grpo_trainer"):
" if isinstance(data_collator, DataCollatorForSeq2Seq):\n"\
" data_collator = DataCollatorForSeq2Seq(__tokenizer.tokenizer)\n"\
" else:\n"\
" data_collator = TransformersDataCollatorForLanguageModeling(__tokenizer.tokenizer, mlm = False)\n"
" data_collator = TransformersDataCollatorForLanguageModeling(__tokenizer.tokenizer, mlm = False, mlm_probability = 0.0)\n"
extra_args += pad_check
pass