From 0a4ddbcf7dd88c69c543a77345594268bdb816a8 Mon Sep 17 00:00:00 2001 From: Daniel Han Date: Wed, 28 May 2025 12:27:55 -0700 Subject: [PATCH] Update rl.py --- unsloth/models/rl.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/unsloth/models/rl.py b/unsloth/models/rl.py index 6da8bf4604..8b6dd16da5 100644 --- a/unsloth/models/rl.py +++ b/unsloth/models/rl.py @@ -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