diff --git a/unsloth/models/rl_replacements.py b/unsloth/models/rl_replacements.py index bcfe4d7774..decaf32096 100644 --- a/unsloth/models/rl_replacements.py +++ b/unsloth/models/rl_replacements.py @@ -177,7 +177,6 @@ def grpo_trainer__get_per_token_logps(function_name, function): if function_name != "_get_per_token_logps": return function def _get_per_token_logps(self, model, input_ids, attention_mask, logits_to_keep): - print(self.args.unsloth_num_chunks) if self.args.unsloth_num_chunks != 1: return None if not hasattr(self, '_autocast_dtype'): self._autocast_dtype = torch.float16 if os.environ.get('ACCELERATE_MIXED_PRECISION', 'fp16') == 'fp16' else torch.bfloat16 @@ -240,7 +239,6 @@ def grpo_trainer_compute_loss(function_name, function): ref_per_token_logps, per_token_logps, input_ids, completion_mask, self.beta, advantages, bsz, ) else: - print(int(self.args.unsloth_num_chunks), end = ",") loss, completion_length, mean_kl = grpo_accumulated_loss( self, _input_ids, logits_to_keep, completion_mask, advantages, n_chunks = self.args.unsloth_num_chunks,