diff --git a/unsloth/models/rl_replacements.py b/unsloth/models/rl_replacements.py index 4d7a4dbe09..aeb5f3e0d4 100644 --- a/unsloth/models/rl_replacements.py +++ b/unsloth/models/rl_replacements.py @@ -108,6 +108,7 @@ def _sft_trainer_compute_loss(self, model, inputs, return_outputs = False, num_i return_outputs = return_outputs, num_items_in_batch = num_items_in_batch, ) + print(loss, outputs) return (loss, outputs) if return_outputs else loss pass