Update modeling_quiet.py

modeling_quiet.py

@@ -1737,10 +1737,10 @@ class QuietForCausalLM(QuietPreTrainedModel, GenerationMixin):
                         shift_labels = labels[..., 1 + shift_amount:].contiguous()
                         # Flatten the tokens
                         loss_fct = CrossEntropyLoss(reduction="none")
-                        print("Shift logits before:", shift_logits)
+                        # print("Shift logits before:", shift_logits)
                         shift_logits = shift_logits.view(-1, self.config.vocab_size)
                         shift_labels = shift_labels.view(-1).clone()
-                        print("shift logits after:", shift_logits)
+                        # print("shift logits after:", shift_logits)
                         # Enable model parallelism
                         shift_labels[shift_labels == self.tokenizer.pad_token_id] = -100
                         shift_labels = shift_labels.to(shift_logits.device)