Update modeling_quiet.py

modeling_quiet.py

@@ -1674,10 +1674,10 @@ class QuietForCausalLM(QuietPreTrainedModel, GenerationMixin):
 					base_attention_mask = base_attention_mask.view(1, 1, seq_len, seq_len)
 					base_attention_mask = base_attention_mask.repeat(input_ids.shape[0], 1, 1, 1)
 					attention_mask = base_attention_mask
-					breakpoint()
+					# breakpoint()
 				elif attention_mask.dim() == 2:
 					if seq_len + past_key_values_length != attention_mask.shape[-1]:
-						breakpoint()
+						# breakpoint()
 						attention_mask = torch.cat(
 							[torch.ones((attention_mask.shape[0], past_key_values_length), dtype=attention_mask.dtype, device=attention_mask.device), attention_mask],
 							dim=-1