Update bert_layers.py

bert_layers.py

@@ -485,6 +485,7 @@ class BertEncoder(nn.Module):
                                                               attn_mask=attention_mask,
                                                               bias=alibi_attn_mask)
             all_attention_weights.append(attention_weights)  # Store attention weights
+            print(all_attention_weights)
     
         if not output_all_encoded_layers:
             all_encoder_layers.append(hidden_states)