Update modeling_bert.py

modeling_bert.py

@@ -537,7 +537,7 @@ class BertOutEffHop(nn.Module):
 
         # Normalize the attention scores to probabilities.
         attention_probs = softmax_1(attention_scores, dim=-1)
-
+        print(softmax_1)
         # This is actually dropping out entire tokens to attend to, which might
         # seem a bit unusual, but is taken from the original Transformer paper.
         attention_probs = self.dropout(attention_probs)