diff --git a/torchnlp/nn/attention.py b/torchnlp/nn/attention.py
index 286e7b8..769bf1c 100755
--- a/torchnlp/nn/attention.py
+++ b/torchnlp/nn/attention.py
@@ -61,9 +61,9 @@ def forward(self, query, context):
         query_len = context.size(1)
 
         if self.attention_type == "general":
-            query = query.view(batch_size * output_len, dimensions)
+            query = query.reshape(batch_size * output_len, dimensions)
             query = self.linear_in(query)
-            query = query.view(batch_size, output_len, dimensions)
+            query = query.reshape(batch_size, output_len, dimensions)
 
         # TODO: Include mask on PADDING_INDEX?