bidirectional rnn

Signed-off-by: Anand Avati <[email protected]>
stanfordmlgroup · Apr 27, 2016 · c77ae22 · c77ae22
1 parent b5ee9e7
commit c77ae22
Show file tree

Hide file tree

Showing 4 changed files with 612 additions and 9 deletions.
diff --git a/nlc.py b/nlc.py
@@ -174,7 +174,7 @@ def train():
 
       # Get a batch and make a step.
       start_time = time.time()
-      encoder_inputs, decoder_inputs, target_weights = model.get_batch(
+      encoder_inputs, decoder_inputs, target_weights, sequence_length = model.get_batch(
           train_set, bucket_id)
       _, step_loss, _ = model.step(sess, encoder_inputs, decoder_inputs,
                                    target_weights, bucket_id, False)
@@ -202,7 +202,7 @@ def train():
           if len(dev_set[bucket_id]) == 0:
             print("  eval: empty bucket %d" % (bucket_id))
             continue
-          encoder_inputs, decoder_inputs, target_weights = model.get_batch(
+          encoder_inputs, decoder_inputs, target_weights, sequence_length = model.get_batch(
               dev_set, bucket_id)
           _, eval_loss, _ = model.step(sess, encoder_inputs, decoder_inputs,
                                        target_weights, bucket_id, True)
@@ -236,7 +236,7 @@ def decode():
       bucket_id = min([b for b in xrange(len(_buckets))
                        if _buckets[b][0] > len(token_ids)])
       # Get a 1-element batch to feed the sentence to the model.
-      encoder_inputs, decoder_inputs, target_weights = model.get_batch(
+      encoder_inputs, decoder_inputs, target_weights, sequence_length = model.get_batch(
           {bucket_id: [(token_ids, [])]}, bucket_id)
       # Get output logits for the sentence.
       _, _, output_logits = model.step(sess, encoder_inputs, decoder_inputs,
@@ -267,7 +267,7 @@ def self_test():
                 [([1, 1, 1, 1, 1], [2, 2, 2, 2, 2]), ([3, 3, 3], [5, 6])])
     for _ in xrange(5):  # Train the fake model for 5 steps.
       bucket_id = random.choice([0, 1])
-      encoder_inputs, decoder_inputs, target_weights = model.get_batch(
+      encoder_inputs, decoder_inputs, target_weights, sequence_length = model.get_batch(
           data_set, bucket_id)
       model.step(sess, encoder_inputs, decoder_inputs, target_weights,
                  bucket_id, False)