07_sentiment_with_doc2vec + skip_gram + tf 1.4.0 = error
ave19 opened this issue · comments
I think the order of parameters probably moved around in tf. But here's what I get:
batch_inputs, batch_labels = text_helpers.generate_batch_data(
text_data, batch_size,
window_size, method='skip_gram')
...errors out with...
Starting Training
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-9-69e5ea759480> in <module>()
13
14 # Run the train step
---> 15 sess.run(train_step, feed_dict=feed_dict)
16
17 # Return the loss
~/env/3.5/lib/python3.5/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
887 try:
888 result = self._run(None, fetches, feed_dict, options_ptr,
--> 889 run_metadata_ptr)
890 if run_metadata:
891 proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)
~/env/3.5/lib/python3.5/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
1094 'Cannot feed value of shape %r for Tensor %r, '
1095 'which has shape %r'
-> 1096 % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape())))
1097 if not self.graph.is_feedable(subfeed_t):
1098 raise ValueError('Tensor %s may not be fed.' % subfeed_t)
ValueError: Cannot feed value of shape (500,) for Tensor 'Placeholder:0', which has shape '(?, 4)'
I don't have a fix, sorry.
Ah yes, thanks for bringing this to my attention. I think this is a quick fix with the parameters. I'm starting a book/code v2 over the next few months. Expect this to be fixed some time soon. Thanks!
This should be fixed. Please repull and try again with the updated code. Let me know / reopen if you still have issues.