"too many values to unpack" Errors

Just putting this out there in case anyone knows what's going on, or is running into something similar.

In multiple, seemingly independent places, I get "too many values to unpack" errors, e.g.

➜  key-value-memory-networks git:(master) ✗ python3 interactive.py -m model_memnn_kv.h5
Using TensorFlow backend.
/usr/local/Cellar/python/3.6.5/Frameworks/Python.framework/Versions/3.6/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88
  return f(*args, **kwds)
Namespace(max_mem_size=100, max_query_len=16, model='model_memnn_kv.h5')
load data...
2018-07-23 23:32:18.557243: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
load pickle/mov_vocab.pickle
load pickle/mov_w2i.pickle
load pickle/mov_i2w.pickle
load pickle/mov_w2i_label.pickle
load pickle/mov_i2w_label.pickle
load pickle/mov_kv_pairs.pickle
load pickle/mov_stopwords.pickle
Question: who directed dancing with wolves
q_tokens: ['who', 'directed', 'dancing', 'with', 'wolves']
--- 1 268582
Traceback (most recent call last):
  File "interactive.py", line 71, in <module>
    predict(q)
  File "interactive.py", line 60, in predict
    data_k, data_v = load_kv_dataset([(None,q_tokens,None)], kv_pairs, stopwords)
  File "/Users/samhavens/key-value-memory-networks/process_data.py", line 213, in load_kv_dataset
    for i, (q, _) in enumerate(data):
ValueError: too many values to unpack (expected 2)

This also happened in the notebook, in the second large code block, the line

query_maxlen = max(map(len, (x for _, x, _ in train_data + test_data)))

For the second example, removing the first underscore seemed to work. I don't know if the data changed shape, or what, but it seemed like train_data is an array of tuples with 2 elements.

jojonki / key-value-memory-networks

"too many values to unpack" Errors #3