Can not run example with pre-built models

[nkrot]

System information

• Ubuntu 18.0.4 (LTS)
• Python 3.6.8
• anago==1.0.8
• Keras==2.2.5
• Keras-Applications==1.0.8
• Keras-Preprocessing==1.1.0
• tensorflow==1.14.0
• tensorflow-estimator==1.14.0

Problem

I try to run pretrained models as shown in README, namely:

url = 'https://s3-ap-northeast-1.amazonaws.com/dev.tech-sketch.jp/chakki/public/conll2003_en.zip'
weights, params, preprocessor = anago.utils.download(url)
model = anago.Sequence.load(weights, params, preprocessor)
text = 'President Obama is speaking at the White House .'
res = model.analyze(text)

and receive the following error (future warnings have been silenced):

Using TensorFlow backend.
/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/sklearn/externals/joblib/__init__.py:15: DeprecationWarning: sklearn.externals.joblib is deprecated in 0.21 and will be removed in 0.23. Please import this functionality directly from joblib, which can be installed with: pip install joblib. If this warning is raised when loading pickled models, you may need to re-serialize those models with scikit-learn 0.21+.
  warnings.warn(msg, category=DeprecationWarning)
WARNING: Logging before flag parsing goes to stderr.
W0827 17:33:44.624381 140560557238080 deprecation_wrapper.py:119] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.

W0827 17:33:44.635842 140560557238080 deprecation_wrapper.py:119] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.

W0827 17:33:44.646406 140560557238080 deprecation_wrapper.py:119] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.

W0827 17:33:44.847993 140560557238080 deprecation.py:323] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:3239: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.where in 2.0, which has the same broadcast rule as np.where
W0827 17:33:45.003281 140560557238080 deprecation_wrapper.py:119] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:148: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.

W0827 17:33:45.008451 140560557238080 deprecation.py:506] From /tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:3733: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.
Instructions for updating:
Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.
Traceback (most recent call last):
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 471, in _apply_op_helper
    as_ref=input_arg.is_ref)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1293, in internal_convert_n_to_tensor
    ctx=ctx))
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1224, in internal_convert_to_tensor
    ret = conversion_func(value, dtype=dtype, name=name, as_ref=as_ref)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1018, in _TensorTensorConversionFunction
    (dtype.name, t.dtype.name, str(t)))
ValueError: Tensor conversion requested dtype bool for Tensor with dtype float32: 'Tensor("crf_1/zeros_like_4:0", shape=(?, ?), dtype=float32)'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "./test_02.py", line 21, in <module>
    model = anago.Sequence.load(weights, params, preprocessor)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/anago/wrapper.py", line 140, in load
    self.model = load_model(weights, params)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/anago/models.py", line 22, in load_model
    model = model_from_json(f.read(), custom_objects={'CRF': CRF})
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/engine/saving.py", line 627, in model_from_json
    return deserialize(config, custom_objects=custom_objects)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/layers/__init__.py", line 168, in deserialize
    printable_module_name='layer')
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/utils/generic_utils.py", line 147, in deserialize_keras_object
    list(custom_objects.items())))
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/engine/network.py", line 1062, in from_config
    process_node(layer, node_data)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/engine/network.py", line 1012, in process_node
    layer(unpack_singleton(input_tensors), **kwargs)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/engine/base_layer.py", line 451, in __call__
    output = self.call(inputs, **kwargs)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/anago/layers.py", line 215, in call
    test_output = self.viterbi_decoding(X, mask)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/anago/layers.py", line 484, in viterbi_decoding
    argmin_tables = self.recursion(input_energy, mask, return_logZ=False)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/anago/layers.py", line 445, in recursion
    mask2 = K.cast(K.concatenate([mask, K.zeros_like(mask[:, :1])], axis=1), K.floatx())
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 2176, in concatenate
    return tf.concat([to_dense(x) for x in tensors], axis)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/util/dispatch.py", line 180, in wrapper
    return target(*args, **kwargs)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/ops/array_ops.py", line 1299, in concat
    return gen_array_ops.concat_v2(values=values, axis=axis, name=name)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/ops/gen_array_ops.py", line 1256, in concat_v2
    "ConcatV2", values=values, axis=axis, name=name)
  File "/tmp/test-anago/venv-1.0.8/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 499, in _apply_op_helper
    raise TypeError("%s that don't all match." % prefix)
TypeError: Tensors in list passed to 'values' of 'ConcatV2' Op have types [bool, float32] that don't all match.

Could you help me understand what i am doing wrong?

Thanks in advance, Nikolai

[nkrot]

Hi. I think I have found the cause. The same cause is also preventing the training on a fresh model.

So... the reason that pip install anago (that installs anago==1.0.8) installs too new libraries of Keras and tensorflow. In fact, I ended up with:

Keras==2.2.5
tensorflow==1.14.0

Which is expected given that the file setup.py (https://github.com/Hironsan/anago/blob/master/setup.py) specifies versions of these libraries in a broad, in fact open-ended, fashion:

    'Keras>=2.2.0', 'h5py>=2.7.1', 'scikit-learn>=0.19.1',
    'numpy>=1.14.3', 'tensorflow>=1.8.0', 'requests>=2.18.4',
    'seqeval>=0.0.3'

Assuming that anago was tested with at least the lowest versions, I downgraded Keras, tensorflow and seqeval libraries.