example model (resnet50_coco_best_v1.2.2.h5) failed to load, while v1.2.1.h5 works.

I followed the “example/ResNet50RetinaNet - COCO 2017” notebook and found the latest model failed to load, while the previous version works fine.

Error is TypeError: Fetch argument None has invalid type <class 'NoneType'>

Environment:

Keras 2.1.2 py 3.6 TF: 1.4

=====

Complete stack trace:

C:\ProgramData\Anaconda3\lib\site-packages\keras-2.1.2-py3.6.egg\keras\models.py:271: UserWarning: Output "non_maximum_suppression_2" missing from loss dictionary. We assume this was done on purpose, and we will not be expecting any data to be passed to "non_maximum_suppression_2" during training.
  sample_weight_mode=sample_weight_mode)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-6-b8566c0d85ae> in <module>()
----> 1 model = keras.models.load_model('snapshots/resnet50_coco_best_v1.2.2.h5', custom_objects=custom_objects)
      2 #print(model.summary())

C:\ProgramData\Anaconda3\lib\site-packages\keras-2.1.2-py3.6.egg\keras\models.py in load_model(filepath, custom_objects, compile)
    284                                        optimizer_weight_names]
    285             try:
--> 286                 model.optimizer.set_weights(optimizer_weight_values)
    287             except ValueError:
    288                 warnings.warn('Error in loading the saved optimizer '

C:\ProgramData\Anaconda3\lib\site-packages\keras-2.1.2-py3.6.egg\keras\optimizers.py in set_weights(self, weights)
     97         params = self.weights
     98         weight_value_tuples = []
---> 99         param_values = K.batch_get_value(params)
    100         for pv, p, w in zip(param_values, params, weights):
    101             if pv.shape != w.shape:

C:\ProgramData\Anaconda3\lib\site-packages\keras-2.1.2-py3.6.egg\keras\backend\tensorflow_backend.py in batch_get_value(ops)
   2208     """
   2209     if ops:
-> 2210         return get_session().run(ops)
   2211     else:
   2212         return []

C:\ProgramData\Anaconda3\lib\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)

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
   1103     # Create a fetch handler to take care of the structure of fetches.
   1104     fetch_handler = _FetchHandler(
-> 1105         self._graph, fetches, feed_dict_tensor, feed_handles=feed_handles)
   1106 
   1107     # Run request and get response.

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in __init__(self, graph, fetches, feeds, feed_handles)
    412     """
    413     with graph.as_default():
--> 414       self._fetch_mapper = _FetchMapper.for_fetch(fetches)
    415     self._fetches = []
    416     self._targets = []

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in for_fetch(fetch)
    232     elif isinstance(fetch, (list, tuple)):
    233       # NOTE(touts): This is also the code path for namedtuples.
--> 234       return _ListFetchMapper(fetch)
    235     elif isinstance(fetch, dict):
    236       return _DictFetchMapper(fetch)

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in __init__(self, fetches)
    339     """
    340     self._fetch_type = type(fetches)
--> 341     self._mappers = [_FetchMapper.for_fetch(fetch) for fetch in fetches]
    342     self._unique_fetches, self._value_indices = _uniquify_fetches(self._mappers)
    343 

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in <listcomp>(.0)
    339     """
    340     self._fetch_type = type(fetches)
--> 341     self._mappers = [_FetchMapper.for_fetch(fetch) for fetch in fetches]
    342     self._unique_fetches, self._value_indices = _uniquify_fetches(self._mappers)
    343 

C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\client\session.py in for_fetch(fetch)
    229     if fetch is None:
    230       raise TypeError('Fetch argument %r has invalid type %r' %
--> 231                       (fetch, type(fetch)))
    232     elif isinstance(fetch, (list, tuple)):
    233       # NOTE(touts): This is also the code path for namedtuples.

TypeError: Fetch argument None has invalid type <class 'NoneType'>

[hgaiser]

You need to train your own model and use that, or download the pretrained model from the README. Then adjust the example code to point to that model and you should be good to go.

[hgaiser]

Closing this due to inactivity of OP.