Getting ValueError while running Regression Estimator - Training process

[rajeshkumargp]

Hi,

I made my learning in Collab. I have used Tensorflow 1.15 Version.

I have followed https://github.com/ksalama/tf-estimator-tutorials/blob/master/01%20-%20Regression/01.0%20-%20Regression%20Data%20Generation.ipynb

Nextly , this script https://github.com/ksalama/tf-estimator-tutorials/blob/master/01%20-%20Regression/02.0%20-%20TF%20Regression%20Model%20-%20Estimator%20APIs%20%2B%20Pandas.ipynb

While running the section 5. Train the Estimator, and the line estimator.train(input_fn = train_input_fn) I got error as

INFO:tensorflow:Calling model_fn.
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-67-d2bca762b908> in <module>()
----> 1 estimator.train(input_fn = train_input_fn)

10 frames
/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/estimator.py in train(self, input_fn, hooks, steps, max_steps, saving_listeners)
    368 
    369       saving_listeners = _check_listeners_type(saving_listeners)
--> 370       loss = self._train_model(input_fn, hooks, saving_listeners)
    371       logging.info('Loss for final step: %s.', loss)
    372       return self

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/estimator.py in _train_model(self, input_fn, hooks, saving_listeners)
   1159       return self._train_model_distributed(input_fn, hooks, saving_listeners)
   1160     else:
-> 1161       return self._train_model_default(input_fn, hooks, saving_listeners)
   1162 
   1163   def _train_model_default(self, input_fn, hooks, saving_listeners):

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/estimator.py in _train_model_default(self, input_fn, hooks, saving_listeners)
   1189       worker_hooks.extend(input_hooks)
   1190       estimator_spec = self._call_model_fn(
-> 1191           features, labels, ModeKeys.TRAIN, self.config)
   1192       global_step_tensor = training_util.get_global_step(g)
   1193       return self._train_with_estimator_spec(estimator_spec, worker_hooks,

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/estimator.py in _call_model_fn(self, features, labels, mode, config)
   1147 
   1148     logging.info('Calling model_fn.')
-> 1149     model_fn_results = self._model_fn(features=features, **kwargs)
   1150     logging.info('Done calling model_fn.')
   1151 

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/canned/dnn.py in _model_fn(features, labels, mode, config)
   1212           input_layer_partitioner=input_layer_partitioner,
   1213           config=config,
-> 1214           batch_norm=batch_norm)
   1215 
   1216     super(DNNRegressor, self).__init__(

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/canned/dnn.py in _dnn_model_fn(features, labels, mode, head, hidden_units, feature_columns, optimizer, activation_fn, dropout, input_layer_partitioner, config, use_tpu, batch_norm)
    461         input_layer_partitioner=input_layer_partitioner,
    462         batch_norm=batch_norm)
--> 463     logits = logit_fn(features=features, mode=mode)
    464 
    465     return _get_dnn_estimator_spec(use_tpu, head, features, labels, mode,

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/canned/dnn.py in dnn_logit_fn(features, mode)
    106         input_layer_partitioner,
    107         batch_norm,
--> 108         name='dnn')
    109     return dnn_model(features, mode)
    110 

/tensorflow-1.15.2/python3.6/tensorflow_estimator/python/estimator/canned/dnn.py in __init__(self, units, hidden_units, feature_columns, activation_fn, dropout, input_layer_partitioner, batch_norm, name, **kwargs)
    184     if feature_column_lib.is_feature_column_v2(feature_columns):
    185       self._input_layer = dense_features.DenseFeatures(
--> 186           feature_columns=feature_columns, name='input_layer')
    187     else:
    188       self._input_layer = feature_column.InputLayer(

/tensorflow-1.15.2/python3.6/tensorflow_core/python/feature_column/dense_features.py in __init__(self, feature_columns, trainable, name, **kwargs)
     84         name=name,
     85         expected_column_type=fc.DenseColumn,
---> 86         **kwargs)
     87 
     88   @property

/tensorflow-1.15.2/python3.6/tensorflow_core/python/feature_column/feature_column_v2.py in __init__(self, feature_columns, expected_column_type, trainable, name, **kwargs)
    370     super(_BaseFeaturesLayer, self).__init__(
    371         name=name, trainable=trainable, **kwargs)
--> 372     self._feature_columns = _normalize_feature_columns(feature_columns)
    373     self._state_manager = _StateManagerImpl(self, self.trainable)
    374     for column in self._feature_columns:

/tensorflow-1.15.2/python3.6/tensorflow_core/python/feature_column/feature_column_v2.py in _normalize_feature_columns(feature_columns)
   2727                        'Given (type {}): {}.'.format(type(column), column))
   2728   if not feature_columns:
-> 2729     raise ValueError('feature_columns must not be empty.')
   2730   name_to_column = {}
   2731   for column in feature_columns:

ValueError: feature_columns must not be empty.

Can you please guide on how to debug this issue.