CODEWITHSUNDEEP
pythontensorflow
Missak Boyajian
Missak Boyajian

Reputation: 2245

TensorFlow - shape does not match the shape stored in checkpoint

I am very new to TensorFlow and I am trying Simple Neural Network on my data. I have a .csv data with 19 columns and the last column being the target column. It is 0 or 1.

I started from here https://www.tensorflow.org/get_started/estimator and trying to modify to suit my data. I produced this.

...

    # Data sets
    IRIS_TRAINING = "Training.csv"
    IRIS_TEST = "Test.csv"

    def main():

      # Load datasets.
      training_set = tf.contrib.learn.datasets.base.load_csv_without_header(
          filename=IRIS_TRAINING,
          target_dtype=np.int,
          features_dtype=np.float32)

      test_set = tf.contrib.learn.datasets.base.load_csv_without_header(
          filename=IRIS_TEST,
          target_dtype=np.int,
          features_dtype=np.float32)

      # Specify that all features have real-value data
      feature_columns = [tf.feature_column.numeric_column("x", shape=[18])]

      **# SOMETHING WRONG HERE** 
      classifier = tf.estimator.DNNClassifier(feature_columns=feature_columns,
                                              hidden_units=[18],
                                              n_classes=2,
                                              model_dir="/tmp/iris_model")
      # Define the training inputs
      train_input_fn = tf.estimator.inputs.numpy_input_fn(
          x={"x": np.array(training_set.data)},
          y=np.array(training_set.target),
          num_epochs=None,
          shuffle=True)

      # Train model.
      classifier.train(input_fn=train_input_fn, steps=2000)

      # Define the test inputs
      test_input_fn = tf.estimator.inputs.numpy_input_fn(
          x={"x": np.array(test_set.data)},
          y=np.array(test_set.target),
          num_epochs=1,
          shuffle=False)

      # Evaluate accuracy.
      accuracy_score = classifier.evaluate(input_fn=test_input_fn)["accuracy"]

      print("\nTest Accuracy: {0:f}\n".format(accuracy_score))



    if __name__ == "__main__":
        main()

I just changed the hidden units to make it 1 layer and changed the shape to 18 because I have 18 features. However, I am getting this error.

InvalidArgumentError (see above for traceback): tensor_name = dnn/hiddenlayer_0/bias/t_0/Adagrad; shape in shape_and_slice spec [18] does not match the shape stored in checkpoint: [10]
         [[Node: save/RestoreV2_1 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2_1/tensor_names, save/RestoreV2_1/shape_and_slices)]]

Upvotes: 1

Views: 3022

Answers (1)

Krishnan R.
Krishnan R.

Reputation: 66

I believe that you problem lies in model_dir="/tmp/iris_model" in tf.estimator.DNNClassifier(). This is actually loading and retraining the model that it saved to that directory the first time you ran it with the Tensorflow example data. Just take out the model_dir="/tmp/iris_model" part and that error should disappear.

Source: https://www.tensorflow.org/api_docs/python/tf/estimator/DNNClassifier

Upvotes: 4

Related Questions