Hello, excuse me. In your blog-Classifying text with TensorFlow Estimators at the section of Loading the data, you set 200 as the number of fixed size of sequences. I found this sequences with a maximum of 2494 and minimum of 11, may i ask you why you set 200? Is there any references about that question?
Thank you very much.

The error comes from here

params = {'embedding_initializer': tf.random_uniform_initializer(-1.0, 1.0)}

# Creating an estimator for random embedding
cnn_classifier = tf.estimator.Estimator(model_fn=cnn_model_fn,
                                        model_dir=os.path.join(model_dir, 'cnn'),
                                        params=params)
train_and_evaluate(cnn_classifier)

This is the error.

ValueError: Mismatched label shape. Classifier configured with n_classes=1. Received 254. Suggested Fix: check your n_classes argument to the estimator and/or the shape of your label.

The code worked perfected with py27

In this function you outputs 3 values [x_train, x_len_train, y_train]:
def train_input_fn(): dataset = tf.data.Dataset.from_tensor_slices((x_train, x_len_train, y_train)) dataset = dataset.shuffle(buffer_size=len(x_train_variable)) dataset = dataset.batch(100) dataset = dataset.map(parser) dataset = dataset.repeat() iterator = dataset.make_one_shot_iterator() return iterator.get_next()

However, in Google document, they say:

The return value must be a two-element tuple organized as follows: :

The first element must be a dict in which each input feature is a key, and then a list of values for the training batch.

The second element is a list of labels for the training batch.

So I don't really understand that how custom Estimator can work with a tuple of 3 values
Thanks in advance

Thank you for this great example on input pipeline.

(x_train_variable, y_train), (x_test_variable, y_test) = imdb.load_data(num_words=vocab_size)

I have a dataset that is 2GB in size. Unlike you I cannot load the whole data in memory. How can we modify this logic to include GB's dataset ?

First thank you for your execellent tutorial.

I met a KerError when I run the nlp_estimators.py at L140, which said it cannot find the key logistic:

predictions = np.array([p['logistic'][0] for p in classifier.predict(input_fn=eval_input_fn)])

I'm using TensorFlow 1.9. Is it supposed to be probabilities?