In [1]:

    

import os
import sys

import tensorflow as tf


    

In [3]:

    

tf_records_dir = '/Users/jorge/data/cats_vs_dogs/'


    

In [4]:

    

# Creates a dataset that reads all of the examples from two files.
tf_trn = [tf_records_dir + "data_train_00000-of-00005.tfrecord",
             tf_records_dir + "data_train_00001-of-00005.tfrecord",
             tf_records_dir + "data_train_00002-of-00005.tfrecord",
             tf_records_dir + "data_train_00003-of-00005.tfrecord",
             tf_records_dir + "data_train_00004-of-00005.tfrecord",]
tf_val = [tf_records_dir + "data_validation_00000-of-00005.tfrecord",
             tf_records_dir + "data_validation_00001-of-00005.tfrecord",
             tf_records_dir + "data_validation_00002-of-00005.tfrecord",
             tf_records_dir + "data_validation_00003-of-00005.tfrecord",
             tf_records_dir + "data_validation_00004-of-00005.tfrecord",]


    

In [ ]:

    

def map_func:


    

In [ ]:

    

# Use a tf.placeholder(tf.string) to represent the filenames,
#  and initialize an iterator from the appropriate filenames

filenames = tf.placeholder(tf.string, shape=[None])
dataset = tf.data.TFRecordDataset(filenames)
dataset = dataset.map(map_func)  # Parse the record into tensors.
dataset = dataset.repeat()  # Repeat the input indefinitely.
dataset = dataset.batch(32)
iterator = dataset.make_initializable_iterator()


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    

# You can feed the initializer with the appropriate filenames for the current
# phase of execution, e.g. training vs. validation.

# Initialize `iterator` with training data.
sess.run(iterator.initializer, feed_dict={filenames: tf_trn})

# Initialize `iterator` with validation data.
sess.run(iterator.initializer, feed_dict={filenames: tf_val})


    

In [ ]: