Training a neural network on MNIST with Keras

This simple example demonstrate how to plug TFDS into a Keras model.



In [0]:

    
import tensorflow.compat.v2 as tf
import tensorflow_datasets as tfds

tfds.disable_progress_bar()
tf.enable_v2_behavior()

Step 1: Create your input pipeline

Build efficient input pipeline using advices from:

Load MNIST

Load with the following arguments:

shuffle_files: The MNIST data is only stored in a single file, but for larger datasets with multiple files on disk, it's good practice to shuffle them when training.
as_supervised: Returns tuple (img, label) instead of dict {'image': img, 'label': label}



In [0]:

    
(ds_train, ds_test), ds_info = tfds.load(
    'mnist',
    split=['train', 'test'],
    shuffle_files=True,
    as_supervised=True,
    with_info=True,
)

Build training pipeline

Apply the following transormations:

ds.map: TFDS provide the images as tf.uint8, while the model expect tf.float32, so normalize images
ds.cache As the dataset fit in memory, cache before shuffling for better performance.
Note: Random transformations should be applied after caching
ds.shuffle: For true randomness, set the shuffle buffer to the full dataset size.
Note: For bigger datasets which do not fit in memory, a standard value is 1000 if your system allows it.
ds.batch: Batch after shuffling to get unique batches at each epoch.
ds.prefetch: Good practice to end the pipeline by prefetching for performances.



In [0]:

    
def normalize_img(image, label):
  """Normalizes images: `uint8` -> `float32`."""
  return tf.cast(image, tf.float32) / 255., label

ds_train = ds_train.map(
    normalize_img, num_parallel_calls=tf.data.experimental.AUTOTUNE)
ds_train = ds_train.cache()
ds_train = ds_train.shuffle(ds_info.splits['train'].num_examples)
ds_train = ds_train.batch(128)
ds_train = ds_train.prefetch(tf.data.experimental.AUTOTUNE)

Build evaluation pipeline

Testing pipeline is similar to the training pipeline, with small differences:

No ds.shuffle() call
Caching is done after batching (as batches can be the same between epoch)



In [0]:

    
ds_test = ds_test.map(
    normalize_img, num_parallel_calls=tf.data.experimental.AUTOTUNE)
ds_test = ds_test.batch(128)
ds_test = ds_test.cache()
ds_test = ds_test.prefetch(tf.data.experimental.AUTOTUNE)

Step 2: Create and train the model

Plug the input pipeline into Keras.



In [0]:

    
model = tf.keras.models.Sequential([
  tf.keras.layers.Flatten(input_shape=(28, 28, 1)),
  tf.keras.layers.Dense(128,activation='relu'),
  tf.keras.layers.Dense(10, activation='softmax')
])
model.compile(
    loss='sparse_categorical_crossentropy',
    optimizer=tf.keras.optimizers.Adam(0.001),
    metrics=['accuracy'],
)

model.fit(
    ds_train,
    epochs=6,
    validation_data=ds_test,
)