A script to test M1 GPU training via the tensorflow-metal plugin

test_tensorflow_metal.py

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

tf.get_logger().setLevel('DEBUG') 

tf.enable_v2_behavior()

from tensorflow.python.framework.ops import disable_eager_execution
disable_eager_execution()

# from tensorflow.python.compiler.mlcompute import mlcompute
# mlcompute.set_mlc_device(device_name='gpu')
import os
os.environ['TF_MLC_DEVICE_TYPE'] = os.environ.get('TF_MLC_DEVICE_TYPE', '1')

if 'mixed_precision' and False:
  from tensorflow import keras
  from tensorflow.keras import layers
  from tensorflow.keras import mixed_precision
  policy = mixed_precision.Policy('mixed_float16')
  mixed_precision.set_global_policy(policy)


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

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

batch_size = 128

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(batch_size)
ds_train = ds_train.prefetch(tf.data.experimental.AUTOTUNE)


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


model = tf.keras.models.Sequential([
  tf.keras.layers.Conv2D(32, kernel_size=(3, 3),
                 activation='relu'),
  tf.keras.layers.Conv2D(64, kernel_size=(3, 3),
                 activation='relu'),
  tf.keras.layers.MaxPooling2D(pool_size=(2, 2)),
#   tf.keras.layers.Dropout(0.25),
  tf.keras.layers.Flatten(),
  tf.keras.layers.Dense(128, activation='relu'),
#   tf.keras.layers.Dropout(0.5),
  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=4,
    validation_data=ds_test,
)

"""
469/469 [==============================] - 12s 20ms/step - batch: 234.0000 - size: 1.0000 - loss: 0.1537 - accuracy: 0.9546 - val_loss: 0.0655 - val_accuracy: 0.9785
Epoch 2/4
469/469 [==============================] - 10s 20ms/step - batch: 234.0000 - size: 1.0000 - loss: 0.0425 - accuracy: 0.9869 - val_loss: 0.0365 - val_accuracy: 0.9875
Epoch 3/4
469/469 [==============================] - 10s 20ms/step - batch: 234.0000 - size: 1.0000 - loss: 0.0268 - accuracy: 0.9919 - val_loss: 0.0427 - val_accuracy: 0.9859
Epoch 4/4
469/469 [==============================] - 10s 20ms/step - batch: 234.0000 - size: 1.0000 - loss: 0.0184 - accuracy: 0.9940 - val_loss: 0.0377 - val_accuracy: 0.9886
"""