tensorflow2_keras_mnist.py

# Copyright 2019 Uber Technologies, Inc. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

import tensorflow as tf
import argparse
import time
# Horovod: initialize Horovod.

try:
    import horovod.tensorflow.keras as hvd
    with_hvd=True
except:
    with_hvd=False
    class Hvd:
        def init():
            print("I could not find Horovod package, will do things sequentially")
        def rank():
            return 0
        def size():
            return 1
    hvd=Hvd;

hvd.init()
t0 = time.time()
parser = argparse.ArgumentParser(description='TensorFlow MNIST Example')
parser.add_argument('--batch_size', type=int, default=64, metavar='N',
                    help='input batch size for training (default: 64)')
parser.add_argument('--epochs', type=int, default=10, metavar='N',
                    help='number of epochs to train (default: 10)')
parser.add_argument('--lr', type=float, default=0.001, metavar='LR',
                    help='learning rate (default: 0.01)')
parser.add_argument('--device', default='cpu',
                    help='Wheter this is running on cpu or gpu')
parser.add_argument('--num_inter', default=2, help='set number inter', type=int)
parser.add_argument('--num_intra', default=0, help='set number intra', type=int)
parser.add_argument('--warmup_epochs', default=3, help='number of warmup epochs', type=int)
parser.add_argument("--log_dir", default='log_dir')
args = parser.parse_args()

# Horovod: pin GPU to be used to process local rank (one GPU per process)

print("I am rank %s of %s" %(hvd.rank(), hvd.size()))
# Horovod: pin GPU to be used to process local rank (one GPU per process)
if args.device == 'cpu':
    tf.config.threading.set_intra_op_parallelism_threads(args.num_intra)
    tf.config.threading.set_inter_op_parallelism_threads(args.num_inter)
else:
    gpus = tf.config.experimental.list_physical_devices('GPU')
    for gpu in gpus:
        tf.config.experimental.set_memory_growth(gpu, True)
    if gpus:
        tf.config.experimental.set_visible_devices(gpus[hvd.local_rank()], 'GPU')

(mnist_images, mnist_labels), _ = \
    tf.keras.datasets.mnist.load_data(path='mnist.npz')

dataset = tf.data.Dataset.from_tensor_slices(
    (tf.cast(mnist_images[..., tf.newaxis] / 255.0, tf.float32),
             tf.cast(mnist_labels, tf.int64))
)
nsamples = len(list(dataset))
dataset = dataset.repeat().shuffle(10000).batch(args.batch_size)
mnist_model = tf.keras.Sequential([
    tf.keras.layers.Conv2D(32, [3, 3], activation='relu'),
    tf.keras.layers.Conv2D(64, [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')
])

# Horovod: adjust learning rate based on number of GPUs.
scaled_lr = args.lr * hvd.size()
opt = tf.optimizers.Adam(scaled_lr)

# Horovod: add Horovod DistributedOptimizer.
if (with_hvd):
    opt = hvd.DistributedOptimizer(opt)

# Horovod: Specify `experimental_run_tf_function=False` to ensure TensorFlow
# uses hvd.DistributedOptimizer() to compute gradients.
mnist_model.compile(loss=tf.losses.SparseCategoricalCrossentropy(),
                    optimizer=opt,
                    metrics=['accuracy'],
                    experimental_run_tf_function=False)

if (with_hvd):
    callbacks = [
        # Horovod: broadcast initial variable states from rank 0 to all other processes.
        # This is necessary to ensure consistent initialization of all workers when
        # training is started with random weights or restored from a checkpoint.
        hvd.callbacks.BroadcastGlobalVariablesCallback(0),

    # Horovod: average metrics among workers at the end of every epoch.
        #
        # Note: This callback must be in the list before the ReduceLROnPlateau,
        # TensorBoard or other metrics-based callbacks.
        hvd.callbacks.MetricAverageCallback(),

        # Horovod: using `lr = 1.0 * hvd.size()` from the very beginning leads to worse final
        # accuracy. Scale the learning rate `lr = 1.0` ---> `lr = 1.0 * hvd.size()` during
        # the first three epochs. See https://arxiv.org/abs/1706.02677 for details.
        hvd.callbacks.LearningRateWarmupCallback(initial_lr=scaled_lr, warmup_epochs=args.warmup_epochs, verbose=1),
    ]
else:
    callbacks=[]
    # Horovod: save checkpoints only on worker 0 to prevent other workers from corrupting them.

if hvd.rank() == 0:
    #tboard_callback = tf.keras.callbacks.TensorBoard(log_dir = args.log_dir, histogram_freq = 1, profile_batch = 2)
    callbacks.append(tf.keras.callbacks.ModelCheckpoint('./checkpoints/keras_mnist-{epoch}.h5'))
    #callbacks.append(tboard_callback)

# Horovod: write logs on worker 0.
verbose = 1 if hvd.rank() == 0 else 0

# Train the model.
# Horovod: adjust number of steps based on number of GPUs.
with tf.profiler.experimental.Profile(args.log_dir):
    mnist_model.fit(dataset, steps_per_epoch=nsamples // hvd.size() // args.batch_size, callbacks=callbacks, epochs=args.epochs, verbose=verbose)
t1 = time.time()
if (hvd.rank()==0):
    print("Total training time: %s seconds" %(t1 - t0))