def __init__(self, average=True):
"""
Args:
average (bool): whether to average or sum the gradients across processes.
"""
import byteps.tensorflow as bps
self.hvd = bps # BytePS has the same interface as Horovod
self.hvd.allreduce = bps.push_pull # https://github.com/bytedance/byteps/issues/8
# TODO bootstrap env vars
bps.init()
self.is_chief = bps.rank() == 0
self._local_rank = bps.local_rank()
self._rank = bps.rank()
self._average = average
self._compression = None
self._has_compression = False
logger.info("[BytePSTrainer] local rank={}".format(self._local_rank))
SingleCostTrainer.__init__(self)
def __init__(self, average=True):
"""
Args:
average (bool): whether to average or sum the gradients across processes.
"""
import byteps.tensorflow as bps
self.hvd = bps # BytePS has the same interface as Horovod
self.hvd.allreduce = bps.push_pull # https://github.com/bytedance/byteps/issues/8
assert os.environ.get("DMLC_ROLE", None) == "worker"
assert "DMLC_WORKER_ID" in os.environ and "DMLC_NUM_WORKER" in os.environ
bps.init()
self.is_chief = bps.rank() == 0
self._local_rank = bps.local_rank()
self._rank = bps.rank()
self._average = average
self._compression = None
self._has_compression = False
logger.info("[BytePSTrainer] local rank={}".format(self._local_rank))
SingleCostTrainer.__init__(self)
def main(_):
# BytePS: initialize BytePS.
bps.init()
# Keras automatically creates a cache directory in ~/.keras/datasets for
# storing the downloaded MNIST data. This creates a race
# condition among the workers that share the same filesystem. If the
# directory already exists by the time this worker gets around to creating
# it, ignore the resulting exception and continue.
cache_dir = os.path.join(os.path.expanduser('~'), '.keras', 'datasets')
if not os.path.exists(cache_dir):
try:
os.mkdir(cache_dir)
except OSError as e:
if e.errno == errno.EEXIST and os.path.isdir(cache_dir):
pass
else:
raise
# Download and load MNIST dataset.
(x_train, y_train), (x_test, y_test) = \
keras.datasets.mnist.load_data('MNIST-data-%d' % bps.rank())
# The shape of downloaded data is (-1, 28, 28), hence we need to reshape it
# into (-1, 784) to feed into our network. Also, need to normalize the
# features between 0 and 1.
x_train = np.reshape(x_train, (-1, 784)) / 255.0
x_test = np.reshape(x_test, (-1, 784)) / 255.0
# Build model...
with tf.name_scope('input'):
image = tf.placeholder(tf.float32, [None, 784], name='image')
label = tf.placeholder(tf.float32, [None], name='label')
predict, loss = conv_model(image, label, tf.estimator.ModeKeys.TRAIN)
# BytePS: adjust learning rate based on number of GPUs.
opt = tf.train.RMSPropOptimizer(0.001 * bps.size())
# BytePS: add BytePS Distributed Optimizer.
opt = bps.DistributedOptimizer(opt)
global_step = tf.train.get_or_create_global_step()
train_op = opt.minimize(loss, global_step=global_step)
hooks = [
# BytePS: BroadcastGlobalVariablesHook broadcasts 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.
bps.BroadcastGlobalVariablesHook(0),
# BytePS: adjust number of steps based on number of GPUs.
tf.train.StopAtStepHook(last_step=200000 // bps.size()),
tf.train.LoggingTensorHook(tensors={'step': global_step, 'loss': loss},
every_n_iter=10),
]
# BytePS: pin GPU to be used to process local rank (one GPU per process)
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
config.gpu_options.visible_device_list = str(bps.local_rank())
# BytePS: save checkpoints only on worker 0 to prevent other workers from
# corrupting them.
checkpoint_dir = './checkpoints' if bps.rank() == 0 else None
training_batch_generator = train_input_generator(x_train,
y_train, batch_size=100)
# The MonitoredTrainingSession takes care of session initialization,
# restoring from a checkpoint, saving to a checkpoint, and closing when done
# or an error occurs.
with tf.train.MonitoredTrainingSession(checkpoint_dir=checkpoint_dir,
hooks=hooks,
config=config) as mon_sess:
while not mon_sess.should_stop():
# Run a training step synchronously.
image_, label_ = next(training_batch_generator)
mon_sess.run(train_op, feed_dict={image: image_, label: label_})
def log(s, nl=True):
if bps.rank() != 0:
return
print(s, end='\n' if nl else '')
sys.stdout.flush()
def log(s, nl=True):
if bps.rank() != 0:
return
print(s, end='\n' if nl else '')
import tensorflow as tf
import byteps.tensorflow as bps
bps.init()
# BytePS: pin GPU to be used to process local rank (one GPU per process)
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[bps.local_rank()], 'GPU')
# Before launching, need to fist download the dataset to ~/.keras/datasets
(mnist_images, mnist_labels), _ = \
tf.keras.datasets.mnist.load_data(path='mnist-%d.npz' % bps.rank())
dataset = tf.data.Dataset.from_tensor_slices(
(tf.cast(mnist_images[..., tf.newaxis] / 255.0,
tf.float32), tf.cast(mnist_labels, tf.int64)))
dataset = dataset.repeat().shuffle(10000).batch(128)
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')
])
