Horovod ? Tensor flow ? Uber ?

Horovod is a distributed training framework for TensorFlow. The goal of Horovod is to make distributed Deep Learning fast and easy to use.


The primary motivation for this project is to make it easy to take a single-GPU TensorFlow program and successfully train it on many GPUs faster. This has two aspects:

  1. How much modifications does one have to make to a program to make it distributed, and how easy is it to run it.
  2. How much faster would it run in distributed mode?

Internally at Uber they found the MPI model to be much more straightforward and require far less code changes than the Distributed TensorFlow with parameter servers. See the Usage section for more details.

Bench marking

Horovod is fast. Below is a chart representing the benchmark that was done on 32 servers with 4 Pascal GPUs each connected by RoCE-capable 25 Gbit/s network

Horovod achieves 90% scaling efficiency for both Inception V3 and ResNet-101, and 79% scaling efficiency for VGG-16. See the Benchmarks page to find out how to reproduce these numbers.

To use Horovod, make the following additions to your program:

  1. Run hvd.init().
  2. Pin a server GPU to be used by this process using config.gpu_options.visible_device_list. With the typical setup of one GPU per process, this can be set to local rank. In that case, the first process on the server will be allocated the first GPU, second process will be allocated the second GPU and so forth.
  3. Wrap optimizer in hvd.DistributedOptimizer. The distributed optimizer delegates gradient computation to the original optimizer, averages gradients using allreduce or allgather, and then applies those averaged gradients.
  4. Add hvd.BroadcastGlobalVariablesHook(0) to 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. Alternatively, if you're not using MonitoredTrainingSession, you can simply execute the hvd.broadcast_global_variables op after global variables have been initialized.
  5. Modify your code to save checkpoints only on worker 0 to prevent other workers from corrupting them. This can be accomplished by passing checkpoint_dir=None to tf.train.MonitoredTrainingSession if hvd.rank() != 0.

Example (see the examples directory for full training examples)

import tensorflow as tf
import horovod.tensorflow as hvd
# Initialize Horovod
# Pin GPU to be used to process local rank (one GPU per process)
config = tf.ConfigProto()
config.gpu_options.visible_device_list = str(hvd.local_rank())
# Build model...
loss = ...
opt = tf.train.AdagradOptimizer(0.01)
# Add Horovod Distributed Optimizer
opt = hvd.DistributedOptimizer(opt)
# Add hook to broadcast variables from rank 0 to all other processes during
# initialization.
hooks = [hvd.BroadcastGlobalVariablesHook(0)]
# Make training operation
train_op = opt.minimize(loss)
# Save checkpoints only on worker 0 to prevent other workers from corrupting them.
checkpoint_dir = '/tmp/train_logs' if hvd.rank() == 0 else None
# 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) as mon_sess:
while not mon_sess.should_stop():
# Perform synchronous training.

See full training simple and advanced examples.


As a parting note I wish you to remain productive, and optimize your tools as much as you can (without using this as an excuse not to work )