Skip to content
Branch: master
Find file History
cchung100m and wkcn [MXNET-1291] solve pylint errors in examples with issue no.12205 (#13848

* [issue_12205 - PART1]

* [issue_12205 - PART2]

* Update example/cnn_chinese_text_classification/

Co-Authored-By: cchung100m <>

* Update example/distributed_training/

Co-Authored-By: cchung100m <>

* unify the style

unify the style

* unify the style

unify the style

* Update

* unify the style here

unify the style here

* Remove the preposition 'with' in comment

Remove the preposition 'with' in comment 'Train module using Caffe operator in MXNet'

* Unify the style of comments

Unify the style of comments suggested by @sandeep-krishnamurthy

* add

* retrigger the CI
Latest commit f4ab2d7 Mar 7, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information. [MXNET-1291] solve pylint errors in examples with issue no.12205 (#13848 Mar 7, 2019

Distributed Training using Gluon

Deep learning models are usually trained using GPUs because GPUs can do a lot more computations in parallel that CPUs. But even with the modern GPUs, it could take several days to train big models. Training can be done faster by using multiple GPUs like described in this tutorial. However only a certain number of GPUs can be attached to one host (typically 8 or 16). To make the training even faster, we can use multiple GPUs attached to multiple hosts.

In this tutorial, we will show how to train a model faster using multi-host distributed training.

Multiple GPUs connected to multiple hosts

We will use data parallelism to distribute the training which involves splitting the training data across GPUs attached to multiple hosts. Since the hosts are working with different subset of the training data in parallel, the training completes a lot faster.

In this tutorial, we will train a ResNet18 network using CIFAR-10 dataset using two hosts each having four GPUs.

Distributed Training Architecture:

Multihost distributed training involves working with three different types of processes - worker, parameter server and scheduler.

Distributed training architecture

Parameter Server:

The parameters of the model needs to be shared with all hosts since multiple hosts are working together to train one model. To make this sharing efficient, the parameters are split across multiple hosts. A parameter server in each host stores a subset of parameters. In the figure above, parameters are split evenly between the two hosts. At the end of every iteration, each host communicates with every other host to update all parameters of the model.


Each host has a worker process which in each iteration fetches a batch of data, runs forward and backward pass on all GPUs in the host, computes the parameter updates and sends those updates to the parameter servers in each host. Since we have multiple workers to train the model, each worker only needs to process 1/N part of the training data where N is the number of workers.


Scheduler is responsible for scheduling the workers and parameter servers. There is only one scheduler in the entire cluster.

Moving to distributed training: contains code that trains a ResNet18 network using distributed training. In this section we'll walk through parts of that file that are unique to distributed training.

Step 1: Use a distributed key-value store:

Like mentioned above, in distributed training, parameters are split into N parts and distributed across N hosts. This is done automatically by the distributed key-value store. User only needs to create the distributed kv store and ask the Trainer to use the created store.

store = mxnet.kv.create('dist')

It is the job of the trainer to take the gradients computed in the backward pass and update the parameters of the model. We'll tell the trainer to store and update the parameters in the distributed kv store we just created instead of doing it in GPU of CPU memory. For example,

trainer = gluon.Trainer(net.collect_params(),
                        'adam', {'learning_rate': .001},

Step 2: Split the training data:

In distributed training (using data parallelism), training data is split into equal parts across all workers and each worker uses its subset of the training data for training. For example, if we had two machines, each running a worker, each worker managing four GPUs we'll split the data like shown below. Note that we don't split the data depending on the number of GPUs but split it depending on the number of workers.

Splitting data

Each worker can find out the total number of workers in the cluster and its own rank which is an integer between 0 and N-1 where N is the number of workers.

store = kv.create('dist')
print("Total number of workers: %d" % store.num_workers)
print("This worker's rank: %d" % store.rank)
Total number of workers: 2
This worker's rank: 0

Knowing the number of workers and a particular worker's rank, it is easy to split the dataset into partitions and pick one partition to train depending on the rank of the worker. Here is a sampler that does exactly that.

class SplitSampler(
    """ Split the dataset into `num_parts` parts and sample from the part with index `part_index`
    length: int
      Number of examples in the dataset
    num_parts: int
      Partition the data into multiple parts
    part_index: int
      The index of the part to read from
    def __init__(self, length, num_parts=1, part_index=0):
        # Compute the length of each partition
        self.part_len = length // num_parts
        # Compute the start index for this partition
        self.start = self.part_len * part_index
        # Compute the end index for this partition
        self.end = self.start + self.part_len

    def __iter__(self):
        # Extract examples between `start` and `end`, shuffle and return them.
        indices = list(range(self.start, self.end))
        return iter(indices)

    def __len__(self):
        return self.part_len

We can then create a DataLoader using the SplitSampler like shown below:

# Load the training data
train_data =, transform=transform),
                                      sampler=SplitSampler(50000, store.num_workers, store.rank))

Step 3: Training with multiple GPUs

Note that we didn't split the dataset by the number of GPUs. We split it by the number of workers which usually translates to number of machines. It is the worker's responsibility to split the partition it has across multiple GPUs it might have and run the training in parallel across multiple GPUs.

To train with multiple GPUs, we first need to specify the list of GPUs we want to use for training:

ctx = [mx.gpu(i) for i in range(gpus_per_machine)]

We can then train a batch like shown below:

# Train a batch using multiple GPUs
def train_batch(batch, ctx, net, trainer):

    # Split and load data into multiple GPUs
    data = batch[0]
    data = gluon.utils.split_and_load(data, ctx)
    # Split and load label into multiple GPUs
    label = batch[1]
    label = gluon.utils.split_and_load(label, ctx)

    # Run the forward and backward pass
    forward_backward(net, data, label)

    # Update the parameters
    this_batch_size = batch[0].shape[0]

Here is the code that runs the forward (computing loss) and backward (computing gradients) pass on multiple GPUs:

# We'll use cross entropy loss since we are doing multiclass classification
loss = gluon.loss.SoftmaxCrossEntropyLoss()

# Run one forward and backward pass on multiple GPUs
def forward_backward(net, data, label):

    # Ask autograd to remember the forward pass
    with autograd.record():
        # Compute the loss on all GPUs
        losses = [loss(net(X), Y) for X, Y in zip(data, label)]

    # Run the backward pass (calculate gradients) on all GPUs
    for l in losses:

Given train_batch, training an epoch is simple:

for batch in train_data:
    # Train the batch using multiple GPUs
    train_batch(batch, ctx, net, trainer)

Final Step: Launching the distributed training

Note that there are several processes that needs to be launched on multiple machines to do distributed training. One worker and one parameter server needs to be launched on each host. Scheduler needs to be launched on one of the hosts. While this can be done manually, MXNet provides the tool to make this easy.

For example, the following command launches distributed training on two machines:

python ~/mxnet/tools/ -n 2 -s 2 -H hosts \
    --sync-dst-dir /home/ubuntu/cifar10_dist \
    --launcher ssh \
    "python /home/ubuntu/cifar10_dist/"
  • -n 2 specifies the number of workers that must be launched
  • -s 2 specifies the number of parameter servers that must be launched.
  • --sync-dst-dir specifies a destination location where the contents of the current directory will be rsync'd
  • --launcher ssh tells to use ssh to login on each machine in the cluster and launch processes.
  • "python /home/ubuntu/dist/" is the command that will get executed in each of the launched processes.
  • Finally, -H hosts specifies the list of hosts in the cluster to be used for distributed training.

Let's take a look at the hosts file.

~/dist$ cat hosts 

'd1' and 'd2' are the hostnames of the hosts we want to run distributed training using. should be able to ssh into these hosts by providing just the hostname on the command line. For example:

~/dist$ ssh d1
Welcome to Ubuntu 16.04.3 LTS (GNU/Linux 4.4.0-1049-aws x86_64)

 * Documentation:
 * Management:
 * Support:

  Get cloud support with Ubuntu Advantage Cloud Guest:

0 packages can be updated.
0 updates are security updates.

Last login: Wed Jan 31 18:06:45 2018 from

Note that no authentication information was provided to login to the host. This can be done using multiple methods. One easy way is to specify the ssh certificates in ~/.ssh/config. Example:

~$ cat ~/.ssh/config 
Host d1
    port 22
    user ubuntu
    IdentityFile /home/ubuntu/my_key.pem
    IdentitiesOnly yes

Host d2
    port 22
    user ubuntu
    IdentityFile /home/ubuntu/my_key.pem
    IdentitiesOnly yes

A better way is to use ssh agent forwarding. Check this article for more details.

Here is a sample output from running distributed training:

$ python ~/mxnet/tools/ -n 2 -s 2 -H hosts --sync-dst-dir /home/ubuntu/cifar10_dist --launcher ssh "python /home/ubuntu/cifar10_dist/"
2018-06-03 05:30:05,609 INFO rsync /home/ubuntu/cifar10_dist/ -> a1:/home/ubuntu/cifar10_dist
2018-06-03 05:30:05,879 INFO rsync /home/ubuntu/cifar10_dist/ -> a2:/home/ubuntu/cifar10_dist
Epoch 0: Test_acc 0.467400
Epoch 0: Test_acc 0.466800
Epoch 1: Test_acc 0.568500
Epoch 1: Test_acc 0.571300
Epoch 2: Test_acc 0.586300
Epoch 2: Test_acc 0.594000
Epoch 3: Test_acc 0.659200
Epoch 3: Test_acc 0.653300
Epoch 4: Test_acc 0.681200
Epoch 4: Test_acc 0.687900

Note that the output from all hosts are merged and printed to the console.

You can’t perform that action at this time.