batch-shipyard/recipes/PyTorch-GPU

PyTorch-GPU

This recipe shows how to run PyTorch on GPUs using N-series Azure VM instances in an Azure Batch compute pool. This sample executes the MNIST example.

Configuration

Please see refer to this set of sample configuration files for this recipe.

Pool Configuration

The pool configuration should enable the following properties:

• vm_size must be a GPU enabled VM size. Because PyTorch is a GPU-accelerated compute application, you should choose a GPU compute accelerated VM instance size.
• vm_configuration is the VM configuration. Please select an appropriate platform_image with GPU as supported by Batch Shipyard. VM instance size.

Global Configuration

The global configuration should set the following properties:

• docker_images array must have a reference to a valid PyTorch Docker image. pytorch/pytorch can be used for this recipe.

Jobs Configuration

The jobs configuration should set the following properties within the tasks array which should have a task definition containing:

• docker_image should be the name of the Docker image for this container invocation, e.g., pytorch/pytorch
• resource_files array should be populated if you want Azure Batch to handle the download of the training file from the web endpoint:
  • file_path is the local file path which should be set to main.py
  • blob_source is the remote URL of the file to retrieve: https://raw.githubusercontent.com/pytorch/examples/master/mnist/main.py
• command should contain the command to pass to the Docker run invocation. To run the MNIST example, the command would be: python -u main.py