Gradient actions

This page provides an introduction to Gradient Actions.

caution

Workflows are in beta, so try them out and let us know what you think. Hold off on using them for production use cases, as we can't promise bug fixes or new features just yet.

Overview

Gradient Actions are composable building blocks for creating reproducible machine learning Workflows. Actions use the uses and with syntax to specify how a job step executes.

container

uses: container@v1
with:
  image: bash:5
  args: ["echo", "hello", "world"]

The Gradient Action called container@v1 allows you to use an arbitrary Docker container image (in this case the lightweight bash container image) and pass arguments directly to it.

script

uses: script@v1
with:
  script: |-
    echo 'hello world'
    echo $RANDOM
  image: bash:5

If you want to run multiple commands, the script@v1 action allows you to pass a script in a literal-style HereDoc denoted by |-. The pipe character will preserve newlines and the dash will remove extra newlines after the block.

note

The image you provide will need to have bash available in its PATH.

git-checkout

outputs:
  repo:
    type: volume
uses: git-checkout@v1
with:
  url: https://github.com/user/my-public-repo
  ref: 46aa59d6ecc3720ffe2454a6d9d360e6ce75acce # Optional git ref
  path: /outputs/repo # Optional, defaults to exactly one output volume or dataset

In this example, the Gradient Action git-checkout@v1 clones the public GitHub URL https://github.com/user/my-public-repo at ref 46aa... into a volume named repo. The cloned files are accessible at /outputs/<output-name> (in this case, /outputs/repo), and subsequent jobs that specify the checkout job's volume as an input can also access the repository files as read-only at /inputs/<input-name>.

inputs:
  repo: checkout-job.outputs.repo
uses: container@v1
with:
  image: busybox
  args: ["ls", "/inputs/repo"]

Note: To clone a private repository, add your username as a parameter, set a Gradient secret with a GitHub access token value, and add a password parameter:

outputs:
  repo:
    type: volume
uses: git-checkout@v1
with:
  url: https://github.com/user/my-private-repo
  username: paperspace
  password: secret:MY_SECRET_NAME

You can also use path to pick an output target:

outputs:
  repo:
    type: volume
  ds:
    type: dataset
    with:
      ref: my-dataset
uses: git-checkout@v1
with:
  url: https://github.com/user/my-public-repo
  ref: 46aa59d6ecc3720ffe2454a6d9d360e6ce75acce # Optional git ref
  path: /outputs/repo/subfolder

s3-download

outputs:
  s3:
    type: volume
uses: s3-download@v1
with:
  url: s3://bucket/path/
  access-key: MYACCESSKEY
  secret-access-key: secret:MY_SECRET_NAME

The s3-download@v1 Gradient Action copies the contents of an Amazon S3 bucket into an output (in this example, the volume is named s3). Subsequent jobs that specify an input that reference the s3-download job's volume output can access the downloaded files within that job at /inputs/<input-name>.

Note: access-key and secret-access-key are required parameters, and the latter must be a Gradient secret. Optional parameters include region (for AWS buckets), endpoint (for non-AWS buckets), and path (to disambiguate target outputs or to download to a subfolder).

model-create

inputs:
  model:
    type: dataset
    with:
      ref: dsr8k5qzn401lb5:klfoyy9 # Example dataset ref
outputs:
  model-id:
    type: string
uses: create-model@v1
with:
  name: my-model-name
  type: Tensorflow # Tensorflow, ONNX, or Custom

In this example, the create-model@v1 action takes a dataset input named model and outputs a string ID (named model-id) that references a Gradient model. With this reference, the created model can be tested, edited, or deployed in future jobs.