tfjs-node-lambda

Why

It is difficult to run tensorflow in a lambda:

• Size constraints from the lambda environment (<50MB)
• Native code compilation for the lambda environment

This family of packages will handle that heavy lifting for you.

Versions

See the GitHub repository for version information. This package will always contain a matrix of the currently supported AWS Lambda Node environments and the @tensorflow/tfjs-node latest major versions.

Example matrix:

{
  "lambda": ["nodejs10.x", "nodejs12.x"],
  "tensorflow": ["1.7.4", "2.7.0"]
}

Example releases:

nodejs10.x-tf1.7.4.br
nodejs10.x-tf2.7.0.br
nodejs12.x-tf1.7.4.br
nodejs12.x-tf2.7.0.br

Using GitHub actions, these versions are updated daily around midnight. Lock your versions with the --save-exact flag. Double check tfjs-node-lambda and tfjs-node-lambda-releases are the same matching versions.

Installation

npm install --save --save-exact tfjs-node-lambda tfjs-node-lambda-helpers
npm install --save-dev --save-exact tfjs-node-lambda-releases @tensorflow/tfjs-node @tensorflow/tfjs

tfjs-node-lambda

import loadTf from 'tfjs-node-lambda';

Before running loadTf, you must ensure that the correct release is downloaded to the tmp dir. Use the tfjs-node-lambda-helpers to do it for you.

TODO: recognize tfjs-node.br as well as all release package names in the tmp dir.

JavaScript

const tf = await loadTf();

TypeScript

Due to having multiple supported versions of @tensorflow/tfjs-node and an irregular bundle method, you must specify the types:

const tf: typeof import('@tensorflow/tfjs') = await loadTf();

tfjs-node-lambda-helpers

See the npm package for more documentation.

Contributing

We welcome contributions!

Committing

We use husky git hooks to automate commitizen. Due to this issue, we recommend running git commit -m "any or empty message" and following the interactive mode. Thank you!