tfjs-lambda

Usage

const loadTf = require('tfjs-lambda')

const tf = await loadTf()

// ... do something with tf

How it works ?

The package contains a zipped and compressed version of all the dependencies and binaries needed to run @tensorflow/tfjs-node on AWS Lambda (these dependencies are built with Github Actions).

During cold start, the files are deflated in /tmp and required in your node program.

Motivation

@tensorflow/tfjs works with AWS Lambda but the main problem is that it is slow very slow when used in node. On the other hand, @tensorflow/tfjs-node is fast when used with node but it is >140mo and it does not fit under AWS Lambda's size limit (50mo) and it needs to be pre-compiled for lambda for it to work in a lambda environment.

I was looking for an easy way to use tensorflowjs with lambda and I couldn't find any, so I made this package.