JSPM

tensorflow-lambda

1.0.0
  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
    • 0
  • Score
    100M100P100Q21394F
  • License MIT

Package Exports

  • tensorflow-lambda

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (tensorflow-lambda) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

tensorflow-lambda

Usage

const loadTf = require('tensorflow-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.