Skip to content

JellyML is an open-source tool (python API and command line) for effortlessly embedding a snapshot of your code into a checkpoint of a pytorch (and pytorch lightning) model.

License

mmulet/jellyml

main
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 
 
 
 
 

JellyML

JellyML is an open-source tool (python API and command line) for effortlessly embedding a snapshot of your code | into a checkpoint of a pytorch model. Learn more at jellyml.com

pip install jellyml

Structure of the jellyml repository

(Note that the jellyml repository is a monorepo. If you are reading this from the python package source code, go to github.com/mmulet/jellyml to see the whole repository)

  • jellyml is the source for the python package
  • jellyml-lightning is the source for pytorch lightning plugin
  • client is the source for the website
  • dev_server is the source for the development server of the website

Build

jellyml

  1. Make a venv
python3 -m venv venv
# activate the venv ( depends on your shell and OS)
# see https://docs.python.org/3/library/venv.html
# bash
source venv/bin/activate
  1. Install build
pip install build
  1. Build the package
cd jellyml
python -m build
pip install dist/jellyml-0.0.1-py3-none-any.whl

jellyml-lightning

  1. Follow the directions for building and installing jellyml. jellyml is a dependency of jellyml-lightning.
  2. Build the package
cd jellyml-lightning
python -m build
pip install dist/jellyml-lightning-0.0.1-py3-none-any.whl

Website

Build the website

cd client
npm install .
cd ../dev_server
npm install .
npm run build

Dev the website

cd client
npm install .
cd ../dev_server
npm install .
npm run build

Tests

Located in the source files in src/jellyml. Have the prefix test_. Run them as a module

cd src;
python3 -m jellyml.test_all

About

JellyML is an open-source tool (python API and command line) for effortlessly embedding a snapshot of your code into a checkpoint of a pytorch (and pytorch lightning) model.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published