envd

Hey HN! We’re Allen and Ce. We’re building github.com/tensorchord/envd: a command-line tool that helps you create the container-based development environment for AI/ML. Creating development environments is not easy, especially with today's complex systems and dependencies. With everything from Python to CUDA, BASH scripts, and Dockerfiles constantly breaking, it can feel like a nightmare - until now! With envd, you can: • Quickly spin up container-based environments for your projects with a single command. • Reproduce environments exactly as you need - from setting up Python, CUDA, Tensorflow, etc. to installing project-specific dependencies. • Share environments easily with teammates. We built envd to solve our own struggles with managing AI/ML development environments. The tool handles all the gritty details of installing dependencies a

Hi all, I'm the maintainer of https://github.com/kubeflow and https://github.com/tensorchord/envd. I've been part of the kubeflow community for ~5 years. One of the hardest issues I have faced is keeping our development environments in sync. We are often eager to do something speedily on the cloud or debug independently on the MacBooks. This is where envd comes in. It's a command-line utility that helps you create and manage development environments locally, or in the cloud. I can't wait to see what you create with envd. Looking forward to your feedback and contributions to our project. Happy coding!

Recently, I found another tool built for machine learning environment: https://github.com/tensorchord/envd. It integrates with buildkit and can share the config like Python functions cross the team. This can reduce a lot complexity for data scientists.

I do not think Docker is the end game. Firecracker/kata containers neither. https://github.com/tensorchord/envd is my favorite container tool. It provides the build language based on Python and the build is optimized for this scenario. There may be more runtime and image build tools for both containers and VMs. And different tools may be designed for different purposes. That's the future I believe.

Remote development will be popular? Yes. But developing in a monolithic machine may be not. The development environment should be clean and isolated, and products like gitpod and coder is promising. Besides this, maybe you can have a look at https://github.com/tensorchord/envd and https://github.com/okteto/okteto

The devbox is awesome. We use https://github.com/tensorchord/envd to develop AI/ML applications. And it will be great if users can build and deliver devbox images with a CI/CD systems like GitHub Action.