Weekly update on envd (2022.06.27-2022.07.01)

[gaocegege]

⚠️ envd is still under heavy development, and subject to change. it is not feature-complete or production-ready. Please contact us in discord if there is any problem.

envd is a machine learning development environment for data science and AI/ML engineering teams.

pip install --pre --upragde envd

Out-of-box support for Julia is ready 🎉

Features

• Julia is supported https://envd.tensorchord.ai/docs/lang/julia
• A new command envd init is introduced, to ease the setup feat: add envd init #514

  NAME:
     envd init - Initializes the current directory with the build.envd file
  
  USAGE:
     envd init [command options] [arguments...]
  
  OPTIONS:
     --lang value, -l value  language usage. Support Python, R
  

• A new command envd context is introduced, to support remote build feat(CLI): Support context #512. context can be created by envd context create, e.g. envd context create --name test --builder-name test --use --builder kube-pod. With this feature, you can use the buildkit on Kubernetes or the remote docker daemon to build envd images.

    NAME:
    envd context - Manage envd contexts
  
    USAGE:
    envd context command [command options] [arguments...]
  
    COMMANDS:
    create   Create envd context
    ls       List envd contexts
    rm       Remove envd context
    use      Use the specified envd context
    help, h  Shows a list of commands or help for one command
  
    OPTIONS:
    --help, -h  show help (default: false)
  

Bug fixes

• fix: uid corrupted when run envd by root user #522
• Fix Julia multiple pkg installation bug #521

Feasibility research

• feat: Add envd-server runtime proposal #303
• feasibility-research(doc): Remote build with buildkit or kaniko #510