Envd

Name:flytekitplugins-envd
Version:0.0.0+develop
Author:admin@flyte.org
Provides: flytekitplugins.envd
Requires: flytekit>=1.12.0
envd
Python:>=3.9
License:apache2
Source Code: https://github.com/flyteorg/flytekit/tree/master/plugins/flytekit-envd
  • Intended Audience :: Science/Research
  • Intended Audience :: Developers
  • License :: OSI Approved :: Apache Software License
  • Programming Language :: Python :: 3.9
  • Programming Language :: Python :: 3.10
  • Programming Language :: Python :: 3.11
  • Topic :: Scientific/Engineering
  • Topic :: Scientific/Engineering :: Artificial Intelligence
  • Topic :: Software Development
  • Topic :: Software Development :: Libraries
  • Topic :: Software Development :: Libraries :: Python Modules

envd is a command-line tool that helps you create the container-based development environment for AI/ML.

Environments built with envd provide the following features out-of-the-box:

  • Knowledge reuse in your team
  • BuiltKit native, build up to 6x faster
  • Smaller and leaner images

With flytekitplugins-envd, people easily create a docker image for the workflows without writing a docker file.

To install the plugin, run the following command:

pip install flytekitplugins-envd

Example

# from flytekit import task
# from flytekit.image_spec import ImageSpec
#
# @task(image_spec=ImageSpec(packages=["pandas", "numpy"], apt_packages=["git"], registry="flyteorg"))
# def t1() -> str:
#     return "hello"

This plugin also supports install packages from conda:

from flytekit import task, ImageSpec

image_spec = ImageSpec(
    base_image="ubuntu:20.04",
    python_version="3.11",
    packages=["flytekit"],
    conda_packages=["pytorch", "pytorch-cuda=12.1"],
    conda_channels=["pytorch", "nvidia"]
)

@task(container_image=image_spec)
def run_pytorch():
    ...