Skip to content

Latest commit

 

History

History
32 lines (27 loc) · 1.23 KB

README.md

File metadata and controls

32 lines (27 loc) · 1.23 KB
Raw

IceTray docker images for ML tasks

Supported

  • Ubuntu 20.04
  • Ubuntu 22.04

These images add machine learning tools to tagged version of icetray images, along with test-data, etc.

Two flavors are created:

  • icetray-cuda11.X - GPU enabled nvidia image, with copy of installed icetray, tools, etc (Ubuntu 20.04 and 22.04)
  • icetray-ml-base - Adds PyTorch, IceCube low-level-ml to icetray-cuda11 image to serve as base for ML tasks. (20.04 only)

Deploying

Uses docker github actions to build per docker/setup-buildx-action

Versions

  • 20.04 CUDA 11.6 icetray image:
    • icecube/icetray-ml:icetray-cuda11.6-vX.Y.Z-ubuntu20.04
    • icecube/icetray-ml:icetray-cuda11.6-vX.Y.Z-ubuntu20.04-YYYY-MM-DD
    • icecube/icetray-ml:icetray-cuda11.6-current-ubuntu20.04
  • 22.04 CUDA 11.8 icetray image:
    • icecube/icetray-ml:icetray-cuda11.8-vX.Y.Z-ubuntu22.04
    • icecube/icetray-ml:icetray-cuda11.8-vX.Y.Z-ubuntu22.04-YYYY-MM-DD
    • icecube/icetray:icetray-cuda11.8-current-ubuntu22.04
  • 20.04 IceCube ML base image:
    • icecube/icetray-ml:icetray-ml-base-vX.Y.Z-ubuntu20.04
    • icecube/icetray-ml:icetray-ml-base-vX.Y.Z-ubuntu20.0-YYYY-MM-DD
    • icecube/icetray-ml:icetray-ml-base-ubuntu20.04