search

mle-infrastructure / mle-toolbox

Lightweight Tool to Manage Distributed ML Experiments 🛠
https://mle-infrastructure.github.io/mle_toolbox/toolbox/
MIT License
3 stars 1 forks source link

Refactor `multi_runner` into `experiments` #45

Closed RobertTLange closed 3 years ago

RobertTLange commented 3 years ago

As of right now there are three subdirectories which are concerned with the different experiment types:

  1. experiment: Provides the 'shell' for submitting an experiment on different platforms.
  2. multi_runner: Provides helpers which launch single/multi seed/config runs.
  3. launch: Provides the outer entry point for the different experiment types. This is what is ultimately called in mle run.

I think it may make sense to either rename multi_runner to something that makes more sense or to merge it into experiments. We should also think about how this will integrate with the different cloud platforms and how we want to rsync/gsync files.

RobertTLange commented 3 years ago

Addressed in #49.