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

Restructure toolbox config `.toml` #44

Closed RobertTLange closed 3 years ago

RobertTLange commented 3 years ago

Restructure the mle_config.toml in a more elegant way. E.g. add the gcloud_credentials_path to the different resource categories (right now this is stored under gcp which doesn't make sense). Add a local category for running local jobs. Make it easy to add a new type of remote resource (e.g. AWS, Azure, etc.). Potentially differentiate between 3 classes with standardized variables:

  1. Local: For running jobs on different "small" workstations.
  2. Cluster: For running jobs on different clusters with their respective scheduling systems (SGE/Slurm).
  3. Cloud: For running jobs on cloud services (GCP, AWS, Azure).