W&B: Publish metrics

Refs #297

added P1 label

assigned to @mblanco

Demo by:

1. adding "wandb": {"project": "wandb-integration"} to configs/tests.json and using teklia-dan train --config configs/tests.json
2. adding "wandb": {"project": "wandb-integration", "id": "ns2mnl1u", "resume": "must"} and updating "max_nb_epochs": 8 in configs/tests.json and using teklia-dan train --config configs/tests.json again

changed the description

added 1 commit

• 6819edfe - Update documentation

Compare with previous version

changed the description

changed the description

added 1 commit

• d2156927 - FIx lint

Compare with previous version

added 1 commit

• 4bfd378e - Nit

Compare with previous version

requested review from @starride

I tested the code, and it works as expected

• "wandb": {"project": "wandb-integration"} → creates a wandb run with id 9ns6hj6e
• "wandb": {"project": "wandb-integration", "id": "9ns6hj6e", "resume": "must"} → resumes the previous run
• "wandb": {"project": "wandb-integration", "id": "8ns6hj6e", "resume": "must"} → raises an error (incorrect run id)

wandb.errors.UsageError: You provided an invalid value for the `resume` argument. The value 'must' is not a valid option for resuming a run (8ns6hj6e) that does not exist. Please check your inputs and try again with a valid run ID. If you are trying to start a new run, please omit the `resume` argument or use `resume='allow'`.

• "wandb": {"project": "wandb-integration", "resume": "must"} → raises an error (missing run id)

wandb.errors.UsageError: You provided an invalid value for the `resume` argument. The value 'must' is not a valid option for resuming a run (dah1f98s) that does not exist. Please check your inputs and try again with a valid run ID. If you are trying to start a new run, please omit the `resume` argument or use `resume='allow'`.

• "wandb": {"project": "wandb-integration", "id": "8ns6hj6e"} → creates a wandb run with id 8ns6hj6e

approved this merge request

merged

mentioned in commit 516b99c7