Add automatic tracking of validation loss and an optional recommendation

I can access training_losses from report(mach) but I was surprised that there is no validation_losses (although these are certainly computed with a holdout or resampling strategy).

I understand that I can use callbacks for this but it felt asymmetric (and less friendly) to only find training_losses but no num_epochs_finished (although it can be derived from the log anyway) or validation_losses.

As for the optional recommendation, currently the argument controls allows two types of functionality:

• Specifying conditions to stop training
• Specifying callbacks to call while training

I think it's more intuitive to have two separate arguments. Maybe stop_conditions=[...] for the former and callbacks=[...] for the latter.

Start at the report(mach) entry point and inspect how training_losses and num_epochs_finished are assembled. Trace the holdout or resampling path and the controls handling for stopping conditions and callbacks. Done means the reporting additions and the proposed controls API separation have defined behavior and are covered by the project's existing checks.