Skip eval on train and test for self-reporting results

Feature request: allow users to skip eval on train and test

Evaluating on the training and test sets is time-consuming and not necessary for self-reporting results.
We should add a flag that allow the user to skip eval on these datasets, to make scoring faster.

Accordingly, in this scenario we should modify:

goals_reached = (
              train_state['validation_goal_reached'] and
              train_state['test_goal_reached'])

into:

goals_reached = (train_state['validation_goal_reached'])

This would speed up self-evalution even more, by stopping training when validation target is reached, avoiding unnecessary usage of computational resources.

Start by locating the training loop that computes goals_reached and the self-reporting evaluation path described in the issue. Trace how train, validation, and test evaluation are selected, then define the flag's scope and verify that enabling it skips train and test evaluation while stopping on the validation goal.