Create tracker for which modeling and evaluation runs we've done

[jmogarrio]

For isoscape generation, we want to record what dataset was used and what model (specifying a commit).

For isoscape evaluation, we want to record the isoscapes used, commit for the evaluation code, and evaluation results (precision, recall, etc.)

Important to keep track of hyperparameters, though if we store a commit that includes Colab metadata we can reference the hyperparameters in the commit.

[jmogarrio]

I've shared the following tracker, added a comment about it in the results/analysis log.