CatBoost RandomizedSearch stopping and finding out hyperparameters for it

Question

So I started running CatBoost's RandomizedSearch implementation and so far as I see from the log (which I know from verbose=1, down below is the log), there is one model that is the best one for a really long time, and I cannot wait anymore for it to finish the randomized search till the end, and I wanted to see here does anyone knows how I can somehow access hyperparameters of that model after shutting my Python script down. Is there any way to see that?

/catboost_log.txt
...
595:    loss: 7.3805087 best: 6.8218305 (130)   total: 9h 7m 51s    remaining: 4h 16m 27s
596:    loss: 7.3949953 best: 6.8218305 (130)   total: 9h 10m 11s   remaining: 4h 16m 12s
...

Answer 1

If you use jupyter notebook and set plot=True, then you will see parameter values on the plot. You can also look into file catboost_info/catboost_training.json, this file contains information about metrics + parameter values for every iteration of randomized_search.