Member-only story
AWS DeepRacer: What does ‘clone’ model do?
This is a quick post to help those who might be wondering about this question.
When training your models, after training and evaluating has been completed. You have an action option of ‘clone’.
Does this merely copy the configurations or does it build on top of the trained model?
The answer is the latter.
AWS documentation clear this up,
You can also learn how to clone a training model to extend a training session…
If you clone a previously trained model as the starting point of a new round of training, you could improve training efficiency.
An important thing to note if you intend to do this is mentioned in the very next line,
To do this, modify the hyperparameters to make use of already learned knowledge.
Check out this useful table by Daniel Gonzalez of the tradeoff’s for the hyperparametes:
That’s it. Good luck with your model tuning!
