Models for testing training

Jump to bottom

Slin Lee edited this page Sep 10, 2020 · 1 revision

Too many episodes - https://help.pathmind.com/en/articles/4004788-getting-started
Converged - https://help.pathmind.com/en/articles/4318146-product-delivery
No learning - Try https://help.pathmind.com/en/articles/4004788-getting-started with reward = 0 or something that won't give the policy any information.
Max iterations - I'm not sure about this one but maybe try https://help.pathmind.com/en/articles/4004788-getting-started with reward = after[0] - before[0].
Max time - Try https://3.basecamp.com/3684163/buckets/11875773/documents/2065939584#__recording_2980337409 (HVAC Tuple.zip at the bottom of page)