16 Commits
Author SHA1 Message Date
wassname d48ac96e23 misc 2018-07-08 10:20:56 +08:00
wassname 43a1b66e6d tidy 2018-06-25 05:39:35 +08:00
wassname c4105e7209 log reward before scaling 2018-06-24 13:07:09 +08:00
wassname bc7f039b13 fix logging 2018-06-24 13:06:58 +08:00
wassname de6793c373 logging and reduce batch size (for my 8GB gpu) 2018-06-24 13:06:44 +08:00
wassname ef74fb80b2 working with rudder 2018-06-23 19:39:36 +08:00
wassname 271a977610 local sonic, load on reset, backtrack 2018-06-23 18:51:48 +08:00
wassname c48d02a820 use rudder, and log 2018-06-23 18:43:49 +08:00
wassname 7c7ccd0c48 Only need ppo 2018-06-23 18:43:12 +08:00
Alex Nichol 0867d1fcfb lock to baselines version 2018-05-25 11:55:33 -04:00
Alex Nichol 89fb3595fd remove unused variable in jerk
fixes #3
2018-05-03 18:56:41 -07:00
Alex Nichol d616b9c50d fix docstring 2018-03-27 11:09:34 -07:00
Vicki Pfau 2b020a683d Use qualified image names 2018-03-20 10:57:11 -07:00
Alex Nichol 5417c4c69d new backronym 2018-03-16 15:30:26 -07:00
Alex Nichol 5a5976cdf1 drafted rainbow baseline
This is untested at present.
2018-03-14 11:34:43 -07:00
Alex Nichol 588b373ca4 initial JERK and PPO2 dumps 2018-03-02 14:40:36 -08:00