Brandon Rozek
|
66496fe0d8
|
Changes from honors thesis
|
2020-03-23 20:02:06 -04:00 |
|
Brandon Rozek
|
a44b981e55
|
Small import error fix
|
2019-11-17 18:39:28 -05:00 |
|
Brandon Rozek
|
8dd9ca617e
|
Incorporated concepts from the paper "Deep Q-Learning From Demonstrations"
|
2019-11-17 18:36:35 -05:00 |
|
Brandon Rozek
|
744656aaa9
|
Updated configs and fixed threading issues
|
2019-11-05 07:09:49 -05:00 |
|
Brandon Rozek
|
32862e4d79
|
Began separating config & networks, F1 for pausing, text functions, and more sneaky agent stuff
|
2019-10-27 20:42:37 -04:00 |
|
Brandon Rozek
|
d78892e62c
|
SneakyTrain uses separate replay buffer
Scripts were cleaned up considerably and comments were added
|
2019-10-23 21:53:20 -04:00 |
|
|
|
b7aa4a4ec6
|
Updated GymInteract to introduce a form of hidden training between showing the human play the game and the computer
|
2019-10-20 09:06:42 -04:00 |
|
|
|
1bf2c15542
|
Back and forth between computer play and human play while training an agent
|
2019-09-21 19:03:00 -04:00 |
|
