Maxime Chevalier-Boisvert
|
c4f68f309b
Implemented GoToDoor environment
|
7 éve |
Maxime Chevalier-Boisvert
|
28df92e70d
Fixed issues with run_tests.py, grid encode/decode
|
7 éve |
Maxime Chevalier-Boisvert
|
e7e870ce2d
Fixed issues with wrappers
|
7 éve |
Maxime Chevalier-Boisvert
|
360927639b
Added wrapper for one-hot string encoding. Fixed bugs in goto env.
|
7 éve |
Maxime Chevalier-Boisvert
|
116bfb2d4b
Completed implementation of goto env
|
7 éve |
Maxime Chevalier-Boisvert
|
67acc5ff18
Started work on goto env
|
7 éve |
Maxime Chevalier-Boisvert
|
a05dd9456f
Split out simple_envs.py
|
7 éve |
Maxime Chevalier-Boisvert
|
d4d83a9bf6
Added warning if cuda is disabled, to avoid silent failure
|
7 éve |
Maxime Chevalier-Boisvert
|
32535cc145
Changed reward range for fetch environment
|
7 éve |
Maxime Chevalier-Boisvert
|
e02777d81b
Added function to encode observations to FetchEnv
|
7 éve |
Maxime Chevalier-Boisvert
|
16698be044
Added 5x5 config for the fetch environment
|
7 éve |
Maxime Chevalier-Boisvert
|
80b3178610
Moved rl code to pytorch-rl. Fixed warnings. Fixed issue w/ flat obs.
|
7 éve |
Maxime Chevalier-Boisvert
|
c080bf08d8
Added _randPos utility function
|
7 éve |
Maxime Chevalier-Boisvert
|
f47e47272e
Added animation for door-key curriculum
|
7 éve |
Maxime Chevalier-Boisvert
|
dbecad9ad0
Added randomization to DoorKey envs
|
7 éve |
Maxime Chevalier-Boisvert
|
95448c1ebd
Added check in run_tests.py
|
7 éve |
Maxime Chevalier-Boisvert
|
25cc3b5253
Made it so agent can see what it is carrying in observations
|
7 éve |
Maxime Chevalier-Boisvert
|
5daf219a68
Added tests. Moved envs into own source files.
|
7 éve |
Maxime Chevalier-Boisvert
|
4c0fb7cf53
Update README.md
|
7 éve |
Maxime Chevalier-Boisvert
|
7edf1575f1
Update README.md
|
7 éve |
Maxime Chevalier-Boisvert
|
0d9c084e68
Merge pull request #2 from zach-nervana/master
|
7 éve |
Zach Dwiel
|
4cfe559b0e
include environments in install
|
7 éve |
Zach Dwiel
|
31132832e3
update setup.py to actually install the package
|
7 éve |
Maxime Chevalier-Boisvert
|
6092631168
Updated README
|
7 éve |
Maxime Chevalier-Boisvert
|
23d1b6b98d
Added smaller DoorKey environments, exploration bonus wrapper
|
7 éve |
Maxime Chevalier-Boisvert
|
18480bfef3
Changed empty env size
|
7 éve |
Maxime Chevalier-Boisvert
|
f2824c7687
Implemented count-based exploration system (intrinsic motivation)
|
7 éve |
Maxime Chevalier-Boisvert
|
c23608ec9a
Update README.md
|
7 éve |
Maxime Chevalier-Boisvert
|
acaec3c75a
Added figure for empty environment
|
7 éve |
Maxime Chevalier-Boisvert
|
491c79586b
Update README.md
|
7 éve |