Commit Graph

19 Commits

Author SHA1 Message Date
Greg Brockman
8a535ca6f2 Switch to a global PRNG for action/observation spaces (#144)
cf 58e6aa95e5 (commitcomment-17669277)
2016-05-30 18:07:59 -07:00
Greg Brockman
58e6aa95e5 [WIP] add support for seeding environments (#135)
* Make environments seedable

* Fix monitor bugs

- Set monitor_id before setting the infix. This was a bug that would yield incorrect results with multiple monitors.
- Remove extra pid from stats recorder filename. This should be purely cosmetic.

* Start uploading seeds in episode_batch

* Fix _bigint_from_bytes for python3

* Set seed explicitly in random_agent

* Pass through seed argument

* Also pass through random state to spaces

* Pass random state into the observation/action spaces

* Make all _seed methods return the list of used seeds

* Switch over to np.random where possible

* Start hashing seeds, and also seed doom engine

* Fixup seeding determinism in many cases

* Seed before loading the ROM

* Make seeding more Python3 friendly

* Make the MuJoCo skipping a bit more forgiving

* Remove debugging PDB calls

* Make setInt argument into raw bytes

* Validate and upload seeds

* Skip box2d

* Make seeds smaller, and change representation of seeds in upload

* Handle long seeds

* Fix RandomAgent example to be deterministic

* Handle integer types correctly in Python2 and Python3

* Try caching pip

* Try adding swap

* Add df and free calls

* Bump swap

* Bump swap size

* Try setting overcommit

* Try other sysctls

* Try fixing overcommit

* Try just setting overcommit_memory=1

* Add explanatory comment

* Add what's new section to readme

* BUG: Mark ElevatorAction-ram-v0 as non-deterministic for now

* Document seed

* Move nondetermistic check into spec
2016-05-29 09:07:09 -07:00
Dominik
a50113a3a2 Handle draw in go env (#97) 2016-05-23 15:51:57 -07:00
Trevor Blackwell
f094e7d763 Cleanup pr/53 2016-05-10 17:29:02 -07:00
Trevor Blackwell
52dc50eacb Merge branch 'pr/53' 2016-05-10 17:26:15 -07:00
Trevor Blackwell
9287889214 Merge branch 'master' into pr/53
# Conflicts:
#	.gitignore
#	gym/envs/board_game/go.py
#	gym/monitoring/video_recorder.py
#	gym/scoreboard/client/api_requestor.py
2016-05-10 17:22:49 -07:00
Shu Shen
168f25b976 Fix Python3 compatibility in Go env (#64)
Under Python3, user inputs are in string format and needs be
converted
into bytes before passing to pachi-py. The returned __repr__ from
pachi-py for the boards are bytes and needs be converted into string
for
line breaks to work.

Also fixed raw_input and print functions in play_go under
examples/scripts with six package.
2016-05-09 19:18:55 -07:00
Botev
e7f7741548 Same fix for Go for StringIO 2016-05-03 20:13:05 +01:00
Botev
c62b2da337 A few compatibility issues with python3 resolved.
Fixed lot's of issues with the HexEnv.
Added Hex9x9-v0 to registry
2016-05-03 20:10:00 +01:00
Jonas Schneider
fd18ec3507 Merge remote-tracking branch 'origin/master' into py3
Conflicts:
	gym/envs/atari/atari_env.py
2016-04-29 18:07:47 -07:00
Jonathan Ho
d0132e6b4c Clean up resigning logic. Deals with agent resigns. 2016-04-29 15:48:53 -07:00
Jonathan Ho
330a8970a7 _pass_action and _resign_action methods 2016-04-29 15:37:11 -07:00
Jonas Schneider
f8e245d90d Correctly encode pachi_py arguments as binary 2016-04-28 22:32:05 -07:00
aforr
c8ada04b34 Fix victory conditions
Old victory condition was "player is white and white wins." Added "player is black and black wins" and "opponent resigned."
2016-04-28 20:30:50 -04:00
sergii-gavrylov
f6db41be75 minor fix with CreateBoard 2016-04-28 21:28:58 +02:00
Dominik Mueller
d3d23f3455 Fix readme example, documentation and error messages 2016-04-28 14:01:24 +02:00
Jonas Schneider
e8cf5f356c Fix missing StringIO import 2016-04-27 19:09:25 -07:00
Jonas Schneider
5065950a09 Fix Python3 compat of import dependencies 2016-04-27 18:03:29 -07:00
Greg Brockman
e8f2980603 Initial release. Hello world :). 2016-04-27 08:00:58 -07:00