Commit graph

114 commits

Author SHA1 Message Date
Antonin RAFFIN
353ea81080
Fix several VecEnv issues, add fork start method to tests (#43)
* Fix several VecEnv issues, add `fork` start method to tests

* Fix signature
2020-06-04 11:22:12 +02:00
Roland Gavrilescu
bb01253261
Tensorboard integration (#30)
* init commit tensorboard-integration

* Added tb logger to ppo (with output exclusions)

* fixed truncated stdout

* categorize stdout outputs by tag

* separated exclusions from values, added missing logs

* saving exclusions as dict instead of list

* reformatting, auto run indexing

* included renaming suggestions, fixed tests

* tb support for sac

* linting

* moved logging to base class

* tb support for td3

* removed histograms, non-verbose output working

* modifed changelog

* linting

* fixed type error

* moved logger config to utils

* removed episode_rewards log from ppo

* Enable tensorboard in tests

* Remove unused import

* Update logger sub titles

* Minor edit for PPO

* Update logger and tb log folder

* Pass correct logger to Callbacks

* updated docs

* added tb example image to docs

* add support for continuing training in tensorboard

* added tensorboard to docs index

* added tb test

* moved logger config to _setup_learn, updated tests

* accessing verbose from base class

* Update doc and tests

* Rename session -> time

* Update version

* Update logger truncate

* Update types

* Remove duplicated code

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>
2020-06-01 11:55:44 +02:00
Stelios Tymvios
78e8d405d7
Implemented Vectorized Action Noise (#34)
* Implemented Vectorized Action Noise

Vectorized Action Noise allows for multiple instances of
ActionNoiseProcesses to run in parallel. This makes it easier to
run TD3/SAC/DDPG with VecEnv.

* fixed linting issues

* make test function name consistent

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>

* sanity checks and more detailed test

* Update stable_baselines3/common/noise.py

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>

* Added assertion error message in noises setter

* Corrected tests to reflect change to AssertionError from ValueError

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>
2020-05-27 09:53:01 +02:00
Roland Gavrilescu
91adefdb4b
Support for MultiBinary / MultiDiscrete spaces (#13)
* multicategorical dist and test

* fixed List annotation

* bernoulli dist and test

* added distributions to preprocessing (needs testing)

* fixed and tested distributions

* added changelog and fixed ppo policy

* minor fix

* dist fixes, added test_spaces

* clean up

* modified changelog

* additional fixes

* minor changelog mod

* hot encoding fix, flake8 clean up

* lint tests

* preprocessing fix

* fixed bernoulli bug

* removed commented prints

* Update changelog.rst

* included suggested modifications

* linting fix

* increased space dim

* Update doc and tests

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>
2020-05-18 14:42:13 +02:00
Antonin RAFFIN
15ff6d47ee
Documentation update and style fixes (#21)
* Update doc: add gSDE

* Fix codestyle

* Remove travis script

* Add lint check to gitlab
2020-05-15 13:54:06 +02:00
Antonin RAFFIN
54f6f5b6fb
Add flake8 linter and Github CI (#19)
* Cleanup code

* Add flake8 lint and github workflow

* Update build matrix

* Relax precision for python3.7
2020-05-12 17:55:01 +02:00
Antonin RAFFIN
257a40ef4b
Add Gitlab CI (#12)
* Test gitlab-ci

* Try different image

* Add pytest and doc build

* Fix command

* Fix image used for CI

* Seperate pytest builds

* Fix weird seg fault in docker image due to FakeImageEnv

* Fix make command

* [ci skip] Add space in the badges

* Fix CI failures

* Re-install opencv

* Use opencv-headless

* Test with new docker image
2020-05-09 23:10:49 +02:00
Antonin RAFFIN
c20af230f7 Remove SDE support for TD3 2020-05-08 15:00:34 +02:00
Antonin RAFFIN
aa0ff8a59b Update atari test 2020-05-07 16:36:48 +02:00
Antonin RAFFIN
8046a24719 More doc + sync VecEnvs + atari 2020-05-07 16:08:23 +02:00
Antonin RAFFIN
cf1ae840c8 Sync identity envs 2020-05-05 16:52:22 +02:00
Antonin RAFFIN
04d85ac2e2 Fix import in tests 2020-05-05 16:32:08 +02:00
Antonin RAFFIN
2c34a4d694 Sync with Stable-Baselines 2020-05-05 16:28:38 +02:00
Antonin RAFFIN
d542732c8d Rename to stable-baselines3 2020-05-05 15:02:35 +02:00
Antonin RAFFIN
7ae54206ce Reformat and code cleanup 2020-04-23 15:18:21 +02:00
Antonin RAFFIN
f38ddcb278 Allow any number of channels 2020-04-22 16:11:23 +02:00
Antonin RAFFIN
f3cb0688c4 Fix custom optimizer 2020-04-22 13:21:11 +02:00
Antonin RAFFIN
041f2bc59a Cleanup, bug fixes + more tests 2020-04-22 13:14:22 +02:00
Antonin RAFFIN
73fb8d1c63 Add CNN support for TD3 2020-04-22 11:05:46 +02:00
Antonin RAFFIN
8aac9e819d Add VecTransposeImage and fix for SAC 2020-04-21 20:41:58 +02:00
Antonin RAFFIN
93c2a01f91 Start CNN support (failing for SAC) 2020-04-21 16:22:46 +02:00
Antonin RAFFIN
f347474e6a Independent save/load for policies 2020-04-20 15:59:44 +02:00
Antonin RAFFIN
aa1026ee87 Added `optimizer and optimizer_kwargs to policy_kwargs` 2020-04-17 15:13:45 +02:00
Antonin RAFFIN
71ce9ef2f4 Add test for actor 2020-03-31 18:26:26 +02:00
Antonin RAFFIN
c264403816 Rename for consistency
+ add _predict to actors
+ improve sac actor code
2020-03-31 17:48:23 +02:00
Antonin RAFFIN
2bbf6a9462 Minor: remove comment 2020-03-31 16:40:53 +02:00
Antonin RAFFIN
fdecd512db Add save/load weights for policies and refactor action distributions 2020-03-31 16:29:13 +02:00
Antonin RAFFIN
b782f3a208 Fix for test failures 2020-03-31 10:18:56 +02:00
Antonin RAFFIN
fa599c65a6 Add support for Discrete observation spaces 2020-03-25 16:42:05 +01:00
Antonin RAFFIN
4b2092f55a Remove base network 2020-03-23 15:31:14 +01:00
Antonin RAFFIN
dcb54b5301 Remove CEMRL 2020-03-23 14:48:38 +01:00
Antonin RAFFIN
fd9e73cfb8 Fix entropy computation 2020-03-19 10:19:48 +01:00
Antonin RAFFIN
9485b90a41 Sync predict with SB and add version file 2020-03-18 15:11:19 +01:00
Antonin Raffin
70e601c03c Improve code and bump version 2020-03-12 15:34:35 +01:00
Antonin Raffin
b64873ffff Sync callbacks 2020-03-12 12:34:25 +01:00
Antonin Raffin
18f38f8cf5 Reformat 2020-03-12 11:12:10 +01:00
Antonin Raffin
037986a91d Add test for expln 2020-03-11 16:35:13 +01:00
Antonin Raffin
4392759057 Comment unused code 2020-02-14 14:15:55 +01:00
Antonin Raffin
e31b139c47 Add test for predict method 2020-02-14 14:03:41 +01:00
Antonin Raffin
8b559d71ab Remove deprecated monitor format and improve tests 2020-02-14 13:42:16 +01:00
Antonin Raffin
f1a4fa2d3f Improve predict method 2020-02-12 15:25:05 +01:00
Antonin Raffin
7bafdb3a67 Add get_vec_normalize_env() 2020-02-12 11:34:29 +01:00
Antonin Raffin
2ce31c1e21 Fix entropy loss for squashed Gaussian and VecEnv seeding 2020-02-11 17:22:03 +01:00
Antonin Raffin
2afcf395b9 Update tests 2020-02-11 16:42:25 +01:00
Antonin Raffin
b7dcc8d58e Add extend method 2020-02-11 16:40:44 +01:00
Antonin Raffin
75a86881b3 Add save/load for replay buffer 2020-02-05 13:10:02 +01:00
Antonin Raffin
d850a35311 Update tests 2020-02-03 15:57:37 +01:00
Antonin Raffin
ec657cc34e Fix tests and change log_path behavior for EvalCallback 2020-01-31 13:42:04 +01:00
Antonin Raffin
6d59bfd4a0 Merge branch 'master' into feat/callbacks 2020-01-31 13:09:55 +01:00
Dormann, Noah
1f0dd60b97 Fix saving on GPU - Loading on CPU (#45)
* removed policy from save, changed th.loads to map to device

* found hack: catch pickle exception and trying th.load with mapping instead, otherwise raise exception with more information -> loading cuda on cpu raises exception -> leads to th.load with map being called

* deleted todo

* updated changelog

* start of saving refactor

* first working c

* all tests pass, save refactored

* - backwards compatibilty not always
- make pytest all passing
- make typing all passing

* Fixes and simplify the save method

* Remove unused param

* Fix backward compat

* Fix docstring
2020-01-31 13:06:55 +01:00