57 commits

Author	SHA1	Message	Date
Antonin Raffin	8acac6b0f4	Update docstring	2020-02-03 18:31:13 +01:00
Antonin Raffin	16121cf2b8	Create OffPolicyRLModel	2020-02-03 18:18:41 +01:00
Antonin Raffin	5d4e73544c	Fix reset_num_timesteps	2020-01-31 13:16:28 +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
Antonin Raffin	a628354721	Refactor evaluation	2020-01-27 15:53:27 +01:00
Antonin Raffin	b66003cfb3	Add callback support	2020-01-27 14:32:31 +01:00
Antonin Raffin	44fce7c02a	Fix typing errors and typos	2020-01-22 17:17:12 +01:00
Antonin Raffin	9e250b6818	Build doc	2020-01-20 16:19:35 +01:00
Antonin Raffin	c542009641	Clean up code + bug fixes	2020-01-20 11:17:55 +01:00
Antonin Raffin	8831eff163	Unify evaluation	2020-01-07 14:00:03 +01:00
Antonin Raffin	57708a628c	Add value function for SDE + TD3	2019-12-17 15:01:08 +01:00
Antonin Raffin	1d6f9bf100	Add sample freq for SDE	2019-12-17 11:47:21 +01:00
Antonin Raffin	0117cc37f4	Merge branch 'master' into feat/sde-features	2019-12-05 16:33:41 +01:00
Noah Dormann	ff7c4d24f4	deleted types in train_actor td3	2019-12-05 08:46:26 +01:00
Noah Dormann	6928879f5a	Refactored doc-strings	2019-11-28 16:30:13 +01:00
Noah Dormann	e95858784a	Formatted all files	2019-11-28 15:38:04 +01:00
Noah Dormann	9ff59eaf3d	Added attribute self.policy_class to prevent errors when using self.policy as class	2019-11-28 15:25:01 +01:00
Noah Dormann	e26564e0ec	Added function for setting up any attributes that weren't saved and thus not loaded	2019-11-28 13:35:16 +01:00
Antonin Raffin	5483e02d1a	Add SDE support for SAC	2019-11-26 15:26:12 +01:00
Antonin Raffin	03d2ab10f8	Fix clipped action when adapting noise with TD3	2019-11-22 15:04:34 +01:00
Antonin Raffin	604a19fbc3	Cleanup + update doc	2019-11-22 13:33:12 +01:00
Noah Dormann	924ba9aea6	cleaned comments on model specific get and load functions	2019-11-21 16:50:59 +01:00
Noah Dormann	2d72f6d1b5	Added SAC, TD3, A2C ... Missing CEMRL	2019-11-21 16:46:53 +01:00
Noah Dormann	a7655ca6e1	Reformated every file with PEP 8 errors	2019-11-21 13:01:03 +01:00
Antonin Raffin	ad32aa60f3	Add sde scheduler	2019-11-18 16:03:08 +01:00
Antonin Raffin	5278a6f3f8	Testing off policy normalization	2019-11-14 14:35:00 +01:00
Antonin Raffin	da325a0ba7	Solve NaN issue and reduce number of parameters	2019-11-13 13:02:37 +01:00
Antonin RAFFIN	d725d01186	Normalize returns	2019-11-12 22:32:21 +01:00
Antonin Raffin	a08382faab	Add sde update for TD3	2019-11-12 18:37:13 +01:00
Antonin Raffin	db87e0d36a	Quick and dirty SDE version for TD3	2019-11-07 17:31:52 +01:00
Antonin Raffin	d67822718c	Add learning rate schedule	2019-10-28 16:47:13 +01:00
Antonin Raffin	ab64ff464e	Add tensorboard_log dummy arg	2019-10-14 11:09:22 +02:00
Antonin Raffin	b5656531d1	Enable logger for SAC/TD3 + refactor	2019-10-10 13:47:13 +02:00
Antonin Raffin	37ab9d10f1	Rescale actions and add action noise	2019-10-07 16:26:03 +02:00
Antonin RAFFIN	12f854e1aa	Fix learning starts	2019-10-01 21:56:37 +02:00
Antonin Raffin	b4dc9d4e4d	Add doc	2019-09-26 11:46:40 +02:00
Antonin Raffin	322399e8fe	Update collect rollout	2019-09-25 13:20:06 +02:00
Antonin Raffin	6bfbb7198a	Rename seed	2019-09-24 16:59:47 +02:00
Antonin Raffin	32648d9029	Add docstrings	2019-09-24 15:30:58 +02:00
Antonin Raffin	f4fe1362f0	Renaming	2019-09-24 14:53:03 +02:00
Antonin Raffin	d22caac616	Working SAC	2019-09-24 14:15:12 +02:00
Antonin RAFFIN	2469ff3859	Reformat	2019-09-21 17:17:09 +02:00
Antonin Raffin	0e727a5f72	Full compat for VecEnv + bug fixes for cuda	2019-09-20 16:43:19 +02:00
Antonin RAFFIN	cc4380eccd	Add eval env and clip vf	2019-09-19 17:18:41 +02:00
Antonin RAFFIN	26f0c8d8e5	Refactor buffer	2019-09-19 11:43:15 +02:00
Antonin Raffin	54dd7ea60d	Start PPO	2019-09-18 13:10:27 +02:00
Antonin Raffin	2a660e9a41	Update closer to original implementation for CEMRL	2019-09-12 15:38:15 +02:00
Antonin Raffin	f04754afec	Refactor for collecting rollout	2019-09-12 14:00:55 +02:00
Antonin Raffin	5e3a84d551	Refactor policies	2019-09-12 11:19:06 +02:00