|
|
8f244455bd
|
Tweaked README
|
2024-04-03 17:53:56 +02:00 |
|
|
|
0b4956873c
|
Fixed README (command for full cov)
|
2024-04-01 00:35:28 +02:00 |
|
|
|
a2f3926ffa
|
Updated README
|
2024-04-01 00:18:00 +02:00 |
|
|
|
3fa8beb7ae
|
Make pca optional dependency
|
2024-04-01 00:17:50 +02:00 |
|
|
|
67033170cb
|
PCA as dependency
|
2024-04-01 00:10:58 +02:00 |
|
|
|
445eb7407e
|
Updated README
|
2024-04-01 00:10:51 +02:00 |
|
|
|
892b4b2377
|
Added icon
|
2024-04-01 00:10:40 +02:00 |
|
|
|
ada300fa63
|
Tell PCA to use mSqrt when using W2 proj
|
2024-03-30 14:44:20 +01:00 |
|
|
|
4831fbb7e1
|
Support Full Cov in Buffer
|
2024-03-30 14:43:13 +01:00 |
|
|
|
1321e47b81
|
Wrote tests
|
2024-03-30 14:42:21 +01:00 |
|
|
|
eed4363ddd
|
Moved README.md
|
2024-03-30 14:41:59 +01:00 |
|
|
|
85e9e1033d
|
Extended README
|
2024-03-14 17:35:07 +01:00 |
|
|
|
8d4f57a59d
|
Added tests
|
2024-03-14 17:34:59 +01:00 |
|
|
|
39c21ab6b9
|
Fix problem when env is provided just as id
|
2024-03-14 17:34:47 +01:00 |
|
|
|
1c3d3cf6cf
|
Fix typo 2
|
2024-03-09 13:58:56 +01:00 |
|
|
|
4a7fbc2470
|
Fix typo
|
2024-03-09 13:57:52 +01:00 |
|
|
|
e14aae90ac
|
Fix: drop n_envs for DIstributions that dont need to know
|
2024-03-09 13:55:04 +01:00 |
|
|
|
66b8d21398
|
Fix: Crash when no dist_kwargs defined
|
2024-03-09 13:51:38 +01:00 |
|
|
|
8f66a34c29
|
Tell distributions the n_envs (for e.g. Pink Noise pregen)
|
2024-03-09 13:46:23 +01:00 |
|
|
|
b2384e183c
|
Add support for VecEnvs
|
2024-03-09 12:33:20 +01:00 |
|
|
|
6e79fce9ae
|
Unify how init_std is passed into policy
|
2024-01-29 18:11:33 +01:00 |
|
|
|
5fa351db22
|
bugfix: policy_kwargs can be None
|
2024-01-26 13:03:03 +01:00 |
|
|
|
9db4e93fa0
|
Fix issue when policy_kwargs=None and TRPL used
|
2024-01-26 13:01:36 +01:00 |
|
|
|
e788e9f998
|
Fix: Buffer supplying weird shapes for mean and cov
|
2024-01-26 12:40:24 +01:00 |
|
|
|
53505a80ad
|
Another issue with clip_range=None
|
2024-01-23 18:32:22 +01:00 |
|
|
|
8bbb01504b
|
Fix bug when self.max_grad_norm=None
|
2024-01-23 13:32:38 +01:00 |
|
|
|
c67f78159b
|
clip_fraction calc broken for clip_fraction=None
|
2024-01-23 11:46:50 +01:00 |
|
|
|
54b841654b
|
Fix another issue when clip_range=None
|
2024-01-23 11:23:35 +01:00 |
|
|
|
252710489e
|
Fixed error on clip_range None
|
2024-01-23 09:22:58 +01:00 |
|
|
|
e3f4c511bf
|
Better default HPs for TRPL
|
2024-01-23 09:20:34 +01:00 |
|
|
|
5b9f8b028c
|
Minor bug fixes
|
2024-01-22 19:58:08 +01:00 |
|
|
|
3d2b7dfc8f
|
import metastable_projections corretly when accesing via string name
|
2024-01-16 15:35:18 +01:00 |
|
|
|
d5e53a87db
|
Expose TRPL class
|
2024-01-16 15:34:12 +01:00 |
|
|
|
a8930a897a
|
Fix typo: FrobeniusProjectionLay
|
2024-01-16 15:27:20 +01:00 |
|
|
|
0806a24036
|
Implement Importance Sampling for PCA
|
2024-01-16 15:13:06 +01:00 |
|
|
|
1fa66611a3
|
Implement Buffer that stores Distribution params
|
2024-01-16 15:12:25 +01:00 |
|
|
|
5ed5d32083
|
TRPL is da
|
2023-11-19 18:34:41 +01:00 |
|
|
|
865efe4221
|
Updating to new sb3 version
|
2023-11-19 18:34:15 +01:00 |
|
|
|
1c1a909d27
|
Porting to sb3=2.1.0
|
2023-11-11 12:14:30 +01:00 |
|
|
|
3e27ad3766
|
Use ssf also for pca
|
2023-09-07 21:07:46 +02:00 |
|
|
|
f0cd88365e
|
Support for seting act_fn as string
|
2023-09-05 19:12:26 +02:00 |
|
|
|
3c81a15630
|
Fix: Set initial logging std correctly
|
2023-08-22 01:18:44 +02:00 |
|
|
|
a7cbc660e0
|
Fix: Passed wrong args to pca propa net
|
2023-08-22 01:12:50 +02:00 |
|
|
|
0af83eb7dc
|
Make sure pca_policy knowns, we are using pca
|
2023-08-22 01:10:41 +02:00 |
|
|
|
d6c05668cb
|
fixes
|
2023-08-22 01:01:13 +02:00 |
|
|
|
5739f365ab
|
Tell PCA to predict log_std instead of std when using SAC
|
2023-08-22 00:57:34 +02:00 |
|
|
|
3f3912eed1
|
Bug Fix: Logging of std for PCA based SAC
|
2023-08-22 00:30:17 +02:00 |
|
|
|
e39f1573cf
|
Fix: missing brackets
|
2023-08-22 00:21:43 +02:00 |
|
|
|
e6071a546b
|
Fixes for SACPolicy
|
2023-08-22 00:20:42 +02:00 |
|
|
|
28a518fe9d
|
Fix: Referenced old SACPolicy
|
2023-08-22 00:19:13 +02:00 |
|
