stable-baselines3/tests
Adam Gleave 4fb8aec215
Update evaluate_policy type annotation to support policies as well as RL algorithms (#1146)
* Add PolicyPredictor protocol and use it in evaluate_policy

* Update changelog

* Move Protocol to type_aliases to avoid circular import

* Add test for evaluate_policy on BasePolicy

* Remove unused import

* Use typing_extensions

* Move typing_extensions to 3rd party

* Add version range (typing_extensions uses SemVer)

* Import Protocol from typing_extensions only on Python<3.8

Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com>

* Install typing_extensions only on Python<3.8

* Add missing sys import

* Fix import ordering

* Fix observation type hint in predict

Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com>
Co-authored-by: Quentin GALLOUÉDEC <gallouedec.quentin@gmail.com>
2022-11-03 15:36:19 +01:00
..
__init__.py Init: TD3 2019-09-05 17:29:41 +02:00
test_buffers.py Support for device=auto buffers and set it as default value (#1009) 2022-08-16 17:54:55 +02:00
test_callbacks.py Add progress bar callback and argument (#1095) 2022-10-06 18:17:31 +02:00
test_cnn.py Add skip option for VecTransposeImage and bug fix in frame stack (#700) 2021-12-23 17:12:49 +02:00
test_custom_policy.py Gym fixes - Follow up from #705 (#734) 2022-02-04 15:13:57 -08:00
test_deterministic.py Gym fixes - Follow up from #705 (#734) 2022-02-04 15:13:57 -08:00
test_dict_env.py Multiprocessing support for off policy algorithms (#439) 2021-12-01 22:30:09 +01:00
test_distributions.py Fix return type of evaluate_actions (#1118) 2022-10-14 17:45:28 +02:00
test_env_checker.py Gym fixes - Follow up from #705 (#734) 2022-02-04 15:13:57 -08:00
test_envs.py Raise error when observation keys and observation space keys don't match (#1047) 2022-09-05 14:54:58 +02:00
test_gae.py Upgrade code to Python 3.7+ syntax using pyupgrade (#887) 2022-04-25 13:01:38 +03:00
test_her.py Upgrade code to Python 3.7+ syntax using pyupgrade (#887) 2022-04-25 13:01:38 +03:00
test_identity.py Add timeout handling for on-policy algorithms (#658) 2021-11-16 17:19:16 +01:00
test_logger.py Fix duplicate key error in HumanOutputFormat (#1079) 2022-09-28 12:06:07 +02:00
test_monitor.py Automatically create missing directories of `filenames passed to ResultsWriter` (#1072) 2022-09-21 13:14:38 +02:00
test_predict.py Remove deprecated features and attributes (#1104) 2022-10-11 10:55:16 +02:00
test_run.py Remove deprecated features and attributes (#1104) 2022-10-11 10:55:16 +02:00
test_save_load.py Allow model trained with python3.7 to be loaded with python3.8+ without the custom_objects workaround (#1123) 2022-10-17 17:33:47 +02:00
test_sde.py Remove deprecated features and attributes (#1104) 2022-10-11 10:55:16 +02:00
test_spaces.py Handling multi-dimensional action spaces (#971) 2022-08-06 14:19:20 +02:00
test_tensorboard.py Support hparams logging to tensorboard (#984) 2022-08-22 22:06:54 +02:00
test_train_eval_mode.py Remove deprecated features and attributes (#1104) 2022-10-11 10:55:16 +02:00
test_utils.py Update evaluate_policy type annotation to support policies as well as RL algorithms (#1146) 2022-11-03 15:36:19 +01:00
test_vec_check_nan.py Upgrade code to Python 3.7+ syntax using pyupgrade (#887) 2022-04-25 13:01:38 +03:00
test_vec_envs.py Fixed unchecked None value in SubprocVecEnv (#808) 2022-04-12 16:05:40 +02:00
test_vec_extract_dict_obs.py Support for VecMonitor for gym3-style environments (#311) 2021-04-13 18:09:31 +02:00
test_vec_monitor.py Upgrade code to Python 3.7+ syntax using pyupgrade (#887) 2022-04-25 13:01:38 +03:00
test_vec_normalize.py Allow model trained with python3.7 to be loaded with python3.8+ without the custom_objects workaround (#1123) 2022-10-17 17:33:47 +02:00