ChenDRAG e27b5a26f3

Refactor PG algorithm and change behavior of compute_episodic_return (#319) ...

- simplify code
- apply value normalization (global) and adv norm (per-batch) in on-policy algorithms

2021-03-23 22:05:48 +08:00

..

__init__.py

refract test code

2020-03-21 10:58:01 +08:00

test_a2c_with_il.py

Refactor PG algorithm and change behavior of compute_episodic_return (#319)

2021-03-23 22:05:48 +08:00

test_c51.py

Add Timelimit trick to optimize policies (#296)

2021-02-26 13:23:18 +08:00

test_dqn.py

add logger (#295)

2021-02-24 14:48:42 +08:00

test_drqn.py

Remove reward_normaliztion option in offpolicy algorithm (#298)

2021-02-27 11:20:43 +08:00

test_il_bcq.py

add logger (#295)

2021-02-24 14:48:42 +08:00

test_pg.py

Refactor PG algorithm and change behavior of compute_episodic_return (#319)

2021-03-23 22:05:48 +08:00

test_ppo.py

Refactor PG algorithm and change behavior of compute_episodic_return (#319)

2021-03-23 22:05:48 +08:00

test_qrdqn.py

fix qvalue mask_action error for obs_next (#310)

2021-03-15 08:06:24 +08:00

test_sac.py

Remove reward_normaliztion option in offpolicy algorithm (#298)

2021-02-27 11:20:43 +08:00