forked from TensorLayer/tensorlayer3
37 lines
855 B
ReStructuredText
37 lines
855 B
ReStructuredText
API - Reinforcement Learning
|
|
==============================
|
|
|
|
We provide two reinforcement learning libraries:
|
|
|
|
- `RL-tutorial <https://github.com/tensorlayer/tensorlayer/tree/master/examples/reinforcement_learning>`__ for professional users with low-level APIs.
|
|
- `RLzoo <https://rlzoo.readthedocs.io/en/latest/>`__ for simple usage with high-level APIs.
|
|
|
|
.. automodule:: tensorlayer.rein
|
|
|
|
.. autosummary::
|
|
|
|
discount_episode_rewards
|
|
cross_entropy_reward_loss
|
|
log_weight
|
|
choice_action_by_probs
|
|
|
|
|
|
Reward functions
|
|
---------------------
|
|
.. autofunction:: discount_episode_rewards
|
|
|
|
Cost functions
|
|
---------------------
|
|
|
|
Weighted Cross Entropy
|
|
^^^^^^^^^^^^^^^^^^^^^^^^
|
|
.. autofunction:: cross_entropy_reward_loss
|
|
|
|
Log weight
|
|
^^^^^^^^^^^^^^
|
|
.. autofunction:: log_weight
|
|
|
|
Sampling functions
|
|
---------------------
|
|
.. autofunction:: choice_action_by_probs
|