Kittipat Virochsiri bb88559350 Simplifying model API ...

Summary:
- Delete SingleQValue and AllActionQValues; returns tensor directly
- CategoricalDQN is now taking a ModelBase which returns 3-D tensor (batch_size, num_actions, num_atoms)
- FullyConnectedDQN's constructor has one more argument, num_atoms, which default to None. When it's not None, returns 3-D tensor like above
- DuelingQNetwork takes 3 sub networks: shared_network, value_network, advantage_network. The parametric version is split out to ParametricDuelingQNetwork for better readability.
- Introduces a subclass of Sequential conforming to ModelBase API. `input_prototype()` returns the value from the first module.

Reviewed By: czxttkl

Differential Revision: D21336534

fbshipit-source-id: 629c688b0de4754effb30cd61f4a6da3cca021ab

2020-05-07 15:09:36 -07:00

..

ml.rl.evaluation.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.models.rst

Simplifying model API

2020-05-07 15:09:36 -07:00

ml.rl.prediction.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.preprocessing.rst

update sphinx docs file (#163)

2019-10-14 23:23:09 -07:00

ml.rl.readers.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.simulators.rst

update sphinx (#146)

2019-07-18 16:03:39 -07:00

ml.rl.training.gradient_free.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.training.ranking.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.training.rst

Update docs

2019-10-28 23:15:36 -07:00

ml.rl.training.world_model.rst

update sphinx (#146)

2019-07-18 16:03:39 -07:00

ml.rl.workflow.rst

update sphinx (#146)

2019-07-18 16:03:39 -07:00

ml.rst

update sphinx (#146)

2019-07-18 16:03:39 -07:00

modules.rst

Add sphinx documentation. Convert markdown files to rst.

2019-04-19 10:42:57 -07:00