qlearning_family

c51_learner

Distributional Reinforcement Learning (C51DQN) Paper link: http://proceedings.mlr.press/v70/bellemare17a/bellemare17a.pdf Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.c51_learner.C51_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

ddqn_learner

DQN with Double Q-learning (Double DQN) Paper link: https://ojs.aaai.org/index.php/AAAI/article/view/10295 Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.ddqn_learner.DDQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

dqn_learner

Deep Q-Network (DQN) Paper link: https://www.nature.com/articles/nature14236 Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.dqn_learner.DQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

drqn_learner

Deep Recurrent Q-Netwrk (DRQN) Paper link: https://cdn.aaai.org/ocs/11673/11673-51288-1-PB.pdf Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.drqn_learner.DRQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(batch_size, obs_batch, act_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

dueldqn_learner

DQN with Dueling network (Dueling DQN) Paper link: http://proceedings.mlr.press/v48/wangf16.pdf Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.dueldqn_learner.DuelDQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

perdqn_learner

DQN with Prioritized Experience Replay (PER-DQN) Paper link: https://arxiv.org/pdf/1511.05952.pdf Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.perdqn_learner.PerDQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)

qrdqn_learner

DQN with Quantile Regression (QRDQN) Paper link: https://ojs.aaai.org/index.php/AAAI/article/view/11791 Implementation: TensorFlow2

class xuance.tensorflow.learners.qlearning_family.qrdqn_learner.QRDQN_Learner(config: Namespace, policy: tensorflow.keras.Model, callback)

Bases: Learner

forward_fn(obs_batch, act_batch, next_batch, rew_batch, ter_batch)

learn(*inputs)

update(**samples)