Policy Gradient

class PolicyGradient(model, lr)

Bases: parl.core.paddle.algorithm.Algorithm

__init__(model, lr)

Policy gradient algorithm

Parameters

• model (parl.Model) – model defining forward network of policy.

• lr (float) – learning rate.

learn(obs, action, reward)

Update model with policy gradient algorithm

Parameters

• obs (paddle tensor) – shape of (batch_size, obs_dim)

• action (paddle tensor) – shape of (batch_size, 1)

• reward (paddle tensor) – shape of (batch_size, 1)

Returns

shape of (1)

Return type

loss (paddle tensor)

predict(obs)

Predict the probability of actions

Parameters

obs (paddle tensor) – shape of (obs_dim,)

Returns

shape of (action_dim,)

Return type

prob (paddle tensor)