roll_batch#
- hybrid_aif.train.roll_batch(
- system,
- params,
- actor_m,
- critic_m,
- start_beliefs,
- keys,
- *,
- horizon,
- bounds,
- cont,
- code_and_carry,
- aux,
- expert_actions,
- sample_scoring=False,
A batch of sampled imagined trajectories: one
Rolloutwith a leading axis.Shared by
imagine()androllout_returns().actor_m/critic_mare passed explicitly so the gradient can substitute traced copies; the critic is stop-gradiented here, and so is the actor’s conditioning state at every step (the actor’s gradient flows through its reparameterised draw and the dynamics).