torchrl.trainers.algorithms.configs.collectors.CollectorConfig

class torchrl.trainers.algorithms.configs.collectors.CollectorConfig(create_env_fn: ConfigBase = '???', policy: Any = None, policy_factory: Any = None, frames_per_batch: int | None = None, total_frames: int = -1, init_random_frames: int | None = 0, device: str | None = None, storing_device: str | None = None, policy_device: str | None = None, env_device: str | None = None, create_env_kwargs: dict | None = None, max_frames_per_traj: int | None = None, reset_at_each_iter: bool = False, postproc: Any = None, split_trajs: bool = False, exploration_type: str = 'RANDOM', return_same_td: bool = False, reset_when_done: bool = True, interruptor: Any = None, set_truncated: bool = False, use_buffers: bool = False, replay_buffer: Any = None, extend_buffer: bool = False, trust_policy: bool = True, compile_policy: Any = None, cudagraph_policy: Any = None, no_cuda_sync: bool = False, weight_updater: Any = None, weight_sync_schemes: Any = None, weight_recv_schemes: Any = None, track_policy_version: bool = False, worker_idx: int | None = None, trajs_per_batch: int | None = None, trajs_per_write: int | None = None, _target_: str = 'torchrl.collectors.Collector', _partial_: bool = False)

Hydra configuration for Collector.

Every kwarg accepted by Collector.__init__ is exposed as a field here.