reference/generated/torchrl.trainers.algorithms.ReinforceTrainer

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

ReinforceTrainer

class torchrl.trainers.algorithms.ReinforceTrainer(*args, **kwargs)

REINFORCE (policy gradient with baseline) trainer implementation.

See also ReinforceTrainerConfig for the Hydra configuration counterpart.

Warning

This is an experimental/prototype feature. The API may change in future versions. Please report any issues or feedback to help improve this implementation.

This trainer implements the REINFORCE algorithm for training reinforcement learning agents, using a critic network as a baseline for advantage estimation (GAE by default, matching ReinforceLoss). It extends OnPolicyTrainer; see that class for the full list of keyword arguments, covering advantage estimation, replay-buffer wiring, collector weight synchronization and logging.

REINFORCE is a single-pass on-policy algorithm: each collected batch is consumed once. This trainer therefore defaults to 1 epoch per batch.

Examples

>>> # Basic usage with manual configuration
>>> from torchrl.trainers.algorithms.reinforce import ReinforceTrainer
>>> from torchrl.trainers.algorithms.configs import ReinforceTrainerConfig
>>> from hydra.utils import instantiate
>>> config = ReinforceTrainerConfig(...)  # Configure with required parameters
>>> trainer = instantiate(config)
>>> trainer.train()

Note

This trainer requires a configurable environment setup. See the configs module for configuration options.

compute_loss(sub_batch: TensorDictBase, method: str | None = None) → TensorDictBase | tuple[Any, ...]

Evaluate the configured loss through the active execution boundary.

load_from_file(file: str | Path, **kwargs) → Trainer

Loads a file and its state-dict in the trainer.

Keyword arguments are passed to the load() function for legacy torch checkpoints and unified components explicitly saved with the torch state-dict payload format. Unified checkpoints additionally accept strict to control missing or incompatible components. Arguments are ignored when CKPT_BACKEND=memmap.

Note

Unified state-dict components use TensorDict storage by default and do not invoke the pickle loader. For explicit torch payloads and CKPT_BACKEND=torch checkpoints, weights_only=True is the default for safer deserialization. Pass weights_only=False explicitly only if the state dict contains custom objects. On torch < 2.4 the default is weights_only=False because the weights-only unpickler of those versions cannot deserialize the torch.device instances contained in TensorDict state-dicts.

Note

Explicit torch payloads and CKPT_BACKEND=torch checkpoints use mmap=True by default. Pass mmap=False for legacy pre-zipfile torch.save files or file-like objects. On Windows the default is mmap=False because a mapped checkpoint keeps the file locked, preventing deletion or re-save.

Note

Unified checkpoint tensors are mapped to CPU by default. Pass an explicit map_location to select another device mapping.

Note

After restoring an independently registered policy component, the trainer synchronizes the collector once so local policy copies and remote workers observe the restored learner weights.

request_stop(reason: str | None = None) → None

Signal that training should stop at the next loop boundary.