Shortcuts

RayModuleTransformScheme

class torchrl.weight_update.RayModuleTransformScheme(strategy: str = 'tensordict')[source]

Weight synchronization for RayModuleTransform actors.

This scheme is designed specifically for updating models hosted within Ray actors, such as RayModuleTransform instances. It creates a transport that directly calls the actor’s weight update methods.

Parameters:

strategy (str) – The weight transmission strategy (“state_dict” or “tensordict”). Default is “tensordict”.

create_receiver() RayModuleTransformReceiver[source]

Create a specialized receiver for Ray actor communication.

create_sender() RayModuleTransformSender[source]

Create a specialized sender for Ray actor communication.

create_transport(pipe_or_context: Any) TransportBackend[source]

Create RayActorTransport for the given actor.

Parameters:

pipe_or_context – Either a Ray actor reference or a context object from which to extract the actor reference.

Returns:

RayActorTransport configured with the actor reference.

Docs

Access comprehensive developer documentation for PyTorch

View Docs

Tutorials

Get in-depth tutorials for beginners and advanced developers

View Tutorials

Resources

Find development resources and get your questions answered

View Resources