VLLMCollectiveTransport

class torchrl.weight_update.llm.VLLMCollectiveTransport(master_address: str, master_port: int, rank: int | None, world_size: int, device: device | str | int | None = None, vllm_engine: Any | None = None)

Transport for vLLM using vLLM’s native WeightTransferConfig API (vLLM 0.17+).

This transport uses vLLM’s built-in NCCL weight transfer engine to broadcast weights from a trainer (rank 0) to vLLM workers (ranks 1+).

Parameters:

• master_address – Address of the master node for distributed init.

• master_port – Port of the master node for distributed init.

• rank – Rank of this process (0 for trainer, 1+ for vLLM workers).

• world_size – Total number of processes (1 + num_replicas * gpus_per_replica).

• device – Device to use for communication (typically cuda:0).

• vllm_engine – Optional vLLM engine reference (for receiver side).

check_connection() → bool

Check if the communication group is initialized.

init_all_workers_group(model_metadata: dict[str, tuple[dtype, Size]], gpus_per_replica: int | None = None)

Initialize the collective communication group using vLLM’s native API.

Parameters:

• model_metadata – Dict mapping param names to (dtype, shape) tuples.

• gpus_per_replica – GPUs per replica (for rank_offset calculation). Inferred if not provided.

receive_weights(timeout: float | None = None, *, weights: Any = None, model: Any = None, strategy: Any = None) → Any | None

Receive weights from broadcaster.

Returns:

None - vLLM handles weight application internally via native API.

send_weights(model_id: str, weights: Any) → None

Send weights to all workers using vLLM’s native weight transfer API.

Parameters:

• model_id – ID of the model (used for logging).

• weights – TensorDict or dict of weights to broadcast.