diff --git a/docs/training/weight_transfer/base.md b/docs/training/weight_transfer/base.md index 6c768c87fd..ace228b009 100644 --- a/docs/training/weight_transfer/base.md +++ b/docs/training/weight_transfer/base.md @@ -156,5 +156,6 @@ from vllm.distributed.weight_transfer.factory import WeightTransferEngineFactory engine = WeightTransferEngineFactory.create_engine( config=weight_transfer_config, parallel_config=parallel_config, + model=model, ) ``` diff --git a/tests/distributed/test_weight_transfer.py b/tests/distributed/test_weight_transfer.py index 467e3934a0..2df0d9e71c 100644 --- a/tests/distributed/test_weight_transfer.py +++ b/tests/distributed/test_weight_transfer.py @@ -291,7 +291,9 @@ def test_nccl_receive_sparse_weights_without_init_raises(): config = WeightTransferConfig(backend="nccl") parallel_config = create_mock_parallel_config() - engine = NCCLWeightTransferEngine(config, parallel_config) + engine = NCCLWeightTransferEngine( + config, parallel_config, MagicMock(spec=torch.nn.Module) + ) update_info = NCCLWeightTransferUpdateInfo( names=["w"], @@ -526,7 +528,9 @@ def inference_receive_sparse_tensor( parallel_config.data_parallel_rank = 0 parallel_config.data_parallel_index = 0 - engine = NCCLWeightTransferEngine(config, parallel_config) + engine = NCCLWeightTransferEngine( + config, parallel_config, MagicMock(spec=torch.nn.Module) + ) engine.init_transfer_engine( NCCLWeightTransferInitInfo( master_address=master_address,