recv¶

paddle.distributed. recv ( tensor, src=0, group=None, use_calc_stream=True ) [source]

Receive a tensor to the sender.

Parameters

• tensor (Tensor) – The Tensor to receive. Its data type should be float16, float32, float64, int32 or int64.

• src (int) – The source rank id.

• group (Group, optional) – The group instance return by new_group or None for global default group. Default: None.

• use_calc_stream (bool, optional) – Whether to use calculate stream or communication stream. Default: True.

Returns

None.

Examples

# required: distributed
import paddle
from paddle.distributed import init_parallel_env

init_parallel_env()
if paddle.distributed.ParallelEnv().rank == 0:
    data = paddle.to_tensor([7, 8, 9])
    paddle.distributed.send(data, dst=1)
else:
    data = paddle.to_tensor([1,2,3])
    paddle.distributed.recv(data, src=0)
out = data.numpy()