tensorplay.distributed.nn.functional.all_to_all_single

tensorplay.distributed.nn.functional.all_to_all_single(output, input, output_split_sizes=None, input_split_sizes=None, group=None)[source]