[BUG]无法在windows上训练

[unlimit999]

看了下报错原因是不支持nccl，可否提供选项或者自动判断作业系统改用gloo?

[unlimit999]

暫時找到的解決方法...
.\fishenv\Lib\site-packages\lightning\fabric\utilities\distributed.py

找到這個函數

def _init_dist_connection(
    cluster_environment: "ClusterEnvironment",
    torch_distributed_backend: str,
    global_rank: Optional[int] = None,
    world_size: Optional[int] = None,
    **kwargs: Any,
) -> None:
    """Utility function to initialize distributed connection by setting env variables and initializing the distributed
    process group.

    Args:
        cluster_environment: ``ClusterEnvironment`` instance
        torch_distributed_backend: Backend to use (includes `nccl` and `gloo`)
        global_rank: Rank of the current process
        world_size: Number of processes in the group
        kwargs: Kwargs for ``init_process_group``

    Raises:
        RuntimeError:
            If ``torch.distributed`` is not available

    """
    if not torch.distributed.is_available():
        raise RuntimeError("torch.distributed is not available. Cannot initialize distributed process group")
    if torch.distributed.is_initialized():
        log.debug("torch.distributed is already initialized. Exiting early")
        return
    global_rank = global_rank if global_rank is not None else cluster_environment.global_rank()
    world_size = world_size if world_size is not None else cluster_environment.world_size()
    os.environ["MASTER_ADDR"] = cluster_environment.main_address
    os.environ["MASTER_PORT"] = str(cluster_environment.main_port)
    log.info(f"Initializing distributed: GLOBAL_RANK: {global_rank}, MEMBER: {global_rank + 1}/{world_size}")
    #torch.distributed.init_process_group(torch_distributed_backend, rank=global_rank, world_size=world_size, **kwargs)
    torch.distributed.init_process_group(backend=torch_distributed_backend, rank=global_rank, world_size=world_size, **kwargs)

    # On rank=0 let everyone know training is starting
    rank_zero_info(
        f"{'-' * 100}\n"
        f"distributed_backend={torch_distributed_backend}\n"
        f"All distributed processes registered. Starting with {world_size} processes\n"
        f"{'-' * 100}\n"
    )

將torch.distributed.init_process_group(torch_distributed_backend, rank=global_rank, world_size=world_size, **kwargs)這一行第一個參數torch_distributed_backend改成backend=torch_distributed_backend即可正常訓練

[Leng Yue]

目前已经实现了来着, 你可以把配置文件里面的 backend 改成 gloo, 或者用命令行覆盖的.

[xumason]

请问是在这个函数里，把所有的 backend 都改成 gloo 吗？