Skip to content

关于多卡训练卡死的问题 #108

@SEUvictor

Description

@SEUvictor

对代码进行了修改以训练自己数据,发现多卡训练总是不能进行下去,单卡是没问题。调试发现是卡在了下面这一步。
msg = model.load_state_dict(checkpoint, strict=False)
image
请问是为什么呢?
训练命令为:
CUDA_VISIBLE_DEVICES=0,1 python -m torch.distributed.run --nproc_per_node 2 --master_port 12345 main.py

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions