多GPU下finetune报错

ZJU-lishuang / yolov5_prune

yolov5 prune，Support V2, V3, V4 and V6 versions of yolov5

Apache License 2.0

556 stars 136 forks source link

Closed thunder95 closed 2 years ago

thunder95 commented 2 years ago

torch.nn.modules.module.ModuleAttributeError: 'Model' object has no attribute 'module_list'

注释掉后yololayer也会报错

都注释掉后，finetune会出现map为NAN，请教下大佬怎么破啊

thunder95 commented 2 years ago

使用单卡训练的时候， soft的值很高，而且训练少量steps之后，就都变成nan了

ZJU-lishuang commented 2 years ago

多卡训练用torch.distributed.run

ZJU-lishuang commented 2 years ago

蒸馏有问题就先不要用蒸馏选项

thunder95 commented 2 years ago

@ZJU-lishuang 好的谢谢大哥好人一生平安