统计模型flops和params

[caomao1111]

请问paddleseg如何统计模型的flops和params 我原来直接使用paddle.flops统计paddleclas里的模型是没问题的，但统计paddleseg模型时报错 "Your model must be an instance of paddle.nn.Layer or paddle.static.Program."

[shiyutang]

不知道你是统计哪个模型出错呢？ 预期是不会出错的，如果出错可以联系我们修复

[caomao1111]

统计的BiseNetV2

[shiyutang]

能否提供一个复现指令，或者复现脚本？ 我们复现并修复～

[caomao1111]

import paddle from paddleseg.models.bisenet import BiSeNetV2 net=BiSeNetV2 input_sizer=[1,3,512,512] paddle.flops(net, input_size=input_sizer, print_detail=True) 用这个统计paddleclas分类模式是没有出错的

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. Thank you for your contributions.