jirong0214 comments

Repositories
Issues
Comments

Results 1 comments of


                                            jirong0214

Reshape and Concat 层也需要反向传播计算梯度吗？

好的谢谢，我理解了。另外，我使用train.py可以成功构建网络并训练，但是直接运行lib/network.py进行测试网络时，出现：**RuntimeError: expected device cuda:0 and dtype Float but got device cpu and dtype Float**的错误，错误指向了43行 output[:, :, i * Reshape_Concat_Adap.blocksize:(i + 1) * Reshape_Concat_Adap.blocksize, j * Reshape_Concat_Adap.blocksize:(j + 1)...