JNT
Hello,have you solved this problem? I also had this problem and it bothered me for a long time.
> 您好，我在训练前尝试了将所有训练数据归一化，但是在训练过程中还是会提示： Validation is in progress... 0%| | 0/10 [00:00
非常感谢您的回复。我将根据您提供的思路尝试解决这个问题。关于loss:训练loss是随epochs下降的。计算loss时return了三个值，比如：计算complex loss时： def calloss_cplxmse(output, source): # B 2 F T loss = 0 output_real, output_imag = output[:,0], output[:,1] source_real, source_imag = source[:,0], source[:,1] for i in range(output.shape[0]): loss_real = F.mse_loss(output_real[i],...
好的，感谢您的耐心回复，我是菜鸟所以问的比较多哈哈哈
Ok, thank you for your reply.
代码还没跑起来，是在查看代码的时候发现了这个情况。我粘贴了一张图片，不知道你能不能看到。 ![Uploading 16843791894827.png…]()
噢噢好的，谢谢你
I found the problem that caused the error. In the dsconv2d_cplx.py and dsconv2d_real.py file, you maybe use the decoder to restore the original input shape instead of the encoder? This...
aha, speech enhancement