Nick Chen

import change in newer version of python ```python3 from skimage.metrics import peak_signal_noise_ratio as psnr from skimage.metrics import structural_similarity as ssim ```

可以参考这个链接调整语速，不过只能加速不能降速 https://stackoverflow.com/a/60495058

调速也可以后期再调整，效果一般差距也不大。这里面有几个发音人效果不大好可能还是因为训练样本不够多，大厂的ai语音合成还有下面链接这一个的效果应该会好不少。 https://github.com/lturing/tacotronv2_wavernn_chinese

> Thanks for letting me know @zinuoli. The paper mentions that smoothL1 loss was used for training. But as I notice, in the code ([this line](https://github.com/CXH-Research/DocShadow-SD7K/blob/eb0789d46b1c7db79f4116ca65a1c3cfb58b5674/train.py#L47)), it is `MSEloss`. Did...

@baicenxiao I have retrained the model for 50 times and shared the corresponding [wandb log](https://wandb.ai/xuhangc/jung?nw=nwuserxuhangc) log for your reference, with the selection of highest performance. Upon analyzing the results, it...

What is the W and H in your config yml file, the default value of width and height should be 512.

> > links > > Same as your situation, have you found the dataset? No, I have also email the organizer but still no response

> > links > > Same as your situation, have you found the dataset? Hi, Dr.Peng has kindly shared the EBB dataset [here](https://github.com/JuewenPeng/BokehMe/issues/6).

@alloblue0 作者已经把数据集上传至kaggle了 https://www.kaggle.com/datasets/colorlabeilat/seathru-dataset