Rongjiehuang
Loss nan occurred in the fastdiff training time
Hello, Huang ! Will there be a loss nan when training fastdiff without loading preweights? How to handle the parameters or adjust them if this happens.and the loss always nonconvergence。how to handle it？
hello it seems that you have run this experienment, i want to ask you that if you have the link about the LJspeech that have been preprocessed or if you have ever met this problem that "File "/root/tmp/FastDiff-main/data_gen/tts/vocoder_pre_align.py", line 23, in init self.pre_align_args = hparams['pre_align_args'] KeyError: 'pre_align_args'" when you train this on your own dataset? i will appreiate it if you can give a reply thannk u
Hello, I encountered the same issue as you. I wanted to ask if you've found a solution.