LPCNet_pytorch

LPCNet_pytorch copied to clipboard

help

18

请问数据处理后是这样么 data shape：(3200000000,) features shape：(180000000,) 这样执行下去代码会报错： Traceback (most recent call last): File "/Users/bytedance/Desktop/mywork/lpcnet/main.py", line 33, in dataloader = get_data(pcm_file, feature_file) File "/Users/bytedance/Desktop/mywork/lpcnet/utils.py", line 30, in get_data **features = np.reshape(features, (nb_frames*feature_chunk_size,...

sundekai

When generate nnet_data.c & nnet_data.h with current dump_lpcnet.py, the output of lpcnet_demo is bad, is there something wrong with dump_lpcnet.py?

220180928

Mdense layer without softmax and bad testing samplies

1

When checking the code, I find Mdense layer without softmax with a little different from the official code. And when testing, bad samples are generated. Have you ever met with...

Kristopher-Chen

1、在复现lpcnet时，执行test文件的时候最后会出现OverflowError: cannot convert float infinity to integer，具体经查是在进行了十几个frame之后就会出现的数值过大无法执行round，推知此时的采样值已经过大无法继续执行程序，检查程序与论文内容没有出入，不知道为什么会出现这样的问题 2、此前的程序当中训练时无法通过embbeding，经查并对照原文对train时getdata读取的数个data值添加了一步量化使之符合embbeding的参数要求，不知这样处理是否正确，如果不是的话是怎么处理的呢？

sgjh666