wx-wei / harmof0 Goto Github PK

Interesting work! May you provide the training code for fine-tuning? I found the results on my datasets were noisy.

The RAP on the PTDU-TUG DataSet is 0.93, it is very high, I want to refer to this paper.
Could you show me the code how to calculate the RAP on the PTDU-TUG DataSet.

after install the repo, The results in test/a.f0. are not the same with yours in Readme
I got

HarmoF0$ head test/a.f0.txt 
# time frequency activation
0.000 0.000 0.000
0.010 0.000 0.000
0.020 0.000 0.000
0.030 0.000 0.000
0.040 0.000 0.000
0.050 0.000 0.000
0.060 45.586 0.000
0.070 106.869 0.039
0.080 58.270 0.034
```

During test, I got warnings 

```
DeprecationWarning: `np.float` is a deprecated alias for the builtin `float`. To silence this warning, use 
`float` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.float64` here.
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  eps=np.finfo(np.float).eps,
```

Does this mean the install failed? 
Waiting for your reply

大佬您好！我在阅读您文章的时候，发现您文章在阐述 MRDC-Conv的时候，是将基频和它的各次谐波进行了逐元素相加操作，我有些不太能理解，为什么要将基频和它的各次谐波相加呢，我不太能理解这样的含义，所以向您请教一下，希望得到大佬的回复，谢谢！

When loading checkpoint_mir-1k.pth in the /harmof0/checkpoints/ directory, the following error was thrown:

RuntimeError: Error(s) in loading state_dict for HarmoF0:
size mismatch for block_2.0.weight: copying a param with shape torch.Size([64, 32, 1, 3]) from checkpoint, the shape in current model is torch.Size([64, 32, 3, 3]).
size mismatch for block_3.0.weight: copying a param with shape torch.Size([128, 64, 1, 3]) from checkpoint, the shape in current model is torch.Size([128, 64, 3, 3]).
size mismatch for block_4.0.weight: copying a param with shape torch.Size([128, 128, 1, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).

It seems that checkpoint has different network dimensions from the default one. Is that checkpoint still available?