Skip to content

Latest commit

 

History

History
15 lines (12 loc) · 1.6 KB

DiffSinger.md

File metadata and controls

15 lines (12 loc) · 1.6 KB
Raw

Use DDSP Vocoders in DiffSinger (OpenVPI version)

Suppose you have already trained a model called exp/combsub-test/model_100000.pt using the code in this repository, run

python export.py -m exp/combsub-test/model_100000.pt --traced

This will create a .jit format model file in the same directory.

Then, move this .jit model file and the config.yaml together to the checkpoints/ddsp directory of the DiffSinger repository.

Finally, edit the configs/acoustic.yaml file in the DiffSinger repository to enable the DDSP vocoder. the details are:

  1. Set the vocoder option to DDSP.
  2. Set the vocoder_ckpt option to the path of the .jit model. An example may be checkpoints/ddsp/model_100000-traced-torch1.9.1.jit
  3. Check whether other mel related parameters match the parameters in the checkpoints/ddsp/config.yaml file. For the details, the audio_sample_rate,audio_num_mel_bins,hop_size,fft_size,win_size,fmin and fmax in the configs/acoustic.yaml need to match sampling_rate, n_mels, block_size, n_fft, win_length,mel_fmin and mel_fmax in the checkpoints/ddsp/config.yaml, respectively.

After doing all this, DiffSinger's default NSF-HiFiGAN vocoder has been replaced by your own trained DDSP vocoder, and you can perform preprocessing, training or inference normally.