Copying waveglow files @ 2fd4e63 (per collab synthesis.ipynb)

This commit is contained in:
Brandon Thomas committed 2021-06-02 22:26:55 -04:00
1 parent 7744f0e324
commit a1c77fb527
2 files changed
+3 -6

No files matched your search

+2 -2
View File
@@ -12,7 +12,7 @@ using only a single network, trained using only a single cost function:
maximizing the likelihood of the training data, which makes the training
procedure simple and stable.
Our [PyTorch] implementation produces audio samples at a rate of 1200
Our [PyTorch] implementation produces audio samples at a rate of 4850
kHz on an NVIDIA V100 GPU. Mean Opinion Scores show that it delivers audio
quality as good as the best publicly available WaveNet implementation.
@@ -85,7 +85,7 @@ with fused residual and skip connections.
[Glow]: https://blog.openai.com/glow/
[WaveNet]: https://deepmind.com/blog/wavenet-generative-model-raw-audio/
[PyTorch]: http://pytorch.org
[published model]: https://drive.google.com/open?id=1rpK8CzAAirq9sWZhe9nlfvxMF1dRgFbF
[published model]: https://ngc.nvidia.com/catalog/models/nvidia:waveglow_ljs_256channels
[mel-spectrograms]: https://drive.google.com/file/d/1g_VXK2lpP9J25dQFhQwx7doWl_p20fXA/view?usp=sharing
[LJ Speech Data]: https://keithito.com/LJ-Speech-Dataset
[Apex]: https://github.com/nvidia/apex
+1 -4
View File
@@ -61,15 +61,12 @@ def update_model(old_model):
_update_model_res_skip(old_model, new_model)
if hasattr(old_model.WN[0], 'cond_layers'):
_update_model_cond(old_model, new_model)
for m in new_model.modules():
if 'Conv' in str(type(m)) and not hasattr(m, 'padding_mode'):
setattr(m, 'padding_mode', 'zeros')
return new_model
if __name__ == '__main__':
old_model_path = sys.argv[1]
new_model_path = sys.argv[2]
model = torch.load(old_model_path, map_location='cpu')
model = torch.load(old_model_path)
model['model'] = update_model(model['model'])
torch.save(model, new_model_path)