 Corentin Jemine
|
9edc87cf1e
Fixed missing __init__ file in utils
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
9ddc17dea7
Moved root of repo
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
3a62ffcd69
Cleanup
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
8671abe4c2
Wrote more about the synthesizer
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
21d470f467
Removed pretrained encoder models from the git
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
45464506f2
Integrated the vocoder
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
015868e1e9
Reduced overhead of using visdom
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
4f022decdc
Introduction done
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
871666a37b
Two thirds of the introduction done
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
9a2f1ddd22
Moved tacotron2 under SV2TTS
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
6190c6e934
Reimplemented the training process for the encoder
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
c6ab1141ee
Moved encoder to SV2TTS. Deleted old encoder trained models
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
be861eae02
Few changes regarding the synthesizing of the vocoder training data. Uploaded trained models of the vocoder to the git.
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
4cd60d96b6
Setup the training configuration for ASR
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
6ac1ff26c9
Included wave-rnn in the git, with trained models.
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
9e88e3b163
Cleaning up. Updated README
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
0e8571dbc1
First prototype
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
4116e06d80
Reverted relative imports, separated the encoder package and the encoder scripts instead. Improved on the umap demo
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
4d21f6b0b7
Renamed SV2TTS to encoder. Finished basic interfacing of the encoder. Changed all imports of the encoder to be relative imports in order to use it as a package
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
04169a38c8
Included Rayhane-mamah/Tacotron-2 to the git, begun injecting speaker embeddings in Tacotron
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
12ba55c06f
Removed backups from the git
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
dfa3a5c589
Backing up decent results
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
2f555e3c5a
Added a rough demo, put the saved models on the git
|
%!s(int64=5) %!d(string=hai) anos |
|
Corentin Jemine
|
6e1918f8f4
Implemented model saving and loading
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
048f981930
Data processing, data loading and data debugging procedures done
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
eb0406f0e8
Added Tacotron 1 & 2
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
510575cd11
Wrote section on Deep Voice 1
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
8f486f81de
Cleared some confusion on the HMM-DNN approach
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
953436e048
Cleanup
|
%!s(int64=6) %!d(string=hai) anos |
|
Corentin Jemine
|
bd52fdcf70
Initial commit
|
%!s(int64=6) %!d(string=hai) anos |
