Comments (4)
Hi, thank you, nice to hear
At first, I will ask you a silly question: why do you need a multilingual model? Isn't an multispeaker English text-to-speech model enough for you? Btw I think I have a monolingual monospeaker English model trained using this code on my local machine. If you want, I can share it.
Your plan is ok, but:
a) It was, it actually does just spectrogram inversion, so it might somehow work anyway. You can also use a different model with publicly available pretrained weights (such as WaveGlow).
b-c) You do not need to use Common Voice, English is data rich and has a wide variety of multispeaker datasets. You can for example use LJ Speech and a subset of VCTK to replace CSS10 and CommonVoice. But you should make sure that these English examples are normalized similarly to the rest of the data used for training (i.e, the same punctuation usage, distriution of sentence lenghts and audio durations, ...)
The current code is not very fine-tuning-friendly, unfortunately
from multilingual_text_to_speech.
hey!
Thank you for getting back to me so quickly and for the feedback. I really appreciate it. I will try to go ahead and train the model on normalized data from the suggested datasets.
A part of the research team is working on using multi-speaker English text-to-speech model so I just wanted to explore a different path. That being said, if you could share the monolingual mono-speaker English model trained using this code on your local machine -- I'll go ahead and test it out to see if its produces the kind of results we want it to. And of course, if we end up using either model for the research, we will make sure to give you the appropriate credit in the paper.
Best,
Vic
from multilingual_text_to_speech.
Oh, I see. The english checkpoint is available here. It produces linear spectrograms (not mel), so you should use the Griffin-Lim algorithm for conversion to audio.
Hope it helps
from multilingual_text_to_speech.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
from multilingual_text_to_speech.
Related Issues (20)
- WaveRNN + CUDA 11 + pytorch==1.10.0 == grad norm is NaN? Help needed. HOT 3
- Continue training with an exiting checkpoint but with new speaker id HOT 2
- How to apply this tool to my custom Japanese data? HOT 3
- Adding support for windows sapi5 or android HOT 4
- Voice cloning attempts HOT 1
- Model is much slower on CPU HOT 6
- No softmax layer in the classifier? HOT 1
- Params.py issue
- torch version issue HOT 4
- can't run train.py HOT 1
- data HOT 3
- When I try to train it, I got the following error: HOT 6
- How "Pronunciation control" can be implemented? HOT 1
- batchnorm1D on padded values results in large activation scaling HOT 3
- Project dependencies may have API risk issues HOT 2
- about ยต and variances ฯ HOT 5
- Dataset with various sample rates and frequency bins HOT 1
- preprocess Error HOT 1
- Can we get a cloned voicie in Real Time ? HOT 1
- is the pretrained model support speech generation in Hebrew? HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from multilingual_text_to_speech.