Comments (11)
@roodrallec , your notebook is excellent. I would like to include it as a runtime example for TTS-Cube, if this is ok with you. I'm actually working on a better model which uses g2p and global style tokens for English. I have some examples here, but it's still in progress. However, they seem better than the current model.
Also, I'm training a waveglow as a vocoder, to see if it provides better e2e results.
from tts-cube.
Sure, use it as you please. Waveglow samples sound awesome... Good luck!
from tts-cube.
Hi @roodrallec ,
I've just updated the examples for English and pushed the models. It would be great if you could update your Notebook to use this branch instead. You have to copy the nn and pnn networks from the models file (not just the encoder). Also, for synthesis, you need to specify the g2p model and speaker:
python3 cube/synthesis.py --target-sample-rate=16000 --g2p-model=data/models/en-g2p --input-file=../test.en.1 --output-file=../en/cat.test.en.1.wav --speaker=cat --use-gpu
The list of speakers is:
SPEAKER:cat
SPEAKER:clb
SPEAKER:lnh
SPEAKER:eey
SPEAKER:aew
SPEAKER:fem
SPEAKER:bdl
SPEAKER:ksp
SPEAKER:rms
SPEAKER:axb
SPEAKER:jmk
SPEAKER:ahw
SPEAKER:gka
SPEAKER:ljm
SPEAKER:rxr
SPEAKER:awb
SPEAKER:lj
SPEAKER:slp
SPEAKER:slt
SPEAKER:aup
from tts-cube.
@roodrallec : I actually changed your notebook a little. Thanks again for sharing. I would really like to include this as a tutorial, but I think it would be fair for you to do the pull request. This is my current version of code: tts-cube-test
from tts-cube.
Sure, which folder would you like it in?
from tts-cube.
The examples folder would be great. I think it would be ok to also link this from README.md
Thanks again!
from tts-cube.
I think I need push access to the repo, unless there's a magic way to create a PR without permissions from an open issue.
from tts-cube.
I just added you as a collaborator. Normally, you can always create a PR from your fork, if tou are not a collaborator on a repo.
from tts-cube.
Awesome, all done, thanks. I was going to use TTS cube as one of the components for my master thesis, but ran out of time unfortunately. Did you create TTS cube as part of a research project?
from tts-cube.
It was just for fun, I guess
from tts-cube.
Do you think this issue can be closed?
from tts-cube.
Related Issues (20)
- How to use the G2P model HOT 1
- What is BeeCoder? HOT 5
- Negative loss when training step2 HOT 19
- Pretrained text encoder for included IAF model HOT 2
- English model and hardware requirements HOT 26
- Training times HOT 1
- Is there any interest in providing a model trained in Brazilian Portuguese? HOT 2
- Integration with LPCNET HOT 1
- what is the present inference for generating 10sec audio using vocoder? HOT 2
- how to synthesize from melspectogram directly without using encoder HOT 3
- some words are missing during synthesizing HOT 5
- what should the development set's content be in a speech dataset and g2p? HOT 5
- colab notebook missing command to enter the github folder
- How to train non-English?
- How can I synthesize my own text to speech?
- cum folosesc vocea in romana?
- melgan vocoder is fast, let's integrate it? HOT 1
- Install fails on ARCH during pip command HOT 1
- Fine-tuning/Speaker adaptation HOT 15
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 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. 📊📈🎉
-
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 ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from tts-cube.