Comments (4)
What do you think it is... Christmas? 😄 Was finetuning not enough? haha
Multiple additional models might be complicated, the loader situation may get messy. Its not impossible, but Ill think around it. There are quire a few complexities to the loaders in the "multiple other models" scenario.
It would be relatively easy to do a "detect an extra model in this location and make an extra gradio radio box appear for that 1x model"... I think. But I am talking for 1x extra model in 1x location. Though still, Id need to think around that one.
Would 1x extra model work? or are you looking at literally 2 to ??? extra models?
Would you train the same model over and over (different voices), so just need that 1x model.
from alltalk_tts.
Initially I thought it'd be like a model selector in Auto1111's stable diffusion webui I suppose.
Would you train the same model over and over (different voices), so just need that 1x model.
Is that possible? So I can train a model with one voice, then use that finetuned model again with a different voice? If that's the case then I suppose a model selector may not be needed, depending on how much other voices effect the previous finetuning.
from alltalk_tts.
Ive just updated finetuning to not only compact your model, but also deal with all the model moving. Its creating a folder in the /models/ folder called trainedmodel
, now that its standardised in some way, I should be able to do something with the gradio interface that looks for a model existing in that folder and if so, gives you an extra loader option. But Ive just spent my day on a load of other bits and Im sure Ive got something else to do. So Ill get there.. but give me time :)
If you do want to compact your existing model go to this link #28
from alltalk_tts.
The model has to be stored in /models/trainedmodel/
This is the default location that finetuning now moves models to after training.
If a model is detected there when AllTalk starts up, it will add the additional loader type into the Gradio interface.
At some point Ill work on the finetuning, so that you can choose to use either the base model OR your already pre-trained model.
from alltalk_tts.
Related Issues (20)
- [GEN] Error during audio generation: 500: Exception while running subprocess: [Errno 2] No such file or directory: 'piper' HOT 1
- problem with Polish language HOT 7
- Standard TTS generation example from the documentation does not work HOT 4
- Need help :( HOT 1
- Support for metavoice-1B HOT 1
- Are you going to add latest Parler TTS large v1 and mini V1 models? HOT 3
- Loading Parlor at 16-bit HOT 1
- Training with ddp doesnt work hangs at init_process_group HOT 1
- Add an option to send a voice (file not name) with the streaming/generation request. HOT 1
- Add ability to request specific audio types HOT 4
- Having narrator enabled in dockerconfig.json results in a silent failure in a clean container HOT 5
- Torch error on text-generation-webui 'Apply requirements' while running at_setup.bat HOT 1
- Different voice for different emotion of character / narrator HOT 5
- API Issues HOT 7
- The "start_alltalk" batch file, but that pops up really briefly and then just disappears again. HOT 2
- AllTalk v2 BETA Download Details & Discussion HOT 4
- ImportError: DLL load failed while importing _imaging: The specified module could not be found. HOT 1
- Connection between SillyTavern & Alltalk works, but Error 500 occured: Model file not found. HOT 3
- DeepSpeed N/A in Gradio Alltalk_v2 HOT 1
- CUDA not detected/python errors when running finetune/diagnostics also throwing error HOT 1
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 alltalk_tts.