-
We need to replace the current fishspeech with better TTS model.
WIP Shortlist of Possible candidates:
- Amphion (https://github.com/open-mmlab/Amphion)
-
I have trin tamil text to speech with my own voice. everything working fine. But when i try train this error message received"
24s
2024-10-26 11:14:16.494389: E external/local_xla/xla/stream_execu…
-
Did a fresh install of AllTalk V2 and Streaming is not working Silly Tavern. Standard generation is working fine.
This is the error i get on the terminal
ERROR: Exception in ASGI applicat…
-
Just a note on poor xttsv2 performance on AMD with ROCm 6.2. Doesn't seem like an issue with AllTalk necessarily.
- EndeavourOS (Arch)
- AMD 7900 XTX
- Beta version
- Standalone version
Run…
-
To make ToirToiSe say the word "Hi." On the fastest setting, 1/2hr -- on 0% of my GPU is utilized, I can run the largest SD models whilst this inferencing is taking place. Any ideas on how to get the …
-
### Describe the bug
Similar to #3787, but also when running `xtts_v2` model with voice cloning (vocoder model), using `device='cpu'` results to the following error:
```
RuntimeError: CUDA error: …
-
Hi! First, i want to thank you for making this repo and sharing your experiences.
I am trying to train XTTS model for Persian language. I read the topics in this repo https://github.com/coqui-ai/TTS/…
-
### Describe the bug
I have fine tuned XTTS v2 model on my own data containing both long and short audios (with the following histogram showing duration in seconds on x-axis. Labels 'old' and 'new' r…
-
**Describe the bug**
Finetuning on GoogleColab fails to launch with the error shown below.
**To Reproduce**
Launch the Colab notebook either from the link in the wiki or using the notebook in th…
-
Hello,
I have seen that you have done some experiments finetuning Tortoise, XTTS and StyleTTS2.
May I ask you - is the quality of the StyleTTS2 finetunes much better than in case of XTTSv2?
…