-
**Describe the bug**
Since the Android version crashes immediately on all my devices, the WSL version is aliased and buggy and the Windows version doesn't compile, I was hoping the web version could …
-
### Describe the bug
When training GlowTTS, learning rate is too low if `scheduler_after_epoch=True`, which is the default. It seems like `NoamLR` is not working properly.
Here's an example te…
-
Because v1.4 was getting too big and some features had unknown timelines and finish dates, I made the decision to move the unsure features (based on when they would be added) into v1.5, the update tha…
-
### Describe the bug
Hi,
I am following the [Multispeaker training documentation](https://tts.readthedocs.io/en/latest/training_a_model.html#multi-speaker-training) on VITS model. So I added the…
-
### Describe the bug
When I train Glow TTS on LJSpeech Spanish set (angelina or victor) from [AI Labs](https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset/), the avg loss stays constant.
…
-
### Describe the bug
It appears that the shared VITS model trained on VCTK does not return the voice of speaker as specfied.
For example, when I specify a female voice (speaker index = p229), it…
-
### Describe the bug
After running the introduction example on a single speaker ljspeech, I switched to mailabs format where the speaker are derived from the folder structure. I am getting an except…
-
**Describe the bug**
check.sh fails to run ith the following compilation error
Click to expand check.sh output
```
+ cargo check --workspace --all-targets --all-features
Checking …
-
### Describe the bug
got debian on VM and installed TTS, tested text synthesiz and all works good, but as soon as I'm trying to train model following training guide for beginners, https://tts.read…
-
### Describe the bug
It seems like default number of warmup steps for lr scheduler in most recipes is wrong (or not?). It is set to 4000 for default `NoamLR` scheduler, however in Trainer, default …