Kelsey

Hi and thank you for this example code. I am trying to solve the same MNIST problem by using a PyTorch model that has been converted with jit.script(). However, jit...

How would you train for BW extension?

2

I'm interested in training to convert 24 kHz mel spectrograms to 48 kHz waveforms (like HIFI-GAN2). Might not work without changing the architecture, but that's ok. How would you modify...

Hi, I ran the following testing code to convert .wav -> mel using librosa and then Univnet with pretrained checkpoint to do the inverse but the results were extremely bad....

Redundant ljspeech vits models

3

Hi all, I noticed that these two model tags link to the same download. Is there a pre-trained ljspeech vits model with space/pauses? kan-bayashi/ljspeech_tts_train_vits_raw_phn_tacotron_g2p_en_no_space_train.total_count.ave kan-bayashi/ljspeech_vits

Run from backend

1

Does anyone have (or can point me to) a workflow to just run the backend of this? I'm having serious issues with installing and running the js/react frontend when I...

Finetune success for accented English?

1

Has anyone had success fine-tuning on a non-US/UK accent, and if so could you share your training config? I fine-tuned 50 epochs with ~30 minutes good quality, Australian English but...

I tried to run the playground with default input text and pre-trained speaker, and got the following error: Error Something went wrong. Reason: CUDA error: device-side assert triggered\nCUDA kernel errors...

adapting the gentle model

1

Hello, I was hoping to train a model to do alignment with Gentle. I saw that you also tried this but had version mismatch issues. Do you have any ideas...