-
Typo somewhere: The readme says "The synthesizer model has been trained on my voice using Tacotron1." while the releases say "v0.1 Latest Pre-trained German synthesizer model based on tacotron2."
…
-
Related to Tacotron2
**Describe the bug**
amp.state_dict() in LoadCheckPoint and SaveCheckPoint error
**To Reproduce**
Steps to reproduce the behavior:
1. normal train step
2.It shows sav…
-
I have completed my training for espnet/ train_pytorch_tacotron2.yaml model and I want to train for 2nd stage that is for train_fastspeech.v2.yaml where can I find the receipe for it?
-
Hi,
the linked pretrained models like `https://issai.nu.edu.kz/wp-content/uploads/2021/04/kaztts_male1_tacotron2_train.loss.ave.zip` can't be downloaded. The links forward to the main page.
gree…
-
Hi, I found the synthesized speech files are different by running inference many times. Their style or intonation always are diverse. How to contral the style or intonation of the synthesized speech…
-
I trained korean tacotron2 with xvector model(espnet2 recipe) and neural vocoder(parallelwavegan repo). Now, trying to generate wav files from arbitrary reference audio, by modifying synth_wav.sh
w…
-
Hello!
You seem to have done quite a bit of vocoder comparisons. I have two questions based on your own personal experience.
- Which vocoder do you feel has the best overall quality (ignoring in…
Coice updated
2 years ago
-
Hi, I am trying to train Tacotron2 on a multi-speaker German dataset (4 speakers, 13 hours of audio material for now). I preprocessed it with the `thorsten` processor and adjusted the config files acc…
-
So I assume, the collab isn't actually finetuning. This makes sense because for some voices I can get close enough for it sounds good. But like let's say my voice. I have no idea why but it keeps maki…
-
Hello, I have a question.
I'm using a dataset with > 17k sentences (about 30 hours audio), 90% for training and 10% for validating.
It's been training for 3 days (using batch_size 8) and reaching …