-
Hello, I have an issue after proceeding inference. After training every time when I try to use model from different checkpoints it gives the silent result.
Below you can see my spectrograms:
![inf…
-
Related to **Model/Framework(s)**
[Tacotron2-Waveglow/PyTorch]
**Describe the bug**
I try inference by pre-trained models is ok:
python inference.py --tacotron2 nvidia_tacotron2pyt_fp16_20190…
-
Hi
I would like to train fastpitch, but the filelists expect durations, mels, pitch, etc extracted from a tacotron model. But how does this work in the multi speaker scenario? Is it expected that w…
-
this umbrella issue tracks my current progress and discuss priority of planned TODOs. It has been closed since all objectives are hit.
# Goal
- [x] achieve a high quality human-like text to spe…
-
Hello everyone!
I know that question has already been asked multiple times, but anyway, DeepSpeech by Mozilla team has its feature, they use CTC as well, though. Therefore, are there any plans on i…
vcjob updated
3 years ago
-
I have recently started to try to implement text-to-speech synthesis with espnet library in my project and have succesfully generated individual audio files from loaded models. If I load two different…
-
Sorry for a noob question, what is the best (in terms of quality) english TTS available pretrained for today? Is it a following combination or there is something better?
1. Tacotron2 | char_train_n…
qo4on updated
3 years ago
-
Below code snippet is from E2E-TTS demo colab. When I load the vocoder, then results of the tacotron model changes. I think this may be something about pytorch, but I tried it in a different scenario.…
-
Using https://github.com/NVIDIA/tacotron2 inference.ipynb and the `default max_decoder_steps=1000`, I can make sentences that are about 5-10 words long with good quality.
Increasing this to =2000 …
-
Chinese TTS now available, thank @azraelkuan for his support :D. The model used Baker dataset here (https://www.data-baker.com/open_source.htmlt). The pretrained model licensed under CC BY-NC-SA 4.0 (…