speedyspeech

speedyspeech copied to clipboard

Time for training teacher model is quite slow.

5

I have trained the teacher model from the scratch with 14900 utterances ~ 40 hours on the Vietnamese dataset. The model takes about 30 minutes for an epoch with GPU...

ductho9799

wav file quality - melgan vs griffinlim

2

used the model for training in another language. from the student model, the generated mel spectrogram seems to be close, but some bluriness is there.  tried...

dsplog

hi Thanks for you I would train with my dataset but when i use duration extractor i have a below error please help me padded_items = torch.stack([torch.nn.functional.pad(torch.as_tensor(x), pad= zeros +...

herokarimpoor

if sample_rate is 16k, the reference hyperparamerter hop_size, win_size ..... how to set

1

jiahong3837

The AudioDataset.py file in the dataset is for creating the datasets files. Is there a script that creates these files(filename|text|tokenized-text) or should it be done manually?

herokarimpoor

I have the sound and text but I don't know how much to give as a(input_model) model and what changes to apply on it.

herokarimpoor

hi can you help me How to prepare the dataset? or Auxiliary link thank for you

herokarimpoor

Bumps [certifi](https://github.com/certifi/python-certifi) from 2020.4.5.1 to 2022.12.7. Commits 9e9e840 2022.12.07 b81bdb2 2022.09.24 939a28f 2022.09.14 aca828a 2022.06.15.2 de0eae1 Only use importlib.resources's new files() / Traversable API on Python ≥3.11 ... b8eb5e9 2022.06.15.1...

dependabot[bot]

Bumps [pillow](https://github.com/python-pillow/Pillow) from 7.1.2 to 9.3.0. Release notes Sourced from pillow's releases. 9.3.0 https://pillow.readthedocs.io/en/stable/releasenotes/9.3.0.html Changes Initialize libtiff buffer when saving #6699 [@​radarhere] Limit SAMPLESPERPIXEL to avoid runtime DOS #6700 [@​wiredfool]...

dependabot[bot]

i read about the SSIM loss and my understanding is that it was originally designed for measuring the similarity between images. It is composed of three components: luminance, contrast, structural...

binbinxue