-
## 論文タイトル(原文まま)
HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis
## 一言でいうと
HiFi-GANは、効率的かつ高忠実度な音声合成を実現するために設計された生成対向ネットワーク(GAN)であり、最新のモデルと比較して高い品質と高速な合成を…
-
Hi authors,
I am trying to train low-bit-rate codecs for 8kHz sample rate audio, specifically targeting bit rates of 3.2 kbps, 1.8 kbps, and 1.2 kbps. I set the hop_size to 240 (which factors as 2 …
-
Hi,
I want to ask whether your code about training Mel-GAN vocoder is support multiple GPUs? In you paper, you use one single GPU training about 14 days. So I want to ask you whether we can use mu…
-
Hi @bigpon
Can you show the training process graph of 'symAD_libritts_24000_hop300' ? It is implemented through the tensorboard command. Because I found that my training graph is chaotic, as shown i…
-
step 1 torch2onnx:
```python3
#!/ssd4/exec/huangps/anaconda3/envs/melgan/bin/python
import torch
import torchvision
import numpy as np
from model.generator import Generator
from utils.hparams…
-
Hello, thanks for your nice implementation of mel-gan.
I guess mel-gan can be used as the universal vocoder, and I thought there were a mention about multi-speaker training scheme in the original p…
-
I'd like to train hifi-gan on a custom dataset with its own set of wav files. For this I need to generate the corresponding mel spectrograms, which the readme says can be done using Tacotron2 although…
-
Hello. I downloaded the pretrained modal `ljspeech v3.1` and when I try to run `python gen_forward.py --alpha 1 --checkpoint pretrained-forward_step90k.pt --input_text 'this is whatever you want it to…
-
Hi @p0p4k, thanks for making this repo!
I am currently trying to train a 44.1kHz English model, but my model is struggling with a rather high duration loss when compared against your TensorBoard lo…
w11wo updated
1 month ago
-
Hi
Thanks for this great implementation.
I trained Tacotron and used WaveRNN as vocoder and result was good. Now I want to use HiFi-GAN as vocoder so I cloned this project and run it with mel-spectr…