-
Hi, I have seen this mentioned a couple times, but not really very touched upon. I notice that the TTS I train has trouble with the F0 of my female speaker (it glottalizes too much to the point where …
-
Running through your pre-trained models, I found that generated audio does not exactly match the input in duration length. For example,
```
wav, sr = load_wav(os.path.join(a.input_wavs_dir, filname…
-
1. document of [torch.nn.CTCLoss](https://pytorch.org/docs/stable/generated/torch.nn.CTCLoss.html#torch.nn.CTCLoss) says: the input is a **logarithmized probabilities**, obtained with `torch.nn.funct…
-
Do you have a pre-training model? I want to save time on training.
And what about your training hours with the epoch=100.
-
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 …
-
Great work! This singing synthesis effect is the best I have heard so far, and the paper is also very scientific. Will you release the code?
-
感谢开源!!最近在关注soft-vc,对于自有数据集建模有两个疑问:
1 自有数据集训练声学模型,用到的units来源,是来自于原作者预训练的hubert_soft吗?这个是基于英文数据集训练的,在做中文VC任务的时候,这个可以直接用于提取units吗?
2 此项目开源的databaker音色的声码器是通用的还是在databaker数据集上finetune过?
-
The hparams.py says `n_frames_per_step=1, # currently only 1 is supported`, but reduction window is very important for them model to pick up alignment. Using a reduction window can be considered as d…
-
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…
-
got prompt
[rgthree] Using rgthree's optimized recursive execution.
(80, 279)
Length of mel chunks: 3
Load model from: E:\VIDEO\ComfyUI_video\ComfyUI\custom_nodes\ComfyUI_wav2lip\Wav2Lip\checkpoin…