sungfeng-huang/meta-tts

Official repository of https://doi.org/10.1109/TASLP.2022.3167258. More up-to-date code is in "refactor" branch.

meta-tts's Issues

Question for Meta-update in Meta-TTS

I have a question about meta update in the outer loop of Meta TTS

I compare the algorithm of MetaTTS and the original of MAML, are they the same in the below red box

I can not find the sum of query loss of all tasks in the repo, can you help to show!

I refer to the implementation of meta-learning for mnist, it seems also sum query loss of all tasks:
https://github.com/learnables/learn2learn/blob/0b9d3a3d540646307ca5debf8ad9c79ffe975e1c/examples/vision/meta_mnist.py#L100

One more question is the MAML from L2L, what is the "out of space" mean:

Thank you SungFeng

What's the difference between the models in the "pretrained model" link?

Hi. The work is amazing. If I want to test the model, which pretrained model should I use? (Because I notice that the filenames in the pretrained model link are quite similar)
I also notice that in the demo page (Section 4.3), you only did parallel voice cloning with unseen speakers, have you tried testing with different text with these speakers?
Thank you very much.

PERSONALIZED LIGHTWEIGHT TEXT-TO-SPEECH: VOICE CLONING WITH ADAPTIVE STRUCTURED PRUNING

I have read your article“PERSONALIZED LIGHTWEIGHT TEXT-TO-SPEECH: VOICE CLONING WITH ADAPTIVE
STRUCTURED PRUNING”. May I ask if the code for this article can be published

update evaluation/README

您好，冒昧打扰，望海涵。
看到您的evaluation/README中有提到Utilizing Self-supervised Representations for MOS Prediction这篇论文，不确定是不是这篇已发布的代码：https://github.com/s3prl/s3prl/tree/master/s3prl/upstream/mos_prediction
如果是的话不知道您是否方便更新一下evaluation/README。
感激不尽。
敬颂春祺

Have you tried to finetune only a few parameters?

Hi, thanks for your amazing work! I noticed you finetune the whole decoder with other layers in your experiment. Have you ever tried to finetune only a few parameters? For example，only the last layer of the model? I want to know how it performs with very few trainable parameters.

LibriTTS-360 pretrained checkpoints

Hello, thanks for great work on MetaTTS paper and repo!

I've noticed there are "dev" configs for training on larger dataset including LibriTTS 360 and 500 subets. Do you have plans for releasing pretrained checkpoint and publishing results obtained with more data?

try HiFi-GAN

I noticed that you used MelGAN in this paper. Have you also tried using HiFi-GAN?

Is 4 positive samples and 4 negative samples enough for computing the threshold for asv acceptant rate?

Hi,

I noticed that in the evaluation/json/VCTK/pair.json you provide 4 positive samples and 4 negative samples per speaker in vctk dataset. I wonder that is it enough for computing a good threshold value for asv acceptant rate?

Thanks

incompatible package versions

感谢您的贡献。我是一名学生，想以您的Meta-TTS模型作为基线模型或者对照组来完成我的毕业论文，因此需要复现您的代码。但在复现过程中出现了一些预料之外的bug比如安装包不兼容情况等，如果可能的话，希望您提供requiremengts.txt中包的版本，或者我应该复现之前版本的代码（可以使用您的预训练模型）。我想在您的预训练模型上进行推理，如果您能给到我一些宝贵的建议我将感激不尽。

8gpu CUDA out of memory

I used 8 GPUs for training, and set shots and queries to 3, meta batchsizes is set to 8. And set batchsize in config/train/base.yaml is set to 48, grad_ acc_ Step is set to 8. But when the code runs, it still reports CUDA out of memory. What else do I need to change?