Kai

> Hi all, we provide the > > spk_id gender mos result in the following files [aishell3_gender.pdf](https://github.com/PaddlePaddle/PaddleSpeech/files/8916101/aishell3_gender.pdf) [VCTK_gender.pdf](https://github.com/PaddlePaddle/PaddleSpeech/files/8916102/VCTK_gender.pdf) > > speaker info from original datasets: [aishell3_spk-info.txt](https://github.com/PaddlePaddle/PaddleSpeech/files/9316002/aishell3_spk-info.txt) [vctk_speaker-info.txt](https://github.com/PaddlePaddle/PaddleSpeech/files/9316003/vctk_speaker-info.txt) 请问团队在训练tts的时候是不是并没有使用aishell3全部的说话人？比如 fastspeech2_aishell3_ckpt_1.1.0 我看...

> Hi @treya-lin @maociao, > > I could not replicate this error on my side. > > How much RAM do you guys have available on the computer that you...

> The quickest fix is to run > > ``` > conda install -c lilcom lilcom > ``` > > and then install lhotse. Hi @csukuangfj I ran this command,...

It can. At least worked when tested on my linux server. My server actually has gpu on it. To disable gpu, I added `os.environ["CUDA_VISIBLE_DEVICES"] = "-1"` before importing torch in...

> Yes it's a good idea but I need time to support it. Hi Any update on this? Is it possible to infer in batch now?

I also wanna know how much vram is required for using the models of different sizes

这个好像不是错吧，bitsandbyte总是在开头打这一段。后面脚本没有继续运行吗

> > 这个好像不是错吧，bitsandbyte总是在开头打这一段。后面脚本没有继续运行吗 > > 问题就是后面没有继续执行，卡一会就没了 我跑了一下我的环境里的，正常的话最后一句就是 ``` CUDA SETUP: Loading binary ... ``` 没有这句 ``` normalizer.cc(51) LOG(INFO) precompiled_charsmap is empty. use identity normalization. ``` 不过你说的它卡在那里是正常的，它做完CUDA SETUP 它就会自己继续运行了，完了就会停，不会打印别的东西，不会报告运行结束，结束后可以查看到结果文件。 你的案例里它停了以后也并没有生成 `--prediction_path`...

Your prompt audio has very strong background noise. If you denoise it, it should get better result. By the way, according to my very brief experience using this repo, only...

这个功能现在有了吗？