Open Ohzyang opened 2 months ago
How to solve this error, is it a problem with torchaudio?And in the first few lines of the picture, why the model in the unet and vae of the cvssp/audioldm-l-full model requires the .safetensors file, but the download from huggingface is the .bin file
Which version of torchaudio are you using? I tested on my machine with torchaudio==0.13.1
, and there are no issues.
the version I use is 0.13.1,it is all the same with the requirements.I don't know what the reason is, But I did not install conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 Pytorch-CUDa =11.6 -c pytorch-c nvidia Since this command is not useful in autoDL, then I use pip install torch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1. Is there any problem with this
每天都要开心鸭 @.***
------------------ 原始邮件 ------------------ 发件人: "xinke-wang/ModaVerse" @.>; 发送时间: 2024年7月8日(星期一) 晚上6:28 @.>; @.**@.>; 主题: Re: [xinke-wang/ModaVerse] Failed to load audio: Failed to load audio from assets/media/audio/cat.wav (Issue #4)
image.png (view on web)
Which version of torchaudio are you using? I tested on my machine with torchaudio==0.13.1, and there are no issues.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
Are you able to run this code from the root dir of the Modaverse project?
python -c "import torchaudio; audio = torchaudio.load('assets/media/audio/cat.wav'); print(audio)"
The expected output should be
(tensor([[ 0.0000e+00, -2.9877e-13, -5.6765e-13, ..., 1.8055e-08,
5.9317e-08, 4.9721e-08],
[ 0.0000e+00, -4.9038e-13, -7.7776e-13, ..., 1.8055e-08,
5.9317e-08, 4.9721e-08]]), 16000)
it is not work. I use pip install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 instead of conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 Pytorch-CUDa =11.6 -c pytorch-c nvidia. Is there any problem with this?
每天都要开心鸭 @.***
------------------ 原始邮件 ------------------ 发件人: "xinke-wang/ModaVerse" @.>; 发送时间: 2024年7月8日(星期一) 晚上6:45 @.>; @.**@.>; 主题: Re: [xinke-wang/ModaVerse] Failed to load audio: Failed to load audio from assets/media/audio/cat.wav (Issue #4)
Are you able to run this code from the root dir of the Modaverse project? python -c "import torchaudio; audio = torchaudio.load('assets/media/audio/cat.wav'); print(audio)"
The expected output should be (tensor([[ 0.0000e+00, -2.9877e-13, -5.6765e-13, ..., 1.8055e-08, 5.9317e-08, 4.9721e-08], [ 0.0000e+00, -4.9038e-13, -7.7776e-13, ..., 1.8055e-08, 5.9317e-08, 4.9721e-08]]), 16000)
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
I am not sure what's wrong with your environment because I am not able to reproduce this issue on my side. Are you able to use torchaudio to load other audio formats, such as MP3, instead of WAV? Are there any error messages that appear?
it is also not work. My environment except conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.6 -c pytorch-c nvidia is different from yours, but everything else is exactly the same. Since the conda command is not useful, I used the pip command. Please tell me whether I need to re-install the environment with the conda command.
Original Email
From:"Xinyu Wang"< @.*** >;
Sent Time:2024/7/8 19:40
To:"xinke-wang/ModaVerse"< @.*** >;
Cc recipient:"Ohzyang"< @. >;"Author"< @. >;
Subject:Re: [xinke-wang/ModaVerse] Failed to load audio: Failed to load audiofrom assets/media/audio/cat.wav (Issue #4)
I am not sure what's wrong with your environment because I am not able to reproduce this issue on my side. Are you able to use torchaudio to load other audio formats, such as MP3, instead of WAV? Are there any error messages that appear?
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>