-
Notifications
You must be signed in to change notification settings - Fork 178
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
install is broken #60
Comments
I have tried re-installing it like 5x now... each time, it messes up somehow and will only use the CPU for inference. DEPRECATION: pesq is being installed using the legacy 'setup.py install' method, because it does not have a 'pyproject.toml' and the 'wheel' package is not installed. pip 23.1 will enforce this behaviour change. A possible replacement is to enable the '--use-pep517' option. Discussion can be found at pypa/pip#8559 × Running setup.py install for pesq did not run successfully. note: This error originates from a subprocess, and is likely not a problem with pip. × Encountered error while trying to install package. note: This is an issue with the package mentioned above, not pip. |
C:\Users\chlyw\Desktop\tortoise-tts-fast\tortoise\utils\audio.py:19: WavFileWarning: Chunk (non-data) not understood, skipping it. |
Running setup.py install for pesq did not run successfully. note: This error originates from a subprocess, and is likely not a problem with pip. × Encountered error while trying to install package. note: This is an issue with the package mentioned above, not pip. [notice] A new release of pip available: 22.3.1 -> 23.0.1 |
c /nologo /O2 /W3 /GL /DNDEBUG /MD -Ipesq -IC:\Users\chlyw\anaconda3\envs\ttts-fast\lib\site-packages\numpy\core\include -IC:\Users\chlyw\anaconda3\envs\ttts-fast\lib\site-packages\numpy\core\include -IC:\Users\chlyw\anaconda3\envs\ttts-fast\include -IC:\Users\chlyw\anaconda3\envs\ttts-fast\Include "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.35.32215\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.35.32215\ATLMFC\include" "-IC:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Auxiliary\VS\include" /Tcpesq/dsp.c /Fobuild\temp.win-amd64-cpython-38\Release\pesq/dsp.obj note: This error originates from a subprocess, and is likely not a problem with pip. × Encountered error while trying to install package. note: This is an issue with the package mentioned above, not pip. |
I give up. I re-tried one last 15'th time through a Python Venv, after installing buildtools for C++ v14+ + windows 10 sdk. I didn't get any critical errors this time, HOWEVER. Streamlit still uses only CPU! My GPU is on 1% utilization. -- (myenv) C:\Users\chlyw\Desktop\tortoise-tts-fast>pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu117 [notice] A new release of pip available: 22.3.1 -> 23.0.1 (myenv) C:\Users\chlyw\Desktop\tortoise-tts-fast>python3 -m pip install -e . [notice] A new release of pip is available: 23.0 -> 23.0.1 (myenv) C:\Users\chlyw\Desktop\tortoise-tts-fast>pip3 install git+https://github.com/152334H/BigVGAN.git [notice] A new release of pip available: 22.3.1 -> 23.0.1 (myenv) C:\Users\chlyw\Desktop\tortoise-tts-fast>streamlit run scripts/app.py You can now view your Streamlit app in your browser. Local URL: http://localhost:8501 C:\Users\chlyw\Desktop\tortoise-tts-fast\tortoise\utils\audio.py:19: WavFileWarning: Chunk (non-data) not understood, skipping it. |
Have you tried not running it in a virtual environment? What GPU do you use? |
I've tried about 15x no joke, tried conda env's, tried python venv's, tried python without venv. Always issues. |
What GPU are you using? |
4090 |
you could always try using Windows Subsystem for Linux on your windows machine and run the program from there. I've noticed Windows is prone to a lot of errors with these types of programs |
Update no luck. Tried Conda/Python through WSL. No install errors, but inference wouldnt run. Massive error count with RuntimeError: cuFFT error: CUFFT_INTERNAL_ERROR |
ERROR: Cannot install -r requirements.txt (line 1064), -r requirements.txt (line 191), -r requirements.txt (line 323), -r requirements.txt (line 401) and scipy>=1.10.1 because these package versions have conflicting dependencies. The conflict is caused by: To fix this you could try to:
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts |
Another thing I thought you could check is seeing if you have the cuda 11.7 toolkit installed on windows , idk if it needs to be in the environment or not I don't know why you're getting that error in Linux and I'm sorry I haven't been more helpful |
Yep I have a bunch of them 11.7 and 11.8 |
FIXED!!!!!!!!!!!!!!!! Use this line instead of the line in the instructions to install an older version of Pytorch> conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c nvidia Make sure Visual Studio 14+ C++ build tools are installed along with Windows 10/11 SDK. |
Could you be a little more specific for us noobs on the set up, you've got it in a conda too right? |
Steps for installing using Conda (in Windows10), WSL, Linux may be different. *Make sure you have Cuda toolkit 11.7 installed on your machine. Open Conda environment (Anaconda/MiniConda)
|
You might good sir, are a boss. |
Do you know why I get this error when starting streamlit run scripts/app.py? |
From ChatGPT The error message indicates that the path to the torchvision package is too long, causing issues when loading the image Python extension. This is a known issue on Windows systems, where the maximum path length is limited to 260 characters by default. There are a few possible solutions to this issue: Enable long file paths on Windows: Starting from Windows 10 version 1607, you can enable long file paths. To do this, you need to update the group policy or the registry. To update the group policy: Press Win + R, type gpedit.msc, and press Enter to open the Local Group Policy Editor. To update the registry: Press Win + R, type regedit, and press Enter to open the Registry Editor. After enabling long file paths, restart your computer for the changes to take effect. Shorten the environment name or the path: You can try shortening the name of your Conda environment or installing Miniconda/Anaconda in a directory with a shorter path. This will help you keep the path length below the Windows limit of 260 characters. Use a different environment manager: As an alternative, you can use a virtual environment manager like venv or virtualenv instead of Conda, which may help you avoid this issue. Remember to back up your data and create a system restore point before making changes to the registry or group policy. |
Ty, tried that already, as I also used chatgpt, Doesn't work and the path length is not more then the 260 chars, so must be a different problem. The weird thing is the interface still starts and runs, but to me there is no difference in speed. |
You knocked it out of the park. It's working beautifully. |
Glad I could help. |
What kind of settings are you running/find the best quality/speed in? |
Christ dude, thank you |
I have made a simple docker project, want to try it ? |
I followed your comment to the letter but trying to use the ui gets me this error. 2024-02-04 14:46:16.774 Uncaught app exception |
Finally fixed my issue by running the commands |
instructions are unclear.
dependencies are wrong/out-of-date and unclear
hash's are off.
nothing works.
The text was updated successfully, but these errors were encountered: