llama fails running on the GPU.
Traced it to torch! Torch is using CUDA 12.1.Tried multiple different approaches where I removed 12.1 to make it use 12.3 downgraded the Nvidia driver. No joy! All help is appreciated. Running on a openSUSE tumbleweed.
PyTorch Version: 2.1.1
CUDA Version: 12.1
CUDA Available: False
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
Defaulting to user installation because normal site-packages is not writeable
Looking in indexes: https://download.pytorch.org/whl/cu121
ERROR: Could not find a version that satisfies the requirement torch (from versions: none)
ERROR: No matching distribution found for torch
You asked your initial question in this topic which is specific to CUDA 12.3. If you really need this version for some reason you would need to build PyTorch from source as indicated in my previous answer.
Now you post a pip install command instead, which would install PyTorch binaries with CUDA dependencies.
Could you explain what you actually want?
How are you importing torch if nothing was installed?
In the end you should either uninstall all PyTorch binaries and install the current stable or nightly release (a simple pip install torch would be enough to install torch==2.1.2+cu121) or create a new and empty virtual environment and install PyTorch there.
brother, have you got the solution ? my cuda version is 12.3 . I’ve installed cuda , cudnn but cant install pytoch. "import torch
print (torch.version)
print (torch.cuda.is_available())
2.3.0+cpu
False " please let us know
my cuda version is 12.3 . I’ve installed cuda , cudnn but cant install pytoch. "import torch
print (torch.version )
print (torch.cuda.is_available())
2.3.0+cpu
False " please let us know
You’ve installed the CPU-only binary and would need to install the PyTorch binary with CUDA runtime dependencies by selecting it in the install matrix on our webite.
thank you for reply. I’ve tried several times to install " stable = windows= conda= python= cuda (12.1) " but the installation was never completed. it says "(‘Connection broken: IncompleteRead(23132496 bytes read, 1201416417 more expected)’, IncompleteRead(23132496 bytes read) & Downloading and Extracting Packages:
An HTTP error occurred when trying to retrieve this URL.
HTTP errors are often intermittent, and a simple retry will get you on your way.
(base) PS C:\Users\hp>" and one more thing – when i checked "conda list " it has the “torch 2.3.0+cu121 pypi_0 pypi & torchaudio 2.3.0+cu121 pypi_0 pypi” but can’t find any pytorch version neither cpu nor cuda. tried both ways “conda install” and “pip3 install”
The error points to a connection issue when trying to download packages from conda, so you might want to check your connection or try to download it later as suggested in the error message.