I have a new Dell XPS 9640 with a NVIDIA RTX 4070. I have installed torch v 2.2.2 and Python 3.12.2, and am using the latest version of VSCode. a simple Python script indicates GPU is not available. What am I missing here?
Assuming you have installed the correct PyTorch binary with CUDA support, you might want to check your NVIDIA drivers and reinstall them if needed.
when I show torch in terminal it reports:
Name: torch
Version: 2.2.2
Summary: Tensors and Dynamic neural networks in Python with strong GPU acceleration
Home-page: https://pytorch.org/
Author: PyTorch Team
Author-email: packages@pytorch.org
License: BSD-3
Location: c:\Users\rdcap\pythonScripts.venv\Lib\site-packages
Requires: filelock, fsspec, jinja2, networkx, sympy, typing-extensions
Required-by: torchaudio, torchvision
nvidia reports:
What does torch.__version__ and torch.version.cuda return?
torch.version says 2.2.2+cpu (oops! shouldn’t it say +GPU?) and if so, how do I get the right version
Yes, it should show a valid CUDA runtime version. Refer to the install instructions. The easiest way would be to simply run pip install torch, which will install the latest stable PyTorch release with CUDA 12.1.
i installed torch with pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu124/torch_stable.html
This install command is invalid and not listed in the support matrix.
The CUDA 12.4 builds are still being worked on, so stick to the provided install commands.
Fixed! I backed up to cuda 12.1 and the gpu is available. thanks for your help
Hey guys,
Just a note of thanks. This info was very helpful.