Search code examples
pytorch

torch.cuda.is_available() returns False even after installing PyTorch with CUDA

I have recently installed PyTorch with CUDA support on my machine, but when I run torch.cuda.is_available(), it returns False. I verified my GPU setup using nvidia-smi, and it seems that my system recognizes the GPU correctly.

Here are the steps I've followed so far:

  1. Installed PyTorch with CUDA support using the command pip install torch torchvision torchaudio.

  2. Confirmed that my GPU is recognized by my system using nvidia-smi.

  3. Verified that the correct CUDA version is installed.

Despite this, running torch.cuda.is_available() still returns False. What could be the reason for this issue, and how can I resolve it?

Additional Information:

  • Operating System: Windows 11

  • Python Version: 3.10

  • PyTorch Version: 2.3.1

  • CUDA Version: 12.5.82

  • GPU Model: NVIDIA GeForce RTX 4060

What I've Tried:

  • Reinstalling PyTorch with CUDA support.

  • Updating GPU drivers.

  • Rebooting the system.

import torch

print(torch.cuda.is_available())  # Returns False
print(torch.cuda.current_device())  # Raises RuntimeError: No CUDA GPUs are available

I have recently installed PyTorch with CUDA support on my machine, but when I run torch.cuda.is_available(), it returns False.

Solution

  • I solved this problem via DELETING ALL TORCH AND CUDA modules on my PC. Then I used Pytorch with Cuda Website and downloaded 11.8 version.