0

I recently switched from windows to ubuntu. I installed miniconda here, created a virtual environment with python 3.8.5, activated that environment and ran

conda install pytorch torchvision torchaudio cudatoolkit=11.0 -c pytorch

I ran torch.cuda.is_available() and it says True, but when I run which CUDA/Cuda/cuda or whereis CUDA/Cuda/cuda, I get nothing. nvcc is also not working. I think I might have to update bash file. Would someone please be able to guide me.

Also, Do I need to install cuDNN to work on my ML models using CUDA or is that not needed.

talonmies
  • 70,661
  • 34
  • 192
  • 269
Manish Goyal
  • 1
  • conda will only install the necessary components required to run the ML frameworks they distribute. The "cudatoolkit" package isn't a complete development toolkit and if you are not going to write CUDA code (and it sounds like you are not), then you don't need anything more than what conda installs, **except for a suitable GPU driver**. See the linked duplicate for more information – talonmies Feb 24 '21 at 02:31
  • Thank You for your response. Can you explain why do I see my CPU utilization 100% and GPU utilization almost 0% when I run my deep learning models on this machine. – Manish Goyal Feb 24 '21 at 04:51
  • Sorry, no. That will be a property of PyTorch and your code – talonmies Feb 24 '21 at 07:29

0 Answers0