anibali/docker-pytorch

Cuda not found after pulling docker image at docker pull anibali/pytorch:1.4.0-cuda10.1

likwifi opened this issue · 2 comments

likwifi commented

I've run command docker pull anibali/pytorch:1.4.0-cuda10.1
after connected to container with command docker run -it anibali/pytorch:1.4.0-cuda10.1 sh
both nvcc and nvidia-smi not found.

Output from my vm's nvidia-smi.

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 435.21 Driver Version: 435.21 CUDA Version: 10.1 |
|-------------------------------+----------------------+----------------------+
| 0 Tesla V100-PCIE... Off | 00000000:37:00.0 Off | 0 |
| N/A 59C P0 46W / 250W | 0MiB / 32510MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
| 1 Tesla V100-PCIE... Off | 00000000:86:00.0 Off | 0 |
| N/A 49C P0 37W / 250W | 0MiB / 32510MiB | 0% Default |
+-------------------------------+----------------------+----------------------+

anibali commented

Have you set up nvidia-docker and tested that it works? If it's set up correctly, you should be able to run the following and get proper nvidia-smi output:

$ docker run -it --gpus=all nvidia/cuda:10.1-base-ubuntu16.04 nvidia-smi

Note the inclusion of the --gpus=all option in the docker run command---you may find that this is required or not depending on how you're set up.

If that doesn't work, then the problem is not specific to this project and you should make sure that you get nvidia-docker working properly first.

likwifi commented

--gpus=all fixed my problem thanks.

  • 🎉1