ml-infra-tools

Tools for building ML environments & infra

prerequisites

• docker
• docker-composer
• Nvidia Driver
• nvidia-docker

How to know docker, Nvidia Driver, Nvidia Docker have been installed correctly

docker run --rm --gpus all nvidia/cuda:11.0-base nvidia-smi

The output will be somethings like

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 450.80.02    Driver Version: 450.80.02    CUDA Version: 11.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce GTX 108...  Off  | 00000000:01:00.0  On |                  N/A |
| 43%   57C    P0    71W / 250W |   3291MiB / 11175MiB |      5%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
+-----------------------------------------------------------------------------+

Configure docker-composer runtime

Update runtime for nvidia-docker

$ cat /etc/docker/daemon.json
{
    "runtimes": {
        "nvidia": {
            "path": "nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}

How to run

The docker images has Jupyter Lab inside, when you start docker image the Jupyter will automatically launch, and bind to 8889 port on your computer.

Happy Coding!!!

$ cd dockerfiles
$ docker-compose -f docker-compose.yml -f dev-gpu.yml up --build

Test GPU inside

• Open localhost:8889
• Open Test.ipynb

• Run the first code block

  import torch
  torch.cuda.current_device()