Neuroevolution of Efficient Models for Image Segmentation (NEMesIS)

Installation

Use your favorite python environment tool and install the packages in requirements.txt

pip3 example

python3 -m venv .venv

source .venv/bin/activate

python3 -m pip install -r requirements.txt

Usage

-c [mandatory]: sets the path to the config file to be used (in yaml or json format).

-g [mandatory]: path to the grammar file to be used.

-r [mandatory]: identifies the run id and seed to be used;

--gpu-enabled [optional]: if this flag is enabled, the run is performed in a gpu

Shell

python3 -m nemesis.main
    -c <config_path>
    -g <grammar_path>
    -r <run>
    --gpu-enabled

VSCode launch

{
    "name": "NEMesIS",
    "type": "debugpy",
    "request": "launch",
    "module": "nemesis.main",
    "console": "integratedTerminal",
    "cwd": "${workspaceFolder}",
    "args": [
        "-c",
        "${cwd}/settings/configs/enc_dec.json",
        "-g",
        "${cwd}/settings/grammars/enc_dec.grammar",
        "--gpu-enabled",
        "-r",
        "${command:pickArgs}"
    ]
}

Docker (Dockerfile for Nvidia, can be easily changed for AMD)

It is necessary to create a volume named datasets to save the used datasets, some may require manual download.

docker build -t nemesis .

docker run -v datasets:/usr/local/app/data --ipc=host --gpus all -it nemesis -c <config_path> -g <grammar_path> -r <run> --gpu-enabled