run_docker_linux_gpu.sh

echo -n "Enter your dataset's absolute path (folder containing images', labels' folders and configuration file): " 
read folder_path
configfile=$folder_path/train_config.json
echo -n "Choose a name for the docker container: " 
read container_name
ports=''
if [ -f $configfile ]; then 
	custom_api=`jq .training.custom_api.enable $configfile`
	if [ "$custom_api" = "true" ]; then
		custom_api_port=`jq .training.custom_api.port $configfile`
		ports="$ports -p $custom_api_port:$custom_api_port"
	fi
	tensorboard=`jq .training.tensorboard.enable $configfile`
	if [ "$tensorboard" = "true" ]; then
		tensorboard_port=`jq .training.tensorboard.port $configfile`
		ports="$ports -p $tensorboard_port:6006"
	fi
	web_ui=`jq .training.web_ui.enable $configfile`
	if [ "$web_ui" = "true" ]; then
		web_ui_port=`jq .training.web_ui.port $configfile`
		ports="$ports -p $web_ui_port:8090"
	fi
	sudo docker run --rm --runtime=nvidia -it -e TRAIN_NAME=$container_name -e TRAIN_START_TIME="$(date '+%Y%m%d_%H:%M:%S')" $ports -v $folder_path:/training/assets -v $(pwd)/trainings:/training/custom_training -v /etc/timezone:/etc/timezone:ro -v /etc/localtime:/etc/localtime:ro --name $container_name darknet_yolo_gpu:1 ; 
else	
	echo "Error: Configuration file not found in the provided path"
fi