configuration of a service

[linhsolar]

@vtn13042000
consider the inference service:
https://github.com/rdsea/ROHE/blob/main/examples/applications/object_detection/kube_deployment/edge-inference-server/server.py

parser.add_argument('--conf', help='configuration file', default="./conf.json")
parser.add_argument('--port', help='default port', default=5002)
args = parser.parse_args()

# Serving port
port = int(args.port)

# Load configuration file
config_file = args.conf
configuration = qoa_utils.load_config(config_file)


# Load Rest service
inferenceService = YoloRestService(configuration)

it receives a config like this for models. Then where are the model storage? A service may receives a configuration about the system (e.g., storage, number of cpus, etc.) and about the application logic (e.g., composition, parameters). But we must have enough info

{
"composition":[
{
"model": "Yolov5",
"parameter": "yolov5l"
},
{
"model": "Yolov8",
"parameter": "yolov8n"
}
]
}