daiedge-vlab 1.2.0

Python client for dAIEdge VLab: simplifies interaction with the API and provides utilities for model benchmarking and on-device training.

dAIEdge VLab

The dAIEdge VLab is a benchmarking platform for edge AI devices. It allows users to run benchmarks on various edge devices and retrieve the results.

Prerequisites

• An account on the dAIEdge VLab platform. You can create an account here.
• Python 3.7 or higher

Setup

Create a file .yaml in the root directory of the project and add the following content to it :

api:
    url: "vlab.daiedge.eu"
    port: "443"
    base_path: ""  # Optional, can be left empty
    ssl: true  # Optional, true by default 
user : 
    email: "your-email"
    password: "your-password"
    

Use the package

Give the .yaml file path to the dAIEdgeVLabAPI constructor. The dAIEdgeVLabAPI object will try to log in to the API using the credentials provided in the .yaml file immediately after the object is created.

from daiedge_vlab import dAIEdgeVLabAPI

SETUP_FILE = "setup.yaml"

TARGET = 'rpi5'
RUNTIME = 'tflite'
MODEL = 'models/small_model.tflite'

if __name__ == '__main__':

    api = dAIEdgeVLabAPI(SETUP_FILE)
    id = api.startBenchmark(TARGET, RUNTIME, MODEL)

    result = api.waitBenchmarkResult(id)
    
    print(result)