Accompanying code for the Deploying a ML Model on Google Compute Engine tutorial on Medium.
.
├── yolo-coco # Pre-trained model files
└── coco.names # Object labels (person, car, etc.)
└── yolov3.cfg # Model configuration
└── yolov3.weights # Model weights
├── app.py # Flask app serving predictions
├── yolo.py # Functions to generate predictions
├── requirements.txt # Dependencies
└── README.md
conda create -n [name of enviroment] python=3.7
pip install -r requirements.txt
cd google-cloud-flask-tutorial
conda activate [name of enviroment]
python app.py
Edit and use request.py
to send a POST request to the app.