Hosting a Machine Learning model over a FASTAPI server.

Purpose: To predict sale price of a used toyota car and to send details of the car through a JSON request.

Step 1: Load the files in the below format

ML_Price_Prediction | |--models | |-- picklefile | |-- ML_full_trained_model.ipynb | |-- price_predictor.py |--v1 |main.py

Step 2: Start uvicorn instance on cmd/powershell

Step 3: From Postman use a request format JSON as below to include car details and POST:

{ "predict": "True", "classify": "False", "predict_input_set": { "model_number": 10, "year": 2012, "mileage": 15106, "tax": 257, "mpg": 22, "enginesize": 1.0, "automatic": 1, "manual": 1, "semiauto": 0, "diesel": 0, "hybrid": 0, "other": 0, "petrol": 1}

}

Sample output(JSON) :

"{"message": " Below is the approx current value for your toyota", "Sale_Price": "[8129.29]"}"