/ml_deployment

Primary LanguageJupyter Notebook

Calories Prediction Interface

This repository contains a simple web interface for predicting calories burned based on total distance and total active minutes using a machine learning model.

Features

  • Allows users to input Total Distance (in miles) and Total Active Minutes.
  • Sends a POST request to a FastAPI endpoint to get predictions.
  • Displays the predicted calories burned to the user.

Requirements

  • Python 3.x
  • FastAPI
  • Pydantic
  • Pandas

Installation

  1. Dependencies:

    pip install -r requirements.txt

Usage

  1. Start the FastAPI server:

    uvicorn ml_api:app --reload

  2. Open your web browser and navigate to http://localhost:8000 to access the interface.

  3. Input the required information and click the "Predict Calories" button to get predictions.

Structure

  • main.py: Contains the FastAPI code for serving the predictions endpoint and the HTML file.
  • model/model.pkl: Pre-trained machine learning model for predicting calories burned.
  • index.html: HTML file containing the user interface elements.
  • requirements.txt: List of Python dependencies.
  • notebooks/: Contain all the required notebooks for train a simple ml models.
  • data_src: All of data used for this project.