P2024 MMMDE4IN19-22-antoine-courbi

This README is meant to explain how to run the application and what features are implemented.

The application is a simple application using flask enabling a user to make a prediction on a trained ML model. The dataset is a stackoverflow topics posts dataset.

The requirements can be installed as follow using command line:

pip install -r requirements.txt

The two following commands can then be used in the predict\predict folder to run the application:

set FLASK_APP=app.py
flask run

After which the application can be accessed at http://127.0.0.1:5000/.

This project is a Flask-based web application designed to perform text predictions. It uses a trained machine learning model to predict the label of a given text input. The application provides a simple interface where users can enter their text and receive a prediction.

The application is structured as follows:

• app.py: The main application file. It initializes the Flask application and defines the routes.
• predict/predict.py: Contains the TextPredictionModel class used for making predictions.
• artefacts_path: A directory containing the trained model and other necessary artefacts.

The application provides a single route (/) that accepts both GET and POST requests. On a GET request, it simply renders the prediction form. On a POST request, it takes the user's text from the form, makes a prediction using the model, and then displays the prediction on the page.

This project includes a suite of unit tests to ensure the functionality of the TextPredictionModel class in the run.py file. These tests are implemented using the unittest framework and are located in the test_predict.py file.

The tests cover the following areas:

• Loading model, parameters, and labels from artefacts

The test_from_artefacts_loads_model_params_and_labels test verifies that the from_artefacts class method of TextPredictionModel correctly loads the model, parameters, and labels from the specified artefacts directory. The test uses mocking to avoid actual file reading and model loading.

• Calling embedding and model prediction

The test_predict_calls_embedding_and_model_predict test checks the predict method of TextPredictionModel. It verifies that the method correctly calls the embed function and the predict method of the model. The test uses a mock model and the embed function is also mocked.

• Format and length of predictions

The test_predict_with_mock_model test checks the predict method of TextPredictionModel and verifies the format and length of the predictions returned by the predict method. It uses a mock model and the embed function is also mocked.

The setUp method is run before each test method, and it sets up a temporary directory for testing. The tearDown method is run after each test method, and it removes the temporary directory.

To run these tests, navigate to the project directory and use the following command in your terminal:

python -m unittest test_predict.py

This will run all the tests and display the results in the terminal.

Thank you for reading this README. I hope you enjoyed it. If you have any remarks or questions, feel free to contact me on LinkedIn.

— ⭐️ —

This README was created during the Poc-To-Prod course

Created by Antoine Courbi