Quora Question Pair Challenge Solution

This repository contains the solution to the Quora Question Pair Challenge, a natural language processing task that consists of predicting whether a pair of questions asked on Quora are duplicates or not.

Collaborators

The solution was developed by the following collaborators:

• @sergibech
• @socalesc
• @davidrosado4
• @Goodjorx
• @sarabase

Repository Structure

The repository is organized as follows:

• train_models.ipynb: a Jupyter notebook that contains the code to preprocess the data and train the models. When executed, it creates a folder called model_artifacts that contains all the necessary information to reproduce the results.
• reproduce_results.ipynb: a Jupyter notebook that contains the code to reproduce the results obtained by our models. It reads from the model_artifacts folder.
• utils.py: a Python module that contains all the helper functions for the two previous notebooks.

Reproducing the Results

To reproduce the results obtained by our models, follow these steps:

1. Clone this repository:

git clone https://github.com/sarabase/quora-question-pairs.git

2. Create a conda environment and install the necessary requirements. Activate the environment:

conda create --name quora_test_env --file requirements.txt
conda activate quora_test_env

3. Run the train_models.ipynb notebook.
4. Open the reproduce_results.ipynb notebook in Jupyter and execute the cells.