Contradictory, My Dear Watson

Overview

This competition challenges you to build a Natural Language Processing (NLP) model to detect relationships between sentences, specifically identifying entailment, contradiction, or neutral connections. You'll work with a multilingual dataset, making it a great opportunity to practice advanced NLP techniques like transformers and TPUs. This is a 'getting started' competition, perfect for learning.

Requirements

This is a 'Getting Started' code competition with an ongoing, rolling timeline and no cash prizes, making it ideal for beginners. Submissions must be made through Kaggle Notebooks, which have a 2-hour run-time limit per session. You can use CPU, GPU, or TPU. Your submission file should be a CSV named 'submission.csv' with 'id' and a 'prediction' column (0 for entailment, 1 for neutral, 2 for contradiction).

Prizes

This is a 'Getting Started' competition and does not offer cash prizes. The goal is to learn and practice NLP skills on Kaggle.