Fine-Tuning Llama 3-8B for Structured Math Reasoning: Fine-tuning Llama3 8b to generate JSON formats for arithmetic questions and process the output to perform calculations.

Supervised-Fine Tuning lab of Llama3-8B on prediction of correctness of answers to math questions

In this competition, participants are tasked with the Supervised-Fine Tuning(SFT) of Llama3-8B model to predict the correctness of answers to math questions. The goal is to assess whether the provided answer to each question is correct or not.

In this competition, participants are tasked with the Supervised-Fine Tuning(SFT) of Llama3-8B model to predict the correctness of answers to math questions. The goal is to assess whether the provided answer to each question is correct or not.

Team of NYU-DL-Fall-24-Kaggle-Competition, named Binary Beats. Our focus is on the Supervised-Fine Tuning (SFT) of the Llama3-8B model to evaluate the correctness of answers to mathematics questions. Our aim is to create a reliable model that can determine if a provided answer to a math problem is correct or incorrect.