Fine-tuning llama-2-7b-chat on science questions

This repository contains the implementation and resources for fine-tuning the LLMa-7b-chat model on science-based multiple choice questions. Our goal is to enhance the model's capability to understand and accurately respond to science prompts, specifically in the context of selecting the best answer from multiple choice options.

model/: Contains the fine-tuned Llama-2-7b-chat model and relevant configurations.

data/: The custom dataset of science-based multiple choice questions.

notebooks/: Jupyter notebooks detailing the fine-tuning process, evaluations, and results visualization.