A fine-tuning library leveraging distilled GPT-4 outputs to fine-tune to GPT-3.5 for better accuracy. - View it on GitHub