Iterative Length-Regularized Direct Preference Optimization: A Case Study on Improving 7B Language Models to GPT-4 Level Paper • 2406.11817 • Published Jun 17 • 13
