OpenAI has incorporated Reinforcement Fine-Tuning (RFT) technology into the o4-mini inference model, enhancing model outputs through the design of scoring functions and minimizing the dependency on labeled data. This groundbreaking technology is poised to make a significant impact in intricate domains such as healthcare, taxation, and law, and has already assisted numerous enterprises in boosting their model performance. Certified organizations now have access to this innovative RFT technology, with a training fee set at $100 per hour, and additional discounts available for data-sharing. This advancement is anticipated to propel the application and advancement of AI across a broader spectrum of fields.
