Fine-tuning becomes available for GPT-3.5 Turbo

OpenAI opened its fine-tuning API for GPT-3.5 Turbo, letting developers customize the model with their own data. On certain narrow tasks a fine-tuned GPT-3.5 Turbo could match or exceed base GPT-4, the company said. Typical uses were steerability and reliable output formatting; some early testers cut prompt length by up to 90% by baking instructions into the model. OpenAI said fine-tuning data belonged to the customer and would not be used to train other models, and that fine-tuning for GPT-4 would follow in the autumn.