OpenAI and Scale are joining forces to help more companies benefit from fine-tuning our most advanced models.
Enterprises expect high performance, manageability, and customization when it comes to deploying ai in production. We recently released tweaks for GPT-3.5 Turbo and will be implementing them in GPT-4 this fall. With tweaks, businesses can now safely customize our most advanced models with proprietary data, making our most powerful models even more useful. As always, data sent in and out of the Tuning API is the property of the customer and is not used by OpenAI or any other organization to train other models.
We are working with Scale as a preferred partner to expand the benefits of our adjustability given their experience helping companies leverage data for ai safely and effectively. Developing robust enterprise-grade functionality requires rigorous data enrichment and model evaluation. Scale customers can now tune OpenAI models just as they would through OpenAI, while benefiting from Scale’s enterprise ai and Data Engine expertise.
Scale has already demonstrated value to customers by tuning GPT-3.5 for Brex. See more details ai-scale-partnership-gpt-3-5-fine-tuning” rel=”noopener noreferrer” target=”_blank”>here.
