OpenAI customers are now able to add their own customization data in the GPT-3.5 Turbo version. This will allow for specific behavioral models, reports TechCrunch.
According to OpenAI, additionally customized GPT-3.5 will be able to meet and even exceed the basic capabilities of GPT-4 in certain tasks.
“Since the release of GPT-3.5 Turbo, developers and businesses have asked for the ability to customize the model to create unique and differentiated experiences for their users,” the company wrote in a blog post published this afternoon. “This update gives developers the ability to customize models that perform better for their use cases and run these custom models at scale.”
Customization will allow the model to better follow certain instructions, such as always responding in a specific language or maintaining a distinctive style. In addition, such customization allows OpenAI customers to shorten text prompts to speed up API calls and reduce costs.
“Early testers have reduced prompt size by up to 90% by fine-tuning instructions into the model itself,” OpenAI claims.
By the way, on July 20, OpenAI stopped using a tool that was supposed to distinguish between human and artificial intelligence. This decision was made due to the low level of accuracy of the classifier.