As businesses face growing pressures from competitors to incorporate new technologies into their practices, OpenAI is releasing a new fine-tuning feature aimed at companies who want to use their own data.
The AI company behind ChatGPT made the announcement on Tuesday (August 20) as they shared the details of the new ‘Fine-tuning’ feature which is now available for GPT-4o.
With the tool, “developers can now fine-tune GPT-4o with custom datasets to get higher performance at a lower cost for their specific use cases.
“Fine-tuning enables the model to customize the structure and tone of responses or to follow complex domain-specific instructions. Developers can already produce strong results for their applications with as little as a few dozen examples in their training data set.”
This capability should make it a whole lot easier for brands and businesses to tailor the AI to their own uses.
In a conversation with Bloomberg, the head of product for OpenAI’s API Olivier Godement said: “We’ve been extremely focused on lowering the bar, the friction, the amount of work it takes to get started.”
Companies to keep ‘full ownership of your business data,’ with fine-tuning feature
With fine-tuning, customers have to upload their data to OpenAI’s servers. The training is then expected to take an hour or two, with it believed to be just text-based data to begin with.
While that may make some companies nervous, the San Francisco-based company has said the “fine-tuned models remain entirely under your control, with full ownership of your business data, including all inputs and outputs.
“This ensures your data is never shared or used to train other models.
“We’ve also implemented layered safety mitigations for fine-tuned models to ensure they aren’t being misused. For example, we continuously run automated safety evals on fine-tuned models and monitor usage to ensure applications adhere to our usage policies.”
The team has said they’re “excited” to see what people choose to build by fine-tuning GPT-4o as customization has now opened up for the first time ever.
Featured Image: Photo by Jernej Furman on Wikimedia Commons