New Delhi, August 22: OpenAI has announced an update with the introduction of fine-tuning capabilities for GPT-4o to enhance the performance and accuracy of applications using the model. The new feature might allow developers to fine-tune custom versions of GPT-4o to better align with specific needs and tasks. The ability to fine-tune GPT-4o is expected to bring benefits to various industries to offer enhanced AI experience.

As per the information available on the official website, Fine-tuning is now available for GPT-4o to allow users to create custom versions of the model that can increase its performance and accuracy for specific applications. By fine-tuning, users can adjust the model to better fit their unique needs and might improve its effectiveness in tasks with different applications. OpenAI Releases GPT-4o System Card, Unveiling Key Insights and Capabilities; Check Details.

OpenAI is offering 1 million training tokens per day for free until September 23. It means that developers can now fine-tune GPT-4o with their own custom datasets. By doing this, they can achieve better performance at a lower cost for their specific needs. Even with just a few dozen examples in their training dataset, developers might achieve results for their applications. Fine-tuning will likely allow the model to adapt its responses to match the desired style and tone or to follow detailed instructions related to a specific field.

Developers using any paid plan can now access GPT-4o fine-tuning. It costs USD 25 per million tokens for training, USD 3.75 per million input tokens, and USD 15 per million output tokens for inference. To start, go to the fine-tuning dashboard, click “create” and choose “gpt-4o-2024-08-06” from the base model menu. OpenAI Rolls Out DALL·E 3 Image Creation With Limitations for ChatGPT Free Users; Check Details and Know How To Use It.

When developers fine-tune models, they have complete control and ownership over their business data. It means that all the inputs and outputs of the fine-tuned models are fully owned by the developers. OpenAI does not share or use the data to train other models to enhance the privacy and security of the developer's data. To further ensure the responsible use of fine-tuned models, OpenAI has implemented safety measures to align with OpenAI's policies.

(The above story first appeared on LatestLY on Aug 22, 2024 07:35 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).