Amazon SageMaker Ground Truth Plus now supports human feedback and fine-tuning data for Generative AI

Amazon SageMaker Ground Truth Plus now enables customers to prepare high-quality, large-scale training datasets to fine-tune foundation models to perform human-like generative AI tasks. It also allows skilled humans to review model outputs to align them with human preferences. Additionally, it enables application builders to customize models using their industry or company data while preparing training datasets. SageMaker Ground Truth Plus provides workflows, labeling interfaces, and a skilled workforce to generate high quality training data to accomplish these tasks in a turnkey fully managed service on behalf of the customer.

Fine-tuning foundation models on generative tasks enables models to interact with humans in a natural way - such as answering questions that a human asks, summarizing text that a human provides, or generating images or videos for which a human provides a description. Aligning models with human preferences ensures that a generative model’s output content is aligned with human preferences such as helpfulness, accuracy, and harmlessness. And, customizing the model for industry or company specific use cases ensures that application builders who leverage pre-existing generative foundation models are able to build their own applications in a way that successfully generates content in their preferred style or voice.

To get started, connect with the SageMaker Ground Truth Plus team here.

SageMaker Ground Truth Plus is available in US East (N. Virginia) region in the AWS console.