Posted On: Apr 23, 2024
We are excited to announce the preview of Custom Model Import for Amazon Bedrock. Now you can import customized models into Amazon Bedrock to accelerate your generative AI application development. This new feature allows you to leverage your prior model customization investments within Amazon Bedrock and consume them in the same fully-managed manner as Bedrock’s existing models. For supported architectures such as Llama, Mistral, or Flan T5, you can now import models customized anywhere and access them on-demand.
As the generative AI space rapidly evolves, new models with open architectures such as Mistral and Llama continue to emerge that offer better price-performance than existing models. Customers often adapt these models to specific use cases to solve specific business needs. These include techniques such as fine-tuning or domain adaptation using proprietary data. Previously, customers had to deploy on self-managed infrastructure for models customized outside Bedrock, creating a disjointed experience for application developers as they switch between different model sources. Now, with this launch, customers can import these models customized outside Bedrock (for supported model architectures) and access them on-demand through Bedrock’s invoke model API, creating a unified experience across base, custom, and imported models on Amazon Bedrock.
This feature is available in preview in the US-East (N.Virginia) region. You can get started by initiating the model import workflow in the custom models page of the Amazon Bedrock console. To learn more, please visit the documentation page on custom model import.