Llama 2-13B now available on Amazon Bedrock

Llama 2-13B now available on Amazon Bedrock

Meta’s Llama 2 Chat Model (13B) is now available on Amazon Bedrock!

  • First public cloud service with fully managed Llama 2 API,
  • Optimized for dialog use cases,
  • Easy integration via Amazon Bedrock API, AWS SDKs, or AWS CLI Experience the future of AI-driven conversations today.

Amazon Bedrock has introduced a groundbreaking service as the first public cloud platform to offer a fully managed API for Llama 2, Meta’s advanced LLM. This development allows entities of various sizes to utilize the Llama 2 Chat models through Amazon Bedrock, eliminating the need for them to handle the complex infrastructure. This represents a significant advancement in making such technology more accessible.

About LLama 2

Llama 2, developed by Meta, is a series of LLMs available to the public. Its base model underwent pre-training on a massive scale, using 2 trillion tokens from publicly accessible online data. Meta reports that training the Llama 2 13B model required a staggering 184,320 GPU hours, equating to over 21 years of continuous single GPU use, not including leap years.

The Llama 2 Chat model, an extension of the base model, is specifically tailored for conversational applications. It has been refined with more than a million human annotations through a process known as reinforcement learning from human feedback (RLHF). Meta has rigorously tested this model to identify and address any performance issues, particularly aiming to prevent problematic responses in chat scenarios, such as offensive or inappropriate content.

Moreover, Meta has committed to fostering a responsible and collaborative AI innovation environment. To support this, they offer a variety of resources for users of Llama 2, including individuals, creators, developers, researchers, academics, and businesses of any scale. Of particular note is the Meta Responsible Use Guide. This guide, aimed at developers, outlines best practices and key considerations for responsibly creating LLM-powered products, spanning from the initial concept to the final deployment. This guide complements AWS’s suite of tools and resources designed for ethical AI development.

Integrating the Llama 2 Chat model into your applications, regardless of the programming language, is now seamless. This can be achieved by utilizing the Amazon Bedrock API, AWS SDKs, or the AWS Command Line Interface (AWS CLI).

Read related articles: