
Meta's Llama 2 foundation models are now available in Amazon SageMaker JumpStart, enabling developers to deploy and use these language models with a few clicks while leveraging SageMaker's MLOps controls and security features.


<div><p>This article announces the availability of Meta's Llama 2 foundation models in Amazon SageMaker JumpStart, enabling easy deployment and use of these language models.</p><ul><li>Llama 2 models available in 7B, 13B, and 70B parameter sizes with pretrained and fine-tuned variations</li><li>Deploy models with few clicks in SageMaker Studio or programmatically via Python SDK</li><li>Fine-tuned versions use supervised fine-tuning and reinforcement learning with human feedback</li><li>Models deployed in secure AWS environment under customer VPC controls</li><li>Access to SageMaker MLOps features including Pipelines, Debugger, and container logs</li><li>Initially available in us-east-1 and us-west-2 regions</li><li>Tuned models for chat assistance; pretrained models adaptable for various NLP tasks</li></ul><p>Llama 2 foundation models are now accessible through SageMaker JumpStart, combining Meta's language models with AWS's MLOps capabilities for streamlined deployment and management.</p></div>


Related articles