Meta releases Llama 2 on Hugging Face

AI Impact Summary

Meta has released Llama 2, a family of open-access large language models ranging from 7B to 70B parameters, available on Hugging Face. This release includes both base models and fine-tuned chat models optimized for dialogue applications using RLHF, offering a commercially usable alternative to closed-source chatbots. The release provides a range of model sizes and configurations, along with tools for inference and fine-tuning, enabling developers to experiment with and deploy Llama 2 models within the Hugging Face ecosystem.