Meta's Llama 2 overtakes open LLM leaderboards, enables commercial use

Meta has released Llama 2, an open-source large language model that has quickly become the state-of-the-art in its weight class, outperforming other open models. The model was pre-trained on 2 trillion tokens with an expanded context length and significant investment in reinforcement learning from human feedback. Llama 2 is available for commercial use, addressing a critical need for organizations that cannot send sensitive data to external API providers and offering an alternative to proprietary models like GPT-4. AI

RANK_REASON Release of a new open-source LLM from a major tech company that claims state-of-the-art performance.

Read on Latent Space Podcast →

• a16z
• Baseten
• Claude 2
• Falcon-40B
• GPT-4
• LLaMA
• Llama 2
• llama.cpp
• Meta
• ollama
• OpenAI
• Replicate
• Scale AI
• Chroma

• model release
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →