LLaMA v2 was just announced. And it's live on Replicate. LLaMA v2 is the latest open source model from Meta. v2 was trained on more data than v1 (~2 trillion tokens) and supports context windows up to 4k tokens. https://
replicate.com/a16z-infra/lla
ma13b-v2-chat
… We think it's on par with GPT-3.5

→ View original post on X — @replicate