Are there any long-context LLaMA 3 variants that the community has trained yet?

→ View original post on X — @mattshumer_