Meta launches Llama 4, advanced multi-modal LLMs achieving SOTA results
Tweet Meta has launched its new Llama 4 series, featuring three models: Scout, Maverick, and Behemoth. Scout and Maverick are …
Llama
Foxconn unveils its own large language model distilled from Meta’s Llama 3.1
Tweet Foxconn has announced its own large language model, FoxBrain, that it would use to streamline its manufacturing processes within …
Filing says Zuckerberg approved Meta’s use of copyrighted material in Llama training
Tweet One of the world's largest "shadow library" was allegedly used to train Llama models, and a new filing now …
Meet Llama 3.1, the world’s largest and most capable open foundation model
Tweet Meta has announced the release of its Llama 3.1 family of AI models, which includes three models: 8B, 70B, …
Meta releases much more powerful Code Llama 70B to rival GitHub Copilot
Tweet Meta releases its new AI-assisted code-writing tool, Code Llama 70B. It is trained on 500B code tokens to generate …