Meta launches Llama 3 model with 8 billion and 70 billion parameters

Meta has officially released two models of Llama 3, including 8 billion and 70 billion parameters. The model implements new features, such as improved inference capabilities, and sets new state-of-the-art levels for specific models. In the coming months, we expect to introduce new features, longer contextual windows, additional model sizes, and enhanced performance + Llama 3 research papers for the community to learn from our work.