The full name of the LLaMA language model is "Large Language Model Meta AI". It is a new large language model (LLM) of Meta. It is a model series, which is divided according to the parameter scale (including 7 billion, 13 billion, 33 billion and 65 billion parameters).
It is worth noting that LaMA-13B (a model with 13 billion parameters), although its model parameters are more than ten times less than OpenAI's GPT-3 (175 billion parameters), can actually surpass GPT-3 in performance. The smaller model also means that developers can run AI assistants like ChatGPT locally on PCs, even smartphones and other devices, without relying on large-scale facilities such as data centers.
At present, a compact version of LLaMA can be used in GitHub Meta has no plan to publish the complete model and weight. Meta provides a Registration page , interested developers can apply for registration to obtain complete code and weight.