LLaMA-13B
Meta · 2023-02-24 · 13B parameters
LLaMA-13B is Meta's 13B pretrained base language model from the original LLaMA family, released in February 2023. It uses a 2,048-token context window and was trained on 1.0 trillion tokens of publicly available text; Meta distributed the weights to approved researchers under a bespoke non-commercial license.