AI Hub
All papers
Training

LLaMA: Open and Efficient Foundation Language Models

Meta·February 27, 2023

Hugo Touvron, Thibaut Lavril, Gautier Izacard

View on arXiv

TL;DR

Releases efficient foundation models that match much larger ones, igniting the open-model boom.

Why it matters

The release whose leaked weights seeded a vast open fine-tuning ecosystem almost overnight.

Related models

Related terms