LLaMA: Open and Efficient Foundation Language Models
We introduce LLaMA, a collection of foundation language models ranging from 7B to 65B parameters. We train our models on trillions of tokens, and show that it is possible to train state-of-the-art models using publicly available datasets exclusively, without resorting to proprietary and inaccessible datasets. In particular, LLaMA-13B outperforms GPT-3 (175B) on most benchmarks, and LLaMA-65B is competitive with the best models, Chinchilla-70B and PaLM-540B. We release all our models to the research community.
It's amazing work. Better they share the weights for research than no sharing at all, but they should share it with a license which allows commercial usage, too. It would feedback educatively, because LeCun struggles to see the full potential LLMs have. Multimodality is handy, not key.