BitNet: Scaling 1-bit Transformers for Large Language Models (2023): https://arxiv.org/abs/2310.11453
I do like the idea of local/on device llms so it's interesting and will keep an eye on it.
