[keep_reading]

Apple has released a paper about running LLMs efficiently with low memory

https://arxiv.org/pdf/2312.11514.pdf