Follow
Quantized Llama models with increased speed and a reduced memory footprint
https://news.ycombinator.com/item?id=41938473
#hackernews #tech
