1.58-bit LLMs: thousands of tokens/sec, reduced computing footprint (huggingface.co) 1 points by Rabot 2y ago ↗ HN
0 comments
[ 3.7 ms ] story [ 13.2 ms ] threadNo comments yet.