A complete Llama2 inference engine that fits in 1356 bytes of x86 assembly (github.com) 27 points by monax 4mo ago ↗ HN
0 comments
[ 4.5 ms ] story [ 16.3 ms ] threadNo comments yet.