models · abacusai
AbacusAI Smaug 72B — DPO-trained on Qwen for best average benchmark across leaderboard.
| Weight file SHA-256 | ad1a35e1c0b63dfc278814b4c9474861880d10afc8728fbac4e9d8e59a00619c |
|---|---|
| Method | hf-published-sha256 |
| Publisher | abacusai |
|---|---|
| Origin | United States |
| Licence | Apache-2.0 |
| Publisher’s own licence statement | other |
| Parameters | 72,288,575,488 (72.3B) |
| Architecture | llama |
| Layers | 80 |
| Hidden size | 8192 |
| Attention | MHA (64 heads, 64 KV) |
| Native precision | BF16 |
| Pulls recorded here | 16,006 |
| Downloads reported upstream | 8,712 |
Registry record as of 2026-06-11. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.