System One typed-decision model (choice / score / noul) — M2' dual encoder + option tower.
| Nagi T4 | Laya | Jev | |
|---|---|---|---|
| hard | 0.481 | ~0.35 | 0.773 |
| ECE raw | 0.070 | 0.081 | 0.121 |
| p50 | 9.2ms | 33ms | 501ms |
Beats Laya on LTO + latency. 55× faster than Jev API. Honest gap: LTO 0.48 vs Jev 0.77. Domain FT (prior): Banking77 0.942. Train: 17k rows / 643 schemas / H100 BF16 3ep.
Have the files? Submit a torrent.
Submit a torrent →Have the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo nagisanzeninz/nagi-t4-m2p-v0 ./model-folder