ANNOUNCE  box benchbox · card 0 (one RTX 3090, 250 W) · arm 11 · train encoder `longtable_voices` · scripts/train.py --task tasks/longtable_voices --epochs 6 --bs 8 --init-from checkpoints/general/model.pt --overwrite · 2026-09-21T22:20:59Z
longtable_voices     heldout n=30  trained 70.0%  openjev 86.7%  gemma4 70.0%  floor 16.7%  14s  598 MB
ANNOUNCE  box benchbox · card 0 (one RTX 3090, 250 W) · arm 11 · train encoder `doorman_gate` · scripts/train.py --task tasks/doorman_gate --epochs 6 --bs 8 --init-from checkpoints/general/model.pt --overwrite · 2026-09-21T22:21:20Z
doorman_gate         heldout n=15  trained 86.7%  openjev 100.0%  gemma4 93.3%  floor 73.3%  10s  598 MB
ANNOUNCE  box benchbox · card 0 (one RTX 3090, 250 W) · arm 11 · train encoder `fieldexam_grounded` · scripts/train.py --task tasks/fieldexam_grounded --epochs 6 --bs 8 --init-from checkpoints/general/model.pt --overwrite · 2026-09-21T22:21:37Z
fieldexam_grounded   heldout n=12  trained 100.0%  openjev 100.0%  gemma4 100.0%  floor 50.0%  10s  598 MB