CoolFace
Datasetpublic

AITeamUIT/eval-gliner2-ner-classweight-3seed

Class-weighted loss — 3 seed x 4 dataset train: --loss-class-weight tren fastino/gliner2-multi-v1, batch 16, patience 2, seeds [42, 43, 44] eval : --threshold 0.7 --use-desc --schema-case original --no-extra-val (giong loss ablation) dataset epochs eval split strict macro F1 lenient micro F1 conll2003 6 test 86.06 +/- 0.95 89.31 +/- 0.95 ontonotes5 3 test 77.47 +/- 0.51 90.30 +/- 0.06 fin 8 test 38.15 +/- 2.40 75.92 +/- 1.36 mit_restaurant 6 test 81.55 +/-… See the full description on the dataset page: https://huggingface.co/datasets/AITeamUIT/eval-gliner2-ner-classweight-3seed.

sourceHugging Faceupdated 1mo agoView on Hugging Face
0likes8downloads
Dataset Card

Class-weighted loss — 3 seed x 4 dataset

  • train: --loss-class-weight tren fastino/gliner2-multi-v1, batch 16, patience 2, seeds [42, 43, 44]
  • eval : --threshold 0.7 --use-desc --schema-case original --no-extra-val (giong loss ablation)
datasetepochseval splitstrict macro F1lenient micro F1
conll20036test86.06 +/- 0.9589.31 +/- 0.95
ontonotes53test77.47 +/- 0.5190.30 +/- 0.06
fin8test38.15 +/- 2.4075.92 +/- 1.36
mit_restaurant6test81.55 +/- 0.2688.25 +/- 0.66

Dong cho tab:macro-f1 (strict macro F1)

latex
Class-weighted           & $ 86.06_{\pm 0.95} $ & $ 77.47_{\pm 0.51} $ & $ 38.15_{\pm 2.40} $ & $ 81.55_{\pm 0.26} $ & 70.81 \\

Checkpoint tung run

generated 2026-08-22T14:20:16