granite-3B-code-base-2k: benchmark results
Provider: IBM. Released 2024-05-06. Access: Open.
Aggregate Rating 1291 ± 33, rank #8858 of 10916 rated models, from 97 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Portuguese NLU - HAREM | 43.25 | Micro F1 excluding MISC (%) | 56.3 |
| EuroEval English NLU - SQuAD | 80.39 | Character F1 (%) | 54.9 |
| EuroEval French NLU - Eltec | 54.52 | Micro F1 excluding MISC (%) | 52.4 |
| EuroEval Norwegian Knowledge - Idioms NO | 3.31 | MCC (x100) | 47.8 |
| EuroEval Swedish NLU - SUC3 | 51.94 | Micro F1 excluding MISC (%) | 43 |
| EuroEval Faroese NLU - FONE | 56.93 | Micro F1 excluding MISC (%) | 41.8 |
| EuroEval Spanish NLU - MLQA ES | 56.38 | Character F1 (%) | 40.6 |
| EuroEval Norwegian NLU - NorNE NB | 56.65 | Micro F1 excluding MISC (%) | 40.3 |
| EuroEval Icelandic NLU - NQII | 36.02 | Character F1 (%) | 38.9 |
| EuroEval Dutch NLU - CoNLL NL | 51.45 | Micro F1 excluding MISC (%) | 37.6 |
| EuroEval Italian NLU - SQuAD IT | 56.95 | Character F1 (%) | 37.2 |
| EuroEval French NLU - FQuAD | 58.7 | Character F1 (%) | 36.2 |
Interactive version: theaggregate.ai/model?slug=granite-3b-code-base-2k · How It Works · Data refreshed daily, snapshot 2026-11-08.