granite-3B-code-instruct-2k: benchmark results
Provider: IBM. Released 2024-05-06. Access: Open.
Aggregate Rating 1234 ± 37, rank #9384 of 10916 rated models, from 70 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Icelandic NLU - ScaLA IS | 2.85 | Matthews correlation coefficient (%) | 47.4 |
| EuroEval English NLU - SQuAD | 78.94 | Character F1 (%) | 45.7 |
| EuroEval Italian NLU - MultiNERD IT | 57.95 | Micro F1 excluding MISC (%) | 44.5 |
| EuroEval Spanish NLU - MLQA ES | 57.68 | Character F1 (%) | 43.3 |
| EuroEval Portuguese NLU - HAREM | 39.46 | Micro F1 excluding MISC (%) | 41.9 |
| EuroEval Icelandic NLU - NQII | 34.26 | Character F1 (%) | 36.7 |
| EuroEval Italian NLU - SQuAD IT | 55.66 | Character F1 (%) | 36.3 |
| EuroEval Norwegian Knowledge - Idioms NO | 0.89 | MCC (x100) | 34.9 |
| EuroEval French NLU - FQuAD | 56.31 | Character F1 (%) | 33.5 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 57.8 | Character F1 (%) | 32.9 |
| EuroEval Portuguese NLU - ScaLA PT | 3.91 | Matthews correlation coefficient (%) | 32.6 |
| EuroEval Italian NLU - ScaLA IT | 6.84 | Matthews correlation coefficient (%) | 32.3 |
Interactive version: theaggregate.ai/model?slug=granite-3b-code-instruct-2k · How It Works · Data refreshed daily, snapshot 2026-11-08.