granite-20B-code-base: benchmark results
Provider: IBM. Released 2024-05-06. Access: Open.
Aggregate Rating 1375 ± 35, rank #7011 of 10916 rated models, from 106 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval English NLU - SQuAD | 85.84 | Character F1 (%) | 90.5 |
| ShaderMatch | 15.74 | Clone Match Rate (%) | 83.3 |
| EuroEval German NLU - Germanquad | 60.9 | Character F1 (%) | 81.2 |
| EuroEval Finnish NLU - Tydiqa FI | 68.61 | Character F1 (%) | 80.7 |
| EuroEval Italian NLU - MultiNERD IT | 74.37 | Micro F1 excluding MISC (%) | 79.2 |
| EuroEval Portuguese NLU - HAREM | 50.48 | Micro F1 excluding MISC (%) | 78.6 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 72.74 | Character F1 (%) | 71.9 |
| EuroEval French NLU - FQuAD | 68.48 | Character F1 (%) | 67.4 |
| EuroEval Spanish NLU - MLQA ES | 62.59 | Character F1 (%) | 66.7 |
| EuroEval Icelandic NLU - MIM-GOLD NER | 55.19 | Micro F1 excluding MISC (%) | 66.4 |
| EuroEval French NLU - Eltec | 59.25 | Micro F1 excluding MISC (%) | 63.4 |
| EuroEval Dutch NLU - CoNLL NL | 60.77 | Micro F1 excluding MISC (%) | 61.9 |
Interactive version: theaggregate.ai/model?slug=granite-20b-code-base · How It Works · Data refreshed daily, snapshot 2026-11-08.