granite-34B-code-base: benchmark results
Provider: IBM. Released 2024-05-06. Access: Open.
Aggregate Rating 1390 ± 38, rank #6532 of 10916 rated models, from 104 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval English NLU - SQuAD | 86.96 | Character F1 (%) | 92.9 |
| EuroEval French NLU - FQuAD | 74.02 | Character F1 (%) | 91.7 |
| EuroEval Spanish NLU - MLQA ES | 67.74 | Character F1 (%) | 91.4 |
| EuroEval German NLU - Germanquad | 65.48 | Character F1 (%) | 90.9 |
| EuroEval Finnish NLU - Tydiqa FI | 71.72 | Character F1 (%) | 87.9 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 75.17 | Character F1 (%) | 85.7 |
| EuroEval Dutch NLU - SQuAD NL | 77.31 | Character F1 (%) | 83.2 |
| EuroEval Portuguese NLU - HAREM | 50.81 | Micro F1 excluding MISC (%) | 80 |
| EuroEval Italian NLU - MultiNERD IT | 74.28 | Micro F1 excluding MISC (%) | 78.7 |
| EuroEval Norwegian NLU - Norquad | 69.23 | Character F1 (%) | 69.6 |
| EuroEval Icelandic NLU - MIM-GOLD NER | 56.2 | Micro F1 excluding MISC (%) | 67.8 |
| EuroEval Finnish NLU - Turku NER FI | 57.37 | Micro F1 excluding MISC (%) | 67.6 |
Interactive version: theaggregate.ai/model?slug=granite-34b-code-base · How It Works · Data refreshed daily, snapshot 2026-11-08.