InstructPalmyra (30B): benchmark results
Provider: Other.
Aggregate Rating 1224 ± 25, rank #9454 of 10916 rated models, from 27 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - QuAC | 43.34 | F1 (%) | 89.2 |
| HELM Classic - NaturalQuestions Open Book | 68.21 | F1 (%) | 87.7 |
| HELM Classic - Entity Data Imputation | 83.27 | Exact Match (%) | 83.3 |
| HELM Classic - CNN/DailyMail | 15.16 | ROUGE-2 (%) | 79.3 |
| HELM Classic - Synthetic Reasoning Natural | 24.68 | F1 (%) | 73.5 |
| HELM Classic - NaturalQuestions Closed Book | 33 | F1 (%) | 66.7 |
| HELM Classic - BBQ | 37.03 | Exact Match (%) | 63.4 |
| RealToxicityPrompts | 0.05 | Toxic fraction | 63.4 |
| HELM Classic - CivilComments | 55.51 | Exact Match (%) | 62.1 |
| HELM Classic - MMLU | 40.27 | Exact Match (%) | 60.6 |
| HELM Classic - RAFT | 65.15 | Exact Match (%) | 60.6 |
| HELM Classic - IMDB | 94.03 | Exact Match (%) | 59.1 |
Interactive version: theaggregate.ai/model?slug=instructpalmyra-30b · How It Works · Data refreshed daily, snapshot 2026-11-08.