calme-2.5-qwen2-7B: benchmark results
MaziyarPanahi's Qwen2 7B fine-tune from the Calme-2 series, a sibling of calme-2.1 aiming to improve the base model across benchmarks. Provider: MaziyarPanahi. Released 2024-06-27. Access: Open.
Aggregate Rating 1445 ± 40, rank #4004 of 10916 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GSM8K (flexible extract) | 56.94 | Exact match, flexible extract (%) | 91.1 |
| Open LLM Leaderboard - MuSR | 45.65 | Score | 86.4 |
| Open Korean LLM Leaderboard | 42.12 | Average Score (%) | 84.9 |
| Open Korean LLM - KorNAT-Helpful | 49.2 | Normalized accuracy (%) | 77.5 |
| Open LLM Leaderboard - MATH Level 5 | 22.58 | Score | 75.8 |
| Open Korean LLM - Ko-EQ-Bench | 22.7 | EQ-Bench score | 68.3 |
| Open Korean LLM - Ko-IFEval | 33.69 | Strict accuracy, prompt/instruction mean (%) | 68.2 |
| Open LLM Leaderboard - GPQA | 31.04 | Score | 67.9 |
| Open Korean LLM - KorNAT-CKA | 25.73 | Normalized accuracy (%) | 60.3 |
| Open LLM Leaderboard - MMLU-Pro | 36.82 | Score | 58.9 |
| Open Korean LLM - Ko-Winogrande | 59.51 | Accuracy (%) | 46.7 |
| Open LLM Leaderboard - BBH | 48.87 | Score | 43.9 |
Interactive version: theaggregate.ai/model?slug=calme-2-5-qwen2-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.