T.E-8.1: benchmark results
Cran-May's bilingual Chinese-English roleplay fine-tune of Qwen2.5 7B Instruct. Provider: Cran-May. Released 2024-09-27. Access: Open.
Aggregate Rating 1527 ± 17, rank #694 of 1656 rated models, from 144 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GSM8K (flexible extract) | 72.71 | Exact match, flexible extract (%) | 98.7 |
| Open Ko-LLM Leaderboard | 49.57 | Average (%) | 96.9 |
| Open Korean LLM - Ko-IFEval | 58.99 | Strict accuracy, prompt/instruction mean (%) | 96.9 |
| Open Korean LLM Leaderboard | 49.57 | Average Score (%) | 96.9 |
| Open LLM Leaderboard - MATH Level 5 | 44.56 | Score | 93.2 |
| Open Korean LLM - Ko-EQ-Bench | 42.97 | EQ-Bench score | 92.7 |
| Open Portuguese LLM - ENEM | 75.86 | Accuracy (%) | 89.7 |
| Open Portuguese LLM - ASSIN2 RTE | 93.42 | Macro F1 (%) | 89.3 |
| Open Portuguese LLM - BLUEX | 66.48 | Accuracy (%) | 88.5 |
| Open Korean LLM - KorNAT-CKA | 31.66 | Normalized accuracy (%) | 87.3 |
| Open Portuguese LLM - OAB Exams | 54.81 | Accuracy (%) | 87.2 |
| Open LLM Leaderboard - IFEval | 70.77 | Score | 84 |
Interactive version: theaggregate.ai/model?slug=t-e-8-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.