Master-Yi-9B: benchmark results
qnguyen3's Master-Yi-9B, ORPO fine-tuned on Yi-9B, using human-collected seed questions with answers regenerated by a mix of open-source LLMs. Provider: qnguyen3. Released 2024-03-06. Access: Open.
Aggregate Rating 1425 ± 15, rank #1102 of 1635 rated models, from 39 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 71.64 | Accuracy (%) (5-shot) | 95.1 |
| Open Medical LLM - MMLU College Biology | 80.56 | Accuracy (%) | 87.5 |
| Open Medical LLM - PubMedQA | 76.2 | Accuracy (%) | 82.6 |
| Open LLM Leaderboard v1 - GSM8K | 65.5 | Accuracy (%) (5-shot) | 80.2 |
| Open Medical LLM - MMLU Anatomy | 67.41 | Accuracy (%) | 76.3 |
| Open Korean LLM - Ko-GPQA-Diamond | 25.25 | Normalized accuracy (%) | 75.9 |
| Open Chinese LLM - CMMLU | 64.24 | Accuracy (%) | 75.7 |
| Open Chinese LLM - C-Eval Semantic | 84.03 | Accuracy (%) | 74.9 |
| Open Medical LLM - MMLU College Medicine | 65.9 | Accuracy (%) | 74.7 |
| Open Chinese LLM Leaderboard | 62.61 | Official Nonzero-Task Mean (%) | 74.3 |
| Open Chinese LLM - ARC Challenge | 56.66 | Accuracy (%) | 72.9 |
| Open Chinese LLM - GSM8K | 53.53 | Accuracy (%) | 72.8 |
Interactive version: theaggregate.ai/model?slug=master-yi-9b · How It Works · Data refreshed daily, snapshot 2026-11-05.