WizardVicuna-open-llama-3B-v2: benchmark results
heegyu's WizardVicuna open-llama 3B v2, a partial fine-tune of OpenLM Research's open_llama_3b_v2 stopped at 3 of 8 planned epochs. Provider: heegyu. Released 2023-08-25. Access: Open.
Aggregate Rating 1154 ± 16, rank #9905 of 10916 rated models, from 77 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - High School Physics | 32.45 | Accuracy (%) | 66 |
| Open Korean LLM - Ko-GPQA-Diamond | 24.24 | Normalized accuracy (%) | 62.4 |
| MMLU-by-task - College Physics | 25.49 | Accuracy (%) | 53.2 |
| MMLU-by-task - High School Statistics | 39.35 | Accuracy (%) | 52.9 |
| MMLU-by-task - Global Facts | 33 | Accuracy (%) | 51.6 |
| Open Korean LLM - Ko-IFEval | 26.94 | Strict accuracy, prompt/instruction mean (%) | 41.1 |
| MMLU-by-task - Security Studies | 40.41 | Accuracy (%) | 40.8 |
| MMLU-by-task - Moral Scenarios | 24.69 | Accuracy (%) | 38.9 |
| MMLU-by-task - Professional Accounting | 29.43 | Accuracy (%) | 37.6 |
| MMLU-by-task - Formal Logic | 26.98 | Accuracy (%) | 35.6 |
| MMLU-by-task - Business Ethics | 35 | Accuracy (%) | 34.3 |
| MMLU-by-task - Electrical Engineering | 31.03 | Accuracy (%) | 34.2 |
Interactive version: theaggregate.ai/model?slug=wizardvicuna-open-llama-3b-v2 · How It Works · Data refreshed daily, snapshot 2026-11-08.