WizardVicuna-3B-0719: benchmark results
heegyu's WizardVicuna 3B (0719 release), a fine-tune of OpenLM Research's open_llama_3b base model. Provider: heegyu. Released 2023-07-23. Access: Open.
Aggregate Rating 1145 ± 16, rank #10038 of 10916 rated models, from 77 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - College Physics | 24.51 | Accuracy (%) | 48.1 |
| MMLU-by-task - TruthfulQA MC1 | 27.29 | Accuracy (%) | 45.3 |
| MMLU-by-task - Moral Scenarios | 24.8 | Accuracy (%) | 41.6 |
| Open Korean LLM - Ko-GPQA-Diamond | 22.73 | Normalized accuracy (%) | 40.9 |
| MMLU-by-task - Abstract Algebra | 28 | Accuracy (%) | 39.6 |
| Open Korean LLM - Ko-IFEval | 26.54 | Strict accuracy, prompt/instruction mean (%) | 39.2 |
| Open Korean LLM - KorNAT-Helpful | 45.48 | Normalized accuracy (%) | 37.1 |
| MMLU-by-task - Conceptual Physics | 32.77 | Accuracy (%) | 33.3 |
| MMLU-by-task - College Medicine | 30.06 | Accuracy (%) | 32.7 |
| MMLU-by-task - Moral Disputes | 30.06 | Accuracy (%) | 32 |
| MMLU-by-task - HellaSwag | 50.48 | Accuracy (%) | 31.5 |
| MMLU-by-task - ARC Challenge | 38.4 | Accuracy (%) | 31.2 |
Interactive version: theaggregate.ai/model?slug=wizardvicuna-3b-0719 · How It Works · Data refreshed daily, snapshot 2026-11-08.