HyperLlama-3.1-8B: benchmark results
bunnycore's model_stock merge of three Llama 3.1 8B fine-tunes: Replete-LLM V2, Ultra-Instruct, and Lexi Uncensored V2. Provider: bunnycore. Released 2024-09-04. Access: Open.
Aggregate Rating 1449 ± 19, rank #1012 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 78.83 | Score | 95.7 |
| Open Korean LLM - Ko-IFEval | 52.93 | Strict accuracy, prompt/instruction mean (%) | 95.4 |
| Open Ko-LLM Leaderboard | 41.87 | Average (%) | 83.7 |
| Open Korean LLM Leaderboard | 41.87 | Average Score (%) | 83.7 |
| Open Korean LLM - Ko-EQ-Bench | 32.85 | EQ-Bench score | 82.2 |
| Open Portuguese LLM - OAB Exams | 51.66 | Accuracy (%) | 80.7 |
| Open Portuguese LLM - ENEM | 71.45 | Accuracy (%) | 80.4 |
| Open Portuguese LLM - BLUEX | 59.39 | Accuracy (%) | 76.4 |
| Open Korean LLM - Ko-GPQA-Diamond | 25.25 | Normalized accuracy (%) | 75.9 |
| Open Portuguese LLM - ASSIN2 RTE | 92.15 | Macro F1 (%) | 74.9 |
| Open Portuguese LLM - FaQuAD NLI | 74 | Macro F1 (%) | 69.8 |
| Open LLM Leaderboard - MATH Level 5 | 18.28 | Score | 68 |
Interactive version: theaggregate.ai/model?slug=hyperllama-3-1-8b · How It Works · Data refreshed daily, snapshot 2026-10-23.