Llama-3.1-SauerkrautLM-8B-Instruct: benchmark results
VAGO solutions' German-English fine-tune of Llama 3.1 8B Instruct using Spectrum training on 25% of the model's layers. Provider: VAGOsolutions. Released 2024-07-25. Access: Open.
Aggregate Rating 1499 ± 19, rank #790 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 80.17 | Score | 97.5 |
| Open Ko-LLM Leaderboard | 47.94 | Average (%) | 95.4 |
| Open Korean LLM Leaderboard | 47.94 | Average Score (%) | 95.4 |
| Open Korean LLM - Ko-IFEval | 51.95 | Strict accuracy, prompt/instruction mean (%) | 95 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 60.8 | Exact match, flexible extract (%) | 94.3 |
| Open Korean LLM - Ko-EQ-Bench | 37.78 | EQ-Bench score | 88.5 |
| Open Korean LLM - KorNAT-CKA | 31.26 | Normalized accuracy (%) | 86.5 |
| Open Portuguese LLM - ENEM | 71.17 | Accuracy (%) | 79.3 |
| Open Portuguese LLM - OAB Exams | 51.07 | Accuracy (%) | 79.1 |
| Open Portuguese LLM - ASSIN2 RTE | 92.4 | Macro F1 (%) | 78.7 |
| Open Korean LLM - KorNAT-Helpful | 48.83 | Normalized accuracy (%) | 73.2 |
| Open LLM Leaderboard - MMLU-Pro | 38.9 | Score | 70.1 |
Interactive version: theaggregate.ai/model?slug=llama-3-1-sauerkrautlm-8b-instruct · How It Works · Data refreshed daily, snapshot 2026-10-23.