Llama 2 70B Guanaco QLoRA FP16: benchmark results
TheBloke's Hugging Face publication of Llama2 70b Guanaco QLoRA, whose original release is Mikael110/llama-2-70b-guanaco-qlora. Conversion publication is credited separately from the original model creator. Provider: TheBloke. Access: Open.
Aggregate Rating 1441 ± 17, rank #4172 of 10916 rated models, from 68 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - Electrical Engineering | 67.59 | Accuracy (%) | 100 |
| MMLU-by-task - Global Facts | 54 | Accuracy (%) | 100 |
| MMLU-by-task - Prehistory | 85.19 | Accuracy (%) | 100 |
| MMLU-by-task - High School Physics | 50.99 | Accuracy (%) | 99.9 |
| MMLU-by-task - Nutrition | 78.43 | Accuracy (%) | 99.9 |
| MMLU-by-task - HellaSwag | 69.53 | Accuracy (%) | 99.7 |
| MMLU-by-task - College Medicine | 68.79 | Accuracy (%) | 99.6 |
| MMLU-by-task - High School Statistics | 62.96 | Accuracy (%) | 99.5 |
| MMLU-by-task - Moral Disputes | 79.77 | Accuracy (%) | 99.5 |
| MMLU-by-task - World Religions | 87.72 | Accuracy (%) | 99.3 |
| MMLU-by-task - Human Aging | 80.72 | Accuracy (%) | 99 |
| MMLU-by-task - Virology | 55.42 | Accuracy (%) | 98.8 |
Interactive version: theaggregate.ai/model?slug=llama-2-70b-guanaco-qlora-fp16 · How It Works · Data refreshed daily, snapshot 2026-11-08.