airoboros-7B-gpt4-fp16: benchmark results
TheBloke's Hugging Face publication of Jon Durbin's Airoboros 7B GPT4, whose original release is jondurbin/airoboros-7b-gpt4. Conversion publication is credited separately from the original model creator. Provider: TheBloke. Access: Open.
Aggregate Rating 1242 ± 16, rank #9321 of 10916 rated models, from 68 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - Global Facts | 40 | Accuracy (%) | 88.7 |
| MMLU-by-task - HellaSwag | 60.06 | Accuracy (%) | 58.1 |
| MMLU-by-task - International Law | 61.98 | Accuracy (%) | 50.5 |
| MMLU-by-task - ARC Challenge | 49.83 | Accuracy (%) | 49.9 |
| MMLU-by-task - Professional Medicine | 45.22 | Accuracy (%) | 48.2 |
| MMLU-by-task - High School Physics | 29.8 | Accuracy (%) | 46.5 |
| MMLU-by-task - Abstract Algebra | 29 | Accuracy (%) | 45.4 |
| MMLU-by-task - Electrical Engineering | 40.69 | Accuracy (%) | 45.3 |
| MMLU-by-task - Elementary Mathematics | 27.78 | Accuracy (%) | 45.3 |
| MMLU-by-task - High School Geography | 49.49 | Accuracy (%) | 44.7 |
| MMLU-by-task - Professional Law | 33.44 | Accuracy (%) | 44.7 |
| MMLU-by-task - Logical Fallacies | 45.4 | Accuracy (%) | 44.4 |
Interactive version: theaggregate.ai/model?slug=airoboros-7b-gpt4-fp16 · How It Works · Data refreshed daily, snapshot 2026-11-08.