llama2-ko-7B-model: benchmark results
A community fine-tune of Llama 2 7B for Korean and English by jb723, trained on the AIHUB Korean-English parallel corpus plus the KoAlpaca and Alpaca datasets. Provider: Other. Released 2023-09-07. Access: Open.
Aggregate Rating 1284 ± 14, rank #8934 of 10916 rated models, from 77 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - Global Facts | 39 | Accuracy (%) | 85.7 |
| MMLU-by-task - College Mathematics | 37 | Accuracy (%) | 85.2 |
| MMLU-by-task - Econometrics | 32.46 | Accuracy (%) | 77.1 |
| MMLU-by-task - HellaSwag | 61.58 | Accuracy (%) | 69 |
| MMLU-by-task - Elementary Mathematics | 31.75 | Accuracy (%) | 65.7 |
| MMLU-by-task - ARC Challenge | 53.5 | Accuracy (%) | 63.2 |
| MMLU-by-task - High School Physics | 31.79 | Accuracy (%) | 61.3 |
| MMLU-by-task - Abstract Algebra | 31 | Accuracy (%) | 60.2 |
| MMLU-by-task - College Chemistry | 36 | Accuracy (%) | 59.4 |
| MMLU-by-task - Conceptual Physics | 41.28 | Accuracy (%) | 58.2 |
| MMLU-by-task - Virology | 42.17 | Accuracy (%) | 56.8 |
| MMLU-by-task - High School Government And Politics | 68.91 | Accuracy (%) | 56.7 |
Interactive version: theaggregate.ai/model?slug=llama2-ko-7b-model · How It Works · Data refreshed daily, snapshot 2026-11-08.