Phi-2
Open Sourceby Microsoft 路 Released 2024-01-01
41.2
avg score
N/A
Input Price
N/A
Output Price
N/A
Context Window
text
Type
Tested on 8 benchmarks with 41.2% average. Top scores: ARC AI2 (67.9%), OpenBookQA (64.8%), BBH (45.9%).
Benchmark Scores
| Benchmark | Category | Score | Bar |
|---|---|---|---|
| ARC AI2 | knowledge | 67.9 | |
| OpenBookQA | knowledge | 64.8 | |
| BBH | reasoning | 45.9 | |
| TriviaQA | knowledge | 45.2 | |
| MMLU | knowledge | 44.5 | |
| HellaSwag | knowledge | 38.1 | |
| ANLI | knowledge | 13.8 | |
| Winogrande | knowledge | 9.4 |