phi-3-medium 14B
Code source ouvertpar Microsoft · Sorti le 2024-01-01
58.6
score moyen
N/A
Prix d'entrée
N/A
Prix de sortie
N/A
Fenêtre de contexte
text
Type
Tested on 10 benchmarks with 58.6% average. Top scores: ARC AI2 (88.8%), OpenBookQA (83.2%), HellaSwag (76.5%).
Scores de benchmark
| Benchmark | Catégorie | Score | Bar |
|---|---|---|---|
| ARC AI2 | knowledge | 88.8 | |
| OpenBookQA | knowledge | 83.2 | |
| HellaSwag | knowledge | 76.5 | |
| BBH | reasoning | 75.2 | |
| TriviaQA | knowledge | 73.9 | |
| MMLU | knowledge | 70.7 | |
| Winogrande | knowledge | 63.0 | |
| ANLI | knowledge | 33.7 | |
| MATH level 5 | math | 17.6 | |
| GPQA diamond | knowledge | 3.5 |