Llama 3 70B Instruct
Open Sourcedi Meta · Rilascio 2024-04-18
32.4
punteggio medio
$0.51/1M
Prezzo Input
$0.74/1M
Prezzo Output
8K tokens (~4 books)
Finestra di Contesto
text
Tipo
Tested on 9 benchmarks with 32.4% average. Top scores: Chatbot Arena Elo — Overall (1275.1%), MMLU (72.4%), Winogrande (67.0%).
Punteggi Benchmark
| Benchmark | Categoria | Punteggio | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1275.1 | |
| MMLU | knowledge | 72.4 | |
| Winogrande | knowledge | 67.0 | |
| CMMLU | knowledge | 36.9 | |
| OpenBookQA | knowledge | 30.1 | |
| MATH level 5 | math | 22.6 | |
| GPQA diamond | knowledge | 20.8 | |
| Cybench | coding | 5.0 | |
| OTIS Mock AIME 2024-2025 | math | 4.2 |