Llama 3.2 3B Instruct
Open Sourcedi Meta · Rilascio 2024-09-25
24.2
punteggio medio
$0.05/1M
Prezzo Input
$0.34/1M
Prezzo Output
80K tokens (~40 books)
Finestra di Contesto
text
Tipo
Tested on 7 benchmarks with 24.2% average. Top scores: Chatbot Arena Elo — Overall (1165.7%), IFEval (73.9%), MMLU-PRO (24.4%).
Punteggi Benchmark
| Benchmark | Categoria | Punteggio | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1165.7 | |
| IFEval | language | 73.9 | |
| MMLU-PRO | knowledge | 24.4 | |
| BBH (HuggingFace) | general | 24.1 | |
| MATH Level 5 | math | 17.7 | |
| GPQA | knowledge | 3.8 | |
| MUSR | reasoning | 1.4 |
Modelli Simili
U
StarCoder 2 15BUnknown
24.3
U
Baichuan1-7BUnknown
23.7
Meta
23.6
Mistral AI
23.5