Qwen2 Instruct 72B
AlibabaQwenOpen Weighttongyi-qianwen
Descripción
Qwen2-72B-Instruct is an instruction-tuned language model with 72 billion parameters, supporting a context length of up to 131,072 tokens. It's part of the new Qwen2 series, which has surpassed most open-source models and demonstrates competitiveness against proprietary models across various benchmarks.
Fecha de lanzamiento
2024-06-07
Parámetros
72.0B
Longitud del contexto
—
Modalidades
—
Radar de capacidades
23
general
16
coding
36
reasoning
27
science
30
agents
0
multimodal
Rankings
| Dominio | #Posición | Puntuación | Fuente |
|---|---|---|---|
| Ranking de codificación | 533 | 17.0 | AA |
| Ranking general | 525 | 26.0 | AA |
| Ciencia | 582 | 18.0 | AA |
Puntuaciones de benchmarks (LLM Stats)
(LLM Stats (zeroeval))General
C-Eval
83.8%Aut.
MMLU
82.3%Aut.
MultiPL-E
69.2%Aut.
TruthfulQA
54.8%Aut.
Language
CMMLU
90.1%Aut.
MMLU-Pro
64.4%Aut.
Math
GSM8k
91.1%Aut.
MATH
59.7%Aut.
TheoremQA
44.4%Aut.
Reasoning
HellaSwagAI2 (2019)
87.6%Aut.
HumanEvalOpenAI (2021)
86.0%Aut.
Winogrande
85.1%Aut.
BBH
82.4%Aut.
MBPP
0.80 / 100Aut.
EvalPlus
0.79 / 100Aut.
ARC-C
68.9%Aut.
GPQANYU + Cohere + Anthropic (2023)
42.4%Aut.
Índices de evaluación AA
(Artificial Analysis)Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))70.1
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))62.2
Gpqa(NYU + Cohere + Anthropic (2023))37.1
Livecodebench(UC Berkeley + MIT + Cornell (2024))15.9
Aime(MAA (Mathematical Association of America))14.7
Intelligence Index(Artificial Analysis)6.3
Hle(Center for AI Safety + Scale AI (2025))3.7
Puntuaciones por categoría LLM Stats
(LLM Stats (zeroeval))Language80
Code80
Legal70
Math70
Reasoning70
General70
Healthcare70
Finance60
Physics40
Biology40
Chemistry40
Precios
Precio de entradaGratis
Precio de salidaGratis
Precio mixto (3:1)Gratis
Velocidad
Tokens/seg0.0
Retraso del primer token0.00s
Tiempo hasta la respuesta0.00s
Ranking de Precios por Proveedor
No hay datos de proveedores disponibles