Key Specifications

Vendoralibaba
Version2.5-7b
Release Date2024-09-19
Context Window131072 tokens
Input Modalitiestext
Output Modalitiestext
LicenseQwen License
Documentationhttps://qwenlm.github.io/

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU73.6%2024-09-195-shotview
HUMANEVAL66.5pass@12024-09-19view
GSM8K63.7%2024-09-190-shot CoTview
MATH41.9%2024-09-190-shot CoTview
BBH61.9%2024-09-193-shot CoTview
GPQA28.5%2024-09-190-shotview
IFEVAL55.4%2024-09-19prompt_strictview
ARC88.4%2024-09-19challengeview
MUSR46.9%2024-09-190-shotview
WINOGRANDE71.4%2024-09-190-shotview

Pricing

TierPriceCurrency
Input$0.15 / MtokUSD
Output$0.15 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://help.aliyun.com/zh/model-studio/getting-started/models · as of 2024-09-19

Compliance

  • Data Residency: CN
  • SOC2: ✗
  • HIPAA: ✗
  • GDPR: ✗
  • ISO 27001: ✗

Qwen2.5 7B

Modellöversikt

阿里巴巴 Qwen2.5 7B 经济型开源模型, 131K 上下文, 7B 参数, 性能超越 Llama 3 8B, 适合本地部署。

Kärnspecifikationer

LeverantörVersionReleasedatumKontextfönsterInmatningsmodaliteterUtmatningsmodaliteterLicens
Alibaba2.5-7b2024-09-19131KtexttextQwen License

Benchmarkprestanda

BenchmarkPoängEnhetAnteckningar
MMLU (Massive Multitask Language Understanding)73.6%5-shot
HumanEval66.5pass@1
GSM8K (Grade School Math 8K)63.7%0-shot CoT
MATH41.9%0-shot CoT
BBH (BIG-Bench Hard)61.9%3-shot CoT
GPQA28.5%0-shot
IFEval55.4%prompt_strict
ARC88.4%challenge
MUSR46.9%0-shot
WinoGrande71.4%0-shot

Priser

InmatningUtmatningCacheläsningCacheskrivning

per miljon tokens

Styrkor

  • 可靠的通用模型。

Svagheter

  • 闭源专有模型,不支持自托管。

Användningsområden

  • 通用对话与问答

Referenser