Qwen1.5 14B: Complete Benchmark Performance Guide

alibaba 1.5-14b 2024-02-25 Qwen License

In-depth analysis of Qwen1.5 14B performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendoralibaba
Version1.5-14b
Release Date2024-02-25
Context Window32768 tokens
Input Modalitiestext
Output Modalitiestext
LicenseQwen License
Documentationhttps://qwenlm.github.io/

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU78.4%2024-02-255-shotview
HUMANEVAL65.8pass@12024-02-25—view
GSM8K84%2024-02-250-shot CoTview
MATH46.2%2024-02-250-shot CoTview
BBH71.6%2024-02-253-shot CoTview
GPQA33.9%2024-02-250-shotview
IFEVAL79.9%2024-02-25prompt_strictview
ARC92.7%2024-02-25challengeview
MUSR58.7%2024-02-250-shotview
WINOGRANDE78.6%2024-02-250-shotview

Pricing

TierPriceCurrency
Input$0.25 / MtokUSD
Output$0.35 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://help.aliyun.com/zh/model-studio/getting-started/models · as of 2024-02-25

Compliance

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

Qwen1.5 14B

모델 개요

阿里巴巴 Qwen1.5 14B 中型开源模型, 32K 上下文, 平衡性能与资源, 适合企业级应用。

핵심 사양

공급업체버전출시일컨텍스트 창입력 모달리티출력 모달리티라이선스
Alibaba1.5-14b2024-02-2532KtexttextQwen License

벤치마크 성능

벤치마크점수단위비고
MMLU (Massive Multitask Language Understanding)78.4%5-shot
HumanEval65.8pass@1—
GSM8K (Grade School Math 8K)84.0%0-shot CoT
MATH46.2%0-shot CoT
BBH (BIG-Bench Hard)71.6%3-shot CoT
GPQA33.9%0-shot
IFEval79.9%prompt_strict
ARC92.7%challenge
MUSR58.7%0-shot
WinoGrande78.6%0-shot

가격

입력출력캐시 읽기캐시 쓰기
————

백만 토큰당

강점

  • 可靠的通用模型。

약점

  • 闭源专有模型，不支持自托管。

사용 사례

  • 通用对话与问答

참고문헌