Qwen1.5 110B: Complete Benchmark Performance Guide

alibaba 1.5-110b 2024-02-25 Qwen License

In-depth analysis of Qwen1.5 110B performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendoralibaba
Version1.5-110b
Release Date2024-02-25
Context Window32768 tokens
Input Modalitiestext
Output Modalitiestext
LicenseQwen License
Documentationhttps://qwenlm.github.io/

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU86.1%2024-02-255-shotview
HUMANEVAL76.5pass@12024-02-25—view
GSM8K80.6%2024-02-250-shot CoTview
MATH54.1%2024-02-250-shot CoTview
BBH82.3%2024-02-253-shot CoTview
GPQA45.9%2024-02-250-shotview
IFEVAL83.4%2024-02-25prompt_strictview
ARC95.7%2024-02-25challengeview
MUSR63.9%2024-02-250-shotview
WINOGRANDE84%2024-02-250-shotview

Pricing

TierPriceCurrency
Input$0.7 / MtokUSD
Output$1.2 / 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 110B

모델 개요

阿里巴巴 Qwen1.5 110B 开源超大模型, 32K 上下文, 在中文与多语言基准上表现突出。

핵심 사양

공급업체버전출시일컨텍스트 창입력 모달리티출력 모달리티라이선스
Alibaba1.5-110b2024-02-2532KtexttextQwen License

벤치마크 성능

벤치마크점수단위비고
MMLU (Massive Multitask Language Understanding)86.1%5-shot
HumanEval76.5pass@1—
GSM8K (Grade School Math 8K)80.6%0-shot CoT
MATH54.1%0-shot CoT
BBH (BIG-Bench Hard)82.3%3-shot CoT
GPQA45.9%0-shot
IFEval83.4%prompt_strict
ARC95.7%challenge
MUSR63.9%0-shot
WinoGrande84.0%0-shot

가격

입력출력캐시 읽기캐시 쓰기
————

백만 토큰당

강점

  • MMLU score 86.1, strong knowledge reasoning.

약점

  • 闭源专有模型，不支持自托管。

사용 사례

  • 代码生成与调试

참고문헌