StarCoder2 7B: Complete Benchmark Performance Guide

other starcoder2-7b 2024-02-28 BigCode Open Model License

In-depth analysis of StarCoder2 7B performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionstarcoder2-7b
Release Date2024-02-28
Context Window16384 tokens
Input Modalitiestext
Output Modalitiestext
LicenseBigCode Open Model License
Documentationhttps://huggingface.co/models

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU67.4%2024-02-285-shotview
HUMANEVAL74.6pass@12024-02-28—view
GSM8K65%2024-02-280-shot CoTview
MATH30.5%2024-02-280-shot CoTview
BBH60%2024-02-283-shot CoTview
GPQA33.6%2024-02-280-shotview
IFEVAL54%2024-02-28prompt_strictview
ARC90.5%2024-02-28challengeview
MUSR38.3%2024-02-280-shotview
WINOGRANDE73.1%2024-02-280-shotview

Pricing

TierPriceCurrency
Input$0.2 / MtokUSD
Output$0.2 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2024-02-28

Compliance

  • Data Residency: self-host
  • SOC2: ✗
  • HIPAA: ✗
  • GDPR: ✗
  • ISO 27001: ✗

StarCoder2 7B

모델 개요

BigCode StarCoder2 7B 经济型代码模型, 16K 上下文, 7B 参数, 适合本地代码补全与生成。

핵심 사양

공급업체버전출시일컨텍스트 창입력 모달리티출력 모달리티라이선스
Otherstarcoder2-7b2024-02-2816KtexttextBigCode Open Model License

벤치마크 성능

벤치마크점수단위비고
MMLU (Massive Multitask Language Understanding)67.4%5-shot
HumanEval74.6pass@1—
GSM8K (Grade School Math 8K)65.0%0-shot CoT
MATH30.5%0-shot CoT
BBH (BIG-Bench Hard)60.0%3-shot CoT
GPQA33.6%0-shot
IFEval54.0%prompt_strict
ARC90.5%challenge
MUSR38.3%0-shot
WinoGrande73.1%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。
  • 上下文窗口 16K 偏小。

사용 사례

  • 代码生成与调试

참고문헌