Command Light: Complete Benchmark Performance Guide

cohere light 2023-09-14 CC-BY-NC-4.0

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

Key Specifications

Vendorcohere
Versionlight
Release Date2023-09-14
Context Window4000 tokens
Input Modalitiestext
Output Modalitiestext
LicenseCC-BY-NC-4.0
Documentationhttps://docs.cohere.com/docs

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU52.8%2023-09-145-shotview
HUMANEVAL51.2pass@12023-09-14—view
GSM8K39.8%2023-09-140-shot CoTview
MATH24.1%2023-09-140-shot CoTview
BBH55.5%2023-09-143-shot CoTview
GPQA30.1%2023-09-140-shotview
IFEVAL53.9%2023-09-14prompt_strictview
ARC85.9%2023-09-14challengeview
MUSR40.7%2023-09-140-shotview
WINOGRANDE74.7%2023-09-140-shotview

Pricing

TierPriceCurrency
Input$0.5 / MtokUSD
Output$0.5 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://cohere.com/pricing · as of 2023-09-14

Compliance

  • Data Residency: US
  • SOC2: ✓
  • HIPAA: ✗
  • GDPR: ✓
  • ISO 27001: ✓

Command Light

모델 개요

Cohere Command Light 经济型模型, 4K 上下文, 速度快成本低, 适合简单任务与高吞吐场景。

핵심 사양

공급업체버전출시일컨텍스트 창입력 모달리티출력 모달리티라이선스
Coherelight2023-09-144KtexttextCC-BY-NC-4.0

벤치마크 성능

벤치마크점수단위비고
MMLU (Massive Multitask Language Understanding)52.8%5-shot
HumanEval51.2pass@1—
GSM8K (Grade School Math 8K)39.8%0-shot CoT
MATH24.1%0-shot CoT
BBH (BIG-Bench Hard)55.5%3-shot CoT
GPQA30.1%0-shot
IFEval53.9%prompt_strict
ARC85.9%challenge
MUSR40.7%0-shot
WinoGrande74.7%0-shot

가격

입력출력캐시 읽기캐시 쓰기
————

백만 토큰당

강점

  • 可靠的通用模型。

약점

  • MMLU 仅 52.8，知识推理偏弱。
  • 闭源专有模型，不支持自托管。
  • 上下文窗口 4K 偏小。

사용 사례

  • 通用对话与问答

참고문헌