Phi-3 Mini: Complete Benchmark Performance Guide

other phi-3-mini 2024-04-22 MIT

In-depth analysis of Phi-3 Mini performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionphi-3-mini
Release Date2024-04-22
Context Window4096 tokens
Input Modalitiestext
Output Modalitiestext
LicenseMIT
Documentationhttps://huggingface.co/models

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU43.6%2024-04-225-shotview
HUMANEVAL26.9pass@12024-04-22—view
GSM8K40.9%2024-04-220-shot CoTview
MATH17.9%2024-04-220-shot CoTview
BBH55.9%2024-04-223-shot CoTview
GPQA20.7%2024-04-220-shotview
IFEVAL54.9%2024-04-22prompt_strictview
ARC84.4%2024-04-22challengeview
MUSR33.5%2024-04-220-shotview
WINOGRANDE67.1%2024-04-220-shotview

Pricing

TierPriceCurrency
Input$0.1 / MtokUSD
Output$0.1 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2024-04-22

Compliance

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

Phi-3 Mini

Model Overview

Microsoft Phi-3 Mini 3.8B 轻量模型, 4K 上下文 (可扩展 128K), 训练数据高质量, 性能超越 Llama 2 7B。

Core Specifications

VendorVersionRelease DateContext WindowInput ModalitiesOutput ModalitiesLicense
Otherphi-3-mini2024-04-224KtexttextMIT

Benchmark Performance

BenchmarkScoreUnitNotes
MMLU (Massive Multitask Language Understanding)43.6%5-shot
HumanEval26.9pass@1—
GSM8K (Grade School Math 8K)40.9%0-shot CoT
MATH17.9%0-shot CoT
BBH (BIG-Bench Hard)55.9%3-shot CoT
GPQA20.7%0-shot
IFEval54.9%prompt_strict
ARC84.4%challenge
MUSR33.5%0-shot
WinoGrande67.1%0-shot

Pricing

InputOutputCache ReadCache Write
————

per million tokens

Strengths

  • Reliable general-purpose model.

Weaknesses

  • MMLU 43.6, weak knowledge reasoning.
  • HumanEval 26.9, coding weak.
  • Proprietary, not self-hostable.
  • Context window 4K is limited.

Use Cases

  • General chat and Q&A

References