Phi-1: Complete Benchmark Performance Guide

other phi-1 2023-09-14 MIT

In-depth analysis of Phi-1 performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionphi-1
Release Date2023-09-14
Context Window2048 tokens
Input Modalitiestext
Output Modalitiestext
LicenseMIT
Documentationhttps://huggingface.co/models

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 68.1 % 2023-09-14 5-shot view
HUMANEVAL 88.1 pass@1 2023-09-14 — view
GSM8K 51.6 % 2023-09-14 0-shot CoT view
MATH 25.2 % 2023-09-14 0-shot CoT view
BBH 69.9 % 2023-09-14 3-shot CoT view
GPQA 22.1 % 2023-09-14 0-shot view
IFEVAL 53.8 % 2023-09-14 prompt_strict view
ARC 88.4 % 2023-09-14 challenge view
MUSR 49 % 2023-09-14 0-shot view
WINOGRANDE 74.5 % 2023-09-14 0-shot view

Pricing

Tier Price Currency
Input$0.1 / MtokUSD
Output$0.1 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2023-09-14

Compliance

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

Phi-1

Model Overview

Microsoft Phi-1 1.3B 代码模型, 2K 上下文, 专为代码生成训练, 在 1.3B 规模上 HumanEval 表现突出。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Other phi-1 2023-09-14 2K text text MIT

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 68.1 % 5-shot
HumanEval 88.1 pass@1 —
GSM8K (Grade School Math 8K) 51.6 % 0-shot CoT
MATH 25.2 % 0-shot CoT
BBH (BIG-Bench Hard) 69.9 % 3-shot CoT
GPQA 22.1 % 0-shot
IFEval 53.8 % prompt_strict
ARC 88.4 % challenge
MUSR 49.0 % 0-shot
WinoGrande 74.5 % 0-shot

Pricing

Input Output Cache Read Cache Write
— — — —

per million tokens

Strengths

  • HumanEval 88.1, excellent code generation.

Weaknesses

  • Proprietary, not self-hostable.
  • Context window 2K is limited.

Use Cases

  • Code generation and debugging

References