Microsoft WizardLM 2 8x22B: Complete Benchmark Performance Guide

other wizardlm-2-8x22b 2024-04-15 Apache 2.0

In-depth analysis of Microsoft WizardLM 2 8x22B performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionwizardlm-2-8x22b
Release Date2024-04-15
Context Window65536 tokens
Input Modalitiestext
Output Modalitiestext
LicenseApache 2.0
Documentationhttps://huggingface.co/models

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU78.9%2024-04-155-shotview
HUMANEVAL77.9pass@12024-04-15—view
GSM8K83.7%2024-04-150-shot CoTview
MATH39.8%2024-04-150-shot CoTview
BBH73.6%2024-04-153-shot CoTview
GPQA39.3%2024-04-150-shotview
IFEVAL75.3%2024-04-15prompt_strictview
ARC92.1%2024-04-15challengeview
MUSR58%2024-04-150-shotview
WINOGRANDE81.6%2024-04-150-shotview

Pricing

TierPriceCurrency
Input$1.2 / MtokUSD
Output$1.2 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2024-04-15

Compliance

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

Microsoft WizardLM 2 8x22B

Aperçu du modèle

Microsoft WizardLM 2 8x22B 微调模型, 64K 上下文, 基于 Mixtral 8x22B, 在 Arena 上接近 GPT-4。

Spécifications principales

FournisseurVersionDate de sortieFenêtre de contexteModalités d’entréeModalités de sortieLicence
Otherwizardlm-2-8x22b2024-04-1565KtexttextApache 2.0

Performance aux benchmarks

BenchmarkScoreUnitéNotes
MMLU (Massive Multitask Language Understanding)78.9%5-shot
HumanEval77.9pass@1—
GSM8K (Grade School Math 8K)83.7%0-shot CoT
MATH39.8%0-shot CoT
BBH (BIG-Bench Hard)73.6%3-shot CoT
GPQA39.3%0-shot
IFEval75.3%prompt_strict
ARC92.1%challenge
MUSR58.0%0-shot
WinoGrande81.6%0-shot

Tarification

EntréeSortieLecture cacheÉcriture cache
————

par million de jetons

Forces

  • 采用 MoE 混合专家架构。

Faiblesses

  • 闭源专有模型，不支持自托管。

Cas d’usage

  • 代码生成与调试

Références