Orca 2 13B: Complete Benchmark Performance Guide

other orca-2-13b 2023-11-20 MIT

In-depth analysis of Orca 2 13B performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionorca-2-13b
Release Date2023-11-20
Context Window4096 tokens
Input Modalitiestext
Output Modalitiestext
LicenseMIT
Documentationhttps://huggingface.co/models

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 78.4 % 2023-11-20 5-shot view
HUMANEVAL 74.3 pass@1 2023-11-20 — view
GSM8K 84.1 % 2023-11-20 0-shot CoT view
MATH 42.3 % 2023-11-20 0-shot CoT view
BBH 70.8 % 2023-11-20 3-shot CoT view
GPQA 42.5 % 2023-11-20 0-shot view
IFEVAL 74.2 % 2023-11-20 prompt_strict view
ARC 91.4 % 2023-11-20 challenge view
MUSR 59.8 % 2023-11-20 0-shot view
WINOGRANDE 80.2 % 2023-11-20 0-shot view

Pricing

Tier Price Currency
Input$0.3 / MtokUSD
Output$0.3 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2023-11-20

Compliance

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

Orca 2 13B

Model Overview

Microsoft Orca 2 13B 推理增强模型, 4K 上下文, 基于 Llama 2 微调, 通过渐进式复杂任务训练提升推理。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Other orca-2-13b 2023-11-20 4K text text MIT

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 78.4 % 5-shot
HumanEval 74.3 pass@1 —
GSM8K (Grade School Math 8K) 84.1 % 0-shot CoT
MATH 42.3 % 0-shot CoT
BBH (BIG-Bench Hard) 70.8 % 3-shot CoT
GPQA 42.5 % 0-shot
IFEval 74.2 % prompt_strict
ARC 91.4 % challenge
MUSR 59.8 % 0-shot
WinoGrande 80.2 % 0-shot

Pricing

Input Output Cache Read Cache Write
— — — —

per million tokens

Strengths

  • Reliable general-purpose model.

Weaknesses

  • Proprietary, not self-hostable.
  • Context window 4K is limited.

Use Cases

  • Code generation and debugging

References