OpenChat 3.5 1210: Complete Benchmark Performance Guide

other openchat-3-5-1210 2023-12-10 Apache 2.0

In-depth analysis of OpenChat 3.5 1210 performance on MMLU, HumanEval, GSM8K, MATH, BBH benchmarks, including pricing and use cases.

Key Specifications

Vendorother
Versionopenchat-3-5-1210
Release Date2023-12-10
Context Window8192 tokens
Input Modalitiestext
Output Modalitiestext
LicenseApache 2.0
Documentationhttps://huggingface.co/models

Benchmark Performance

BenchmarkScoreUnitEvaluated AtNotesSource
MMLU70.5%2023-12-105-shotview
HUMANEVAL41.9pass@12023-12-10—view
GSM8K55%2023-12-100-shot CoTview
MATH15.2%2023-12-100-shot CoTview
BBH60.4%2023-12-103-shot CoTview
GPQA20.9%2023-12-100-shotview
IFEVAL50.4%2023-12-10prompt_strictview
ARC88.1%2023-12-10challengeview
MUSR32.6%2023-12-100-shotview
WINOGRANDE69.3%2023-12-100-shotview

Pricing

TierPriceCurrency
Input$0.18 / MtokUSD
Output$0.18 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2023-12-10

Compliance

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

OpenChat 3.5 1210

Modellöversikt

OpenChat 3.5 1210 对话模型, 8K 上下文, 基于 Mistral 7B C-RLFT 微调, 性能超越 ChatGPT 3.5。

Kärnspecifikationer

LeverantörVersionReleasedatumKontextfönsterInmatningsmodaliteterUtmatningsmodaliteterLicens
Otheropenchat-3-5-12102023-12-108KtexttextApache 2.0

Benchmarkprestanda

BenchmarkPoängEnhetAnteckningar
MMLU (Massive Multitask Language Understanding)70.5%5-shot
HumanEval41.9pass@1—
GSM8K (Grade School Math 8K)55.0%0-shot CoT
MATH15.2%0-shot CoT
BBH (BIG-Bench Hard)60.4%3-shot CoT
GPQA20.9%0-shot
IFEval50.4%prompt_strict
ARC88.1%challenge
MUSR32.6%0-shot
WinoGrande69.3%0-shot

Priser

InmatningUtmatningCacheläsningCacheskrivning
————

per miljon tokens

Styrkor

  • 可靠的通用模型。

Svagheter

  • HumanEval 41.9，代码能力较弱。
  • 闭源专有模型，不支持自托管。
  • 上下文窗口 8K 偏小。

Användningsområden

  • 通用对话与问答

Referenser