Key Specifications

Vendorother
Versionfalcon-180b
Release Date2023-09-06
Context Window2048 tokens
Input Modalitiestext
Output Modalitiestext
LicenseTII Falcon LLM License
Documentationhttps://huggingface.co/models

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 76.7 % 2023-09-06 5-shot view
HUMANEVAL 68.9 pass@1 2023-09-06 view
GSM8K 77.5 % 2023-09-06 0-shot CoT view
MATH 44.8 % 2023-09-06 0-shot CoT view
BBH 76.4 % 2023-09-06 3-shot CoT view
GPQA 43.2 % 2023-09-06 0-shot view
IFEVAL 70.5 % 2023-09-06 prompt_strict view
ARC 91.6 % 2023-09-06 challenge view
MUSR 59.1 % 2023-09-06 0-shot view
WINOGRANDE 80 % 2023-09-06 0-shot view

Pricing

Tier Price Currency
Input$1 / MtokUSD
Output$1 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://huggingface.co/models · as of 2023-09-06

Compliance

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

Falcon 180B

Model Overview

TII Falcon 180B 大型开源模型, 2K 上下文, 1800 亿参数, 在开源模型中曾排名第一, 排在 GPT-4 之后。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Other falcon-180b 2023-09-06 2K text text TII Falcon LLM License

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 76.7 % 5-shot
HumanEval 68.9 pass@1
GSM8K (Grade School Math 8K) 77.5 % 0-shot CoT
MATH 44.8 % 0-shot CoT
BBH (BIG-Bench Hard) 76.4 % 3-shot CoT
GPQA 43.2 % 0-shot
IFEval 70.5 % prompt_strict
ARC 91.6 % challenge
MUSR 59.1 % 0-shot
WinoGrande 80.0 % 0-shot

Pricing

Input Output Cache Read Cache Write

per million tokens

Strengths

  • Reliable general-purpose model.

Weaknesses

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

Use Cases

  • General chat and Q&A

References