OpenFinLLM Leaderboard Documentation

Introduction

  • Overview
  • Model List
  • Demo

Financial Question Trees

  • Overview
  • BloombergGPT Datasets
    • FPB (Financial PhraseBank)
    • FiQA-SA (Sentiment Analysis)
    • Headlines
    • NER (Named Entity Recognition)
    • CONVFINQA
  • Information Extraction (IE)
  • Textual Analysis (TA)
  • Question Answering (QA)
  • Text Generation (TG)
  • Risk Management (RM)
  • Forecasting (FO)
  • Decision-Making (DM)
  • Other Datasets

Real-world Use Cases

  • Simple Questions
  • Efficient Legal Consultation Prep
  • Financial Document Analysis
  • AI-Driven Regulatory Compliance in Capital Markets
  • XBRL Analytics

Finagents

  • Tutor Agent
  • Search Agent
  • Trading Agent
  • XBRL Agent
  • ECC Analyzer

References

  • References

Tutorials

  • FPB Dataset Evaluation with O3-Mini
  • Benchmark DeepSeek on Financial Sentiment Analysis
  • Benchmark DeepSeek on Financial Questions with RAG
  • Benchmark DeepSeek on Financial Sentiment Analysis (few-shot)
  • Benchmark Llama-3.2 on Financial Sentiment Analysis (zeroshot)
  • Benchmark GPT-4o on Financial Sentiment Analysis (zeroshot)
  • Evaluate Deepseek using Framework

Experiment Settings

  • Install Cuda
  • Obtaining DeepSeek API Key
  • Get Model API keys

Industry Collaboration

  • FAQs
  • Contributions
OpenFinLLM Leaderboard Documentation
  • BloombergGPT Datasets
  • Edit on GitHub

BloombergGPT Datasets

This section contains datasets specifically related to BloombergGPT and its evaluation tasks.

  • FPB (Financial PhraseBank)
    • Goal
    • Description
    • Task Description
    • Evaluation Metrics
    • Why Use FPB?
    • References
  • FiQA-SA (Sentiment Analysis)
    • Description
    • Task Description
    • Evaluation Metrics
    • References
  • Headlines
    • Description
  • NER (Named Entity Recognition)
    • Description
    • Task Description
    • Evaluation Metrics
    • References
  • CONVFINQA
    • Description
Previous Next

© Copyright 2025, OpenFinLLM Leaderboard.

Built with Sphinx using a theme provided by Read the Docs.