FinGPT - A large-scale open-source model in the financial field that can predict stock price movements.

What is FinGPT?

FinGPT is an open-source, large-scale pre-trained language model for the financial field, driving financial innovation based on natural language processing technology. Through reinforcement learning and human feedback (RLHF) techniques, it can learn individual preferences and provide personalized investment advice. FinGPT supports various financial tasks, such as sentiment analysis, relation extraction, title classification, and named entity recognition, with a wide range of data sources, including financial news websites, social media, and financial regulatory agency websites.

Main functions of FinGPT

• Financial Sentiment AnalysisUsing natural language processing technology, we can perform sentiment analysis on financial texts (such as financial news and social media comments) to determine whether they are positive, negative, or neutral.
• Financial relationship extractionExtract relationships between financial entities from the text, such as partnerships and mergers between companies.
• Financial Title ClassificationClassify financial news headlines and determine their respective financial themes (such as stock market, monetary policy, industry dynamics, etc.).
• Financial Named Entity RecognitionIt identifies financial entities in text, such as company names, stock codes, and names of financial products.
• Market forecastIt combines historical data and real-time information to predict market trends and stock price movements.
• Personalized investment adviceBy learning users' investment preferences and risk tolerance, we provide personalized investment advice.
• Data-driven model trainingIt supports low-rank adaptation (LoRA) and reinforcement learning techniques, enabling rapid adaptation to new data and reducing training costs.
• Multilingual supportIt supports financial text processing in multiple languages, covering global financial markets.

FinGPT's technical principles

• end-to-end frameworkFinGPT employs a complete end-to-end framework, comprising four layers:
  • Data source layerWe acquire real-time financial data from multiple channels to ensure comprehensive market coverage.
  • Data Engineering LayerIt cleans, preprocesses, and extracts features from real-time data to address the issues of high time sensitivity and low signal-to-noise ratio in financial data.
  • LLMs layersBased on pre-trained LLMs, fine-tuning is performed through low-rank adaptation (LoRA) and stock-price-based reinforcement learning (RLSP) to adapt to the dynamic changes in the financial field.
  • Application layerIt offers a variety of financial applications, such as sentiment analysis, market forecasting, and investment advice.
• Pre-trained models and fine-tuningFinGPT is based on a pre-trained Transformer architecture and learns rich financial knowledge and language patterns by training on large-scale financial text data. Fine-tuning techniques include:
  • Low-rank adaptation (LoRA)By reducing the number of trainable parameters, training costs can be significantly reduced. For example, the number of trainable parameters can be reduced from 6.17 billion to 3.67 million while maintaining model performance.
  • Stock Price-Based Reinforcement Learning (RLSP)This approach uses stock price changes as feedback signals, replacing traditional reinforcement learning based on human feedback. This allows the model to learn in real-time how the market reacts to news and events, improving predictive accuracy.
• Real-time data processingFinGPT's data engineering layer supports real-time data processing, including:
  • Data cleaning: Remove irrelevant data, handle missing values, and normalize text.
  • Feature extractionUse techniques such as TF-IDF, bag-of-words model or Word2Vec to transform raw data into input that machine learning models can understand.
  • Sentiment AnalysisThe cleaned data is classified into sentiment categories to determine whether it is positive, negative, or neutral.

FinGPT's project address

• Github repository:https://github.com/AI4Finance-Foundation/FinGPT
• arXiv technical paper:https://arxiv.org/pdf/2306.06031

Application scenarios of FinGPT

• Investment AnalysisFinGPT can help analyze the sentiment and fundamental information of investment targets such as stocks and funds, providing investors with a more comprehensive basis for decision-making.
• Market ResearchBy rapidly processing unstructured data such as news and social media, FinGPT can identify market sentiment and help investors and financial institutions grasp market dynamics.
• Automated Quantitative TradingFinGPT's predictive capabilities can be combined with quantitative trading strategies to achieve AI-based trading strategy optimization, thereby improving trading efficiency and returns.
• Intelligent predictionBased on the FinGPT-Forecaster module, users can obtain predictions of stock price trends, and the functionality can be easily accessed via HuggingFace Spaces.
• Sentiment AnalysisFinGPT can perform accurate sentiment analysis on information from financial news and social media to assist in decision-making.