Qlib provides a modular machine learning framework for quantitative investment research, spanning dataset construction, model training, backtesting, and evaluation. It is designed for researchers and engineers building and testing models on historical market data.
Project overview
The framework offers qrun, an automated tool that executes a complete quantitative research workflow, alongside support for reinforcement learning to model continuous trading decisions.
Project type
Workflow · Model Runtime · Data Processing
Use cases
Data Analysis
Deployment
Refer to project documentation
License
MIT
Best for
Researchers, data teams, and AI engineers who need a modular framework to build complete quantitative research workflows and backtest models using historical market data.
Key capabilities
Provides a tool named qrun to run the whole workflow automatically, including building datasets, training models, running backtests, and performing evaluation.
Supports reinforcement learning to model continuous investment decisions and optimize trading strategies by learning from interactions with the environment.
Limitations and risks
The run_all_model.py script only supports Linux currently.
The TFT model only supports Python 3.6~3.7 due to a tensorflow==1.15.0 limitation.
Pandas groupby break changes can cause errors in specific RL and benchmark scripts.
Getting started
Setup requires managing Python environments, installing via pip, downloading data separately via scripts, initializing Qlib with the data path, and executing qrun on a config file.
Alternatives and comparisons
A multi-agent LLM framework that deploys specialized agents to collaboratively evaluate market conditions and inform trading decisions.
GitHub project description: Qlib is an AI-oriented Quant investment platform that aims to use AI tech to empower Quant Research, from exploring ideas to implementing productions.
README: Qlib provides a strong infrastructure to support Quant research.
README: Qlib provides a tool named `qrun` to run the whole workflow automatically (including building dataset, training models, backtest and evaluation).
README: Qlib also provides a modularized interface to allow researchers to build their own workflow by code.