Knowledge library

Qlib separates forecasting signals from portfolio construction. A strategy turns prediction scores into trading decisions, while a weight-based base class lets users specify target holdings and delegates order generation to the framework. The documented…

Qlib is introduced as a modular platform for researching quantitative investment strategies with AI and machine learning. Its components are loosely coupled, so parts of the platform can be used independently. The architecture is organized into…

This QlibRL example describes how to configure and run a reinforcement learning workflow for executing orders in one asset. The training setup defines a simulator with 30-minute steps, a categorical action space, a full-history state representation, a…

This example outlines an end-to-end workflow for training and evaluating reinforcement learning agents for order execution. It covers preparing five-minute HS300 data and order files, configuring PPO and OPDS training tasks, saving checkpoints, and running a…

This overview introduces reinforcement learning as a way to learn sequential decisions by interacting with an environment and maximizing accumulated reward. It outlines the agent, environment, policy, and reward, and explains how delayed feedback differs…

This document presents a framework for testing trading decisions made at multiple time scales together. It argues that portfolio selection and intraday order execution should interact within one backtest because execution quality can change which…

This example shows a two-stage workflow for keeping predictions current with Qlib’s online model tools. First, it trains a model using a CSI 300 gradient-boosting task configuration and marks the resulting model as the online model. Second, it calls the…

This configuration specifies a Qlib research workflow for China A-share stocks in the CSI 300 universe. It uses Alpha360 features, robust feature normalization and missing-value filling, and cross-sectional rank normalization for labels. The prediction…

This configuration describes an order-execution backtest using five-minute market data and an order file. The main strategy uses a recurrent network with a PPO policy, a categorical action interpreter, and a state interpreter that supplies recent intraday…

The document introduces Qlib’s online-serving components for applying trained models to current market data. It describes a workflow that can produce predictions in live conditions and support real trading based on those predictions. The named components are…

This configuration specifies a reinforcement learning setup for order execution using Proximal Policy Optimization. It defines a categorical action interpreter, a recurrent network, and a full-history state representation built from intraday and prior-day…

QlibRL organizes a reinforcement learning workflow for trading around an environment wrapper that connects a policy to a market simulator. The wrapper accepts actions, advances the simulated market, and returns updated states and rewards. Its components…

This Qlib example demonstrates an end-to-end simulation of rolling online model workflows. It initializes Chinese-market data and experiment settings, generates rolling tasks at a configurable step, and connects those tasks to an online manager and a…

This notebook walks through a minimal Qlib reinforcement learning setup, connecting a simulator, state and action interpreters, a reward function, a policy, a dataset, and training and backtest workflows. Its simulator runs for a fixed number of steps,…

The guide lays out different entry paths for using QlibRL, depending on whether the reader is new to reinforcement learning, researches RL algorithms, or already has quantitative finance experience. It recommends learning RL fundamentals, understanding…

This Qlib documentation explains how a client can access market data managed on a central server. The client configuration points to a provider location, a local mount path, and a data service endpoint; NFS mounts the shared files, while a Flask service…

This configuration specifies a Qlib experiment for generating equity signals on the CSI 300 universe, using the Shanghai Shenzhen 300 index as its benchmark. It sets a historical data range and separates training, validation, and test periods. The data…

This configuration specifies a Chinese equities prediction and portfolio backtest using Qlib, LightGBM, and the Alpha158 feature handler. It pairs daily labels with one-minute features, resampling the minute data at 14:56. The listed data span begins in 2008…

This Qlib workflow example combines an LightGBM model trained on Alpha158 features with a top-ranked stock strategy for the CSI 300 universe. It sets training, validation, and test periods, then demonstrates nested execution across daily, 30-minute, and…

This example explains nested decision execution in Qlib backtesting: one strategy forms a portfolio at a slower frequency while another handles orders at a faster frequency. The first workflow generates portfolios weekly with DropoutTopkStrategy, described…