This repository contains the code to instantiate, train, and evaluate the deep neural network architecture described in DeepLOB model described in the paper DeepLOB: Deep Convolutional Neural Networks for Limit Order Books by Z. Zhang et al.

The complete pipeline to download the data, implement the dataset I/O, instantiate and train the model is contained in this notebook.

This implementation is based on Python 3.8. To install all the requirements:

$ pip install -r requirements.txt

To train the model from scratch, just run the entire notebook. The model automatically logs some metrics (e.g., loss, accuracy, etc.) using WandB, so make sure to log in before you start the training process.

However, the model has already been trained and you can find here the WandB logs for it.