Thanks for the great work! I am very interested in this project and would like to use it for my research. Is there a timeline for open-sourcing the training and inference code?