This release introduces a generator called ForestDiffusion from the paper "Generating and Imputing Tabular Data via Diffusion and Flow-based XGBoost Models".
Installation: pip install tabgan
Data generation: ForestDiffusionGenerator().generate_data_pipe(train, target, test, )