cg_300_integrated_training_integrin_dmap_decoder_main.ipynb: Notebook to train conditional Wasserstein generative adversarial network with gradient penality to learn the inverse mapping between the unified latent space of all four metastable states and 300 bead CG configurations.
wgangp_b_100_e1000.ckpt: Trained model. Notebooks illustrating its usage in generating intermediate structures are available in inference.