-
Notifications
You must be signed in to change notification settings - Fork 5
Open
Description
Thank you so much for sharing this nice repo on Diffusion model for time series data.
When running TimeAutoDiff-ddpm.ipynb, I observe that there is an extra positional argument (time_info.to(device)) in
ds = tae.train_autoencoder(real_df1, processed_data, time_info.to(device), channels, hidden_size, num_layers, lr, weight_decay, n_epochs, \ batch_size, threshold, min_beta, max_beta, emb_dim, time_dim, lat_dim, device)
So I am wondering if this will affect the whole Jupyter notebook or I can just delete the element?
Metadata
Metadata
Assignees
Labels
No labels