Chris Gervais
05/11/2023, 2:41 PMCristian (Nixtla)
05/11/2023, 2:50 PMwindows_batch_size
(less windows per batch)
⢠Reduce input_size
(shorter input window)
⢠Reduce hidden_size
(smaller model)val_size>0
), you can also set early_stop_patience_steps
larger than 0, to stop the training if the validation loss is not improving.Chris Gervais
05/12/2023, 5:36 PMKin Gtz. Olivares
05/13/2023, 5:48 PM