Hi, I've stumbled across the paper where HealthGAN is used for generating time-series data for the MIMIC-III Time Series Benchmark dataset. I'm trying to implement HealthGAN using PyTorch for said data and would like to inquire the dimensions regarding the input and hidden, i.e. are they (Batch, Seq_Len, Dim) or (Batch, ?). Another question is the hyperparameters for reproducing the results for the model, e.g. number of layers, hidden dimension etc. Thanks.
Hi, I've stumbled across the paper where HealthGAN is used for generating time-series data for the MIMIC-III Time Series Benchmark dataset. I'm trying to implement HealthGAN using PyTorch for said data and would like to inquire the dimensions regarding the input and hidden, i.e. are they (Batch, Seq_Len, Dim) or (Batch, ?). Another question is the hyperparameters for reproducing the results for the model, e.g. number of layers, hidden dimension etc. Thanks.