model.get_normalized_expression issue of MultiVI

[ZennFangZhou]

Hello, developer! I'm conducting MultiVI analysis on my CITE-seq data with 30,000 cells and 6,000 genes. When I ran the model.get_normalized_expression function on the trained model, system memory usage quickly surged from 10GB to over 100GB, causing a software crash. This didn't happen when I ran the tutorial data. How can I resolve this?

[ZennFangZhou]

all right, when i set the para to rna_denoised, protein_denoised =model.get_normalized_expression(mdata,n_samples=25, return_mean=True,gene_list=rna.var_names.to_list()),i deal with the problem