search

jsxlei / SCALEX

Online single-cell data integration through projecting heterogeneous datasets into a common cell-embedding space
BSD 3-Clause "New" or "Revised" License
71 stars 18 forks source link

model.train() setting #18

Closed GlancerZ closed 1 year ago

GlancerZ commented 1 year ago

Why do you use model.train in the inference part (encode batch)?

jsxlei commented 1 year ago

Thank you for your interest. Here, model.train means using data-specific mean and var in the batch_norm, otherwise it will use the mean and var from the training data. Anyway, the model.train or model.eval does not make big difference.