“It,” that's an implementation which doesn't bind to Python Torch, ... train_dl <- dataloader(train_ds, batch_size = 32, shuffle = TRUE) ...
確定! 回上一頁