... see what format it is in: cifar10_train.take(1) The output tells us that each image in the dataset is of format <DatasetV1Adapter shapes: {image: (32, ...
確定! 回上一頁