Tensorflow dataset space describes the use of the API, dataset API mainly for reading data. This course on house prices predicted by the issue of combat in detail how to use tf.dataset read csv file and tfrecord file.
In tensorflow tf.data for processing data it is mainly used to read the data, and to do some preprocessing in some data then the training procedure to superimpose the data train.
Actual content:
Dataset base API.
Dataset read csv file.
Dataset read tfrecord file.
API list:
Dataset基础使用(tf.data.Dataset.from_tensor_slices; repeat,batch,interleave,map,shuffle,list_files);
csv(tf.data.TextLineDataset,tf.io.decode_csv);
Tfrecord(tf.train.FloatLlist,tf.train.Int64,tf.trian.BytesList; tf.train.Feature,tf.train.Features,tf.train.Example; example.SerializeToString; tf.io.ParseSingExample; tf.io.VarLenFeature,tf.io.FixedLenFeature; tf.data.TFRecordDataset,tf.io.TFRecordOptions)
tf_data basis using the API
Generate csv file