I follow the code from this link Basic regression: Predict fuel efficiency | TensorFlow Core
error show in normalizer.adapt(np.array(train_features))
#train feature is my data
only this column has string value and cannot convert while i am using stackoverflow and etc.
tf.keras.layers.Normalization performs feature wise normalization only on numerical features. You can use
tf.keras.layers.StringLookup which converts string categorical values an encoded representation that can be read by an
Embedding layer or
Dense layer. Thank you.
Thanks Chunduriv, I am very thankful for your answer but I solve this error last night. The issue is that some column are in other format string and int. So, first I can change into float64 all columns and then run, it was successfully solve.
And one more thing I forget that I changes all the data type float64 in dataframe but run the command in train_features so my friend check the code and tell me that you cannot change the data type in train_features, so I change the datatype in train_features and it was solved.