在keras.sequential中获得很多警告和错误

如何解决在keras.sequential中获得很多警告和错误

我曾经使用以下功能，但是现在运行它时，升级到Tensorflow 2.3后，我收到许多警告和dtypes令人困惑的通知。我不确定如何修复功能以正确运行keras顺序代码以生成正确的预测。

这是我的实现方式

feature_columns = []

for col in numericColumns:
  col = feature_column.numeric_column(col)
  feature_columns.append(col)

for col in categoricalColumns:
  col = feature_column.indicator_column(feature_column.categorical_column_with_vocabulary_list(col,df_train[col].unique()))
  feature_columns.append(col)

feature_layer = tf.keras.layers.DenseFeatures(feature_columns)


from tensorflow.keras import optimizers

def build_model():
  
  model = keras.Sequential([
    feature_layer,layers.Dense(10,activation='relu'),layers.Dense(1)
  ])

  optimizer = tf.optimizers.RMSprop(learning_rate=lr_schedule)
  model.compile(loss='mse',optimizer=optimizer,metrics=['mse'])
  return model

model = build_model()

这是我调用该函数的方式：

# train the model
history = model.fit(train_ds,validation_data=val_ds,epochs=30)

这是我得到的杂乱的回报，我正在寻找帮助以弄清楚如何修复或重写：

Epoch 1/30
WARNING:tensorflow:Layers in a Sequential model should only have a single input tensor,but we receive a <class 'dict'> input: {'MSSubClass': <tf.Tensor 'ExpandDims_49:0' shape=(None,1) dtype=float64>,'MSZoning': <tf.Tensor 'ExpandDims_50:0' shape=(None,1) dtype=int32>,'LotFrontage': <tf.Tensor 'ExpandDims_46:0' shape=(None,'LotArea': <tf.Tensor 'ExpandDims_44:0' shape=(None,'Street': <tf.Tensor 'ExpandDims_66:0' shape=(None,'LotShape': <tf.Tensor 'ExpandDims_47:0' shape=(None,'LandContour': <tf.Tensor 'ExpandDims_42:0' shape=(None,'Utilities': <tf.Tensor 'ExpandDims_69:0' shape=(None,'LotConfig': <tf.Tensor 'ExpandDims_45:0' shape=(None,'LandSlope': <tf.Tensor 'ExpandDims_43:0' shape=(None,'Neighborhood': <tf.Tensor 'ExpandDims_55:0' shape=(None,'Condition1': <tf.Tensor 'ExpandDims_16:0' shape=(None,'Condition2': <tf.Tensor 'ExpandDims_17:0' shape=(None,'BldgType': <tf.Tensor 'ExpandDims_4:0' shape=(None,'HouseStyle': <tf.Tensor 'ExpandDims_39:0' shape=(None,'OverallQual': <tf.Tensor 'ExpandDims_58:0' shape=(None,'OverallCond': <tf.Tensor 'ExpandDims_57:0' shape=(None,'YearBuilt': <tf.Tensor 'ExpandDims_71:0' shape=(None,'YearRemodAdd': <tf.Tensor 'ExpandDims_72:0' shape=(None,'RoofStyle': <tf.Tensor 'ExpandDims_62:0' shape=(None,'RoofMatl': <tf.Tensor 'ExpandDims_61:0' shape=(None,'Exterior1st': <tf.Tensor 'ExpandDims_22:0' shape=(None,'Exterior2nd': <tf.Tensor 'ExpandDims_23:0' shape=(None,'MasVnrType': <tf.Tensor 'ExpandDims_52:0' shape=(None,'MasVnrArea': <tf.Tensor 'ExpandDims_51:0' shape=(None,'ExterQual': <tf.Tensor 'ExpandDims_21:0' shape=(None,'ExterCond': <tf.Tensor 'ExpandDims_20:0' shape=(None,'Foundation': <tf.Tensor 'ExpandDims_25:0' shape=(None,'BsmtQual': <tf.Tensor 'ExpandDims_13:0' shape=(None,'BsmtCond': <tf.Tensor 'ExpandDims_5:0' shape=(None,1) dtype=int64>,'BsmtExposure': <tf.Tensor 'ExpandDims_6:0' shape=(None,'BsmtFinType1': <tf.Tensor 'ExpandDims_9:0' shape=(None,'BsmtFinSF1': <tf.Tensor 'ExpandDims_7:0' shape=(None,'BsmtFinType2': <tf.Tensor 'ExpandDims_10:0' shape=(None,'BsmtFinSF2': <tf.Tensor 'ExpandDims_8:0' shape=(None,'BsmtUnfSF': <tf.Tensor 'ExpandDims_14:0' shape=(None,'TotalBsmtSF': <tf.Tensor 'ExpandDims_68:0' shape=(None,'Heating': <tf.Tensor 'ExpandDims_37:0' shape=(None,'HeatingQC': <tf.Tensor 'ExpandDims_38:0' shape=(None,'CentralAir': <tf.Tensor 'ExpandDims_15:0' shape=(None,'Electrical': <tf.Tensor 'ExpandDims_18:0' shape=(None,'1stFlrSF': <tf.Tensor 'ExpandDims:0' shape=(None,'2ndFlrSF': <tf.Tensor 'ExpandDims_1:0' shape=(None,'LowQualFinSF': <tf.Tensor 'ExpandDims_48:0' shape=(None,'GrLivArea': <tf.Tensor 'ExpandDims_35:0' shape=(None,'BsmtFullBath': <tf.Tensor 'ExpandDims_11:0' shape=(None,'BsmtHalfBath': <tf.Tensor 'ExpandDims_12:0' shape=(None,'FullBath': <tf.Tensor 'ExpandDims_26:0' shape=(None,'HalfBath': <tf.Tensor 'ExpandDims_36:0' shape=(None,'BedroomAbvGr': <tf.Tensor 'ExpandDims_3:0' shape=(None,'KitchenAbvGr': <tf.Tensor 'ExpandDims_40:0' shape=(None,'KitchenQual': <tf.Tensor 'ExpandDims_41:0' shape=(None,'TotRmsAbvGrd': <tf.Tensor 'ExpandDims_67:0' shape=(None,'Functional': <tf.Tensor 'ExpandDims_27:0' shape=(None,'Fireplaces': <tf.Tensor 'ExpandDims_24:0' shape=(None,'GarageType': <tf.Tensor 'ExpandDims_33:0' shape=(None,'GarageYrBlt': <tf.Tensor 'ExpandDims_34:0' shape=(None,'GarageFinish': <tf.Tensor 'ExpandDims_31:0' shape=(None,'GarageCars': <tf.Tensor 'ExpandDims_29:0' shape=(None,'GarageArea': <tf.Tensor 'ExpandDims_28:0' shape=(None,'GarageQual': <tf.Tensor 'ExpandDims_32:0' shape=(None,'GarageCond': <tf.Tensor 'ExpandDims_30:0' shape=(None,'PavedDrive': <tf.Tensor 'ExpandDims_59:0' shape=(None,'WoodDeckSF': <tf.Tensor 'ExpandDims_70:0' shape=(None,'OpenPorchSF': <tf.Tensor 'ExpandDims_56:0' shape=(None,'EnclosedPorch': <tf.Tensor 'ExpandDims_19:0' shape=(None,'3SsnPorch': <tf.Tensor 'ExpandDims_2:0' shape=(None,'ScreenPorch': <tf.Tensor 'ExpandDims_65:0' shape=(None,'PoolArea': <tf.Tensor 'ExpandDims_60:0' shape=(None,'MiscVal': <tf.Tensor 'ExpandDims_53:0' shape=(None,'MoSold': <tf.Tensor 'ExpandDims_54:0' shape=(None,'YrSold': <tf.Tensor 'ExpandDims_73:0' shape=(None,'SaleType': <tf.Tensor 'ExpandDims_64:0' shape=(None,'SaleCondition': <tf.Tensor 'ExpandDims_63:0' shape=(None,1) dtype=int32>}
Consider rewriting this model with the Functional API.
WARNING:tensorflow:Layers in a Sequential model should only have a single input tensor,1) dtype=int32>}
Consider rewriting this model with the Functional API.
35/37 [===========================>..] - ETA: 0s - loss: 120.7742 - mse: 120.7742WARNING:tensorflow:Layers in a Sequential model should only have a single input tensor,1) dtype=int32>}
Consider rewriting this model with the Functional API.