TensorFlow报错：Normalization层维度不匹配(ValueError)求助

问题根源 #

代码存在一处关键逻辑错误：

• 你先通过train_features.pop('HCO3')移除了训练集的标签列，此时train_features是7维特征数据
• 但随后执行train_features = np.asarray(train_dataset.copy()).astype('float32')，这里的train_dataset是原始数据集（包含'HCO3'标签列），导致训练用的train_features变成8维
• 测试用的test_features是执行test_features.pop('HCO3')后的7维数据，Normalization层适配了8维的训练数据，测试时输入7维数据自然触发维度不匹配错误

修正后的代码 #

#Split labels
train_features = train_dataset.copy()
test_features = test_dataset.copy()

train_labels = train_features.pop('HCO3')
test_labels = test_features.pop('HCO3')

# 修正：使用已经移除标签的train_features转换数组，而非原始数据集
train_features = np.asarray(train_features).astype('float32')
# 同步修正测试集特征的格式，确保和训练集一致
test_features = np.asarray(test_features).astype('float32')

#Normalization
normalizer = tf.keras.layers.Normalization(axis=-1)
normalizer.adapt(np.array(train_features))
first = np.array(train_features[:1])

linear_model = tf.keras.Sequential([
    normalizer,
    layers.Dense(units=1)
])
#Compilation
linear_model.compile(
    optimizer=tf.keras.optimizers.Adam(learning_rate=0.1),
    loss='mean_absolute_error'
)
history = linear_model.fit(
    train_features,
    train_labels,
    epochs=100,
    # Suppress logging.
    verbose=0,
    # Calculate validation results on 20% of the training data.
    validation_split = 0.2)

#Track error for later
test_results = {}
test_results['linear_model'] = linear_model.evaluate(test_features, test_labels, verbose = 0)

关键修正点 #

• 将train_features = np.asarray(train_dataset.copy()).astype('float32')改为train_features = np.asarray(train_features).astype('float32')，确保训练特征是移除标签后的7维数据
• 新增test_features = np.asarray(test_features).astype('float32')，让测试特征的格式和训练特征完全对齐，避免潜在的类型或维度问题

内容的提问来源于stack exchange，提问作者Rena W