tensorflow-ValueError logits and labels must have the same shape, received ((None, 240, 1) vs (None,)).

如果输入的 shape 都没问题，那应该是没加 Flatten 的原因

model = tf.keras.models.Sequential([
    # input_shape 保持和数据集一致
    tf.keras.layers.Dense(32,input_shape=(240, 4), activation='relu'),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(1, activation='sigmoid'),
])