Xây dựng Sequential Model (Dense Layer) với TensorFlow/Keras Python

Mô tả

Tạo model Sequential bằng cách thêm các Dense layer rõ ràng, sau đó gọi model trên dữ liệu test.

Cách 1: Tạo layer riêng lẻ rồi kết hợp (Functional style)

import tensorflow as tf
from tensorflow.keras import layers

# Tạo 3 Dense layer riêng biệt
layer1 = layers.Dense(2, activation="relu", name="layer_1")
layer2 = layers.Dense(3, activation="relu", name="layer_2")
layer3 = layers.Dense(4, name="layer_3")

print("Three dense layers created")

# Gọi tuần tự (functional call)
x = tf.ones((4, 4))  # Batch 4, input dim 4
y = layer3(layer2(layer1(x)))

print("Model called on test data")
print(f"Output shape: {y.shape}")  # (4, 4)

# Xem layers
print("Layers:", [layer.name for layer in [layer1, layer2, layer3]])

Cách 2: Sequential constructor với list layers (Khuyên dùng)

import tensorflow as tf
from tensorflow.keras import layers, Sequential

model = Sequential([
    layers.Dense(2, activation="relu", name="layer_1"),
    layers.Dense(3, activation="relu", name="layer_2"),
    layers.Dense(4, name="layer_3"),
], name="my_sequential_model")

print("Model summary:")
model.summary()

# Gọi trên test data
x = tf.ones((2, 4))
y = model(x)
print(f"\nOutput shape: {y.shape}")  # (2, 4)

Đầu ra model.summary()

Model: "my_sequential_model"
_________________________________________________________________
 Layer (type)                Output Shape              Param #
=================================================================
 layer_1 (Dense)             (None, 2)                 10
 layer_2 (Dense)             (None, 3)                 9
 layer_3 (Dense)             (None, 4)                 16
=================================================================
Total params: 35
Trainable params: 35
Non-trainable params: 0
_________________________________________________________________

Cách 3: Add layer lần lượt (Incremental)

model = Sequential(name="incremental_model")
model.add(layers.Dense(2, activation="relu", input_shape=(4,)))
model.add(layers.Dense(3, activation="relu"))
model.add(layers.Dense(4))

print("Dense layers added")
model.summary()

So sánh 3 cách

Compile & Train

model.compile(optimizer='adam', loss='mse', metrics=['mae'])
history = model.fit(X_train, y_train, epochs=10, validation_split=0.2)