作者: nkovela1
创建时间 2022/09/19
上次修改时间 2022/09/26
描述:关于如何在多个 Keras 模型之间共享自定义训练步骤的指南。
此示例展示了如何使用“训练器模式”创建自定义训练步骤,然后可以在多个 Keras 模型之间共享。此模式覆盖了 keras.Model
类的 train_step()
方法,允许进行超出简单监督学习的训练循环。
训练器模式也可以轻松地适应具有更大自定义训练步骤的更复杂模型,例如 此端到端 GAN 模型,方法是将自定义训练步骤放在训练器类的定义中。
import os
os.environ["KERAS_BACKEND"] = "tensorflow"
import tensorflow as tf
import keras
# Load MNIST dataset and standardize the data
mnist = keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
可以通过覆盖 Model
子类的 train_step()
和 test_step()
方法来创建自定义训练和评估步骤。
class MyTrainer(keras.Model):
def __init__(self, model):
super().__init__()
self.model = model
# Create loss and metrics here.
self.loss_fn = keras.losses.SparseCategoricalCrossentropy()
self.accuracy_metric = keras.metrics.SparseCategoricalAccuracy()
@property
def metrics(self):
# List metrics here.
return [self.accuracy_metric]
def train_step(self, data):
x, y = data
with tf.GradientTape() as tape:
y_pred = self.model(x, training=True) # Forward pass
# Compute loss value
loss = self.loss_fn(y, y_pred)
# Compute gradients
trainable_vars = self.trainable_variables
gradients = tape.gradient(loss, trainable_vars)
# Update weights
self.optimizer.apply_gradients(zip(gradients, trainable_vars))
# Update metrics
for metric in self.metrics:
metric.update_state(y, y_pred)
# Return a dict mapping metric names to current value.
return {m.name: m.result() for m in self.metrics}
def test_step(self, data):
x, y = data
# Inference step
y_pred = self.model(x, training=False)
# Update metrics
for metric in self.metrics:
metric.update_state(y, y_pred)
return {m.name: m.result() for m in self.metrics}
def call(self, x):
# Equivalent to `call()` of the wrapped keras.Model
x = self.model(x)
return x
让我们定义两个不同的模型,它们可以共享我们的训练器类及其自定义 train_step()
。
# A model defined using Sequential API
model_a = keras.models.Sequential(
[
keras.layers.Flatten(input_shape=(28, 28)),
keras.layers.Dense(256, activation="relu"),
keras.layers.Dropout(0.2),
keras.layers.Dense(10, activation="softmax"),
]
)
# A model defined using Functional API
func_input = keras.Input(shape=(28, 28, 1))
x = keras.layers.Flatten(input_shape=(28, 28))(func_input)
x = keras.layers.Dense(512, activation="relu")(x)
x = keras.layers.Dropout(0.4)(x)
func_output = keras.layers.Dense(10, activation="softmax")(x)
model_b = keras.Model(func_input, func_output)
trainer_1 = MyTrainer(model_a)
trainer_2 = MyTrainer(model_b)
trainer_1.compile(optimizer=keras.optimizers.SGD())
trainer_1.fit(
x_train, y_train, epochs=5, batch_size=64, validation_data=(x_test, y_test)
)
trainer_2.compile(optimizer=keras.optimizers.Adam())
trainer_2.fit(
x_train, y_train, epochs=5, batch_size=64, validation_data=(x_test, y_test)
)
Epoch 1/5
...
Epoch 4/5
938/938 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - sparse_categorical_accuracy: 0.9770 - val_sparse_categorical_accuracy: 0.9770
Epoch 5/5
938/938 ━━━━━━━━━━━━━━━━━━━━ 1s 2ms/step - sparse_categorical_accuracy: 0.9805 - val_sparse_categorical_accuracy: 0.9789
<keras.src.callbacks.history.History at 0x7efe405fe560>