无
Trainer 模式
作者： nkovela1
创建日期 2022/09/19
最后修改日期 2022/09/26
描述： 关于如何跨多个 Keras 模型共享自定义训练步骤的指南。
简介
此示例演示了如何使用“Trainer 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。此模式会覆盖 keras.Model 类的 train_step() 方法，从而实现超越纯监督学习的训练循环。
通过将自定义训练步骤放在 Trainer 类定义中，Trainer 模式还可以轻松适应更复杂的模型，其中包含更大的自定义训练步骤，例如 这个端到端的 GAN 模型。
设置
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
定义 Trainer 类
通过覆盖 Model 子类的 train_step() 和 test_step() 方法，可以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
定义多个模型以共享自定义训练步骤
我们来定义两个可以共享我们的 Trainer 类及其自定义 train_step() 的不同模型。
# 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)
从模型创建 Trainer 类对象
trainer_1 = MyTrainer(model_a)
trainer_2 = MyTrainer(model_b)
编译并对 MNIST 数据集拟合模型。
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>