作者: Kenneth Borup
创建日期 2020/09/01
最后修改日期 2020/09/01
描述: 经典知识蒸馏的实现。
知识蒸馏是一种模型压缩过程,其中训练一个小型(学生)模型来匹配一个大型预训练(教师)模型。 通过最小化损失函数,旨在匹配软化的教师 logits 以及真实标签,将知识从教师模型转移到学生模型。
通过在 softmax 中应用“温度”缩放函数来软化 logits,有效地平滑概率分布并揭示教师学习到的类间关系。
参考
import os
import keras
from keras import layers
from keras import ops
import numpy as np
Distiller()
类自定义 Distiller()
类,覆盖了 Model
的方法 compile
,compute_loss
和 call
。 为了使用蒸馏器,我们需要
temperature
,用于软化的学生预测和软化的教师标签之间的差异alpha
因子来衡量学生损失和蒸馏损失的权重在 compute_loss
方法中,我们执行教师和学生的前向传递,通过 alpha
和 1 - alpha
分别对 student_loss
和 distillation_loss
进行加权来计算损失。 注意:仅更新学生权重。
class Distiller(keras.Model):
def __init__(self, student, teacher):
super().__init__()
self.teacher = teacher
self.student = student
def compile(
self,
optimizer,
metrics,
student_loss_fn,
distillation_loss_fn,
alpha=0.1,
temperature=3,
):
"""Configure the distiller.
Args:
optimizer: Keras optimizer for the student weights
metrics: Keras metrics for evaluation
student_loss_fn: Loss function of difference between student
predictions and ground-truth
distillation_loss_fn: Loss function of difference between soft
student predictions and soft teacher predictions
alpha: weight to student_loss_fn and 1-alpha to distillation_loss_fn
temperature: Temperature for softening probability distributions.
Larger temperature gives softer distributions.
"""
super().compile(optimizer=optimizer, metrics=metrics)
self.student_loss_fn = student_loss_fn
self.distillation_loss_fn = distillation_loss_fn
self.alpha = alpha
self.temperature = temperature
def compute_loss(
self, x=None, y=None, y_pred=None, sample_weight=None, allow_empty=False
):
teacher_pred = self.teacher(x, training=False)
student_loss = self.student_loss_fn(y, y_pred)
distillation_loss = self.distillation_loss_fn(
ops.softmax(teacher_pred / self.temperature, axis=1),
ops.softmax(y_pred / self.temperature, axis=1),
) * (self.temperature**2)
loss = self.alpha * student_loss + (1 - self.alpha) * distillation_loss
return loss
def call(self, x):
return self.student(x)
最初,我们创建一个教师模型和一个较小的学生模型。 两个模型都是使用 Sequential()
创建的卷积神经网络,但可以是任何 Keras 模型。
# Create the teacher
teacher = keras.Sequential(
[
keras.Input(shape=(28, 28, 1)),
layers.Conv2D(256, (3, 3), strides=(2, 2), padding="same"),
layers.LeakyReLU(negative_slope=0.2),
layers.MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding="same"),
layers.Conv2D(512, (3, 3), strides=(2, 2), padding="same"),
layers.Flatten(),
layers.Dense(10),
],
name="teacher",
)
# Create the student
student = keras.Sequential(
[
keras.Input(shape=(28, 28, 1)),
layers.Conv2D(16, (3, 3), strides=(2, 2), padding="same"),
layers.LeakyReLU(negative_slope=0.2),
layers.MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding="same"),
layers.Conv2D(32, (3, 3), strides=(2, 2), padding="same"),
layers.Flatten(),
layers.Dense(10),
],
name="student",
)
# Clone student for later comparison
student_scratch = keras.models.clone_model(student)
用于训练教师和蒸馏教师的数据集是 MNIST,对于任何其他数据集(例如 CIFAR-10),该过程都是等效的,只需选择合适的模型即可。 学生和教师都在训练集上进行训练,并在测试集上进行评估。
# Prepare the train and test dataset.
batch_size = 64
(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()
# Normalize data
x_train = x_train.astype("float32") / 255.0
x_train = np.reshape(x_train, (-1, 28, 28, 1))
x_test = x_test.astype("float32") / 255.0
x_test = np.reshape(x_test, (-1, 28, 28, 1))
在知识蒸馏中,我们假设教师经过训练并且是固定的。 因此,我们首先以通常的方式在训练集上训练教师模型。
# Train teacher as usual
teacher.compile(
optimizer=keras.optimizers.Adam(),
loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=[keras.metrics.SparseCategoricalAccuracy()],
)
# Train and evaluate teacher on data.
teacher.fit(x_train, y_train, epochs=5)
teacher.evaluate(x_test, y_test)
Epoch 1/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 8s 3ms/step - loss: 0.2408 - sparse_categorical_accuracy: 0.9259
Epoch 2/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 5s 3ms/step - loss: 0.0912 - sparse_categorical_accuracy: 0.9726
Epoch 3/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 7s 4ms/step - loss: 0.0758 - sparse_categorical_accuracy: 0.9777
Epoch 4/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 5s 3ms/step - loss: 0.0690 - sparse_categorical_accuracy: 0.9797
Epoch 5/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 5s 3ms/step - loss: 0.0582 - sparse_categorical_accuracy: 0.9825
313/313 ━━━━━━━━━━━━━━━━━━━━ 1s 3ms/step - loss: 0.0931 - sparse_categorical_accuracy: 0.9760
[0.09044107794761658, 0.978100061416626]
我们已经训练了教师模型,我们只需要初始化一个 Distiller(student, teacher)
实例,使用所需的损失、超参数和优化器对其进行 compile()
,并将教师蒸馏到学生。
# Initialize and compile distiller
distiller = Distiller(student=student, teacher=teacher)
distiller.compile(
optimizer=keras.optimizers.Adam(),
metrics=[keras.metrics.SparseCategoricalAccuracy()],
student_loss_fn=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
distillation_loss_fn=keras.losses.KLDivergence(),
alpha=0.1,
temperature=10,
)
# Distill teacher to student
distiller.fit(x_train, y_train, epochs=3)
# Evaluate student on test dataset
distiller.evaluate(x_test, y_test)
Epoch 1/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 8s 3ms/step - loss: 1.8752 - sparse_categorical_accuracy: 0.7357
Epoch 2/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 6s 3ms/step - loss: 0.0333 - sparse_categorical_accuracy: 0.9475
Epoch 3/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 6s 3ms/step - loss: 0.0223 - sparse_categorical_accuracy: 0.9621
313/313 ━━━━━━━━━━━━━━━━━━━━ 2s 4ms/step - loss: 0.0189 - sparse_categorical_accuracy: 0.9629
[0.017046602442860603, 0.969200074672699]
我们还可以从头开始训练一个等效的学生模型,而无需教师,以便评估通过知识蒸馏获得的性能提升。
# Train student as doen usually
student_scratch.compile(
optimizer=keras.optimizers.Adam(),
loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=[keras.metrics.SparseCategoricalAccuracy()],
)
# Train and evaluate student trained from scratch.
student_scratch.fit(x_train, y_train, epochs=3)
student_scratch.evaluate(x_test, y_test)
Epoch 1/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 4s 1ms/step - loss: 0.5111 - sparse_categorical_accuracy: 0.8460
Epoch 2/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 3s 1ms/step - loss: 0.1039 - sparse_categorical_accuracy: 0.9687
Epoch 3/3
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 3s 1ms/step - loss: 0.0748 - sparse_categorical_accuracy: 0.9780
313/313 ━━━━━━━━━━━━━━━━━━━━ 1s 2ms/step - loss: 0.0744 - sparse_categorical_accuracy: 0.9737
[0.0629437193274498, 0.9778000712394714]
如果教师模型被训练了 5 个完整 epoch,而学生模型在这个教师模型上被蒸馏了 3 个完整 epoch,那么在这个示例中,您应该体验到与从头开始训练相同的学生模型相比,甚至与教师模型本身相比,性能有所提升。 您应该期望教师的准确率在 97.6% 左右,从头开始训练的学生的准确率在 97.6% 左右,而经过蒸馏的学生模型准确率应该在 98.1% 左右。 删除或尝试不同的种子以使用不同的权重初始化。