tensorflow 2.0 实战-Minst
2. 手写字体识别
- 数据载入
- 搭建全连接网络
- 计算输出和 Loss
- 计算梯度与回传
- Loop
- 完整代码
数据载入
import os
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers, optimizers, datasets
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
(x, y), (x_val, y_val) = datasets.mnist.load_data() # train and test
x = tf.convert_to_tensor(x, dtype=tf.float32) / 255. # normalize （broadcasting）
y = tf.convert_to_tensor(y, dtype=tf.int32)
y = tf.one_hot(y, depth=10)
print(x.shape, y.shape) # (60000, 28, 28) (60000, 10)
train_dataset = tf.data.Dataset.from_tensor_slices((x, y))
train_dataset = train_dataset.batch(200) # setting batch
# for step, (x, y) in enumerate(train_dataset):
# print (step, x.shape, y.shape)
0 (200, 28, 28) (200, 10)
1 (200, 28, 28) (200, 10)
2 (200, 28, 28) (200, 10)
3 (200, 28, 28) (200, 10)
4 (200, 28, 28) (200, 10)
5 (200, 28, 28) (200, 10)
6 (200, 28, 28) (200, 10)
......
295 (200, 28, 28) (200, 10)
296 (200, 28, 28) (200, 10)
297 (200, 28, 28) (200, 10)
298 (200, 28, 28) (200, 10)
299 (200, 28, 28) (200, 10)
搭建全连接网络
o u t = r e l u { r e l u { r e l u { x @ w 1 + b 1 } @ w 2 + b 2 } @ w 3 + b 3 } out = relu\{relu\{relu\{x@w_1+b_1\}@w_2+b_2\}@w_3+b_3\} out=relu{relu{relu{x@w1+b1}@w2+b2}@w3+b3}
model = keras.Sequential([
layers.Dense(512, activation='relu'),
layers.Dense(256, activation='relu'),
layers.Dense(10)])
optimizer = optimizers.SGD(learning_rate=0.001)
计算输出和 Loss
with tf.GradientTape() as tape:
# [b, 28, 28] => [b, 784]
x = tf.reshape(x, (-1, 28 * 28))
# Step1. compute output
# [b, 784] => [b, 10]
out = model(x)
# Step2. compute loss
loss = tf.reduce_sum(tf.square(out - y)) / x.shape[0]
计算梯度与回传
# Step3. optimize and update w1, w2, w3, b1, b2, b3
grads = tape.gradient(loss, model.trainable_variables)
# w' = w - lr * grad
optimizer.apply_gradients(zip(grads, model.trainable_variables))
Loop
def train_epoch(epoch):
# Step4.loop
for step, (x, y) in enumerate(train_dataset):
with tf.GradientTape() as tape:
# [b, 28, 28] => [b, 784]
x = tf.reshape(x, (-1, 28 * 28))
# Step1. compute output
# [b, 784] => [b, 10]
out = model(x)
# Step2. compute loss
loss = tf.reduce_sum(tf.square(out - y)) / x.shape[0]
# Step3. optimize and update w1, w2, w3, b1, b2, b3
grads = tape.gradient(loss, model.trainable_variables)
# w' = w - lr * grad
optimizer.apply_gradients(zip(grads, model.trainable_variables))
if step % 100 == 0:
print(epoch, step, 'loss:', loss.numpy())
完整代码
import os
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers, optimizers, datasets
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
(x, y), (x_val, y_val) = datasets.mnist.load_data()
x = tf.convert_to_tensor(x, dtype=tf.float32) / 255.
y = tf.convert_to_tensor(y, dtype=tf.int32)
y = tf.one_hot(y, depth=10)
print(x.shape, y.shape)
train_dataset = tf.data.Dataset.from_tensor_slices((x, y))
train_dataset = train_dataset.batch(200)
model = keras.Sequential([
layers.Dense(512, activation='relu'),
layers.Dense(256, activation='relu'),
layers.Dense(10)])
optimizer = optimizers.SGD(learning_rate=0.001)
def train_epoch(epoch):
# Step4.loop
for step, (x, y) in enumerate(train_dataset):
with tf.GradientTape() as tape:
# [b, 28, 28] => [b, 784]
x = tf.reshape(x, (-1, 28 * 28))
# Step1. compute output
# [b, 784] => [b, 10]
out = model(x)
# Step2. compute loss
loss = tf.reduce_sum(tf.square(out - y)) / x.shape[0]
# Step3. optimize and update w1, w2, w3, b1, b2, b3
grads = tape.gradient(loss, model.trainable_variables)
# w' = w - lr * grad
optimizer.apply_gradients(zip(grads, model.trainable_variables))
if step % 100 == 0:
print(epoch, step, 'loss:', loss.numpy())
def train():
for epoch in range(30):
train_epoch(epoch)
if __name__ == '__main__':
train()