keras高层API

keras高层API

keras高层API主要用于代码的简化，即将训练，预测，误差计算等代码简化

1. 数值更新

• 数值指准确率和损失函数
• 按batch计算的准确率和损失函数不如按epoch计算的更有说服力
• 主要流程：新建一个Metrics；更新参数（传入需要组成一个epoch时在输出的参数）；传出结果；清空Metrics
• 建立一个Metrics：

acc_meter = metrics.Accuracy()
loss_meter = metrics.Mean()

• 更新参数：

loss_meter.update_state(loss)
acc_meter.update_state(y,pred)

• 传出结果：

print(step, 'loss:', loss_meter.result().numpy())
    ...
print(step, 'acc:', total_correct/total, acc_meter.result().numpy())   

• 整体代码：

import  tensorflow as tf
from    tensorflow.keras import datasets, layers, optimizers, Sequential, metrics


def preprocess(x, y):

    x = tf.cast(x, dtype=tf.float32) / 255.
    y = tf.cast(y, dtype=tf.int32)

    return x,y


batchsz = 128
(x, y), (x_val, y_val) = datasets.mnist.load_data()
print('datasets:', x.shape, y.shape, x.min(), x.max())



db = tf.data.Dataset.from_tensor_slices((x,y))
db = db.map(preprocess).shuffle(60000).batch(batchsz).repeat(10)

ds_val = tf.data.Dataset.from_tensor_slices((x_val, y_val))
ds_val = ds_val.map(preprocess).batch(batchsz) 




network = Sequential([layers.Dense(256, activation='relu'),
                     layers.Dense(128, activation='relu'),
                     layers.Dense(64, activation='relu'),
                     layers.Dense(32, activation='relu'),
                     layers.Dense(10)])
network.build(input_shape=(None, 28*28))
network.summary()

optimizer = optimizers.Adam(lr=0.01)

acc_meter = metrics.Accuracy()
loss_meter = metrics.Mean()


for step, (x,y) in enumerate(db):

    with tf.GradientTape() as tape:
        # [b, 28, 28] => [b, 784]
        x = tf.reshape(x, (-1, 28*28))
        # [b, 784] => [b, 10]
        out = network(x)
        # [b] => [b, 10]
        y_onehot = tf.one_hot(y, depth=10) 
        # [b]
        loss = tf.reduce_mean(tf.losses.categorical_crossentropy(y_onehot, out, from_logits=True))

        loss_meter.update_state(loss)

 

    grads = tape.gradient(loss, network.trainable_variables)
    optimizer.apply_gradients(zip(grads, network.trainable_variables))


    if step % 100 == 0:

        print(step, 'loss:', loss_meter.result().numpy()) 
        loss_meter.reset_states()


    # evaluate
    if step % 500 == 0:
        total, total_correct = 0., 0
        acc_meter.reset_states()

        for step, (x, y) in enumerate(ds_val): 
            # [b, 28, 28] => [b, 784]
            x = tf.reshape(x, (-1, 28*28))
            # [b, 784] => [b, 10]
            out = network(x) 


            # [b, 10] => [b] 
            pred = tf.argmax(out, axis=1) 
            pred = tf.cast(pred, dtype=tf.int32)
            # bool type 
            correct = tf.equal(pred, y)
            # bool tensor => int tensor => numpy
            total_correct += tf.reduce_sum(tf.cast(correct, dtype=tf.int32)).numpy()
            total += x.shape[0]

            acc_meter.update_state(y, pred)


        print(step, 'Evaluate Acc:', total_correct/total, acc_meter.result().numpy())

2. 快捷训练方法

• 主要包括compile；fit；evaluate；predict
• compile：损失函数，优化器，评估指标的选择

network.compile(optimizer=optimizers.Adam(lr=1e-3),
                loss=tf.losses.CategoricalCrossentropy(from_logits=True),
                metrics=['accuracy']
        )

• fit： 完成标准的train流程

network.fit(db, epochs=10)

• evaluate：测试

network.compile(optimizer=optimizers.Adam(lr=1e-3),
                loss=tf.losses.CategoricalCrossentropy(from_logits=True),
                metrics=['accuracy']
        )
network.fit(db, epochs=10, validation_data=ds_val, validation_steps=2)
network.evaluate(ds_val)

• predict：对新的sample的预测

sample = next(iter(ds_val))
x = sample[0]
y = sample[1]
pred = network.predict(x)
y = tf.argmax(y, axis=1)
pred = tf.argmax(pred, axis=1)

• 测试：需要对fit添加新的参数，其中为在哪个集上做测试，为每隔几个db做一次测试，这个测试的结果时插在db之间的，这样的方式适合于在训练到多少准确率的情况下停止训练的情况

network.compile(optimizer=optimizers.Adam(lr=1e-3),
                loss=tf.losses.CategoricalCrossentropy(from_logits=True),
                metrics=['accuracy']
        )
network.fit(db, epochs=10, validation_data=ds_val, validation_steps=2)