恢复已保存的TensorFlow模型以评估测试集

Question

我看过一些关于恢复模型和导出图表的文档页面的帖子,但我想我错过了一些东西.TFGoogle

我使用此Gist中的代码来保存模型以及定义模型的此utils文件

现在我想恢复它并运行以前看不见的测试数据,如下所示:

def evaluate(X_data, y_data):
    num_examples = len(X_data)
    total_accuracy = 0
    total_loss = 0
    sess = tf.get_default_session()
    acc_steps = len(X_data) // BATCH_SIZE
    for i in range(acc_steps):
        batch_x, batch_y = next_batch(X_val, Y_val, BATCH_SIZE)

        loss, accuracy = sess.run([loss_value, acc], feed_dict={
                images_placeholder: batch_x,
                labels_placeholder: batch_y,
                keep_prob: 0.5
                })
        total_accuracy += (accuracy * len(batch_x))
        total_loss += (loss * len(batch_x))
    return (total_accuracy / num_examples, total_loss / num_examples)

## re-execute the code that defines the model

# Image Tensor
images_placeholder = tf.placeholder(tf.float32, shape=[None, 32, 32, 3], name='x')

gray = tf.image.rgb_to_grayscale(images_placeholder, name='gray')

gray /= 255.

# Label Tensor
labels_placeholder = tf.placeholder(tf.float32, shape=(None, 43), name='y')

# dropout Tensor
keep_prob = tf.placeholder(tf.float32, name='drop')

# construct model
logits = inference(gray, keep_prob)

# calculate loss
loss_value = loss(logits, labels_placeholder)

# training
train_op = training(loss_value, 0.001)

# accuracy
acc = accuracy(logits, labels_placeholder)

with tf.Session() as sess:
    loader = tf.train.import_meta_graph('gtsd.meta')
    loader.restore(sess, tf.train.latest_checkpoint('./'))
    sess.run(tf.initialize_all_variables())   
    test_accuracy = evaluate(X_test, y_test)
    print("Test Accuracy = {:.3f}".format(test_accuracy[0]))

我的测试精度只有3%.但是,如果我没有关闭笔记本并在训练模型后立即运行测试代码,我的准确率为95%.

这让我相信我没有正确加载模型？

Answer 1

这两个问题产生了问题:

loader.restore(sess, tf.train.latest_checkpoint('./'))
sess.run(tf.initialize_all_variables())   

第一行从检查点加载已保存的模型.第二行重新初始化模型中的所有变量(例如权重矩阵,卷积滤波器和偏置矢量),通常是随机数,并覆盖加载的值.

解决方案很简单:删除第二行(sess.run(tf.initialize_all_variables())),评估将继续使用从检查点加载的训练值.

PS.这种变化很可能会给你一个关于"未初始化变量"的错误.在这种情况下,您应该执行sess.run(tf.initialize_all_variables())以初始化在执行之前未保存在检查点中的任何变量loader.restore(sess, tf.train.latest_checkpoint('./')).