TensorFlow自帶一個非常好的視覺化工具tensorboard，可以從極客學院檢視它的詳細介紹。

tensorboard介面，視覺化訓練過程。

• 使用tensorboard需要首先定義好需要觀察的tensor的名字，以及運算過程的名字。

#在宣告張量時，輸入引數name即為顯示在tensorboard上張量的名字
x = tf.placeholder(tf.float32, [None, 784], name='InputData')  #輸入張量的名字
y = tf.placeholder(tf.float32, [None, 10], name='LabelData')   #輸出張量的名字
 

w = tf.Variable(tf.random_normal([n_input, n_hidden_1]), name='w')  #權值張量的名字

#運算過程的名字，使用tf.name_scope('op_name')來定義
with tf.name_scope('Loss'):
    loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))

• 使用tf.summary記錄張量或者標量的值，以便用圖形顯示

四個記錄變數的函式

tf.summary.scalar('Loss'
 
, loss)         #記錄標量
tf.summary.histogram('var_name', var)   #變數的直方圖
tf.summary.distribution                 #變數的分佈圖
tf.summary.image                        #記錄影象

• 在張量的名字以及要觀察的變數定義好之後，需要定義運算，在每次訓練時，執行該運算就可以記錄資料了

#使用tf.summary.merge_all()定義運算
merged_summary_op = tf.summary.merge_all()

#使用tf.summary.FileWriter(path, graph)將日誌寫入到檔案中
 

#path:日誌檔案儲存路徑， graph:需要儲存的圖
summary_writer = tf.summary.FileWriter(logs_path, graph=tf.get_default_graph())

#每次訓練時執行運算與寫入函式
with tf.Session() as sess:
    summary = sess.run(merged_summary_op, feed_dict)  #執行運算
    summary_writer.add_summary(summary, curr_step)    #寫入檔案

• 程式執行完後，在命令列輸入以下命令後，根據輸出提示在瀏覽器中輸入地址即可玩耍tensorboard了

tensorboard --logdir=logs_path

from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
import tensorflow as tf

learning_rate = 0.001
train_epochs = 20
batch_size = 64
logs_path = 'tmp/tensorflow_logs/mlp_mnist'   #日誌儲存路徑

n_input = 784
n_hidden1 = 100
n_hidden2 = 100
n_classes = 10

#name引數，記錄變數名字
x = tf.placeholder(tf.float32, shape=[None, n_input], name='InputData')
y = tf.placeholder(tf.float32, shape=[None, n_classes], name='LabelData')
weights = {'w1': tf.Variable(tf.random_normal([n_input, n_hidden1]), name='W1'),
           'w2': tf.Variable(tf.random_normal([n_hidden1, n_hidden2]), name='W2'),
           'w3': tf.Variable(tf.random_normal([n_hidden2, n_classes]), name='W3')}
biases = {'b1': tf.Variable(tf.random_normal([n_hidden1]), name='b1'),
          'b2': tf.Variable(tf.random_normal([n_hidden2]), name='b2'),
          'b3': tf.Variable(tf.random_normal([n_classes]), name='b3')}

def inference(input_x):
    layer_1 = tf.nn.relu(tf.matmul(x, weights['w1']) + biases['b1'])
    tf.summary.histogram('layer_1', layer_1)   #記錄變數直方圖
    layer_2 = tf.nn.relu(tf.matmul(layer_1, weights['w2']) + biases['b2'])
    tf.summary.histogram('layer_2', layer_2)   #記錄變數直方圖
    out_layer = tf.matmul(layer_2, weights['w3']) + biases['b3']
    return out_layer

#定義計算過程的名字
with tf.name_scope('Inference'):
    logits = inference(x)
with tf.name_scope('Loss'):
    loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=y))
with tf.name_scope('Optimizer'):
    optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
    train_op = optimizer.minimize(loss)
with tf.name_scope('Accuracy'):
    pre_correct = tf.equal(tf.argmax(y, 1), tf.argmax(tf.nn.softmax(logits), 1))
    accuracy = tf.reduce_mean(tf.cast(pre_correct, tf.float32))

#記錄張量的資料
tf.summary.scalar("Loss", loss)
tf.summary.scalar("Accuracy", accuracy)

init = tf.global_variables_initializer()
merged_summary_op = tf.summary.merge_all()   #定義記錄運算

with tf.Session() as sess:
    sess.run(init)
    summary_writer = tf.summary.FileWriter(logs_path, graph=tf.get_default_graph())   #建立寫物件
    total_batch = int(mnist.train.num_examples / batch_size)

    for epoch in range(train_epochs):
        for batch in range(total_batch):
            batch_x, batch_y = mnist.train.next_batch(batch_size)
            _, loss_, summary = sess.run([train_op, loss, merged_summary_op], feed_dict={x:batch_x, y:batch_y})   #執行記錄運算
            summary_writer.add_summary(summary, epoch * total_batch + batch)     #將日誌寫入檔案
        if epoch % 5 == 0:
            loss_, acc = sess.run([loss, accuracy], feed_dict={x:batch_x, y:batch_y})
            print("epoch {},  loss {:.4f}, acc {:.3f}".format(epoch, loss_, acc))

    print("optimizer finished!")

    #計算測試集的準確度
    test_acc = sess.run(accuracy, feed_dict={x:mnist.test.images, y:mnist.test.labels})
    print('test accuracy', test_acc)

    #命令列輸入tensorboard --logdir==tmp/tensorflow_logs