TensorFlow: snippet: TensorBoard

x = tf.placeholder(tf.float32, [None, 784], name="x-input")
W = tf.Variable(tf.zeros([784,10]), name="weights")
b = tf.Variable(tf.zeros([10], name="bias"))
 
with tf.name_scope("Wx_b") as scope:
  y = tf.nn.softmax(tf.matmul(x,W) + b)
 
w_hist = tf.histogram_summary("weights", W)
b_hist = tf.histogram_summary("biases", b)
y_hist = tf.histogram_summary("y", y)
 
y_ = tf.placeholder(tf.float32, [None,10], name="y-input")

with tf.name_scope("xent") as scope:
  cross_entropy = -tf.reduce_sum(y_*tf.log(y))
  ce_summ = tf.scalar_summary("cross entropy", cross_entropy)
with tf.name_scope("train") as scope:
  train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)
 
with tf.name_scope("test") as scope:
  correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))
  accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
  accuracy_summary = tf.scalar_summary("accuracy", accuracy)
 
merged = tf.merge_all_summaries()
writer = tf.train.SummaryWriter("/tmp/mnist_logs", sess.graph_def)
tf.initialize_all_variables().run()
 
for i in range(1000):
  if i % 10 == 0:
    feed = {x: mnist.test.images, y_: mnist.test.labels}
    result = sess.run([merged, accuracy], feed_dict=feed)
    summary_str = result[0]
    acc = result[1]
    writer.add_summary(summary_str, i)
    print("Accuracy at step %s: %s" % (i, acc))
  else:
    batch_xs, batch_ys = mnist.train.next_batch(100)
    feed = {x: batch_xs, y_: batch_ys}
    sess.run(train_step, feed_dict=feed)
 
print(accuracy.eval({x: mnist.test.images, y_: mnist.test.labels}))

8 行目: データ収集のための summary operations を追加します。

25 行目: 全ての要約をマージしてそれらを /tmp/mnist_logs に書き出します。

30 行目: 要約データと正解率を記録します。

【参考】
（翻訳/解説）TensorFlow : How To : 学習を視覚化する

以上