Tensorboard

1.启动tensorboard

tensorboard --logdir="./logs" --port 6006

2.显示graph

import tensorflow as tf  
a=tf.constant(2)  
b=tf.constant(3)  
x=tf.add(a,b)  
sess = tf.Session()
writer=tf.summary.FileWriter('./logs',sess.graph)  
writer.close()  

3.查看变量

import tensorflow as tf

a = tf.constant(1.0,name="a")    
b = tf.random_normal(shape=[], mean=0, stddev=1, name='b')    
print a.get_shape()
print b.get_shape()

x=tf.add(a,b)  

tf.summary.scalar('x', x)
tf.summary.histogram('x_h', x)

merged = tf.summary.merge_all()

sess = tf.Session()
writer=tf.summary.FileWriter('./logs',sess.graph)  

step = 0
while True:
    mysum = sess.run(merged)
    writer.add_summary(mysum,step)
    step+=1

writer.close()  

更多参考资料：http://www.jianshu.com/p/d059ffea9ec0