''' Created on May 24, 2017 @author: p0079482 ''' #使用程序输出日志 import tensorflow as tf with tf.Session() as sess: tf.initialize_all_variables().run() for i in range(TRAINING_STEPS): xs,ys=mnist.train.next_batch(BATCH_SIZE) #每1000轮记录一次运行状态 if i%1000==0: #配置运行时需要记录的信息 run_options = tf.RunOptions(trace_level=tf.RunOptions.FULL_TRACE) #运行时记录运行信息的proto run_metadata=tf.RunMetadata() #将配置信息和记录运行信息的proto传入运行的过程,从而记录运行时每一个节点的时间、空间开销信息 _,loss_value,step=sess.run([train_op,loss,global_step], feed_dict={x:xs,y_:ys}, options=run_options,run_metadata=run_metadata) #将节点在运行时的信息写入日志文件 train_writer.add_run_metadata(run_metadata,'step%03d'%i) print("After %d training step(s),loss on training batch is %g."%(step,loss_value)) else: _,loss_value,step=sess.run([train_op,loss,global_step],feed_dict={x:xs,y_:ys})