tensorflow学习笔记（四）tensorboard数据可视

tf.histogram_summary(layer_name+"/weights",Weights)       #name命名，Weights赋值

tf.scalar_summary('loss',loss)          #命名和赋值

#合并到Summary中

merged = tf.merge_all_summaries()

#选定可视化存储目录

writer = tf.train.SummaryWriter("/目录",sess.graph)

result = sess.run(merged)#merged也是需要run的

writer.add_summary(result,i)