学习tensorflow遇到的问题

1. 怎样在代码中打印tesor的形状等：

在下载的源码中加入print，报 SyntaxError

print W.shape
          ^

SyntaxError: invalid syntax

from __future__ import print_function  是这个导致的具体原因还不知道

这样就可以很清楚的看到 tensor形状的变化：CNN, MaxPool等

(?, 784)
reshape/Reshape:0
(?, 28, 28, 1)
conv1/Relu:0
(?, 28, 28, 32)
pool1/MaxPool:0
(?, 14, 14, 32)
conv2/Relu:0
(?, 14, 14, 64)
pool2/MaxPool:0
(?, 7, 7, 64)
fc1/Reshape:0
(?, 3136)
fc1/Relu:0
(?, 1024)
dropout/dropout/mul:0
(?, 1024)
fc2/add:0
(?, 10)

2.怎样得到模型(变量的值)

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")

但保存再文件中的都是二进制，看不出是什么参数，怎么处理？

3. 用 TensorBoard 可视化、图表展示

merged_summary_op = tf.merge_all_summaries()
summary_writer = tf.train.SummaryWriter('/tmp/mnist_logs', sess.graph)
total_step = 0
while training:
  total_step += 1
  session.run(training_op)
  if total_step % 100 == 0:
    summary_str = session.run(merged_summary_op)
    summary_writer.add_summary(summary_str, total_step)

pip安装了 TensorBoard

然后：

tensorboard --logdir=/path/to/log-directory