读取tensorflow checkpoint文件的变量名和值

	使用tensorflow过程中，我们训练完成的参数结果都会保持checkpoint文件，那么我们到底如何查看他呢？下面给出代码

import tensorflow as tf
reader =tf.train.NewCheckpointReader("./train_dir/model-299")
# NewCheck =tf.train.NewCheckpointReader("./inception_v3/inception_v3.ckpt")
var_to_shape_map = reader.get_variable_to_shape_map()

# 输出权重tensor名字和值
for key in var_to_shape_map:
    print("tensor_name: ", key,reader.get_tensor(key))

我这里用的是Google的inceptionv3 做的测试 ，部分结果如下：

tensor_name: InceptionV3/AuxLogits/Conv2d_2a_5x5/weights/RMSProp (5, 5, 128, 768)
tensor_name: InceptionV3/Mixed_5c/Branch_0/Conv2d_0a_1x1/BatchNorm/beta (64,)
tensor_name: InceptionV3/AuxLogits/Conv2d_1b_1x1/BatchNorm/beta (128,)
tensor_name: InceptionV3/AuxLogits/Conv2d_2a_5x5/BatchNorm/moving_variance (768,)
tensor_name: InceptionV3/AuxLogits/Conv2d_2a_5x5/weights (5, 5, 128, 768)
tensor_name: InceptionV3/AuxLogits/Conv2d_1b_1x1/BatchNorm/moving_mean (128,)
.
.
.