Tensorflow打印网络结构与变量

最新推荐文章于 2022-12-11 22:22:11 发布
最新推荐文章于 2022-12-11 22:22:11 发布
阅读量4.9k 收藏 2
点赞数 4
分类专栏： 深度学习框架 文章标签： Tensorflow 网络可视化
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/LCCFlccf/article/details/102267093
版权

在用tensorflow搭建好网络之后，如果可视化一下网络的结构与变量，会对网络结构有一个更直观的了解。

网络结构与变量的可视化方法如下：

在搭建好网络之后，如下代码可以打印出网络的变量


variables = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES)
for v in variables:
    print(v)

活动的输出如下：

<tf.Variable 'resnet_v2_152/conv1/weights:0' shape=(7, 7, 3, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/conv1/biases:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/preact/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/preact/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/shortcut/weights:0' shape=(1, 1, 64, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/shortcut/biases:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv1/weights:0' shape=(1, 1, 64, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv2/weights:0' shape=(3, 3, 64, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv3/weights:0' shape=(1, 1, 64, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_1/bottleneck_v2/conv3/biases:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/preact/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/preact/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv1/weights:0' shape=(1, 1, 256, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv2/weights:0' shape=(3, 3, 64, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv3/weights:0' shape=(1, 1, 64, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_2/bottleneck_v2/conv3/biases:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/preact/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/preact/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv1/weights:0' shape=(1, 1, 256, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv2/weights:0' shape=(3, 3, 64, 64) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(64,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv3/weights:0' shape=(1, 1, 64, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block1/unit_3/bottleneck_v2/conv3/biases:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/preact/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/preact/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/shortcut/weights:0' shape=(1, 1, 256, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/shortcut/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv1/weights:0' shape=(1, 1, 256, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_1/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_2/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_3/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_4/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_5/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_6/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_7/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv2/weights:0' shape=(3, 3, 128, 128) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(128,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv3/weights:0' shape=(1, 1, 128, 512) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block2/unit_8/bottleneck_v2/conv3/biases:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/preact/gamma:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/preact/beta:0' shape=(512,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/shortcut/weights:0' shape=(1, 1, 512, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/shortcut/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv1/weights:0' shape=(1, 1, 512, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_1/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_2/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_3/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_4/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_5/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_6/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_7/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_8/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_9/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_10/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_11/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_12/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_13/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_14/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_15/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_16/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_17/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_18/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_19/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv3/weights:0' shape=(1, 1, 256, 1024) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_20/bottleneck_v2/conv3/biases:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/preact/gamma:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/preact/beta:0' shape=(1024,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv1/weights:0' shape=(1, 1, 1024, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv1/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv1/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv2/weights:0' shape=(3, 3, 256, 256) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv2/BatchNorm/gamma:0' shape=(256,) dtype=float32_ref>
<tf.Variable 'resnet_v2_152/block3/unit_21/bottleneck_v2/conv2/BatchNorm/beta:0' shape=(256,) dtype=float32_ref>

 

LCCFlccf
关注
  • 4
    点赞
  • 踩
  • 2
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
专栏目录
tensorflow打印模型图_从Tensorflow模型文件中解析并显示网络结构图（pb模型篇）...
weixin_39866966的博客
12-20 493
最近看到一个巨牛的人工智能教程，分享一下给大家。教程不仅是零基础，通俗易懂，而且非常风趣幽默，像看小说一样！觉得太牛了，所以分享给大家。平时碎片时间可以当小说看，【点这里可以去膜拜一下大神的“小说”】。Tensorflow官方提供的Tensorboard可以可视化神经网络结构图，但是说实话，我几乎从来不用。主要是因为Tensorboard中查看到的图结构太混乱了，包含了网络中所有的计算节点(读取数...
打印tensorflow结构
山中有石为玉
05-10 6715
代码： import tensorflow as tf sess=tf.Session with tf.Graph().as_default(): with tf.gfile.FastGFile('./models/emotion_model.pb', 'rb') as modelfile: graph_def=tf.GraphDef() graph_...
初学tensorflow，生成一个自己的网络结构，并用tensorboard可视化（代码）
11-20
初学tensorflow，生成一个自己的网络结构，并用tensorboard可视化（代码）
tensorflow-tensorboard打印网络结构
秀呀秀的博客
09-24 1789
from __future__ import division, print_function import tensorflow as tf import numpy as np import argparse import cv2 import os os.environ['CUDA_VISIBLE_DEVICES'] = "7" from utils.misc_utils import parse_anchors, read_class_names from utils.nms_utils imp.
打印出tensorflow2.x的神经网络结构
爱听许嵩歌的博客
12-09 4834
使用tensorflow中plot_model函数，以VGG-16网络为例子，全部代码如下。 from tensorflow.keras.models import Sequential from tensorflow.keras.layers import Dense, Conv1D, MaxPool1D, LSTM, BatchNormalization, Dropout, Flatten, Reshape from tensorflow.keras.utils import plot_model m
打印tensorflow恢复模型中所有变量与操作节点方式
09-16
本文将详细讨论如何打印恢复模型中的所有变量与操作节点，以及`.ckpt`、`.ckpt.meta`、`.ckpt.index`和`.pb`文件的关系。 首先，当我们使用`tf.train.Saver()`保存模型时，会产生一系列的文件，包括： 1. `.ckpt`...
从训练好的tensorflow模型中打印训练变量实例
09-18
本文将详细讲解如何从训练好的TensorFlow模型中提取并打印这些训练变量。 首先，我们需要导入必要的库，即TensorFlow库，用于操作和读取模型： ```python import tensorflow as tf ``` 当你完成了一次TensorFlow...
tensorflow 获取变量&打印权值的实例讲解
09-20
在某些情况下，例如使用预定义的库（如tensorflow的slim库）构建神经网络层时，我们可能无法直接访问到内部创建的变量，因为它们是由库自动定义的。以下是如何在这样的场景下获取和打印变量的值的实例讲解。 首先，...
TensorFlow实现打印每一层的输出
12-23
在test.py中可以通过如下代码直接生成带weight的pb文件，也可以通过tf官方的freeze_graph.py将ckpt转为pb文件。 constant_graph = graph_util.convert_variables_to_constants(sess, sess.graph_def,['net_loss/inference/encode/conv_output/conv_output']) with tf.gfile.FastGFile('net_model.pb', mode='wb') as f: f.write(constant_graph.SerializeT
TensorFlow 输出checkpoint 中的变量名与变量值方式
09-17
### TensorFlow 输出checkpoint中的变量名与变量值方法 在深度学习领域，TensorFlow作为一个非常流行的开源机器学习框架，被广泛应用于各种复杂模型的构建与训练之中。对于开发者而言，能够有效地管理和查看模型...
tensorflow输出网络结构_【机器学习Tensorflow实践】神经网络搭建流程（四）
weixin_39788969的博客
11-25 445
tf.keras 搭建神经网络keras介绍tf.keras 是 tensorflow2 引入的高封装度的框架， 可以用于快速搭建神经网络模型， keras 为支持快速实验而生，能够把想法迅速转换为结果， 是深度学习框架之中最终易上手的一个，它提供了一致而简洁的 API，能够极大地减少一般应用下的工作量，提高代码地封装程度和复用性。搭建神经网络六步法第一步：import 相关模块，如 i...
Tensorflow 网络基本结构
最新发布
zzj的博客
12-11 1086
Tensorflow 网络基本结构
tensorflow打印模型结构_深度学习入门 | 第三章：神经网络的TensorFlow实现（2）
weixin_39968995的博客
11-27 944
点击关注了解更多精彩内容！！3.3线性回归模型的Tensorflow实现深度学习是一种特殊的非线性回归分析方法，之所以“特殊”，是因为它常常涉及非结构化数据，如图像、声音、语言。本节首先复习线性回归的基础知识，然后以图像作为非结构化的变量，以连续型因变量作为，在TensorFlow的框架下建立到的线性回归模型。最后通过一个美食评分案例介绍代码实现的步骤。3.3.1 线性回归模型线性回归(...
tensorflow打印模型图_从Tensorflow模型文件中解析并显示网络结构图（CKPT模型篇）...
weixin_39926540的博客
12-20 1301
最近看到一个巨牛的人工智能教程，分享一下给大家。教程不仅是零基础，通俗易懂，而且非常风趣幽默，像看小说一样！觉得太牛了，所以分享给大家。平时碎片时间可以当小说看，【点这里可以去膜拜一下大神的“小说”】。上一篇文章《从Tensorflow模型文件中解析并显示网络结构图(pb模型篇)》中介绍了如何从pb模型文件中提取网络结构图并实现可视化,本文介绍如何从CKPT模型文件中提取网络结构图并实现可视化。理...
tensorflow打印模型结构_Tensorflow上手1: Print与py_func
weixin_39915605的博客
12-10 1548
这个公众号终于又要开始更新了，这几个月每天都在研究Tensorflow，虽然之前工作中也用它训练了不少模型，可是大部分模型都非常简单．最近接触了更多的深度学习模型后，对Tensorflow有了一些新的体会，于是打算写一系列相关文章，记录自己的学习心得．目前计划有简单的函数介绍，关于Keras的使用感触，分布式训练与优化，以及CUDA编程和Tensorflow的自定义函数．今天就先从简单的...
tensorflow打印出模型中任意特征层（三）（tensorflow模型可视化）
qq_38939498的博客
03-07 1579
tensorflow打印出模型中任意特征层（三）（tensorflow模型可视化） 1、前面二个内容中将了使用model.add()和class()搭建的网络，将其模型的任何一层的特征给打印出来，那么这次的内容将介绍如何使用tensorflow的高级API搭建的模型，将模型的特征打印出来。 如果你是使用这种API(如下图)搭建的API，那么前两次的内容不适用与本次API搭建的可视化。 x = Conv1D(filters=None,kernel_size=None,strides=None,padding=
从零使用TensorFlow搭建CNN（卷积）神经网络
hyk的博客
10-17 8233
本节内容主要向大家介绍如何使用TensorFlow快速搭建自己的卷积神经网络，并通过cifar数据集训练验证。文章最后会有相关内容知识点的补给。
TensorFlow实战：单变量线性回归与人工数据集
"本教程主要介绍了如何使用TensorFlow在Jupyter环境中进行单变量线性回归的实战，通过生成人工数据集并训练模型来求解线性回归的权重。 首先，线性回归是一种基本的统计和机器学习方法，其目标是找到一个最佳的直线...
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值