成功把ckpt模型文件转换为pb文件

最新推荐文章于 2022-03-04 08:48:49 发布
最新推荐文章于 2022-03-04 08:48:49 发布
阅读量459 收藏
点赞数
分类专栏： 深度学习 tensorflow python
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/qq_30336973/article/details/110911146
版权
tensorflow 同时被 3 个专栏收录
24 篇文章 0 订阅
订阅专栏
python
17 篇文章 1 订阅
订阅专栏
深度学习
13 篇文章 0 订阅
订阅专栏

这里放出成功后的打印结果，如想了解转换具体步骤，查看篇博客

/home/gqw/anaconda3/envs/tensorflow/bin/python /home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/models-master/research/object_detection/export_inference_graph.py
/home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/absl/flags/_validators.py:356: UserWarning: Flag --pipeline_config_path has a non-None default value; therefore, mark_flag_as_required will pass even if flag is not specified in the command line!
  'command line!' % flag_name)
/home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/absl/flags/_validators.py:356: UserWarning: Flag --trained_checkpoint_prefix has a non-None default value; therefore, mark_flag_as_required will pass even if flag is not specified in the command line!
  'command line!' % flag_name)
/home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/absl/flags/_validators.py:356: UserWarning: Flag --output_directory has a non-None default value; therefore, mark_flag_as_required will pass even if flag is not specified in the command line!
  'command line!' % flag_name)
WARNING:tensorflow:From /home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/models-master/research/object_detection/exporter.py:330: get_or_create_global_step (from tensorflow.contrib.framework.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.train.get_or_create_global_step
WARNING:tensorflow:From /home/gqw/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/models-master/research/object_detection/exporter.py:484: print_model_analysis (from tensorflow.contrib.tfprof.model_analyzer) is deprecated and will be removed after 2018-01-01.
Instructions for updating:
Use `tf.profiler.profile(graph, run_meta, op_log, cmd, options)`. Build `options` with `tf.profiler.ProfileOptionBuilder`. See README.md for details
198 ops no flops stats due to incomplete shapes.
Parsing Inputs...
Incomplete shape.

=========================Options=============================
-max_depth                  10000
-min_bytes                  0
-min_peak_bytes             0
-min_residual_bytes         0
-min_output_bytes           0
-min_micros                 0
-min_accelerator_micros     0
-min_cpu_micros             0
-min_params                 0
-min_float_ops              0
-min_occurrence             0
-step                       -1
-order_by                   name
-account_type_regexes       _trainable_variables
-start_name_regexes         .*
-trim_name_regexes          .*BatchNorm.*
-show_name_regexes          .*
-hide_name_regexes          
-account_displayed_op_only  true
-select                     params
-output                     stdout:

==================Model Analysis Report======================
Incomplete shape.

Doc:
scope: The nodes in the model graph are organized by their names, which is hierarchical like filesystem.
param: Number of parameters (in the Variable).

Profile:
node name | # parameters
_TFProfRoot (--/31.53m params)
  FeatureExtractor (--/26.73m params)
    FeatureExtractor/resnet_v1_50 (--/26.73m params)
      FeatureExtractor/resnet_v1_50/block1 (--/212.99k params)
        FeatureExtractor/resnet_v1_50/block1/unit_1 (--/73.73k params)
          FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1 (--/73.73k params)
            FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv1 (--/4.10k params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv1/weights (1x1x64x64, 4.10k/4.10k params)
            FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv2 (--/36.86k params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv2/weights (3x3x64x64, 36.86k/36.86k params)
            FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv3 (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/conv3/weights (1x1x64x256, 16.38k/16.38k params)
            FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/shortcut (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/shortcut/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_1/bottleneck_v1/shortcut/weights (1x1x64x256, 16.38k/16.38k params)
        FeatureExtractor/resnet_v1_50/block1/unit_2 (--/69.63k params)
          FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1 (--/69.63k params)
            FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv1 (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv1/weights (1x1x256x64, 16.38k/16.38k params)
            FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv2 (--/36.86k params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv2/weights (3x3x64x64, 36.86k/36.86k params)
            FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv3 (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_2/bottleneck_v1/conv3/weights (1x1x64x256, 16.38k/16.38k params)
        FeatureExtractor/resnet_v1_50/block1/unit_3 (--/69.63k params)
          FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1 (--/69.63k params)
            FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv1 (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv1/weights (1x1x256x64, 16.38k/16.38k params)
            FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv2 (--/36.86k params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv2/weights (3x3x64x64, 36.86k/36.86k params)
            FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv3 (--/16.38k params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block1/unit_3/bottleneck_v1/conv3/weights (1x1x64x256, 16.38k/16.38k params)
      FeatureExtractor/resnet_v1_50/block2 (--/1.21m params)
        FeatureExtractor/resnet_v1_50/block2/unit_1 (--/376.83k params)
          FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1 (--/376.83k params)
            FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv1 (--/32.77k params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv1/weights (1x1x256x128, 32.77k/32.77k params)
            FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv2 (--/147.46k params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv2/weights (3x3x128x128, 147.46k/147.46k params)
            FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv3 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/conv3/weights (1x1x128x512, 65.54k/65.54k params)
            FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/shortcut (--/131.07k params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/shortcut/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_1/bottleneck_v1/shortcut/weights (1x1x256x512, 131.07k/131.07k params)
        FeatureExtractor/resnet_v1_50/block2/unit_2 (--/278.53k params)
          FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1 (--/278.53k params)
            FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv1 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv1/weights (1x1x512x128, 65.54k/65.54k params)
            FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv2 (--/147.46k params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv2/weights (3x3x128x128, 147.46k/147.46k params)
            FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv3 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_2/bottleneck_v1/conv3/weights (1x1x128x512, 65.54k/65.54k params)
        FeatureExtractor/resnet_v1_50/block2/unit_3 (--/278.53k params)
          FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1 (--/278.53k params)
            FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv1 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv1/weights (1x1x512x128, 65.54k/65.54k params)
            FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv2 (--/147.46k params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv2/weights (3x3x128x128, 147.46k/147.46k params)
            FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv3 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_3/bottleneck_v1/conv3/weights (1x1x128x512, 65.54k/65.54k params)
        FeatureExtractor/resnet_v1_50/block2/unit_4 (--/278.53k params)
          FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1 (--/278.53k params)
            FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv1 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv1/weights (1x1x512x128, 65.54k/65.54k params)
            FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv2 (--/147.46k params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv2/weights (3x3x128x128, 147.46k/147.46k params)
            FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv3 (--/65.54k params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block2/unit_4/bottleneck_v1/conv3/weights (1x1x128x512, 65.54k/65.54k params)
      FeatureExtractor/resnet_v1_50/block3 (--/7.08m params)
        FeatureExtractor/resnet_v1_50/block3/unit_1 (--/1.51m params)
          FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1 (--/1.51m params)
            FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv1 (--/131.07k params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv1/weights (1x1x512x256, 131.07k/131.07k params)
            FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/shortcut (--/524.29k params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/shortcut/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_1/bottleneck_v1/shortcut/weights (1x1x512x1024, 524.29k/524.29k params)
        FeatureExtractor/resnet_v1_50/block3/unit_2 (--/1.11m params)
          FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1 (--/1.11m params)
            FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv1 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv1/weights (1x1x1024x256, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_2/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
        FeatureExtractor/resnet_v1_50/block3/unit_3 (--/1.11m params)
          FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1 (--/1.11m params)
            FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv1 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv1/weights (1x1x1024x256, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_3/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
        FeatureExtractor/resnet_v1_50/block3/unit_4 (--/1.11m params)
          FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1 (--/1.11m params)
            FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv1 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv1/weights (1x1x1024x256, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_4/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
        FeatureExtractor/resnet_v1_50/block3/unit_5 (--/1.11m params)
          FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1 (--/1.11m params)
            FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv1 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv1/weights (1x1x1024x256, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_5/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
        FeatureExtractor/resnet_v1_50/block3/unit_6 (--/1.11m params)
          FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1 (--/1.11m params)
            FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv1 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv1/weights (1x1x1024x256, 262.14k/262.14k params)
            FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv2 (--/589.82k params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv2/weights (3x3x256x256, 589.82k/589.82k params)
            FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv3 (--/262.14k params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block3/unit_6/bottleneck_v1/conv3/weights (1x1x256x1024, 262.14k/262.14k params)
      FeatureExtractor/resnet_v1_50/block4 (--/14.94m params)
        FeatureExtractor/resnet_v1_50/block4/unit_1 (--/6.03m params)
          FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1 (--/6.03m params)
            FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv1 (--/524.29k params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv1/weights (1x1x1024x512, 524.29k/524.29k params)
            FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv2 (--/2.36m params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv2/weights (3x3x512x512, 2.36m/2.36m params)
            FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv3 (--/1.05m params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/conv3/weights (1x1x512x2048, 1.05m/1.05m params)
            FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/shortcut (--/2.10m params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/shortcut/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_1/bottleneck_v1/shortcut/weights (1x1x1024x2048, 2.10m/2.10m params)
        FeatureExtractor/resnet_v1_50/block4/unit_2 (--/4.46m params)
          FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1 (--/4.46m params)
            FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv1 (--/1.05m params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv1/weights (1x1x2048x512, 1.05m/1.05m params)
            FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv2 (--/2.36m params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv2/weights (3x3x512x512, 2.36m/2.36m params)
            FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv3 (--/1.05m params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_2/bottleneck_v1/conv3/weights (1x1x512x2048, 1.05m/1.05m params)
        FeatureExtractor/resnet_v1_50/block4/unit_3 (--/4.46m params)
          FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1 (--/4.46m params)
            FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv1 (--/1.05m params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv1/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv1/weights (1x1x2048x512, 1.05m/1.05m params)
            FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv2 (--/2.36m params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv2/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv2/weights (3x3x512x512, 2.36m/2.36m params)
            FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv3 (--/1.05m params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv3/BatchNorm (--/0 params)
              FeatureExtractor/resnet_v1_50/block4/unit_3/bottleneck_v1/conv3/weights (1x1x512x2048, 1.05m/1.05m params)
      FeatureExtractor/resnet_v1_50/conv1 (--/9.41k params)
        FeatureExtractor/resnet_v1_50/conv1/BatchNorm (--/0 params)
        FeatureExtractor/resnet_v1_50/conv1/weights (7x7x3x64, 9.41k/9.41k params)
      FeatureExtractor/resnet_v1_50/fpn (--/3.28m params)
        FeatureExtractor/resnet_v1_50/fpn/bottom_up_block5 (--/589.82k params)
          FeatureExtractor/resnet_v1_50/fpn/bottom_up_block5/BatchNorm (--/0 params)
          FeatureExtractor/resnet_v1_50/fpn/bottom_up_block5/weights (3x3x256x256, 589.82k/589.82k params)
        FeatureExtractor/resnet_v1_50/fpn/bottom_up_block6 (--/589.82k params)
          FeatureExtractor/resnet_v1_50/fpn/bottom_up_block6/BatchNorm (--/0 params)
          FeatureExtractor/resnet_v1_50/fpn/bottom_up_block6/weights (3x3x256x256, 589.82k/589.82k params)
        FeatureExtractor/resnet_v1_50/fpn/projection_1 (--/131.33k params)
          FeatureExtractor/resnet_v1_50/fpn/projection_1/biases (256, 256/256 params)
          FeatureExtractor/resnet_v1_50/fpn/projection_1/weights (1x1x512x256, 131.07k/131.07k params)
        FeatureExtractor/resnet_v1_50/fpn/projection_2 (--/262.40k params)
          FeatureExtractor/resnet_v1_50/fpn/projection_2/biases (256, 256/256 params)
          FeatureExtractor/resnet_v1_50/fpn/projection_2/weights (1x1x1024x256, 262.14k/262.14k params)
        FeatureExtractor/resnet_v1_50/fpn/projection_3 (--/524.54k params)
          FeatureExtractor/resnet_v1_50/fpn/projection_3/biases (256, 256/256 params)
          FeatureExtractor/resnet_v1_50/fpn/projection_3/weights (1x1x2048x256, 524.29k/524.29k params)
        FeatureExtractor/resnet_v1_50/fpn/smoothing_1 (--/589.82k params)
          FeatureExtractor/resnet_v1_50/fpn/smoothing_1/BatchNorm (--/0 params)
          FeatureExtractor/resnet_v1_50/fpn/smoothing_1/weights (3x3x256x256, 589.82k/589.82k params)
        FeatureExtractor/resnet_v1_50/fpn/smoothing_2 (--/589.82k params)
          FeatureExtractor/resnet_v1_50/fpn/smoothing_2/BatchNorm (--/0 params)
          FeatureExtractor/resnet_v1_50/fpn/smoothing_2/weights (3x3x256x256, 589.82k/589.82k params)
  WeightSharedConvolutionalBoxPredictor (--/4.80m params)
    WeightSharedConvolutionalBoxPredictor/BoxPredictionTower (--/2.36m params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_0 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_0/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_0/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_1 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_1/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_1/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_2 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_2/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_2/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_3 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_3/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/BoxPredictionTower/conv2d_3/weights (3x3x256x256, 589.82k/589.82k params)
    WeightSharedConvolutionalBoxPredictor/BoxPredictor (--/55.32k params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictor/biases (24, 24/24 params)
      WeightSharedConvolutionalBoxPredictor/BoxPredictor/weights (3x3x256x24, 55.30k/55.30k params)
    WeightSharedConvolutionalBoxPredictor/ClassPredictionTower (--/2.36m params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_0 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_0/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_0/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_1 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_1/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_1/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_2 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_2/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_2/weights (3x3x256x256, 589.82k/589.82k params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_3 (--/589.82k params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_3/BatchNorm (--/0 params)
        WeightSharedConvolutionalBoxPredictor/ClassPredictionTower/conv2d_3/weights (3x3x256x256, 589.82k/589.82k params)
    WeightSharedConvolutionalBoxPredictor/ClassPredictor (--/27.66k params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictor/biases (12, 12/12 params)
      WeightSharedConvolutionalBoxPredictor/ClassPredictor/weights (3x3x256x12, 27.65k/27.65k params)

======================End of Report==========================
198 ops no flops stats due to incomplete shapes.
Parsing Inputs...
Incomplete shape.
Incomplete shape.

=========================Options=============================
-max_depth                  10000
-min_bytes                  0
-min_peak_bytes             0
-min_residual_bytes         0
-min_output_bytes           0
-min_micros                 0
-min_accelerator_micros     0
-min_cpu_micros             0
-min_params                 0
-min_float_ops              1
-min_occurrence             0
-step                       -1
-order_by                   float_ops
-account_type_regexes       .*
-start_name_regexes         .*
-trim_name_regexes          .*BatchNorm.*,.*Initializer.*,.*Regularizer.*,.*BiasAdd.*
-show_name_regexes          .*
-hide_name_regexes          
-account_displayed_op_only  true
-select                     float_ops
-output                     stdout:

==================Model Analysis Report======================

Doc:
scope: The nodes in the model graph are organized by their names, which is hierarchical like filesystem.
flops: Number of float operations. Note: Please read the implementation for the math behind it.

Profile:
node name | # float_ops
_TFProfRoot (--/614.61k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/add_2 (76.80k/76.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/sub (76.80k/76.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_5 (76.80k/76.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_6 (76.80k/76.80k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/Scale/mul_3 (38.40k/38.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/Scale/mul_2 (38.40k/38.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/Scale/mul_1 (38.40k/38.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/Scale/mul (38.40k/38.40k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_5 (19.20k/19.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_6 (19.20k/19.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/sub (19.20k/19.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/add_2 (19.20k/19.20k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/Scale/mul_3 (9.60k/9.60k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/Scale/mul_2 (9.60k/9.60k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/Scale/mul_1 (9.60k/9.60k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/Scale/mul (9.60k/9.60k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/sub (4.80k/4.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_6 (4.80k/4.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/add_2 (4.80k/4.80k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_5 (4.80k/4.80k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/Scale/mul_1 (2.40k/2.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/Scale/mul_2 (2.40k/2.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/Scale/mul_3 (2.40k/2.40k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/Scale/mul (2.40k/2.40k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/add_2 (1.20k/1.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/sub (1.20k/1.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_6 (1.20k/1.20k flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_5 (1.20k/1.20k flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/Scale/mul_2 (600/600 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/Scale/mul_1 (600/600 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/Scale/mul (600/600 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/Scale/mul_3 (600/600 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/sub (300/300 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_6 (300/300 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_5 (300/300 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/add_2 (300/300 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/Scale/mul (150/150 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/Scale/mul_3 (150/150 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/Scale/mul_2 (150/150 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/Scale/mul_1 (150/150 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_4 (80/80 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/add_1 (80/80 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_3 (80/80 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/add (80/80 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/add (40/40 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_3 (40/40 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_4 (40/40 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/add_1 (40/40 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_4 (20/20 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_3 (20/20 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/add_1 (20/20 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/add (20/20 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/add_1 (10/10 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/add (10/10 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_3 (10/10 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_4 (10/10 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_2 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_1 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_1 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/mul_2 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/truediv (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_1 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/truediv (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul_2 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_1 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/mul (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/truediv (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/truediv (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_2 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul_1 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/mul (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/truediv (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/mul_2 (6/6 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/add_1 (5/5 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_4 (5/5 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/mul_3 (5/5 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/add (5/5 flops)
  Postprocessor/Decode/get_center_coordinates_and_sizes/transpose/sub_1 (2/2 flops)
  Postprocessor/Decode/transpose_1/sub_1 (2/2 flops)
  Postprocessor/Decode/transpose/sub_1 (2/2 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_3 (1/1 flops)
  Preprocessor/map/while/Less (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_2 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_13 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_12 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_11 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_10 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_6 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_5 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_4 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_3 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_2 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/Greater (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_4 (1/1 flops)
  Preprocessor/map/while/add (1/1 flops)
  Preprocessor/map/while/Less_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/add_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_5 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_6 (1/1 flops)
  Postprocessor/Decode/transpose_1/sub (1/1 flops)
  Postprocessor/Decode/transpose/sub (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_7 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_8 (1/1 flops)
  Preprocessor/map/while/add_1 (1/1 flops)
  Postprocessor/Decode/get_center_coordinates_and_sizes/transpose/sub (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/PadOrClipBoxList/sub_9 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/add (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/ones/Less (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/truediv (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/truediv_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_2/truediv (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/truediv_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_1/truediv (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/truediv_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates/truediv (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_4/assert_equal/Equal (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_3/truediv_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/assert_equal/Equal (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_3/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/assert_equal/Equal (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_2/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/assert_equal/Equal (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator_1/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/assert_equal/Equal (1/1 flops)
  MultiscaleGridAnchorGenerator/GridAnchorGenerator/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/assert_equal/Equal (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/add (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/SortByField_1/Equal (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/SortByField/Equal (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/Minimum_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/Minimum (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/Greater (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/Less_1 (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/Less (1/1 flops)
  Postprocessor/BatchMultiClassNonMaxSuppression/map/while/MultiClassNonMaxSuppression/sub (1/1 flops)
  MultiscaleGridAnchorGenerator/add_4 (1/1 flops)
  MultiscaleGridAnchorGenerator/add_3 (1/1 flops)
  MultiscaleGridAnchorGenerator/add_2 (1/1 flops)
  MultiscaleGridAnchorGenerator/add_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/add (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/truediv_1 (1/1 flops)
  MultiscaleGridAnchorGenerator/ToNormalizedCoordinates_4/truediv (1/1 flops)

======================End of Report==========================
2020-12-08 21:03:57.626315: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA
2020-12-08 21:03:59.294301: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:897] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-12-08 21:03:59.294886: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1392] Found device 0 with properties: 
name: GeForce GTX 1050 Ti major: 6 minor: 1 memoryClockRate(GHz): 1.62
pciBusID: 0000:01:00.0
totalMemory: 3.95GiB freeMemory: 3.42GiB
2020-12-08 21:03:59.294930: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1471] Adding visible gpu devices: 0
2020-12-08 21:04:06.910855: I tensorflow/core/common_runtime/gpu/gpu_device.cc:952] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-12-08 21:04:06.910928: I tensorflow/core/common_runtime/gpu/gpu_device.cc:958]      0 
2020-12-08 21:04:06.910944: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0:   N 
2020-12-08 21:04:06.911281: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1084] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3145 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1)
2020-12-08 21:04:11.344492: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1471] Adding visible gpu devices: 0
2020-12-08 21:04:11.344600: I tensorflow/core/common_runtime/gpu/gpu_device.cc:952] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-12-08 21:04:11.344622: I tensorflow/core/common_runtime/gpu/gpu_device.cc:958]      0 
2020-12-08 21:04:11.344628: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0:   N 
2020-12-08 21:04:11.344720: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1084] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3145 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1)
2020-12-08 21:04:16.440743: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1471] Adding visible gpu devices: 0
2020-12-08 21:04:16.440810: I tensorflow/core/common_runtime/gpu/gpu_device.cc:952] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-12-08 21:04:16.440825: I tensorflow/core/common_runtime/gpu/gpu_device.cc:958]      0 
2020-12-08 21:04:16.440835: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0:   N 
2020-12-08 21:04:16.440993: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1084] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3145 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1)

Process finished with exit code 0
泡泡里的月亮
关注
专栏目录
tensorflow的ckpt文件转pb模型文件
10-28
使用这个python程序可以直接将ckpt文件转成固化的模型文件，用于预测结果，注意修改路径
TensorFlow把ckpt转化为pb文件代码
12-13
把tensorflow保存的checkpoint类型模型冻结，转化为.pb模型输出。调用方法参见我的博文https://blog.csdn.net/weixin_41864878/article/details/84957681
tensorflow模型文件(ckpt)转pb文件的方法（不知道输出节点名）
12-20
网上关于tensorflow模型文件ckpt格式转pb文件的帖子很多，本人几乎尝试了所有方法，最后终于成功了，现总结如下。方法无外乎下面两种： 使用tensorflow.python.tools.freeze_graph.freeze_graph 使用graph_util.convert_variables_to_constants 1、tensorflow模型的文件解读 使用tensorflow训练好的模型会自动保存为四个文件，如下 checkpoint：记录近几次训练好的模型结果（名称）。 xxx.data-00000-of-00001: 模型的所有变量的值(weights, b
ckpt 转 pb
salewang的博客
03-04 1513
正常保存的ckpt 有5个文件 第三个是权重（其他的去网上查下） 这是我的环境 absl-py (1.0.0) astor (0.8.1) certifi (2016.2.28) gast (0.5.3) grpcio (1.12.1) h5py (2.7.0) Keras-Applications (1.0.8) Keras-Preprocessing (1.1.2) Markdown (2.6.9) numpy (1.13.3) pip (21.3.1) protobuf (3.7.1.
ckpt转pb
chenaxin的博客
09-10 489
有时需要将tensorflow训练得到的ckpt固化成pb使用。Tensorflow提供了相关的固化命令脚本，下面是我搜集到的转换步骤。参考https://www.yanxishe.com/columnDetail/15278 一、tensorflow保存下的内容 用 tf.train.Saver.save() 方式保存下来的checkpoint会产生四个文件： checkpoint 记录了部分已存储和最近存储的模型： model_checkpoint_path: “mtcnn-3000000” all_
TensorFlow把ckpt转化为pb文件代码进阶版
12-13
把tensorflow保存的checkpoint类型模型冻结，转化为.pb模型输出。调用方法参见我的博文https://blog.csdn.net/weixin_41864878/article/details/84957681
基于python将tensorflow的ckpt模型转化为pb模型项目源码
03-24
本项目源码将介绍如何使用Python将TensorFlow的CKPT模型转换为PB模型。 首先，我们需要理解CKPT文件。CKPT文件由三部分组成：`.index`、`.meta`和`.data-xxxxxx-of-xxxxxx`。`.index`文件记录了权重数据的位置信息...
TensorFlow：详解如何将ckpt模型转换为pb文件
"本文主要探讨如何使用TensorFlow将ckpt模型转换为.pb文件，这是一个常见的需求，特别是当需要在不同环境中部署模型或者简化模型文件管理时。.pb文件是一种包含了模型结构和权重数据的二进制格式，易于在不支持...
深度学习之格式转换笔记(二)：CKPT 转换成 PB格式文件
源的博客
01-22 1666
我们使用tf.train.saver()保存模型时会产生多个文件，也就是说把计算图的结构和图上参数取值分成了不同的文件存储。这也是在tensorflow中常用的保存方式。 保存文件的代码: import tensorflow as tf # 声明两个变量 v1 = tf.Variable(tf.random_normal([1, 2]), name="v1") v2 = tf.Variable(tf.random_normal([2, 3]), name="v2") init_op = tf.global_
ckpt转pb的代码 pb2.py 适用于Semantic-Segmentation-Suite
08-15
Semantic-Segmentation-Suite 的ckpt转 pb代码，根据最后一层更改。
tensorflow-ckpt转pb
秀呀秀的博客
09-24 264
from __future__ import division, print_function import os import sys import tensorflow as tf import numpy as np from utils.nms_utils import gpu_nms from model import yolov3 from utils.misc_utils import parse_anchors, load_weights from utils.data_aug impo.
.ckpt模型转.pb模型（查找ckpt节点名称）-Tensorflow
whocare的博客
03-29 944
ckpt模型和pb模型介绍： ckpt： 1.这种模型文件是依赖 TensorFlow 的，只能在其框架下使用； 2.在恢复模型之前还需要再定义一遍网络结构，然后才能把变量的值恢复到网络中。 pb： 1.谷歌推荐的保存模型的方式是保存模型为 PB 文件，它具有语言独立性，可独立运行，封闭的序列化格式，任何语言都可以解析它，它允许其他语言和深度学习框架读取、继续训练和迁移 TensorFlow 的模型； 2. 保存为 PB 文件时候，模型的变量都会变成固定的，导致模型的大小会大大减小，适合在手机端运行。 所以
Tensorflow .ckpt模型转换为.pb模型
Devin的博客
09-28 265
https://blog.csdn.net/xichen9581/article/details/101004610
C++学习之路 | PTA乙级—— 1006 换个格式输出整数 (15分)（精简）
Eighteen
09-22 200
1006 换个格式输出整数 (15分) 让我们用字母 B 来表示“百”、字母 S 表示“十”，用 12…n 来表示不为零的个位数字 n（<10），换个格式来输出任一个不超过 3 位的正整数。例如 234 应该被输出为 BBSSS1234，因为它有 2 个“百”、3 个“十”、以及个位的 4。 输入格式： 每个测试输入包含 1 个测试用例，给出正整数 n（<1000）。 输出格式： 每个测试用例的输出占一行，用规定的格式输出 n。 输入样例 1： 234 输出样例 1： BBSSS1234 输入样
tensorflow ckpt转化到pb
m0_37400316的博客
12-10 116
import tensorflow as tf from tensorflow.python.framework import graph_util def ckpt2pb(): with tf.Graph().as_default() as graph_old: isess = tf.InteractiveSession() ckpt_filename = './yolov3.ckpt' isess.run(tf.global_variable.
【Tensorflow】.ckpt模型转换为.pb模型（BERT）
xichen9581的博客
09-19 9728
目录 1. 模型文件转换 1.1 获取模型中节点名称 1.2 将ckpt模型转换为pb模型 1.3 查看.ckpt文件保存的tensor信息 2. 模型文件可视化 2.1 ckpt模型可视化 2.2 pb模型可视化 Google_BERT 1. 模型文件转换 .indel是对应模型的索引文件，保存.data文件与.meta文件中图的结构关系； .data-0000...
tensorflow | tensorflow实现将ckpt转pb文件
东南风的博客
01-06 6736
   本博客实现将自己训练保存的ckpt模型转换为pb文件，该方法适用于任何ckpt模型，当然你需要确定ckpt模型输入/输出的节点名称。 目录 tensorflow实现将ckpt转pb文件 一、CKPT 转换成 PB格式 二、 pb模型预测 三、源码下载和资料推荐     1、训练方法     2、本博客Github地址     3、将模型移植Android的方法    使用 ...
[Tensorflow]模型持久化的原理，将CKPT转为pb文件，使用pb模型预测
yegeli的博客
12-02 1201
tensorflow：模型持久化的原理，将CKPT转为pb文件，使用pb模型预测 ​ 由题目就可以看出，本节内容分为三部分，第一部分就是如何将训练好的模型持久化，并学习模型持久化的原理，第二部分就是如何将CKPT转化为pb文件，第三部分就是如何使用pb模型进行预测。 ​ 这里新增一个h5转tflite，h5转pb，pb转 tflite的文件代码，代码直接展示，不写什么了，我感觉其实也没有必要写什么了，该说的都说了，只不过h5模型是Keras中训练的。 一、模型持久化 ​ 为了让训练得到的模型保存下来方便下次
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