Does keras.layers.ZeroPadding2D is supported by TensorRT currently?

1036758468 · June 30, 2021, 9:37am

Description

When I try to convert onnx (converted from keras) to tensorrt engine, this error occurs.

IndexError: Attribute not found: pads.
I found my keras model contains keras.layers.ZeroPadding2D layers (keras version is 2.2.4). Does keras.layers.ZeroPadding2D is supported by TensorRT?

Environment

TensorRT Version : 7.2.34.
NVIDIA GPU : Tesla P100
NVIDIA Driver Version : 440.33.01
CUDA Version : 10.2
CUDNN Version : 7.6.5
Operating System : CentOS
Python Version (if applicable) : 3.6
Tensorflow Version (if applicable) : 1.14.0

spolisetty · June 30, 2021, 10:03am

Hi @1036758468,

Yes it’s supported. Please refer onnx-tensorrt/operators.md at master · onnx/onnx-tensorrt · GitHub to check ONNX - TRT supported operators.

Please let us know which version of Opset are you using. We recommend you to please try on latest TensorRT version 8.0 EA. If you still face this issue, please share verbose logs and ONNX model to try from our end for better assistance.

Thank you.

NVES · June 30, 2021, 10:08am

Hi,
Request you to share the ONNX model and the script if not shared already so that we can assist you better.
Alongside you can try few things:

validating your model with the below snippet

check_model.py

import sys
import onnx
filename = yourONNXmodel
model = onnx.load(filename)
onnx.checker.check_model(model).
2) Try running your model with trtexec command.

In case you are still facing issue, request you to share the trtexec “”–verbose"" log for further debugging
Thanks!

1036758468 · June 30, 2021, 10:33am

Hi,

The maximum opset needed by this model is only 11. The verbose log of keras to onnx is as follows:

Processing a keras layer - (conv_81: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_80/LeakyRelu:0
output: conv_81/BiasAdd:0
Processing a keras layer - (conv_93: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_92/LeakyRelu:0
output: conv_93/BiasAdd:0
Processing a keras layer - (conv_105: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_104/LeakyRelu:0
output: conv_105/BiasAdd:0
Processing a keras layer - (leaky_80: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_80/cond/Merge:0
output: leaky_80/LeakyRelu:0
Processing a keras layer - (leaky_92: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_92/cond/Merge:0
output: leaky_92/LeakyRelu:0
Processing a keras layer - (leaky_104: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_104/cond/Merge:0
output: leaky_104/LeakyRelu:0
Processing a keras layer - (bnorm_80: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_80/convolution:0
output: bnorm_80/cond/Merge:0
Processing a keras layer - (bnorm_92: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_92/convolution:0
output: bnorm_92/cond/Merge:0
Processing a keras layer - (bnorm_104: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_104/convolution:0
output: bnorm_104/cond/Merge:0
Processing a keras layer - (conv_80: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_79/LeakyRelu:0
output: conv_80/convolution:0
Processing a keras layer - (conv_92: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_91/LeakyRelu:0
output: conv_92/convolution:0
Processing a keras layer - (conv_104: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_103/LeakyRelu:0
output: conv_104/convolution:0
Processing a keras layer - (leaky_79: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_79/cond/Merge:0
output: leaky_79/LeakyRelu:0
Processing a keras layer - (leaky_91: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_91/cond/Merge:0
output: leaky_91/LeakyRelu:0
Processing a keras layer - (leaky_103: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_103/cond/Merge:0
output: leaky_103/LeakyRelu:0
Processing a keras layer - (bnorm_79: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_79/convolution:0
output: bnorm_79/cond/Merge:0
Processing a keras layer - (bnorm_91: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_91/convolution:0
output: bnorm_91/cond/Merge:0
Processing a keras layer - (bnorm_103: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_103/convolution:0
output: bnorm_103/cond/Merge:0
Processing a keras layer - (conv_79: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_78/LeakyRelu:0
output: conv_79/convolution:0
Processing a keras layer - (conv_91: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_90/LeakyRelu:0
output: conv_91/convolution:0
Processing a keras layer - (conv_103: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_102/LeakyRelu:0
output: conv_103/convolution:0
Processing a keras layer - (leaky_78: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_78/cond/Merge:0
output: leaky_78/LeakyRelu:0
Processing a keras layer - (leaky_90: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_90/cond/Merge:0
output: leaky_90/LeakyRelu:0
Processing a keras layer - (leaky_102: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_102/cond/Merge:0
output: leaky_102/LeakyRelu:0
Processing a keras layer - (bnorm_78: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_78/convolution:0
output: bnorm_78/cond/Merge:0
Processing a keras layer - (bnorm_90: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_90/convolution:0
output: bnorm_90/cond/Merge:0
Processing a keras layer - (bnorm_102: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_102/convolution:0
output: bnorm_102/cond/Merge:0
Processing a keras layer - (conv_78: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_77/LeakyRelu:0
output: conv_78/convolution:0
Processing a keras layer - (conv_90: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_89/LeakyRelu:0
output: conv_90/convolution:0
Processing a keras layer - (conv_102: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_101/LeakyRelu:0
output: conv_102/convolution:0
Processing a keras layer - (leaky_77: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_77/cond/Merge:0
output: leaky_77/LeakyRelu:0
Processing a keras layer - (leaky_89: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_89/cond/Merge:0
output: leaky_89/LeakyRelu:0
Processing a keras layer - (leaky_101: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_101/cond/Merge:0
output: leaky_101/LeakyRelu:0
Processing a keras layer - (bnorm_77: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_77/convolution:0
output: bnorm_77/cond/Merge:0
Processing a keras layer - (bnorm_89: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_89/convolution:0
output: bnorm_89/cond/Merge:0
Processing a keras layer - (bnorm_101: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_101/convolution:0
output: bnorm_101/cond/Merge:0
Processing a keras layer - (conv_77: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_76/LeakyRelu:0
output: conv_77/convolution:0
Processing a keras layer - (conv_89: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_88/LeakyRelu:0
output: conv_89/convolution:0
Processing a keras layer - (conv_101: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_100/LeakyRelu:0
output: conv_101/convolution:0
Processing a keras layer - (leaky_76: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_76/cond/Merge:0
output: leaky_76/LeakyRelu:0
Processing a keras layer - (leaky_88: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_88/cond/Merge:0
output: leaky_88/LeakyRelu:0
Processing a keras layer - (leaky_100: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_100/cond/Merge:0
output: leaky_100/LeakyRelu:0
Processing a keras layer - (bnorm_76: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_76/convolution:0
output: bnorm_76/cond/Merge:0
Processing a keras layer - (bnorm_88: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_88/convolution:0
output: bnorm_88/cond/Merge:0
Processing a keras layer - (bnorm_100: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_100/convolution:0
output: bnorm_100/cond/Merge:0
Processing a keras layer - (conv_76: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_75/LeakyRelu:0
output: conv_76/convolution:0
Processing a keras layer - (conv_88: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_87/LeakyRelu:0
output: conv_88/convolution:0
Processing a keras layer - (conv_100: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_99/LeakyRelu:0
output: conv_100/convolution:0
Processing a keras layer - (leaky_75: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_75/cond/Merge:0
output: leaky_75/LeakyRelu:0
Processing a keras layer - (leaky_87: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_87/cond/Merge:0
output: leaky_87/LeakyRelu:0
Processing a keras layer - (leaky_99: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_99/cond/Merge:0
output: leaky_99/LeakyRelu:0
Processing a keras layer - (bnorm_75: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_75/convolution:0
output: bnorm_75/cond/Merge:0
Processing a keras layer - (bnorm_87: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_87/convolution:0
output: bnorm_87/cond/Merge:0
Processing a keras layer - (bnorm_99: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_99/convolution:0
output: bnorm_99/cond/Merge:0
Processing a keras layer - (conv_75: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_23/add:0
output: conv_75/convolution:0
Processing a keras layer - (conv_87: <class ‘keras.layers.convolutional.Conv2D’>)
input : concatenate_1/concat:0
output: conv_87/convolution:0
Processing a keras layer - (conv_99: <class ‘keras.layers.convolutional.Conv2D’>)
input : concatenate_2/concat:0
output: conv_99/convolution:0
Processing a keras layer - (add_23: <class ‘keras.layers.merge.Add’>)
input : add_22/add:0
input : leaky_73/LeakyRelu:0
output: add_23/add:0
Processing a keras layer - (concatenate_1: <class ‘keras.layers.merge.Concatenate’>)
input : up_sampling2d_1/ResizeNearestNeighbor:0
input : add_19/add:0
output: concatenate_1/concat:0
Processing a keras layer - (concatenate_2: <class ‘keras.layers.merge.Concatenate’>)
input : up_sampling2d_2/ResizeNearestNeighbor:0
input : add_11/add:0
output: concatenate_2/concat:0
Processing a keras layer - (add_22: <class ‘keras.layers.merge.Add’>)
input : add_21/add:0
input : leaky_70/LeakyRelu:0
output: add_22/add:0
Processing a keras layer - (leaky_73: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_73/cond/Merge:0
output: leaky_73/LeakyRelu:0
Processing a keras layer - (up_sampling2d_1: <class ‘keras.layers.convolutional.UpSampling2D’>)
input : leaky_84/LeakyRelu:0
output: up_sampling2d_1/ResizeNearestNeighbor:0
Processing a keras layer - (add_19: <class ‘keras.layers.merge.Add’>)
input : add_18/add:0
input : leaky_60/LeakyRelu:0
output: add_19/add:0
Processing a keras layer - (up_sampling2d_2: <class ‘keras.layers.convolutional.UpSampling2D’>)
input : leaky_96/LeakyRelu:0
output: up_sampling2d_2/ResizeNearestNeighbor:0
Processing a keras layer - (add_11: <class ‘keras.layers.merge.Add’>)
input : add_10/add:0
input : leaky_35/LeakyRelu:0
output: add_11/add:0
Processing a keras layer - (add_21: <class ‘keras.layers.merge.Add’>)
input : add_20/add:0
input : leaky_67/LeakyRelu:0
output: add_21/add:0
Processing a keras layer - (leaky_70: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_70/cond/Merge:0
output: leaky_70/LeakyRelu:0
Processing a keras layer - (bnorm_73: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_73/convolution:0
output: bnorm_73/cond/Merge:0
Processing a keras layer - (leaky_84: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_84/cond/Merge:0
output: leaky_84/LeakyRelu:0
Processing a keras layer - (add_18: <class ‘keras.layers.merge.Add’>)
input : add_17/add:0
input : leaky_57/LeakyRelu:0
output: add_18/add:0
Processing a keras layer - (leaky_60: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_60/cond/Merge:0
output: leaky_60/LeakyRelu:0
Processing a keras layer - (leaky_96: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_96/cond/Merge:0
output: leaky_96/LeakyRelu:0
Processing a keras layer - (add_10: <class ‘keras.layers.merge.Add’>)
input : add_9/add:0
input : leaky_32/LeakyRelu:0
output: add_10/add:0
Processing a keras layer - (leaky_35: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_35/cond/Merge:0
output: leaky_35/LeakyRelu:0
Processing a keras layer - (add_20: <class ‘keras.layers.merge.Add’>)
input : leaky_62/LeakyRelu:0
input : leaky_64/LeakyRelu:0
output: add_20/add:0
Processing a keras layer - (leaky_67: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_67/cond/Merge:0
output: leaky_67/LeakyRelu:0
Processing a keras layer - (bnorm_70: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_70/convolution:0
output: bnorm_70/cond/Merge:0
Processing a keras layer - (conv_73: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_72/LeakyRelu:0
output: conv_73/convolution:0
Processing a keras layer - (bnorm_84: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_84/convolution:0
output: bnorm_84/cond/Merge:0
Processing a keras layer - (add_17: <class ‘keras.layers.merge.Add’>)
input : add_16/add:0
input : leaky_54/LeakyRelu:0
output: add_17/add:0
Processing a keras layer - (leaky_57: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_57/cond/Merge:0
output: leaky_57/LeakyRelu:0
Processing a keras layer - (bnorm_60: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_60/convolution:0
output: bnorm_60/cond/Merge:0
Processing a keras layer - (bnorm_96: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_96/convolution:0
output: bnorm_96/cond/Merge:0
Processing a keras layer - (add_9: <class ‘keras.layers.merge.Add’>)
input : add_8/add:0
input : leaky_29/LeakyRelu:0
output: add_9/add:0
Processing a keras layer - (leaky_32: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_32/cond/Merge:0
output: leaky_32/LeakyRelu:0
Processing a keras layer - (bnorm_35: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_35/convolution:0
output: bnorm_35/cond/Merge:0
Processing a keras layer - (leaky_62: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_62/cond/Merge:0
output: leaky_62/LeakyRelu:0
Processing a keras layer - (leaky_64: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_64/cond/Merge:0
output: leaky_64/LeakyRelu:0
Processing a keras layer - (bnorm_67: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_67/convolution:0
output: bnorm_67/cond/Merge:0
Processing a keras layer - (conv_70: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_69/LeakyRelu:0
output: conv_70/convolution:0
Processing a keras layer - (leaky_72: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_72/cond/Merge:0
output: leaky_72/LeakyRelu:0
Processing a keras layer - (conv_84: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_79/LeakyRelu:0
output: conv_84/convolution:0
Processing a keras layer - (add_16: <class ‘keras.layers.merge.Add’>)
input : add_15/add:0
input : leaky_51/LeakyRelu:0
output: add_16/add:0
Processing a keras layer - (leaky_54: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_54/cond/Merge:0
output: leaky_54/LeakyRelu:0
Processing a keras layer - (bnorm_57: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_57/convolution:0
output: bnorm_57/cond/Merge:0
Processing a keras layer - (conv_60: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_59/LeakyRelu:0
output: conv_60/convolution:0
Processing a keras layer - (conv_96: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_91/LeakyRelu:0
output: conv_96/convolution:0
Processing a keras layer - (add_8: <class ‘keras.layers.merge.Add’>)
input : add_7/add:0
input : leaky_26/LeakyRelu:0
output: add_8/add:0
Processing a keras layer - (leaky_29: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_29/cond/Merge:0
output: leaky_29/LeakyRelu:0
Processing a keras layer - (bnorm_32: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_32/convolution:0
output: bnorm_32/cond/Merge:0
Processing a keras layer - (conv_35: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_34/LeakyRelu:0
output: conv_35/convolution:0
Processing a keras layer - (bnorm_62: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_62/convolution:0
output: bnorm_62/cond/Merge:0
Processing a keras layer - (bnorm_64: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_64/convolution:0
output: bnorm_64/cond/Merge:0
Processing a keras layer - (conv_67: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_66/LeakyRelu:0
output: conv_67/convolution:0
Processing a keras layer - (leaky_69: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_69/cond/Merge:0
output: leaky_69/LeakyRelu:0
Processing a keras layer - (bnorm_72: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_72/convolution:0
output: bnorm_72/cond/Merge:0
Processing a keras layer - (add_15: <class ‘keras.layers.merge.Add’>)
input : add_14/add:0
input : leaky_48/LeakyRelu:0
output: add_15/add:0
Processing a keras layer - (leaky_51: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_51/cond/Merge:0
output: leaky_51/LeakyRelu:0
Processing a keras layer - (bnorm_54: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_54/convolution:0
output: bnorm_54/cond/Merge:0
Processing a keras layer - (conv_57: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_56/LeakyRelu:0
output: conv_57/convolution:0
Processing a keras layer - (leaky_59: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_59/cond/Merge:0
output: leaky_59/LeakyRelu:0
Processing a keras layer - (add_7: <class ‘keras.layers.merge.Add’>)
input : add_6/add:0
input : leaky_23/LeakyRelu:0
output: add_7/add:0
Processing a keras layer - (leaky_26: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_26/cond/Merge:0
output: leaky_26/LeakyRelu:0
Processing a keras layer - (bnorm_29: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_29/convolution:0
output: bnorm_29/cond/Merge:0
Processing a keras layer - (conv_32: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_31/LeakyRelu:0
output: conv_32/convolution:0
Processing a keras layer - (leaky_34: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_34/cond/Merge:0
output: leaky_34/LeakyRelu:0
Processing a keras layer - (conv_62: <class ‘keras.layers.convolutional.Conv2D’>)
input : zero_padding2d_5/Pad:0
output: conv_62/convolution:0
Processing a keras layer - (conv_64: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_63/LeakyRelu:0
output: conv_64/convolution:0
Processing a keras layer - (leaky_66: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_66/cond/Merge:0
output: leaky_66/LeakyRelu:0
Processing a keras layer - (bnorm_69: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_69/convolution:0
output: bnorm_69/cond/Merge:0
Processing a keras layer - (conv_72: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_22/add:0
output: conv_72/convolution:0
Processing a keras layer - (add_14: <class ‘keras.layers.merge.Add’>)
input : add_13/add:0
input : leaky_45/LeakyRelu:0
output: add_14/add:0
Processing a keras layer - (leaky_48: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_48/cond/Merge:0
output: leaky_48/LeakyRelu:0
Processing a keras layer - (bnorm_51: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_51/convolution:0
output: bnorm_51/cond/Merge:0
Processing a keras layer - (conv_54: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_53/LeakyRelu:0
output: conv_54/convolution:0
Processing a keras layer - (leaky_56: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_56/cond/Merge:0
output: leaky_56/LeakyRelu:0
Processing a keras layer - (bnorm_59: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_59/convolution:0
output: bnorm_59/cond/Merge:0
Processing a keras layer - (add_6: <class ‘keras.layers.merge.Add’>)
input : add_5/add:0
input : leaky_20/LeakyRelu:0
output: add_6/add:0
Processing a keras layer - (leaky_23: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_23/cond/Merge:0
output: leaky_23/LeakyRelu:0
Processing a keras layer - (bnorm_26: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_26/convolution:0
output: bnorm_26/cond/Merge:0
Processing a keras layer - (conv_29: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_28/LeakyRelu:0
output: conv_29/convolution:0
Processing a keras layer - (leaky_31: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_31/cond/Merge:0
output: leaky_31/LeakyRelu:0
Processing a keras layer - (bnorm_34: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_34/convolution:0
output: bnorm_34/cond/Merge:0
Processing a keras layer - (zero_padding2d_5: <class ‘keras.layers.convolutional.ZeroPadding2D’>)
input : add_19/add:0
output: zero_padding2d_5/Pad:0
Processing a keras layer - (leaky_63: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_63/cond/Merge:0
output: leaky_63/LeakyRelu:0
Processing a keras layer - (bnorm_66: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_66/convolution:0
output: bnorm_66/cond/Merge:0
Processing a keras layer - (conv_69: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_21/add:0
output: conv_69/convolution:0
Processing a keras layer - (add_13: <class ‘keras.layers.merge.Add’>)
input : add_12/add:0
input : leaky_42/LeakyRelu:0
output: add_13/add:0
Processing a keras layer - (leaky_45: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_45/cond/Merge:0
output: leaky_45/LeakyRelu:0
Processing a keras layer - (bnorm_48: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_48/convolution:0
output: bnorm_48/cond/Merge:0
Processing a keras layer - (conv_51: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_50/LeakyRelu:0
output: conv_51/convolution:0
Processing a keras layer - (leaky_53: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_53/cond/Merge:0
output: leaky_53/LeakyRelu:0
Processing a keras layer - (bnorm_56: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_56/convolution:0
output: bnorm_56/cond/Merge:0
Processing a keras layer - (conv_59: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_18/add:0
output: conv_59/convolution:0
Processing a keras layer - (add_5: <class ‘keras.layers.merge.Add’>)
input : add_4/add:0
input : leaky_17/LeakyRelu:0
output: add_5/add:0
Processing a keras layer - (leaky_20: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_20/cond/Merge:0
output: leaky_20/LeakyRelu:0
Processing a keras layer - (bnorm_23: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_23/convolution:0
output: bnorm_23/cond/Merge:0
Processing a keras layer - (conv_26: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_25/LeakyRelu:0
output: conv_26/convolution:0
Processing a keras layer - (leaky_28: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_28/cond/Merge:0
output: leaky_28/LeakyRelu:0
Processing a keras layer - (bnorm_31: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_31/convolution:0
output: bnorm_31/cond/Merge:0
Processing a keras layer - (conv_34: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_10/add:0
output: conv_34/convolution:0
Processing a keras layer - (bnorm_63: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_63/convolution:0
output: bnorm_63/cond/Merge:0
Processing a keras layer - (conv_66: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_20/add:0
output: conv_66/convolution:0
Processing a keras layer - (add_12: <class ‘keras.layers.merge.Add’>)
input : leaky_37/LeakyRelu:0
input : leaky_39/LeakyRelu:0
output: add_12/add:0
Processing a keras layer - (leaky_42: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_42/cond/Merge:0
output: leaky_42/LeakyRelu:0
Processing a keras layer - (bnorm_45: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_45/convolution:0
output: bnorm_45/cond/Merge:0
Processing a keras layer - (conv_48: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_47/LeakyRelu:0
output: conv_48/convolution:0
Processing a keras layer - (leaky_50: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_50/cond/Merge:0
output: leaky_50/LeakyRelu:0
Processing a keras layer - (bnorm_53: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_53/convolution:0
output: bnorm_53/cond/Merge:0
Processing a keras layer - (conv_56: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_17/add:0
output: conv_56/convolution:0
Processing a keras layer - (add_4: <class ‘keras.layers.merge.Add’>)
input : leaky_12/LeakyRelu:0
input : leaky_14/LeakyRelu:0
output: add_4/add:0
Processing a keras layer - (leaky_17: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_17/cond/Merge:0
output: leaky_17/LeakyRelu:0
Processing a keras layer - (bnorm_20: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_20/convolution:0
output: bnorm_20/cond/Merge:0
Processing a keras layer - (conv_23: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_22/LeakyRelu:0
output: conv_23/convolution:0
Processing a keras layer - (leaky_25: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_25/cond/Merge:0
output: leaky_25/LeakyRelu:0
Processing a keras layer - (bnorm_28: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_28/convolution:0
output: bnorm_28/cond/Merge:0
Processing a keras layer - (conv_31: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_9/add:0
output: conv_31/convolution:0
Processing a keras layer - (conv_63: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_62/LeakyRelu:0
output: conv_63/convolution:0
Processing a keras layer - (leaky_37: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_37/cond/Merge:0
output: leaky_37/LeakyRelu:0
Processing a keras layer - (leaky_39: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_39/cond/Merge:0
output: leaky_39/LeakyRelu:0
Processing a keras layer - (bnorm_42: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_42/convolution:0
output: bnorm_42/cond/Merge:0
Processing a keras layer - (conv_45: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_44/LeakyRelu:0
output: conv_45/convolution:0
Processing a keras layer - (leaky_47: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_47/cond/Merge:0
output: leaky_47/LeakyRelu:0
Processing a keras layer - (bnorm_50: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_50/convolution:0
output: bnorm_50/cond/Merge:0
Processing a keras layer - (conv_53: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_16/add:0
output: conv_53/convolution:0
Processing a keras layer - (leaky_12: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_12/cond/Merge:0
output: leaky_12/LeakyRelu:0
Processing a keras layer - (leaky_14: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_14/cond/Merge:0
output: leaky_14/LeakyRelu:0
Processing a keras layer - (bnorm_17: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_17/convolution:0
output: bnorm_17/cond/Merge:0
Processing a keras layer - (conv_20: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_19/LeakyRelu:0
output: conv_20/convolution:0
Processing a keras layer - (leaky_22: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_22/cond/Merge:0
output: leaky_22/LeakyRelu:0
Processing a keras layer - (bnorm_25: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_25/convolution:0
output: bnorm_25/cond/Merge:0
Processing a keras layer - (conv_28: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_8/add:0
output: conv_28/convolution:0
Processing a keras layer - (bnorm_37: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_37/convolution:0
output: bnorm_37/cond/Merge:0
Processing a keras layer - (bnorm_39: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_39/convolution:0
output: bnorm_39/cond/Merge:0
Processing a keras layer - (conv_42: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_41/LeakyRelu:0
output: conv_42/convolution:0
Processing a keras layer - (leaky_44: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_44/cond/Merge:0
output: leaky_44/LeakyRelu:0
Processing a keras layer - (bnorm_47: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_47/convolution:0
output: bnorm_47/cond/Merge:0
Processing a keras layer - (conv_50: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_15/add:0
output: conv_50/convolution:0
Processing a keras layer - (bnorm_12: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_12/convolution:0
output: bnorm_12/cond/Merge:0
Processing a keras layer - (bnorm_14: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_14/convolution:0
output: bnorm_14/cond/Merge:0
Processing a keras layer - (conv_17: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_16/LeakyRelu:0
output: conv_17/convolution:0
Processing a keras layer - (leaky_19: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_19/cond/Merge:0
output: leaky_19/LeakyRelu:0
Processing a keras layer - (bnorm_22: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_22/convolution:0
output: bnorm_22/cond/Merge:0
Processing a keras layer - (conv_25: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_7/add:0
output: conv_25/convolution:0
Processing a keras layer - (conv_37: <class ‘keras.layers.convolutional.Conv2D’>)
input : zero_padding2d_4/Pad:0
output: conv_37/convolution:0
Processing a keras layer - (conv_39: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_38/LeakyRelu:0
output: conv_39/convolution:0
Processing a keras layer - (leaky_41: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_41/cond/Merge:0
output: leaky_41/LeakyRelu:0
Processing a keras layer - (bnorm_44: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_44/convolution:0
output: bnorm_44/cond/Merge:0
Processing a keras layer - (conv_47: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_14/add:0
output: conv_47/convolution:0
Processing a keras layer - (conv_12: <class ‘keras.layers.convolutional.Conv2D’>)
input : zero_padding2d_3/Pad:0
output: conv_12/convolution:0
Processing a keras layer - (conv_14: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_13/LeakyRelu:0
output: conv_14/convolution:0
Processing a keras layer - (leaky_16: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_16/cond/Merge:0
output: leaky_16/LeakyRelu:0
Processing a keras layer - (bnorm_19: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_19/convolution:0
output: bnorm_19/cond/Merge:0
Processing a keras layer - (conv_22: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_6/add:0
output: conv_22/convolution:0
Processing a keras layer - (zero_padding2d_4: <class ‘keras.layers.convolutional.ZeroPadding2D’>)
input : add_11/add:0
output: zero_padding2d_4/Pad:0
Processing a keras layer - (leaky_38: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_38/cond/Merge:0
output: leaky_38/LeakyRelu:0
Processing a keras layer - (bnorm_41: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_41/convolution:0
output: bnorm_41/cond/Merge:0
Processing a keras layer - (conv_44: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_13/add:0
output: conv_44/convolution:0
Processing a keras layer - (zero_padding2d_3: <class ‘keras.layers.convolutional.ZeroPadding2D’>)
input : add_3/add:0
output: zero_padding2d_3/Pad:0
Processing a keras layer - (leaky_13: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_13/cond/Merge:0
output: leaky_13/LeakyRelu:0
Processing a keras layer - (bnorm_16: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_16/convolution:0
output: bnorm_16/cond/Merge:0
Processing a keras layer - (conv_19: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_5/add:0
output: conv_19/convolution:0
Processing a keras layer - (bnorm_38: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_38/convolution:0
output: bnorm_38/cond/Merge:0
Processing a keras layer - (conv_41: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_12/add:0
output: conv_41/convolution:0
Processing a keras layer - (add_3: <class ‘keras.layers.merge.Add’>)
input : add_2/add:0
input : leaky_10/LeakyRelu:0
output: add_3/add:0
Processing a keras layer - (bnorm_13: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_13/convolution:0
output: bnorm_13/cond/Merge:0
Processing a keras layer - (conv_16: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_4/add:0
output: conv_16/convolution:0
Processing a keras layer - (conv_38: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_37/LeakyRelu:0
output: conv_38/convolution:0
Processing a keras layer - (add_2: <class ‘keras.layers.merge.Add’>)
input : leaky_5/LeakyRelu:0
input : leaky_7/LeakyRelu:0
output: add_2/add:0
Processing a keras layer - (leaky_10: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_10/cond/Merge:0
output: leaky_10/LeakyRelu:0
Processing a keras layer - (conv_13: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_12/LeakyRelu:0
output: conv_13/convolution:0
Processing a keras layer - (leaky_5: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_5/cond/Merge:0
output: leaky_5/LeakyRelu:0
Processing a keras layer - (leaky_7: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_7/cond/Merge:0
output: leaky_7/LeakyRelu:0
Processing a keras layer - (bnorm_10: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_10/convolution:0
output: bnorm_10/cond/Merge:0
Processing a keras layer - (bnorm_5: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_5/convolution:0
output: bnorm_5/cond/Merge:0
Processing a keras layer - (bnorm_7: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_7/convolution:0
output: bnorm_7/cond/Merge:0
Processing a keras layer - (conv_10: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_9/LeakyRelu:0
output: conv_10/convolution:0
Processing a keras layer - (conv_5: <class ‘keras.layers.convolutional.Conv2D’>)
input : zero_padding2d_2/Pad:0
output: conv_5/convolution:0
Processing a keras layer - (conv_7: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_6/LeakyRelu:0
output: conv_7/convolution:0
Processing a keras layer - (leaky_9: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_9/cond/Merge:0
output: leaky_9/LeakyRelu:0
Processing a keras layer - (zero_padding2d_2: <class ‘keras.layers.convolutional.ZeroPadding2D’>)
input : add_1/add:0
output: zero_padding2d_2/Pad:0
Processing a keras layer - (leaky_6: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_6/cond/Merge:0
output: leaky_6/LeakyRelu:0
Processing a keras layer - (bnorm_9: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_9/convolution:0
output: bnorm_9/cond/Merge:0
Processing a keras layer - (add_1: <class ‘keras.layers.merge.Add’>)
input : leaky_1/LeakyRelu:0
input : leaky_3/LeakyRelu:0
output: add_1/add:0
Processing a keras layer - (bnorm_6: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_6/convolution:0
output: bnorm_6/cond/Merge:0
Processing a keras layer - (conv_9: <class ‘keras.layers.convolutional.Conv2D’>)
input : add_2/add:0
output: conv_9/convolution:0
Processing a keras layer - (leaky_1: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_1/cond/Merge:0
output: leaky_1/LeakyRelu:0
Processing a keras layer - (leaky_3: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_3/cond/Merge:0
output: leaky_3/LeakyRelu:0
Processing a keras layer - (conv_6: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_5/LeakyRelu:0
output: conv_6/convolution:0
Processing a keras layer - (bnorm_1: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_1/convolution:0
output: bnorm_1/cond/Merge:0
Processing a keras layer - (bnorm_3: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_3/convolution:0
output: bnorm_3/cond/Merge:0
Processing a keras layer - (conv_1: <class ‘keras.layers.convolutional.Conv2D’>)
input : zero_padding2d_1/Pad:0
output: conv_1/convolution:0
Processing a keras layer - (conv_3: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_2/LeakyRelu:0
output: conv_3/convolution:0
Processing a keras layer - (zero_padding2d_1: <class ‘keras.layers.convolutional.ZeroPadding2D’>)
input : leaky_0/LeakyRelu:0
output: zero_padding2d_1/Pad:0
Processing a keras layer - (leaky_2: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_2/cond/Merge:0
output: leaky_2/LeakyRelu:0
Processing a keras layer - (leaky_0: <class ‘keras.layers.advanced_activations.LeakyReLU’>)
input : bnorm_0/cond/Merge:0
output: leaky_0/LeakyRelu:0
Processing a keras layer - (bnorm_2: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_2/convolution:0
output: bnorm_2/cond/Merge:0
Processing a keras layer - (bnorm_0: <class ‘keras.layers.normalization.BatchNormalization’>)
input : conv_0/convolution:0
output: bnorm_0/cond/Merge:0
Processing a keras layer - (conv_2: <class ‘keras.layers.convolutional.Conv2D’>)
input : leaky_1/LeakyRelu:0
output: conv_2/convolution:0
Processing a keras layer - (conv_0: <class ‘keras.layers.convolutional.Conv2D’>)
input : input_1:0
output: conv_0/convolution:0
var: input_1
var: input_1:0
var: input_1:01
var: conv_0/convolution:0
var: bnorm_0/cond/Merge:0
var: leaky_0/LeakyRelu:0
var: zero_padding2d_1/Pad:0
var: conv_1/convolution:0
var: bnorm_1/cond/Merge:0
var: leaky_1/LeakyRelu:0
var: add_1/add:0
var: conv_2/convolution:0
var: zero_padding2d_2/Pad:0
var: bnorm_2/cond/Merge:0
var: conv_5/convolution:0
var: leaky_2/LeakyRelu:0
var: bnorm_5/cond/Merge:0
var: conv_3/convolution:0
var: leaky_5/LeakyRelu:0
var: bnorm_3/cond/Merge:0
var: add_2/add:0
var: conv_6/convolution:0
var: leaky_3/LeakyRelu:0
var: add_3/add:0
var: conv_9/convolution:0
var: bnorm_6/cond/Merge:0
var: zero_padding2d_3/Pad:0
var: bnorm_9/cond/Merge:0
var: leaky_6/LeakyRelu:0
var: conv_12/convolution:0
var: leaky_9/LeakyRelu:0
var: conv_7/convolution:0
var: bnorm_12/cond/Merge:0
var: conv_10/convolution:0
var: bnorm_7/cond/Merge:0
var: leaky_12/LeakyRelu:0
var: bnorm_10/cond/Merge:0
var: leaky_7/LeakyRelu:0
var: add_4/add:0
var: conv_13/convolution:0
var: leaky_10/LeakyRelu:0
var: add_5/add:0
var: conv_16/convolution:0
var: bnorm_13/cond/Merge:0
var: add_6/add:0
var: conv_19/convolution:0
var: bnorm_16/cond/Merge:0
var: leaky_13/LeakyRelu:0
var: add_7/add:0
var: conv_22/convolution:0
var: bnorm_19/cond/Merge:0
var: leaky_16/LeakyRelu:0
var: conv_14/convolution:0
var: add_8/add:0
var: conv_25/convolution:0
var: bnorm_22/cond/Merge:0
var: leaky_19/LeakyRelu:0
var: conv_17/convolution:0
var: bnorm_14/cond/Merge:0
var: add_9/add:0
var: conv_28/convolution:0
var: bnorm_25/cond/Merge:0
var: leaky_22/LeakyRelu:0
var: conv_20/convolution:0
var: bnorm_17/cond/Merge:0
var: leaky_14/LeakyRelu:0
var: add_10/add:0
var: conv_31/convolution:0
var: bnorm_28/cond/Merge:0
var: leaky_25/LeakyRelu:0
var: conv_23/convolution:0
var: bnorm_20/cond/Merge:0
var: leaky_17/LeakyRelu:0
var: add_11/add:0
var: conv_34/convolution:0
var: bnorm_31/cond/Merge:0
var: leaky_28/LeakyRelu:0
var: conv_26/convolution:0
var: bnorm_23/cond/Merge:0
var: leaky_20/LeakyRelu:0
var: concatenate_2/concat:0
var: zero_padding2d_4/Pad:0
var: bnorm_34/cond/Merge:0
var: leaky_31/LeakyRelu:0
var: conv_29/convolution:0
var: bnorm_26/cond/Merge:0
var: leaky_23/LeakyRelu:0
var: conv_99/convolution:0
var: conv_37/convolution:0
var: leaky_34/LeakyRelu:0
var: conv_32/convolution:0
var: bnorm_29/cond/Merge:0
var: leaky_26/LeakyRelu:0
var: bnorm_99/cond/Merge:0
var: bnorm_37/cond/Merge:0
var: conv_35/convolution:0
var: bnorm_32/cond/Merge:0
var: leaky_29/LeakyRelu:0
var: leaky_99/LeakyRelu:0
var: leaky_37/LeakyRelu:0
var: bnorm_35/cond/Merge:0
var: leaky_32/LeakyRelu:0
var: conv_100/convolution:0
var: add_12/add:0
var: conv_38/convolution:0
var: leaky_35/LeakyRelu:0
var: bnorm_100/cond/Merge:0
var: add_13/add:0
var: conv_41/convolution:0
var: bnorm_38/cond/Merge:0
var: leaky_100/LeakyRelu:0
var: add_14/add:0
var: conv_44/convolution:0
var: bnorm_41/cond/Merge:0
var: leaky_38/LeakyRelu:0
var: conv_101/convolution:0
var: add_15/add:0
var: conv_47/convolution:0
var: bnorm_44/cond/Merge:0
var: leaky_41/LeakyRelu:0
var: conv_39/convolution:0
var: bnorm_101/cond/Merge:0
var: add_16/add:0
var: conv_50/convolution:0
var: bnorm_47/cond/Merge:0
var: leaky_44/LeakyRelu:0
var: conv_42/convolution:0
var: bnorm_39/cond/Merge:0
var: leaky_101/LeakyRelu:0
var: add_17/add:0
var: conv_53/convolution:0
var: bnorm_50/cond/Merge:0
var: leaky_47/LeakyRelu:0
var: conv_45/convolution:0
var: bnorm_42/cond/Merge:0
var: leaky_39/LeakyRelu:0
var: conv_102/convolution:0
var: add_18/add:0
var: conv_56/convolution:0
var: bnorm_53/cond/Merge:0
var: leaky_50/LeakyRelu:0
var: conv_48/convolution:0
var: bnorm_45/cond/Merge:0
var: leaky_42/LeakyRelu:0
var: bnorm_102/cond/Merge:0
var: add_19/add:0
var: conv_59/convolution:0
var: bnorm_56/cond/Merge:0
var: leaky_53/LeakyRelu:0
var: conv_51/convolution:0
var: bnorm_48/cond/Merge:0
var: leaky_45/LeakyRelu:0
var: leaky_102/LeakyRelu:0
var: concatenate_1/concat:0
var: zero_padding2d_5/Pad:0
var: bnorm_59/cond/Merge:0
var: leaky_56/LeakyRelu:0
var: conv_54/convolution:0
var: bnorm_51/cond/Merge:0
var: leaky_48/LeakyRelu:0
var: conv_103/convolution:0
var: conv_87/convolution:0
var: conv_62/convolution:0
var: leaky_59/LeakyRelu:0
var: conv_57/convolution:0
var: bnorm_54/cond/Merge:0
var: leaky_51/LeakyRelu:0
var: bnorm_103/cond/Merge:0
var: bnorm_87/cond/Merge:0
var: bnorm_62/cond/Merge:0
var: conv_60/convolution:0
var: bnorm_57/cond/Merge:0
var: leaky_54/LeakyRelu:0
var: leaky_103/LeakyRelu:0
var: leaky_87/LeakyRelu:0
var: leaky_62/LeakyRelu:0
var: bnorm_60/cond/Merge:0
var: leaky_57/LeakyRelu:0
var: conv_104/convolution:0
var: conv_88/convolution:0
var: add_20/add:0
var: conv_63/convolution:0
var: leaky_60/LeakyRelu:0
var: bnorm_104/cond/Merge:0
var: bnorm_88/cond/Merge:0
var: add_21/add:0
var: conv_66/convolution:0
var: bnorm_63/cond/Merge:0
var: leaky_104/LeakyRelu:0
var: leaky_88/LeakyRelu:0
var: add_22/add:0
var: conv_69/convolution:0
var: bnorm_66/cond/Merge:0
var: leaky_63/LeakyRelu:0
var: conv_105/BiasAdd:01
var: conv_89/convolution:0
var: add_23/add:0
var: conv_72/convolution:0
var: bnorm_69/cond/Merge:0
var: leaky_66/LeakyRelu:0
var: conv_64/convolution:0
var: conv_105/BiasAdd:0
var: bnorm_89/cond/Merge:0
var: conv_75/convolution:0
var: bnorm_72/cond/Merge:0
var: leaky_69/LeakyRelu:0
var: conv_67/convolution:0
var: bnorm_64/cond/Merge:0
var: conv_105
var: leaky_89/LeakyRelu:0
var: bnorm_75/cond/Merge:0
var: leaky_72/LeakyRelu:0
var: conv_70/convolution:0
var: bnorm_67/cond/Merge:0
var: leaky_64/LeakyRelu:0
var: conv_90/convolution:0
var: leaky_75/LeakyRelu:0
var: conv_73/convolution:0
var: bnorm_70/cond/Merge:0
var: leaky_67/LeakyRelu:0
var: bnorm_90/cond/Merge:0
var: conv_76/convolution:0
var: bnorm_73/cond/Merge:0
var: leaky_70/LeakyRelu:0
var: leaky_90/LeakyRelu:0
var: bnorm_76/cond/Merge:0
var: leaky_73/LeakyRelu:0
var: conv_91/convolution:0
var: leaky_76/LeakyRelu:0
var: bnorm_91/cond/Merge:0
var: conv_77/convolution:0
var: leaky_91/LeakyRelu:0
var: bnorm_77/cond/Merge:0
var: conv_92/convolution:0
var: conv_96/convolution:0
var: leaky_77/LeakyRelu:0
var: bnorm_92/cond/Merge:0
var: bnorm_96/cond/Merge:0
var: conv_78/convolution:0
var: leaky_92/LeakyRelu:0
var: leaky_96/LeakyRelu:0
var: bnorm_78/cond/Merge:0
var: conv_93/BiasAdd:01
var: up_sampling2d_2/ResizeNearestNeighbor:0
var: leaky_78/LeakyRelu:0
var: conv_93/BiasAdd:0
var: conv_79/convolution:0
var: conv_93
var: bnorm_79/cond/Merge:0
var: leaky_79/LeakyRelu:0
var: conv_80/convolution:0
var: conv_84/convolution:0
var: bnorm_80/cond/Merge:0
var: bnorm_84/cond/Merge:0
var: leaky_80/LeakyRelu:0
var: leaky_84/LeakyRelu:0
var: conv_81/BiasAdd:01
var: up_sampling2d_1/ResizeNearestNeighbor:0
var: conv_81/BiasAdd:0
var: conv_81
Converting the operator (Identity): Identity
Converting the operator (Identity1): Identity
Converting the operator (Identity2): Identity
Converting the operator (Identity3): Identity
Converting the operator (Identity4): Identity
Converting the operator (Identity5): Identity
Converting the operator (Identity6): Identity
Converting the operator (conv_81): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_93): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_105): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_80): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_92): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_104): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_80): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_92): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_104): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_80): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_92): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_104): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_79): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_91): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_103): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_79): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_91): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_103): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_79): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_91): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_103): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_78): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_90): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_102): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_78): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_90): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_102): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_78): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_90): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_102): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_77): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_89): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_101): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_77): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_89): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_101): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_77): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_89): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_101): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_76): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_88): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_100): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_76): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_88): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_100): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_76): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_88): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_100): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_75): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_87): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_99): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_75): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_87): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_99): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_75): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_87): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_99): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_23): <class ‘keras.layers.merge.Add’>
Converting the operator (concatenate_1): <class ‘keras.layers.merge.Concatenate’>
Converting the operator (concatenate_2): <class ‘keras.layers.merge.Concatenate’>
Converting the operator (add_22): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_73): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (up_sampling2d_1): <class ‘keras.layers.convolutional.UpSampling2D’>
Converting the operator (add_19): <class ‘keras.layers.merge.Add’>
Converting the operator (up_sampling2d_2): <class ‘keras.layers.convolutional.UpSampling2D’>
Converting the operator (add_11): <class ‘keras.layers.merge.Add’>
Converting the operator (add_21): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_70): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_73): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (leaky_84): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (add_18): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_60): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_96): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (add_10): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_35): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (add_20): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_67): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_70): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_73): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_84): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (add_17): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_57): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_60): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_96): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (add_9): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_32): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_35): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (leaky_62): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_64): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_67): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_70): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_72): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (conv_84): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_16): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_54): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_57): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_60): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_96): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_8): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_29): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_32): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_35): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_62): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_64): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_67): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_69): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_72): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (add_15): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_51): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_54): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_57): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_59): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (add_7): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_26): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_29): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_32): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_34): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (conv_62): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_64): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_66): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_69): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_72): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_14): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_48): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_51): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_54): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_56): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_59): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (add_6): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_23): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_26): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_29): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_31): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_34): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (zero_padding2d_5): <class ‘keras.layers.convolutional.ZeroPadding2D’>
Converting the operator (leaky_63): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_66): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_69): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_13): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_45): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_48): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_51): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_53): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_56): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_59): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_5): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_20): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_23): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_26): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_28): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_31): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_34): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_63): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_66): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_12): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_42): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_45): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_48): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_50): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_53): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_56): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_4): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_17): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_20): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_23): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_25): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_28): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_31): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_63): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_37): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_39): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_42): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_45): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_47): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_50): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_53): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_12): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_14): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_17): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_20): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_22): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_25): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_28): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_37): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_39): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_42): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_44): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_47): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_50): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_12): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_14): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_17): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_19): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_22): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_25): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_37): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_39): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_41): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_44): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_47): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_12): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_14): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_16): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_19): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_22): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (zero_padding2d_4): <class ‘keras.layers.convolutional.ZeroPadding2D’>
Converting the operator (leaky_38): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_41): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_44): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (zero_padding2d_3): <class ‘keras.layers.convolutional.ZeroPadding2D’>
Converting the operator (leaky_13): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_16): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_19): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_38): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_41): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_3): <class ‘keras.layers.merge.Add’>
Converting the operator (bnorm_13): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_16): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_38): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (add_2): <class ‘keras.layers.merge.Add’>
Converting the operator (leaky_10): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (conv_13): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_5): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_7): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_10): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_5): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_7): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_10): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_5): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_7): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_9): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (zero_padding2d_2): <class ‘keras.layers.convolutional.ZeroPadding2D’>
Converting the operator (leaky_6): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_9): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (add_1): <class ‘keras.layers.merge.Add’>
Converting the operator (bnorm_6): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_9): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (leaky_1): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_3): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (conv_6): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (bnorm_1): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_3): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_1): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_3): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (zero_padding2d_1): <class ‘keras.layers.convolutional.ZeroPadding2D’>
Converting the operator (leaky_2): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (leaky_0): <class ‘keras.layers.advanced_activations.LeakyReLU’>
Converting the operator (bnorm_2): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (bnorm_0): <class ‘keras.layers.normalization.BatchNormalization’>
Converting the operator (conv_2): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (conv_0): <class ‘keras.layers.convolutional.Conv2D’>
Converting the operator (Identity7): Identity
The maximum opset needed by this model is only 11.

spolisetty · June 30, 2021, 10:36am

@1036758468,

Looks like you shared different logs(keras2onnx), We request you to share ONNX to TensorRT conversion logs and ONNX model.

Thank you.

1036758468 · June 30, 2021, 11:11am

TensorRT logs are as follows(I used python API to build engine). I’m sorry I can’t share my onnx model with you because of the company privacy issue.

[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::GridAnchor_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::NMS_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Reorg_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Region_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Clip_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::LReLU_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::PriorBox_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Normalize_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::RPROI_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::BatchedNMS_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::FlattenConcat_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::CropAndResize
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::DetectionLayer_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Proposal
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::ProposalLayer_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::PyramidROIAlign_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::ResizeNearest_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::Split
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::SpecialSlice_TRT
[TensorRT] VERBOSE: Plugin creator registration succeeded - ONNXTRT_NAMESPACE::InstanceNormalization_TRT
[TensorRT] VERBOSE: ModelImporter.cpp:203: Adding network input: input_1 with dtype: float32, dimensions: (-1, -1, -1, 3)
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: input_1 for ONNX tensor: input_1
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_81/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_81/bias:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_93/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_93/bias:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_105/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_105/bias:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale1
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias1
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean1
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var1
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale2
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias2
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean2
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var2
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_80/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_92/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_104/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale3
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias3
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean3
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var3
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale4
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias4
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean4
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var4
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale5
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias5
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean5
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var5
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_79/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_91/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_103/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale6
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias6
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean6
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var6
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale7
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias7
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean7
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var7
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale8
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias8
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean8
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var8
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_78/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_90/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_102/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale9
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias9
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean9
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var9
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale10
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias10
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean10
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var10
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale11
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias11
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean11
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var11
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_77/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_89/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_101/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale12
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias12
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean12
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var12
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale13
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias13
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean13
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var13
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale14
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias14
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean14
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var14
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_76/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_88/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_100/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale15
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias15
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean15
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var15
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale16
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias16
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean16
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var16
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale17
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias17
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean17
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var17
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_75/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_87/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_99/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Resize_roi
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Resize_scales
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Resize1_roi
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Resize1_scales
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale18
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias18
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean18
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var18
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale19
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias19
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean19
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var19
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_73/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale20
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias20
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean20
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var20
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale21
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias21
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean21
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var21
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale22
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias22
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean22
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var22
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale23
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias23
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean23
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var23
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale24
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias24
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean24
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var24
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_70/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_84/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale25
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias25
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean25
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var25
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_60/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_96/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale26
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias26
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean26
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var26
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_35/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale27
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias27
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean27
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var27
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale28
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias28
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean28
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var28
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_67/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale29
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias29
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean29
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var29
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale30
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias30
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean30
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var30
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_57/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale31
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias31
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean31
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var31
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_32/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_62/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_64/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale32
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias32
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean32
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var32
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_72/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale33
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias33
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean33
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var33
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_54/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale34
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias34
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean34
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var34
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale35
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias35
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean35
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var35
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_29/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale36
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias36
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean36
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var36
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad_pads
[TensorRT] WARNING: onnx2trt_utils.cpp:198: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad_value
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale37
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias37
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean37
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var37
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_69/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale38
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias38
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean38
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var38
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_51/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale39
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias39
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean39
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var39
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_59/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale40
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias40
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean40
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var40
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_26/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale41
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias41
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean41
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var41
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_34/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale42
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias42
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean42
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var42
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_66/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale43
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias43
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean43
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var43
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_48/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale44
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias44
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean44
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var44
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_56/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale45
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias45
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean45
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var45
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_23/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale46
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias46
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean46
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var46
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_31/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_63/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale47
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias47
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean47
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var47
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_45/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale48
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias48
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean48
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var48
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_53/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale49
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias49
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean49
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var49
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_20/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale50
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias50
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean50
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var50
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_28/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale51
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias51
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean51
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var51
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale52
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias52
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean52
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var52
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_42/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale53
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias53
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean53
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var53
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_50/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale54
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias54
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean54
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var54
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale55
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias55
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean55
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var55
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_17/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale56
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias56
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean56
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var56
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_25/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_37/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_39/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale57
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias57
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean57
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var57
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_47/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_12/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_14/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale58
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias58
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean58
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var58
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_22/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad1_pads
[TensorRT] WARNING: onnx2trt_utils.cpp:198: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad1_value
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale59
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias59
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean59
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var59
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_44/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad2_pads
[TensorRT] WARNING: onnx2trt_utils.cpp:198: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad2_value
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale60
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias60
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean60
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var60
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_19/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale61
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias61
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean61
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var61
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_41/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale62
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias62
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean62
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var62
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_16/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_38/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_13/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale63
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias63
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean63
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var63
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale64
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias64
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean64
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var64
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale65
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias65
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean65
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var65
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_10/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_5/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_7/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad3_pads
[TensorRT] WARNING: onnx2trt_utils.cpp:198: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad3_value
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale66
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias66
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean66
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var66
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale67
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias67
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean67
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var67
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_9/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_6/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale68
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias68
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean68
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var68
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale69
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias69
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean69
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var69
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_1/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_3/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad4_pads
[TensorRT] WARNING: onnx2trt_utils.cpp:198: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: Pad4_value
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale70
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias70
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean70
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var70
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: scale71
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: bias71
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: mean71
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: var71
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_2/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:90: Importing initializer: conv_0/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Identity8 [Identity]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: input_1
[TensorRT] VERBOSE: ModelImporter.cpp:129: Identity8 [Identity] inputs: [input_1 → (-1, -1, -1, 3)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Identity8 for ONNX node: Identity8
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: input_1:0 for ONNX tensor: input_1:0
[TensorRT] VERBOSE: ModelImporter.cpp:180: Identity8 [Identity] outputs: [input_1:0 → (-1, -1, -1, 3)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Identity90 [Identity]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: input_1:0
[TensorRT] VERBOSE: ModelImporter.cpp:129: Identity90 [Identity] inputs: [input_1:0 → (-1, -1, -1, 3)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Identity90 for ONNX node: Identity90
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: input_1:01 for ONNX tensor: input_1:01
[TensorRT] VERBOSE: ModelImporter.cpp:180: Identity90 [Identity] outputs: [input_1:01 → (-1, -1, -1, 3)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Transpose306 [Transpose]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: input_1:01
[TensorRT] VERBOSE: ModelImporter.cpp:129: Transpose306 [Transpose] inputs: [input_1:01 → (-1, -1, -1, 3)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Transpose306 for ONNX node: Transpose306
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: adjusted_input74 for ONNX tensor: adjusted_input74
[TensorRT] VERBOSE: ModelImporter.cpp:180: Transpose306 [Transpose] outputs: [adjusted_input74 → (-1, 3, -1, -1)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: conv_0 [Conv]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: adjusted_input74
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: conv_0/kernel:0
[TensorRT] VERBOSE: ModelImporter.cpp:129: conv_0 [Conv] inputs: [adjusted_input74 → (-1, 3, -1, -1)], [conv_0/kernel:0 → (32, 3, 3, 3)],
[TensorRT] VERBOSE: builtin_op_importers.cpp:442: Convolution input dimensions: (-1, 3, -1, -1)
[TensorRT] VERBOSE: Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[TensorRT] VERBOSE: builtin_op_importers.cpp:524: Using kernel: (3, 3), strides: (1, 1), padding: (0, 0), dilations: (1, 1), numOutputs: 32
[TensorRT] VERBOSE: builtin_op_importers.cpp:525: Convolution output dimensions: (-1, 32, -1, -1)
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: conv_0 for ONNX node: conv_0
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: convolution_output74 for ONNX tensor: convolution_output74
[TensorRT] VERBOSE: ModelImporter.cpp:180: conv_0 [Conv] outputs: [convolution_output74 → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Transpose307 [Transpose]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: convolution_output74
[TensorRT] VERBOSE: ModelImporter.cpp:129: Transpose307 [Transpose] inputs: [convolution_output74 → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Transpose307 for ONNX node: Transpose307
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: transpose_output74 for ONNX tensor: transpose_output74
[TensorRT] VERBOSE: ModelImporter.cpp:180: Transpose307 [Transpose] outputs: [transpose_output74 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Identity89 [Identity]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: transpose_output74
[TensorRT] VERBOSE: ModelImporter.cpp:129: Identity89 [Identity] inputs: [transpose_output74 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Identity89 for ONNX node: Identity89
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: conv_0/convolution:0 for ONNX tensor: conv_0/convolution:0
[TensorRT] VERBOSE: ModelImporter.cpp:180: Identity89 [Identity] outputs: [conv_0/convolution:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Transpose302 [Transpose]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: conv_0/convolution:0
[TensorRT] VERBOSE: ModelImporter.cpp:129: Transpose302 [Transpose] inputs: [conv_0/convolution:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Transpose302 for ONNX node: Transpose302
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: conv_0/convolution:0_transposed for ONNX tensor: conv_0/convolution:0_transposed
[TensorRT] VERBOSE: ModelImporter.cpp:180: Transpose302 [Transpose] outputs: [conv_0/convolution:0_transposed → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: bnorm_0 [BatchNormalization]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: conv_0/convolution:0_transposed
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: scale71
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: bias71
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: mean71
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: var71
[TensorRT] VERBOSE: ModelImporter.cpp:129: bnorm_0 [BatchNormalization] inputs: [conv_0/convolution:0_transposed → (-1, 32, -1, -1)], [scale71 → (32)], [bias71 → (32)], [mean71 → (32)], [var71 → (32)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: bnorm_0 for ONNX node: bnorm_0
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: batch_norm_output_buffer71 for ONNX tensor: batch_norm_output_buffer71
[TensorRT] VERBOSE: ModelImporter.cpp:180: bnorm_0 [BatchNormalization] outputs: [batch_norm_output_buffer71 → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Transpose303 [Transpose]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: batch_norm_output_buffer71
[TensorRT] VERBOSE: ModelImporter.cpp:129: Transpose303 [Transpose] inputs: [batch_norm_output_buffer71 → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Transpose303 for ONNX node: Transpose303
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: bnorm_0/cond/Merge:0 for ONNX tensor: bnorm_0/cond/Merge:0
[TensorRT] VERBOSE: ModelImporter.cpp:180: Transpose303 [Transpose] outputs: [bnorm_0/cond/Merge:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: leaky_0 [LeakyRelu]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: bnorm_0/cond/Merge:0
[TensorRT] VERBOSE: ModelImporter.cpp:129: leaky_0 [LeakyRelu] inputs: [bnorm_0/cond/Merge:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: leaky_0 for ONNX node: leaky_0
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: leaky_0/LeakyRelu:0 for ONNX tensor: leaky_0/LeakyRelu:0
[TensorRT] VERBOSE: ModelImporter.cpp:180: leaky_0 [LeakyRelu] outputs: [leaky_0/LeakyRelu:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Transpose298 [Transpose]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: leaky_0/LeakyRelu:0
[TensorRT] VERBOSE: ModelImporter.cpp:129: Transpose298 [Transpose] inputs: [leaky_0/LeakyRelu:0 → (-1, -1, -1, 32)],
[TensorRT] VERBOSE: ImporterContext.hpp:122: Registering layer: Transpose298 for ONNX node: Transpose298
[TensorRT] VERBOSE: ImporterContext.hpp:97: Registering tensor: leaky_0/LeakyRelu:0_permuted for ONNX tensor: leaky_0/LeakyRelu:0_permuted
[TensorRT] VERBOSE: ModelImporter.cpp:180: Transpose298 [Transpose] outputs: [leaky_0/LeakyRelu:0_permuted → (-1, 32, -1, -1)],
[TensorRT] VERBOSE: ModelImporter.cpp:107: Parsing node: Pad4 [Pad]
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: leaky_0/LeakyRelu:0_permuted
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: Pad4_pads
[TensorRT] VERBOSE: ModelImporter.cpp:123: Searching for input: Pad4_value
[TensorRT] VERBOSE: ModelImporter.cpp:129: Pad4 [Pad] inputs: [leaky_0/LeakyRelu:0_permuted → (-1, 32, -1, -1)], [Pad4_pads → (8)], [Pad4_value → ()],
Traceback (most recent call last):
File “/xxx/TensorRT/TensorRT-7.0.0.11/mg_predict_vs_tensorrt.py”, line 154, in
main()
File “/xxx/TensorRT/TensorRT-7.0.0.11/mg_predict_vs_tensorrt.py”, line 124, in main
with build_engine_onnx(onnx_model_file) as engine, engine.create_execution_context() as context:
File “/xxx/TensorRT/TensorRT-7.0.0.11/mg_predict_vs_tensorrt.py”, line 50, in build_engine_onnx
if not parser.parse(model.read()):
IndexError: Attribute not found: pads

spolisetty · July 1, 2021, 12:41pm

@1036758468,

Please check whether “pads” exists for the layer in ONNX model. If not there, this is TF/Keras to ONNX converter issue.

1036758468 · July 2, 2021, 2:04am

Hi, Thanks. I found the problem. I used TensorRT 7.2.3.4 to build engine from onnx model, but in my environment path, I set LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/xxx/TensorRT/TensorRT-7.0.0.11/lib. So, TensorRT actually uses files in TensorRT-7.0.0.11/lib, which leading the “Attribute not found” error somehow.

TomK · October 12, 2021, 7:47pm