search

PINTO0309 / openvino2tensorflow

This script converts the ONNX/OpenVINO IR model to Tensorflow's saved_model, tflite, h5, tfjs, tftrt(TensorRT), CoreML, EdgeTPU, ONNX and pb. PyTorch (NCHW) -> ONNX (NCHW) -> OpenVINO (NCHW) -> openvino2tensorflow -> Tensorflow/Keras (NHWC/NCHW) -> TFLite (NHWC/NCHW). And the conversion from .pb to saved_model and from saved_model to .pb and from .pb to .tflite and saved_model to .tflite and saved_model to onnx. Support for building environments with Docker. It is possible to directly access the host PC GUI and the camera to verify the operation. NVIDIA GPU (dGPU) support. Intel iHD GPU (iGPU) support.
MIT License
334 stars 40 forks source link

tf.constant with unsupported shape #72

Closed asonee closed 2 years ago

asonee commented 2 years ago

Thanks for your great work! When I use this to do convert openvino to tensorflow, It occurs error: Eager execution of tf.constant with unsupported shape , and the PRELU op convert failed.

convert shell script :

INTEL_OPENVINO_DIR=/opt/intel/openvino_2021
H=112
W=112
MODEL=gender_mobile2

openvino2tensorflow \
--model_path openvino/${MODEL}/FP32/${MODEL}_${H}x${W}.xml \
--output_saved_model \
--output_pb

Issue Details

image

![Uploading image.png…]()

openvino models

FP32.zip

PINTO0309 commented 2 years ago 
$ pip list | grep openvino2
openvino2tensorflow      1.21.3

$ openvino2tensorflow \
--model_path gender_mobile2_112x112.xml \
--output_no_quant_float32_tflite

Screenshot 2021-09-27 23:18:35

asonee commented 2 years ago

It seems very strange, When I update openvino2tensorflow, it occurs errors too.

image image

PINTO0309 commented 2 years ago

I do not know why the error occurs only in your environment.

user@776cfeff6ae2:~/workdir$ openvino2tensorflow --model_path gender_mobile2_112x112.xml --output_no_quant_float32_tflite
TensorFlow/Keras model building process starts ======================================
Layer structure =====================================================================
layer_type: Input
layer_id: 0
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 112, 112, 3), dtype=tf.float32, name='input_0'), name='input_0', description="created by layer 'input_0'")
====================================================================================
layer_type: Const
layer_id: 1
tf_layers_dict: (32, 3, 3, 3)
====================================================================================
layer_type: Convolution
layer_id: 2
input_layer0: layer_id=0: KerasTensor(type_spec=TensorSpec(shape=(1, 112, 112, 3), dtype=tf.float32, name='input_0'), name='input_0', description="created by layer 'input_0'")
input_layer1: layer_id=1: Const(ndarray).shape (32, 3, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='conv2d/Conv2D:0', description="created by layer 'conv2d'")
====================================================================================
layer_type: Const
layer_id: 3
tf_layers_dict: (1, 32, 1, 1)
====================================================================================
layer_type: Add
layer_id: 4
input_layer0: layer_id=2: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='conv2d/Conv2D:0', description="created by layer 'conv2d'")
input_layer1: layer_id=3: Const(ndarray).shape (1, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add/Add:0', description="created by layer 'tf.math.add'")
====================================================================================
layer_type: Const
layer_id: 5
tf_layers_dict: (32, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 6
input_layer0: layer_id=4: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add/Add:0', description="created by layer 'tf.math.add'")
input_layer1: layer_id=5: Const(ndarray).shape (32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_1/add:0', description="created by layer 'p_re_lu_1'")
====================================================================================
layer_type: Const
layer_id: 7
tf_layers_dict: (32, 32, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 8
input_layer0: layer_id=6: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_1/add:0', description="created by layer 'p_re_lu_1'")
input_layer1: layer_id=7: Const(ndarray).shape (32, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='conv2d_1/Conv2D:0', description="created by layer 'conv2d_1'")
====================================================================================
layer_type: Const
layer_id: 9
tf_layers_dict: (1, 32, 1, 1)
====================================================================================
layer_type: Add
layer_id: 10
input_layer0: layer_id=8: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='conv2d_1/Conv2D:0', description="created by layer 'conv2d_1'")
input_layer1: layer_id=9: Const(ndarray).shape (1, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add_1/Add:0', description="created by layer 'tf.math.add_1'")
====================================================================================
layer_type: Const
layer_id: 11
tf_layers_dict: (32, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 12
input_layer0: layer_id=10: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add_1/Add:0', description="created by layer 'tf.math.add_1'")
input_layer1: layer_id=11: Const(ndarray).shape (32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_3/add:0', description="created by layer 'p_re_lu_3'")
====================================================================================
layer_type: Const
layer_id: 13
tf_layers_dict: (32, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 14
input_layer0: layer_id=12: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_3/add:0', description="created by layer 'p_re_lu_3'")
input_layer1: layer_id=13: Const(ndarray).shape (32, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='depthwise_conv2d/depthwise:0', description="created by layer 'depthwise_conv2d'")
====================================================================================
layer_type: Const
layer_id: 15
tf_layers_dict: (1, 32, 1, 1)
====================================================================================
layer_type: Add
layer_id: 16
input_layer0: layer_id=14: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='depthwise_conv2d/depthwise:0', description="created by layer 'depthwise_conv2d'")
input_layer1: layer_id=15: Const(ndarray).shape (1, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add_2/Add:0', description="created by layer 'tf.math.add_2'")
====================================================================================
layer_type: Const
layer_id: 17
tf_layers_dict: (32, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 18
input_layer0: layer_id=16: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='tf.math.add_2/Add:0', description="created by layer 'tf.math.add_2'")
input_layer1: layer_id=17: Const(ndarray).shape (32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_5/add:0', description="created by layer 'p_re_lu_5'")
====================================================================================
layer_type: Const
layer_id: 19
tf_layers_dict: (16, 32, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 20
input_layer0: layer_id=18: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_5/add:0', description="created by layer 'p_re_lu_5'")
input_layer1: layer_id=19: Const(ndarray).shape (16, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='conv2d_2/Conv2D:0', description="created by layer 'conv2d_2'")
====================================================================================
layer_type: Const
layer_id: 21
tf_layers_dict: (1, 16, 1, 1)
====================================================================================
layer_type: Add
layer_id: 22
input_layer0: layer_id=20: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='conv2d_2/Conv2D:0', description="created by layer 'conv2d_2'")
input_layer1: layer_id=21: Const(ndarray).shape (1, 16, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_3/Add:0', description="created by layer 'tf.math.add_3'")
====================================================================================
layer_type: Const
layer_id: 23
tf_layers_dict: (16, 32, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 24
input_layer0: layer_id=6: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 32), dtype=tf.float32, name=None), name='p_re_lu_1/add:0', description="created by layer 'p_re_lu_1'")
input_layer1: layer_id=23: Const(ndarray).shape (16, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='conv2d_3/Conv2D:0', description="created by layer 'conv2d_3'")
====================================================================================
layer_type: Const
layer_id: 25
tf_layers_dict: (1, 16, 1, 1)
====================================================================================
layer_type: Add
layer_id: 26
input_layer0: layer_id=24: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='conv2d_3/Conv2D:0', description="created by layer 'conv2d_3'")
input_layer1: layer_id=25: Const(ndarray).shape (1, 16, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_4/Add:0', description="created by layer 'tf.math.add_4'")
====================================================================================
layer_type: Add
layer_id: 27
input_layer0: layer_id=22: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_3/Add:0', description="created by layer 'tf.math.add_3'")
input_layer1: layer_id=26: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_4/Add:0', description="created by layer 'tf.math.add_4'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_5/Add:0', description="created by layer 'tf.math.add_5'")
====================================================================================
layer_type: Const
layer_id: 28
tf_layers_dict: (96, 16, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 29
input_layer0: layer_id=27: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 16), dtype=tf.float32, name=None), name='tf.math.add_5/Add:0', description="created by layer 'tf.math.add_5'")
input_layer1: layer_id=28: Const(ndarray).shape (96, 16, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='conv2d_4/Conv2D:0', description="created by layer 'conv2d_4'")
====================================================================================
layer_type: Const
layer_id: 30
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 31
input_layer0: layer_id=29: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='conv2d_4/Conv2D:0', description="created by layer 'conv2d_4'")
input_layer1: layer_id=30: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='tf.math.add_6/Add:0', description="created by layer 'tf.math.add_6'")
====================================================================================
layer_type: Const
layer_id: 32
tf_layers_dict: (96, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 33
input_layer0: layer_id=31: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='tf.math.add_6/Add:0', description="created by layer 'tf.math.add_6'")
input_layer1: layer_id=32: Const(ndarray).shape (96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='p_re_lu_7/add:0', description="created by layer 'p_re_lu_7'")
====================================================================================
layer_type: Const
layer_id: 34
tf_layers_dict: (96, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 35
input_layer0: layer_id=33: KerasTensor(type_spec=TensorSpec(shape=(1, 56, 56, 96), dtype=tf.float32, name=None), name='p_re_lu_7/add:0', description="created by layer 'p_re_lu_7'")
input_layer1: layer_id=34: Const(ndarray).shape (96, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='depthwise_conv2d_1/depthwise:0', description="created by layer 'depthwise_conv2d_1'")
====================================================================================
layer_type: Const
layer_id: 36
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 37
input_layer0: layer_id=35: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='depthwise_conv2d_1/depthwise:0', description="created by layer 'depthwise_conv2d_1'")
input_layer1: layer_id=36: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='tf.math.add_7/Add:0', description="created by layer 'tf.math.add_7'")
====================================================================================
layer_type: Const
layer_id: 38
tf_layers_dict: (96, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 39
input_layer0: layer_id=37: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='tf.math.add_7/Add:0', description="created by layer 'tf.math.add_7'")
input_layer1: layer_id=38: Const(ndarray).shape (96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='p_re_lu_9/add:0', description="created by layer 'p_re_lu_9'")
====================================================================================
layer_type: Const
layer_id: 40
tf_layers_dict: (24, 96, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 41
input_layer0: layer_id=39: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 96), dtype=tf.float32, name=None), name='p_re_lu_9/add:0', description="created by layer 'p_re_lu_9'")
input_layer1: layer_id=40: Const(ndarray).shape (24, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='conv2d_5/Conv2D:0', description="created by layer 'conv2d_5'")
====================================================================================
layer_type: Const
layer_id: 42
tf_layers_dict: (1, 24, 1, 1)
====================================================================================
layer_type: Add
layer_id: 43
input_layer0: layer_id=41: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='conv2d_5/Conv2D:0', description="created by layer 'conv2d_5'")
input_layer1: layer_id=42: Const(ndarray).shape (1, 24, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_8/Add:0', description="created by layer 'tf.math.add_8'")
====================================================================================
layer_type: Const
layer_id: 44
tf_layers_dict: (144, 24, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 45
input_layer0: layer_id=43: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_8/Add:0', description="created by layer 'tf.math.add_8'")
input_layer1: layer_id=44: Const(ndarray).shape (144, 24, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='conv2d_6/Conv2D:0', description="created by layer 'conv2d_6'")
====================================================================================
layer_type: Const
layer_id: 46
tf_layers_dict: (1, 144, 1, 1)
====================================================================================
layer_type: Add
layer_id: 47
input_layer0: layer_id=45: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='conv2d_6/Conv2D:0', description="created by layer 'conv2d_6'")
input_layer1: layer_id=46: Const(ndarray).shape (1, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_9/Add:0', description="created by layer 'tf.math.add_9'")
====================================================================================
layer_type: Const
layer_id: 48
tf_layers_dict: (144, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 49
input_layer0: layer_id=47: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_9/Add:0', description="created by layer 'tf.math.add_9'")
input_layer1: layer_id=48: Const(ndarray).shape (144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_11/add:0', description="created by layer 'p_re_lu_11'")
====================================================================================
layer_type: Const
layer_id: 50
tf_layers_dict: (144, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 51
input_layer0: layer_id=49: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_11/add:0', description="created by layer 'p_re_lu_11'")
input_layer1: layer_id=50: Const(ndarray).shape (144, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='depthwise_conv2d_2/depthwise:0', description="created by layer 'depthwise_conv2d_2'")
====================================================================================
layer_type: Const
layer_id: 52
tf_layers_dict: (1, 144, 1, 1)
====================================================================================
layer_type: Add
layer_id: 53
input_layer0: layer_id=51: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='depthwise_conv2d_2/depthwise:0', description="created by layer 'depthwise_conv2d_2'")
input_layer1: layer_id=52: Const(ndarray).shape (1, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_10/Add:0', description="created by layer 'tf.math.add_10'")
====================================================================================
layer_type: Const
layer_id: 54
tf_layers_dict: (144, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 55
input_layer0: layer_id=53: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_10/Add:0', description="created by layer 'tf.math.add_10'")
input_layer1: layer_id=54: Const(ndarray).shape (144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_13/add:0', description="created by layer 'p_re_lu_13'")
====================================================================================
layer_type: Const
layer_id: 56
tf_layers_dict: (24, 144, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 57
input_layer0: layer_id=55: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_13/add:0', description="created by layer 'p_re_lu_13'")
input_layer1: layer_id=56: Const(ndarray).shape (24, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='conv2d_7/Conv2D:0', description="created by layer 'conv2d_7'")
====================================================================================
layer_type: Const
layer_id: 58
tf_layers_dict: (1, 24, 1, 1)
====================================================================================
layer_type: Add
layer_id: 59
input_layer0: layer_id=57: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='conv2d_7/Conv2D:0', description="created by layer 'conv2d_7'")
input_layer1: layer_id=58: Const(ndarray).shape (1, 24, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_11/Add:0', description="created by layer 'tf.math.add_11'")
====================================================================================
layer_type: Add
layer_id: 60
input_layer0: layer_id=43: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_8/Add:0', description="created by layer 'tf.math.add_8'")
input_layer1: layer_id=59: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_11/Add:0', description="created by layer 'tf.math.add_11'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_12/Add:0', description="created by layer 'tf.math.add_12'")
====================================================================================
layer_type: Const
layer_id: 61
tf_layers_dict: (144, 24, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 62
input_layer0: layer_id=60: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 24), dtype=tf.float32, name=None), name='tf.math.add_12/Add:0', description="created by layer 'tf.math.add_12'")
input_layer1: layer_id=61: Const(ndarray).shape (144, 24, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='conv2d_8/Conv2D:0', description="created by layer 'conv2d_8'")
====================================================================================
layer_type: Const
layer_id: 63
tf_layers_dict: (1, 144, 1, 1)
====================================================================================
layer_type: Add
layer_id: 64
input_layer0: layer_id=62: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='conv2d_8/Conv2D:0', description="created by layer 'conv2d_8'")
input_layer1: layer_id=63: Const(ndarray).shape (1, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_13/Add:0', description="created by layer 'tf.math.add_13'")
====================================================================================
layer_type: Const
layer_id: 65
tf_layers_dict: (144, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 66
input_layer0: layer_id=64: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='tf.math.add_13/Add:0', description="created by layer 'tf.math.add_13'")
input_layer1: layer_id=65: Const(ndarray).shape (144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_15/add:0', description="created by layer 'p_re_lu_15'")
====================================================================================
layer_type: Const
layer_id: 67
tf_layers_dict: (144, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 68
input_layer0: layer_id=66: KerasTensor(type_spec=TensorSpec(shape=(1, 28, 28, 144), dtype=tf.float32, name=None), name='p_re_lu_15/add:0', description="created by layer 'p_re_lu_15'")
input_layer1: layer_id=67: Const(ndarray).shape (144, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='depthwise_conv2d_3/depthwise:0', description="created by layer 'depthwise_conv2d_3'")
====================================================================================
layer_type: Const
layer_id: 69
tf_layers_dict: (1, 144, 1, 1)
====================================================================================
layer_type: Add
layer_id: 70
input_layer0: layer_id=68: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='depthwise_conv2d_3/depthwise:0', description="created by layer 'depthwise_conv2d_3'")
input_layer1: layer_id=69: Const(ndarray).shape (1, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='tf.math.add_14/Add:0', description="created by layer 'tf.math.add_14'")
====================================================================================
layer_type: Const
layer_id: 71
tf_layers_dict: (144, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 72
input_layer0: layer_id=70: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='tf.math.add_14/Add:0', description="created by layer 'tf.math.add_14'")
input_layer1: layer_id=71: Const(ndarray).shape (144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='p_re_lu_17/add:0', description="created by layer 'p_re_lu_17'")
====================================================================================
layer_type: Const
layer_id: 73
tf_layers_dict: (32, 144, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 74
input_layer0: layer_id=72: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 144), dtype=tf.float32, name=None), name='p_re_lu_17/add:0', description="created by layer 'p_re_lu_17'")
input_layer1: layer_id=73: Const(ndarray).shape (32, 144, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='conv2d_9/Conv2D:0', description="created by layer 'conv2d_9'")
====================================================================================
layer_type: Const
layer_id: 75
tf_layers_dict: (1, 32, 1, 1)
====================================================================================
layer_type: Add
layer_id: 76
input_layer0: layer_id=74: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='conv2d_9/Conv2D:0', description="created by layer 'conv2d_9'")
input_layer1: layer_id=75: Const(ndarray).shape (1, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_15/Add:0', description="created by layer 'tf.math.add_15'")
====================================================================================
layer_type: Const
layer_id: 77
tf_layers_dict: (192, 32, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 78
input_layer0: layer_id=76: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_15/Add:0', description="created by layer 'tf.math.add_15'")
input_layer1: layer_id=77: Const(ndarray).shape (192, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='conv2d_10/Conv2D:0', description="created by layer 'conv2d_10'")
====================================================================================
layer_type: Const
layer_id: 79
tf_layers_dict: (1, 192, 1, 1)
====================================================================================
layer_type: Add
layer_id: 80
input_layer0: layer_id=78: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='conv2d_10/Conv2D:0', description="created by layer 'conv2d_10'")
input_layer1: layer_id=79: Const(ndarray).shape (1, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_16/Add:0', description="created by layer 'tf.math.add_16'")
====================================================================================
layer_type: Const
layer_id: 81
tf_layers_dict: (192, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 82
input_layer0: layer_id=80: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_16/Add:0', description="created by layer 'tf.math.add_16'")
input_layer1: layer_id=81: Const(ndarray).shape (192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_19/add:0', description="created by layer 'p_re_lu_19'")
====================================================================================
layer_type: Const
layer_id: 83
tf_layers_dict: (192, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 84
input_layer0: layer_id=82: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_19/add:0', description="created by layer 'p_re_lu_19'")
input_layer1: layer_id=83: Const(ndarray).shape (192, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='depthwise_conv2d_4/depthwise:0', description="created by layer 'depthwise_conv2d_4'")
====================================================================================
layer_type: Const
layer_id: 85
tf_layers_dict: (1, 192, 1, 1)
====================================================================================
layer_type: Add
layer_id: 86
input_layer0: layer_id=84: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='depthwise_conv2d_4/depthwise:0', description="created by layer 'depthwise_conv2d_4'")
input_layer1: layer_id=85: Const(ndarray).shape (1, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_17/Add:0', description="created by layer 'tf.math.add_17'")
====================================================================================
layer_type: Const
layer_id: 87
tf_layers_dict: (192, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 88
input_layer0: layer_id=86: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_17/Add:0', description="created by layer 'tf.math.add_17'")
input_layer1: layer_id=87: Const(ndarray).shape (192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_21/add:0', description="created by layer 'p_re_lu_21'")
====================================================================================
layer_type: Const
layer_id: 89
tf_layers_dict: (32, 192, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 90
input_layer0: layer_id=88: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_21/add:0', description="created by layer 'p_re_lu_21'")
input_layer1: layer_id=89: Const(ndarray).shape (32, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='conv2d_11/Conv2D:0', description="created by layer 'conv2d_11'")
====================================================================================
layer_type: Const
layer_id: 91
tf_layers_dict: (1, 32, 1, 1)
====================================================================================
layer_type: Add
layer_id: 92
input_layer0: layer_id=90: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='conv2d_11/Conv2D:0', description="created by layer 'conv2d_11'")
input_layer1: layer_id=91: Const(ndarray).shape (1, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_18/Add:0', description="created by layer 'tf.math.add_18'")
====================================================================================
layer_type: Add
layer_id: 93
input_layer0: layer_id=76: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_15/Add:0', description="created by layer 'tf.math.add_15'")
input_layer1: layer_id=92: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_18/Add:0', description="created by layer 'tf.math.add_18'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_19/Add:0', description="created by layer 'tf.math.add_19'")
====================================================================================
layer_type: Const
layer_id: 94
tf_layers_dict: (192, 32, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 95
input_layer0: layer_id=93: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 32), dtype=tf.float32, name=None), name='tf.math.add_19/Add:0', description="created by layer 'tf.math.add_19'")
input_layer1: layer_id=94: Const(ndarray).shape (192, 32, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='conv2d_12/Conv2D:0', description="created by layer 'conv2d_12'")
====================================================================================
layer_type: Const
layer_id: 96
tf_layers_dict: (1, 192, 1, 1)
====================================================================================
layer_type: Add
layer_id: 97
input_layer0: layer_id=95: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='conv2d_12/Conv2D:0', description="created by layer 'conv2d_12'")
input_layer1: layer_id=96: Const(ndarray).shape (1, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_20/Add:0', description="created by layer 'tf.math.add_20'")
====================================================================================
layer_type: Const
layer_id: 98
tf_layers_dict: (192, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 99
input_layer0: layer_id=97: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='tf.math.add_20/Add:0', description="created by layer 'tf.math.add_20'")
input_layer1: layer_id=98: Const(ndarray).shape (192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_23/add:0', description="created by layer 'p_re_lu_23'")
====================================================================================
layer_type: Const
layer_id: 100
tf_layers_dict: (192, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 101
input_layer0: layer_id=99: KerasTensor(type_spec=TensorSpec(shape=(1, 14, 14, 192), dtype=tf.float32, name=None), name='p_re_lu_23/add:0', description="created by layer 'p_re_lu_23'")
input_layer1: layer_id=100: Const(ndarray).shape (192, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='depthwise_conv2d_5/depthwise:0', description="created by layer 'depthwise_conv2d_5'")
====================================================================================
layer_type: Const
layer_id: 102
tf_layers_dict: (1, 192, 1, 1)
====================================================================================
layer_type: Add
layer_id: 103
input_layer0: layer_id=101: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='depthwise_conv2d_5/depthwise:0', description="created by layer 'depthwise_conv2d_5'")
input_layer1: layer_id=102: Const(ndarray).shape (1, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='tf.math.add_21/Add:0', description="created by layer 'tf.math.add_21'")
====================================================================================
layer_type: Const
layer_id: 104
tf_layers_dict: (192, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 105
input_layer0: layer_id=103: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='tf.math.add_21/Add:0', description="created by layer 'tf.math.add_21'")
input_layer1: layer_id=104: Const(ndarray).shape (192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='p_re_lu_25/add:0', description="created by layer 'p_re_lu_25'")
====================================================================================
layer_type: Const
layer_id: 106
tf_layers_dict: (64, 192, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 107
input_layer0: layer_id=105: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 192), dtype=tf.float32, name=None), name='p_re_lu_25/add:0', description="created by layer 'p_re_lu_25'")
input_layer1: layer_id=106: Const(ndarray).shape (64, 192, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='conv2d_13/Conv2D:0', description="created by layer 'conv2d_13'")
====================================================================================
layer_type: Const
layer_id: 108
tf_layers_dict: (1, 64, 1, 1)
====================================================================================
layer_type: Add
layer_id: 109
input_layer0: layer_id=107: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='conv2d_13/Conv2D:0', description="created by layer 'conv2d_13'")
input_layer1: layer_id=108: Const(ndarray).shape (1, 64, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_22/Add:0', description="created by layer 'tf.math.add_22'")
====================================================================================
layer_type: Const
layer_id: 110
tf_layers_dict: (384, 64, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 111
input_layer0: layer_id=109: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_22/Add:0', description="created by layer 'tf.math.add_22'")
input_layer1: layer_id=110: Const(ndarray).shape (384, 64, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='conv2d_14/Conv2D:0', description="created by layer 'conv2d_14'")
====================================================================================
layer_type: Const
layer_id: 112
tf_layers_dict: (1, 384, 1, 1)
====================================================================================
layer_type: Add
layer_id: 113
input_layer0: layer_id=111: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='conv2d_14/Conv2D:0', description="created by layer 'conv2d_14'")
input_layer1: layer_id=112: Const(ndarray).shape (1, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_23/Add:0', description="created by layer 'tf.math.add_23'")
====================================================================================
layer_type: Const
layer_id: 114
tf_layers_dict: (384, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 115
input_layer0: layer_id=113: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_23/Add:0', description="created by layer 'tf.math.add_23'")
input_layer1: layer_id=114: Const(ndarray).shape (384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_27/add:0', description="created by layer 'p_re_lu_27'")
====================================================================================
layer_type: Const
layer_id: 116
tf_layers_dict: (384, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 117
input_layer0: layer_id=115: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_27/add:0', description="created by layer 'p_re_lu_27'")
input_layer1: layer_id=116: Const(ndarray).shape (384, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='depthwise_conv2d_6/depthwise:0', description="created by layer 'depthwise_conv2d_6'")
====================================================================================
layer_type: Const
layer_id: 118
tf_layers_dict: (1, 384, 1, 1)
====================================================================================
layer_type: Add
layer_id: 119
input_layer0: layer_id=117: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='depthwise_conv2d_6/depthwise:0', description="created by layer 'depthwise_conv2d_6'")
input_layer1: layer_id=118: Const(ndarray).shape (1, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_24/Add:0', description="created by layer 'tf.math.add_24'")
====================================================================================
layer_type: Const
layer_id: 120
tf_layers_dict: (384, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 121
input_layer0: layer_id=119: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_24/Add:0', description="created by layer 'tf.math.add_24'")
input_layer1: layer_id=120: Const(ndarray).shape (384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_29/add:0', description="created by layer 'p_re_lu_29'")
====================================================================================
layer_type: Const
layer_id: 122
tf_layers_dict: (64, 384, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 123
input_layer0: layer_id=121: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_29/add:0', description="created by layer 'p_re_lu_29'")
input_layer1: layer_id=122: Const(ndarray).shape (64, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='conv2d_15/Conv2D:0', description="created by layer 'conv2d_15'")
====================================================================================
layer_type: Const
layer_id: 124
tf_layers_dict: (1, 64, 1, 1)
====================================================================================
layer_type: Add
layer_id: 125
input_layer0: layer_id=123: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='conv2d_15/Conv2D:0', description="created by layer 'conv2d_15'")
input_layer1: layer_id=124: Const(ndarray).shape (1, 64, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_25/Add:0', description="created by layer 'tf.math.add_25'")
====================================================================================
layer_type: Add
layer_id: 126
input_layer0: layer_id=109: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_22/Add:0', description="created by layer 'tf.math.add_22'")
input_layer1: layer_id=125: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_25/Add:0', description="created by layer 'tf.math.add_25'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_26/Add:0', description="created by layer 'tf.math.add_26'")
====================================================================================
layer_type: Const
layer_id: 127
tf_layers_dict: (384, 64, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 128
input_layer0: layer_id=126: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_26/Add:0', description="created by layer 'tf.math.add_26'")
input_layer1: layer_id=127: Const(ndarray).shape (384, 64, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='conv2d_16/Conv2D:0', description="created by layer 'conv2d_16'")
====================================================================================
layer_type: Const
layer_id: 129
tf_layers_dict: (1, 384, 1, 1)
====================================================================================
layer_type: Add
layer_id: 130
input_layer0: layer_id=128: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='conv2d_16/Conv2D:0', description="created by layer 'conv2d_16'")
input_layer1: layer_id=129: Const(ndarray).shape (1, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_27/Add:0', description="created by layer 'tf.math.add_27'")
====================================================================================
layer_type: Const
layer_id: 131
tf_layers_dict: (384, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 132
input_layer0: layer_id=130: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_27/Add:0', description="created by layer 'tf.math.add_27'")
input_layer1: layer_id=131: Const(ndarray).shape (384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_31/add:0', description="created by layer 'p_re_lu_31'")
====================================================================================
layer_type: Const
layer_id: 133
tf_layers_dict: (384, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 134
input_layer0: layer_id=132: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_31/add:0', description="created by layer 'p_re_lu_31'")
input_layer1: layer_id=133: Const(ndarray).shape (384, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='depthwise_conv2d_7/depthwise:0', description="created by layer 'depthwise_conv2d_7'")
====================================================================================
layer_type: Const
layer_id: 135
tf_layers_dict: (1, 384, 1, 1)
====================================================================================
layer_type: Add
layer_id: 136
input_layer0: layer_id=134: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='depthwise_conv2d_7/depthwise:0', description="created by layer 'depthwise_conv2d_7'")
input_layer1: layer_id=135: Const(ndarray).shape (1, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_28/Add:0', description="created by layer 'tf.math.add_28'")
====================================================================================
layer_type: Const
layer_id: 137
tf_layers_dict: (384, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 138
input_layer0: layer_id=136: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='tf.math.add_28/Add:0', description="created by layer 'tf.math.add_28'")
input_layer1: layer_id=137: Const(ndarray).shape (384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_33/add:0', description="created by layer 'p_re_lu_33'")
====================================================================================
layer_type: Const
layer_id: 139
tf_layers_dict: (96, 384, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 140
input_layer0: layer_id=138: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 384), dtype=tf.float32, name=None), name='p_re_lu_33/add:0', description="created by layer 'p_re_lu_33'")
input_layer1: layer_id=139: Const(ndarray).shape (96, 384, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_17/Conv2D:0', description="created by layer 'conv2d_17'")
====================================================================================
layer_type: Const
layer_id: 141
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 142
input_layer0: layer_id=140: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_17/Conv2D:0', description="created by layer 'conv2d_17'")
input_layer1: layer_id=141: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_29/Add:0', description="created by layer 'tf.math.add_29'")
====================================================================================
layer_type: Const
layer_id: 143
tf_layers_dict: (96, 64, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 144
input_layer0: layer_id=126: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 64), dtype=tf.float32, name=None), name='tf.math.add_26/Add:0', description="created by layer 'tf.math.add_26'")
input_layer1: layer_id=143: Const(ndarray).shape (96, 64, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_18/Conv2D:0', description="created by layer 'conv2d_18'")
====================================================================================
layer_type: Const
layer_id: 145
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 146
input_layer0: layer_id=144: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_18/Conv2D:0', description="created by layer 'conv2d_18'")
input_layer1: layer_id=145: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_30/Add:0', description="created by layer 'tf.math.add_30'")
====================================================================================
layer_type: Add
layer_id: 147
input_layer0: layer_id=142: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_29/Add:0', description="created by layer 'tf.math.add_29'")
input_layer1: layer_id=146: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_30/Add:0', description="created by layer 'tf.math.add_30'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_31/Add:0', description="created by layer 'tf.math.add_31'")
====================================================================================
layer_type: Const
layer_id: 148
tf_layers_dict: (576, 96, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 149
input_layer0: layer_id=147: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_31/Add:0', description="created by layer 'tf.math.add_31'")
input_layer1: layer_id=148: Const(ndarray).shape (576, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_19/Conv2D:0', description="created by layer 'conv2d_19'")
====================================================================================
layer_type: Const
layer_id: 150
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 151
input_layer0: layer_id=149: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_19/Conv2D:0', description="created by layer 'conv2d_19'")
input_layer1: layer_id=150: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_32/Add:0', description="created by layer 'tf.math.add_32'")
====================================================================================
layer_type: Const
layer_id: 152
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 153
input_layer0: layer_id=151: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_32/Add:0', description="created by layer 'tf.math.add_32'")
input_layer1: layer_id=152: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_35/add:0', description="created by layer 'p_re_lu_35'")
====================================================================================
layer_type: Const
layer_id: 154
tf_layers_dict: (576, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 155
input_layer0: layer_id=153: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_35/add:0', description="created by layer 'p_re_lu_35'")
input_layer1: layer_id=154: Const(ndarray).shape (576, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_8/depthwise:0', description="created by layer 'depthwise_conv2d_8'")
====================================================================================
layer_type: Const
layer_id: 156
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 157
input_layer0: layer_id=155: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_8/depthwise:0', description="created by layer 'depthwise_conv2d_8'")
input_layer1: layer_id=156: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_33/Add:0', description="created by layer 'tf.math.add_33'")
====================================================================================
layer_type: Const
layer_id: 158
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 159
input_layer0: layer_id=157: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_33/Add:0', description="created by layer 'tf.math.add_33'")
input_layer1: layer_id=158: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_37/add:0', description="created by layer 'p_re_lu_37'")
====================================================================================
layer_type: Const
layer_id: 160
tf_layers_dict: (96, 576, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 161
input_layer0: layer_id=159: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_37/add:0', description="created by layer 'p_re_lu_37'")
input_layer1: layer_id=160: Const(ndarray).shape (96, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_20/Conv2D:0', description="created by layer 'conv2d_20'")
====================================================================================
layer_type: Const
layer_id: 162
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 163
input_layer0: layer_id=161: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_20/Conv2D:0', description="created by layer 'conv2d_20'")
input_layer1: layer_id=162: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_34/Add:0', description="created by layer 'tf.math.add_34'")
====================================================================================
layer_type: Add
layer_id: 164
input_layer0: layer_id=147: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_31/Add:0', description="created by layer 'tf.math.add_31'")
input_layer1: layer_id=163: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_34/Add:0', description="created by layer 'tf.math.add_34'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_35/Add:0', description="created by layer 'tf.math.add_35'")
====================================================================================
layer_type: Const
layer_id: 165
tf_layers_dict: (576, 96, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 166
input_layer0: layer_id=164: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_35/Add:0', description="created by layer 'tf.math.add_35'")
input_layer1: layer_id=165: Const(ndarray).shape (576, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_21/Conv2D:0', description="created by layer 'conv2d_21'")
====================================================================================
layer_type: Const
layer_id: 167
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 168
input_layer0: layer_id=166: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_21/Conv2D:0', description="created by layer 'conv2d_21'")
input_layer1: layer_id=167: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_36/Add:0', description="created by layer 'tf.math.add_36'")
====================================================================================
layer_type: Const
layer_id: 169
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 170
input_layer0: layer_id=168: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_36/Add:0', description="created by layer 'tf.math.add_36'")
input_layer1: layer_id=169: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_39/add:0', description="created by layer 'p_re_lu_39'")
====================================================================================
layer_type: Const
layer_id: 171
tf_layers_dict: (576, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 172
input_layer0: layer_id=170: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_39/add:0', description="created by layer 'p_re_lu_39'")
input_layer1: layer_id=171: Const(ndarray).shape (576, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_9/depthwise:0', description="created by layer 'depthwise_conv2d_9'")
====================================================================================
layer_type: Const
layer_id: 173
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 174
input_layer0: layer_id=172: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_9/depthwise:0', description="created by layer 'depthwise_conv2d_9'")
input_layer1: layer_id=173: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_37/Add:0', description="created by layer 'tf.math.add_37'")
====================================================================================
layer_type: Const
layer_id: 175
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 176
input_layer0: layer_id=174: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_37/Add:0', description="created by layer 'tf.math.add_37'")
input_layer1: layer_id=175: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_41/add:0', description="created by layer 'p_re_lu_41'")
====================================================================================
layer_type: Const
layer_id: 177
tf_layers_dict: (96, 576, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 178
input_layer0: layer_id=176: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_41/add:0', description="created by layer 'p_re_lu_41'")
input_layer1: layer_id=177: Const(ndarray).shape (96, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_22/Conv2D:0', description="created by layer 'conv2d_22'")
====================================================================================
layer_type: Const
layer_id: 179
tf_layers_dict: (1, 96, 1, 1)
====================================================================================
layer_type: Add
layer_id: 180
input_layer0: layer_id=178: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='conv2d_22/Conv2D:0', description="created by layer 'conv2d_22'")
input_layer1: layer_id=179: Const(ndarray).shape (1, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_38/Add:0', description="created by layer 'tf.math.add_38'")
====================================================================================
layer_type: Add
layer_id: 181
input_layer0: layer_id=164: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_35/Add:0', description="created by layer 'tf.math.add_35'")
input_layer1: layer_id=180: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_38/Add:0', description="created by layer 'tf.math.add_38'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_39/Add:0', description="created by layer 'tf.math.add_39'")
====================================================================================
layer_type: Const
layer_id: 182
tf_layers_dict: (576, 96, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 183
input_layer0: layer_id=181: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 96), dtype=tf.float32, name=None), name='tf.math.add_39/Add:0', description="created by layer 'tf.math.add_39'")
input_layer1: layer_id=182: Const(ndarray).shape (576, 96, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_23/Conv2D:0', description="created by layer 'conv2d_23'")
====================================================================================
layer_type: Const
layer_id: 184
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 185
input_layer0: layer_id=183: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='conv2d_23/Conv2D:0', description="created by layer 'conv2d_23'")
input_layer1: layer_id=184: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_40/Add:0', description="created by layer 'tf.math.add_40'")
====================================================================================
layer_type: Const
layer_id: 186
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 187
input_layer0: layer_id=185: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='tf.math.add_40/Add:0', description="created by layer 'tf.math.add_40'")
input_layer1: layer_id=186: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_43/add:0', description="created by layer 'p_re_lu_43'")
====================================================================================
layer_type: Const
layer_id: 188
tf_layers_dict: (576, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 189
input_layer0: layer_id=187: KerasTensor(type_spec=TensorSpec(shape=(1, 7, 7, 576), dtype=tf.float32, name=None), name='p_re_lu_43/add:0', description="created by layer 'p_re_lu_43'")
input_layer1: layer_id=188: Const(ndarray).shape (576, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_10/depthwise:0', description="created by layer 'depthwise_conv2d_10'")
====================================================================================
layer_type: Const
layer_id: 190
tf_layers_dict: (1, 576, 1, 1)
====================================================================================
layer_type: Add
layer_id: 191
input_layer0: layer_id=189: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='depthwise_conv2d_10/depthwise:0', description="created by layer 'depthwise_conv2d_10'")
input_layer1: layer_id=190: Const(ndarray).shape (1, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='tf.math.add_41/Add:0', description="created by layer 'tf.math.add_41'")
====================================================================================
layer_type: Const
layer_id: 192
tf_layers_dict: (576, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 193
input_layer0: layer_id=191: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='tf.math.add_41/Add:0', description="created by layer 'tf.math.add_41'")
input_layer1: layer_id=192: Const(ndarray).shape (576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='p_re_lu_45/add:0', description="created by layer 'p_re_lu_45'")
====================================================================================
layer_type: Const
layer_id: 194
tf_layers_dict: (160, 576, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 195
input_layer0: layer_id=193: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 576), dtype=tf.float32, name=None), name='p_re_lu_45/add:0', description="created by layer 'p_re_lu_45'")
input_layer1: layer_id=194: Const(ndarray).shape (160, 576, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_24/Conv2D:0', description="created by layer 'conv2d_24'")
====================================================================================
layer_type: Const
layer_id: 196
tf_layers_dict: (1, 160, 1, 1)
====================================================================================
layer_type: Add
layer_id: 197
input_layer0: layer_id=195: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_24/Conv2D:0', description="created by layer 'conv2d_24'")
input_layer1: layer_id=196: Const(ndarray).shape (1, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_42/Add:0', description="created by layer 'tf.math.add_42'")
====================================================================================
layer_type: Const
layer_id: 198
tf_layers_dict: (960, 160, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 199
input_layer0: layer_id=197: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_42/Add:0', description="created by layer 'tf.math.add_42'")
input_layer1: layer_id=198: Const(ndarray).shape (960, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_25/Conv2D:0', description="created by layer 'conv2d_25'")
====================================================================================
layer_type: Const
layer_id: 200
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 201
input_layer0: layer_id=199: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_25/Conv2D:0', description="created by layer 'conv2d_25'")
input_layer1: layer_id=200: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_43/Add:0', description="created by layer 'tf.math.add_43'")
====================================================================================
layer_type: Const
layer_id: 202
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 203
input_layer0: layer_id=201: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_43/Add:0', description="created by layer 'tf.math.add_43'")
input_layer1: layer_id=202: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_47/add:0', description="created by layer 'p_re_lu_47'")
====================================================================================
layer_type: Const
layer_id: 204
tf_layers_dict: (960, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 205
input_layer0: layer_id=203: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_47/add:0', description="created by layer 'p_re_lu_47'")
input_layer1: layer_id=204: Const(ndarray).shape (960, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_11/depthwise:0', description="created by layer 'depthwise_conv2d_11'")
====================================================================================
layer_type: Const
layer_id: 206
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 207
input_layer0: layer_id=205: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_11/depthwise:0', description="created by layer 'depthwise_conv2d_11'")
input_layer1: layer_id=206: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_44/Add:0', description="created by layer 'tf.math.add_44'")
====================================================================================
layer_type: Const
layer_id: 208
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 209
input_layer0: layer_id=207: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_44/Add:0', description="created by layer 'tf.math.add_44'")
input_layer1: layer_id=208: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_49/add:0', description="created by layer 'p_re_lu_49'")
====================================================================================
layer_type: Const
layer_id: 210
tf_layers_dict: (160, 960, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 211
input_layer0: layer_id=209: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_49/add:0', description="created by layer 'p_re_lu_49'")
input_layer1: layer_id=210: Const(ndarray).shape (160, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_26/Conv2D:0', description="created by layer 'conv2d_26'")
====================================================================================
layer_type: Const
layer_id: 212
tf_layers_dict: (1, 160, 1, 1)
====================================================================================
layer_type: Add
layer_id: 213
input_layer0: layer_id=211: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_26/Conv2D:0', description="created by layer 'conv2d_26'")
input_layer1: layer_id=212: Const(ndarray).shape (1, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_45/Add:0', description="created by layer 'tf.math.add_45'")
====================================================================================
layer_type: Add
layer_id: 214
input_layer0: layer_id=197: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_42/Add:0', description="created by layer 'tf.math.add_42'")
input_layer1: layer_id=213: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_45/Add:0', description="created by layer 'tf.math.add_45'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_46/Add:0', description="created by layer 'tf.math.add_46'")
====================================================================================
layer_type: Const
layer_id: 215
tf_layers_dict: (960, 160, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 216
input_layer0: layer_id=214: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_46/Add:0', description="created by layer 'tf.math.add_46'")
input_layer1: layer_id=215: Const(ndarray).shape (960, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_27/Conv2D:0', description="created by layer 'conv2d_27'")
====================================================================================
layer_type: Const
layer_id: 217
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 218
input_layer0: layer_id=216: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_27/Conv2D:0', description="created by layer 'conv2d_27'")
input_layer1: layer_id=217: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_47/Add:0', description="created by layer 'tf.math.add_47'")
====================================================================================
layer_type: Const
layer_id: 219
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 220
input_layer0: layer_id=218: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_47/Add:0', description="created by layer 'tf.math.add_47'")
input_layer1: layer_id=219: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_51/add:0', description="created by layer 'p_re_lu_51'")
====================================================================================
layer_type: Const
layer_id: 221
tf_layers_dict: (960, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 222
input_layer0: layer_id=220: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_51/add:0', description="created by layer 'p_re_lu_51'")
input_layer1: layer_id=221: Const(ndarray).shape (960, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_12/depthwise:0', description="created by layer 'depthwise_conv2d_12'")
====================================================================================
layer_type: Const
layer_id: 223
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 224
input_layer0: layer_id=222: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_12/depthwise:0', description="created by layer 'depthwise_conv2d_12'")
input_layer1: layer_id=223: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_48/Add:0', description="created by layer 'tf.math.add_48'")
====================================================================================
layer_type: Const
layer_id: 225
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 226
input_layer0: layer_id=224: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_48/Add:0', description="created by layer 'tf.math.add_48'")
input_layer1: layer_id=225: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_53/add:0', description="created by layer 'p_re_lu_53'")
====================================================================================
layer_type: Const
layer_id: 227
tf_layers_dict: (160, 960, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 228
input_layer0: layer_id=226: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_53/add:0', description="created by layer 'p_re_lu_53'")
input_layer1: layer_id=227: Const(ndarray).shape (160, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_28/Conv2D:0', description="created by layer 'conv2d_28'")
====================================================================================
layer_type: Const
layer_id: 229
tf_layers_dict: (1, 160, 1, 1)
====================================================================================
layer_type: Add
layer_id: 230
input_layer0: layer_id=228: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='conv2d_28/Conv2D:0', description="created by layer 'conv2d_28'")
input_layer1: layer_id=229: Const(ndarray).shape (1, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_49/Add:0', description="created by layer 'tf.math.add_49'")
====================================================================================
layer_type: Add
layer_id: 231
input_layer0: layer_id=214: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_46/Add:0', description="created by layer 'tf.math.add_46'")
input_layer1: layer_id=230: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_49/Add:0', description="created by layer 'tf.math.add_49'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_50/Add:0', description="created by layer 'tf.math.add_50'")
====================================================================================
layer_type: Const
layer_id: 232
tf_layers_dict: (960, 160, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 233
input_layer0: layer_id=231: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_50/Add:0', description="created by layer 'tf.math.add_50'")
input_layer1: layer_id=232: Const(ndarray).shape (960, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_29/Conv2D:0', description="created by layer 'conv2d_29'")
====================================================================================
layer_type: Const
layer_id: 234
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 235
input_layer0: layer_id=233: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='conv2d_29/Conv2D:0', description="created by layer 'conv2d_29'")
input_layer1: layer_id=234: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_51/Add:0', description="created by layer 'tf.math.add_51'")
====================================================================================
layer_type: Const
layer_id: 236
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 237
input_layer0: layer_id=235: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_51/Add:0', description="created by layer 'tf.math.add_51'")
input_layer1: layer_id=236: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_55/add:0', description="created by layer 'p_re_lu_55'")
====================================================================================
layer_type: Const
layer_id: 238
tf_layers_dict: (960, 1, 1, 3, 3)
====================================================================================
layer_type: GroupConvolution
layer_id: 239
input_layer0: layer_id=237: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_55/add:0', description="created by layer 'p_re_lu_55'")
input_layer1: layer_id=238: Const(ndarray).shape (960, 1, 1, 3, 3)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_13/depthwise:0', description="created by layer 'depthwise_conv2d_13'")
====================================================================================
layer_type: Const
layer_id: 240
tf_layers_dict: (1, 960, 1, 1)
====================================================================================
layer_type: Add
layer_id: 241
input_layer0: layer_id=239: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='depthwise_conv2d_13/depthwise:0', description="created by layer 'depthwise_conv2d_13'")
input_layer1: layer_id=240: Const(ndarray).shape (1, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_52/Add:0', description="created by layer 'tf.math.add_52'")
====================================================================================
layer_type: Const
layer_id: 242
tf_layers_dict: (960, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 243
input_layer0: layer_id=241: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='tf.math.add_52/Add:0', description="created by layer 'tf.math.add_52'")
input_layer1: layer_id=242: Const(ndarray).shape (960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_57/add:0', description="created by layer 'p_re_lu_57'")
====================================================================================
layer_type: Const
layer_id: 244
tf_layers_dict: (240, 960, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 245
input_layer0: layer_id=243: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 960), dtype=tf.float32, name=None), name='p_re_lu_57/add:0', description="created by layer 'p_re_lu_57'")
input_layer1: layer_id=244: Const(ndarray).shape (240, 960, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='conv2d_30/Conv2D:0', description="created by layer 'conv2d_30'")
====================================================================================
layer_type: Const
layer_id: 246
tf_layers_dict: (1, 240, 1, 1)
====================================================================================
layer_type: Add
layer_id: 247
input_layer0: layer_id=245: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='conv2d_30/Conv2D:0', description="created by layer 'conv2d_30'")
input_layer1: layer_id=246: Const(ndarray).shape (1, 240, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_53/Add:0', description="created by layer 'tf.math.add_53'")
====================================================================================
layer_type: Const
layer_id: 248
tf_layers_dict: (240, 160, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 249
input_layer0: layer_id=231: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 160), dtype=tf.float32, name=None), name='tf.math.add_50/Add:0', description="created by layer 'tf.math.add_50'")
input_layer1: layer_id=248: Const(ndarray).shape (240, 160, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='conv2d_31/Conv2D:0', description="created by layer 'conv2d_31'")
====================================================================================
layer_type: Const
layer_id: 250
tf_layers_dict: (1, 240, 1, 1)
====================================================================================
layer_type: Add
layer_id: 251
input_layer0: layer_id=249: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='conv2d_31/Conv2D:0', description="created by layer 'conv2d_31'")
input_layer1: layer_id=250: Const(ndarray).shape (1, 240, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_54/Add:0', description="created by layer 'tf.math.add_54'")
====================================================================================
layer_type: Add
layer_id: 252
input_layer0: layer_id=247: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_53/Add:0', description="created by layer 'tf.math.add_53'")
input_layer1: layer_id=251: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_54/Add:0', description="created by layer 'tf.math.add_54'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_55/Add:0', description="created by layer 'tf.math.add_55'")
====================================================================================
layer_type: Const
layer_id: 253
tf_layers_dict: (512, 240, 1, 1)
====================================================================================
layer_type: Convolution
layer_id: 254
input_layer0: layer_id=252: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 240), dtype=tf.float32, name=None), name='tf.math.add_55/Add:0', description="created by layer 'tf.math.add_55'")
input_layer1: layer_id=253: Const(ndarray).shape (512, 240, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='conv2d_32/Conv2D:0', description="created by layer 'conv2d_32'")
====================================================================================
layer_type: Const
layer_id: 255
tf_layers_dict: (1, 512, 1, 1)
====================================================================================
layer_type: Add
layer_id: 256
input_layer0: layer_id=254: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='conv2d_32/Conv2D:0', description="created by layer 'conv2d_32'")
input_layer1: layer_id=255: Const(ndarray).shape (1, 512, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='tf.math.add_56/Add:0', description="created by layer 'tf.math.add_56'")
====================================================================================
layer_type: Const
layer_id: 257
tf_layers_dict: (512, 1, 1)
====================================================================================
layer_type: PReLU
layer_id: 258
input_layer0: layer_id=256: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='tf.math.add_56/Add:0', description="created by layer 'tf.math.add_56'")
input_layer1: layer_id=257: Const(ndarray).shape (512, 1, 1)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='p_re_lu_59/add:0', description="created by layer 'p_re_lu_59'")
====================================================================================
layer_type: AvgPool
layer_id: 259
input_layer0: layer_id=258: KerasTensor(type_spec=TensorSpec(shape=(1, 4, 4, 512), dtype=tf.float32, name=None), name='p_re_lu_59/add:0', description="created by layer 'p_re_lu_59'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 1, 1, 512), dtype=tf.float32, name=None), name='average_pooling2d/AvgPool:0', description="created by layer 'average_pooling2d'")
====================================================================================
layer_type: ShapeOf
layer_id: 260
input_layer0: layer_id=259: KerasTensor(type_spec=TensorSpec(shape=(1, 1, 1, 512), dtype=tf.float32, name=None), name='average_pooling2d/AvgPool:0', description="created by layer 'average_pooling2d'")
tf_layers_dict: [  1   1   1 512]
====================================================================================
layer_type: Const
layer_id: 261
tf_layers_dict: (1,)
====================================================================================
layer_type: Const
layer_id: 262
tf_layers_dict: (1,)
====================================================================================
layer_type: Gather
layer_id: 263
input_layer0: layer_id=260: [  1   1   1 512]
input_layer1: layer_id=261: Const(ndarray).shape (1,)
input_layer2: layer_id=262: Const(ndarray).shape (1,)
tf_layers_dict: [1]
====================================================================================
layer_type: Const
layer_id: 264
tf_layers_dict: (1,)
====================================================================================
layer_type: Unsqueeze
layer_id: 265
input_layer0: layer_id=263: [1]
input_layer1: layer_id=264: Const(ndarray).shape (1,)
tf_layers_dict: [1]
====================================================================================
layer_type: Const
layer_id: 266
tf_layers_dict: (1,)
====================================================================================
layer_type: Concat
layer_id: 267
input_layer0: layer_id=265: [1]
input_layer1: layer_id=266: Const(ndarray).shape (1,)
tf_layers_dict: [ 1 -1]
====================================================================================
layer_type: Reshape
layer_id: 268
input_layer0: layer_id=259: KerasTensor(type_spec=TensorSpec(shape=(1, 1, 1, 512), dtype=tf.float32, name=None), name='average_pooling2d/AvgPool:0', description="created by layer 'average_pooling2d'")
input_layer1: layer_id=267: [ 1 -1]
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 512), dtype=tf.float32, name=None), name='tf.reshape/Reshape:0', description="created by layer 'tf.reshape'")
====================================================================================
layer_type: Const
layer_id: 269
tf_layers_dict: (256, 512)
====================================================================================
layer_type: MatMul
layer_id: 270
input_layer0: layer_id=268: KerasTensor(type_spec=TensorSpec(shape=(1, 512), dtype=tf.float32, name=None), name='tf.reshape/Reshape:0', description="created by layer 'tf.reshape'")
input_layer1: layer_id=269: Const(ndarray).shape (256, 512)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 256), dtype=tf.float32, name=None), name='tf.linalg.matmul/MatMul:0', description="created by layer 'tf.linalg.matmul'")
====================================================================================
layer_type: Const
layer_id: 271
tf_layers_dict: (1, 256)
====================================================================================
layer_type: Add
layer_id: 272
input_layer0: layer_id=270: KerasTensor(type_spec=TensorSpec(shape=(1, 256), dtype=tf.float32, name=None), name='tf.linalg.matmul/MatMul:0', description="created by layer 'tf.linalg.matmul'")
input_layer1: layer_id=271: Const(ndarray).shape (1, 256)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 256), dtype=tf.float32, name=None), name='tf.math.add_57/Add:0', description="created by layer 'tf.math.add_57'")
====================================================================================
layer_type: Const
layer_id: 273
tf_layers_dict: (2, 256)
====================================================================================
layer_type: MatMul
layer_id: 274
input_layer0: layer_id=272: KerasTensor(type_spec=TensorSpec(shape=(1, 256), dtype=tf.float32, name=None), name='tf.math.add_57/Add:0', description="created by layer 'tf.math.add_57'")
input_layer1: layer_id=273: Const(ndarray).shape (2, 256)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 2), dtype=tf.float32, name=None), name='tf.linalg.matmul_1/MatMul:0', description="created by layer 'tf.linalg.matmul_1'")
====================================================================================
layer_type: Const
layer_id: 275
tf_layers_dict: (1, 2)
====================================================================================
layer_type: Add
layer_id: 276
input_layer0: layer_id=274: KerasTensor(type_spec=TensorSpec(shape=(1, 2), dtype=tf.float32, name=None), name='tf.linalg.matmul_1/MatMul:0', description="created by layer 'tf.linalg.matmul_1'")
input_layer1: layer_id=275: Const(ndarray).shape (1, 2)
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 2), dtype=tf.float32, name=None), name='tf.math.add_58/Add:0', description="created by layer 'tf.math.add_58'")
====================================================================================
layer_type: Result
layer_id: 277
input_layer0: layer_id=276: KerasTensor(type_spec=TensorSpec(shape=(1, 2), dtype=tf.float32, name=None), name='tf.math.add_58/Add:0', description="created by layer 'tf.math.add_58'")
tf_layers_dict: KerasTensor(type_spec=TensorSpec(shape=(1, 2), dtype=tf.float32, name=None), name='tf.identity/Identity:0', description="created by layer 'tf.identity'")
====================================================================================
TensorFlow/Keras model building process complete!
tflite Float32 convertion started ===================================================
tflite Float32 convertion complete! - saved_model/model_float32.tflite
All the conversion process is finished! =============================================

https://github.com/PINTO0309/openvino2tensorflow/blob/40684a2990503b6da73637e1aa19f887afee0408/openvino2tensorflow/openvino2tensorflow.py#L1169-L1178