652 lines
115 KiB
Python
652 lines
115 KiB
Python
|
import numpy as np
|
||
|
|
import torch
|
||
|
|
import torch.nn as nn
|
||
|
|
import torch.nn.functional as F
|
||
|
|
import math
|
||
|
|
|
||
|
|
_weights_dict = dict()
|
||
|
|
|
||
|
|
def load_weights(weight_file):
|
||
|
|
if weight_file == None:
|
||
|
|
return
|
||
|
|
|
||
|
|
try:
|
||
|
|
weights_dict = np.load(weight_file, allow_pickle=True).item()
|
||
|
|
except:
|
||
|
|
weights_dict = np.load(weight_file, allow_pickle=True, encoding='bytes').item()
|
||
|
|
|
||
|
|
return weights_dict
|
||
|
|
|
||
|
|
class KitModel(nn.Module):
|
||
|
|
|
||
|
|
|
||
|
|
def __init__(self, weight_file):
|
||
|
|
super(KitModel, self).__init__()
|
||
|
|
global _weights_dict
|
||
|
|
_weights_dict = load_weights(weight_file)
|
||
|
|
|
||
|
|
self.resnet_v2_101_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/conv1/Conv2D', in_channels=3, out_channels=64, kernel_size=(7, 7), stride=(2, 2), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_1/bottleneck_v2/preact/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_shortcut_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_1/bottleneck_v2/shortcut/Conv2D', in_channels=64, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_1/bottleneck_v2/conv1/Conv2D', in_channels=64, out_channels=64, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_1/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_1/bottleneck_v2/conv2/Conv2D', in_channels=64, out_channels=64, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_1/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_1/bottleneck_v2/conv3/Conv2D', in_channels=64, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_2/bottleneck_v2/preact/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_2/bottleneck_v2/conv1/Conv2D', in_channels=256, out_channels=64, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_2/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_2/bottleneck_v2/conv2/Conv2D', in_channels=64, out_channels=64, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_2/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_2/bottleneck_v2/conv3/Conv2D', in_channels=64, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_3/bottleneck_v2/preact/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_3/bottleneck_v2/conv1/Conv2D', in_channels=256, out_channels=64, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_3/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_3/bottleneck_v2/conv2/Conv2D', in_channels=64, out_channels=64, kernel_size=(3, 3), stride=(2, 2), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block1/unit_3/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=64, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block1/unit_3/bottleneck_v2/conv3/Conv2D', in_channels=64, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_1/bottleneck_v2/preact/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_shortcut_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_1/bottleneck_v2/shortcut/Conv2D', in_channels=256, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_1/bottleneck_v2/conv1/Conv2D', in_channels=256, out_channels=128, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_1/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_1/bottleneck_v2/conv2/Conv2D', in_channels=128, out_channels=128, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_1/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_1/bottleneck_v2/conv3/Conv2D', in_channels=128, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_2/bottleneck_v2/preact/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_2/bottleneck_v2/conv1/Conv2D', in_channels=512, out_channels=128, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_2/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_2/bottleneck_v2/conv2/Conv2D', in_channels=128, out_channels=128, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_2/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_2/bottleneck_v2/conv3/Conv2D', in_channels=128, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_3/bottleneck_v2/preact/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_3/bottleneck_v2/conv1/Conv2D', in_channels=512, out_channels=128, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_3/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_3/bottleneck_v2/conv2/Conv2D', in_channels=128, out_channels=128, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_3/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_3/bottleneck_v2/conv3/Conv2D', in_channels=128, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_4/bottleneck_v2/preact/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_4/bottleneck_v2/conv1/Conv2D', in_channels=512, out_channels=128, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_4/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_4/bottleneck_v2/conv2/Conv2D', in_channels=128, out_channels=128, kernel_size=(3, 3), stride=(2, 2), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block2/unit_4/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=128, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block2/unit_4/bottleneck_v2/conv3/Conv2D', in_channels=128, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_1/bottleneck_v2/preact/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_shortcut_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_1/bottleneck_v2/shortcut/Conv2D', in_channels=512, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_1/bottleneck_v2/conv1/Conv2D', in_channels=512, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_1/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_1/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_1/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_1/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_2/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_2/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_2/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_2/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_2/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_2/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_3/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_3/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_3/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_3/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_3/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_3/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_4/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_4/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_4/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_4/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_4/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_4/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_5/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_5/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_5/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_5/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_5/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_5/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_6/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_6/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_6/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_6/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_6/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_6/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_7/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_7/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_7/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_7/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_7/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_7/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_8/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_8/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_8/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_8/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_8/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_8/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_9/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_9/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_9/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_9/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_9/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_9/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_10/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_10/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_10/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_10/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_10/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_10/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_11/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_11/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_11/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_11/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_11/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_11/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_12/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_12/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_12/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_12/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_12/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_12/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_13/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_13/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_13/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_13/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_13/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_13/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_14/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_14/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_14/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_14/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_14/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_14/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_15/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_15/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_15/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_15/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_15/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_15/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_16/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_16/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_16/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_16/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_16/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_16/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_17/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_17/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_17/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_17/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_17/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_17/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_18/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_18/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_18/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_18/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_18/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_18/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_19/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_19/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_19/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_19/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_19/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_19/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_20/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_20/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_20/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_20/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_20/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_20/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_21/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_21/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_21/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_21/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_21/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_21/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_22/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_22/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_22/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_22/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_22/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_22/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_23/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_23/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=256, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_23/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_23/bottleneck_v2/conv2/Conv2D', in_channels=256, out_channels=256, kernel_size=(3, 3), stride=(2, 2), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block3/unit_23/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=256, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block3/unit_23/bottleneck_v2/conv3/Conv2D', in_channels=256, out_channels=1024, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_1/bottleneck_v2/preact/FusedBatchNorm', num_features=1024, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_shortcut_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_1/bottleneck_v2/shortcut/Conv2D', in_channels=1024, out_channels=2048, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_1/bottleneck_v2/conv1/Conv2D', in_channels=1024, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_1/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_1/bottleneck_v2/conv2/Conv2D', in_channels=512, out_channels=512, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_1/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_1/bottleneck_v2/conv3/Conv2D', in_channels=512, out_channels=2048, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_2/bottleneck_v2/preact/FusedBatchNorm', num_features=2048, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_2/bottleneck_v2/conv1/Conv2D', in_channels=2048, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_2/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_2/bottleneck_v2/conv2/Conv2D', in_channels=512, out_channels=512, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_2/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_2/bottleneck_v2/conv3/Conv2D', in_channels=512, out_channels=2048, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_3/bottleneck_v2/preact/FusedBatchNorm', num_features=2048, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_3/bottleneck_v2/conv1/Conv2D', in_channels=2048, out_channels=512, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_3/bottleneck_v2/conv1/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_3/bottleneck_v2/conv2/Conv2D', in_channels=512, out_channels=512, kernel_size=(3, 3), stride=(1, 1), groups=1, bias=None)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/block4/unit_3/bottleneck_v2/conv2/BatchNorm/FusedBatchNorm', num_features=512, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv3_Conv2D = self.__conv(2, name='resnet_v2_101/block4/unit_3/bottleneck_v2/conv3/Conv2D', in_channels=512, out_channels=2048, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
self.resnet_v2_101_postnorm_FusedBatchNorm = self.__batch_normalization(2, 'resnet_v2_101/postnorm/FusedBatchNorm', num_features=2048, eps=1.0009999641624745e-05, momentum=0.0)
|
||
|
|
self.resnet_v2_101_logits_Conv2D = self.__conv(2, name='resnet_v2_101/logits/Conv2D', in_channels=2048, out_channels=1001, kernel_size=(1, 1), stride=(1, 1), groups=1, bias=True)
|
||
|
|
|
||
|
|
def forward(self, x):
|
||
|
|
resnet_v2_101_Pad = F.pad(x, (3, 3, 3, 3), mode = 'constant', value = 0)
|
||
|
|
resnet_v2_101_conv1_Conv2D = self.resnet_v2_101_conv1_Conv2D(resnet_v2_101_Pad)
|
||
|
|
resnet_v2_101_pool1_MaxPool_pad = F.pad(resnet_v2_101_conv1_Conv2D, (0, 1, 0, 1), value=float('-inf'))
|
||
|
|
resnet_v2_101_pool1_MaxPool, resnet_v2_101_pool1_MaxPool_idx = F.max_pool2d(resnet_v2_101_pool1_MaxPool_pad, kernel_size=(3, 3), stride=(2, 2), padding=0, ceil_mode=False, return_indices=True)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block1_unit_1_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_pool1_MaxPool)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block1_unit_1_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_shortcut_Conv2D = self.resnet_v2_101_block1_unit_1_bottleneck_v2_shortcut_Conv2D(resnet_v2_101_block1_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block1_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block1_unit_1_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block1_unit_1_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block1_unit_1_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block1_unit_1_bottleneck_v2_add = resnet_v2_101_block1_unit_1_bottleneck_v2_shortcut_Conv2D + resnet_v2_101_block1_unit_1_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block1_unit_2_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block1_unit_1_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block1_unit_2_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block1_unit_2_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block1_unit_2_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block1_unit_2_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block1_unit_2_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block1_unit_2_bottleneck_v2_add = resnet_v2_101_block1_unit_1_bottleneck_v2_add + resnet_v2_101_block1_unit_2_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block1_unit_3_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block1_unit_2_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_shortcut_MaxPool, resnet_v2_101_block1_unit_3_bottleneck_v2_shortcut_MaxPool_idx = F.max_pool2d(resnet_v2_101_block1_unit_2_bottleneck_v2_add, kernel_size=(1, 1), stride=(2, 2), padding=0, ceil_mode=False, return_indices=True)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block1_unit_3_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block1_unit_3_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_Pad = F.pad(resnet_v2_101_block1_unit_3_bottleneck_v2_conv1_Relu, (1, 1, 1, 1), mode = 'constant', value = 0)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block1_unit_3_bottleneck_v2_Pad)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block1_unit_3_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block1_unit_3_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block1_unit_3_bottleneck_v2_add = resnet_v2_101_block1_unit_3_bottleneck_v2_shortcut_MaxPool + resnet_v2_101_block1_unit_3_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block2_unit_1_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block1_unit_3_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block2_unit_1_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_shortcut_Conv2D = self.resnet_v2_101_block2_unit_1_bottleneck_v2_shortcut_Conv2D(resnet_v2_101_block2_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block2_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block2_unit_1_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block2_unit_1_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block2_unit_1_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block2_unit_1_bottleneck_v2_add = resnet_v2_101_block2_unit_1_bottleneck_v2_shortcut_Conv2D + resnet_v2_101_block2_unit_1_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block2_unit_2_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block2_unit_1_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block2_unit_2_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block2_unit_2_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block2_unit_2_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block2_unit_2_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block2_unit_2_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block2_unit_2_bottleneck_v2_add = resnet_v2_101_block2_unit_1_bottleneck_v2_add + resnet_v2_101_block2_unit_2_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block2_unit_3_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block2_unit_2_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block2_unit_3_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block2_unit_3_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block2_unit_3_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block2_unit_3_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block2_unit_3_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block2_unit_3_bottleneck_v2_add = resnet_v2_101_block2_unit_2_bottleneck_v2_add + resnet_v2_101_block2_unit_3_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block2_unit_4_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block2_unit_3_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_shortcut_MaxPool, resnet_v2_101_block2_unit_4_bottleneck_v2_shortcut_MaxPool_idx = F.max_pool2d(resnet_v2_101_block2_unit_3_bottleneck_v2_add, kernel_size=(1, 1), stride=(2, 2), padding=0, ceil_mode=False, return_indices=True)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block2_unit_4_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block2_unit_4_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_Pad = F.pad(resnet_v2_101_block2_unit_4_bottleneck_v2_conv1_Relu, (1, 1, 1, 1), mode = 'constant', value = 0)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block2_unit_4_bottleneck_v2_Pad)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block2_unit_4_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block2_unit_4_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block2_unit_4_bottleneck_v2_add = resnet_v2_101_block2_unit_4_bottleneck_v2_shortcut_MaxPool + resnet_v2_101_block2_unit_4_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_1_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block2_unit_4_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_1_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_shortcut_Conv2D = self.resnet_v2_101_block3_unit_1_bottleneck_v2_shortcut_Conv2D(resnet_v2_101_block3_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_1_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_1_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_1_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_1_bottleneck_v2_add = resnet_v2_101_block3_unit_1_bottleneck_v2_shortcut_Conv2D + resnet_v2_101_block3_unit_1_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_2_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_1_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_2_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_2_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_2_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_2_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_2_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_2_bottleneck_v2_add = resnet_v2_101_block3_unit_1_bottleneck_v2_add + resnet_v2_101_block3_unit_2_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_3_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_2_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_3_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_3_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_3_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_3_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_3_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_3_bottleneck_v2_add = resnet_v2_101_block3_unit_2_bottleneck_v2_add + resnet_v2_101_block3_unit_3_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_4_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_3_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_4_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_4_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_4_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_4_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_4_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_4_bottleneck_v2_add = resnet_v2_101_block3_unit_3_bottleneck_v2_add + resnet_v2_101_block3_unit_4_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_5_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_4_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_5_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_5_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_5_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_5_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_5_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_5_bottleneck_v2_add = resnet_v2_101_block3_unit_4_bottleneck_v2_add + resnet_v2_101_block3_unit_5_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_6_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_5_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_6_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_6_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_6_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_6_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_6_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_6_bottleneck_v2_add = resnet_v2_101_block3_unit_5_bottleneck_v2_add + resnet_v2_101_block3_unit_6_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_7_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_6_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_7_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_7_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_7_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_7_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_7_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_7_bottleneck_v2_add = resnet_v2_101_block3_unit_6_bottleneck_v2_add + resnet_v2_101_block3_unit_7_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_8_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_7_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_8_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_8_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_8_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_8_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_8_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_8_bottleneck_v2_add = resnet_v2_101_block3_unit_7_bottleneck_v2_add + resnet_v2_101_block3_unit_8_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_9_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_8_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_9_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_9_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_9_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_9_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_9_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_9_bottleneck_v2_add = resnet_v2_101_block3_unit_8_bottleneck_v2_add + resnet_v2_101_block3_unit_9_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_10_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_9_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_10_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_10_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_10_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_10_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_10_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_10_bottleneck_v2_add = resnet_v2_101_block3_unit_9_bottleneck_v2_add + resnet_v2_101_block3_unit_10_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_11_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_10_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_11_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_11_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_11_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_11_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_11_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_11_bottleneck_v2_add = resnet_v2_101_block3_unit_10_bottleneck_v2_add + resnet_v2_101_block3_unit_11_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_12_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_11_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_12_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_12_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_12_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_12_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_12_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_12_bottleneck_v2_add = resnet_v2_101_block3_unit_11_bottleneck_v2_add + resnet_v2_101_block3_unit_12_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_13_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_12_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_13_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_13_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_13_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_13_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_13_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_13_bottleneck_v2_add = resnet_v2_101_block3_unit_12_bottleneck_v2_add + resnet_v2_101_block3_unit_13_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_14_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_13_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_14_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_14_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_14_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_14_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_14_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_14_bottleneck_v2_add = resnet_v2_101_block3_unit_13_bottleneck_v2_add + resnet_v2_101_block3_unit_14_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_15_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_14_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_15_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_15_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_15_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_15_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_15_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_15_bottleneck_v2_add = resnet_v2_101_block3_unit_14_bottleneck_v2_add + resnet_v2_101_block3_unit_15_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_16_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_15_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_16_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_16_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_16_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_16_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_16_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_16_bottleneck_v2_add = resnet_v2_101_block3_unit_15_bottleneck_v2_add + resnet_v2_101_block3_unit_16_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_17_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_16_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_17_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_17_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_17_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_17_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_17_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_17_bottleneck_v2_add = resnet_v2_101_block3_unit_16_bottleneck_v2_add + resnet_v2_101_block3_unit_17_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_18_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_17_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_18_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_18_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_18_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_18_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_18_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_18_bottleneck_v2_add = resnet_v2_101_block3_unit_17_bottleneck_v2_add + resnet_v2_101_block3_unit_18_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_19_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_18_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_19_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_19_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_19_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_19_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_19_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_19_bottleneck_v2_add = resnet_v2_101_block3_unit_18_bottleneck_v2_add + resnet_v2_101_block3_unit_19_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_20_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_19_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_20_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_20_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_20_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_20_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_20_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_20_bottleneck_v2_add = resnet_v2_101_block3_unit_19_bottleneck_v2_add + resnet_v2_101_block3_unit_20_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_21_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_20_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_21_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_21_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_21_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_21_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_21_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_21_bottleneck_v2_add = resnet_v2_101_block3_unit_20_bottleneck_v2_add + resnet_v2_101_block3_unit_21_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_22_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_21_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_22_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_22_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block3_unit_22_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_22_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_22_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_22_bottleneck_v2_add = resnet_v2_101_block3_unit_21_bottleneck_v2_add + resnet_v2_101_block3_unit_22_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block3_unit_23_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_22_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_shortcut_MaxPool, resnet_v2_101_block3_unit_23_bottleneck_v2_shortcut_MaxPool_idx = F.max_pool2d(resnet_v2_101_block3_unit_22_bottleneck_v2_add, kernel_size=(1, 1), stride=(2, 2), padding=0, ceil_mode=False, return_indices=True)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block3_unit_23_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block3_unit_23_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_Pad = F.pad(resnet_v2_101_block3_unit_23_bottleneck_v2_conv1_Relu, (1, 1, 1, 1), mode = 'constant', value = 0)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block3_unit_23_bottleneck_v2_Pad)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block3_unit_23_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block3_unit_23_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block3_unit_23_bottleneck_v2_add = resnet_v2_101_block3_unit_23_bottleneck_v2_shortcut_MaxPool + resnet_v2_101_block3_unit_23_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block4_unit_1_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block3_unit_23_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block4_unit_1_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_shortcut_Conv2D = self.resnet_v2_101_block4_unit_1_bottleneck_v2_shortcut_Conv2D(resnet_v2_101_block4_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block4_unit_1_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block4_unit_1_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block4_unit_1_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block4_unit_1_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block4_unit_1_bottleneck_v2_add = resnet_v2_101_block4_unit_1_bottleneck_v2_shortcut_Conv2D + resnet_v2_101_block4_unit_1_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block4_unit_2_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block4_unit_1_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block4_unit_2_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block4_unit_2_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block4_unit_2_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block4_unit_2_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block4_unit_2_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block4_unit_2_bottleneck_v2_add = resnet_v2_101_block4_unit_1_bottleneck_v2_add + resnet_v2_101_block4_unit_2_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_preact_FusedBatchNorm = self.resnet_v2_101_block4_unit_3_bottleneck_v2_preact_FusedBatchNorm(resnet_v2_101_block4_unit_2_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_preact_Relu = F.relu(resnet_v2_101_block4_unit_3_bottleneck_v2_preact_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Conv2D = self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Conv2D(resnet_v2_101_block4_unit_3_bottleneck_v2_preact_Relu)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Relu = F.relu(resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D_pad = F.pad(resnet_v2_101_block4_unit_3_bottleneck_v2_conv1_Relu, (1, 1, 1, 1))
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D = self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D(resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D_pad)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm = self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm(resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Conv2D)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Relu = F.relu(resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_BatchNorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_conv3_Conv2D = self.resnet_v2_101_block4_unit_3_bottleneck_v2_conv3_Conv2D(resnet_v2_101_block4_unit_3_bottleneck_v2_conv2_Relu)
|
||
|
|
resnet_v2_101_block4_unit_3_bottleneck_v2_add = resnet_v2_101_block4_unit_2_bottleneck_v2_add + resnet_v2_101_block4_unit_3_bottleneck_v2_conv3_Conv2D
|
||
|
|
resnet_v2_101_postnorm_FusedBatchNorm = self.resnet_v2_101_postnorm_FusedBatchNorm(resnet_v2_101_block4_unit_3_bottleneck_v2_add)
|
||
|
|
resnet_v2_101_postnorm_Relu = F.relu(resnet_v2_101_postnorm_FusedBatchNorm)
|
||
|
|
resnet_v2_101_pool5 = torch.mean(resnet_v2_101_postnorm_Relu, 3, True)
|
||
|
|
resnet_v2_101_pool5 = torch.mean(resnet_v2_101_pool5, 2, True)
|
||
|
|
resnet_v2_101_logits_Conv2D = self.resnet_v2_101_logits_Conv2D(resnet_v2_101_pool5)
|
||
|
|
resnet_v2_101_SpatialSqueeze = torch.squeeze(resnet_v2_101_logits_Conv2D)
|
||
|
|
MMdnn_Output_input = [resnet_v2_101_SpatialSqueeze]
|
||
|
|
return MMdnn_Output_input
|
||
|
|
|
||
|
|
|
||
|
|
@staticmethod
|
||
|
|
def __conv(dim, name, **kwargs):
|
||
|
|
if dim == 1: layer = nn.Conv1d(**kwargs)
|
||
|
|
elif dim == 2: layer = nn.Conv2d(**kwargs)
|
||
|
|
elif dim == 3: layer = nn.Conv3d(**kwargs)
|
||
|
|
else: raise NotImplementedError()
|
||
|
|
|
||
|
|
layer.state_dict()['weight'].copy_(torch.from_numpy(_weights_dict[name]['weights']))
|
||
|
|
if 'bias' in _weights_dict[name]:
|
||
|
|
layer.state_dict()['bias'].copy_(torch.from_numpy(_weights_dict[name]['bias']))
|
||
|
|
return layer
|
||
|
|
|
||
|
|
@staticmethod
|
||
|
|
def __batch_normalization(dim, name, **kwargs):
|
||
|
|
if dim == 0 or dim == 1: layer = nn.BatchNorm1d(**kwargs)
|
||
|
|
elif dim == 2: layer = nn.BatchNorm2d(**kwargs)
|
||
|
|
elif dim == 3: layer = nn.BatchNorm3d(**kwargs)
|
||
|
|
else: raise NotImplementedError()
|
||
|
|
|
||
|
|
if 'scale' in _weights_dict[name]:
|
||
|
|
layer.state_dict()['weight'].copy_(torch.from_numpy(_weights_dict[name]['scale']))
|
||
|
|
else:
|
||
|
|
layer.weight.data.fill_(1)
|
||
|
|
|
||
|
|
if 'bias' in _weights_dict[name]:
|
||
|
|
layer.state_dict()['bias'].copy_(torch.from_numpy(_weights_dict[name]['bias']))
|
||
|
|
else:
|
||
|
|
layer.bias.data.fill_(0)
|
||
|
|
|
||
|
|
layer.state_dict()['running_mean'].copy_(torch.from_numpy(_weights_dict[name]['mean']))
|
||
|
|
layer.state_dict()['running_var'].copy_(torch.from_numpy(_weights_dict[name]['var']))
|
||
|
|
return layer
|
||
|
|
|