Understranding HRNET with code

• base part
• OCR module
• Basic and Bottleneck Module
• HighResolution Module
• relu
• fuse_layer
• Test model

Hrnet + ocr module is as follows, all the codes borrow from :

1. https://github.com/HRNet/HRNet-Semantic-Segmentation/blob/HRNet-OCR/lib/models/seg_hrnet_ocr.py
2. https://github.com/openseg-group/openseg.pytorch

base part

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import os
import logging
import functools

import numpy as np

import torch
import torch.nn as nn
import torch._utils
import torch.nn.functional as F
import sys
sys.path.append(
    '/home/ubuntu/ds/segmentation/HRNet-Semantic-Segmentation/lib/models/')
# from .bn_helper import BatchNorm2d, BatchNorm2d_class, relu_inplace

from bn_helper import BatchNorm2d, BatchNorm2d_class, relu_inplace

ALIGN_CORNERS = True
BN_MOMENTUM = 0.1
logger = logging.getLogger(__name__)
sys.path.append('/home/ubuntu/ds/segmentation/HRNet-Semantic-Segmentation/lib/')
from config import config
from config import update_config
from config.default import _C as cfg
# config file
path = '/home/ubuntu/ds/segmentation/HRNet-Semantic-Segmentation/experiments/cityscapes/seg_hrnet_ocr_w48_trainval_512x1024_sgd_lr1e-2_wd5e-4_bs_12_epoch484.yaml'
config.merge_from_file(path)

class ModuleHelper:

    @staticmethod
    def BNReLU(num_features, bn_type=None, **kwargs):
        return nn.Sequential(
            BatchNorm2d(num_features, **kwargs),
            nn.ReLU()
        )

    @staticmethod
    def BatchNorm2d(*args, **kwargs):
        return BatchNorm2d


def conv3x3(in_planes, out_planes, stride=1):
    """3x3 convolution with padding"""
    return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
                     padding=1, bias=False)

config

CfgNode({'OUTPUT_DIR': 'output', 'LOG_DIR': 'log', 'GPUS': (0,), 'WORKERS': 1, 'PRINT_FREQ': 10, 'AUTO_RESUME': False, 'PIN_MEMORY': True, 'RANK': 0, 'CUDNN': CfgNode({'BENCHMARK': True, 'DETERMINISTIC': False, 'ENABLED': True}), 'MODEL': CfgNode({'NAME': 'seg_hrnet_ocr', 'PRETRAINED': 'pretrained_models/hrnetv2_w48_imagenet_pretrained.pth', 'ALIGN_CORNERS': True, 'NUM_OUTPUTS': 2, 'EXTRA': CfgNode({'FINAL_CONV_KERNEL': 1, 'STAGE1': CfgNode({'NUM_MODULES': 1, 'NUM_RANCHES': 1, 'BLOCK': 'BOTTLENECK', 'NUM_BLOCKS': [4], 'NUM_CHANNELS': [64], 'FUSE_METHOD': 'SUM'}), 'STAGE2': CfgNode({'NUM_MODULES': 1, 'NUM_BRANCHES': 2, 'BLOCK': 'BASIC', 'NUM_BLOCKS': [4, 4], 'NUM_CHANNELS': [48, 96], 'FUSE_METHOD': 'SUM'}), 'STAGE3': CfgNode({'NUM_MODULES': 4, 'NUM_BRANCHES': 3, 'BLOCK': 'BASIC', 'NUM_BLOCKS': [4, 4, 4], 'NUM_CHANNELS': [48, 96, 192], 'FUSE_METHOD': 'SUM'}), 'STAGE4': CfgNode({'NUM_MODULES': 3, 'NUM_BRANCHES': 4, 'BLOCK': 'BASIC', 'NUM_BLOCKS': [4, 4, 4, 4], 'NUM_CHANNELS': [48, 96, 192, 384], 'FUSE_METHOD': 'SUM'})}), 'OCR': CfgNode({'MID_CHANNELS': 512, 'KEY_CHANNELS': 256, 'DROPOUT': 0.05, 'SCALE': 1})}), 'LOSS': CfgNode({'USE_OHEM': False, 'OHEMTHRES': 0.9, 'OHEMKEEP': 131072, 'CLASS_BALANCE': False, 'BALANCE_WEIGHTS': [0.4, 1]}), 'DATASET': CfgNode({'ROOT': '', 'DATASET': 'cityscapesEXTRA_TRAIN_SET', 'NUM_CLASSES': 19, 'TRAIN_SET': 'list/cityscapes/trainval.lst', 'EXTRA_TRAIN_SET': '', 'TEST_SET': 'list/cityscapes/val.lst'}), 'TRAIN': CfgNode({'FREEZE_LAYERS': '', 'FREEZE_EPOCHS': -1, 'NONBACKBONE_KEYWORDS': [], 'NONBACKBONE_MULT': 10, 'IMAGE_SIZE': [1024, 512], 'BASE_SIZE': 2048, 'DOWNSAMPLERATE': 1, 'FLIP': True, 'MULTI_SCALE': True, 'SCALE_FACTOR': 16, 'RANDOM_BRIGHTNESS': False, 'RANDOM_BRIGHTNESS_SHIFT_VALUE': 10, 'LR_FACTOR': 0.1, 'LR_STEP': [90, 110], 'LR': 0.01, 'EXTRA_LR': 0.001, 'OPTIMIZER': 'sgd', 'MOMENTUM': 0.9, 'WD': 0.0005, 'NESTEROV': False, 'IGNORE_LABEL': 255, 'BEGIN_EPOCH': 0, 'END_EPOCH': 484, 'EXTRA_EPOCH': 0, 'RESUME': True, 'BATCH_SIZE_PER_GPU': 3, 'SHUFFLE': True, 'NUM_SAMPLES': 0}), 'TEST': CfgNode({'IMAGE_SIZE': [2048, 1024], 'BASE_SIZE': 2048, 'BATCH_SIZE_PER_GPU': 4, 'NUM_SAMPLES': 0, 'MODEL_FILE': '', 'FLIP_TEST': False, 'MULTI_SCALE': False, 'SCALE_LIST': [1], 'OUTPUT_INDEX': -1}), 'DEBUG': CfgNode({'DEBUG': False, 'SAVE_BATCH_IMAGES_GT': False, 'SAVE_BATCH_IMAGES_PRED': False, 'SAVE_HEATMAPS_GT': False, 'SAVE_HEATMAPS_PRED': False})})

for i in range(1,4):
    print(i)

1
2
3

OCR module

class SpatialGather_Module(nn.Module):
    """
        Aggregate the context features according to the initial 
        predicted probability distribution.
        Employ the soft-weighted method to aggregate the context.
    """

    def __init__(self, cls_num=0, scale=1):
        super(SpatialGather_Module, self).__init__()
        self.cls_num = cls_num
        self.scale = scale

    def forward(self, feats, probs):
        batch_size, c, h, w = probs.size(0), probs.size(
            1), probs.size(2), probs.size(3)
        probs = probs.view(batch_size, c, -1)
        feats = feats.view(batch_size, feats.size(1), -1)
        feats = feats.permute(0, 2, 1)  # batch x hw x c
        probs = F.softmax(self.scale * probs, dim=2)  # batch x k x hw
        ocr_context = torch.matmul(probs, feats)\
            .permute(0, 2, 1).unsqueeze(3)  # batch x k x c
        return ocr_context


class _ObjectAttentionBlock(nn.Module):
    '''
    The basic implementation for object context block
    Input:
        N X C X H X W
    Parameters:
        in_channels       : the dimension of the input feature map
        key_channels      : the dimension after the key/query transform
        scale             : choose the scale to downsample the input feature maps (save memory cost)
        bn_type           : specify the bn type
    Return:
        N X C X H X W
    '''

    def __init__(self,
                 in_channels,
                 key_channels,
                 scale=1,
                 bn_type=None):
        super(_ObjectAttentionBlock, self).__init__()
        self.scale = scale
        self.in_channels = in_channels
        self.key_channels = key_channels
        self.pool = nn.MaxPool2d(kernel_size=(scale, scale))
        self.f_pixel = nn.Sequential(
            nn.Conv2d(in_channels=self.in_channels, out_channels=self.key_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
            nn.Conv2d(in_channels=self.key_channels, out_channels=self.key_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
        )
        self.f_object = nn.Sequential(
            nn.Conv2d(in_channels=self.in_channels, out_channels=self.key_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
            nn.Conv2d(in_channels=self.key_channels, out_channels=self.key_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
        )
        self.f_down = nn.Sequential(
            nn.Conv2d(in_channels=self.in_channels, out_channels=self.key_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
        )
        self.f_up = nn.Sequential(
            nn.Conv2d(in_channels=self.key_channels, out_channels=self.in_channels,
                      kernel_size=1, stride=1, padding=0, bias=False),
            ModuleHelper.BNReLU(self.in_channels, bn_type=bn_type),
        )

    def forward(self, x, proxy):
        batch_size, h, w = x.size(0), x.size(2), x.size(3)
        if self.scale > 1:
            x = self.pool(x)

        query = self.f_pixel(x).view(batch_size, self.key_channels, -1)
        query = query.permute(0, 2, 1)
        key = self.f_object(proxy).view(batch_size, self.key_channels, -1)
        value = self.f_down(proxy).view(batch_size, self.key_channels, -1)
        value = value.permute(0, 2, 1)

        sim_map = torch.matmul(query, key)
        sim_map = (self.key_channels**-.5) * sim_map
        sim_map = F.softmax(sim_map, dim=-1)

        # add bg context ...
        context = torch.matmul(sim_map, value)
        context = context.permute(0, 2, 1).contiguous()
        context = context.view(batch_size, self.key_channels, *x.size()[2:])
        context = self.f_up(context)
        if self.scale > 1:
            context = F.interpolate(input=context, size=(
                h, w), mode='bilinear', align_corners=ALIGN_CORNERS)

        return context


class ObjectAttentionBlock2D(_ObjectAttentionBlock):
    def __init__(self,
                 in_channels,
                 key_channels,
                 scale=1,
                 bn_type=None):
        super(ObjectAttentionBlock2D, self).__init__(in_channels,
                                                     key_channels,
                                                     scale,
                                                     bn_type=bn_type)


class SpatialOCR_Module(nn.Module):
    """
    Implementation of the OCR module:
    We aggregate the global object representation to update the representation for each pixel.
    """

    def __init__(self,
                 in_channels,
                 key_channels,
                 out_channels,
                 scale=1,
                 dropout=0.1,
                 bn_type=None):
        super(SpatialOCR_Module, self).__init__()
        self.object_context_block = ObjectAttentionBlock2D(in_channels,
                                                           key_channels,
                                                           scale,
                                                           bn_type)
        _in_channels = 2 * in_channels

        self.conv_bn_dropout = nn.Sequential(
            nn.Conv2d(_in_channels, out_channels,
                      kernel_size=1, padding=0, bias=False),
            ModuleHelper.BNReLU(out_channels, bn_type=bn_type),
            nn.Dropout2d(dropout)
        )

    def forward(self, feats, proxy_feats):
        context = self.object_context_block(feats, proxy_feats)

        output = self.conv_bn_dropout(torch.cat([context, feats], 1))

        return output

Basic and Bottleneck Module

class BasicBlock(nn.Module):
    expansion = 1

    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = BatchNorm2d(planes, momentum=BN_MOMENTUM)
        self.relu = nn.ReLU(inplace=relu_inplace)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = BatchNorm2d(planes, momentum=BN_MOMENTUM)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out = out + residual
        out = self.relu(out)

        return out


class Bottleneck(nn.Module):
    expansion = 4

    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(Bottleneck, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = BatchNorm2d(planes, momentum=BN_MOMENTUM)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
                               padding=1, bias=False)
        self.bn2 = BatchNorm2d(planes, momentum=BN_MOMENTUM)
        self.conv3 = nn.Conv2d(planes, planes * self.expansion, kernel_size=1,
                               bias=False)
        self.bn3 = BatchNorm2d(planes * self.expansion,
                               momentum=BN_MOMENTUM)
        self.relu = nn.ReLU(inplace=relu_inplace)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)

        out = self.conv3(out)
        out = self.bn3(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out = out + residual
        out = self.relu(out)

        return out

HighResolution Module

class HighResolutionModule(nn.Module):
    def __init__(self, num_branches, blocks, num_blocks, num_inchannels,
                 num_channels, fuse_method, multi_scale_output=True):
        super(HighResolutionModule, self).__init__()
        self._check_branches(
            num_branches, blocks, num_blocks, num_inchannels, num_channels)

        self.num_inchannels = num_inchannels
        self.fuse_method = fuse_method
        self.num_branches = num_branches

        self.multi_scale_output = multi_scale_output

        self.branches = self._make_branches(
            num_branches, blocks, num_blocks, num_channels)
        self.fuse_layers = self._make_fuse_layers()
        self.relu = nn.ReLU(inplace=relu_inplace)

    def _check_branches(self, num_branches, blocks, num_blocks,
                        num_inchannels, num_channels):
        if num_branches != len(num_blocks):
            error_msg = 'NUM_BRANCHES({}) <> NUM_BLOCKS({})'.format(
                num_branches, len(num_blocks))
            logger.error(error_msg)
            raise ValueError(error_msg)

        if num_branches != len(num_channels):
            error_msg = 'NUM_BRANCHES({}) <> NUM_CHANNELS({})'.format(
                num_branches, len(num_channels))
            logger.error(error_msg)
            raise ValueError(error_msg)

        if num_branches != len(num_inchannels):
            error_msg = 'NUM_BRANCHES({}) <> NUM_INCHANNELS({})'.format(
                num_branches, len(num_inchannels))
            logger.error(error_msg)
            raise ValueError(error_msg)

    def _make_one_branch(self, branch_index, block, num_blocks, num_channels,
                         stride=1):
        downsample = None
        if stride != 1 or \
           self.num_inchannels[branch_index] != num_channels[branch_index] * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(self.num_inchannels[branch_index],
                          num_channels[branch_index] * block.expansion,
                          kernel_size=1, stride=stride, bias=False),
                BatchNorm2d(num_channels[branch_index] * block.expansion,
                            momentum=BN_MOMENTUM),
            )

        layers = []
        layers.append(block(self.num_inchannels[branch_index],
                            num_channels[branch_index], stride, downsample))
        self.num_inchannels[branch_index] = \
            num_channels[branch_index] * block.expansion
        for i in range(1, num_blocks[branch_index]):
            layers.append(block(self.num_inchannels[branch_index],
                                num_channels[branch_index]))

        return nn.Sequential(*layers)

    def _make_branches(self, num_branches, block, num_blocks, num_channels):
        branches = []

        for i in range(num_branches):
            branches.append(
                self._make_one_branch(i, block, num_blocks, num_channels))

        return nn.ModuleList(branches)

    def _make_fuse_layers(self):
        if self.num_branches == 1:
            return None

        num_branches = self.num_branches
        num_inchannels = self.num_inchannels
        fuse_layers = []
        for i in range(num_branches if self.multi_scale_output else 1):
            fuse_layer = []
            for j in range(num_branches):
                if j > i:
                    fuse_layer.append(nn.Sequential(
                        nn.Conv2d(num_inchannels[j],
                                  num_inchannels[i],
                                  1,
                                  1,
                                  0,
                                  bias=False),
                        BatchNorm2d(num_inchannels[i], momentum=BN_MOMENTUM)))
                elif j == i:
                    fuse_layer.append(None)
                else:
                    conv3x3s = []
                    for k in range(i-j):
                        if k == i - j - 1:
                            num_outchannels_conv3x3 = num_inchannels[i]
                            conv3x3s.append(nn.Sequential(
                                nn.Conv2d(num_inchannels[j],
                                          num_outchannels_conv3x3,
                                          3, 2, 1, bias=False),
                                BatchNorm2d(num_outchannels_conv3x3,
                                            momentum=BN_MOMENTUM)))
                        else:
                            num_outchannels_conv3x3 = num_inchannels[j]
                            conv3x3s.append(nn.Sequential(
                                nn.Conv2d(num_inchannels[j],
                                          num_outchannels_conv3x3,
                                          3, 2, 1, bias=False),
                                BatchNorm2d(num_outchannels_conv3x3,
                                            momentum=BN_MOMENTUM),
                                nn.ReLU(inplace=relu_inplace)))
                    fuse_layer.append(nn.Sequential(*conv3x3s))
            fuse_layers.append(nn.ModuleList(fuse_layer))

        return nn.ModuleList(fuse_layers)

    def get_num_inchannels(self):
        return self.num_inchannels

    def forward(self, x):
        if self.num_branches == 1:
            return [self.branches[0](x[0])]

        for i in range(self.num_branches):
            x[i] = self.branches[i](x[i])

        x_fuse = []
        for i in range(len(self.fuse_layers)):
            y = x[0] if i == 0 else self.fuse_layers[i][0](x[0])
            for j in range(1, self.num_branches):
                if i == j:
                    y = y + x[j]
                elif j > i:
                    width_output = x[i].shape[-1]
                    height_output = x[i].shape[-2]
                    y = y + F.interpolate(
                        self.fuse_layers[i][j](x[j]),
                        size=[height_output, width_output],
                        mode='bilinear', align_corners=ALIGN_CORNERS)
                else:
                    y = y + self.fuse_layers[i][j](x[j])
            x_fuse.append(self.relu(y))

        return x_fuse


blocks_dict = {
    'BASIC': BasicBlock,
    'BOTTLENECK': Bottleneck
}

class HighResolutionNet(nn.Module):

    def __init__(self, config, **kwargs):
        global ALIGN_CORNERS
        extra = config.MODEL.EXTRA
        super(HighResolutionNet, self).__init__()
        ALIGN_CORNERS = config.MODEL.ALIGN_CORNERS

        # stem net
        self.conv1 = nn.Conv2d(3, 64, kernel_size=3, stride=2, padding=1,
                               bias=False)
        self.bn1 = BatchNorm2d(64, momentum=BN_MOMENTUM)
        self.conv2 = nn.Conv2d(64, 64, kernel_size=3, stride=2, padding=1,
                               bias=False)
        self.bn2 = BatchNorm2d(64, momentum=BN_MOMENTUM)
        self.relu = nn.ReLU(inplace=relu_inplace)

        self.stage1_cfg = extra['STAGE1']
        num_channels = self.stage1_cfg['NUM_CHANNELS'][0]
        block = blocks_dict[self.stage1_cfg['BLOCK']]
        num_blocks = self.stage1_cfg['NUM_BLOCKS'][0]
        self.layer1 = self._make_layer(block, 64, num_channels, num_blocks)
        stage1_out_channel = block.expansion*num_channels

        self.stage2_cfg = extra['STAGE2']
        num_channels = self.stage2_cfg['NUM_CHANNELS']
        block = blocks_dict[self.stage2_cfg['BLOCK']]
        num_channels = [
            num_channels[i] * block.expansion for i in range(len(num_channels))]
        self.transition1 = self._make_transition_layer(
            [stage1_out_channel], num_channels)
        self.stage2, pre_stage_channels = self._make_stage(
            self.stage2_cfg, num_channels)

        self.stage3_cfg = extra['STAGE3']
        num_channels = self.stage3_cfg['NUM_CHANNELS']
        block = blocks_dict[self.stage3_cfg['BLOCK']]
        num_channels = [
            num_channels[i] * block.expansion for i in range(len(num_channels))]
        self.transition2 = self._make_transition_layer(
            pre_stage_channels, num_channels)
        self.stage3, pre_stage_channels = self._make_stage(
            self.stage3_cfg, num_channels)

        self.stage4_cfg = extra['STAGE4']
        num_channels = self.stage4_cfg['NUM_CHANNELS']
        block = blocks_dict[self.stage4_cfg['BLOCK']]
        num_channels = [
            num_channels[i] * block.expansion for i in range(len(num_channels))]
        self.transition3 = self._make_transition_layer(
            pre_stage_channels, num_channels)
        self.stage4, pre_stage_channels = self._make_stage(
            self.stage4_cfg, num_channels, multi_scale_output=True)

        last_inp_channels = np.int(np.sum(pre_stage_channels))
        ocr_mid_channels = config.MODEL.OCR.MID_CHANNELS
        ocr_key_channels = config.MODEL.OCR.KEY_CHANNELS

        self.conv3x3_ocr = nn.Sequential(
            nn.Conv2d(last_inp_channels, ocr_mid_channels,
                      kernel_size=3, stride=1, padding=1),
            BatchNorm2d(ocr_mid_channels),
            nn.ReLU(inplace=relu_inplace),
        )
        self.ocr_gather_head = SpatialGather_Module(config.DATASET.NUM_CLASSES)

        self.ocr_distri_head = SpatialOCR_Module(in_channels=ocr_mid_channels,
                                                 key_channels=ocr_key_channels,
                                                 out_channels=ocr_mid_channels,
                                                 scale=1,
                                                 dropout=0.05,
                                                 )
        self.cls_head = nn.Conv2d(
            ocr_mid_channels, config.DATASET.NUM_CLASSES, kernel_size=1, stride=1, padding=0, bias=True)

        self.aux_head = nn.Sequential(
            nn.Conv2d(last_inp_channels, last_inp_channels,
                      kernel_size=1, stride=1, padding=0),
            BatchNorm2d(last_inp_channels),
            nn.ReLU(inplace=relu_inplace),
            nn.Conv2d(last_inp_channels, config.DATASET.NUM_CLASSES,
                      kernel_size=1, stride=1, padding=0, bias=True)
        )

    def _make_transition_layer(
            self, num_channels_pre_layer, num_channels_cur_layer):
        num_branches_cur = len(num_channels_cur_layer)
        num_branches_pre = len(num_channels_pre_layer)
        transition_layers = []
        for i in range(num_branches_cur):
            if i < num_branches_pre:
                if num_channels_cur_layer[i] != num_channels_pre_layer[i]:
                    transition_layers.append(nn.Sequential(
                        nn.Conv2d(num_channels_pre_layer[i],
                                  num_channels_cur_layer[i],
                                  3,
                                  1,
                                  1,
                                  bias=False),
                        BatchNorm2d(
                            num_channels_cur_layer[i], momentum=BN_MOMENTUM),
                        nn.ReLU(inplace=relu_inplace)))
                else:
                    transition_layers.append(None)
            else:
                conv3x3s = []
                for j in range(i+1-num_branches_pre):
                    inchannels = num_channels_pre_layer[-1]
                    outchannels = num_channels_cur_layer[i] \
                        if j == i-num_branches_pre else inchannels
                    conv3x3s.append(nn.Sequential(
                        nn.Conv2d(
                            inchannels, outchannels, 3, 2, 1, bias=False),
                        BatchNorm2d(outchannels, momentum=BN_MOMENTUM),
                        nn.ReLU(inplace=relu_inplace)))
                transition_layers.append(nn.Sequential(*conv3x3s))

        return nn.ModuleList(transition_layers)

    def _make_layer(self, block, inplanes, planes, blocks, stride=1):
        downsample = None
        if stride != 1 or inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(inplanes, planes * block.expansion,
                          kernel_size=1, stride=stride, bias=False),
                BatchNorm2d(planes * block.expansion, momentum=BN_MOMENTUM),
            )

        layers = []
        layers.append(block(inplanes, planes, stride, downsample))
        inplanes = planes * block.expansion
        for i in range(1, blocks):
            layers.append(block(inplanes, planes))

        return nn.Sequential(*layers)

    def _make_stage(self, layer_config, num_inchannels,
                    multi_scale_output=True):
        num_modules = layer_config['NUM_MODULES']
        num_branches = layer_config['NUM_BRANCHES']
        num_blocks = layer_config['NUM_BLOCKS']
        num_channels = layer_config['NUM_CHANNELS']
        block = blocks_dict[layer_config['BLOCK']]
        fuse_method = layer_config['FUSE_METHOD']

        modules = []
        for i in range(num_modules):
            # multi_scale_output is only used last module
            if not multi_scale_output and i == num_modules - 1:
                reset_multi_scale_output = False
            else:
                reset_multi_scale_output = True
            modules.append(
                HighResolutionModule(num_branches,
                                     block,
                                     num_blocks,
                                     num_inchannels,
                                     num_channels,
                                     fuse_method,
                                     reset_multi_scale_output)
            )
            num_inchannels = modules[-1].get_num_inchannels()

        return nn.Sequential(*modules), num_inchannels

    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.conv2(x)
        x = self.bn2(x)
        x = self.relu(x)
        x = self.layer1(x)

        x_list = []
        for i in range(self.stage2_cfg['NUM_BRANCHES']):
            if self.transition1[i] is not None:
                x_list.append(self.transition1[i](x))
            else:
                x_list.append(x)
        y_list = self.stage2(x_list)

        x_list = []
        for i in range(self.stage3_cfg['NUM_BRANCHES']):
            if self.transition2[i] is not None:
                if i < self.stage2_cfg['NUM_BRANCHES']:
                    x_list.append(self.transition2[i](y_list[i]))
                else:
                    x_list.append(self.transition2[i](y_list[-1]))
            else:
                x_list.append(y_list[i])
        y_list = self.stage3(x_list)

        x_list = []
        for i in range(self.stage4_cfg['NUM_BRANCHES']):
            if self.transition3[i] is not None:
                if i < self.stage3_cfg['NUM_BRANCHES']:
                    x_list.append(self.transition3[i](y_list[i]))
                else:
                    x_list.append(self.transition3[i](y_list[-1]))
            else:
                x_list.append(y_list[i])
        x = self.stage4(x_list)

        # Upsampling
        x0_h, x0_w = x[0].size(2), x[0].size(3)
        x1 = F.interpolate(x[1], size=(x0_h, x0_w),
                           mode='bilinear', align_corners=ALIGN_CORNERS)
        x2 = F.interpolate(x[2], size=(x0_h, x0_w),
                           mode='bilinear', align_corners=ALIGN_CORNERS)
        x3 = F.interpolate(x[3], size=(x0_h, x0_w),
                           mode='bilinear', align_corners=ALIGN_CORNERS)

        feats = torch.cat([x[0], x1, x2, x3], 1)

        out_aux_seg = []

        # ocr
        out_aux = self.aux_head(feats)
        # compute contrast feature
        feats = self.conv3x3_ocr(feats)

        context = self.ocr_gather_head(feats, out_aux)
        feats = self.ocr_distri_head(feats, context)

        out = self.cls_head(feats)

        out_aux_seg.append(out_aux)
        out_aux_seg.append(out)

        return out_aux_seg

    def init_weights(self, pretrained='',):
        logger.info('=> init weights from normal distribution')
        for name, m in self.named_modules():
            if any(part in name for part in {'cls', 'aux', 'ocr'}):
                # print('skipped', name)
                continue
            if isinstance(m, nn.Conv2d):
                nn.init.normal_(m.weight, std=0.001)
            elif isinstance(m, BatchNorm2d_class):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)
        if os.path.isfile(pretrained):
            pretrained_dict = torch.load(
                pretrained, map_location={'cuda:0': 'cpu'})
            logger.info('=> loading pretrained model {}'.format(pretrained))
            model_dict = self.state_dict()
            pretrained_dict = {k.replace('last_layer', 'aux_head').replace(
                'model.', ''): v for k, v in pretrained_dict.items()}
            print(set(model_dict) - set(pretrained_dict))
            print(set(pretrained_dict) - set(model_dict))
            pretrained_dict = {k: v for k, v in pretrained_dict.items()
                               if k in model_dict.keys()}
            # for k, _ in pretrained_dict.items():
            # logger.info(
            #     '=> loading {} pretrained model {}'.format(k, pretrained))
            model_dict.update(pretrained_dict)
            self.load_state_dict(model_dict)
        elif pretrained:
            raise RuntimeError('No such file {}'.format(pretrained))

relu

fuse_layer

Test model

def get_seg_model(cfg, **kwargs):
    model = HighResolutionNet(cfg, **kwargs)
    # model.init_weights(cfg.MODEL.PRETRAINED)

    return model


model = get_seg_model(config)
model.cuda()
x = torch.randn(1, 3, 256, 256).cuda()
out = model(x)
print(out)

HighResolutionNet(
  (conv1): Conv2d(3, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
  (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
  (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
  (relu): ReLU(inplace=True)
  (layer1): Sequential(
    (0): Bottleneck(
      (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (relu): ReLU(inplace=True)
      (downsample): Sequential(
        (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      )
    )
    (1): Bottleneck(
      (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (relu): ReLU(inplace=True)
    )
    (2): Bottleneck(
      (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (relu): ReLU(inplace=True)
    )
    (3): Bottleneck(
      (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (relu): ReLU(inplace=True)
    )
  )
  (transition1): ModuleList(
    (0): Sequential(
      (0): Conv2d(256, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
      (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (2): ReLU(inplace=True)
    )
    (1): Sequential(
      (0): Sequential(
        (0): Conv2d(256, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (2): ReLU(inplace=True)
      )
    )
  )
  (stage2): Sequential(
    (0): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
        )
      )
      (relu): ReLU(inplace=True)
    )
  )
  (transition2): ModuleList(
    (0): None
    (1): None
    (2): Sequential(
      (0): Sequential(
        (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (2): ReLU(inplace=True)
      )
    )
  )
  (stage3): Sequential(
    (0): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
        )
      )
      (relu): ReLU(inplace=True)
    )
    (1): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
        )
      )
      (relu): ReLU(inplace=True)
    )
    (2): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
        )
      )
      (relu): ReLU(inplace=True)
    )
    (3): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
        )
      )
      (relu): ReLU(inplace=True)
    )
  )
  (transition3): ModuleList(
    (0): None
    (1): None
    (2): None
    (3): Sequential(
      (0): Sequential(
        (0): Conv2d(192, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (2): ReLU(inplace=True)
      )
    )
  )
  (stage4): Sequential(
    (0): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
          (3): Sequential(
            (0): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (2): Sequential(
              (0): Conv2d(48, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(96, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): Sequential(
            (0): Sequential(
              (0): Conv2d(192, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (3): None
        )
      )
      (relu): ReLU(inplace=True)
    )
    (1): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
          (3): Sequential(
            (0): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (2): Sequential(
              (0): Conv2d(48, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(96, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): Sequential(
            (0): Sequential(
              (0): Conv2d(192, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (3): None
        )
      )
      (relu): ReLU(inplace=True)
    )
    (2): HighResolutionModule(
      (branches): ModuleList(
        (0): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(48, 48, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): Sequential(
          (0): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): BasicBlock(
            (conv1): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (relu): ReLU(inplace=True)
            (conv2): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
            (bn2): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
      )
      (fuse_layers): ModuleList(
        (0): ModuleList(
          (0): None
          (1): Sequential(
            (0): Conv2d(96, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): Sequential(
            (0): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (1): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): None
          (2): Sequential(
            (0): Conv2d(192, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): Sequential(
            (0): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): None
          (3): Sequential(
            (0): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): ModuleList(
          (0): Sequential(
            (0): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(48, 48, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (2): Sequential(
              (0): Conv2d(48, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (1): Sequential(
            (0): Sequential(
              (0): Conv2d(96, 96, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(96, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
              (2): ReLU(inplace=True)
            )
            (1): Sequential(
              (0): Conv2d(96, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (2): Sequential(
            (0): Sequential(
              (0): Conv2d(192, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
              (1): BatchNorm2d(384, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            )
          )
          (3): None
        )
      )
      (relu): ReLU(inplace=True)
    )
  )
  (conv3x3_ocr): Sequential(
    (0): Conv2d(720, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (2): ReLU(inplace=True)
  )
  (ocr_gather_head): SpatialGather_Module()
  (ocr_distri_head): SpatialOCR_Module(
    (object_context_block): ObjectAttentionBlock2D(
      (pool): MaxPool2d(kernel_size=(1, 1), stride=(1, 1), padding=0, dilation=1, ceil_mode=False)
      (f_pixel): Sequential(
        (0): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): Sequential(
          (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
        (2): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (3): Sequential(
          (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
      )
      (f_object): Sequential(
        (0): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): Sequential(
          (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
        (2): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (3): Sequential(
          (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
      )
      (f_down): Sequential(
        (0): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): Sequential(
          (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
      )
      (f_up): Sequential(
        (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): Sequential(
          (0): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          (1): ReLU()
        )
      )
    )
    (conv_bn_dropout): Sequential(
      (0): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (1): Sequential(
        (0): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (1): ReLU()
      )
      (2): Dropout2d(p=0.05, inplace=False)
    )
  )
  (cls_head): Conv2d(512, 19, kernel_size=(1, 1), stride=(1, 1))
  (aux_head): Sequential(
    (0): Conv2d(720, 720, kernel_size=(1, 1), stride=(1, 1))
    (1): BatchNorm2d(720, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (2): ReLU(inplace=True)
    (3): Conv2d(720, 19, kernel_size=(1, 1), stride=(1, 1))
  )
)

[tensor([[[[ 0.0126,  0.2089,  0.1555,  ..., -0.2633, -0.2207, -0.1966],
          [-0.0301,  0.0580,  0.3048,  ..., -0.0795, -0.3500, -0.5949],
          [-0.0560,  0.4462,  0.6888,  ..., -0.0914, -0.1904, -0.5955],
          ...,
          [-0.2828, -0.3316, -0.3327,  ...,  0.0599, -0.1506,  0.1004],
          [-0.4286, -0.1721, -0.4987,  ...,  0.1643, -0.1669, -0.2276],
          [-0.4279, -0.4016, -0.3705,  ..., -0.2004, -0.3496, -0.6257]],

         [[ 0.2175,  0.1642,  0.2870,  ..., -0.2048, -0.1047, -0.1548],
          [ 0.1737,  0.1060,  0.1426,  ..., -0.0630, -0.2284, -0.1860],
          [ 0.1276, -0.0596, -0.0157,  ..., -0.3222, -0.4208, -0.2417],
          ...,
          [ 0.2149, -0.1225,  0.2284,  ..., -0.0518, -0.3060, -0.1962],
          [-0.1730, -0.2476,  0.0653,  ..., -0.0104, -0.0105, -0.0807],
          [-0.3364, -0.3119, -0.2409,  ..., -0.0378,  0.0649,  0.0647]],

         [[-0.5266, -0.3146, -0.0793,  ..., -0.0764,  0.0503,  0.1208],
          [-0.6032, -0.5813, -0.3442,  ..., -0.4756, -0.2503, -0.1254],
          [-0.2866, -0.6633, -0.3280,  ..., -0.3490, -0.3198,  0.1013],
          ...,
          [-0.2102, -0.1872, -0.3686,  ...,  0.2620,  0.3569,  0.5164],
          [-0.2391, -0.0143, -0.3167,  ...,  0.1354,  0.4852,  0.4345],
          [-0.0980, -0.0999, -0.2332,  ..., -0.0388,  0.2441,  0.2310]],

         ...,

         [[-0.1700, -0.0286, -0.0042,  ..., -0.8255, -0.8084, -0.5332],
          [ 0.0808,  0.1889,  0.0577,  ..., -0.6874, -0.4808, -0.2507],
          [ 0.1011,  0.1252, -0.0381,  ..., -0.4298, -0.4084, -0.1023],
          ...,
          [ 0.2143,  0.0511,  0.0697,  ..., -0.8343, -0.7336, -0.4068],
          [ 0.3050, -0.1730, -0.1157,  ..., -0.8594, -0.7523, -0.6235],
          [ 0.1338, -0.2147, -0.3771,  ..., -0.7364, -0.7153, -0.7533]],

         [[-0.6849, -0.3476, -0.0539,  ..., -0.2995, -0.1954, -0.1797],
          [-0.6074, -0.4295, -0.3755,  ..., -0.4582, -0.4791, -0.5893],
          [-0.5042, -0.6442, -0.4120,  ..., -0.7364, -0.8572, -0.6877],
          ...,
          [ 0.1772,  0.1758,  0.1408,  ...,  0.2219,  0.3596,  0.1636],
          [ 0.0977,  0.2228,  0.0980,  ..., -0.2784,  0.1537,  0.0426],
          [ 0.2082,  0.1112,  0.3127,  ..., -0.1324,  0.2484,  0.0712]],

         [[-0.7520, -0.7920, -0.8023,  ..., -0.3344, -0.4756, -0.3786],
          [-0.5891, -0.5307, -0.6568,  ..., -0.4531, -0.4720, -0.4871],
          [-0.7524, -0.7254, -0.7203,  ..., -0.5278, -0.4140, -0.3738],
          ...,
          [-0.3742, -0.3159, -0.2781,  ..., -0.8083, -0.8444, -1.0273],
          [-0.3527, -0.2352, -0.1589,  ..., -0.6653, -0.9925, -1.2293],
          [-0.4631, -0.5654, -0.5520,  ..., -1.1085, -1.1969, -1.6035]]]],
       device='cuda:0', grad_fn=<CudnnConvolutionBackward>), tensor([[[[-1.1921e+00, -8.7566e-01, -7.3220e-01,  ..., -7.0549e-01,
           -1.0782e+00,  1.6466e-01],
          [-5.1973e-01, -3.2829e-01, -4.0623e-02,  ..., -3.4792e-01,
           -2.0699e-01,  6.5108e-02],
          [-5.8162e-01,  4.2323e-03,  1.0330e-01,  ..., -7.2836e-03,
            5.7834e-02, -4.0695e-02],
          ...,
          [-2.0400e-01, -1.0644e-01, -5.4366e-02,  ..., -1.6851e-01,
           -1.5228e-01,  4.7855e-01],
          [-4.7569e-01,  9.6399e-02, -2.8926e-01,  ..., -8.0138e-02,
           -5.0628e-01,  3.7228e-01],
          [ 9.3950e-02, -4.9347e-02, -1.0575e-01,  ..., -9.6701e-01,
           -1.0447e+00, -2.0369e-02]],

         [[ 1.3028e+00,  6.7632e-01,  9.9330e-01,  ...,  9.1765e-01,
            8.9304e-01,  9.7930e-01],
          [ 6.8199e-01,  4.9716e-01,  3.4814e-01,  ...,  7.7114e-01,
            6.6421e-01,  3.5449e-01],
          [ 9.0915e-01,  1.0235e-01,  1.8720e-02,  ...,  4.4257e-01,
            3.4798e-01,  2.1372e-01],
          ...,
          [ 9.0561e-01, -1.4279e-01,  5.0179e-02,  ...,  1.8157e-01,
           -1.3743e-01,  2.9420e-01],
          [ 9.3803e-01, -1.0436e-02,  1.4767e-01,  ...,  4.9976e-02,
            1.0739e-01,  2.6171e-01],
          [ 9.2503e-01,  3.5870e-01,  4.8831e-01,  ...,  8.4631e-01,
            3.8659e-01,  3.3774e-01]],

         [[ 2.0981e-01,  6.1833e-01,  9.6794e-01,  ...,  2.8238e-01,
            2.5978e-01,  4.8613e-01],
          [-1.2846e-01,  6.7296e-02, -6.3639e-02,  ..., -4.2966e-01,
            3.1935e-01,  5.5187e-01],
          [ 2.5797e-01, -1.2030e-01, -2.0252e-01,  ...,  1.9528e-01,
            1.5962e-01,  3.3269e-01],
          ...,
          [-2.0601e-02,  8.2864e-02, -3.6000e-01,  ...,  3.5094e-01,
           -5.5201e-02, -2.5831e-02],
          [-3.8858e-01, -4.6351e-02, -2.8292e-01,  ...,  3.5836e-02,
           -2.4179e-01, -1.3999e-03],
          [-8.9333e-02,  4.3436e-01,  4.7767e-01,  ...,  9.1226e-01,
            7.1973e-01,  5.8763e-01]],

         ...,

         [[ 8.5679e-01,  1.1779e-01,  3.7987e-01,  ...,  3.8945e-01,
            5.4756e-01,  5.8113e-01],
          [ 2.1517e-01,  2.9140e-01,  8.3521e-02,  ...,  4.1642e-01,
            4.8541e-02,  5.7063e-01],
          [ 4.2755e-01,  8.5772e-02,  4.9619e-02,  ...,  1.8989e-01,
            1.8816e-03,  4.7896e-01],
          ...,
          [ 3.2219e-02, -3.5875e-02,  2.6908e-01,  ...,  2.4972e-01,
            2.1703e-01,  8.5617e-01],
          [ 3.2784e-01, -6.2063e-02,  1.2819e-01,  ...,  1.4086e-01,
            2.0888e-01,  1.0915e+00],
          [ 2.8844e-01,  5.4034e-02,  1.2540e-02,  ...,  1.4383e+00,
            1.2895e+00,  1.3019e+00]],

         [[ 3.3014e-01,  2.7361e-01, -6.8319e-01,  ...,  1.3882e-01,
            9.3643e-02,  1.0667e-02],
          [ 2.2844e-01, -1.8923e-01, -3.9863e-01,  ..., -5.1912e-01,
           -7.6968e-01, -4.6036e-01],
          [ 6.4837e-02, -1.6802e-01, -1.0226e-01,  ..., -7.3331e-01,
           -7.6095e-01, -6.2935e-01],
          ...,
          [ 2.4629e-01, -3.2840e-01,  6.8417e-02,  ..., -2.9780e-01,
           -2.2339e-01, -6.2761e-01],
          [-2.2035e-01, -2.5319e-01,  7.7725e-03,  ..., -1.8956e-01,
           -4.1416e-01, -6.3720e-01],
          [-3.2191e-01, -2.5960e-01, -4.0099e-01,  ..., -5.5858e-01,
           -6.1042e-01, -3.6042e-01]],

         [[-2.0594e-01,  3.6297e-01,  7.4933e-01,  ...,  6.1509e-01,
            6.3861e-01,  9.7792e-02],
          [ 7.0288e-01,  1.9002e-01,  2.3223e-02,  ...,  1.9716e-01,
            2.1533e-01,  1.1496e-01],
          [ 5.7000e-01,  2.8020e-01, -5.3618e-02,  ...,  1.0633e-01,
            3.1564e-01,  2.8380e-01],
          ...,
          [ 5.9550e-01,  1.5407e-01,  3.3551e-01,  ..., -1.9185e-02,
            1.3940e-01, -5.7018e-01],
          [ 4.7579e-01,  2.6538e-01,  4.5185e-01,  ...,  1.8412e-01,
            2.5192e-01, -2.6754e-01],
          [ 7.6439e-01, -1.3286e-01, -4.0669e-01,  ..., -8.3105e-01,
           -6.9897e-01,  2.7346e-01]]]], device='cuda:0',
       grad_fn=<CudnnConvolutionBackward>)]