2023/05/22 14:24:39 - mmengine - INFO - ------------------------------------------------------------ System environment: sys.platform: linux Python: 3.10.0 (default, Mar 3 2022, 09:58:08) [GCC 7.5.0] CUDA available: True numpy_random_seed: 866926666 GPU 0: NVIDIA A10 CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 11.7, V11.7.99 GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0 PyTorch: 2.0.0+cu117 PyTorch compiling details: PyTorch built with: - GCC 9.3 - C++ Version: 201703 - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 11.7 - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86 - CuDNN 8.5 - Magma 2.6.1 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.0.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.15.1+cu117 OpenCV: 4.7.0 MMEngine: 0.7.3 Runtime environment: cudnn_benchmark: False mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0} dist_cfg: {'backend': 'nccl'} seed: None Distributed launcher: none Distributed training: False GPU number: 1 ------------------------------------------------------------ 2023/05/22 14:24:39 - mmengine - INFO - Config: default_scope = 'mmpose' default_hooks = dict( timer=dict(type='IterTimerHook', _scope_='mmpose'), logger=dict(type='LoggerHook', interval=50, _scope_='mmpose'), param_scheduler=dict(type='ParamSchedulerHook', _scope_='mmpose'), checkpoint=dict( type='CheckpointHook', interval=1, _scope_='mmpose', save_best='NME', rule='less'), sampler_seed=dict(type='DistSamplerSeedHook', _scope_='mmpose'), visualization=dict( type='PoseVisualizationHook', enable=False, _scope_='mmpose')) custom_hooks = [dict(type='SyncBuffersHook', _scope_='mmpose')] env_cfg = dict( cudnn_benchmark=False, mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), dist_cfg=dict(backend='nccl')) vis_backends = [dict(type='LocalVisBackend', _scope_='mmpose')] visualizer = dict( type='PoseLocalVisualizer', vis_backends=[dict(type='LocalVisBackend')], name='visualizer', _scope_='mmpose') log_processor = dict( type='LogProcessor', window_size=50, by_epoch=True, num_digits=6, _scope_='mmpose') log_level = 'INFO' load_from = None resume = False backend_args = dict(backend='local') train_cfg = dict(by_epoch=True, max_epochs=80, val_interval=1) val_cfg = dict() test_cfg = dict() custom_imports = dict(imports=['codecs2', 'models']) optim_wrapper = dict( optimizer=dict(type='AdamW', lr=0.002, weight_decay=0.0005), type='AmpOptimWrapper', loss_scale='dynamic') param_scheduler = [ dict( type='LinearLR', begin=0, end=500, start_factor=0.001, by_epoch=False), dict( type='MultiStepLR', begin=0, end=80, milestones=[40, 60], gamma=0.1, by_epoch=True) ] auto_scale_lr = dict(base_batch_size=512) codec = dict( type='SKPSHeatmap', input_size=(256, 256), heatmap_size=(64, 64), sigma=2) model = dict( type='TopdownPoseEstimator', data_preprocessor=dict( type='PoseDataPreprocessor', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], bgr_to_rgb=True), backbone=dict( type='HRNet', in_channels=3, extra=dict( stage1=dict( num_modules=1, num_branches=1, block='BOTTLENECK', num_blocks=(4, ), num_channels=(64, )), stage2=dict( num_modules=1, num_branches=2, block='BASIC', num_blocks=(4, 4), num_channels=(18, 36)), stage3=dict( num_modules=4, num_branches=3, block='BASIC', num_blocks=(4, 4, 4), num_channels=(18, 36, 72)), stage4=dict( num_modules=3, num_branches=4, block='BASIC', num_blocks=(4, 4, 4, 4), num_channels=(18, 36, 72, 144), multiscale_output=True), upsample=dict(mode='bilinear', align_corners=False)), init_cfg=dict( type='Pretrained', checkpoint='open-mmlab://msra/hrnetv2_w18')), neck=dict(type='FeatureMapProcessor', concat=True), head=dict( type='SKPSHead', in_channels=270, out_channels=98, conv_out_channels=(270, ), conv_kernel_sizes=(1, ), heatmap_loss=dict(type='AdaptiveWingLoss', use_target_weight=True), offside_loss=dict(type='AdaptiveWingLoss', use_target_weight=True), decoder=dict( type='SKPSHeatmap', input_size=(256, 256), heatmap_size=(64, 64), sigma=2)), test_cfg=dict(flip_test=True, flip_mode='heatmap', shift_heatmap=True)) dataset_type = 'WFLWDataset' data_mode = 'topdown' data_root = './data/wflw/' train_pipeline = [ dict(type='LoadImage'), dict(type='GetBBoxCenterScale'), dict(type='RandomFlip', direction='horizontal'), dict( type='Albumentation', transforms=[ dict(type='RandomBrightnessContrast', p=0.5), dict(type='HueSaturationValue', p=0.5), dict(type='GaussianBlur', p=0.5), dict(type='GaussNoise', p=0.1), dict( type='CoarseDropout', max_holes=8, max_height=0.2, max_width=0.2, min_holes=1, min_height=0.1, min_width=0.1, p=0.5) ]), dict( type='RandomBBoxTransform', shift_prob=0.0, rotate_factor=45, scale_factor=(0.75, 1.25), scale_prob=1.0), dict(type='TopdownAffine', input_size=(256, 256)), dict( type='GenerateTarget', encoder=dict( type='SKPSHeatmap', input_size=(256, 256), heatmap_size=(64, 64), sigma=2)), dict(type='PackPoseInputs') ] val_pipeline = [ dict(type='LoadImage'), dict(type='GetBBoxCenterScale'), dict(type='TopdownAffine', input_size=(256, 256)), dict(type='PackPoseInputs') ] train_dataloader = dict( batch_size=64, num_workers=4, persistent_workers=True, sampler=dict(type='DefaultSampler', shuffle=True), dataset=dict( type='WFLWDataset', data_root='./data/wflw/', data_mode='topdown', ann_file='annotations/face_landmarks_wflw_train.json', data_prefix=dict(img='images/'), pipeline=[ dict(type='LoadImage'), dict(type='GetBBoxCenterScale'), dict(type='RandomFlip', direction='horizontal'), dict( type='Albumentation', transforms=[ dict(type='RandomBrightnessContrast', p=0.5), dict(type='HueSaturationValue', p=0.5), dict(type='GaussianBlur', p=0.5), dict(type='GaussNoise', p=0.1), dict( type='CoarseDropout', max_holes=8, max_height=0.2, max_width=0.2, min_holes=1, min_height=0.1, min_width=0.1, p=0.5) ]), dict( type='RandomBBoxTransform', shift_prob=0.0, rotate_factor=45, scale_factor=(0.75, 1.25), scale_prob=1.0), dict(type='TopdownAffine', input_size=(256, 256)), dict( type='GenerateTarget', encoder=dict( type='SKPSHeatmap', input_size=(256, 256), heatmap_size=(64, 64), sigma=2)), dict(type='PackPoseInputs') ])) val_dataloader = dict( batch_size=32, num_workers=4, persistent_workers=True, drop_last=False, sampler=dict(type='DefaultSampler', shuffle=False, round_up=False), dataset=dict( type='WFLWDataset', data_root='./data/wflw/', data_mode='topdown', ann_file='annotations/face_landmarks_wflw_test.json', data_prefix=dict(img='images/'), test_mode=True, pipeline=[ dict(type='LoadImage'), dict(type='GetBBoxCenterScale'), dict(type='TopdownAffine', input_size=(256, 256)), dict(type='PackPoseInputs') ])) test_dataloader = dict( batch_size=32, num_workers=4, persistent_workers=True, drop_last=False, sampler=dict(type='DefaultSampler', shuffle=False, round_up=False), dataset=dict( type='WFLWDataset', data_root='./data/wflw/', data_mode='topdown', ann_file='annotations/face_landmarks_wflw_test.json', data_prefix=dict(img='images/'), test_mode=True, pipeline=[ dict(type='LoadImage'), dict(type='GetBBoxCenterScale'), dict(type='TopdownAffine', input_size=(256, 256)), dict(type='PackPoseInputs') ])) val_evaluator = dict(type='NME', norm_mode='keypoint_distance') test_evaluator = dict(type='NME', norm_mode='keypoint_distance') launcher = 'none' work_dir = './work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256' 2023/05/22 14:24:41 - mmengine - INFO - Distributed training is not used, all SyncBatchNorm (SyncBN) layers in the model will be automatically reverted to BatchNormXd layers if they are used. 2023/05/22 14:24:41 - mmengine - INFO - Hooks will be executed in the following order: before_run: (VERY_HIGH ) RuntimeInfoHook (BELOW_NORMAL) LoggerHook -------------------- before_train: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (VERY_LOW ) CheckpointHook -------------------- before_train_epoch: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (NORMAL ) DistSamplerSeedHook -------------------- before_train_iter: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook -------------------- after_train_iter: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (BELOW_NORMAL) LoggerHook (LOW ) ParamSchedulerHook (VERY_LOW ) CheckpointHook -------------------- after_train_epoch: (NORMAL ) IterTimerHook (NORMAL ) SyncBuffersHook (LOW ) ParamSchedulerHook (VERY_LOW ) CheckpointHook -------------------- before_val_epoch: (NORMAL ) IterTimerHook (NORMAL ) SyncBuffersHook -------------------- before_val_iter: (NORMAL ) IterTimerHook -------------------- after_val_iter: (NORMAL ) IterTimerHook (NORMAL ) PoseVisualizationHook (BELOW_NORMAL) LoggerHook -------------------- after_val_epoch: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (BELOW_NORMAL) LoggerHook (LOW ) ParamSchedulerHook (VERY_LOW ) CheckpointHook -------------------- after_train: (VERY_LOW ) CheckpointHook -------------------- before_test_epoch: (NORMAL ) IterTimerHook -------------------- before_test_iter: (NORMAL ) IterTimerHook -------------------- after_test_iter: (NORMAL ) IterTimerHook (NORMAL ) PoseVisualizationHook (BELOW_NORMAL) LoggerHook -------------------- after_test_epoch: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (BELOW_NORMAL) LoggerHook -------------------- after_run: (BELOW_NORMAL) LoggerHook -------------------- 2023/05/22 14:24:44 - mmengine - WARNING - The prefix is not set in metric class NME. 2023/05/22 14:24:44 - mmengine - INFO - load model from: open-mmlab://msra/hrnetv2_w18 2023/05/22 14:24:44 - mmengine - INFO - Loads checkpoint by openmmlab backend from path: open-mmlab://msra/hrnetv2_w18 2023/05/22 14:24:44 - mmengine - WARNING - The model and loaded state dict do not match exactly unexpected key in source state_dict: incre_modules.0.0.conv1.weight, incre_modules.0.0.bn1.weight, incre_modules.0.0.bn1.bias, incre_modules.0.0.bn1.running_mean, incre_modules.0.0.bn1.running_var, incre_modules.0.0.bn1.num_batches_tracked, incre_modules.0.0.conv2.weight, incre_modules.0.0.bn2.weight, incre_modules.0.0.bn2.bias, incre_modules.0.0.bn2.running_mean, incre_modules.0.0.bn2.running_var, incre_modules.0.0.bn2.num_batches_tracked, incre_modules.0.0.conv3.weight, incre_modules.0.0.bn3.weight, incre_modules.0.0.bn3.bias, incre_modules.0.0.bn3.running_mean, incre_modules.0.0.bn3.running_var, incre_modules.0.0.bn3.num_batches_tracked, incre_modules.0.0.downsample.0.weight, incre_modules.0.0.downsample.1.weight, incre_modules.0.0.downsample.1.bias, incre_modules.0.0.downsample.1.running_mean, incre_modules.0.0.downsample.1.running_var, incre_modules.0.0.downsample.1.num_batches_tracked, incre_modules.1.0.conv1.weight, incre_modules.1.0.bn1.weight, incre_modules.1.0.bn1.bias, incre_modules.1.0.bn1.running_mean, incre_modules.1.0.bn1.running_var, incre_modules.1.0.bn1.num_batches_tracked, incre_modules.1.0.conv2.weight, incre_modules.1.0.bn2.weight, incre_modules.1.0.bn2.bias, incre_modules.1.0.bn2.running_mean, incre_modules.1.0.bn2.running_var, incre_modules.1.0.bn2.num_batches_tracked, incre_modules.1.0.conv3.weight, incre_modules.1.0.bn3.weight, incre_modules.1.0.bn3.bias, incre_modules.1.0.bn3.running_mean, incre_modules.1.0.bn3.running_var, incre_modules.1.0.bn3.num_batches_tracked, incre_modules.1.0.downsample.0.weight, incre_modules.1.0.downsample.1.weight, incre_modules.1.0.downsample.1.bias, incre_modules.1.0.downsample.1.running_mean, incre_modules.1.0.downsample.1.running_var, incre_modules.1.0.downsample.1.num_batches_tracked, incre_modules.2.0.conv1.weight, incre_modules.2.0.bn1.weight, incre_modules.2.0.bn1.bias, incre_modules.2.0.bn1.running_mean, incre_modules.2.0.bn1.running_var, incre_modules.2.0.bn1.num_batches_tracked, incre_modules.2.0.conv2.weight, incre_modules.2.0.bn2.weight, incre_modules.2.0.bn2.bias, incre_modules.2.0.bn2.running_mean, incre_modules.2.0.bn2.running_var, incre_modules.2.0.bn2.num_batches_tracked, incre_modules.2.0.conv3.weight, incre_modules.2.0.bn3.weight, incre_modules.2.0.bn3.bias, incre_modules.2.0.bn3.running_mean, incre_modules.2.0.bn3.running_var, incre_modules.2.0.bn3.num_batches_tracked, incre_modules.2.0.downsample.0.weight, incre_modules.2.0.downsample.1.weight, incre_modules.2.0.downsample.1.bias, incre_modules.2.0.downsample.1.running_mean, incre_modules.2.0.downsample.1.running_var, incre_modules.2.0.downsample.1.num_batches_tracked, incre_modules.3.0.conv1.weight, incre_modules.3.0.bn1.weight, incre_modules.3.0.bn1.bias, incre_modules.3.0.bn1.running_mean, incre_modules.3.0.bn1.running_var, incre_modules.3.0.bn1.num_batches_tracked, incre_modules.3.0.conv2.weight, incre_modules.3.0.bn2.weight, incre_modules.3.0.bn2.bias, incre_modules.3.0.bn2.running_mean, incre_modules.3.0.bn2.running_var, incre_modules.3.0.bn2.num_batches_tracked, incre_modules.3.0.conv3.weight, incre_modules.3.0.bn3.weight, incre_modules.3.0.bn3.bias, incre_modules.3.0.bn3.running_mean, incre_modules.3.0.bn3.running_var, incre_modules.3.0.bn3.num_batches_tracked, incre_modules.3.0.downsample.0.weight, incre_modules.3.0.downsample.1.weight, incre_modules.3.0.downsample.1.bias, incre_modules.3.0.downsample.1.running_mean, incre_modules.3.0.downsample.1.running_var, incre_modules.3.0.downsample.1.num_batches_tracked, downsamp_modules.0.0.weight, downsamp_modules.0.0.bias, downsamp_modules.0.1.weight, downsamp_modules.0.1.bias, downsamp_modules.0.1.running_mean, downsamp_modules.0.1.running_var, downsamp_modules.0.1.num_batches_tracked, downsamp_modules.1.0.weight, downsamp_modules.1.0.bias, downsamp_modules.1.1.weight, downsamp_modules.1.1.bias, downsamp_modules.1.1.running_mean, downsamp_modules.1.1.running_var, downsamp_modules.1.1.num_batches_tracked, downsamp_modules.2.0.weight, downsamp_modules.2.0.bias, downsamp_modules.2.1.weight, downsamp_modules.2.1.bias, downsamp_modules.2.1.running_mean, downsamp_modules.2.1.running_var, downsamp_modules.2.1.num_batches_tracked, final_layer.0.weight, final_layer.0.bias, final_layer.1.weight, final_layer.1.bias, final_layer.1.running_mean, final_layer.1.running_var, final_layer.1.num_batches_tracked, classifier.weight, classifier.bias Name of parameter - Initialization information backbone.conv1.weight - torch.Size([64, 3, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.bn1.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.bn1.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.bn2.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.bn2.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.conv1.weight - torch.Size([64, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn1.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn1.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn2.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn2.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn3.weight - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.bn3.bias - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.downsample.0.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.downsample.1.weight - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.0.downsample.1.bias - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.conv1.weight - torch.Size([64, 256, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn1.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn1.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn2.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn2.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn3.weight - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.1.bn3.bias - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.conv1.weight - torch.Size([64, 256, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn1.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn1.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn2.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn2.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn3.weight - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.2.bn3.bias - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.conv1.weight - torch.Size([64, 256, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn1.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn1.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn2.weight - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn2.bias - torch.Size([64]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn3.weight - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.layer1.3.bn3.bias - torch.Size([256]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.0.0.weight - torch.Size([18, 256, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.1.0.0.weight - torch.Size([36, 256, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.1.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition1.1.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage2.0.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition2.2.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition2.2.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition2.2.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.0.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.1.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.2.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage3.3.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition3.3.0.0.weight - torch.Size([144, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition3.3.0.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.transition3.3.0.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.0.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.1.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.2.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.branches.3.3.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.3.0.weight - torch.Size([18, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.3.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.0.3.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.3.0.weight - torch.Size([36, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.3.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.1.3.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.3.0.weight - torch.Size([72, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.3.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.2.3.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.1.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.2.0.weight - torch.Size([144, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.2.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.0.2.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.0.0.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.1.0.weight - torch.Size([144, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.1.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.1.1.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.2.0.0.weight - torch.Size([144, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.2.0.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.0.fuse_layers.3.2.0.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.0.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.1.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.2.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.branches.3.3.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.3.0.weight - torch.Size([18, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.3.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.0.3.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.3.0.weight - torch.Size([36, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.3.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.1.3.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.3.0.weight - torch.Size([72, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.3.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.2.3.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.1.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.2.0.weight - torch.Size([144, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.2.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.0.2.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.0.0.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.1.0.weight - torch.Size([144, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.1.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.1.1.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.2.0.0.weight - torch.Size([144, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.2.0.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.1.fuse_layers.3.2.0.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.0.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.1.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.2.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.conv1.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.bn1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.bn1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.conv2.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.bn2.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.0.3.bn2.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.0.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.1.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.2.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.conv1.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.bn1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.bn1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.conv2.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.bn2.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.1.3.bn2.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.0.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.1.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.2.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.conv1.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.bn1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.bn1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.conv2.weight - torch.Size([72, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.bn2.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.2.3.bn2.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.0.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.1.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.2.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.conv1.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.bn1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.bn1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.conv2.weight - torch.Size([144, 144, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.bn2.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.branches.3.3.bn2.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.1.0.weight - torch.Size([18, 36, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.2.0.weight - torch.Size([18, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.2.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.2.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.3.0.weight - torch.Size([18, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.3.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.0.3.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.0.0.0.weight - torch.Size([36, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.0.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.0.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.2.0.weight - torch.Size([36, 72, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.2.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.2.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.3.0.weight - torch.Size([36, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.3.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.1.3.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.1.0.weight - torch.Size([72, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.1.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.0.1.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.1.0.0.weight - torch.Size([72, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.1.0.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.1.0.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.3.0.weight - torch.Size([72, 144, 1, 1]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.3.1.weight - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.2.3.1.bias - torch.Size([72]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.0.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.0.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.0.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.1.0.weight - torch.Size([18, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.1.1.weight - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.1.1.bias - torch.Size([18]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.2.0.weight - torch.Size([144, 18, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.2.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.0.2.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.0.0.weight - torch.Size([36, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.0.1.weight - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.0.1.bias - torch.Size([36]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.1.0.weight - torch.Size([144, 36, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.1.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.1.1.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.2.0.0.weight - torch.Size([144, 72, 3, 3]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.2.0.1.weight - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 backbone.stage4.2.fuse_layers.3.2.0.1.bias - torch.Size([144]): PretrainedInit: load from open-mmlab://msra/hrnetv2_w18 head.conv_layers.0.weight - torch.Size([270, 270, 1, 1]): NormalInit: mean=0, std=0.01, bias=0 head.conv_layers.0.bias - torch.Size([270]): NormalInit: mean=0, std=0.01, bias=0 head.conv_layers.1.weight - torch.Size([270]): The value is the same before and after calling `init_weights` of TopdownPoseEstimator head.conv_layers.1.bias - torch.Size([270]): The value is the same before and after calling `init_weights` of TopdownPoseEstimator head.final_layer.weight - torch.Size([294, 270, 1, 1]): NormalInit: mean=0, std=0.01, bias=0 head.final_layer.bias - torch.Size([294]): NormalInit: mean=0, std=0.01, bias=0 2023/05/22 14:24:44 - mmengine - WARNING - "FileClient" will be deprecated in future. Please use io functions in https://mmengine.readthedocs.io/en/latest/api/fileio.html#file-io 2023/05/22 14:24:44 - mmengine - WARNING - "HardDiskBackend" is the alias of "LocalBackend" and the former will be deprecated in future. 2023/05/22 14:24:44 - mmengine - INFO - Checkpoints will be saved to /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256. 2023/05/22 14:26:34 - mmengine - INFO - Epoch(train) [1][ 50/118] lr: 1.981964e-04 eta: 5:42:12 time: 2.186608 data_time: 1.325928 memory: 10706 loss: 0.152405 loss/heatmap: 0.095323 loss/offside: 0.057082 2023/05/22 14:27:39 - mmengine - INFO - Epoch(train) [1][100/118] lr: 3.983968e-04 eta: 4:32:11 time: 1.310512 data_time: 0.507769 memory: 10706 loss: 0.089042 loss/heatmap: 0.036436 loss/offside: 0.052606 2023/05/22 14:28:04 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:28:04 - mmengine - INFO - Saving checkpoint at 1 epochs 2023/05/22 14:28:33 - mmengine - INFO - Epoch(val) [1][50/79] eta: 0:00:15 time: 0.549599 data_time: 0.030405 memory: 10706 2023/05/22 14:28:48 - mmengine - INFO - Evaluating NME... 2023/05/22 14:28:48 - mmengine - INFO - Epoch(val) [1][79/79] NME: 0.148680 data_time: 0.023552 time: 0.535792 2023/05/22 14:28:48 - mmengine - INFO - The best checkpoint with 0.1487 NME at 1 epoch is saved to best_NME_epoch_1.pth. 2023/05/22 14:30:05 - mmengine - INFO - Epoch(train) [2][ 50/118] lr: 6.706693e-04 eta: 4:14:21 time: 1.526263 data_time: 0.720371 memory: 10706 loss: 0.063270 loss/heatmap: 0.027929 loss/offside: 0.035341 2023/05/22 14:31:13 - mmengine - INFO - Epoch(train) [2][100/118] lr: 8.708697e-04 eta: 4:03:24 time: 1.374068 data_time: 0.573683 memory: 10706 loss: 0.054676 loss/heatmap: 0.023971 loss/offside: 0.030705 2023/05/22 14:31:32 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:31:32 - mmengine - INFO - Saving checkpoint at 2 epochs 2023/05/22 14:32:01 - mmengine - INFO - Epoch(val) [2][50/79] eta: 0:00:15 time: 0.549701 data_time: 0.022955 memory: 10706 2023/05/22 14:32:16 - mmengine - INFO - Evaluating NME... 2023/05/22 14:32:16 - mmengine - INFO - Epoch(val) [2][79/79] NME: 0.070201 data_time: 0.018368 time: 0.535800 2023/05/22 14:32:16 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_1.pth is removed 2023/05/22 14:32:17 - mmengine - INFO - The best checkpoint with 0.0702 NME at 2 epoch is saved to best_NME_epoch_2.pth. 2023/05/22 14:34:02 - mmengine - INFO - Epoch(train) [3][ 50/118] lr: 1.143142e-03 eta: 4:10:20 time: 2.097894 data_time: 1.297588 memory: 10706 loss: 0.049100 loss/heatmap: 0.021318 loss/offside: 0.027782 2023/05/22 14:35:24 - mmengine - INFO - Epoch(train) [3][100/118] lr: 1.343343e-03 eta: 4:09:14 time: 1.652967 data_time: 0.845425 memory: 10706 loss: 0.046482 loss/heatmap: 0.020008 loss/offside: 0.026474 2023/05/22 14:35:43 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:35:43 - mmengine - INFO - Saving checkpoint at 3 epochs 2023/05/22 14:36:13 - mmengine - INFO - Epoch(val) [3][50/79] eta: 0:00:16 time: 0.585068 data_time: 0.020932 memory: 10706 2023/05/22 14:36:30 - mmengine - INFO - Evaluating NME... 2023/05/22 14:36:30 - mmengine - INFO - Epoch(val) [3][79/79] NME: 0.068425 data_time: 0.017249 time: 0.572388 2023/05/22 14:36:30 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_2.pth is removed 2023/05/22 14:36:30 - mmengine - INFO - The best checkpoint with 0.0684 NME at 3 epoch is saved to best_NME_epoch_3.pth. 2023/05/22 14:37:50 - mmengine - INFO - Epoch(train) [4][ 50/118] lr: 1.615615e-03 eta: 4:02:26 time: 1.597838 data_time: 0.795565 memory: 10706 loss: 0.045370 loss/heatmap: 0.019296 loss/offside: 0.026074 2023/05/22 14:39:12 - mmengine - INFO - Epoch(train) [4][100/118] lr: 1.815816e-03 eta: 4:01:38 time: 1.642714 data_time: 0.841823 memory: 10706 loss: 0.043225 loss/heatmap: 0.018288 loss/offside: 0.024937 2023/05/22 14:39:31 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:39:31 - mmengine - INFO - Saving checkpoint at 4 epochs 2023/05/22 14:40:03 - mmengine - INFO - Epoch(val) [4][50/79] eta: 0:00:17 time: 0.620601 data_time: 0.021580 memory: 10706 2023/05/22 14:40:19 - mmengine - INFO - Evaluating NME... 2023/05/22 14:40:19 - mmengine - INFO - Epoch(val) [4][79/79] NME: 0.055600 data_time: 0.017547 time: 0.586249 2023/05/22 14:40:19 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_3.pth is removed 2023/05/22 14:40:20 - mmengine - INFO - The best checkpoint with 0.0556 NME at 4 epoch is saved to best_NME_epoch_4.pth. 2023/05/22 14:41:35 - mmengine - INFO - Epoch(train) [5][ 50/118] lr: 2.000000e-03 eta: 3:55:31 time: 1.505055 data_time: 0.699937 memory: 10706 loss: 0.042184 loss/heatmap: 0.017795 loss/offside: 0.024389 2023/05/22 14:42:39 - mmengine - INFO - Epoch(train) [5][100/118] lr: 2.000000e-03 eta: 3:50:20 time: 1.285686 data_time: 0.483594 memory: 10706 loss: 0.041428 loss/heatmap: 0.017165 loss/offside: 0.024263 2023/05/22 14:43:18 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:43:18 - mmengine - INFO - Saving checkpoint at 5 epochs 2023/05/22 14:43:47 - mmengine - INFO - Epoch(val) [5][50/79] eta: 0:00:16 time: 0.556618 data_time: 0.021365 memory: 10706 2023/05/22 14:44:02 - mmengine - INFO - Evaluating NME... 2023/05/22 14:44:02 - mmengine - INFO - Epoch(val) [5][79/79] NME: 0.054456 data_time: 0.017424 time: 0.538277 2023/05/22 14:44:02 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_4.pth is removed 2023/05/22 14:44:02 - mmengine - INFO - The best checkpoint with 0.0545 NME at 5 epoch is saved to best_NME_epoch_5.pth. 2023/05/22 14:45:27 - mmengine - INFO - Epoch(train) [6][ 50/118] lr: 2.000000e-03 eta: 3:52:36 time: 1.702297 data_time: 0.901821 memory: 10706 loss: 0.039775 loss/heatmap: 0.016345 loss/offside: 0.023431 2023/05/22 14:46:32 - mmengine - INFO - Epoch(train) [6][100/118] lr: 2.000000e-03 eta: 3:48:09 time: 1.290126 data_time: 0.493235 memory: 10706 loss: 0.039295 loss/heatmap: 0.016092 loss/offside: 0.023203 2023/05/22 14:46:54 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:46:54 - mmengine - INFO - Saving checkpoint at 6 epochs 2023/05/22 14:47:22 - mmengine - INFO - Epoch(val) [6][50/79] eta: 0:00:15 time: 0.535654 data_time: 0.019616 memory: 10706 2023/05/22 14:47:37 - mmengine - INFO - Evaluating NME... 2023/05/22 14:47:37 - mmengine - INFO - Epoch(val) [6][79/79] NME: 0.051857 data_time: 0.016198 time: 0.522096 2023/05/22 14:47:37 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_5.pth is removed 2023/05/22 14:47:37 - mmengine - INFO - The best checkpoint with 0.0519 NME at 6 epoch is saved to best_NME_epoch_6.pth. 2023/05/22 14:49:09 - mmengine - INFO - Epoch(train) [7][ 50/118] lr: 2.000000e-03 eta: 3:47:57 time: 1.845852 data_time: 1.039817 memory: 10706 loss: 0.038776 loss/heatmap: 0.015810 loss/offside: 0.022965 2023/05/22 14:50:26 - mmengine - INFO - Epoch(train) [7][100/118] lr: 2.000000e-03 eta: 3:46:15 time: 1.531838 data_time: 0.727238 memory: 10706 loss: 0.038263 loss/heatmap: 0.015566 loss/offside: 0.022697 2023/05/22 14:50:57 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:50:57 - mmengine - INFO - Saving checkpoint at 7 epochs 2023/05/22 14:51:26 - mmengine - INFO - Epoch(val) [7][50/79] eta: 0:00:16 time: 0.555376 data_time: 0.020565 memory: 10706 2023/05/22 14:51:41 - mmengine - INFO - Evaluating NME... 2023/05/22 14:51:41 - mmengine - INFO - Epoch(val) [7][79/79] NME: 0.049395 data_time: 0.017085 time: 0.539861 2023/05/22 14:51:41 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_6.pth is removed 2023/05/22 14:51:42 - mmengine - INFO - The best checkpoint with 0.0494 NME at 7 epoch is saved to best_NME_epoch_7.pth. 2023/05/22 14:52:57 - mmengine - INFO - Epoch(train) [8][ 50/118] lr: 2.000000e-03 eta: 3:44:19 time: 1.498886 data_time: 0.701064 memory: 10706 loss: 0.037477 loss/heatmap: 0.015138 loss/offside: 0.022339 2023/05/22 14:54:03 - mmengine - INFO - Epoch(train) [8][100/118] lr: 2.000000e-03 eta: 3:41:10 time: 1.330659 data_time: 0.525056 memory: 10706 loss: 0.037011 loss/heatmap: 0.014966 loss/offside: 0.022044 2023/05/22 14:54:20 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:54:20 - mmengine - INFO - Saving checkpoint at 8 epochs 2023/05/22 14:54:48 - mmengine - INFO - Epoch(val) [8][50/79] eta: 0:00:15 time: 0.543100 data_time: 0.019745 memory: 10706 2023/05/22 14:55:04 - mmengine - INFO - Evaluating NME... 2023/05/22 14:55:04 - mmengine - INFO - Epoch(val) [8][79/79] NME: 0.054507 data_time: 0.016386 time: 0.534208 2023/05/22 14:56:40 - mmengine - INFO - Epoch(train) [9][ 50/118] lr: 2.000000e-03 eta: 3:40:26 time: 1.924789 data_time: 1.125852 memory: 10706 loss: 0.037094 loss/heatmap: 0.014954 loss/offside: 0.022139 2023/05/22 14:56:50 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:57:42 - mmengine - INFO - Epoch(train) [9][100/118] lr: 2.000000e-03 eta: 3:36:57 time: 1.240503 data_time: 0.432254 memory: 10706 loss: 0.036106 loss/heatmap: 0.014486 loss/offside: 0.021620 2023/05/22 14:58:07 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 14:58:07 - mmengine - INFO - Saving checkpoint at 9 epochs 2023/05/22 14:58:35 - mmengine - INFO - Epoch(val) [9][50/79] eta: 0:00:15 time: 0.549979 data_time: 0.020572 memory: 10706 2023/05/22 14:58:50 - mmengine - INFO - Evaluating NME... 2023/05/22 14:58:50 - mmengine - INFO - Epoch(val) [9][79/79] NME: 0.049701 data_time: 0.016905 time: 0.533360 2023/05/22 15:00:13 - mmengine - INFO - Epoch(train) [10][ 50/118] lr: 2.000000e-03 eta: 3:35:24 time: 1.649481 data_time: 0.844547 memory: 10706 loss: 0.036074 loss/heatmap: 0.014425 loss/offside: 0.021649 2023/05/22 15:01:15 - mmengine - INFO - Epoch(train) [10][100/118] lr: 2.000000e-03 eta: 3:32:16 time: 1.241788 data_time: 0.441978 memory: 10706 loss: 0.035357 loss/heatmap: 0.014144 loss/offside: 0.021213 2023/05/22 15:01:40 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:01:40 - mmengine - INFO - Saving checkpoint at 10 epochs 2023/05/22 15:02:07 - mmengine - INFO - Epoch(val) [10][50/79] eta: 0:00:15 time: 0.535075 data_time: 0.020000 memory: 10706 2023/05/22 15:02:22 - mmengine - INFO - Evaluating NME... 2023/05/22 15:02:22 - mmengine - INFO - Epoch(val) [10][79/79] NME: 0.046529 data_time: 0.016635 time: 0.525775 2023/05/22 15:02:22 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_7.pth is removed 2023/05/22 15:02:23 - mmengine - INFO - The best checkpoint with 0.0465 NME at 10 epoch is saved to best_NME_epoch_10.pth. 2023/05/22 15:03:44 - mmengine - INFO - Epoch(train) [11][ 50/118] lr: 2.000000e-03 eta: 3:30:38 time: 1.616856 data_time: 0.814345 memory: 10706 loss: 0.035332 loss/heatmap: 0.014133 loss/offside: 0.021199 2023/05/22 15:04:45 - mmengine - INFO - Epoch(train) [11][100/118] lr: 2.000000e-03 eta: 3:27:41 time: 1.225302 data_time: 0.428909 memory: 10706 loss: 0.035000 loss/heatmap: 0.013965 loss/offside: 0.021035 2023/05/22 15:05:15 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:05:15 - mmengine - INFO - Saving checkpoint at 11 epochs 2023/05/22 15:05:44 - mmengine - INFO - Epoch(val) [11][50/79] eta: 0:00:16 time: 0.564926 data_time: 0.020553 memory: 10706 2023/05/22 15:05:59 - mmengine - INFO - Evaluating NME... 2023/05/22 15:05:59 - mmengine - INFO - Epoch(val) [11][79/79] NME: 0.047591 data_time: 0.016903 time: 0.543465 2023/05/22 15:07:17 - mmengine - INFO - Epoch(train) [12][ 50/118] lr: 2.000000e-03 eta: 3:26:19 time: 1.545344 data_time: 0.731555 memory: 10706 loss: 0.035164 loss/heatmap: 0.014009 loss/offside: 0.021155 2023/05/22 15:08:35 - mmengine - INFO - Epoch(train) [12][100/118] lr: 2.000000e-03 eta: 3:25:10 time: 1.556460 data_time: 0.747231 memory: 10706 loss: 0.034941 loss/heatmap: 0.013923 loss/offside: 0.021018 2023/05/22 15:08:52 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:08:52 - mmengine - INFO - Saving checkpoint at 12 epochs 2023/05/22 15:09:20 - mmengine - INFO - Epoch(val) [12][50/79] eta: 0:00:15 time: 0.543336 data_time: 0.022905 memory: 10706 2023/05/22 15:09:35 - mmengine - INFO - Evaluating NME... 2023/05/22 15:09:35 - mmengine - INFO - Epoch(val) [12][79/79] NME: 0.047301 data_time: 0.018451 time: 0.532865 2023/05/22 15:11:05 - mmengine - INFO - Epoch(train) [13][ 50/118] lr: 2.000000e-03 eta: 3:23:41 time: 1.798722 data_time: 0.996998 memory: 10706 loss: 0.034510 loss/heatmap: 0.013720 loss/offside: 0.020790 2023/05/22 15:11:55 - mmengine - INFO - Epoch(train) [13][100/118] lr: 2.000000e-03 eta: 3:20:06 time: 1.001660 data_time: 0.200027 memory: 10706 loss: 0.033642 loss/heatmap: 0.013359 loss/offside: 0.020283 2023/05/22 15:12:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:12:12 - mmengine - INFO - Saving checkpoint at 13 epochs 2023/05/22 15:12:41 - mmengine - INFO - Epoch(val) [13][50/79] eta: 0:00:16 time: 0.555617 data_time: 0.019441 memory: 10706 2023/05/22 15:12:56 - mmengine - INFO - Evaluating NME... 2023/05/22 15:12:56 - mmengine - INFO - Epoch(val) [13][79/79] NME: 0.045756 data_time: 0.018419 time: 0.542058 2023/05/22 15:12:56 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_10.pth is removed 2023/05/22 15:12:56 - mmengine - INFO - The best checkpoint with 0.0458 NME at 13 epoch is saved to best_NME_epoch_13.pth. 2023/05/22 15:14:09 - mmengine - INFO - Epoch(train) [14][ 50/118] lr: 2.000000e-03 eta: 3:17:14 time: 1.446921 data_time: 0.648222 memory: 10706 loss: 0.033829 loss/heatmap: 0.013406 loss/offside: 0.020423 2023/05/22 15:15:18 - mmengine - INFO - Epoch(train) [14][100/118] lr: 2.000000e-03 eta: 3:15:31 time: 1.393021 data_time: 0.592842 memory: 10706 loss: 0.034203 loss/heatmap: 0.013553 loss/offside: 0.020650 2023/05/22 15:15:38 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:15:38 - mmengine - INFO - Saving checkpoint at 14 epochs 2023/05/22 15:16:06 - mmengine - INFO - Epoch(val) [14][50/79] eta: 0:00:15 time: 0.541788 data_time: 0.028015 memory: 10706 2023/05/22 15:16:21 - mmengine - INFO - Evaluating NME... 2023/05/22 15:16:21 - mmengine - INFO - Epoch(val) [14][79/79] NME: 0.046300 data_time: 0.021653 time: 0.526870 2023/05/22 15:17:36 - mmengine - INFO - Epoch(train) [15][ 50/118] lr: 2.000000e-03 eta: 3:13:13 time: 1.497261 data_time: 0.698250 memory: 10706 loss: 0.034019 loss/heatmap: 0.013465 loss/offside: 0.020554 2023/05/22 15:18:42 - mmengine - INFO - Epoch(train) [15][100/118] lr: 2.000000e-03 eta: 3:11:20 time: 1.323907 data_time: 0.526750 memory: 10706 loss: 0.033769 loss/heatmap: 0.013313 loss/offside: 0.020456 2023/05/22 15:18:58 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:18:58 - mmengine - INFO - Saving checkpoint at 15 epochs 2023/05/22 15:19:26 - mmengine - INFO - Epoch(val) [15][50/79] eta: 0:00:15 time: 0.540908 data_time: 0.022887 memory: 10706 2023/05/22 15:19:41 - mmengine - INFO - Evaluating NME... 2023/05/22 15:19:41 - mmengine - INFO - Epoch(val) [15][79/79] NME: 0.045781 data_time: 0.018229 time: 0.526395 2023/05/22 15:20:58 - mmengine - INFO - Epoch(train) [16][ 50/118] lr: 2.000000e-03 eta: 3:09:02 time: 1.535411 data_time: 0.737637 memory: 10706 loss: 0.033762 loss/heatmap: 0.013331 loss/offside: 0.020430 2023/05/22 15:21:52 - mmengine - INFO - Epoch(train) [16][100/118] lr: 2.000000e-03 eta: 3:06:26 time: 1.084841 data_time: 0.284469 memory: 10706 loss: 0.033049 loss/heatmap: 0.013057 loss/offside: 0.019992 2023/05/22 15:22:08 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:22:08 - mmengine - INFO - Saving checkpoint at 16 epochs 2023/05/22 15:22:38 - mmengine - INFO - Epoch(val) [16][50/79] eta: 0:00:16 time: 0.580389 data_time: 0.020706 memory: 10706 2023/05/22 15:22:54 - mmengine - INFO - Evaluating NME... 2023/05/22 15:22:54 - mmengine - INFO - Epoch(val) [16][79/79] NME: 0.044643 data_time: 0.017159 time: 0.560668 2023/05/22 15:22:54 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_13.pth is removed 2023/05/22 15:22:55 - mmengine - INFO - The best checkpoint with 0.0446 NME at 16 epoch is saved to best_NME_epoch_16.pth. 2023/05/22 15:24:03 - mmengine - INFO - Epoch(train) [17][ 50/118] lr: 2.000000e-03 eta: 3:03:48 time: 1.378139 data_time: 0.569836 memory: 10706 loss: 0.033998 loss/heatmap: 0.013430 loss/offside: 0.020569 2023/05/22 15:25:16 - mmengine - INFO - Epoch(train) [17][100/118] lr: 2.000000e-03 eta: 3:02:29 time: 1.444611 data_time: 0.640335 memory: 10706 loss: 0.032868 loss/heatmap: 0.012922 loss/offside: 0.019946 2023/05/22 15:25:34 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:25:39 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:25:39 - mmengine - INFO - Saving checkpoint at 17 epochs 2023/05/22 15:26:07 - mmengine - INFO - Epoch(val) [17][50/79] eta: 0:00:15 time: 0.550271 data_time: 0.019838 memory: 10706 2023/05/22 15:26:22 - mmengine - INFO - Evaluating NME... 2023/05/22 15:26:22 - mmengine - INFO - Epoch(val) [17][79/79] NME: 0.045536 data_time: 0.016356 time: 0.532920 2023/05/22 15:27:44 - mmengine - INFO - Epoch(train) [18][ 50/118] lr: 2.000000e-03 eta: 3:01:06 time: 1.628234 data_time: 0.829672 memory: 10706 loss: 0.032780 loss/heatmap: 0.012877 loss/offside: 0.019903 2023/05/22 15:28:53 - mmengine - INFO - Epoch(train) [18][100/118] lr: 2.000000e-03 eta: 2:59:38 time: 1.388194 data_time: 0.586636 memory: 10706 loss: 0.032799 loss/heatmap: 0.012881 loss/offside: 0.019918 2023/05/22 15:29:11 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:29:11 - mmengine - INFO - Saving checkpoint at 18 epochs 2023/05/22 15:29:40 - mmengine - INFO - Epoch(val) [18][50/79] eta: 0:00:16 time: 0.563313 data_time: 0.019972 memory: 10706 2023/05/22 15:29:57 - mmengine - INFO - Evaluating NME... 2023/05/22 15:29:57 - mmengine - INFO - Epoch(val) [18][79/79] NME: 0.044246 data_time: 0.016629 time: 0.556603 2023/05/22 15:29:57 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_16.pth is removed 2023/05/22 15:29:57 - mmengine - INFO - The best checkpoint with 0.0442 NME at 18 epoch is saved to best_NME_epoch_18.pth. 2023/05/22 15:31:16 - mmengine - INFO - Epoch(train) [19][ 50/118] lr: 2.000000e-03 eta: 2:57:47 time: 1.568160 data_time: 0.761899 memory: 10706 loss: 0.032839 loss/heatmap: 0.012913 loss/offside: 0.019926 2023/05/22 15:32:07 - mmengine - INFO - Epoch(train) [19][100/118] lr: 2.000000e-03 eta: 2:55:23 time: 1.029752 data_time: 0.228620 memory: 10706 loss: 0.032559 loss/heatmap: 0.012751 loss/offside: 0.019808 2023/05/22 15:32:32 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:32:32 - mmengine - INFO - Saving checkpoint at 19 epochs 2023/05/22 15:33:03 - mmengine - INFO - Epoch(val) [19][50/79] eta: 0:00:17 time: 0.594365 data_time: 0.020814 memory: 10706 2023/05/22 15:33:21 - mmengine - INFO - Evaluating NME... 2023/05/22 15:33:21 - mmengine - INFO - Epoch(val) [19][79/79] NME: 0.043888 data_time: 0.017225 time: 0.594778 2023/05/22 15:33:21 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_18.pth is removed 2023/05/22 15:33:21 - mmengine - INFO - The best checkpoint with 0.0439 NME at 19 epoch is saved to best_NME_epoch_19.pth. 2023/05/22 15:34:36 - mmengine - INFO - Epoch(train) [20][ 50/118] lr: 2.000000e-03 eta: 2:53:46 time: 1.499228 data_time: 0.699881 memory: 10706 loss: 0.032469 loss/heatmap: 0.012755 loss/offside: 0.019714 2023/05/22 15:35:32 - mmengine - INFO - Epoch(train) [20][100/118] lr: 2.000000e-03 eta: 2:51:41 time: 1.117077 data_time: 0.319825 memory: 10706 loss: 0.032725 loss/heatmap: 0.012802 loss/offside: 0.019924 2023/05/22 15:36:03 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:36:03 - mmengine - INFO - Saving checkpoint at 20 epochs 2023/05/22 15:36:35 - mmengine - INFO - Epoch(val) [20][50/79] eta: 0:00:18 time: 0.624147 data_time: 0.021068 memory: 10706 2023/05/22 15:36:54 - mmengine - INFO - Evaluating NME... 2023/05/22 15:36:54 - mmengine - INFO - Epoch(val) [20][79/79] NME: 0.044145 data_time: 0.017445 time: 0.621905 2023/05/22 15:37:49 - mmengine - INFO - Epoch(train) [21][ 50/118] lr: 2.000000e-03 eta: 2:49:26 time: 1.105501 data_time: 0.311335 memory: 10706 loss: 0.032669 loss/heatmap: 0.012796 loss/offside: 0.019873 2023/05/22 15:39:01 - mmengine - INFO - Epoch(train) [21][100/118] lr: 2.000000e-03 eta: 2:48:12 time: 1.434863 data_time: 0.637360 memory: 10706 loss: 0.032411 loss/heatmap: 0.012704 loss/offside: 0.019707 2023/05/22 15:39:25 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:39:25 - mmengine - INFO - Saving checkpoint at 21 epochs 2023/05/22 15:39:56 - mmengine - INFO - Epoch(val) [21][50/79] eta: 0:00:17 time: 0.616084 data_time: 0.020774 memory: 10706 2023/05/22 15:40:12 - mmengine - INFO - Evaluating NME... 2023/05/22 15:40:12 - mmengine - INFO - Epoch(val) [21][79/79] NME: 0.044012 data_time: 0.016911 time: 0.583351 2023/05/22 15:41:30 - mmengine - INFO - Epoch(train) [22][ 50/118] lr: 2.000000e-03 eta: 2:46:42 time: 1.547541 data_time: 0.748418 memory: 10706 loss: 0.032602 loss/heatmap: 0.012779 loss/offside: 0.019823 2023/05/22 15:42:49 - mmengine - INFO - Epoch(train) [22][100/118] lr: 2.000000e-03 eta: 2:45:50 time: 1.594417 data_time: 0.790943 memory: 10706 loss: 0.031839 loss/heatmap: 0.012462 loss/offside: 0.019377 2023/05/22 15:43:10 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:43:10 - mmengine - INFO - Saving checkpoint at 22 epochs 2023/05/22 15:43:38 - mmengine - INFO - Epoch(val) [22][50/79] eta: 0:00:15 time: 0.537054 data_time: 0.019480 memory: 10706 2023/05/22 15:43:53 - mmengine - INFO - Evaluating NME... 2023/05/22 15:43:53 - mmengine - INFO - Epoch(val) [22][79/79] NME: 0.044070 data_time: 0.015958 time: 0.523358 2023/05/22 15:45:01 - mmengine - INFO - Epoch(train) [23][ 50/118] lr: 2.000000e-03 eta: 2:43:46 time: 1.360084 data_time: 0.557146 memory: 10706 loss: 0.031896 loss/heatmap: 0.012456 loss/offside: 0.019441 2023/05/22 15:46:01 - mmengine - INFO - Epoch(train) [23][100/118] lr: 2.000000e-03 eta: 2:42:02 time: 1.197173 data_time: 0.394509 memory: 10706 loss: 0.031936 loss/heatmap: 0.012465 loss/offside: 0.019471 2023/05/22 15:46:17 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:46:17 - mmengine - INFO - Saving checkpoint at 23 epochs 2023/05/22 15:46:46 - mmengine - INFO - Epoch(val) [23][50/79] eta: 0:00:16 time: 0.570788 data_time: 0.019718 memory: 10706 2023/05/22 15:47:01 - mmengine - INFO - Evaluating NME... 2023/05/22 15:47:01 - mmengine - INFO - Epoch(val) [23][79/79] NME: 0.043770 data_time: 0.016199 time: 0.545612 2023/05/22 15:47:01 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_19.pth is removed 2023/05/22 15:47:02 - mmengine - INFO - The best checkpoint with 0.0438 NME at 23 epoch is saved to best_NME_epoch_23.pth. 2023/05/22 15:48:37 - mmengine - INFO - Epoch(train) [24][ 50/118] lr: 2.000000e-03 eta: 2:40:57 time: 1.904219 data_time: 1.104531 memory: 10706 loss: 0.031372 loss/heatmap: 0.012258 loss/offside: 0.019114 2023/05/22 15:49:43 - mmengine - INFO - Epoch(train) [24][100/118] lr: 2.000000e-03 eta: 2:39:31 time: 1.331411 data_time: 0.529870 memory: 10706 loss: 0.031675 loss/heatmap: 0.012356 loss/offside: 0.019319 2023/05/22 15:50:08 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:50:08 - mmengine - INFO - Saving checkpoint at 24 epochs 2023/05/22 15:50:36 - mmengine - INFO - Epoch(val) [24][50/79] eta: 0:00:15 time: 0.537895 data_time: 0.019883 memory: 10706 2023/05/22 15:50:51 - mmengine - INFO - Evaluating NME... 2023/05/22 15:50:51 - mmengine - INFO - Epoch(val) [24][79/79] NME: 0.043367 data_time: 0.016405 time: 0.530486 2023/05/22 15:50:51 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_23.pth is removed 2023/05/22 15:50:52 - mmengine - INFO - The best checkpoint with 0.0434 NME at 24 epoch is saved to best_NME_epoch_24.pth. 2023/05/22 15:52:03 - mmengine - INFO - Epoch(train) [25][ 50/118] lr: 2.000000e-03 eta: 2:37:47 time: 1.430900 data_time: 0.632031 memory: 10706 loss: 0.031609 loss/heatmap: 0.012332 loss/offside: 0.019277 2023/05/22 15:52:58 - mmengine - INFO - Epoch(train) [25][100/118] lr: 2.000000e-03 eta: 2:35:58 time: 1.108741 data_time: 0.309907 memory: 10706 loss: 0.031429 loss/heatmap: 0.012272 loss/offside: 0.019156 2023/05/22 15:53:23 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:53:23 - mmengine - INFO - Saving checkpoint at 25 epochs 2023/05/22 15:53:51 - mmengine - INFO - Epoch(val) [25][50/79] eta: 0:00:15 time: 0.540483 data_time: 0.019517 memory: 10706 2023/05/22 15:54:07 - mmengine - INFO - Evaluating NME... 2023/05/22 15:54:07 - mmengine - INFO - Epoch(val) [25][79/79] NME: 0.042962 data_time: 0.016167 time: 0.536329 2023/05/22 15:54:07 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_24.pth is removed 2023/05/22 15:54:07 - mmengine - INFO - The best checkpoint with 0.0430 NME at 25 epoch is saved to best_NME_epoch_25.pth. 2023/05/22 15:55:14 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:55:14 - mmengine - INFO - Epoch(train) [26][ 50/118] lr: 2.000000e-03 eta: 2:34:06 time: 1.337915 data_time: 0.529006 memory: 10706 loss: 0.031949 loss/heatmap: 0.012456 loss/offside: 0.019494 2023/05/22 15:56:13 - mmengine - INFO - Epoch(train) [26][100/118] lr: 2.000000e-03 eta: 2:32:28 time: 1.183179 data_time: 0.381609 memory: 10706 loss: 0.031920 loss/heatmap: 0.012420 loss/offside: 0.019500 2023/05/22 15:56:34 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:56:34 - mmengine - INFO - Saving checkpoint at 26 epochs 2023/05/22 15:57:02 - mmengine - INFO - Epoch(val) [26][50/79] eta: 0:00:15 time: 0.539808 data_time: 0.019519 memory: 10706 2023/05/22 15:57:17 - mmengine - INFO - Evaluating NME... 2023/05/22 15:57:17 - mmengine - INFO - Epoch(val) [26][79/79] NME: 0.043347 data_time: 0.016101 time: 0.529422 2023/05/22 15:58:25 - mmengine - INFO - Epoch(train) [27][ 50/118] lr: 2.000000e-03 eta: 2:30:33 time: 1.353654 data_time: 0.552104 memory: 10706 loss: 0.031509 loss/heatmap: 0.012264 loss/offside: 0.019245 2023/05/22 15:59:17 - mmengine - INFO - Epoch(train) [27][100/118] lr: 2.000000e-03 eta: 2:28:43 time: 1.046152 data_time: 0.252248 memory: 10706 loss: 0.031410 loss/heatmap: 0.012236 loss/offside: 0.019174 2023/05/22 15:59:34 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 15:59:34 - mmengine - INFO - Saving checkpoint at 27 epochs 2023/05/22 16:00:02 - mmengine - INFO - Epoch(val) [27][50/79] eta: 0:00:15 time: 0.540267 data_time: 0.019472 memory: 10706 2023/05/22 16:00:17 - mmengine - INFO - Evaluating NME... 2023/05/22 16:00:17 - mmengine - INFO - Epoch(val) [27][79/79] NME: 0.043128 data_time: 0.015948 time: 0.525467 2023/05/22 16:01:11 - mmengine - INFO - Epoch(train) [28][ 50/118] lr: 2.000000e-03 eta: 2:26:17 time: 1.080442 data_time: 0.282228 memory: 10706 loss: 0.030998 loss/heatmap: 0.012048 loss/offside: 0.018950 2023/05/22 16:02:14 - mmengine - INFO - Epoch(train) [28][100/118] lr: 2.000000e-03 eta: 2:24:51 time: 1.251001 data_time: 0.449853 memory: 10706 loss: 0.031701 loss/heatmap: 0.012318 loss/offside: 0.019383 2023/05/22 16:02:38 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:02:38 - mmengine - INFO - Saving checkpoint at 28 epochs 2023/05/22 16:03:07 - mmengine - INFO - Epoch(val) [28][50/79] eta: 0:00:16 time: 0.575092 data_time: 0.020296 memory: 10706 2023/05/22 16:03:23 - mmengine - INFO - Evaluating NME... 2023/05/22 16:03:23 - mmengine - INFO - Epoch(val) [28][79/79] NME: 0.043563 data_time: 0.016551 time: 0.551376 2023/05/22 16:04:43 - mmengine - INFO - Epoch(train) [29][ 50/118] lr: 2.000000e-03 eta: 2:23:30 time: 1.610291 data_time: 0.810823 memory: 10706 loss: 0.031466 loss/heatmap: 0.012263 loss/offside: 0.019203 2023/05/22 16:05:55 - mmengine - INFO - Epoch(train) [29][100/118] lr: 2.000000e-03 eta: 2:22:22 time: 1.438381 data_time: 0.639524 memory: 10706 loss: 0.030789 loss/heatmap: 0.011986 loss/offside: 0.018804 2023/05/22 16:06:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:06:12 - mmengine - INFO - Saving checkpoint at 29 epochs 2023/05/22 16:06:40 - mmengine - INFO - Epoch(val) [29][50/79] eta: 0:00:15 time: 0.540063 data_time: 0.019087 memory: 10706 2023/05/22 16:06:55 - mmengine - INFO - Evaluating NME... 2023/05/22 16:06:55 - mmengine - INFO - Epoch(val) [29][79/79] NME: 0.043614 data_time: 0.015743 time: 0.525800 2023/05/22 16:08:10 - mmengine - INFO - Epoch(train) [30][ 50/118] lr: 2.000000e-03 eta: 2:20:39 time: 1.505351 data_time: 0.706100 memory: 10706 loss: 0.031044 loss/heatmap: 0.012063 loss/offside: 0.018981 2023/05/22 16:09:08 - mmengine - INFO - Epoch(train) [30][100/118] lr: 2.000000e-03 eta: 2:19:07 time: 1.161669 data_time: 0.362967 memory: 10706 loss: 0.030564 loss/heatmap: 0.011844 loss/offside: 0.018720 2023/05/22 16:09:28 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:09:28 - mmengine - INFO - Saving checkpoint at 30 epochs 2023/05/22 16:09:56 - mmengine - INFO - Epoch(val) [30][50/79] eta: 0:00:15 time: 0.541079 data_time: 0.022533 memory: 10706 2023/05/22 16:10:11 - mmengine - INFO - Evaluating NME... 2023/05/22 16:10:11 - mmengine - INFO - Epoch(val) [30][79/79] NME: 0.042706 data_time: 0.017865 time: 0.525558 2023/05/22 16:10:11 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_25.pth is removed 2023/05/22 16:10:11 - mmengine - INFO - The best checkpoint with 0.0427 NME at 30 epoch is saved to best_NME_epoch_30.pth. 2023/05/22 16:11:30 - mmengine - INFO - Epoch(train) [31][ 50/118] lr: 2.000000e-03 eta: 2:17:35 time: 1.564945 data_time: 0.767243 memory: 10706 loss: 0.031070 loss/heatmap: 0.012070 loss/offside: 0.019000 2023/05/22 16:12:54 - mmengine - INFO - Epoch(train) [31][100/118] lr: 2.000000e-03 eta: 2:16:47 time: 1.692464 data_time: 0.892069 memory: 10706 loss: 0.030525 loss/heatmap: 0.011824 loss/offside: 0.018701 2023/05/22 16:13:15 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:13:15 - mmengine - INFO - Saving checkpoint at 31 epochs 2023/05/22 16:13:43 - mmengine - INFO - Epoch(val) [31][50/79] eta: 0:00:15 time: 0.540969 data_time: 0.019501 memory: 10706 2023/05/22 16:13:58 - mmengine - INFO - Evaluating NME... 2023/05/22 16:13:58 - mmengine - INFO - Epoch(val) [31][79/79] NME: 0.042544 data_time: 0.016034 time: 0.526979 2023/05/22 16:13:58 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_30.pth is removed 2023/05/22 16:13:59 - mmengine - INFO - The best checkpoint with 0.0425 NME at 31 epoch is saved to best_NME_epoch_31.pth. 2023/05/22 16:15:08 - mmengine - INFO - Epoch(train) [32][ 50/118] lr: 2.000000e-03 eta: 2:15:01 time: 1.378967 data_time: 0.575801 memory: 10706 loss: 0.031122 loss/heatmap: 0.012080 loss/offside: 0.019043 2023/05/22 16:16:07 - mmengine - INFO - Epoch(train) [32][100/118] lr: 2.000000e-03 eta: 2:13:34 time: 1.194890 data_time: 0.397951 memory: 10706 loss: 0.030318 loss/heatmap: 0.011728 loss/offside: 0.018590 2023/05/22 16:16:27 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:16:27 - mmengine - INFO - Saving checkpoint at 32 epochs 2023/05/22 16:16:54 - mmengine - INFO - Epoch(val) [32][50/79] eta: 0:00:15 time: 0.536293 data_time: 0.019109 memory: 10706 2023/05/22 16:17:09 - mmengine - INFO - Evaluating NME... 2023/05/22 16:17:09 - mmengine - INFO - Epoch(val) [32][79/79] NME: 0.042802 data_time: 0.015752 time: 0.522647 2023/05/22 16:18:23 - mmengine - INFO - Epoch(train) [33][ 50/118] lr: 2.000000e-03 eta: 2:11:53 time: 1.473127 data_time: 0.679491 memory: 10706 loss: 0.030818 loss/heatmap: 0.011954 loss/offside: 0.018864 2023/05/22 16:19:30 - mmengine - INFO - Epoch(train) [33][100/118] lr: 2.000000e-03 eta: 2:10:38 time: 1.343280 data_time: 0.542728 memory: 10706 loss: 0.030329 loss/heatmap: 0.011729 loss/offside: 0.018600 2023/05/22 16:19:53 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:19:53 - mmengine - INFO - Saving checkpoint at 33 epochs 2023/05/22 16:20:21 - mmengine - INFO - Epoch(val) [33][50/79] eta: 0:00:15 time: 0.535643 data_time: 0.019340 memory: 10706 2023/05/22 16:20:37 - mmengine - INFO - Evaluating NME... 2023/05/22 16:20:37 - mmengine - INFO - Epoch(val) [33][79/79] NME: 0.042964 data_time: 0.016158 time: 0.536435 2023/05/22 16:21:47 - mmengine - INFO - Epoch(train) [34][ 50/118] lr: 2.000000e-03 eta: 2:08:59 time: 1.407311 data_time: 0.609353 memory: 10706 loss: 0.030788 loss/heatmap: 0.011936 loss/offside: 0.018852 2023/05/22 16:22:54 - mmengine - INFO - Epoch(train) [34][100/118] lr: 2.000000e-03 eta: 2:07:44 time: 1.338974 data_time: 0.536102 memory: 10706 loss: 0.030280 loss/heatmap: 0.011728 loss/offside: 0.018551 2023/05/22 16:23:00 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:23:11 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:23:11 - mmengine - INFO - Saving checkpoint at 34 epochs 2023/05/22 16:23:39 - mmengine - INFO - Epoch(val) [34][50/79] eta: 0:00:15 time: 0.540763 data_time: 0.020189 memory: 10706 2023/05/22 16:23:54 - mmengine - INFO - Evaluating NME... 2023/05/22 16:23:54 - mmengine - INFO - Epoch(val) [34][79/79] NME: 0.043122 data_time: 0.016401 time: 0.525948 2023/05/22 16:25:01 - mmengine - INFO - Epoch(train) [35][ 50/118] lr: 2.000000e-03 eta: 2:05:52 time: 1.340666 data_time: 0.544699 memory: 10706 loss: 0.030657 loss/heatmap: 0.011848 loss/offside: 0.018809 2023/05/22 16:26:01 - mmengine - INFO - Epoch(train) [35][100/118] lr: 2.000000e-03 eta: 2:04:29 time: 1.197212 data_time: 0.398451 memory: 10706 loss: 0.030498 loss/heatmap: 0.011795 loss/offside: 0.018703 2023/05/22 16:26:19 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:26:19 - mmengine - INFO - Saving checkpoint at 35 epochs 2023/05/22 16:26:47 - mmengine - INFO - Epoch(val) [35][50/79] eta: 0:00:15 time: 0.542235 data_time: 0.020149 memory: 10706 2023/05/22 16:27:02 - mmengine - INFO - Evaluating NME... 2023/05/22 16:27:02 - mmengine - INFO - Epoch(val) [35][79/79] NME: 0.042314 data_time: 0.016619 time: 0.528823 2023/05/22 16:27:02 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_31.pth is removed 2023/05/22 16:27:02 - mmengine - INFO - The best checkpoint with 0.0423 NME at 35 epoch is saved to best_NME_epoch_35.pth. 2023/05/22 16:28:07 - mmengine - INFO - Epoch(train) [36][ 50/118] lr: 2.000000e-03 eta: 2:02:37 time: 1.285280 data_time: 0.490786 memory: 10706 loss: 0.030557 loss/heatmap: 0.011831 loss/offside: 0.018726 2023/05/22 16:29:08 - mmengine - INFO - Epoch(train) [36][100/118] lr: 2.000000e-03 eta: 2:01:17 time: 1.231203 data_time: 0.441429 memory: 10706 loss: 0.030639 loss/heatmap: 0.011848 loss/offside: 0.018791 2023/05/22 16:29:26 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:29:26 - mmengine - INFO - Saving checkpoint at 36 epochs 2023/05/22 16:29:54 - mmengine - INFO - Epoch(val) [36][50/79] eta: 0:00:15 time: 0.540393 data_time: 0.019453 memory: 10706 2023/05/22 16:30:09 - mmengine - INFO - Evaluating NME... 2023/05/22 16:30:09 - mmengine - INFO - Epoch(val) [36][79/79] NME: 0.041916 data_time: 0.016000 time: 0.525091 2023/05/22 16:30:09 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_35.pth is removed 2023/05/22 16:30:10 - mmengine - INFO - The best checkpoint with 0.0419 NME at 36 epoch is saved to best_NME_epoch_36.pth. 2023/05/22 16:31:05 - mmengine - INFO - Epoch(train) [37][ 50/118] lr: 2.000000e-03 eta: 1:59:16 time: 1.113915 data_time: 0.323816 memory: 10706 loss: 0.030443 loss/heatmap: 0.011780 loss/offside: 0.018663 2023/05/22 16:31:54 - mmengine - INFO - Epoch(train) [37][100/118] lr: 2.000000e-03 eta: 1:57:43 time: 0.978741 data_time: 0.177886 memory: 10706 loss: 0.029824 loss/heatmap: 0.011532 loss/offside: 0.018292 2023/05/22 16:32:10 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:32:10 - mmengine - INFO - Saving checkpoint at 37 epochs 2023/05/22 16:32:38 - mmengine - INFO - Epoch(val) [37][50/79] eta: 0:00:15 time: 0.539687 data_time: 0.019873 memory: 10706 2023/05/22 16:32:53 - mmengine - INFO - Evaluating NME... 2023/05/22 16:32:53 - mmengine - INFO - Epoch(val) [37][79/79] NME: 0.042197 data_time: 0.016250 time: 0.525056 2023/05/22 16:34:12 - mmengine - INFO - Epoch(train) [38][ 50/118] lr: 2.000000e-03 eta: 1:56:09 time: 1.572946 data_time: 0.771102 memory: 10706 loss: 0.030389 loss/heatmap: 0.011750 loss/offside: 0.018640 2023/05/22 16:35:04 - mmengine - INFO - Epoch(train) [38][100/118] lr: 2.000000e-03 eta: 1:54:40 time: 1.041995 data_time: 0.237554 memory: 10706 loss: 0.030097 loss/heatmap: 0.011607 loss/offside: 0.018490 2023/05/22 16:35:22 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:35:22 - mmengine - INFO - Saving checkpoint at 38 epochs 2023/05/22 16:35:49 - mmengine - INFO - Epoch(val) [38][50/79] eta: 0:00:15 time: 0.541559 data_time: 0.020794 memory: 10706 2023/05/22 16:36:05 - mmengine - INFO - Evaluating NME... 2023/05/22 16:36:05 - mmengine - INFO - Epoch(val) [38][79/79] NME: 0.041788 data_time: 0.016899 time: 0.527745 2023/05/22 16:36:05 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_36.pth is removed 2023/05/22 16:36:05 - mmengine - INFO - The best checkpoint with 0.0418 NME at 38 epoch is saved to best_NME_epoch_38.pth. 2023/05/22 16:37:01 - mmengine - INFO - Epoch(train) [39][ 50/118] lr: 2.000000e-03 eta: 1:52:44 time: 1.117088 data_time: 0.316500 memory: 10706 loss: 0.029812 loss/heatmap: 0.011504 loss/offside: 0.018308 2023/05/22 16:37:55 - mmengine - INFO - Epoch(train) [39][100/118] lr: 2.000000e-03 eta: 1:51:20 time: 1.088348 data_time: 0.290462 memory: 10706 loss: 0.030297 loss/heatmap: 0.011683 loss/offside: 0.018614 2023/05/22 16:38:18 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:38:18 - mmengine - INFO - Saving checkpoint at 39 epochs 2023/05/22 16:38:46 - mmengine - INFO - Epoch(val) [39][50/79] eta: 0:00:15 time: 0.541577 data_time: 0.019590 memory: 10706 2023/05/22 16:39:01 - mmengine - INFO - Evaluating NME... 2023/05/22 16:39:01 - mmengine - INFO - Epoch(val) [39][79/79] NME: 0.042422 data_time: 0.016154 time: 0.528066 2023/05/22 16:40:10 - mmengine - INFO - Epoch(train) [40][ 50/118] lr: 2.000000e-03 eta: 1:49:43 time: 1.366512 data_time: 0.565296 memory: 10706 loss: 0.029864 loss/heatmap: 0.011531 loss/offside: 0.018333 2023/05/22 16:41:01 - mmengine - INFO - Epoch(train) [40][100/118] lr: 2.000000e-03 eta: 1:48:18 time: 1.038322 data_time: 0.239212 memory: 10706 loss: 0.029712 loss/heatmap: 0.011453 loss/offside: 0.018259 2023/05/22 16:41:20 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:41:20 - mmengine - INFO - Saving checkpoint at 40 epochs 2023/05/22 16:41:49 - mmengine - INFO - Epoch(val) [40][50/79] eta: 0:00:16 time: 0.561349 data_time: 0.019828 memory: 10706 2023/05/22 16:42:04 - mmengine - INFO - Evaluating NME... 2023/05/22 16:42:04 - mmengine - INFO - Epoch(val) [40][79/79] NME: 0.042179 data_time: 0.016150 time: 0.538378 2023/05/22 16:43:12 - mmengine - INFO - Epoch(train) [41][ 50/118] lr: 2.000000e-04 eta: 1:46:38 time: 1.369828 data_time: 0.574672 memory: 10706 loss: 0.028620 loss/heatmap: 0.011017 loss/offside: 0.017603 2023/05/22 16:44:03 - mmengine - INFO - Epoch(train) [41][100/118] lr: 2.000000e-04 eta: 1:45:13 time: 1.021091 data_time: 0.226258 memory: 10706 loss: 0.027791 loss/heatmap: 0.010679 loss/offside: 0.017112 2023/05/22 16:44:25 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:44:25 - mmengine - INFO - Saving checkpoint at 41 epochs 2023/05/22 16:44:53 - mmengine - INFO - Epoch(val) [41][50/79] eta: 0:00:15 time: 0.539442 data_time: 0.019340 memory: 10706 2023/05/22 16:45:08 - mmengine - INFO - Evaluating NME... 2023/05/22 16:45:08 - mmengine - INFO - Epoch(val) [41][79/79] NME: 0.040008 data_time: 0.015863 time: 0.525659 2023/05/22 16:45:08 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_38.pth is removed 2023/05/22 16:45:09 - mmengine - INFO - The best checkpoint with 0.0400 NME at 41 epoch is saved to best_NME_epoch_41.pth. 2023/05/22 16:46:22 - mmengine - INFO - Epoch(train) [42][ 50/118] lr: 2.000000e-04 eta: 1:43:43 time: 1.477128 data_time: 0.671872 memory: 10706 loss: 0.027453 loss/heatmap: 0.010533 loss/offside: 0.016920 2023/05/22 16:47:16 - mmengine - INFO - Epoch(train) [42][100/118] lr: 2.000000e-04 eta: 1:42:21 time: 1.070065 data_time: 0.272844 memory: 10706 loss: 0.027109 loss/heatmap: 0.010388 loss/offside: 0.016720 2023/05/22 16:47:37 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:47:37 - mmengine - INFO - Saving checkpoint at 42 epochs 2023/05/22 16:48:05 - mmengine - INFO - Epoch(val) [42][50/79] eta: 0:00:15 time: 0.540050 data_time: 0.019379 memory: 10706 2023/05/22 16:48:20 - mmengine - INFO - Evaluating NME... 2023/05/22 16:48:20 - mmengine - INFO - Epoch(val) [42][79/79] NME: 0.039841 data_time: 0.016076 time: 0.525522 2023/05/22 16:48:20 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_41.pth is removed 2023/05/22 16:48:21 - mmengine - INFO - The best checkpoint with 0.0398 NME at 42 epoch is saved to best_NME_epoch_42.pth. 2023/05/22 16:49:21 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:49:27 - mmengine - INFO - Epoch(train) [43][ 50/118] lr: 2.000000e-04 eta: 1:40:43 time: 1.318262 data_time: 0.523922 memory: 10706 loss: 0.026799 loss/heatmap: 0.010257 loss/offside: 0.016542 2023/05/22 16:50:30 - mmengine - INFO - Epoch(train) [43][100/118] lr: 2.000000e-04 eta: 1:39:31 time: 1.276085 data_time: 0.477382 memory: 10706 loss: 0.026948 loss/heatmap: 0.010312 loss/offside: 0.016636 2023/05/22 16:50:52 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:50:52 - mmengine - INFO - Saving checkpoint at 43 epochs 2023/05/22 16:51:20 - mmengine - INFO - Epoch(val) [43][50/79] eta: 0:00:15 time: 0.545492 data_time: 0.022904 memory: 10706 2023/05/22 16:51:35 - mmengine - INFO - Evaluating NME... 2023/05/22 16:51:35 - mmengine - INFO - Epoch(val) [43][79/79] NME: 0.039633 data_time: 0.018121 time: 0.529380 2023/05/22 16:51:35 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_42.pth is removed 2023/05/22 16:51:35 - mmengine - INFO - The best checkpoint with 0.0396 NME at 43 epoch is saved to best_NME_epoch_43.pth. 2023/05/22 16:52:44 - mmengine - INFO - Epoch(train) [44][ 50/118] lr: 2.000000e-04 eta: 1:37:57 time: 1.386239 data_time: 0.584982 memory: 10706 loss: 0.027173 loss/heatmap: 0.010433 loss/offside: 0.016740 2023/05/22 16:53:39 - mmengine - INFO - Epoch(train) [44][100/118] lr: 2.000000e-04 eta: 1:36:38 time: 1.087468 data_time: 0.293968 memory: 10706 loss: 0.027086 loss/heatmap: 0.010373 loss/offside: 0.016713 2023/05/22 16:53:56 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:53:56 - mmengine - INFO - Saving checkpoint at 44 epochs 2023/05/22 16:54:23 - mmengine - INFO - Epoch(val) [44][50/79] eta: 0:00:15 time: 0.522789 data_time: 0.019717 memory: 10706 2023/05/22 16:54:38 - mmengine - INFO - Evaluating NME... 2023/05/22 16:54:38 - mmengine - INFO - Epoch(val) [44][79/79] NME: 0.039608 data_time: 0.016276 time: 0.515793 2023/05/22 16:54:38 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_43.pth is removed 2023/05/22 16:54:39 - mmengine - INFO - The best checkpoint with 0.0396 NME at 44 epoch is saved to best_NME_epoch_44.pth. 2023/05/22 16:55:57 - mmengine - INFO - Epoch(train) [45][ 50/118] lr: 2.000000e-04 eta: 1:35:08 time: 1.567912 data_time: 0.768490 memory: 10706 loss: 0.026975 loss/heatmap: 0.010336 loss/offside: 0.016640 2023/05/22 16:56:52 - mmengine - INFO - Epoch(train) [45][100/118] lr: 2.000000e-04 eta: 1:33:49 time: 1.083595 data_time: 0.290821 memory: 10706 loss: 0.026572 loss/heatmap: 0.010150 loss/offside: 0.016423 2023/05/22 16:57:08 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 16:57:08 - mmengine - INFO - Saving checkpoint at 45 epochs 2023/05/22 16:57:36 - mmengine - INFO - Epoch(val) [45][50/79] eta: 0:00:15 time: 0.543542 data_time: 0.019481 memory: 10706 2023/05/22 16:57:50 - mmengine - INFO - Evaluating NME... 2023/05/22 16:57:50 - mmengine - INFO - Epoch(val) [45][79/79] NME: 0.039626 data_time: 0.015942 time: 0.519232 2023/05/22 16:59:00 - mmengine - INFO - Epoch(train) [46][ 50/118] lr: 2.000000e-04 eta: 1:32:12 time: 1.402956 data_time: 0.609482 memory: 10706 loss: 0.026727 loss/heatmap: 0.010219 loss/offside: 0.016509 2023/05/22 16:59:51 - mmengine - INFO - Epoch(train) [46][100/118] lr: 2.000000e-04 eta: 1:30:52 time: 1.019401 data_time: 0.222661 memory: 10706 loss: 0.026801 loss/heatmap: 0.010259 loss/offside: 0.016542 2023/05/22 17:00:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:00:12 - mmengine - INFO - Saving checkpoint at 46 epochs 2023/05/22 17:00:39 - mmengine - INFO - Epoch(val) [46][50/79] eta: 0:00:15 time: 0.519342 data_time: 0.019121 memory: 10706 2023/05/22 17:00:53 - mmengine - INFO - Evaluating NME... 2023/05/22 17:00:53 - mmengine - INFO - Epoch(val) [46][79/79] NME: 0.039425 data_time: 0.015723 time: 0.504035 2023/05/22 17:00:53 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_44.pth is removed 2023/05/22 17:00:54 - mmengine - INFO - The best checkpoint with 0.0394 NME at 46 epoch is saved to best_NME_epoch_46.pth. 2023/05/22 17:01:55 - mmengine - INFO - Epoch(train) [47][ 50/118] lr: 2.000000e-04 eta: 1:29:13 time: 1.226374 data_time: 0.429698 memory: 10706 loss: 0.026393 loss/heatmap: 0.010084 loss/offside: 0.016308 2023/05/22 17:02:48 - mmengine - INFO - Epoch(train) [47][100/118] lr: 2.000000e-04 eta: 1:27:55 time: 1.052413 data_time: 0.246208 memory: 10706 loss: 0.026562 loss/heatmap: 0.010142 loss/offside: 0.016420 2023/05/22 17:03:04 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:03:04 - mmengine - INFO - Saving checkpoint at 47 epochs 2023/05/22 17:03:31 - mmengine - INFO - Epoch(val) [47][50/79] eta: 0:00:15 time: 0.521965 data_time: 0.019571 memory: 10706 2023/05/22 17:03:45 - mmengine - INFO - Evaluating NME... 2023/05/22 17:03:45 - mmengine - INFO - Epoch(val) [47][79/79] NME: 0.039385 data_time: 0.016128 time: 0.507278 2023/05/22 17:03:45 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_46.pth is removed 2023/05/22 17:03:46 - mmengine - INFO - The best checkpoint with 0.0394 NME at 47 epoch is saved to best_NME_epoch_47.pth. 2023/05/22 17:05:14 - mmengine - INFO - Epoch(train) [48][ 50/118] lr: 2.000000e-04 eta: 1:26:32 time: 1.769694 data_time: 0.971246 memory: 10706 loss: 0.026495 loss/heatmap: 0.010129 loss/offside: 0.016366 2023/05/22 17:06:20 - mmengine - INFO - Epoch(train) [48][100/118] lr: 2.000000e-04 eta: 1:25:23 time: 1.315745 data_time: 0.513171 memory: 10706 loss: 0.026448 loss/heatmap: 0.010096 loss/offside: 0.016353 2023/05/22 17:06:36 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:06:36 - mmengine - INFO - Saving checkpoint at 48 epochs 2023/05/22 17:07:03 - mmengine - INFO - Epoch(val) [48][50/79] eta: 0:00:15 time: 0.535090 data_time: 0.019607 memory: 10706 2023/05/22 17:07:19 - mmengine - INFO - Evaluating NME... 2023/05/22 17:07:19 - mmengine - INFO - Epoch(val) [48][79/79] NME: 0.039312 data_time: 0.016238 time: 0.529242 2023/05/22 17:07:19 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_47.pth is removed 2023/05/22 17:07:20 - mmengine - INFO - The best checkpoint with 0.0393 NME at 48 epoch is saved to best_NME_epoch_48.pth. 2023/05/22 17:08:24 - mmengine - INFO - Epoch(train) [49][ 50/118] lr: 2.000000e-04 eta: 1:23:44 time: 1.279977 data_time: 0.486850 memory: 10706 loss: 0.026851 loss/heatmap: 0.010265 loss/offside: 0.016585 2023/05/22 17:09:20 - mmengine - INFO - Epoch(train) [49][100/118] lr: 2.000000e-04 eta: 1:22:29 time: 1.133168 data_time: 0.337080 memory: 10706 loss: 0.026244 loss/heatmap: 0.010008 loss/offside: 0.016236 2023/05/22 17:09:39 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:09:39 - mmengine - INFO - Saving checkpoint at 49 epochs 2023/05/22 17:10:07 - mmengine - INFO - Epoch(val) [49][50/79] eta: 0:00:15 time: 0.550970 data_time: 0.020434 memory: 10706 2023/05/22 17:10:23 - mmengine - INFO - Evaluating NME... 2023/05/22 17:10:23 - mmengine - INFO - Epoch(val) [49][79/79] NME: 0.039286 data_time: 0.016875 time: 0.536520 2023/05/22 17:10:23 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_48.pth is removed 2023/05/22 17:10:23 - mmengine - INFO - The best checkpoint with 0.0393 NME at 49 epoch is saved to best_NME_epoch_49.pth. 2023/05/22 17:11:33 - mmengine - INFO - Epoch(train) [50][ 50/118] lr: 2.000000e-04 eta: 1:20:56 time: 1.405125 data_time: 0.609477 memory: 10706 loss: 0.026137 loss/heatmap: 0.009955 loss/offside: 0.016182 2023/05/22 17:12:24 - mmengine - INFO - Epoch(train) [50][100/118] lr: 2.000000e-04 eta: 1:19:39 time: 1.013956 data_time: 0.213620 memory: 10706 loss: 0.026252 loss/heatmap: 0.010017 loss/offside: 0.016235 2023/05/22 17:12:41 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:12:41 - mmengine - INFO - Saving checkpoint at 50 epochs 2023/05/22 17:13:08 - mmengine - INFO - Epoch(val) [50][50/79] eta: 0:00:15 time: 0.517777 data_time: 0.019446 memory: 10706 2023/05/22 17:13:22 - mmengine - INFO - Evaluating NME... 2023/05/22 17:13:22 - mmengine - INFO - Epoch(val) [50][79/79] NME: 0.039166 data_time: 0.016019 time: 0.503838 2023/05/22 17:13:22 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_49.pth is removed 2023/05/22 17:13:23 - mmengine - INFO - The best checkpoint with 0.0392 NME at 50 epoch is saved to best_NME_epoch_50.pth. 2023/05/22 17:14:44 - mmengine - INFO - Epoch(train) [51][ 50/118] lr: 2.000000e-04 eta: 1:18:12 time: 1.624628 data_time: 0.831506 memory: 10706 loss: 0.026206 loss/heatmap: 0.009990 loss/offside: 0.016216 2023/05/22 17:15:50 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:15:50 - mmengine - INFO - Epoch(train) [51][100/118] lr: 2.000000e-04 eta: 1:17:04 time: 1.312432 data_time: 0.508891 memory: 10706 loss: 0.026160 loss/heatmap: 0.009962 loss/offside: 0.016198 2023/05/22 17:16:17 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:16:17 - mmengine - INFO - Saving checkpoint at 51 epochs 2023/05/22 17:16:43 - mmengine - INFO - Epoch(val) [51][50/79] eta: 0:00:15 time: 0.520035 data_time: 0.019736 memory: 10706 2023/05/22 17:16:58 - mmengine - INFO - Evaluating NME... 2023/05/22 17:16:58 - mmengine - INFO - Epoch(val) [51][79/79] NME: 0.039081 data_time: 0.016310 time: 0.506221 2023/05/22 17:16:58 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_50.pth is removed 2023/05/22 17:16:58 - mmengine - INFO - The best checkpoint with 0.0391 NME at 51 epoch is saved to best_NME_epoch_51.pth. 2023/05/22 17:18:05 - mmengine - INFO - Epoch(train) [52][ 50/118] lr: 2.000000e-04 eta: 1:15:33 time: 1.331189 data_time: 0.531251 memory: 10706 loss: 0.025843 loss/heatmap: 0.009843 loss/offside: 0.016000 2023/05/22 17:19:18 - mmengine - INFO - Epoch(train) [52][100/118] lr: 2.000000e-04 eta: 1:14:29 time: 1.462753 data_time: 0.660349 memory: 10706 loss: 0.026575 loss/heatmap: 0.010156 loss/offside: 0.016419 2023/05/22 17:19:36 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:19:36 - mmengine - INFO - Saving checkpoint at 52 epochs 2023/05/22 17:20:03 - mmengine - INFO - Epoch(val) [52][50/79] eta: 0:00:15 time: 0.524003 data_time: 0.020130 memory: 10706 2023/05/22 17:20:18 - mmengine - INFO - Evaluating NME... 2023/05/22 17:20:18 - mmengine - INFO - Epoch(val) [52][79/79] NME: 0.039082 data_time: 0.016475 time: 0.508772 2023/05/22 17:21:38 - mmengine - INFO - Epoch(train) [53][ 50/118] lr: 2.000000e-04 eta: 1:13:01 time: 1.598968 data_time: 0.801053 memory: 10706 loss: 0.026462 loss/heatmap: 0.010086 loss/offside: 0.016376 2023/05/22 17:22:35 - mmengine - INFO - Epoch(train) [53][100/118] lr: 2.000000e-04 eta: 1:11:49 time: 1.145064 data_time: 0.341613 memory: 10706 loss: 0.026108 loss/heatmap: 0.009943 loss/offside: 0.016165 2023/05/22 17:22:52 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:22:52 - mmengine - INFO - Saving checkpoint at 53 epochs 2023/05/22 17:23:19 - mmengine - INFO - Epoch(val) [53][50/79] eta: 0:00:15 time: 0.521173 data_time: 0.019316 memory: 10706 2023/05/22 17:23:34 - mmengine - INFO - Evaluating NME... 2023/05/22 17:23:34 - mmengine - INFO - Epoch(val) [53][79/79] NME: 0.039138 data_time: 0.016018 time: 0.509370 2023/05/22 17:24:50 - mmengine - INFO - Epoch(train) [54][ 50/118] lr: 2.000000e-04 eta: 1:10:19 time: 1.525871 data_time: 0.724004 memory: 10706 loss: 0.025825 loss/heatmap: 0.009827 loss/offside: 0.015998 2023/05/22 17:25:53 - mmengine - INFO - Epoch(train) [54][100/118] lr: 2.000000e-04 eta: 1:09:09 time: 1.262608 data_time: 0.463490 memory: 10706 loss: 0.026092 loss/heatmap: 0.009926 loss/offside: 0.016165 2023/05/22 17:26:14 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:26:14 - mmengine - INFO - Saving checkpoint at 54 epochs 2023/05/22 17:26:42 - mmengine - INFO - Epoch(val) [54][50/79] eta: 0:00:15 time: 0.532588 data_time: 0.019340 memory: 10706 2023/05/22 17:26:57 - mmengine - INFO - Evaluating NME... 2023/05/22 17:26:57 - mmengine - INFO - Epoch(val) [54][79/79] NME: 0.039112 data_time: 0.015972 time: 0.523926 2023/05/22 17:28:04 - mmengine - INFO - Epoch(train) [55][ 50/118] lr: 2.000000e-04 eta: 1:07:36 time: 1.339088 data_time: 0.541044 memory: 10706 loss: 0.025620 loss/heatmap: 0.009737 loss/offside: 0.015884 2023/05/22 17:29:06 - mmengine - INFO - Epoch(train) [55][100/118] lr: 2.000000e-04 eta: 1:06:27 time: 1.249669 data_time: 0.446219 memory: 10706 loss: 0.026029 loss/heatmap: 0.009908 loss/offside: 0.016122 2023/05/22 17:29:23 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:29:23 - mmengine - INFO - Saving checkpoint at 55 epochs 2023/05/22 17:29:50 - mmengine - INFO - Epoch(val) [55][50/79] eta: 0:00:15 time: 0.528948 data_time: 0.020535 memory: 10706 2023/05/22 17:30:05 - mmengine - INFO - Evaluating NME... 2023/05/22 17:30:05 - mmengine - INFO - Epoch(val) [55][79/79] NME: 0.039153 data_time: 0.016780 time: 0.514331 2023/05/22 17:31:27 - mmengine - INFO - Epoch(train) [56][ 50/118] lr: 2.000000e-04 eta: 1:04:59 time: 1.640794 data_time: 0.834207 memory: 10706 loss: 0.026115 loss/heatmap: 0.009941 loss/offside: 0.016174 2023/05/22 17:32:31 - mmengine - INFO - Epoch(train) [56][100/118] lr: 2.000000e-04 eta: 1:03:50 time: 1.286813 data_time: 0.485570 memory: 10706 loss: 0.026047 loss/heatmap: 0.009911 loss/offside: 0.016137 2023/05/22 17:32:50 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:32:50 - mmengine - INFO - Saving checkpoint at 56 epochs 2023/05/22 17:33:16 - mmengine - INFO - Epoch(val) [56][50/79] eta: 0:00:15 time: 0.523307 data_time: 0.023125 memory: 10706 2023/05/22 17:33:31 - mmengine - INFO - Evaluating NME... 2023/05/22 17:33:31 - mmengine - INFO - Epoch(val) [56][79/79] NME: 0.039083 data_time: 0.018458 time: 0.507491 2023/05/22 17:34:34 - mmengine - INFO - Epoch(train) [57][ 50/118] lr: 2.000000e-04 eta: 1:02:15 time: 1.271397 data_time: 0.470694 memory: 10706 loss: 0.026059 loss/heatmap: 0.009919 loss/offside: 0.016141 2023/05/22 17:35:24 - mmengine - INFO - Epoch(train) [57][100/118] lr: 2.000000e-04 eta: 1:01:00 time: 0.986202 data_time: 0.184968 memory: 10706 loss: 0.025593 loss/heatmap: 0.009729 loss/offside: 0.015864 2023/05/22 17:35:44 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:35:44 - mmengine - INFO - Saving checkpoint at 57 epochs 2023/05/22 17:36:11 - mmengine - INFO - Epoch(val) [57][50/79] eta: 0:00:15 time: 0.522716 data_time: 0.019864 memory: 10706 2023/05/22 17:36:25 - mmengine - INFO - Evaluating NME... 2023/05/22 17:36:25 - mmengine - INFO - Epoch(val) [57][79/79] NME: 0.039021 data_time: 0.016462 time: 0.508857 2023/05/22 17:36:25 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_51.pth is removed 2023/05/22 17:36:26 - mmengine - INFO - The best checkpoint with 0.0390 NME at 57 epoch is saved to best_NME_epoch_57.pth. 2023/05/22 17:37:44 - mmengine - INFO - Epoch(train) [58][ 50/118] lr: 2.000000e-04 eta: 0:59:32 time: 1.569241 data_time: 0.772215 memory: 10706 loss: 0.025759 loss/heatmap: 0.009796 loss/offside: 0.015963 2023/05/22 17:38:42 - mmengine - INFO - Epoch(train) [58][100/118] lr: 2.000000e-04 eta: 0:58:22 time: 1.152009 data_time: 0.355790 memory: 10706 loss: 0.025717 loss/heatmap: 0.009774 loss/offside: 0.015943 2023/05/22 17:38:58 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:38:58 - mmengine - INFO - Saving checkpoint at 58 epochs 2023/05/22 17:39:25 - mmengine - INFO - Epoch(val) [58][50/79] eta: 0:00:15 time: 0.519280 data_time: 0.019414 memory: 10706 2023/05/22 17:39:38 - mmengine - INFO - Evaluating NME... 2023/05/22 17:39:39 - mmengine - INFO - Epoch(val) [58][79/79] NME: 0.039154 data_time: 0.015890 time: 0.497731 2023/05/22 17:40:51 - mmengine - INFO - Epoch(train) [59][ 50/118] lr: 2.000000e-04 eta: 0:56:49 time: 1.443804 data_time: 0.641077 memory: 10706 loss: 0.025908 loss/heatmap: 0.009849 loss/offside: 0.016059 2023/05/22 17:41:55 - mmengine - INFO - Epoch(train) [59][100/118] lr: 2.000000e-04 eta: 0:55:41 time: 1.278843 data_time: 0.480073 memory: 10706 loss: 0.025746 loss/heatmap: 0.009785 loss/offside: 0.015960 2023/05/22 17:42:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:42:12 - mmengine - INFO - Saving checkpoint at 59 epochs 2023/05/22 17:42:39 - mmengine - INFO - Epoch(val) [59][50/79] eta: 0:00:15 time: 0.523195 data_time: 0.019024 memory: 10706 2023/05/22 17:42:54 - mmengine - INFO - Evaluating NME... 2023/05/22 17:42:54 - mmengine - INFO - Epoch(val) [59][79/79] NME: 0.039158 data_time: 0.015672 time: 0.515600 2023/05/22 17:43:43 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:43:57 - mmengine - INFO - Epoch(train) [60][ 50/118] lr: 2.000000e-04 eta: 0:54:07 time: 1.252493 data_time: 0.450395 memory: 10706 loss: 0.025484 loss/heatmap: 0.009669 loss/offside: 0.015814 2023/05/22 17:44:54 - mmengine - INFO - Epoch(train) [60][100/118] lr: 2.000000e-04 eta: 0:52:56 time: 1.136540 data_time: 0.328863 memory: 10706 loss: 0.025865 loss/heatmap: 0.009827 loss/offside: 0.016037 2023/05/22 17:45:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:45:12 - mmengine - INFO - Saving checkpoint at 60 epochs 2023/05/22 17:45:40 - mmengine - INFO - Epoch(val) [60][50/79] eta: 0:00:15 time: 0.541631 data_time: 0.023102 memory: 10706 2023/05/22 17:45:54 - mmengine - INFO - Evaluating NME... 2023/05/22 17:45:54 - mmengine - INFO - Epoch(val) [60][79/79] NME: 0.039051 data_time: 0.018376 time: 0.516005 2023/05/22 17:47:01 - mmengine - INFO - Epoch(train) [61][ 50/118] lr: 2.000000e-05 eta: 0:51:24 time: 1.340799 data_time: 0.540981 memory: 10706 loss: 0.025795 loss/heatmap: 0.009803 loss/offside: 0.015992 2023/05/22 17:47:54 - mmengine - INFO - Epoch(train) [61][100/118] lr: 2.000000e-05 eta: 0:50:13 time: 1.058898 data_time: 0.262470 memory: 10706 loss: 0.025127 loss/heatmap: 0.009542 loss/offside: 0.015585 2023/05/22 17:48:12 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:48:12 - mmengine - INFO - Saving checkpoint at 61 epochs 2023/05/22 17:48:41 - mmengine - INFO - Epoch(val) [61][50/79] eta: 0:00:16 time: 0.565313 data_time: 0.020707 memory: 10706 2023/05/22 17:48:57 - mmengine - INFO - Evaluating NME... 2023/05/22 17:48:57 - mmengine - INFO - Epoch(val) [61][79/79] NME: 0.038957 data_time: 0.016991 time: 0.554874 2023/05/22 17:48:57 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_57.pth is removed 2023/05/22 17:48:57 - mmengine - INFO - The best checkpoint with 0.0390 NME at 61 epoch is saved to best_NME_epoch_61.pth. 2023/05/22 17:50:07 - mmengine - INFO - Epoch(train) [62][ 50/118] lr: 2.000000e-05 eta: 0:48:41 time: 1.399077 data_time: 0.595686 memory: 10706 loss: 0.025504 loss/heatmap: 0.009704 loss/offside: 0.015800 2023/05/22 17:51:02 - mmengine - INFO - Epoch(train) [62][100/118] lr: 2.000000e-05 eta: 0:47:31 time: 1.088854 data_time: 0.283956 memory: 10706 loss: 0.025206 loss/heatmap: 0.009582 loss/offside: 0.015624 2023/05/22 17:51:19 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:51:19 - mmengine - INFO - Saving checkpoint at 62 epochs 2023/05/22 17:51:47 - mmengine - INFO - Epoch(val) [62][50/79] eta: 0:00:15 time: 0.541741 data_time: 0.019709 memory: 10706 2023/05/22 17:52:02 - mmengine - INFO - Evaluating NME... 2023/05/22 17:52:02 - mmengine - INFO - Epoch(val) [62][79/79] NME: 0.038963 data_time: 0.016342 time: 0.537333 2023/05/22 17:53:09 - mmengine - INFO - Epoch(train) [63][ 50/118] lr: 2.000000e-05 eta: 0:45:58 time: 1.334961 data_time: 0.537262 memory: 10706 loss: 0.025148 loss/heatmap: 0.009557 loss/offside: 0.015591 2023/05/22 17:54:06 - mmengine - INFO - Epoch(train) [63][100/118] lr: 2.000000e-05 eta: 0:44:49 time: 1.136397 data_time: 0.338046 memory: 10706 loss: 0.025107 loss/heatmap: 0.009532 loss/offside: 0.015575 2023/05/22 17:54:23 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:54:23 - mmengine - INFO - Saving checkpoint at 63 epochs 2023/05/22 17:54:51 - mmengine - INFO - Epoch(val) [63][50/79] eta: 0:00:15 time: 0.534959 data_time: 0.022365 memory: 10706 2023/05/22 17:55:06 - mmengine - INFO - Evaluating NME... 2023/05/22 17:55:06 - mmengine - INFO - Epoch(val) [63][79/79] NME: 0.038973 data_time: 0.018024 time: 0.527243 2023/05/22 17:56:07 - mmengine - INFO - Epoch(train) [64][ 50/118] lr: 2.000000e-05 eta: 0:43:16 time: 1.230725 data_time: 0.436444 memory: 10706 loss: 0.025411 loss/heatmap: 0.009659 loss/offside: 0.015752 2023/05/22 17:57:08 - mmengine - INFO - Epoch(train) [64][100/118] lr: 2.000000e-05 eta: 0:42:08 time: 1.216710 data_time: 0.414711 memory: 10706 loss: 0.025531 loss/heatmap: 0.009704 loss/offside: 0.015827 2023/05/22 17:57:27 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 17:57:27 - mmengine - INFO - Saving checkpoint at 64 epochs 2023/05/22 17:57:55 - mmengine - INFO - Epoch(val) [64][50/79] eta: 0:00:16 time: 0.560739 data_time: 0.019778 memory: 10706 2023/05/22 17:58:10 - mmengine - INFO - Evaluating NME... 2023/05/22 17:58:10 - mmengine - INFO - Epoch(val) [64][79/79] NME: 0.038924 data_time: 0.016250 time: 0.537793 2023/05/22 17:58:10 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_61.pth is removed 2023/05/22 17:58:11 - mmengine - INFO - The best checkpoint with 0.0389 NME at 64 epoch is saved to best_NME_epoch_64.pth. 2023/05/22 17:59:12 - mmengine - INFO - Epoch(train) [65][ 50/118] lr: 2.000000e-05 eta: 0:40:35 time: 1.224686 data_time: 0.424921 memory: 10706 loss: 0.025405 loss/heatmap: 0.009652 loss/offside: 0.015753 2023/05/22 18:00:08 - mmengine - INFO - Epoch(train) [65][100/118] lr: 2.000000e-05 eta: 0:39:27 time: 1.127815 data_time: 0.326229 memory: 10706 loss: 0.025194 loss/heatmap: 0.009576 loss/offside: 0.015618 2023/05/22 18:00:26 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:00:26 - mmengine - INFO - Saving checkpoint at 65 epochs 2023/05/22 18:00:54 - mmengine - INFO - Epoch(val) [65][50/79] eta: 0:00:15 time: 0.542640 data_time: 0.019791 memory: 10706 2023/05/22 18:01:09 - mmengine - INFO - Evaluating NME... 2023/05/22 18:01:09 - mmengine - INFO - Epoch(val) [65][79/79] NME: 0.038915 data_time: 0.016299 time: 0.527677 2023/05/22 18:01:09 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_64.pth is removed 2023/05/22 18:01:10 - mmengine - INFO - The best checkpoint with 0.0389 NME at 65 epoch is saved to best_NME_epoch_65.pth. 2023/05/22 18:02:07 - mmengine - INFO - Epoch(train) [66][ 50/118] lr: 2.000000e-05 eta: 0:37:53 time: 1.139467 data_time: 0.339238 memory: 10706 loss: 0.025716 loss/heatmap: 0.009790 loss/offside: 0.015927 2023/05/22 18:03:11 - mmengine - INFO - Epoch(train) [66][100/118] lr: 2.000000e-05 eta: 0:36:47 time: 1.284995 data_time: 0.479262 memory: 10706 loss: 0.025268 loss/heatmap: 0.009598 loss/offside: 0.015670 2023/05/22 18:03:34 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:03:34 - mmengine - INFO - Saving checkpoint at 66 epochs 2023/05/22 18:04:02 - mmengine - INFO - Epoch(val) [66][50/79] eta: 0:00:15 time: 0.541344 data_time: 0.019261 memory: 10706 2023/05/22 18:04:17 - mmengine - INFO - Evaluating NME... 2023/05/22 18:04:17 - mmengine - INFO - Epoch(val) [66][79/79] NME: 0.038864 data_time: 0.015913 time: 0.524771 2023/05/22 18:04:17 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_65.pth is removed 2023/05/22 18:04:17 - mmengine - INFO - The best checkpoint with 0.0389 NME at 66 epoch is saved to best_NME_epoch_66.pth. 2023/05/22 18:05:13 - mmengine - INFO - Epoch(train) [67][ 50/118] lr: 2.000000e-05 eta: 0:35:15 time: 1.118070 data_time: 0.323095 memory: 10706 loss: 0.025290 loss/heatmap: 0.009605 loss/offside: 0.015685 2023/05/22 18:06:12 - mmengine - INFO - Epoch(train) [67][100/118] lr: 2.000000e-05 eta: 0:34:07 time: 1.174675 data_time: 0.373265 memory: 10706 loss: 0.025176 loss/heatmap: 0.009563 loss/offside: 0.015613 2023/05/22 18:06:31 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:06:31 - mmengine - INFO - Saving checkpoint at 67 epochs 2023/05/22 18:06:59 - mmengine - INFO - Epoch(val) [67][50/79] eta: 0:00:15 time: 0.539718 data_time: 0.019278 memory: 10706 2023/05/22 18:07:14 - mmengine - INFO - Evaluating NME... 2023/05/22 18:07:14 - mmengine - INFO - Epoch(val) [67][79/79] NME: 0.038902 data_time: 0.015884 time: 0.526005 2023/05/22 18:08:08 - mmengine - INFO - Epoch(train) [68][ 50/118] lr: 2.000000e-05 eta: 0:32:34 time: 1.088681 data_time: 0.291445 memory: 10706 loss: 0.025349 loss/heatmap: 0.009635 loss/offside: 0.015714 2023/05/22 18:08:51 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:08:57 - mmengine - INFO - Epoch(train) [68][100/118] lr: 2.000000e-05 eta: 0:31:26 time: 0.977692 data_time: 0.177486 memory: 10706 loss: 0.025473 loss/heatmap: 0.009691 loss/offside: 0.015781 2023/05/22 18:09:13 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:09:13 - mmengine - INFO - Saving checkpoint at 68 epochs 2023/05/22 18:09:41 - mmengine - INFO - Epoch(val) [68][50/79] eta: 0:00:15 time: 0.546341 data_time: 0.020492 memory: 10706 2023/05/22 18:09:57 - mmengine - INFO - Evaluating NME... 2023/05/22 18:09:57 - mmengine - INFO - Epoch(val) [68][79/79] NME: 0.038875 data_time: 0.016722 time: 0.542824 2023/05/22 18:10:49 - mmengine - INFO - Epoch(train) [69][ 50/118] lr: 2.000000e-05 eta: 0:29:53 time: 1.043717 data_time: 0.249177 memory: 10706 loss: 0.025095 loss/heatmap: 0.009531 loss/offside: 0.015565 2023/05/22 18:11:36 - mmengine - INFO - Epoch(train) [69][100/118] lr: 2.000000e-05 eta: 0:28:44 time: 0.930328 data_time: 0.135599 memory: 10706 loss: 0.025165 loss/heatmap: 0.009549 loss/offside: 0.015616 2023/05/22 18:11:52 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:11:52 - mmengine - INFO - Saving checkpoint at 69 epochs 2023/05/22 18:12:21 - mmengine - INFO - Epoch(val) [69][50/79] eta: 0:00:15 time: 0.549123 data_time: 0.020388 memory: 10706 2023/05/22 18:12:36 - mmengine - INFO - Evaluating NME... 2023/05/22 18:12:36 - mmengine - INFO - Epoch(val) [69][79/79] NME: 0.038903 data_time: 0.016748 time: 0.533834 2023/05/22 18:13:30 - mmengine - INFO - Epoch(train) [70][ 50/118] lr: 2.000000e-05 eta: 0:27:12 time: 1.082773 data_time: 0.278860 memory: 10706 loss: 0.025346 loss/heatmap: 0.009622 loss/offside: 0.015724 2023/05/22 18:14:17 - mmengine - INFO - Epoch(train) [70][100/118] lr: 2.000000e-05 eta: 0:26:04 time: 0.937690 data_time: 0.139280 memory: 10706 loss: 0.025070 loss/heatmap: 0.009527 loss/offside: 0.015543 2023/05/22 18:14:33 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:14:33 - mmengine - INFO - Saving checkpoint at 70 epochs 2023/05/22 18:15:04 - mmengine - INFO - Epoch(val) [70][50/79] eta: 0:00:17 time: 0.605439 data_time: 0.021051 memory: 10706 2023/05/22 18:15:21 - mmengine - INFO - Evaluating NME... 2023/05/22 18:15:21 - mmengine - INFO - Epoch(val) [70][79/79] NME: 0.038884 data_time: 0.017205 time: 0.592859 2023/05/22 18:16:13 - mmengine - INFO - Epoch(train) [71][ 50/118] lr: 2.000000e-05 eta: 0:24:32 time: 1.048295 data_time: 0.248712 memory: 10706 loss: 0.025312 loss/heatmap: 0.009621 loss/offside: 0.015691 2023/05/22 18:17:15 - mmengine - INFO - Epoch(train) [71][100/118] lr: 2.000000e-05 eta: 0:23:27 time: 1.223104 data_time: 0.419901 memory: 10706 loss: 0.025049 loss/heatmap: 0.009505 loss/offside: 0.015545 2023/05/22 18:17:31 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:17:31 - mmengine - INFO - Saving checkpoint at 71 epochs 2023/05/22 18:17:59 - mmengine - INFO - Epoch(val) [71][50/79] eta: 0:00:15 time: 0.544401 data_time: 0.020190 memory: 10706 2023/05/22 18:18:14 - mmengine - INFO - Evaluating NME... 2023/05/22 18:18:14 - mmengine - INFO - Epoch(val) [71][79/79] NME: 0.038912 data_time: 0.016739 time: 0.533428 2023/05/22 18:19:24 - mmengine - INFO - Epoch(train) [72][ 50/118] lr: 2.000000e-05 eta: 0:21:58 time: 1.395730 data_time: 0.597645 memory: 10706 loss: 0.025388 loss/heatmap: 0.009642 loss/offside: 0.015745 2023/05/22 18:20:16 - mmengine - INFO - Epoch(train) [72][100/118] lr: 2.000000e-05 eta: 0:20:51 time: 1.043378 data_time: 0.196699 memory: 10706 loss: 0.025621 loss/heatmap: 0.009747 loss/offside: 0.015874 2023/05/22 18:20:33 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:20:33 - mmengine - INFO - Saving checkpoint at 72 epochs 2023/05/22 18:21:01 - mmengine - INFO - Epoch(val) [72][50/79] eta: 0:00:15 time: 0.533990 data_time: 0.019994 memory: 10706 2023/05/22 18:21:16 - mmengine - INFO - Evaluating NME... 2023/05/22 18:21:16 - mmengine - INFO - Epoch(val) [72][79/79] NME: 0.038872 data_time: 0.016495 time: 0.523036 2023/05/22 18:22:16 - mmengine - INFO - Epoch(train) [73][ 50/118] lr: 2.000000e-05 eta: 0:19:21 time: 1.203217 data_time: 0.336032 memory: 10706 loss: 0.025389 loss/heatmap: 0.009653 loss/offside: 0.015736 2023/05/22 18:23:11 - mmengine - INFO - Epoch(train) [73][100/118] lr: 2.000000e-05 eta: 0:18:15 time: 1.085684 data_time: 0.282900 memory: 10706 loss: 0.025160 loss/heatmap: 0.009554 loss/offside: 0.015606 2023/05/22 18:23:28 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:23:28 - mmengine - INFO - Saving checkpoint at 73 epochs 2023/05/22 18:23:56 - mmengine - INFO - Epoch(val) [73][50/79] eta: 0:00:15 time: 0.547138 data_time: 0.019808 memory: 10706 2023/05/22 18:24:11 - mmengine - INFO - Evaluating NME... 2023/05/22 18:24:11 - mmengine - INFO - Epoch(val) [73][79/79] NME: 0.038895 data_time: 0.016387 time: 0.524484 2023/05/22 18:25:12 - mmengine - INFO - Epoch(train) [74][ 50/118] lr: 2.000000e-05 eta: 0:16:46 time: 1.228542 data_time: 0.356372 memory: 10706 loss: 0.025263 loss/heatmap: 0.009593 loss/offside: 0.015670 2023/05/22 18:26:04 - mmengine - INFO - Epoch(train) [74][100/118] lr: 2.000000e-05 eta: 0:15:40 time: 1.037228 data_time: 0.201831 memory: 10706 loss: 0.025355 loss/heatmap: 0.009640 loss/offside: 0.015715 2023/05/22 18:26:20 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:26:20 - mmengine - INFO - Saving checkpoint at 74 epochs 2023/05/22 18:26:47 - mmengine - INFO - Epoch(val) [74][50/79] eta: 0:00:15 time: 0.517705 data_time: 0.019407 memory: 10706 2023/05/22 18:27:01 - mmengine - INFO - Evaluating NME... 2023/05/22 18:27:01 - mmengine - INFO - Epoch(val) [74][79/79] NME: 0.038873 data_time: 0.015975 time: 0.504745 2023/05/22 18:28:09 - mmengine - INFO - Epoch(train) [75][ 50/118] lr: 2.000000e-05 eta: 0:14:12 time: 1.350505 data_time: 0.553269 memory: 10706 loss: 0.025575 loss/heatmap: 0.009733 loss/offside: 0.015842 2023/05/22 18:29:09 - mmengine - INFO - Epoch(train) [75][100/118] lr: 2.000000e-05 eta: 0:13:07 time: 1.210028 data_time: 0.404010 memory: 10706 loss: 0.025060 loss/heatmap: 0.009516 loss/offside: 0.015544 2023/05/22 18:29:28 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:29:28 - mmengine - INFO - Saving checkpoint at 75 epochs 2023/05/22 18:29:54 - mmengine - INFO - Epoch(val) [75][50/79] eta: 0:00:14 time: 0.515939 data_time: 0.019464 memory: 10706 2023/05/22 18:30:09 - mmengine - INFO - Evaluating NME... 2023/05/22 18:30:09 - mmengine - INFO - Epoch(val) [75][79/79] NME: 0.038843 data_time: 0.016056 time: 0.504339 2023/05/22 18:30:09 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_66.pth is removed 2023/05/22 18:30:09 - mmengine - INFO - The best checkpoint with 0.0388 NME at 75 epoch is saved to best_NME_epoch_75.pth. 2023/05/22 18:31:12 - mmengine - INFO - Epoch(train) [76][ 50/118] lr: 2.000000e-05 eta: 0:11:38 time: 1.249537 data_time: 0.455016 memory: 10706 loss: 0.025132 loss/heatmap: 0.009544 loss/offside: 0.015589 2023/05/22 18:32:07 - mmengine - INFO - Epoch(train) [76][100/118] lr: 2.000000e-05 eta: 0:10:33 time: 1.097016 data_time: 0.291350 memory: 10706 loss: 0.025159 loss/heatmap: 0.009552 loss/offside: 0.015607 2023/05/22 18:32:23 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:32:23 - mmengine - INFO - Saving checkpoint at 76 epochs 2023/05/22 18:32:50 - mmengine - INFO - Epoch(val) [76][50/79] eta: 0:00:15 time: 0.524890 data_time: 0.022663 memory: 10706 2023/05/22 18:33:05 - mmengine - INFO - Evaluating NME... 2023/05/22 18:33:05 - mmengine - INFO - Epoch(val) [76][79/79] NME: 0.038880 data_time: 0.018121 time: 0.509413 2023/05/22 18:33:47 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:34:03 - mmengine - INFO - Epoch(train) [77][ 50/118] lr: 2.000000e-05 eta: 0:09:05 time: 1.169358 data_time: 0.368165 memory: 10706 loss: 0.025524 loss/heatmap: 0.009709 loss/offside: 0.015815 2023/05/22 18:34:50 - mmengine - INFO - Epoch(train) [77][100/118] lr: 2.000000e-05 eta: 0:07:59 time: 0.934771 data_time: 0.132900 memory: 10706 loss: 0.025585 loss/heatmap: 0.009748 loss/offside: 0.015838 2023/05/22 18:35:06 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:35:06 - mmengine - INFO - Saving checkpoint at 77 epochs 2023/05/22 18:35:33 - mmengine - INFO - Epoch(val) [77][50/79] eta: 0:00:15 time: 0.526790 data_time: 0.019404 memory: 10706 2023/05/22 18:35:48 - mmengine - INFO - Evaluating NME... 2023/05/22 18:35:48 - mmengine - INFO - Epoch(val) [77][79/79] NME: 0.038865 data_time: 0.015931 time: 0.510318 2023/05/22 18:36:48 - mmengine - INFO - Epoch(train) [78][ 50/118] lr: 2.000000e-05 eta: 0:06:31 time: 1.199361 data_time: 0.402420 memory: 10706 loss: 0.025126 loss/heatmap: 0.009552 loss/offside: 0.015573 2023/05/22 18:37:38 - mmengine - INFO - Epoch(train) [78][100/118] lr: 2.000000e-05 eta: 0:05:26 time: 1.015729 data_time: 0.217038 memory: 10706 loss: 0.025085 loss/heatmap: 0.009522 loss/offside: 0.015564 2023/05/22 18:37:55 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:37:55 - mmengine - INFO - Saving checkpoint at 78 epochs 2023/05/22 18:38:24 - mmengine - INFO - Epoch(val) [78][50/79] eta: 0:00:16 time: 0.552456 data_time: 0.020382 memory: 10706 2023/05/22 18:38:39 - mmengine - INFO - Evaluating NME... 2023/05/22 18:38:39 - mmengine - INFO - Epoch(val) [78][79/79] NME: 0.038878 data_time: 0.016722 time: 0.537172 2023/05/22 18:39:35 - mmengine - INFO - Epoch(train) [79][ 50/118] lr: 2.000000e-05 eta: 0:03:59 time: 1.121260 data_time: 0.330961 memory: 10706 loss: 0.025228 loss/heatmap: 0.009580 loss/offside: 0.015648 2023/05/22 18:40:23 - mmengine - INFO - Epoch(train) [79][100/118] lr: 2.000000e-05 eta: 0:02:54 time: 0.959320 data_time: 0.165693 memory: 10706 loss: 0.024949 loss/heatmap: 0.009471 loss/offside: 0.015477 2023/05/22 18:40:42 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:40:42 - mmengine - INFO - Saving checkpoint at 79 epochs 2023/05/22 18:41:12 - mmengine - INFO - Epoch(val) [79][50/79] eta: 0:00:16 time: 0.581271 data_time: 0.025670 memory: 10706 2023/05/22 18:41:28 - mmengine - INFO - Evaluating NME... 2023/05/22 18:41:28 - mmengine - INFO - Epoch(val) [79][79/79] NME: 0.038839 data_time: 0.022640 time: 0.562505 2023/05/22 18:41:28 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_75.pth is removed 2023/05/22 18:41:28 - mmengine - INFO - The best checkpoint with 0.0388 NME at 79 epoch is saved to best_NME_epoch_79.pth. 2023/05/22 18:42:31 - mmengine - INFO - Epoch(train) [80][ 50/118] lr: 2.000000e-05 eta: 0:01:27 time: 1.248634 data_time: 0.454304 memory: 10706 loss: 0.025470 loss/heatmap: 0.009680 loss/offside: 0.015790 2023/05/22 18:43:22 - mmengine - INFO - Epoch(train) [80][100/118] lr: 2.000000e-05 eta: 0:00:23 time: 1.035301 data_time: 0.242939 memory: 10706 loss: 0.025102 loss/heatmap: 0.009543 loss/offside: 0.015559 2023/05/22 18:43:39 - mmengine - INFO - Exp name: td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256_20230522_142437 2023/05/22 18:43:39 - mmengine - INFO - Saving checkpoint at 80 epochs 2023/05/22 18:44:08 - mmengine - INFO - Epoch(val) [80][50/79] eta: 0:00:16 time: 0.559335 data_time: 0.024229 memory: 10706 2023/05/22 18:44:23 - mmengine - INFO - Evaluating NME... 2023/05/22 18:44:23 - mmengine - INFO - Epoch(val) [80][79/79] NME: 0.038822 data_time: 0.022105 time: 0.545289 2023/05/22 18:44:24 - mmengine - INFO - The previous best checkpoint /root/mmpose/projects/skps/work_dirs/td-hm_hrnetv2-w18_skps-1xb64-80e_wflw-256x256/best_NME_epoch_79.pth is removed 2023/05/22 18:44:24 - mmengine - INFO - The best checkpoint with 0.0388 NME at 80 epoch is saved to best_NME_epoch_80.pth.