2023/02/02 19:52:55 - mmengine - INFO - ------------------------------------------------------------ System environment: sys.platform: linux Python: 3.8.15 (default, Nov 24 2022, 15:19:38) [GCC 11.2.0] CUDA available: True numpy_random_seed: 1208803239 GPU 0,1,2,3,4,5,6,7: NVIDIA A100-SXM4-80GB CUDA_HOME: /nvme/softwares/cuda-11.3 NVCC: Cuda compilation tools, release 11.3, V11.3.109 GCC: gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-44) PyTorch: 1.11.0+cu113 PyTorch compiling details: PyTorch built with: - GCC 7.3 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v2.5.2 (Git Hash a9302535553c73243c632ad3c4c80beec3d19a1e) - OpenMP 201511 (a.k.a. OpenMP 4.5) - LAPACK is enabled (usually provided by MKL) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 11.3 - 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 8.2 - Magma 2.5.2 - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.3, CUDNN_VERSION=8.2.0, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -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 -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.11.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, TorchVision: 0.12.0+cu113 OpenCV: 4.6.0 MMEngine: 0.5.0 Runtime environment: cudnn_benchmark: False mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0} dist_cfg: {'backend': 'nccl'} seed: None deterministic: False Distributed launcher: pytorch Distributed training: True GPU number: 8 ------------------------------------------------------------ 2023/02/02 19:52:55 - mmengine - INFO - Config: dataset_type = 'InShop' data_preprocessor = dict( num_classes=3997, mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='RandomCrop', crop_size=448), dict(type='RandomFlip', prob=0.5, direction='horizontal'), dict(type='PackClsInputs') ] test_pipeline = [ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ] train_dataloader = dict( pin_memory=True, persistent_workers=True, collate_fn=dict(type='default_collate'), batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='train', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='RandomCrop', crop_size=448), dict(type='RandomFlip', prob=0.5, direction='horizontal'), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=True)) query_dataloader = dict( batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='query', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=False)) gallery_dataloader = dict( batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='gallery', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=False)) val_dataloader = dict( pin_memory=True, persistent_workers=True, collate_fn=dict(type='default_collate'), batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='query', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=False)) val_evaluator = dict(type='RetrievalRecall', topk=1) test_dataloader = dict( pin_memory=True, persistent_workers=True, collate_fn=dict(type='default_collate'), batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='query', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=False)) test_evaluator = dict(type='RetrievalRecall', topk=1) optim_wrapper = dict( optimizer=dict( type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0005, nesterov=True)) param_scheduler = [ dict( type='LinearLR', start_factor=0.01, by_epoch=True, begin=0, end=5, convert_to_iter_based=True), dict(type='CosineAnnealingLR', T_max=45, by_epoch=True, begin=5, end=50) ] train_cfg = dict(by_epoch=True, max_epochs=50, val_interval=1) val_cfg = dict() test_cfg = dict() auto_scale_lr = dict(base_batch_size=256, enable=True) default_scope = 'mmcls' default_hooks = dict( timer=dict(type='IterTimerHook'), logger=dict(type='LoggerHook', interval=20), param_scheduler=dict(type='ParamSchedulerHook'), checkpoint=dict( type='CheckpointHook', interval=1, save_best='auto', max_keep_ckpts=3, rule='greater'), sampler_seed=dict(type='DistSamplerSeedHook'), visualization=dict(type='VisualizationHook', enable=False)) 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')] visualizer = dict( type='ClsVisualizer', vis_backends=[dict(type='LocalVisBackend')]) log_level = 'INFO' load_from = None resume = False randomness = dict(seed=None, deterministic=False) pretrained = './resnet50_3rdparty-mill_in21k_20220331-faac000b.pth' model = dict( type='ImageToImageRetriever', image_encoder=[ dict( type='ResNet', depth=50, init_cfg=dict( type='Pretrained', checkpoint= './resnet50_3rdparty-mill_in21k_20220331-faac000b.pth', prefix='backbone')), dict(type='GlobalAveragePooling') ], head=dict( type='ArcFaceClsHead', num_classes=3997, in_channels=2048, loss=dict(type='CrossEntropyLoss', loss_weight=1.0), init_cfg=None), prototype=dict( batch_size=32, num_workers=4, dataset=dict( type='InShop', data_root='data/inshop', split='gallery', pipeline=[ dict(type='LoadImageFromFile'), dict(type='Resize', scale=512), dict(type='CenterCrop', crop_size=448), dict(type='PackClsInputs') ]), sampler=dict(type='DefaultSampler', shuffle=False))) custom_hooks = [ dict(type='PrepareProtoBeforeValLoopHook'), dict(type='SyncBuffersHook') ] launcher = 'pytorch' work_dir = 'inshop74-1' 2023/02/02 19:52:55 - 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: (NORMAL ) PrepareProtoBeforeValLoopHook -------------------- before_val_epoch: (NORMAL ) IterTimerHook -------------------- before_val_iter: (NORMAL ) IterTimerHook -------------------- after_val_iter: (NORMAL ) IterTimerHook (NORMAL ) VisualizationHook (BELOW_NORMAL) LoggerHook -------------------- after_val_epoch: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (BELOW_NORMAL) LoggerHook (LOW ) ParamSchedulerHook (VERY_LOW ) CheckpointHook -------------------- before_test_epoch: (NORMAL ) IterTimerHook -------------------- before_test_iter: (NORMAL ) IterTimerHook -------------------- after_test_iter: (NORMAL ) IterTimerHook (NORMAL ) VisualizationHook (BELOW_NORMAL) LoggerHook -------------------- after_test_epoch: (VERY_HIGH ) RuntimeInfoHook (NORMAL ) IterTimerHook (BELOW_NORMAL) LoggerHook -------------------- after_run: (BELOW_NORMAL) LoggerHook -------------------- 2023/02/02 19:52:56 - mmengine - INFO - LR is set based on batch size of 256 and the current batch size is 256. Scaling the original LR by 1.0. 2023/02/02 19:52:56 - mmengine - INFO - load backbone in model from: ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth Name of parameter - Initialization information image_encoder.0.conv1.weight - torch.Size([64, 3, 7, 7]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.bn1.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.bn1.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.conv1.weight - torch.Size([64, 64, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn1.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn1.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn2.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn2.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn3.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.bn3.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.downsample.0.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.downsample.1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.0.downsample.1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.conv1.weight - torch.Size([64, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn1.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn1.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn2.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn2.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn3.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.1.bn3.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.conv1.weight - torch.Size([64, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn1.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn1.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.conv2.weight - torch.Size([64, 64, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn2.weight - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn2.bias - torch.Size([64]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.conv3.weight - torch.Size([256, 64, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn3.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer1.2.bn3.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.conv1.weight - torch.Size([128, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn1.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn1.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.conv2.weight - torch.Size([128, 128, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn2.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn2.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.conv3.weight - torch.Size([512, 128, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn3.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.bn3.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.downsample.0.weight - torch.Size([512, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.downsample.1.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.0.downsample.1.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.conv1.weight - torch.Size([128, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn1.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn1.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.conv2.weight - torch.Size([128, 128, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn2.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn2.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.conv3.weight - torch.Size([512, 128, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn3.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.1.bn3.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.conv1.weight - torch.Size([128, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn1.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn1.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.conv2.weight - torch.Size([128, 128, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn2.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn2.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.conv3.weight - torch.Size([512, 128, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn3.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.2.bn3.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.conv1.weight - torch.Size([128, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn1.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn1.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.conv2.weight - torch.Size([128, 128, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn2.weight - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn2.bias - torch.Size([128]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.conv3.weight - torch.Size([512, 128, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn3.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer2.3.bn3.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.conv1.weight - torch.Size([256, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.downsample.0.weight - torch.Size([1024, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.downsample.1.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.0.downsample.1.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.conv1.weight - torch.Size([256, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.1.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.conv1.weight - torch.Size([256, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.2.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.conv1.weight - torch.Size([256, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.3.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.conv1.weight - torch.Size([256, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.4.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.conv1.weight - torch.Size([256, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn1.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn1.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.conv2.weight - torch.Size([256, 256, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn2.weight - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn2.bias - torch.Size([256]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.conv3.weight - torch.Size([1024, 256, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn3.weight - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer3.5.bn3.bias - torch.Size([1024]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.conv1.weight - torch.Size([512, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn1.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn1.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.conv2.weight - torch.Size([512, 512, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn2.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn2.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.conv3.weight - torch.Size([2048, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn3.weight - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.bn3.bias - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.downsample.0.weight - torch.Size([2048, 1024, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.downsample.1.weight - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.0.downsample.1.bias - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.conv1.weight - torch.Size([512, 2048, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn1.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn1.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.conv2.weight - torch.Size([512, 512, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn2.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn2.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.conv3.weight - torch.Size([2048, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn3.weight - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.1.bn3.bias - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.conv1.weight - torch.Size([512, 2048, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn1.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn1.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.conv2.weight - torch.Size([512, 512, 3, 3]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn2.weight - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn2.bias - torch.Size([512]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.conv3.weight - torch.Size([2048, 512, 1, 1]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn3.weight - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth image_encoder.0.layer4.2.bn3.bias - torch.Size([2048]): PretrainedInit: load from ./resnet50_3rdparty-mill_in21k_20220331-faac000b.pth head.norm_product.weight - torch.Size([3997, 2048]): The value is the same before and after calling `init_weights` of ImageToImageRetriever 2023/02/02 19:52:56 - mmengine - INFO - Checkpoints will be saved to /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1. 2023/02/02 19:53:00 - mmengine - INFO - Epoch(train) [1][ 20/102] lr: 9.3910e-04 eta: 0:16:26 time: 0.1496 data_time: 0.0004 memory: 11138 loss: 39.6562 2023/02/02 19:53:03 - mmengine - INFO - Epoch(train) [1][ 40/102] lr: 1.7171e-03 eta: 0:14:30 time: 0.1497 data_time: 0.0004 memory: 11138 loss: 39.3445 2023/02/02 19:53:06 - mmengine - INFO - Epoch(train) [1][ 60/102] lr: 2.4951e-03 eta: 0:13:50 time: 0.1504 data_time: 0.0004 memory: 11138 loss: 39.0008 2023/02/02 19:53:09 - mmengine - INFO - Epoch(train) [1][ 80/102] lr: 3.2731e-03 eta: 0:13:28 time: 0.1496 data_time: 0.0004 memory: 11138 loss: 38.8909 2023/02/02 19:53:12 - mmengine - INFO - Epoch(train) [1][100/102] lr: 4.0511e-03 eta: 0:13:13 time: 0.1494 data_time: 0.0003 memory: 11138 loss: 38.9795 2023/02/02 19:53:12 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:53:12 - mmengine - INFO - Saving checkpoint at 1 epochs 2023/02/02 19:53:20 - mmengine - INFO - Epoch(val) [1][20/56] eta: 0:00:03 time: 0.0823 data_time: 0.0003 memory: 11138 2023/02/02 19:53:22 - mmengine - INFO - Epoch(val) [1][40/56] eta: 0:00:01 time: 0.0794 data_time: 0.0003 memory: 3822 2023/02/02 19:53:24 - mmengine - INFO - Epoch(val) [1][56/56] retrieval/Recall@1: 31.4953 2023/02/02 19:53:24 - mmengine - INFO - The best checkpoint with 31.4953 retrieval/Recall@1 at 1 epoch is saved to best_retrieval/Recall@1_epoch_1.pth. 2023/02/02 19:53:27 - mmengine - INFO - Epoch(train) [2][ 20/102] lr: 4.9069e-03 eta: 0:13:01 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 38.6856 2023/02/02 19:53:30 - mmengine - INFO - Epoch(train) [2][ 40/102] lr: 5.6849e-03 eta: 0:12:53 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 38.4067 2023/02/02 19:53:33 - mmengine - INFO - Epoch(train) [2][ 60/102] lr: 6.4629e-03 eta: 0:12:46 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 38.2257 2023/02/02 19:53:36 - mmengine - INFO - Epoch(train) [2][ 80/102] lr: 7.2409e-03 eta: 0:12:40 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 37.7834 2023/02/02 19:53:39 - mmengine - INFO - Epoch(train) [2][100/102] lr: 8.0189e-03 eta: 0:12:34 time: 0.1492 data_time: 0.0003 memory: 11240 loss: 37.5010 2023/02/02 19:53:40 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:53:40 - mmengine - INFO - Saving checkpoint at 2 epochs 2023/02/02 19:53:47 - mmengine - INFO - Epoch(val) [2][20/56] eta: 0:00:03 time: 0.0835 data_time: 0.0003 memory: 11240 2023/02/02 19:53:48 - mmengine - INFO - Epoch(val) [2][40/56] eta: 0:00:01 time: 0.0796 data_time: 0.0003 memory: 3822 2023/02/02 19:53:50 - mmengine - INFO - Epoch(val) [2][56/56] retrieval/Recall@1: 42.8401 2023/02/02 19:53:50 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_1.pth is removed 2023/02/02 19:53:51 - mmengine - INFO - The best checkpoint with 42.8401 retrieval/Recall@1 at 2 epoch is saved to best_retrieval/Recall@1_epoch_2.pth. 2023/02/02 19:53:54 - mmengine - INFO - Epoch(train) [3][ 20/102] lr: 8.8747e-03 eta: 0:12:28 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 36.4508 2023/02/02 19:53:57 - mmengine - INFO - Epoch(train) [3][ 40/102] lr: 9.6527e-03 eta: 0:12:24 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 36.0694 2023/02/02 19:54:00 - mmengine - INFO - Epoch(train) [3][ 60/102] lr: 1.0431e-02 eta: 0:12:20 time: 0.1500 data_time: 0.0004 memory: 11240 loss: 35.6526 2023/02/02 19:54:03 - mmengine - INFO - Epoch(train) [3][ 80/102] lr: 1.1209e-02 eta: 0:12:15 time: 0.1496 data_time: 0.0005 memory: 11240 loss: 35.0507 2023/02/02 19:54:06 - mmengine - INFO - Epoch(train) [3][100/102] lr: 1.1987e-02 eta: 0:12:11 time: 0.1493 data_time: 0.0003 memory: 11240 loss: 34.6292 2023/02/02 19:54:06 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:54:06 - mmengine - INFO - Saving checkpoint at 3 epochs 2023/02/02 19:54:13 - mmengine - INFO - Epoch(val) [3][20/56] eta: 0:00:03 time: 0.0819 data_time: 0.0003 memory: 11240 2023/02/02 19:54:15 - mmengine - INFO - Epoch(val) [3][40/56] eta: 0:00:01 time: 0.0791 data_time: 0.0003 memory: 3822 2023/02/02 19:54:17 - mmengine - INFO - Epoch(val) [3][56/56] retrieval/Recall@1: 54.4591 2023/02/02 19:54:17 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_2.pth is removed 2023/02/02 19:54:18 - mmengine - INFO - The best checkpoint with 54.4591 retrieval/Recall@1 at 3 epoch is saved to best_retrieval/Recall@1_epoch_3.pth. 2023/02/02 19:54:21 - mmengine - INFO - Epoch(train) [4][ 20/102] lr: 1.2842e-02 eta: 0:12:06 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 33.4045 2023/02/02 19:54:24 - mmengine - INFO - Epoch(train) [4][ 40/102] lr: 1.3620e-02 eta: 0:12:03 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 32.8383 2023/02/02 19:54:27 - mmengine - INFO - Epoch(train) [4][ 60/102] lr: 1.4398e-02 eta: 0:11:59 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 32.6000 2023/02/02 19:54:30 - mmengine - INFO - Epoch(train) [4][ 80/102] lr: 1.5176e-02 eta: 0:11:55 time: 0.1500 data_time: 0.0004 memory: 11240 loss: 31.9749 2023/02/02 19:54:33 - mmengine - INFO - Epoch(train) [4][100/102] lr: 1.5954e-02 eta: 0:11:52 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 31.7699 2023/02/02 19:54:33 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:54:33 - mmengine - INFO - Saving checkpoint at 4 epochs 2023/02/02 19:54:40 - mmengine - INFO - Epoch(val) [4][20/56] eta: 0:00:03 time: 0.0800 data_time: 0.0002 memory: 11240 2023/02/02 19:54:41 - mmengine - INFO - Epoch(val) [4][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 19:54:44 - mmengine - INFO - Epoch(val) [4][56/56] retrieval/Recall@1: 52.6867 2023/02/02 19:54:47 - mmengine - INFO - Epoch(train) [5][ 20/102] lr: 1.6810e-02 eta: 0:11:48 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 29.3499 2023/02/02 19:54:50 - mmengine - INFO - Epoch(train) [5][ 40/102] lr: 1.7588e-02 eta: 0:11:45 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 29.5080 2023/02/02 19:54:53 - mmengine - INFO - Epoch(train) [5][ 60/102] lr: 1.8366e-02 eta: 0:11:41 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 29.4199 2023/02/02 19:54:56 - mmengine - INFO - Epoch(train) [5][ 80/102] lr: 1.9144e-02 eta: 0:11:38 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 28.3161 2023/02/02 19:54:59 - mmengine - INFO - Epoch(train) [5][100/102] lr: 1.9922e-02 eta: 0:11:35 time: 0.1495 data_time: 0.0003 memory: 11240 loss: 28.1206 2023/02/02 19:54:59 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:54:59 - mmengine - INFO - Saving checkpoint at 5 epochs 2023/02/02 19:55:06 - mmengine - INFO - Epoch(val) [5][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 19:55:08 - mmengine - INFO - Epoch(val) [5][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 19:55:10 - mmengine - INFO - Epoch(val) [5][56/56] retrieval/Recall@1: 64.9599 2023/02/02 19:55:10 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_3.pth is removed 2023/02/02 19:55:10 - mmengine - INFO - The best checkpoint with 64.9599 retrieval/Recall@1 at 5 epoch is saved to best_retrieval/Recall@1_epoch_5.pth. 2023/02/02 19:55:13 - mmengine - INFO - Epoch(train) [6][ 20/102] lr: 2.0000e-02 eta: 0:11:31 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 25.3717 2023/02/02 19:55:16 - mmengine - INFO - Epoch(train) [6][ 40/102] lr: 2.0000e-02 eta: 0:11:27 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 24.5296 2023/02/02 19:55:19 - mmengine - INFO - Epoch(train) [6][ 60/102] lr: 2.0000e-02 eta: 0:11:24 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 24.7016 2023/02/02 19:55:22 - mmengine - INFO - Epoch(train) [6][ 80/102] lr: 2.0000e-02 eta: 0:11:21 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 23.4871 2023/02/02 19:55:25 - mmengine - INFO - Epoch(train) [6][100/102] lr: 2.0000e-02 eta: 0:11:18 time: 0.1496 data_time: 0.0003 memory: 11240 loss: 23.6110 2023/02/02 19:55:26 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:55:26 - mmengine - INFO - Saving checkpoint at 6 epochs 2023/02/02 19:55:33 - mmengine - INFO - Epoch(val) [6][20/56] eta: 0:00:03 time: 0.0822 data_time: 0.0003 memory: 11240 2023/02/02 19:55:34 - mmengine - INFO - Epoch(val) [6][40/56] eta: 0:00:01 time: 0.0793 data_time: 0.0003 memory: 3822 2023/02/02 19:55:37 - mmengine - INFO - Epoch(val) [6][56/56] retrieval/Recall@1: 62.6248 2023/02/02 19:55:40 - mmengine - INFO - Epoch(train) [7][ 20/102] lr: 1.9976e-02 eta: 0:11:14 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 21.0233 2023/02/02 19:55:43 - mmengine - INFO - Epoch(train) [7][ 40/102] lr: 1.9976e-02 eta: 0:11:11 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 20.6550 2023/02/02 19:55:46 - mmengine - INFO - Epoch(train) [7][ 60/102] lr: 1.9976e-02 eta: 0:11:08 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 19.7886 2023/02/02 19:55:49 - mmengine - INFO - Epoch(train) [7][ 80/102] lr: 1.9976e-02 eta: 0:11:05 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 19.3746 2023/02/02 19:55:52 - mmengine - INFO - Epoch(train) [7][100/102] lr: 1.9976e-02 eta: 0:11:02 time: 0.1497 data_time: 0.0003 memory: 11240 loss: 19.1447 2023/02/02 19:55:52 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:55:52 - mmengine - INFO - Saving checkpoint at 7 epochs 2023/02/02 19:55:59 - mmengine - INFO - Epoch(val) [7][20/56] eta: 0:00:03 time: 0.0822 data_time: 0.0003 memory: 11240 2023/02/02 19:56:00 - mmengine - INFO - Epoch(val) [7][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 19:56:03 - mmengine - INFO - Epoch(val) [7][56/56] retrieval/Recall@1: 66.0641 2023/02/02 19:56:03 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_5.pth is removed 2023/02/02 19:56:03 - mmengine - INFO - The best checkpoint with 66.0641 retrieval/Recall@1 at 7 epoch is saved to best_retrieval/Recall@1_epoch_7.pth. 2023/02/02 19:56:06 - mmengine - INFO - Epoch(train) [8][ 20/102] lr: 1.9903e-02 eta: 0:10:58 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 16.1340 2023/02/02 19:56:09 - mmengine - INFO - Epoch(train) [8][ 40/102] lr: 1.9903e-02 eta: 0:10:55 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 15.9370 2023/02/02 19:56:12 - mmengine - INFO - Epoch(train) [8][ 60/102] lr: 1.9903e-02 eta: 0:10:52 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 15.8115 2023/02/02 19:56:15 - mmengine - INFO - Epoch(train) [8][ 80/102] lr: 1.9903e-02 eta: 0:10:49 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 15.7493 2023/02/02 19:56:18 - mmengine - INFO - Epoch(train) [8][100/102] lr: 1.9903e-02 eta: 0:10:45 time: 0.1495 data_time: 0.0003 memory: 11240 loss: 16.2621 2023/02/02 19:56:18 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:56:18 - mmengine - INFO - Saving checkpoint at 8 epochs 2023/02/02 19:56:26 - mmengine - INFO - Epoch(val) [8][20/56] eta: 0:00:03 time: 0.0819 data_time: 0.0003 memory: 11240 2023/02/02 19:56:27 - mmengine - INFO - Epoch(val) [8][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 19:56:29 - mmengine - INFO - Epoch(val) [8][56/56] retrieval/Recall@1: 76.1429 2023/02/02 19:56:29 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_7.pth is removed 2023/02/02 19:56:30 - mmengine - INFO - The best checkpoint with 76.1429 retrieval/Recall@1 at 8 epoch is saved to best_retrieval/Recall@1_epoch_8.pth. 2023/02/02 19:56:33 - mmengine - INFO - Epoch(train) [9][ 20/102] lr: 1.9781e-02 eta: 0:10:42 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 12.7136 2023/02/02 19:56:36 - mmengine - INFO - Epoch(train) [9][ 40/102] lr: 1.9781e-02 eta: 0:10:39 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 12.9898 2023/02/02 19:56:39 - mmengine - INFO - Epoch(train) [9][ 60/102] lr: 1.9781e-02 eta: 0:10:36 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 12.6148 2023/02/02 19:56:42 - mmengine - INFO - Epoch(train) [9][ 80/102] lr: 1.9781e-02 eta: 0:10:33 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 12.2282 2023/02/02 19:56:45 - mmengine - INFO - Epoch(train) [9][100/102] lr: 1.9781e-02 eta: 0:10:29 time: 0.1496 data_time: 0.0003 memory: 11240 loss: 12.6431 2023/02/02 19:56:45 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:56:45 - mmengine - INFO - Saving checkpoint at 9 epochs 2023/02/02 19:56:52 - mmengine - INFO - Epoch(val) [9][20/56] eta: 0:00:03 time: 0.0809 data_time: 0.0003 memory: 11240 2023/02/02 19:56:54 - mmengine - INFO - Epoch(val) [9][40/56] eta: 0:00:01 time: 0.0780 data_time: 0.0003 memory: 3822 2023/02/02 19:56:56 - mmengine - INFO - Epoch(val) [9][56/56] retrieval/Recall@1: 70.6921 2023/02/02 19:56:59 - mmengine - INFO - Epoch(train) [10][ 20/102] lr: 1.9613e-02 eta: 0:10:26 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 11.1438 2023/02/02 19:57:02 - mmengine - INFO - Epoch(train) [10][ 40/102] lr: 1.9613e-02 eta: 0:10:23 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 11.0372 2023/02/02 19:57:05 - mmengine - INFO - Epoch(train) [10][ 60/102] lr: 1.9613e-02 eta: 0:10:20 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 11.5807 2023/02/02 19:57:08 - mmengine - INFO - Epoch(train) [10][ 80/102] lr: 1.9613e-02 eta: 0:10:17 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 10.8350 2023/02/02 19:57:09 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:57:11 - mmengine - INFO - Epoch(train) [10][100/102] lr: 1.9613e-02 eta: 0:10:14 time: 0.1496 data_time: 0.0003 memory: 11240 loss: 11.1855 2023/02/02 19:57:11 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:57:11 - mmengine - INFO - Saving checkpoint at 10 epochs 2023/02/02 19:57:19 - mmengine - INFO - Epoch(val) [10][20/56] eta: 0:00:03 time: 0.0823 data_time: 0.0003 memory: 11240 2023/02/02 19:57:20 - mmengine - INFO - Epoch(val) [10][40/56] eta: 0:00:01 time: 0.0788 data_time: 0.0003 memory: 3822 2023/02/02 19:57:23 - mmengine - INFO - Epoch(val) [10][56/56] retrieval/Recall@1: 68.3851 2023/02/02 19:57:26 - mmengine - INFO - Epoch(train) [11][ 20/102] lr: 1.9397e-02 eta: 0:10:10 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 9.2148 2023/02/02 19:57:29 - mmengine - INFO - Epoch(train) [11][ 40/102] lr: 1.9397e-02 eta: 0:10:07 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 8.4416 2023/02/02 19:57:32 - mmengine - INFO - Epoch(train) [11][ 60/102] lr: 1.9397e-02 eta: 0:10:04 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 8.2512 2023/02/02 19:57:35 - mmengine - INFO - Epoch(train) [11][ 80/102] lr: 1.9397e-02 eta: 0:10:01 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 8.9758 2023/02/02 19:57:38 - mmengine - INFO - Epoch(train) [11][100/102] lr: 1.9397e-02 eta: 0:09:58 time: 0.1495 data_time: 0.0003 memory: 11240 loss: 8.8910 2023/02/02 19:57:38 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:57:38 - mmengine - INFO - Saving checkpoint at 11 epochs 2023/02/02 19:57:45 - mmengine - INFO - Epoch(val) [11][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 19:57:47 - mmengine - INFO - Epoch(val) [11][40/56] eta: 0:00:01 time: 0.0792 data_time: 0.0003 memory: 3822 2023/02/02 19:57:49 - mmengine - INFO - Epoch(val) [11][56/56] retrieval/Recall@1: 67.2035 2023/02/02 19:57:52 - mmengine - INFO - Epoch(train) [12][ 20/102] lr: 1.9135e-02 eta: 0:09:55 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 7.5530 2023/02/02 19:57:55 - mmengine - INFO - Epoch(train) [12][ 40/102] lr: 1.9135e-02 eta: 0:09:52 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 7.5742 2023/02/02 19:57:58 - mmengine - INFO - Epoch(train) [12][ 60/102] lr: 1.9135e-02 eta: 0:09:49 time: 0.1500 data_time: 0.0004 memory: 11240 loss: 7.9788 2023/02/02 19:58:01 - mmengine - INFO - Epoch(train) [12][ 80/102] lr: 1.9135e-02 eta: 0:09:46 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 6.6821 2023/02/02 19:58:04 - mmengine - INFO - Epoch(train) [12][100/102] lr: 1.9135e-02 eta: 0:09:43 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 7.3478 2023/02/02 19:58:04 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:58:04 - mmengine - INFO - Saving checkpoint at 12 epochs 2023/02/02 19:58:11 - mmengine - INFO - Epoch(val) [12][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 19:58:13 - mmengine - INFO - Epoch(val) [12][40/56] eta: 0:00:01 time: 0.0787 data_time: 0.0003 memory: 3822 2023/02/02 19:58:15 - mmengine - INFO - Epoch(val) [12][56/56] retrieval/Recall@1: 73.2170 2023/02/02 19:58:18 - mmengine - INFO - Epoch(train) [13][ 20/102] lr: 1.8829e-02 eta: 0:09:39 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 5.3958 2023/02/02 19:58:21 - mmengine - INFO - Epoch(train) [13][ 40/102] lr: 1.8829e-02 eta: 0:09:36 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 5.9914 2023/02/02 19:58:24 - mmengine - INFO - Epoch(train) [13][ 60/102] lr: 1.8829e-02 eta: 0:09:33 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 6.1648 2023/02/02 19:58:27 - mmengine - INFO - Epoch(train) [13][ 80/102] lr: 1.8829e-02 eta: 0:09:30 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 6.4624 2023/02/02 19:58:30 - mmengine - INFO - Epoch(train) [13][100/102] lr: 1.8829e-02 eta: 0:09:27 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 6.9988 2023/02/02 19:58:30 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:58:30 - mmengine - INFO - Saving checkpoint at 13 epochs 2023/02/02 19:58:37 - mmengine - INFO - Epoch(val) [13][20/56] eta: 0:00:03 time: 0.0815 data_time: 0.0003 memory: 11240 2023/02/02 19:58:39 - mmengine - INFO - Epoch(val) [13][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 19:58:41 - mmengine - INFO - Epoch(val) [13][56/56] retrieval/Recall@1: 77.1557 2023/02/02 19:58:41 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_8.pth is removed 2023/02/02 19:58:42 - mmengine - INFO - The best checkpoint with 77.1557 retrieval/Recall@1 at 13 epoch is saved to best_retrieval/Recall@1_epoch_13.pth. 2023/02/02 19:58:45 - mmengine - INFO - Epoch(train) [14][ 20/102] lr: 1.8480e-02 eta: 0:09:24 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 5.6272 2023/02/02 19:58:48 - mmengine - INFO - Epoch(train) [14][ 40/102] lr: 1.8480e-02 eta: 0:09:21 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 4.6094 2023/02/02 19:58:51 - mmengine - INFO - Epoch(train) [14][ 60/102] lr: 1.8480e-02 eta: 0:09:18 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 5.1318 2023/02/02 19:58:54 - mmengine - INFO - Epoch(train) [14][ 80/102] lr: 1.8480e-02 eta: 0:09:15 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 5.5701 2023/02/02 19:58:57 - mmengine - INFO - Epoch(train) [14][100/102] lr: 1.8480e-02 eta: 0:09:12 time: 0.1495 data_time: 0.0003 memory: 11240 loss: 5.8726 2023/02/02 19:58:57 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:58:57 - mmengine - INFO - Saving checkpoint at 14 epochs 2023/02/02 19:59:04 - mmengine - INFO - Epoch(val) [14][20/56] eta: 0:00:03 time: 0.0874 data_time: 0.0003 memory: 11240 2023/02/02 19:59:06 - mmengine - INFO - Epoch(val) [14][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 19:59:08 - mmengine - INFO - Epoch(val) [14][56/56] retrieval/Recall@1: 80.8131 2023/02/02 19:59:08 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_13.pth is removed 2023/02/02 19:59:08 - mmengine - INFO - The best checkpoint with 80.8131 retrieval/Recall@1 at 14 epoch is saved to best_retrieval/Recall@1_epoch_14.pth. 2023/02/02 19:59:11 - mmengine - INFO - Epoch(train) [15][ 20/102] lr: 1.8090e-02 eta: 0:09:08 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 4.5207 2023/02/02 19:59:14 - mmengine - INFO - Epoch(train) [15][ 40/102] lr: 1.8090e-02 eta: 0:09:05 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 4.0687 2023/02/02 19:59:17 - mmengine - INFO - Epoch(train) [15][ 60/102] lr: 1.8090e-02 eta: 0:09:02 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 4.7351 2023/02/02 19:59:20 - mmengine - INFO - Epoch(train) [15][ 80/102] lr: 1.8090e-02 eta: 0:08:59 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 4.3602 2023/02/02 19:59:23 - mmengine - INFO - Epoch(train) [15][100/102] lr: 1.8090e-02 eta: 0:08:56 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 4.6484 2023/02/02 19:59:24 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:59:24 - mmengine - INFO - Saving checkpoint at 15 epochs 2023/02/02 19:59:31 - mmengine - INFO - Epoch(val) [15][20/56] eta: 0:00:03 time: 0.0827 data_time: 0.0003 memory: 11240 2023/02/02 19:59:32 - mmengine - INFO - Epoch(val) [15][40/56] eta: 0:00:01 time: 0.0793 data_time: 0.0003 memory: 3822 2023/02/02 19:59:34 - mmengine - INFO - Epoch(val) [15][56/56] retrieval/Recall@1: 79.8284 2023/02/02 19:59:38 - mmengine - INFO - Epoch(train) [16][ 20/102] lr: 1.7660e-02 eta: 0:08:53 time: 0.1496 data_time: 0.0005 memory: 11240 loss: 4.6310 2023/02/02 19:59:41 - mmengine - INFO - Epoch(train) [16][ 40/102] lr: 1.7660e-02 eta: 0:08:50 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 4.2298 2023/02/02 19:59:44 - mmengine - INFO - Epoch(train) [16][ 60/102] lr: 1.7660e-02 eta: 0:08:47 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 4.1776 2023/02/02 19:59:47 - mmengine - INFO - Epoch(train) [16][ 80/102] lr: 1.7660e-02 eta: 0:08:44 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 4.0761 2023/02/02 19:59:50 - mmengine - INFO - Epoch(train) [16][100/102] lr: 1.7660e-02 eta: 0:08:41 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 4.3864 2023/02/02 19:59:50 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 19:59:50 - mmengine - INFO - Saving checkpoint at 16 epochs 2023/02/02 19:59:57 - mmengine - INFO - Epoch(val) [16][20/56] eta: 0:00:03 time: 0.0821 data_time: 0.0003 memory: 11240 2023/02/02 19:59:59 - mmengine - INFO - Epoch(val) [16][40/56] eta: 0:00:01 time: 0.0794 data_time: 0.0003 memory: 3822 2023/02/02 20:00:01 - mmengine - INFO - Epoch(val) [16][56/56] retrieval/Recall@1: 79.5541 2023/02/02 20:00:04 - mmengine - INFO - Epoch(train) [17][ 20/102] lr: 1.7193e-02 eta: 0:08:38 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 3.9755 2023/02/02 20:00:07 - mmengine - INFO - Epoch(train) [17][ 40/102] lr: 1.7193e-02 eta: 0:08:35 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 2.4597 2023/02/02 20:00:10 - mmengine - INFO - Epoch(train) [17][ 60/102] lr: 1.7193e-02 eta: 0:08:32 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 3.6425 2023/02/02 20:00:13 - mmengine - INFO - Epoch(train) [17][ 80/102] lr: 1.7193e-02 eta: 0:08:28 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 3.3715 2023/02/02 20:00:16 - mmengine - INFO - Epoch(train) [17][100/102] lr: 1.7193e-02 eta: 0:08:25 time: 0.1496 data_time: 0.0003 memory: 11240 loss: 3.6276 2023/02/02 20:00:16 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:00:16 - mmengine - INFO - Saving checkpoint at 17 epochs 2023/02/02 20:00:23 - mmengine - INFO - Epoch(val) [17][20/56] eta: 0:00:03 time: 0.0816 data_time: 0.0003 memory: 11240 2023/02/02 20:00:25 - mmengine - INFO - Epoch(val) [17][40/56] eta: 0:00:01 time: 0.0787 data_time: 0.0003 memory: 3822 2023/02/02 20:00:27 - mmengine - INFO - Epoch(val) [17][56/56] retrieval/Recall@1: 81.8188 2023/02/02 20:00:27 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_14.pth is removed 2023/02/02 20:00:28 - mmengine - INFO - The best checkpoint with 81.8188 retrieval/Recall@1 at 17 epoch is saved to best_retrieval/Recall@1_epoch_17.pth. 2023/02/02 20:00:31 - mmengine - INFO - Epoch(train) [18][ 20/102] lr: 1.6691e-02 eta: 0:08:22 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 3.2611 2023/02/02 20:00:34 - mmengine - INFO - Epoch(train) [18][ 40/102] lr: 1.6691e-02 eta: 0:08:19 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 3.1260 2023/02/02 20:00:37 - mmengine - INFO - Epoch(train) [18][ 60/102] lr: 1.6691e-02 eta: 0:08:16 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 2.9063 2023/02/02 20:00:40 - mmengine - INFO - Epoch(train) [18][ 80/102] lr: 1.6691e-02 eta: 0:08:13 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 3.1598 2023/02/02 20:00:43 - mmengine - INFO - Epoch(train) [18][100/102] lr: 1.6691e-02 eta: 0:08:10 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 2.9009 2023/02/02 20:00:43 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:00:43 - mmengine - INFO - Saving checkpoint at 18 epochs 2023/02/02 20:00:50 - mmengine - INFO - Epoch(val) [18][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 20:00:51 - mmengine - INFO - Epoch(val) [18][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:00:54 - mmengine - INFO - Epoch(val) [18][56/56] retrieval/Recall@1: 81.9947 2023/02/02 20:00:54 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_17.pth is removed 2023/02/02 20:00:54 - mmengine - INFO - The best checkpoint with 81.9947 retrieval/Recall@1 at 18 epoch is saved to best_retrieval/Recall@1_epoch_18.pth. 2023/02/02 20:00:57 - mmengine - INFO - Epoch(train) [19][ 20/102] lr: 1.6157e-02 eta: 0:08:07 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 2.8548 2023/02/02 20:01:00 - mmengine - INFO - Epoch(train) [19][ 40/102] lr: 1.6157e-02 eta: 0:08:04 time: 0.1496 data_time: 0.0005 memory: 11240 loss: 2.5654 2023/02/02 20:01:03 - mmengine - INFO - Epoch(train) [19][ 60/102] lr: 1.6157e-02 eta: 0:08:01 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 3.0368 2023/02/02 20:01:06 - mmengine - INFO - Epoch(train) [19][ 80/102] lr: 1.6157e-02 eta: 0:07:58 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 2.6522 2023/02/02 20:01:09 - mmengine - INFO - Epoch(train) [19][100/102] lr: 1.6157e-02 eta: 0:07:55 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 3.0858 2023/02/02 20:01:09 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:01:09 - mmengine - INFO - Saving checkpoint at 19 epochs 2023/02/02 20:01:17 - mmengine - INFO - Epoch(val) [19][20/56] eta: 0:00:03 time: 0.0821 data_time: 0.0003 memory: 11240 2023/02/02 20:01:18 - mmengine - INFO - Epoch(val) [19][40/56] eta: 0:00:01 time: 0.0793 data_time: 0.0003 memory: 3822 2023/02/02 20:01:20 - mmengine - INFO - Epoch(val) [19][56/56] retrieval/Recall@1: 83.3169 2023/02/02 20:01:20 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_18.pth is removed 2023/02/02 20:01:21 - mmengine - INFO - The best checkpoint with 83.3169 retrieval/Recall@1 at 19 epoch is saved to best_retrieval/Recall@1_epoch_19.pth. 2023/02/02 20:01:24 - mmengine - INFO - Epoch(train) [20][ 20/102] lr: 1.5592e-02 eta: 0:07:51 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 2.3422 2023/02/02 20:01:27 - mmengine - INFO - Epoch(train) [20][ 40/102] lr: 1.5592e-02 eta: 0:07:48 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.9664 2023/02/02 20:01:30 - mmengine - INFO - Epoch(train) [20][ 60/102] lr: 1.5592e-02 eta: 0:07:45 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 2.0819 2023/02/02 20:01:30 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:01:33 - mmengine - INFO - Epoch(train) [20][ 80/102] lr: 1.5592e-02 eta: 0:07:42 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 2.0043 2023/02/02 20:01:36 - mmengine - INFO - Epoch(train) [20][100/102] lr: 1.5592e-02 eta: 0:07:39 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 2.6060 2023/02/02 20:01:36 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:01:36 - mmengine - INFO - Saving checkpoint at 20 epochs 2023/02/02 20:01:43 - mmengine - INFO - Epoch(val) [20][20/56] eta: 0:00:03 time: 0.0821 data_time: 0.0003 memory: 11240 2023/02/02 20:01:45 - mmengine - INFO - Epoch(val) [20][40/56] eta: 0:00:01 time: 0.0798 data_time: 0.0006 memory: 3822 2023/02/02 20:01:47 - mmengine - INFO - Epoch(val) [20][56/56] retrieval/Recall@1: 83.3943 2023/02/02 20:01:47 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_19.pth is removed 2023/02/02 20:01:48 - mmengine - INFO - The best checkpoint with 83.3943 retrieval/Recall@1 at 20 epoch is saved to best_retrieval/Recall@1_epoch_20.pth. 2023/02/02 20:01:51 - mmengine - INFO - Epoch(train) [21][ 20/102] lr: 1.5000e-02 eta: 0:07:36 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 2.2391 2023/02/02 20:01:54 - mmengine - INFO - Epoch(train) [21][ 40/102] lr: 1.5000e-02 eta: 0:07:33 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 2.4144 2023/02/02 20:01:57 - mmengine - INFO - Epoch(train) [21][ 60/102] lr: 1.5000e-02 eta: 0:07:30 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.8599 2023/02/02 20:02:00 - mmengine - INFO - Epoch(train) [21][ 80/102] lr: 1.5000e-02 eta: 0:07:27 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.9543 2023/02/02 20:02:03 - mmengine - INFO - Epoch(train) [21][100/102] lr: 1.5000e-02 eta: 0:07:24 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 2.0816 2023/02/02 20:02:03 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:02:03 - mmengine - INFO - Saving checkpoint at 21 epochs 2023/02/02 20:02:10 - mmengine - INFO - Epoch(val) [21][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:02:12 - mmengine - INFO - Epoch(val) [21][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:02:14 - mmengine - INFO - Epoch(val) [21][56/56] retrieval/Recall@1: 75.8194 2023/02/02 20:02:17 - mmengine - INFO - Epoch(train) [22][ 20/102] lr: 1.4384e-02 eta: 0:07:20 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.6234 2023/02/02 20:02:20 - mmengine - INFO - Epoch(train) [22][ 40/102] lr: 1.4384e-02 eta: 0:07:17 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 1.6490 2023/02/02 20:02:23 - mmengine - INFO - Epoch(train) [22][ 60/102] lr: 1.4384e-02 eta: 0:07:14 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 2.1433 2023/02/02 20:02:26 - mmengine - INFO - Epoch(train) [22][ 80/102] lr: 1.4384e-02 eta: 0:07:11 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.7547 2023/02/02 20:02:29 - mmengine - INFO - Epoch(train) [22][100/102] lr: 1.4384e-02 eta: 0:07:08 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 1.8317 2023/02/02 20:02:29 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:02:29 - mmengine - INFO - Saving checkpoint at 22 epochs 2023/02/02 20:02:36 - mmengine - INFO - Epoch(val) [22][20/56] eta: 0:00:03 time: 0.0823 data_time: 0.0003 memory: 11240 2023/02/02 20:02:38 - mmengine - INFO - Epoch(val) [22][40/56] eta: 0:00:01 time: 0.0793 data_time: 0.0003 memory: 3822 2023/02/02 20:02:40 - mmengine - INFO - Epoch(val) [22][56/56] retrieval/Recall@1: 78.3303 2023/02/02 20:02:43 - mmengine - INFO - Epoch(train) [23][ 20/102] lr: 1.3746e-02 eta: 0:07:05 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 1.7580 2023/02/02 20:02:46 - mmengine - INFO - Epoch(train) [23][ 40/102] lr: 1.3746e-02 eta: 0:07:02 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.9826 2023/02/02 20:02:49 - mmengine - INFO - Epoch(train) [23][ 60/102] lr: 1.3746e-02 eta: 0:06:59 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.4907 2023/02/02 20:02:52 - mmengine - INFO - Epoch(train) [23][ 80/102] lr: 1.3746e-02 eta: 0:06:56 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.6529 2023/02/02 20:02:55 - mmengine - INFO - Epoch(train) [23][100/102] lr: 1.3746e-02 eta: 0:06:53 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 1.0242 2023/02/02 20:02:55 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:02:55 - mmengine - INFO - Saving checkpoint at 23 epochs 2023/02/02 20:03:03 - mmengine - INFO - Epoch(val) [23][20/56] eta: 0:00:03 time: 0.0818 data_time: 0.0003 memory: 11240 2023/02/02 20:03:04 - mmengine - INFO - Epoch(val) [23][40/56] eta: 0:00:01 time: 0.0794 data_time: 0.0003 memory: 3822 2023/02/02 20:03:06 - mmengine - INFO - Epoch(val) [23][56/56] retrieval/Recall@1: 84.5548 2023/02/02 20:03:06 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_20.pth is removed 2023/02/02 20:03:07 - mmengine - INFO - The best checkpoint with 84.5548 retrieval/Recall@1 at 23 epoch is saved to best_retrieval/Recall@1_epoch_23.pth. 2023/02/02 20:03:10 - mmengine - INFO - Epoch(train) [24][ 20/102] lr: 1.3090e-02 eta: 0:06:50 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.6729 2023/02/02 20:03:13 - mmengine - INFO - Epoch(train) [24][ 40/102] lr: 1.3090e-02 eta: 0:06:47 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 1.0430 2023/02/02 20:03:16 - mmengine - INFO - Epoch(train) [24][ 60/102] lr: 1.3090e-02 eta: 0:06:44 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.2431 2023/02/02 20:03:19 - mmengine - INFO - Epoch(train) [24][ 80/102] lr: 1.3090e-02 eta: 0:06:41 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.1299 2023/02/02 20:03:22 - mmengine - INFO - Epoch(train) [24][100/102] lr: 1.3090e-02 eta: 0:06:38 time: 0.1494 data_time: 0.0004 memory: 11240 loss: 1.2103 2023/02/02 20:03:22 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:03:22 - mmengine - INFO - Saving checkpoint at 24 epochs 2023/02/02 20:03:29 - mmengine - INFO - Epoch(val) [24][20/56] eta: 0:00:03 time: 0.0898 data_time: 0.0003 memory: 11240 2023/02/02 20:03:31 - mmengine - INFO - Epoch(val) [24][40/56] eta: 0:00:01 time: 0.0788 data_time: 0.0003 memory: 3822 2023/02/02 20:03:33 - mmengine - INFO - Epoch(val) [24][56/56] retrieval/Recall@1: 77.5074 2023/02/02 20:03:36 - mmengine - INFO - Epoch(train) [25][ 20/102] lr: 1.2419e-02 eta: 0:06:34 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.0762 2023/02/02 20:03:39 - mmengine - INFO - Epoch(train) [25][ 40/102] lr: 1.2419e-02 eta: 0:06:31 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6961 2023/02/02 20:03:42 - mmengine - INFO - Epoch(train) [25][ 60/102] lr: 1.2419e-02 eta: 0:06:28 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.3425 2023/02/02 20:03:45 - mmengine - INFO - Epoch(train) [25][ 80/102] lr: 1.2419e-02 eta: 0:06:25 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 1.1649 2023/02/02 20:03:48 - mmengine - INFO - Epoch(train) [25][100/102] lr: 1.2419e-02 eta: 0:06:22 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 1.1782 2023/02/02 20:03:48 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:03:48 - mmengine - INFO - Saving checkpoint at 25 epochs 2023/02/02 20:03:56 - mmengine - INFO - Epoch(val) [25][20/56] eta: 0:00:03 time: 0.0819 data_time: 0.0003 memory: 11240 2023/02/02 20:03:57 - mmengine - INFO - Epoch(val) [25][40/56] eta: 0:00:01 time: 0.0792 data_time: 0.0003 memory: 3822 2023/02/02 20:03:59 - mmengine - INFO - Epoch(val) [25][56/56] retrieval/Recall@1: 84.7869 2023/02/02 20:03:59 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_23.pth is removed 2023/02/02 20:04:00 - mmengine - INFO - The best checkpoint with 84.7869 retrieval/Recall@1 at 25 epoch is saved to best_retrieval/Recall@1_epoch_25.pth. 2023/02/02 20:04:03 - mmengine - INFO - Epoch(train) [26][ 20/102] lr: 1.1736e-02 eta: 0:06:19 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 1.0849 2023/02/02 20:04:06 - mmengine - INFO - Epoch(train) [26][ 40/102] lr: 1.1736e-02 eta: 0:06:16 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.8963 2023/02/02 20:04:09 - mmengine - INFO - Epoch(train) [26][ 60/102] lr: 1.1736e-02 eta: 0:06:13 time: 0.1500 data_time: 0.0006 memory: 11240 loss: 1.0903 2023/02/02 20:04:12 - mmengine - INFO - Epoch(train) [26][ 80/102] lr: 1.1736e-02 eta: 0:06:10 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.2255 2023/02/02 20:04:15 - mmengine - INFO - Epoch(train) [26][100/102] lr: 1.1736e-02 eta: 0:06:07 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.9867 2023/02/02 20:04:15 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:04:15 - mmengine - INFO - Saving checkpoint at 26 epochs 2023/02/02 20:04:22 - mmengine - INFO - Epoch(val) [26][20/56] eta: 0:00:03 time: 0.0822 data_time: 0.0003 memory: 11240 2023/02/02 20:04:24 - mmengine - INFO - Epoch(val) [26][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 20:04:26 - mmengine - INFO - Epoch(val) [26][56/56] retrieval/Recall@1: 85.1034 2023/02/02 20:04:26 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_25.pth is removed 2023/02/02 20:04:27 - mmengine - INFO - The best checkpoint with 85.1034 retrieval/Recall@1 at 26 epoch is saved to best_retrieval/Recall@1_epoch_26.pth. 2023/02/02 20:04:30 - mmengine - INFO - Epoch(train) [27][ 20/102] lr: 1.1045e-02 eta: 0:06:04 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 1.0643 2023/02/02 20:04:33 - mmengine - INFO - Epoch(train) [27][ 40/102] lr: 1.1045e-02 eta: 0:06:01 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.8222 2023/02/02 20:04:36 - mmengine - INFO - Epoch(train) [27][ 60/102] lr: 1.1045e-02 eta: 0:05:58 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6625 2023/02/02 20:04:39 - mmengine - INFO - Epoch(train) [27][ 80/102] lr: 1.1045e-02 eta: 0:05:55 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 1.1126 2023/02/02 20:04:42 - mmengine - INFO - Epoch(train) [27][100/102] lr: 1.1045e-02 eta: 0:05:52 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 0.6974 2023/02/02 20:04:42 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:04:42 - mmengine - INFO - Saving checkpoint at 27 epochs 2023/02/02 20:04:49 - mmengine - INFO - Epoch(val) [27][20/56] eta: 0:00:03 time: 0.0811 data_time: 0.0003 memory: 11240 2023/02/02 20:04:51 - mmengine - INFO - Epoch(val) [27][40/56] eta: 0:00:01 time: 0.0782 data_time: 0.0003 memory: 3822 2023/02/02 20:04:53 - mmengine - INFO - Epoch(val) [27][56/56] retrieval/Recall@1: 87.8323 2023/02/02 20:04:53 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_26.pth is removed 2023/02/02 20:04:53 - mmengine - INFO - The best checkpoint with 87.8323 retrieval/Recall@1 at 27 epoch is saved to best_retrieval/Recall@1_epoch_27.pth. 2023/02/02 20:04:57 - mmengine - INFO - Epoch(train) [28][ 20/102] lr: 1.0349e-02 eta: 0:05:48 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6941 2023/02/02 20:05:00 - mmengine - INFO - Epoch(train) [28][ 40/102] lr: 1.0349e-02 eta: 0:05:45 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.7221 2023/02/02 20:05:03 - mmengine - INFO - Epoch(train) [28][ 60/102] lr: 1.0349e-02 eta: 0:05:42 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.7404 2023/02/02 20:05:06 - mmengine - INFO - Epoch(train) [28][ 80/102] lr: 1.0349e-02 eta: 0:05:39 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6467 2023/02/02 20:05:09 - mmengine - INFO - Epoch(train) [28][100/102] lr: 1.0349e-02 eta: 0:05:36 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.6852 2023/02/02 20:05:09 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:05:09 - mmengine - INFO - Saving checkpoint at 28 epochs 2023/02/02 20:05:16 - mmengine - INFO - Epoch(val) [28][20/56] eta: 0:00:03 time: 0.0819 data_time: 0.0003 memory: 11240 2023/02/02 20:05:17 - mmengine - INFO - Epoch(val) [28][40/56] eta: 0:00:01 time: 0.0860 data_time: 0.0003 memory: 3822 2023/02/02 20:05:19 - mmengine - INFO - Epoch(val) [28][56/56] retrieval/Recall@1: 85.3214 2023/02/02 20:05:23 - mmengine - INFO - Epoch(train) [29][ 20/102] lr: 9.6510e-03 eta: 0:05:33 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6872 2023/02/02 20:05:26 - mmengine - INFO - Epoch(train) [29][ 40/102] lr: 9.6510e-03 eta: 0:05:30 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.9656 2023/02/02 20:05:29 - mmengine - INFO - Epoch(train) [29][ 60/102] lr: 9.6510e-03 eta: 0:05:27 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.5859 2023/02/02 20:05:32 - mmengine - INFO - Epoch(train) [29][ 80/102] lr: 9.6510e-03 eta: 0:05:24 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.5944 2023/02/02 20:05:35 - mmengine - INFO - Epoch(train) [29][100/102] lr: 9.6510e-03 eta: 0:05:21 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.5383 2023/02/02 20:05:35 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:05:35 - mmengine - INFO - Saving checkpoint at 29 epochs 2023/02/02 20:05:42 - mmengine - INFO - Epoch(val) [29][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:05:44 - mmengine - INFO - Epoch(val) [29][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:05:46 - mmengine - INFO - Epoch(val) [29][56/56] retrieval/Recall@1: 87.8675 2023/02/02 20:05:46 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_27.pth is removed 2023/02/02 20:05:46 - mmengine - INFO - The best checkpoint with 87.8675 retrieval/Recall@1 at 29 epoch is saved to best_retrieval/Recall@1_epoch_29.pth. 2023/02/02 20:05:49 - mmengine - INFO - Epoch(train) [30][ 20/102] lr: 8.9547e-03 eta: 0:05:18 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 0.6352 2023/02/02 20:05:52 - mmengine - INFO - Epoch(train) [30][ 40/102] lr: 8.9547e-03 eta: 0:05:15 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.4156 2023/02/02 20:05:53 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:05:55 - mmengine - INFO - Epoch(train) [30][ 60/102] lr: 8.9547e-03 eta: 0:05:12 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.5469 2023/02/02 20:05:58 - mmengine - INFO - Epoch(train) [30][ 80/102] lr: 8.9547e-03 eta: 0:05:09 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.5587 2023/02/02 20:06:01 - mmengine - INFO - Epoch(train) [30][100/102] lr: 8.9547e-03 eta: 0:05:06 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.5945 2023/02/02 20:06:02 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:06:02 - mmengine - INFO - Saving checkpoint at 30 epochs 2023/02/02 20:06:09 - mmengine - INFO - Epoch(val) [30][20/56] eta: 0:00:03 time: 0.0814 data_time: 0.0003 memory: 11240 2023/02/02 20:06:10 - mmengine - INFO - Epoch(val) [30][40/56] eta: 0:00:01 time: 0.0850 data_time: 0.0003 memory: 3822 2023/02/02 20:06:12 - mmengine - INFO - Epoch(val) [30][56/56] retrieval/Recall@1: 83.7319 2023/02/02 20:06:16 - mmengine - INFO - Epoch(train) [31][ 20/102] lr: 8.2635e-03 eta: 0:05:03 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.4299 2023/02/02 20:06:19 - mmengine - INFO - Epoch(train) [31][ 40/102] lr: 8.2635e-03 eta: 0:05:00 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.6732 2023/02/02 20:06:22 - mmengine - INFO - Epoch(train) [31][ 60/102] lr: 8.2635e-03 eta: 0:04:57 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.3765 2023/02/02 20:06:25 - mmengine - INFO - Epoch(train) [31][ 80/102] lr: 8.2635e-03 eta: 0:04:54 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.4451 2023/02/02 20:06:28 - mmengine - INFO - Epoch(train) [31][100/102] lr: 8.2635e-03 eta: 0:04:51 time: 0.1498 data_time: 0.0004 memory: 11240 loss: 0.5763 2023/02/02 20:06:28 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:06:28 - mmengine - INFO - Saving checkpoint at 31 epochs 2023/02/02 20:06:35 - mmengine - INFO - Epoch(val) [31][20/56] eta: 0:00:03 time: 0.0816 data_time: 0.0003 memory: 11240 2023/02/02 20:06:36 - mmengine - INFO - Epoch(val) [31][40/56] eta: 0:00:01 time: 0.0788 data_time: 0.0003 memory: 3822 2023/02/02 20:06:39 - mmengine - INFO - Epoch(val) [31][56/56] retrieval/Recall@1: 87.4525 2023/02/02 20:06:42 - mmengine - INFO - Epoch(train) [32][ 20/102] lr: 7.5808e-03 eta: 0:04:47 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.4837 2023/02/02 20:06:45 - mmengine - INFO - Epoch(train) [32][ 40/102] lr: 7.5808e-03 eta: 0:04:44 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.4503 2023/02/02 20:06:48 - mmengine - INFO - Epoch(train) [32][ 60/102] lr: 7.5808e-03 eta: 0:04:41 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2898 2023/02/02 20:06:51 - mmengine - INFO - Epoch(train) [32][ 80/102] lr: 7.5808e-03 eta: 0:04:38 time: 0.1502 data_time: 0.0005 memory: 11240 loss: 0.4535 2023/02/02 20:06:54 - mmengine - INFO - Epoch(train) [32][100/102] lr: 7.5808e-03 eta: 0:04:35 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.3006 2023/02/02 20:06:54 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:06:54 - mmengine - INFO - Saving checkpoint at 32 epochs 2023/02/02 20:07:01 - mmengine - INFO - Epoch(val) [32][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 20:07:03 - mmengine - INFO - Epoch(val) [32][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:07:05 - mmengine - INFO - Epoch(val) [32][56/56] retrieval/Recall@1: 86.9602 2023/02/02 20:07:08 - mmengine - INFO - Epoch(train) [33][ 20/102] lr: 6.9098e-03 eta: 0:04:32 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.3015 2023/02/02 20:07:11 - mmengine - INFO - Epoch(train) [33][ 40/102] lr: 6.9098e-03 eta: 0:04:29 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2541 2023/02/02 20:07:14 - mmengine - INFO - Epoch(train) [33][ 60/102] lr: 6.9098e-03 eta: 0:04:26 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.4294 2023/02/02 20:07:17 - mmengine - INFO - Epoch(train) [33][ 80/102] lr: 6.9098e-03 eta: 0:04:23 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2175 2023/02/02 20:07:20 - mmengine - INFO - Epoch(train) [33][100/102] lr: 6.9098e-03 eta: 0:04:20 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.3449 2023/02/02 20:07:20 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:07:20 - mmengine - INFO - Saving checkpoint at 33 epochs 2023/02/02 20:07:27 - mmengine - INFO - Epoch(val) [33][20/56] eta: 0:00:03 time: 0.0814 data_time: 0.0003 memory: 11240 2023/02/02 20:07:29 - mmengine - INFO - Epoch(val) [33][40/56] eta: 0:00:01 time: 0.0791 data_time: 0.0003 memory: 3822 2023/02/02 20:07:31 - mmengine - INFO - Epoch(val) [33][56/56] retrieval/Recall@1: 87.5862 2023/02/02 20:07:34 - mmengine - INFO - Epoch(train) [34][ 20/102] lr: 6.2539e-03 eta: 0:04:17 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2988 2023/02/02 20:07:37 - mmengine - INFO - Epoch(train) [34][ 40/102] lr: 6.2539e-03 eta: 0:04:14 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.3265 2023/02/02 20:07:40 - mmengine - INFO - Epoch(train) [34][ 60/102] lr: 6.2539e-03 eta: 0:04:11 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.2369 2023/02/02 20:07:43 - mmengine - INFO - Epoch(train) [34][ 80/102] lr: 6.2539e-03 eta: 0:04:08 time: 0.1504 data_time: 0.0005 memory: 11240 loss: 0.3163 2023/02/02 20:07:46 - mmengine - INFO - Epoch(train) [34][100/102] lr: 6.2539e-03 eta: 0:04:05 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.2417 2023/02/02 20:07:47 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:07:47 - mmengine - INFO - Saving checkpoint at 34 epochs 2023/02/02 20:07:54 - mmengine - INFO - Epoch(val) [34][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 20:07:55 - mmengine - INFO - Epoch(val) [34][40/56] eta: 0:00:01 time: 0.0852 data_time: 0.0002 memory: 3822 2023/02/02 20:07:57 - mmengine - INFO - Epoch(val) [34][56/56] retrieval/Recall@1: 87.8534 2023/02/02 20:08:01 - mmengine - INFO - Epoch(train) [35][ 20/102] lr: 5.6163e-03 eta: 0:04:01 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.4230 2023/02/02 20:08:04 - mmengine - INFO - Epoch(train) [35][ 40/102] lr: 5.6163e-03 eta: 0:03:58 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2110 2023/02/02 20:08:07 - mmengine - INFO - Epoch(train) [35][ 60/102] lr: 5.6163e-03 eta: 0:03:55 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1967 2023/02/02 20:08:10 - mmengine - INFO - Epoch(train) [35][ 80/102] lr: 5.6163e-03 eta: 0:03:52 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2211 2023/02/02 20:08:13 - mmengine - INFO - Epoch(train) [35][100/102] lr: 5.6163e-03 eta: 0:03:49 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.1691 2023/02/02 20:08:13 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:08:13 - mmengine - INFO - Saving checkpoint at 35 epochs 2023/02/02 20:08:20 - mmengine - INFO - Epoch(val) [35][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:08:21 - mmengine - INFO - Epoch(val) [35][40/56] eta: 0:00:01 time: 0.0788 data_time: 0.0003 memory: 3822 2023/02/02 20:08:24 - mmengine - INFO - Epoch(val) [35][56/56] retrieval/Recall@1: 87.9589 2023/02/02 20:08:24 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_29.pth is removed 2023/02/02 20:08:24 - mmengine - INFO - The best checkpoint with 87.9589 retrieval/Recall@1 at 35 epoch is saved to best_retrieval/Recall@1_epoch_35.pth. 2023/02/02 20:08:27 - mmengine - INFO - Epoch(train) [36][ 20/102] lr: 5.0000e-03 eta: 0:03:46 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 0.3555 2023/02/02 20:08:30 - mmengine - INFO - Epoch(train) [36][ 40/102] lr: 5.0000e-03 eta: 0:03:43 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.3214 2023/02/02 20:08:33 - mmengine - INFO - Epoch(train) [36][ 60/102] lr: 5.0000e-03 eta: 0:03:40 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.1186 2023/02/02 20:08:36 - mmengine - INFO - Epoch(train) [36][ 80/102] lr: 5.0000e-03 eta: 0:03:37 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1773 2023/02/02 20:08:39 - mmengine - INFO - Epoch(train) [36][100/102] lr: 5.0000e-03 eta: 0:03:34 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.2001 2023/02/02 20:08:39 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:08:39 - mmengine - INFO - Saving checkpoint at 36 epochs 2023/02/02 20:08:46 - mmengine - INFO - Epoch(val) [36][20/56] eta: 0:00:03 time: 0.0815 data_time: 0.0003 memory: 11240 2023/02/02 20:08:48 - mmengine - INFO - Epoch(val) [36][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 20:08:50 - mmengine - INFO - Epoch(val) [36][56/56] retrieval/Recall@1: 88.0996 2023/02/02 20:08:50 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_35.pth is removed 2023/02/02 20:08:51 - mmengine - INFO - The best checkpoint with 88.0996 retrieval/Recall@1 at 36 epoch is saved to best_retrieval/Recall@1_epoch_36.pth. 2023/02/02 20:08:54 - mmengine - INFO - Epoch(train) [37][ 20/102] lr: 4.4081e-03 eta: 0:03:31 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2151 2023/02/02 20:08:57 - mmengine - INFO - Epoch(train) [37][ 40/102] lr: 4.4081e-03 eta: 0:03:28 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.3533 2023/02/02 20:09:00 - mmengine - INFO - Epoch(train) [37][ 60/102] lr: 4.4081e-03 eta: 0:03:25 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1769 2023/02/02 20:09:03 - mmengine - INFO - Epoch(train) [37][ 80/102] lr: 4.4081e-03 eta: 0:03:22 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2167 2023/02/02 20:09:06 - mmengine - INFO - Epoch(train) [37][100/102] lr: 4.4081e-03 eta: 0:03:19 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 0.2636 2023/02/02 20:09:06 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:09:06 - mmengine - INFO - Saving checkpoint at 37 epochs 2023/02/02 20:09:13 - mmengine - INFO - Epoch(val) [37][20/56] eta: 0:00:03 time: 0.0816 data_time: 0.0002 memory: 11240 2023/02/02 20:09:15 - mmengine - INFO - Epoch(val) [37][40/56] eta: 0:00:01 time: 0.0844 data_time: 0.0003 memory: 3822 2023/02/02 20:09:17 - mmengine - INFO - Epoch(val) [37][56/56] retrieval/Recall@1: 88.5568 2023/02/02 20:09:17 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_36.pth is removed 2023/02/02 20:09:18 - mmengine - INFO - The best checkpoint with 88.5568 retrieval/Recall@1 at 37 epoch is saved to best_retrieval/Recall@1_epoch_37.pth. 2023/02/02 20:09:21 - mmengine - INFO - Epoch(train) [38][ 20/102] lr: 3.8434e-03 eta: 0:03:15 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1512 2023/02/02 20:09:24 - mmengine - INFO - Epoch(train) [38][ 40/102] lr: 3.8434e-03 eta: 0:03:12 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2583 2023/02/02 20:09:27 - mmengine - INFO - Epoch(train) [38][ 60/102] lr: 3.8434e-03 eta: 0:03:09 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2134 2023/02/02 20:09:30 - mmengine - INFO - Epoch(train) [38][ 80/102] lr: 3.8434e-03 eta: 0:03:06 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1322 2023/02/02 20:09:33 - mmengine - INFO - Epoch(train) [38][100/102] lr: 3.8434e-03 eta: 0:03:03 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.2279 2023/02/02 20:09:33 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:09:33 - mmengine - INFO - Saving checkpoint at 38 epochs 2023/02/02 20:09:40 - mmengine - INFO - Epoch(val) [38][20/56] eta: 0:00:03 time: 0.0819 data_time: 0.0003 memory: 11240 2023/02/02 20:09:42 - mmengine - INFO - Epoch(val) [38][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 20:09:44 - mmengine - INFO - Epoch(val) [38][56/56] retrieval/Recall@1: 88.1840 2023/02/02 20:09:47 - mmengine - INFO - Epoch(train) [39][ 20/102] lr: 3.3087e-03 eta: 0:03:00 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2542 2023/02/02 20:09:50 - mmengine - INFO - Epoch(train) [39][ 40/102] lr: 3.3087e-03 eta: 0:02:57 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2952 2023/02/02 20:09:53 - mmengine - INFO - Epoch(train) [39][ 60/102] lr: 3.3087e-03 eta: 0:02:54 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1063 2023/02/02 20:09:56 - mmengine - INFO - Epoch(train) [39][ 80/102] lr: 3.3087e-03 eta: 0:02:51 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2589 2023/02/02 20:09:59 - mmengine - INFO - Epoch(train) [39][100/102] lr: 3.3087e-03 eta: 0:02:48 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.1302 2023/02/02 20:09:59 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:09:59 - mmengine - INFO - Saving checkpoint at 39 epochs 2023/02/02 20:10:06 - mmengine - INFO - Epoch(val) [39][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:10:08 - mmengine - INFO - Epoch(val) [39][40/56] eta: 0:00:01 time: 0.0787 data_time: 0.0003 memory: 3822 2023/02/02 20:10:10 - mmengine - INFO - Epoch(val) [39][56/56] retrieval/Recall@1: 88.5005 2023/02/02 20:10:13 - mmengine - INFO - Epoch(train) [40][ 20/102] lr: 2.8066e-03 eta: 0:02:45 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2027 2023/02/02 20:10:13 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:10:16 - mmengine - INFO - Epoch(train) [40][ 40/102] lr: 2.8066e-03 eta: 0:02:42 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1027 2023/02/02 20:10:19 - mmengine - INFO - Epoch(train) [40][ 60/102] lr: 2.8066e-03 eta: 0:02:39 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2019 2023/02/02 20:10:22 - mmengine - INFO - Epoch(train) [40][ 80/102] lr: 2.8066e-03 eta: 0:02:36 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1556 2023/02/02 20:10:25 - mmengine - INFO - Epoch(train) [40][100/102] lr: 2.8066e-03 eta: 0:02:33 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.1120 2023/02/02 20:10:25 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:10:25 - mmengine - INFO - Saving checkpoint at 40 epochs 2023/02/02 20:10:32 - mmengine - INFO - Epoch(val) [40][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 20:10:34 - mmengine - INFO - Epoch(val) [40][40/56] eta: 0:00:01 time: 0.0858 data_time: 0.0003 memory: 3822 2023/02/02 20:10:36 - mmengine - INFO - Epoch(val) [40][56/56] retrieval/Recall@1: 89.5274 2023/02/02 20:10:36 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_37.pth is removed 2023/02/02 20:10:37 - mmengine - INFO - The best checkpoint with 89.5274 retrieval/Recall@1 at 40 epoch is saved to best_retrieval/Recall@1_epoch_40.pth. 2023/02/02 20:10:40 - mmengine - INFO - Epoch(train) [41][ 20/102] lr: 2.3396e-03 eta: 0:02:29 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2597 2023/02/02 20:10:43 - mmengine - INFO - Epoch(train) [41][ 40/102] lr: 2.3396e-03 eta: 0:02:26 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1781 2023/02/02 20:10:46 - mmengine - INFO - Epoch(train) [41][ 60/102] lr: 2.3396e-03 eta: 0:02:23 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1021 2023/02/02 20:10:49 - mmengine - INFO - Epoch(train) [41][ 80/102] lr: 2.3396e-03 eta: 0:02:20 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1391 2023/02/02 20:10:52 - mmengine - INFO - Epoch(train) [41][100/102] lr: 2.3396e-03 eta: 0:02:17 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.1908 2023/02/02 20:10:52 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:10:52 - mmengine - INFO - Saving checkpoint at 41 epochs 2023/02/02 20:10:59 - mmengine - INFO - Epoch(val) [41][20/56] eta: 0:00:03 time: 0.0820 data_time: 0.0003 memory: 11240 2023/02/02 20:11:01 - mmengine - INFO - Epoch(val) [41][40/56] eta: 0:00:01 time: 0.0792 data_time: 0.0003 memory: 3822 2023/02/02 20:11:03 - mmengine - INFO - Epoch(val) [41][56/56] retrieval/Recall@1: 89.9072 2023/02/02 20:11:03 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_40.pth is removed 2023/02/02 20:11:04 - mmengine - INFO - The best checkpoint with 89.9072 retrieval/Recall@1 at 41 epoch is saved to best_retrieval/Recall@1_epoch_41.pth. 2023/02/02 20:11:07 - mmengine - INFO - Epoch(train) [42][ 20/102] lr: 1.9098e-03 eta: 0:02:14 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1291 2023/02/02 20:11:10 - mmengine - INFO - Epoch(train) [42][ 40/102] lr: 1.9098e-03 eta: 0:02:11 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1549 2023/02/02 20:11:13 - mmengine - INFO - Epoch(train) [42][ 60/102] lr: 1.9098e-03 eta: 0:02:08 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2651 2023/02/02 20:11:16 - mmengine - INFO - Epoch(train) [42][ 80/102] lr: 1.9098e-03 eta: 0:02:05 time: 0.1502 data_time: 0.0005 memory: 11240 loss: 0.1141 2023/02/02 20:11:19 - mmengine - INFO - Epoch(train) [42][100/102] lr: 1.9098e-03 eta: 0:02:02 time: 0.1501 data_time: 0.0004 memory: 11240 loss: 0.0935 2023/02/02 20:11:19 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:11:19 - mmengine - INFO - Saving checkpoint at 42 epochs 2023/02/02 20:11:26 - mmengine - INFO - Epoch(val) [42][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:11:28 - mmengine - INFO - Epoch(val) [42][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:11:30 - mmengine - INFO - Epoch(val) [42][56/56] retrieval/Recall@1: 89.5274 2023/02/02 20:11:33 - mmengine - INFO - Epoch(train) [43][ 20/102] lr: 1.5195e-03 eta: 0:01:59 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2134 2023/02/02 20:11:36 - mmengine - INFO - Epoch(train) [43][ 40/102] lr: 1.5195e-03 eta: 0:01:56 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1777 2023/02/02 20:11:39 - mmengine - INFO - Epoch(train) [43][ 60/102] lr: 1.5195e-03 eta: 0:01:53 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1354 2023/02/02 20:11:42 - mmengine - INFO - Epoch(train) [43][ 80/102] lr: 1.5195e-03 eta: 0:01:50 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1118 2023/02/02 20:11:45 - mmengine - INFO - Epoch(train) [43][100/102] lr: 1.5195e-03 eta: 0:01:47 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 0.1955 2023/02/02 20:11:45 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:11:45 - mmengine - INFO - Saving checkpoint at 43 epochs 2023/02/02 20:11:52 - mmengine - INFO - Epoch(val) [43][20/56] eta: 0:00:03 time: 0.0816 data_time: 0.0003 memory: 11240 2023/02/02 20:11:54 - mmengine - INFO - Epoch(val) [43][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:11:56 - mmengine - INFO - Epoch(val) [43][56/56] retrieval/Recall@1: 89.8228 2023/02/02 20:11:59 - mmengine - INFO - Epoch(train) [44][ 20/102] lr: 1.1705e-03 eta: 0:01:44 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.2327 2023/02/02 20:12:02 - mmengine - INFO - Epoch(train) [44][ 40/102] lr: 1.1705e-03 eta: 0:01:41 time: 0.1497 data_time: 0.0005 memory: 11240 loss: 0.1356 2023/02/02 20:12:05 - mmengine - INFO - Epoch(train) [44][ 60/102] lr: 1.1705e-03 eta: 0:01:38 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1412 2023/02/02 20:12:08 - mmengine - INFO - Epoch(train) [44][ 80/102] lr: 1.1705e-03 eta: 0:01:35 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1147 2023/02/02 20:12:11 - mmengine - INFO - Epoch(train) [44][100/102] lr: 1.1705e-03 eta: 0:01:32 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 0.1046 2023/02/02 20:12:11 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:12:11 - mmengine - INFO - Saving checkpoint at 44 epochs 2023/02/02 20:12:18 - mmengine - INFO - Epoch(val) [44][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:12:20 - mmengine - INFO - Epoch(val) [44][40/56] eta: 0:00:01 time: 0.0790 data_time: 0.0003 memory: 3822 2023/02/02 20:12:22 - mmengine - INFO - Epoch(val) [44][56/56] retrieval/Recall@1: 90.2237 2023/02/02 20:12:22 - mmengine - INFO - The previous best checkpoint /mm_model/yuzhaohui/mmlab/okotaku/inshop74-1/best_retrieval/Recall@1_epoch_41.pth is removed 2023/02/02 20:12:23 - mmengine - INFO - The best checkpoint with 90.2237 retrieval/Recall@1 at 44 epoch is saved to best_retrieval/Recall@1_epoch_44.pth. 2023/02/02 20:12:26 - mmengine - INFO - Epoch(train) [45][ 20/102] lr: 8.6455e-04 eta: 0:01:28 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.0401 2023/02/02 20:12:29 - mmengine - INFO - Epoch(train) [45][ 40/102] lr: 8.6455e-04 eta: 0:01:25 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1980 2023/02/02 20:12:32 - mmengine - INFO - Epoch(train) [45][ 60/102] lr: 8.6455e-04 eta: 0:01:22 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1217 2023/02/02 20:12:35 - mmengine - INFO - Epoch(train) [45][ 80/102] lr: 8.6455e-04 eta: 0:01:19 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.0571 2023/02/02 20:12:38 - mmengine - INFO - Epoch(train) [45][100/102] lr: 8.6455e-04 eta: 0:01:16 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.1875 2023/02/02 20:12:38 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:12:38 - mmengine - INFO - Saving checkpoint at 45 epochs 2023/02/02 20:12:45 - mmengine - INFO - Epoch(val) [45][20/56] eta: 0:00:03 time: 0.0816 data_time: 0.0003 memory: 11240 2023/02/02 20:12:47 - mmengine - INFO - Epoch(val) [45][40/56] eta: 0:00:01 time: 0.0787 data_time: 0.0003 memory: 3822 2023/02/02 20:12:49 - mmengine - INFO - Epoch(val) [45][56/56] retrieval/Recall@1: 89.9986 2023/02/02 20:12:52 - mmengine - INFO - Epoch(train) [46][ 20/102] lr: 6.0307e-04 eta: 0:01:13 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1064 2023/02/02 20:12:55 - mmengine - INFO - Epoch(train) [46][ 40/102] lr: 6.0307e-04 eta: 0:01:10 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2322 2023/02/02 20:12:58 - mmengine - INFO - Epoch(train) [46][ 60/102] lr: 6.0307e-04 eta: 0:01:07 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.0635 2023/02/02 20:13:01 - mmengine - INFO - Epoch(train) [46][ 80/102] lr: 6.0307e-04 eta: 0:01:04 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.1023 2023/02/02 20:13:04 - mmengine - INFO - Epoch(train) [46][100/102] lr: 6.0307e-04 eta: 0:01:01 time: 0.1499 data_time: 0.0004 memory: 11240 loss: 0.1576 2023/02/02 20:13:04 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:13:04 - mmengine - INFO - Saving checkpoint at 46 epochs 2023/02/02 20:13:11 - mmengine - INFO - Epoch(val) [46][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:13:13 - mmengine - INFO - Epoch(val) [46][40/56] eta: 0:00:01 time: 0.0789 data_time: 0.0003 memory: 3822 2023/02/02 20:13:15 - mmengine - INFO - Epoch(val) [46][56/56] retrieval/Recall@1: 90.1463 2023/02/02 20:13:18 - mmengine - INFO - Epoch(train) [47][ 20/102] lr: 3.8738e-04 eta: 0:00:58 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1252 2023/02/02 20:13:21 - mmengine - INFO - Epoch(train) [47][ 40/102] lr: 3.8738e-04 eta: 0:00:55 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1863 2023/02/02 20:13:24 - mmengine - INFO - Epoch(train) [47][ 60/102] lr: 3.8738e-04 eta: 0:00:52 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1596 2023/02/02 20:13:27 - mmengine - INFO - Epoch(train) [47][ 80/102] lr: 3.8738e-04 eta: 0:00:49 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1908 2023/02/02 20:13:30 - mmengine - INFO - Epoch(train) [47][100/102] lr: 3.8738e-04 eta: 0:00:46 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.1808 2023/02/02 20:13:30 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:13:30 - mmengine - INFO - Saving checkpoint at 47 epochs 2023/02/02 20:13:38 - mmengine - INFO - Epoch(val) [47][20/56] eta: 0:00:03 time: 0.0822 data_time: 0.0003 memory: 11240 2023/02/02 20:13:39 - mmengine - INFO - Epoch(val) [47][40/56] eta: 0:00:01 time: 0.0793 data_time: 0.0003 memory: 3822 2023/02/02 20:13:41 - mmengine - INFO - Epoch(val) [47][56/56] retrieval/Recall@1: 90.1815 2023/02/02 20:13:44 - mmengine - INFO - Epoch(train) [48][ 20/102] lr: 2.1852e-04 eta: 0:00:42 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1017 2023/02/02 20:13:47 - mmengine - INFO - Epoch(train) [48][ 40/102] lr: 2.1852e-04 eta: 0:00:39 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2018 2023/02/02 20:13:50 - mmengine - INFO - Epoch(train) [48][ 60/102] lr: 2.1852e-04 eta: 0:00:36 time: 0.1501 data_time: 0.0005 memory: 11240 loss: 0.1452 2023/02/02 20:13:53 - mmengine - INFO - Epoch(train) [48][ 80/102] lr: 2.1852e-04 eta: 0:00:33 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.2366 2023/02/02 20:13:56 - mmengine - INFO - Epoch(train) [48][100/102] lr: 2.1852e-04 eta: 0:00:30 time: 0.1497 data_time: 0.0004 memory: 11240 loss: 0.1686 2023/02/02 20:13:57 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:13:57 - mmengine - INFO - Saving checkpoint at 48 epochs 2023/02/02 20:14:04 - mmengine - INFO - Epoch(val) [48][20/56] eta: 0:00:03 time: 0.0818 data_time: 0.0003 memory: 11240 2023/02/02 20:14:05 - mmengine - INFO - Epoch(val) [48][40/56] eta: 0:00:01 time: 0.0785 data_time: 0.0003 memory: 3822 2023/02/02 20:14:07 - mmengine - INFO - Epoch(val) [48][56/56] retrieval/Recall@1: 90.1041 2023/02/02 20:14:11 - mmengine - INFO - Epoch(train) [49][ 20/102] lr: 9.7319e-05 eta: 0:00:27 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.0899 2023/02/02 20:14:14 - mmengine - INFO - Epoch(train) [49][ 40/102] lr: 9.7319e-05 eta: 0:00:24 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1456 2023/02/02 20:14:17 - mmengine - INFO - Epoch(train) [49][ 60/102] lr: 9.7319e-05 eta: 0:00:21 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.0876 2023/02/02 20:14:20 - mmengine - INFO - Epoch(train) [49][ 80/102] lr: 9.7319e-05 eta: 0:00:18 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.0924 2023/02/02 20:14:23 - mmengine - INFO - Epoch(train) [49][100/102] lr: 9.7319e-05 eta: 0:00:15 time: 0.1495 data_time: 0.0004 memory: 11240 loss: 0.1256 2023/02/02 20:14:23 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:14:23 - mmengine - INFO - Saving checkpoint at 49 epochs 2023/02/02 20:14:30 - mmengine - INFO - Epoch(val) [49][20/56] eta: 0:00:03 time: 0.0818 data_time: 0.0003 memory: 11240 2023/02/02 20:14:32 - mmengine - INFO - Epoch(val) [49][40/56] eta: 0:00:01 time: 0.0788 data_time: 0.0003 memory: 3822 2023/02/02 20:14:34 - mmengine - INFO - Epoch(val) [49][56/56] retrieval/Recall@1: 90.1815 2023/02/02 20:14:34 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:14:37 - mmengine - INFO - Epoch(train) [50][ 20/102] lr: 2.4359e-05 eta: 0:00:12 time: 0.1499 data_time: 0.0005 memory: 11240 loss: 0.1453 2023/02/02 20:14:40 - mmengine - INFO - Epoch(train) [50][ 40/102] lr: 2.4359e-05 eta: 0:00:09 time: 0.1498 data_time: 0.0005 memory: 11240 loss: 0.1708 2023/02/02 20:14:43 - mmengine - INFO - Epoch(train) [50][ 60/102] lr: 2.4359e-05 eta: 0:00:06 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.0566 2023/02/02 20:14:46 - mmengine - INFO - Epoch(train) [50][ 80/102] lr: 2.4359e-05 eta: 0:00:03 time: 0.1500 data_time: 0.0005 memory: 11240 loss: 0.1906 2023/02/02 20:14:49 - mmengine - INFO - Epoch(train) [50][100/102] lr: 2.4359e-05 eta: 0:00:00 time: 0.1496 data_time: 0.0004 memory: 11240 loss: 0.0991 2023/02/02 20:14:49 - mmengine - INFO - Exp name: resnet50-mil-arcface_8xb32_inshop_20230202_195247 2023/02/02 20:14:49 - mmengine - INFO - Saving checkpoint at 50 epochs 2023/02/02 20:14:56 - mmengine - INFO - Epoch(val) [50][20/56] eta: 0:00:03 time: 0.0817 data_time: 0.0003 memory: 11240 2023/02/02 20:14:58 - mmengine - INFO - Epoch(val) [50][40/56] eta: 0:00:01 time: 0.0787 data_time: 0.0003 memory: 3822 2023/02/02 20:15:00 - mmengine - INFO - Epoch(val) [50][56/56] retrieval/Recall@1: 90.1815