2021-06-05 18:27:22,954 - mmaction - INFO - Environment info: ------------------------------------------------------------ sys.platform: linux Python: 3.7.6 (default, Jan 8 2020, 19:59:22) [GCC 7.3.0] CUDA available: True GPU 0,1,2,3,4,5,6,7: GeForce GTX 1080 Ti CUDA_HOME: /mnt/lustrenvme/share/polaris/dep/cuda-9.0-cudnn7.6.5 NVCC: Cuda compilation tools, release 9.0, V9.0.176 GCC: gcc (GCC) 5.4.0 PyTorch: 1.5.0 PyTorch compiling details: PyTorch built with: - GCC 5.4 - C++ Version: 201402 - Intel(R) Math Kernel Library Version 2020.0.1 Product Build 20200208 for Intel(R) 64 architecture applications - Intel(R) MKL-DNN v0.21.1 (Git Hash 912ce228837d1ce28e1a61806118835de03f5751) - OpenMP 201307 (a.k.a. OpenMP 4.0) - NNPACK is enabled - CPU capability usage: AVX2 - CUDA Runtime 9.0 - NVCC architecture flags: -gencode;arch=compute_52,code=sm_52;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_70,code=compute_70 - CuDNN 7.6.5 - Magma 2.5.0 - Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_INTERNAL_THREADPOOL_IMPL -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-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=ON, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF, TorchVision: 0.6.0 OpenCV: 4.2.0 MMCV: 1.3.1 MMCV Compiler: GCC 5.4 MMCV CUDA Compiler: 9.0 MMAction2: 0.15.0+985d284 ------------------------------------------------------------ 2021-06-05 18:27:22,955 - mmaction - INFO - Distributed training: True 2021-06-05 18:27:23,954 - mmaction - INFO - Config: model = dict( type='Recognizer2D', backbone=dict( type='ResNetTSM', pretrained='torchvision://resnet50', depth=50, norm_eval=False, shift_div=8, num_segments=8), cls_head=dict( type='TSMHead', num_classes=101, in_channels=2048, spatial_type='avg', consensus=dict(type='AvgConsensus', dim=1), dropout_ratio=0.5, init_std=0.001, is_shift=True, num_segments=8), train_cfg=None, test_cfg=dict(average_clips='prob')) optimizer = dict( type='SGD', constructor='TSMOptimizerConstructor', paramwise_cfg=dict(fc_lr5=True), lr=0.0015, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=dict(max_norm=20, norm_type=2)) lr_config = dict(policy='step', step=[10, 20]) total_epochs = 25 checkpoint_config = dict(interval=1) log_config = dict(interval=20, hooks=[dict(type='TextLoggerHook')]) dist_params = dict(backend='nccl') log_level = 'INFO' load_from = 'https://download.openmmlab.com/mmaction/recognition/tsm/tsm_r50_256p_1x1x8_50e_kinetics400_rgb/tsm_r50_256p_1x1x8_50e_kinetics400_rgb_20200726-020785e2.pth' resume_from = None workflow = [('train', 1)] split = 1 dataset_type = 'RawframeDataset' data_root = 'data/ucf101/rawframes_train' data_root_val = 'data/ucf101/rawframes_val' ann_file_train = 'data/ucf101/ucf101_train_split_1_rawframes.txt' ann_file_val = 'data/ucf101/ucf101_val_split_1_rawframes.txt' ann_file_test = 'data/ucf101/ucf101_val_split_1_rawframes.txt' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False) mc_cfg = dict( server_list_cfg='/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3') train_pipeline = [ dict(type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg='/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict( type='MultiScaleCrop', input_size=224, scales=(1, 0.875, 0.75, 0.66), random_crop=False, max_wh_scale_gap=1, num_fixed_crops=13), dict(type='Resize', scale=(224, 224), keep_ratio=False), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs', 'label']) ] val_pipeline = [ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8, test_mode=True), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg='/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict(type='CenterCrop', crop_size=224), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs']) ] test_pipeline = [ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8, test_mode=True), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg='/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict(type='CenterCrop', crop_size=224), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs']) ] data = dict( videos_per_gpu=12, workers_per_gpu=4, train=dict( type='RawframeDataset', ann_file='data/ucf101/ucf101_train_split_1_rawframes.txt', data_prefix='data/ucf101/rawframes_train', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg= '/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict( type='MultiScaleCrop', input_size=224, scales=(1, 0.875, 0.75, 0.66), random_crop=False, max_wh_scale_gap=1, num_fixed_crops=13), dict(type='Resize', scale=(224, 224), keep_ratio=False), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs', 'label']) ]), val=dict( type='RawframeDataset', ann_file='data/ucf101/ucf101_val_split_1_rawframes.txt', data_prefix='data/ucf101/rawframes_val', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8, test_mode=True), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg= '/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict(type='CenterCrop', crop_size=224), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs']) ]), test=dict( type='RawframeDataset', ann_file='data/ucf101/ucf101_val_split_1_rawframes.txt', data_prefix='data/ucf101/rawframes_val', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=8, test_mode=True), dict( type='RawFrameDecode', io_backend='memcached', server_list_cfg= '/mnt/lustre/share/memcached_client/server_list.conf', client_cfg='/mnt/lustre/share/memcached_client/client.conf', sys_path='/mnt/lustre/share/pymc/py3'), dict(type='Resize', scale=(-1, 256)), dict(type='CenterCrop', crop_size=224), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_bgr=False), dict(type='FormatShape', input_format='NCHW'), dict(type='Collect', keys=['imgs', 'label'], meta_keys=[]), dict(type='ToTensor', keys=['imgs']) ])) evaluation = dict( interval=1, metrics=['top_k_accuracy', 'mean_class_accuracy']) work_dir = './work_dirs/tsm_r50_1x1x8_25e_ucf101_rgb/' gpu_ids = range(0, 8) omnisource = False module_hooks = [] 2021-06-05 18:27:43,099 - mmaction - INFO - These parameters in pretrained checkpoint are not loaded: {'fc.bias', 'fc.weight'} 2021-06-05 18:27:45,345 - mmaction - INFO - load checkpoint from https://download.openmmlab.com/mmaction/recognition/tsm/tsm_r50_256p_1x1x8_50e_kinetics400_rgb/tsm_r50_256p_1x1x8_50e_kinetics400_rgb_20200726-020785e2.pth 2021-06-05 18:27:45,345 - mmaction - INFO - Use load_from_http loader 2021-06-05 18:27:45,876 - mmaction - WARNING - The model and loaded state dict do not match exactly size mismatch for cls_head.fc_cls.weight: copying a param with shape torch.Size([400, 2048]) from checkpoint, the shape in current model is torch.Size([101, 2048]). size mismatch for cls_head.fc_cls.bias: copying a param with shape torch.Size([400]) from checkpoint, the shape in current model is torch.Size([101]). 2021-06-05 18:27:45,880 - mmaction - INFO - Start running, host: linjintao@SH-IDC1-10-5-30-105, work_dir: /mnt/lustre/linjintao/try/mmaction2_dev/work_dirs/tsm_r50_1x1x8_25e_ucf101_rgb 2021-06-05 18:27:45,881 - mmaction - INFO - workflow: [('train', 1)], max: 25 epochs 2021-06-05 18:35:41,819 - mmaction - INFO - Epoch [1][20/100] lr: 1.500e-03, eta: 16:23:30, time: 23.795, data_time: 22.643, memory: 10389, top1_acc: 0.1932, top5_acc: 0.3547, loss_cls: 4.4498, loss: 4.4498, grad_norm: 1.7201 2021-06-05 18:35:57,638 - mmaction - INFO - Epoch [1][40/100] lr: 1.500e-03, eta: 8:24:02, time: 0.793, data_time: 0.004, memory: 10389, top1_acc: 0.3995, top5_acc: 0.6260, loss_cls: 3.8660, loss: 3.8660, grad_norm: 1.9433 2021-06-05 18:36:13,447 - mmaction - INFO - Epoch [1][60/100] lr: 1.500e-03, eta: 5:44:00, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.6135, top5_acc: 0.8812, loss_cls: 3.0746, loss: 3.0746, grad_norm: 2.1506 2021-06-05 18:36:29,284 - mmaction - INFO - Epoch [1][80/100] lr: 1.500e-03, eta: 4:23:52, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.7104, top5_acc: 0.9370, loss_cls: 2.3694, loss: 2.3694, grad_norm: 2.2005 2021-06-05 18:36:44,686 - mmaction - INFO - Epoch [1][100/100] lr: 1.500e-03, eta: 3:35:31, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.7666, top5_acc: 0.9576, loss_cls: 1.8167, loss: 1.8167, grad_norm: 2.1881 2021-06-05 18:36:49,245 - mmaction - INFO - Saving checkpoint at 1 epochs 2021-06-05 18:44:32,973 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 18:44:33,048 - mmaction - INFO - top1_acc 0.8232 top5_acc 0.9794 2021-06-05 18:44:33,048 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 18:44:33,106 - mmaction - INFO - mean_acc 0.8108 2021-06-05 18:44:33,965 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_1.pth. 2021-06-05 18:44:33,965 - mmaction - INFO - Best top1_acc is 0.8232 at 1 epoch. 2021-06-05 18:44:33,968 - mmaction - INFO - Epoch(val) [1][100] top1_acc: 0.8232, top5_acc: 0.9794, mean_class_accuracy: 0.8108 2021-06-05 18:52:19,535 - mmaction - INFO - Epoch [2][20/100] lr: 1.500e-03, eta: 5:31:59, time: 23.276, data_time: 22.364, memory: 10389, top1_acc: 0.8365, top5_acc: 0.9734, loss_cls: 1.4050, loss: 1.4050, grad_norm: 2.1354 2021-06-05 18:52:35,394 - mmaction - INFO - Epoch [2][40/100] lr: 1.500e-03, eta: 4:46:37, time: 0.794, data_time: 0.003, memory: 10389, top1_acc: 0.8562, top5_acc: 0.9839, loss_cls: 1.1204, loss: 1.1204, grad_norm: 2.0173 2021-06-05 18:52:51,235 - mmaction - INFO - Epoch [2][60/100] lr: 1.500e-03, eta: 4:12:32, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.8656, top5_acc: 0.9812, loss_cls: 1.0012, loss: 1.0012, grad_norm: 2.0744 2021-06-05 18:53:07,103 - mmaction - INFO - Epoch [2][80/100] lr: 1.500e-03, eta: 3:45:58, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.8802, top5_acc: 0.9854, loss_cls: 0.8560, loss: 0.8560, grad_norm: 2.0294 2021-06-05 18:53:22,552 - mmaction - INFO - Epoch [2][100/100] lr: 1.500e-03, eta: 3:24:34, time: 0.773, data_time: 0.001, memory: 10389, top1_acc: 0.8847, top5_acc: 0.9823, loss_cls: 0.7740, loss: 0.7740, grad_norm: 1.9904 2021-06-05 18:53:27,119 - mmaction - INFO - Saving checkpoint at 2 epochs 2021-06-05 19:01:00,182 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:01:00,266 - mmaction - INFO - top1_acc 0.8908 top5_acc 0.9921 2021-06-05 19:01:00,266 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:01:00,331 - mmaction - INFO - mean_acc 0.8881 2021-06-05 19:01:01,182 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_2.pth. 2021-06-05 19:01:01,182 - mmaction - INFO - Best top1_acc is 0.8908 at 2 epoch. 2021-06-05 19:01:01,195 - mmaction - INFO - Epoch(val) [2][100] top1_acc: 0.8908, top5_acc: 0.9921, mean_class_accuracy: 0.8881 2021-06-05 19:08:41,924 - mmaction - INFO - Epoch [3][20/100] lr: 1.500e-03, eta: 4:23:56, time: 23.035, data_time: 22.121, memory: 10389, top1_acc: 0.8901, top5_acc: 0.9865, loss_cls: 0.6920, loss: 0.6920, grad_norm: 1.9232 2021-06-05 19:08:57,736 - mmaction - INFO - Epoch [3][40/100] lr: 1.500e-03, eta: 4:02:18, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.8932, top5_acc: 0.9896, loss_cls: 0.6378, loss: 0.6378, grad_norm: 1.9505 2021-06-05 19:09:13,565 - mmaction - INFO - Epoch [3][60/100] lr: 1.500e-03, eta: 3:43:57, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.8865, top5_acc: 0.9844, loss_cls: 0.6234, loss: 0.6234, grad_norm: 1.9279 2021-06-05 19:09:29,421 - mmaction - INFO - Epoch [3][80/100] lr: 1.500e-03, eta: 3:28:12, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9036, top5_acc: 0.9917, loss_cls: 0.5440, loss: 0.5440, grad_norm: 1.8578 2021-06-05 19:09:44,852 - mmaction - INFO - Epoch [3][100/100] lr: 1.500e-03, eta: 3:14:27, time: 0.772, data_time: 0.001, memory: 10389, top1_acc: 0.9131, top5_acc: 0.9887, loss_cls: 0.5325, loss: 0.5325, grad_norm: 1.8891 2021-06-05 19:09:49,707 - mmaction - INFO - Saving checkpoint at 3 epochs 2021-06-05 19:17:30,754 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:17:30,828 - mmaction - INFO - top1_acc 0.9035 top5_acc 0.9915 2021-06-05 19:17:30,829 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:17:30,886 - mmaction - INFO - mean_acc 0.9012 2021-06-05 19:17:31,735 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_3.pth. 2021-06-05 19:17:31,735 - mmaction - INFO - Best top1_acc is 0.9035 at 3 epoch. 2021-06-05 19:17:31,748 - mmaction - INFO - Epoch(val) [3][100] top1_acc: 0.9035, top5_acc: 0.9915, mean_class_accuracy: 0.9012 2021-06-05 19:25:18,323 - mmaction - INFO - Epoch [4][20/100] lr: 1.500e-03, eta: 3:53:36, time: 23.324, data_time: 22.415, memory: 10389, top1_acc: 0.9167, top5_acc: 0.9901, loss_cls: 0.4733, loss: 0.4733, grad_norm: 1.8174 2021-06-05 19:25:34,156 - mmaction - INFO - Epoch [4][40/100] lr: 1.500e-03, eta: 3:39:32, time: 0.794, data_time: 0.004, memory: 10389, top1_acc: 0.9151, top5_acc: 0.9906, loss_cls: 0.4545, loss: 0.4545, grad_norm: 1.7817 2021-06-05 19:25:49,973 - mmaction - INFO - Epoch [4][60/100] lr: 1.500e-03, eta: 3:26:59, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9135, top5_acc: 0.9896, loss_cls: 0.4488, loss: 0.4488, grad_norm: 1.7964 2021-06-05 19:26:05,825 - mmaction - INFO - Epoch [4][80/100] lr: 1.500e-03, eta: 3:15:43, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9250, top5_acc: 0.9932, loss_cls: 0.4065, loss: 0.4065, grad_norm: 1.7435 2021-06-05 19:26:21,243 - mmaction - INFO - Epoch [4][100/100] lr: 1.500e-03, eta: 3:05:32, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9254, top5_acc: 0.9930, loss_cls: 0.4114, loss: 0.4114, grad_norm: 1.8119 2021-06-05 19:26:26,284 - mmaction - INFO - Saving checkpoint at 4 epochs 2021-06-05 19:33:57,040 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:33:57,126 - mmaction - INFO - top1_acc 0.9159 top5_acc 0.9910 2021-06-05 19:33:57,126 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:33:57,191 - mmaction - INFO - mean_acc 0.9155 2021-06-05 19:33:58,046 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_4.pth. 2021-06-05 19:33:58,046 - mmaction - INFO - Best top1_acc is 0.9159 at 4 epoch. 2021-06-05 19:33:58,059 - mmaction - INFO - Epoch(val) [4][100] top1_acc: 0.9159, top5_acc: 0.9910, mean_class_accuracy: 0.9155 2021-06-05 19:41:23,089 - mmaction - INFO - Epoch [5][20/100] lr: 1.500e-03, eta: 3:31:44, time: 22.247, data_time: 21.335, memory: 10389, top1_acc: 0.9396, top5_acc: 0.9922, loss_cls: 0.3734, loss: 0.3734, grad_norm: 1.7601 2021-06-05 19:41:38,897 - mmaction - INFO - Epoch [5][40/100] lr: 1.500e-03, eta: 3:21:24, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9219, top5_acc: 0.9917, loss_cls: 0.3828, loss: 0.3828, grad_norm: 1.7894 2021-06-05 19:41:54,698 - mmaction - INFO - Epoch [5][60/100] lr: 1.500e-03, eta: 3:11:57, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9432, top5_acc: 0.9911, loss_cls: 0.3425, loss: 0.3425, grad_norm: 1.6779 2021-06-05 19:42:10,549 - mmaction - INFO - Epoch [5][80/100] lr: 1.500e-03, eta: 3:03:15, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9432, top5_acc: 0.9958, loss_cls: 0.3345, loss: 0.3345, grad_norm: 1.6731 2021-06-05 19:42:25,924 - mmaction - INFO - Epoch [5][100/100] lr: 1.500e-03, eta: 2:55:13, time: 0.769, data_time: 0.001, memory: 10389, top1_acc: 0.9340, top5_acc: 0.9973, loss_cls: 0.3417, loss: 0.3417, grad_norm: 1.7525 2021-06-05 19:42:30,719 - mmaction - INFO - Saving checkpoint at 5 epochs 2021-06-05 19:49:54,193 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:49:54,279 - mmaction - INFO - top1_acc 0.9284 top5_acc 0.9944 2021-06-05 19:49:54,279 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:49:54,344 - mmaction - INFO - mean_acc 0.9267 2021-06-05 19:49:55,194 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_5.pth. 2021-06-05 19:49:55,194 - mmaction - INFO - Best top1_acc is 0.9284 at 5 epoch. 2021-06-05 19:49:55,207 - mmaction - INFO - Epoch(val) [5][100] top1_acc: 0.9284, top5_acc: 0.9944, mean_class_accuracy: 0.9267 2021-06-05 19:57:10,881 - mmaction - INFO - Epoch [6][20/100] lr: 1.500e-03, eta: 3:14:26, time: 21.782, data_time: 20.899, memory: 10389, top1_acc: 0.9443, top5_acc: 0.9958, loss_cls: 0.3011, loss: 0.3011, grad_norm: 1.6217 2021-06-05 19:57:26,706 - mmaction - INFO - Epoch [6][40/100] lr: 1.500e-03, eta: 3:06:18, time: 0.793, data_time: 0.003, memory: 10389, top1_acc: 0.9505, top5_acc: 0.9938, loss_cls: 0.2841, loss: 0.2841, grad_norm: 1.6110 2021-06-05 19:57:42,524 - mmaction - INFO - Epoch [6][60/100] lr: 1.500e-03, eta: 2:58:43, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9417, top5_acc: 0.9927, loss_cls: 0.3155, loss: 0.3155, grad_norm: 1.7146 2021-06-05 19:57:58,351 - mmaction - INFO - Epoch [6][80/100] lr: 1.500e-03, eta: 2:51:39, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9406, top5_acc: 0.9953, loss_cls: 0.3047, loss: 0.3047, grad_norm: 1.7243 2021-06-05 19:58:13,752 - mmaction - INFO - Epoch [6][100/100] lr: 1.500e-03, eta: 2:45:01, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9555, top5_acc: 0.9968, loss_cls: 0.2758, loss: 0.2758, grad_norm: 1.6643 2021-06-05 19:58:16,893 - mmaction - INFO - Saving checkpoint at 6 epochs 2021-06-05 20:05:42,460 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:05:42,534 - mmaction - INFO - top1_acc 0.9276 top5_acc 0.9944 2021-06-05 20:05:42,534 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:05:42,599 - mmaction - INFO - mean_acc 0.9272 2021-06-05 20:05:42,600 - mmaction - INFO - Epoch(val) [6][100] top1_acc: 0.9276, top5_acc: 0.9944, mean_class_accuracy: 0.9272 2021-06-05 20:13:30,828 - mmaction - INFO - Epoch [7][20/100] lr: 1.500e-03, eta: 3:01:40, time: 23.409, data_time: 22.498, memory: 10389, top1_acc: 0.9536, top5_acc: 0.9943, loss_cls: 0.2647, loss: 0.2647, grad_norm: 1.6176 2021-06-05 20:13:46,644 - mmaction - INFO - Epoch [7][40/100] lr: 1.500e-03, eta: 2:54:53, time: 0.792, data_time: 0.004, memory: 10389, top1_acc: 0.9484, top5_acc: 0.9938, loss_cls: 0.2770, loss: 0.2770, grad_norm: 1.7111 2021-06-05 20:14:02,485 - mmaction - INFO - Epoch [7][60/100] lr: 1.500e-03, eta: 2:48:30, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9578, top5_acc: 0.9964, loss_cls: 0.2436, loss: 0.2436, grad_norm: 1.5756 2021-06-05 20:14:18,325 - mmaction - INFO - Epoch [7][80/100] lr: 1.500e-03, eta: 2:42:29, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9521, top5_acc: 0.9974, loss_cls: 0.2516, loss: 0.2516, grad_norm: 1.6974 2021-06-05 20:14:33,739 - mmaction - INFO - Epoch [7][100/100] lr: 1.500e-03, eta: 2:36:46, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9582, top5_acc: 0.9930, loss_cls: 0.2546, loss: 0.2546, grad_norm: 1.6796 2021-06-05 20:14:38,632 - mmaction - INFO - Saving checkpoint at 7 epochs 2021-06-05 20:22:11,738 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:22:11,822 - mmaction - INFO - top1_acc 0.9276 top5_acc 0.9950 2021-06-05 20:22:11,822 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:22:11,887 - mmaction - INFO - mean_acc 0.9277 2021-06-05 20:22:11,889 - mmaction - INFO - Epoch(val) [7][100] top1_acc: 0.9276, top5_acc: 0.9950, mean_class_accuracy: 0.9277 2021-06-05 20:29:37,522 - mmaction - INFO - Epoch [8][20/100] lr: 1.500e-03, eta: 2:49:04, time: 22.280, data_time: 21.373, memory: 10389, top1_acc: 0.9568, top5_acc: 0.9958, loss_cls: 0.2394, loss: 0.2394, grad_norm: 1.6138 2021-06-05 20:29:53,331 - mmaction - INFO - Epoch [8][40/100] lr: 1.500e-03, eta: 2:43:17, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9510, top5_acc: 0.9969, loss_cls: 0.2314, loss: 0.2314, grad_norm: 1.6462 2021-06-05 20:30:09,145 - mmaction - INFO - Epoch [8][60/100] lr: 1.500e-03, eta: 2:37:47, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9589, top5_acc: 0.9979, loss_cls: 0.2198, loss: 0.2198, grad_norm: 1.5943 2021-06-05 20:30:24,982 - mmaction - INFO - Epoch [8][80/100] lr: 1.500e-03, eta: 2:32:33, time: 0.792, data_time: 0.002, memory: 10389, top1_acc: 0.9656, top5_acc: 0.9964, loss_cls: 0.2059, loss: 0.2059, grad_norm: 1.5187 2021-06-05 20:30:40,396 - mmaction - INFO - Epoch [8][100/100] lr: 1.500e-03, eta: 2:27:33, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9587, top5_acc: 0.9968, loss_cls: 0.2104, loss: 0.2104, grad_norm: 1.6256 2021-06-05 20:30:45,677 - mmaction - INFO - Saving checkpoint at 8 epochs 2021-06-05 20:38:05,140 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:38:05,209 - mmaction - INFO - top1_acc 0.9355 top5_acc 0.9960 2021-06-05 20:38:05,210 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:38:05,263 - mmaction - INFO - mean_acc 0.9348 2021-06-05 20:38:06,108 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_8.pth. 2021-06-05 20:38:06,108 - mmaction - INFO - Best top1_acc is 0.9355 at 8 epoch. 2021-06-05 20:38:06,121 - mmaction - INFO - Epoch(val) [8][100] top1_acc: 0.9355, top5_acc: 0.9960, mean_class_accuracy: 0.9348 2021-06-05 20:45:20,486 - mmaction - INFO - Epoch [9][20/100] lr: 1.500e-03, eta: 2:37:05, time: 21.716, data_time: 20.840, memory: 10389, top1_acc: 0.9594, top5_acc: 0.9969, loss_cls: 0.2069, loss: 0.2069, grad_norm: 1.5656 2021-06-05 20:45:36,310 - mmaction - INFO - Epoch [9][40/100] lr: 1.500e-03, eta: 2:32:03, time: 0.793, data_time: 0.003, memory: 10389, top1_acc: 0.9682, top5_acc: 0.9995, loss_cls: 0.1921, loss: 0.1921, grad_norm: 1.4721 2021-06-05 20:45:52,117 - mmaction - INFO - Epoch [9][60/100] lr: 1.500e-03, eta: 2:27:13, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9589, top5_acc: 0.9974, loss_cls: 0.2125, loss: 0.2125, grad_norm: 1.6482 2021-06-05 20:46:07,975 - mmaction - INFO - Epoch [9][80/100] lr: 1.500e-03, eta: 2:22:37, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9656, top5_acc: 0.9964, loss_cls: 0.1901, loss: 0.1901, grad_norm: 1.5575 2021-06-05 20:46:23,389 - mmaction - INFO - Epoch [9][100/100] lr: 1.500e-03, eta: 2:18:10, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9662, top5_acc: 0.9957, loss_cls: 0.1872, loss: 0.1872, grad_norm: 1.5274 2021-06-05 20:46:26,365 - mmaction - INFO - Saving checkpoint at 9 epochs 2021-06-05 20:53:56,239 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:53:56,314 - mmaction - INFO - top1_acc 0.9411 top5_acc 0.9955 2021-06-05 20:53:56,314 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:53:56,375 - mmaction - INFO - mean_acc 0.9405 2021-06-05 20:53:57,233 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_9.pth. 2021-06-05 20:53:57,233 - mmaction - INFO - Best top1_acc is 0.9411 at 9 epoch. 2021-06-05 20:53:57,235 - mmaction - INFO - Epoch(val) [9][100] top1_acc: 0.9411, top5_acc: 0.9955, mean_class_accuracy: 0.9405 2021-06-05 21:01:22,344 - mmaction - INFO - Epoch [10][20/100] lr: 1.500e-03, eta: 2:26:13, time: 22.251, data_time: 21.337, memory: 10389, top1_acc: 0.9672, top5_acc: 0.9990, loss_cls: 0.1848, loss: 0.1848, grad_norm: 1.4991 2021-06-05 21:01:38,140 - mmaction - INFO - Epoch [10][40/100] lr: 1.500e-03, eta: 2:21:44, time: 0.791, data_time: 0.003, memory: 10389, top1_acc: 0.9693, top5_acc: 0.9979, loss_cls: 0.1788, loss: 0.1788, grad_norm: 1.5081 2021-06-05 21:01:53,956 - mmaction - INFO - Epoch [10][60/100] lr: 1.500e-03, eta: 2:17:25, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9594, top5_acc: 0.9974, loss_cls: 0.1957, loss: 0.1957, grad_norm: 1.5868 2021-06-05 21:02:09,786 - mmaction - INFO - Epoch [10][80/100] lr: 1.500e-03, eta: 2:13:17, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9693, top5_acc: 0.9984, loss_cls: 0.1675, loss: 0.1675, grad_norm: 1.5272 2021-06-05 21:02:25,161 - mmaction - INFO - Epoch [10][100/100] lr: 1.500e-03, eta: 2:09:17, time: 0.769, data_time: 0.001, memory: 10389, top1_acc: 0.9700, top5_acc: 0.9968, loss_cls: 0.1753, loss: 0.1753, grad_norm: 1.5205 2021-06-05 21:02:29,938 - mmaction - INFO - Saving checkpoint at 10 epochs 2021-06-05 21:10:15,215 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:10:15,291 - mmaction - INFO - top1_acc 0.9421 top5_acc 0.9947 2021-06-05 21:10:15,291 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:10:15,352 - mmaction - INFO - mean_acc 0.9405 2021-06-05 21:10:16,274 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_10.pth. 2021-06-05 21:10:16,274 - mmaction - INFO - Best top1_acc is 0.9421 at 10 epoch. 2021-06-05 21:10:16,287 - mmaction - INFO - Epoch(val) [10][100] top1_acc: 0.9421, top5_acc: 0.9947, mean_class_accuracy: 0.9405 2021-06-05 21:17:49,617 - mmaction - INFO - Epoch [11][20/100] lr: 1.500e-04, eta: 2:16:01, time: 22.661, data_time: 21.746, memory: 10389, top1_acc: 0.9688, top5_acc: 0.9964, loss_cls: 0.1759, loss: 0.1759, grad_norm: 1.5475 2021-06-05 21:18:05,406 - mmaction - INFO - Epoch [11][40/100] lr: 1.500e-04, eta: 2:11:58, time: 0.792, data_time: 0.004, memory: 10389, top1_acc: 0.9677, top5_acc: 0.9979, loss_cls: 0.1638, loss: 0.1638, grad_norm: 1.4196 2021-06-05 21:18:21,219 - mmaction - INFO - Epoch [11][60/100] lr: 1.500e-04, eta: 2:08:04, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9708, top5_acc: 0.9958, loss_cls: 0.1648, loss: 0.1648, grad_norm: 1.4175 2021-06-05 21:18:37,022 - mmaction - INFO - Epoch [11][80/100] lr: 1.500e-04, eta: 2:04:17, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9786, top5_acc: 0.9984, loss_cls: 0.1494, loss: 0.1494, grad_norm: 1.3953 2021-06-05 21:18:52,429 - mmaction - INFO - Epoch [11][100/100] lr: 1.500e-04, eta: 2:00:38, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9726, top5_acc: 0.9989, loss_cls: 0.1536, loss: 0.1536, grad_norm: 1.4121 2021-06-05 21:18:57,281 - mmaction - INFO - Saving checkpoint at 11 epochs 2021-06-05 21:26:29,833 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:26:29,909 - mmaction - INFO - top1_acc 0.9411 top5_acc 0.9955 2021-06-05 21:26:29,909 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:26:29,968 - mmaction - INFO - mean_acc 0.9399 2021-06-05 21:26:29,969 - mmaction - INFO - Epoch(val) [11][100] top1_acc: 0.9411, top5_acc: 0.9955, mean_class_accuracy: 0.9399 2021-06-05 21:34:04,798 - mmaction - INFO - Epoch [12][20/100] lr: 1.500e-04, eta: 2:06:08, time: 22.740, data_time: 21.831, memory: 10389, top1_acc: 0.9750, top5_acc: 0.9958, loss_cls: 0.1563, loss: 0.1563, grad_norm: 1.4125 2021-06-05 21:34:20,616 - mmaction - INFO - Epoch [12][40/100] lr: 1.500e-04, eta: 2:02:26, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9656, top5_acc: 0.9953, loss_cls: 0.1711, loss: 0.1711, grad_norm: 1.4938 2021-06-05 21:34:36,429 - mmaction - INFO - Epoch [12][60/100] lr: 1.500e-04, eta: 1:58:52, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9969, loss_cls: 0.1506, loss: 0.1506, grad_norm: 1.3813 2021-06-05 21:34:52,274 - mmaction - INFO - Epoch [12][80/100] lr: 1.500e-04, eta: 1:55:24, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9661, top5_acc: 0.9984, loss_cls: 0.1753, loss: 0.1753, grad_norm: 1.4779 2021-06-05 21:35:07,661 - mmaction - INFO - Epoch [12][100/100] lr: 1.500e-04, eta: 1:52:02, time: 0.769, data_time: 0.001, memory: 10389, top1_acc: 0.9726, top5_acc: 0.9989, loss_cls: 0.1627, loss: 0.1627, grad_norm: 1.4687 2021-06-05 21:35:12,684 - mmaction - INFO - Saving checkpoint at 12 epochs 2021-06-05 21:42:29,982 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:42:30,059 - mmaction - INFO - top1_acc 0.9442 top5_acc 0.9947 2021-06-05 21:42:30,059 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:42:30,117 - mmaction - INFO - mean_acc 0.9430 2021-06-05 21:42:30,975 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_12.pth. 2021-06-05 21:42:30,975 - mmaction - INFO - Best top1_acc is 0.9442 at 12 epoch. 2021-06-05 21:42:30,977 - mmaction - INFO - Epoch(val) [12][100] top1_acc: 0.9442, top5_acc: 0.9947, mean_class_accuracy: 0.9430 2021-06-05 21:49:45,835 - mmaction - INFO - Epoch [13][20/100] lr: 1.500e-04, eta: 1:56:06, time: 21.741, data_time: 20.870, memory: 10389, top1_acc: 0.9698, top5_acc: 0.9958, loss_cls: 0.1742, loss: 0.1742, grad_norm: 1.5187 2021-06-05 21:50:01,651 - mmaction - INFO - Epoch [13][40/100] lr: 1.500e-04, eta: 1:52:43, time: 0.793, data_time: 0.004, memory: 10389, top1_acc: 0.9776, top5_acc: 0.9979, loss_cls: 0.1467, loss: 0.1467, grad_norm: 1.3702 2021-06-05 21:50:17,487 - mmaction - INFO - Epoch [13][60/100] lr: 1.500e-04, eta: 1:49:25, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9750, top5_acc: 0.9969, loss_cls: 0.1610, loss: 0.1610, grad_norm: 1.4720 2021-06-05 21:50:33,333 - mmaction - INFO - Epoch [13][80/100] lr: 1.500e-04, eta: 1:46:14, time: 0.793, data_time: 0.002, memory: 10389, top1_acc: 0.9693, top5_acc: 0.9974, loss_cls: 0.1614, loss: 0.1614, grad_norm: 1.4220 2021-06-05 21:50:48,729 - mmaction - INFO - Epoch [13][100/100] lr: 1.500e-04, eta: 1:43:07, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9694, top5_acc: 0.9979, loss_cls: 0.1677, loss: 0.1677, grad_norm: 1.5037 2021-06-05 21:50:51,974 - mmaction - INFO - Saving checkpoint at 13 epochs 2021-06-05 21:58:28,407 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:58:28,478 - mmaction - INFO - top1_acc 0.9437 top5_acc 0.9955 2021-06-05 21:58:28,479 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:58:28,533 - mmaction - INFO - mean_acc 0.9426 2021-06-05 21:58:28,534 - mmaction - INFO - Epoch(val) [13][100] top1_acc: 0.9437, top5_acc: 0.9955, mean_class_accuracy: 0.9426 2021-06-05 22:06:09,505 - mmaction - INFO - Epoch [14][20/100] lr: 1.500e-04, eta: 1:46:44, time: 23.047, data_time: 22.131, memory: 10389, top1_acc: 0.9740, top5_acc: 0.9979, loss_cls: 0.1436, loss: 0.1436, grad_norm: 1.3309 2021-06-05 22:06:25,296 - mmaction - INFO - Epoch [14][40/100] lr: 1.500e-04, eta: 1:43:35, time: 0.790, data_time: 0.003, memory: 10389, top1_acc: 0.9729, top5_acc: 0.9969, loss_cls: 0.1653, loss: 0.1653, grad_norm: 1.4224 2021-06-05 22:06:41,105 - mmaction - INFO - Epoch [14][60/100] lr: 1.500e-04, eta: 1:40:31, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9755, top5_acc: 0.9979, loss_cls: 0.1571, loss: 0.1571, grad_norm: 1.4233 2021-06-05 22:06:56,922 - mmaction - INFO - Epoch [14][80/100] lr: 1.500e-04, eta: 1:37:32, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9693, top5_acc: 0.9990, loss_cls: 0.1602, loss: 0.1602, grad_norm: 1.4589 2021-06-05 22:07:12,334 - mmaction - INFO - Epoch [14][100/100] lr: 1.500e-04, eta: 1:34:38, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9764, top5_acc: 0.9968, loss_cls: 0.1546, loss: 0.1546, grad_norm: 1.4300 2021-06-05 22:07:16,937 - mmaction - INFO - Saving checkpoint at 14 epochs 2021-06-05 22:14:54,672 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:14:54,759 - mmaction - INFO - top1_acc 0.9418 top5_acc 0.9952 2021-06-05 22:14:54,760 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:14:54,824 - mmaction - INFO - mean_acc 0.9407 2021-06-05 22:14:54,826 - mmaction - INFO - Epoch(val) [14][100] top1_acc: 0.9418, top5_acc: 0.9952, mean_class_accuracy: 0.9407 2021-06-05 22:22:03,789 - mmaction - INFO - Epoch [15][20/100] lr: 1.500e-04, eta: 1:37:02, time: 21.446, data_time: 20.570, memory: 10389, top1_acc: 0.9708, top5_acc: 0.9953, loss_cls: 0.1686, loss: 0.1686, grad_norm: 1.4955 2021-06-05 22:22:19,581 - mmaction - INFO - Epoch [15][40/100] lr: 1.500e-04, eta: 1:34:07, time: 0.791, data_time: 0.003, memory: 10389, top1_acc: 0.9724, top5_acc: 0.9995, loss_cls: 0.1523, loss: 0.1523, grad_norm: 1.4495 2021-06-05 22:22:35,390 - mmaction - INFO - Epoch [15][60/100] lr: 1.500e-04, eta: 1:31:15, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9672, top5_acc: 0.9974, loss_cls: 0.1685, loss: 0.1685, grad_norm: 1.5337 2021-06-05 22:22:51,235 - mmaction - INFO - Epoch [15][80/100] lr: 1.500e-04, eta: 1:28:28, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9667, top5_acc: 0.9969, loss_cls: 0.1660, loss: 0.1660, grad_norm: 1.4917 2021-06-05 22:23:06,649 - mmaction - INFO - Epoch [15][100/100] lr: 1.500e-04, eta: 1:25:45, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9812, top5_acc: 0.9984, loss_cls: 0.1422, loss: 0.1422, grad_norm: 1.3223 2021-06-05 22:23:09,654 - mmaction - INFO - Saving checkpoint at 15 epochs 2021-06-05 22:30:36,057 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:30:36,127 - mmaction - INFO - top1_acc 0.9405 top5_acc 0.9955 2021-06-05 22:30:36,127 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:30:36,181 - mmaction - INFO - mean_acc 0.9400 2021-06-05 22:30:36,182 - mmaction - INFO - Epoch(val) [15][100] top1_acc: 0.9405, top5_acc: 0.9955, mean_class_accuracy: 0.9400 2021-06-05 22:38:11,419 - mmaction - INFO - Epoch [16][20/100] lr: 1.500e-04, eta: 1:27:49, time: 22.759, data_time: 21.852, memory: 10389, top1_acc: 0.9755, top5_acc: 0.9979, loss_cls: 0.1490, loss: 0.1490, grad_norm: 1.4078 2021-06-05 22:38:27,217 - mmaction - INFO - Epoch [16][40/100] lr: 1.500e-04, eta: 1:25:05, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9656, top5_acc: 0.9969, loss_cls: 0.1724, loss: 0.1724, grad_norm: 1.5723 2021-06-05 22:38:43,024 - mmaction - INFO - Epoch [16][60/100] lr: 1.500e-04, eta: 1:22:24, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9708, top5_acc: 0.9984, loss_cls: 0.1506, loss: 0.1506, grad_norm: 1.4113 2021-06-05 22:38:58,857 - mmaction - INFO - Epoch [16][80/100] lr: 1.500e-04, eta: 1:19:47, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9688, top5_acc: 0.9958, loss_cls: 0.1596, loss: 0.1596, grad_norm: 1.4809 2021-06-05 22:39:14,252 - mmaction - INFO - Epoch [16][100/100] lr: 1.500e-04, eta: 1:17:13, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9726, top5_acc: 0.9968, loss_cls: 0.1502, loss: 0.1502, grad_norm: 1.3877 2021-06-05 22:39:18,820 - mmaction - INFO - Saving checkpoint at 16 epochs 2021-06-05 22:46:53,570 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:46:53,640 - mmaction - INFO - top1_acc 0.9442 top5_acc 0.9942 2021-06-05 22:46:53,640 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:46:53,693 - mmaction - INFO - mean_acc 0.9432 2021-06-05 22:46:53,694 - mmaction - INFO - Epoch(val) [16][100] top1_acc: 0.9442, top5_acc: 0.9942, mean_class_accuracy: 0.9432 2021-06-05 22:54:34,581 - mmaction - INFO - Epoch [17][20/100] lr: 1.500e-04, eta: 1:18:44, time: 23.042, data_time: 22.132, memory: 10389, top1_acc: 0.9750, top5_acc: 0.9990, loss_cls: 0.1506, loss: 0.1506, grad_norm: 1.4183 2021-06-05 22:54:50,425 - mmaction - INFO - Epoch [17][40/100] lr: 1.500e-04, eta: 1:16:09, time: 0.793, data_time: 0.003, memory: 10389, top1_acc: 0.9708, top5_acc: 0.9974, loss_cls: 0.1583, loss: 0.1583, grad_norm: 1.5029 2021-06-05 22:55:06,324 - mmaction - INFO - Epoch [17][60/100] lr: 1.500e-04, eta: 1:13:37, time: 0.794, data_time: 0.001, memory: 10389, top1_acc: 0.9740, top5_acc: 0.9990, loss_cls: 0.1475, loss: 0.1475, grad_norm: 1.4611 2021-06-05 22:55:22,183 - mmaction - INFO - Epoch [17][80/100] lr: 1.500e-04, eta: 1:11:08, time: 0.793, data_time: 0.002, memory: 10389, top1_acc: 0.9708, top5_acc: 0.9958, loss_cls: 0.1495, loss: 0.1495, grad_norm: 1.4188 2021-06-05 22:55:37,596 - mmaction - INFO - Epoch [17][100/100] lr: 1.500e-04, eta: 1:08:42, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9716, top5_acc: 0.9989, loss_cls: 0.1570, loss: 0.1570, grad_norm: 1.4801 2021-06-05 22:55:42,578 - mmaction - INFO - Saving checkpoint at 17 epochs 2021-06-05 23:03:22,025 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:03:22,100 - mmaction - INFO - top1_acc 0.9397 top5_acc 0.9952 2021-06-05 23:03:22,100 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:03:22,161 - mmaction - INFO - mean_acc 0.9392 2021-06-05 23:03:22,162 - mmaction - INFO - Epoch(val) [17][100] top1_acc: 0.9397, top5_acc: 0.9952, mean_class_accuracy: 0.9392 2021-06-05 23:10:58,998 - mmaction - INFO - Epoch [18][20/100] lr: 1.500e-04, eta: 1:09:39, time: 22.840, data_time: 21.921, memory: 10389, top1_acc: 0.9771, top5_acc: 0.9984, loss_cls: 0.1463, loss: 0.1463, grad_norm: 1.4204 2021-06-05 23:11:14,791 - mmaction - INFO - Epoch [18][40/100] lr: 1.500e-04, eta: 1:07:12, time: 0.791, data_time: 0.003, memory: 10389, top1_acc: 0.9812, top5_acc: 0.9974, loss_cls: 0.1426, loss: 0.1426, grad_norm: 1.3284 2021-06-05 23:11:30,595 - mmaction - INFO - Epoch [18][60/100] lr: 1.500e-04, eta: 1:04:48, time: 0.790, data_time: 0.001, memory: 10389, top1_acc: 0.9771, top5_acc: 0.9990, loss_cls: 0.1489, loss: 0.1489, grad_norm: 1.4403 2021-06-05 23:11:46,458 - mmaction - INFO - Epoch [18][80/100] lr: 1.500e-04, eta: 1:02:27, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9729, top5_acc: 0.9979, loss_cls: 0.1573, loss: 0.1573, grad_norm: 1.4641 2021-06-05 23:12:01,853 - mmaction - INFO - Epoch [18][100/100] lr: 1.500e-04, eta: 1:00:08, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9646, top5_acc: 0.9957, loss_cls: 0.1732, loss: 0.1732, grad_norm: 1.5862 2021-06-05 23:12:06,622 - mmaction - INFO - Saving checkpoint at 18 epochs 2021-06-05 23:19:39,949 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:19:40,028 - mmaction - INFO - top1_acc 0.9453 top5_acc 0.9955 2021-06-05 23:19:40,029 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:19:40,088 - mmaction - INFO - mean_acc 0.9448 2021-06-05 23:19:40,946 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_18.pth. 2021-06-05 23:19:40,946 - mmaction - INFO - Best top1_acc is 0.9453 at 18 epoch. 2021-06-05 23:19:40,948 - mmaction - INFO - Epoch(val) [18][100] top1_acc: 0.9453, top5_acc: 0.9955, mean_class_accuracy: 0.9448 2021-06-05 23:27:25,972 - mmaction - INFO - Epoch [19][20/100] lr: 1.500e-04, eta: 1:00:41, time: 23.249, data_time: 22.341, memory: 10389, top1_acc: 0.9719, top5_acc: 0.9984, loss_cls: 0.1507, loss: 0.1507, grad_norm: 1.4224 2021-06-05 23:27:41,775 - mmaction - INFO - Epoch [19][40/100] lr: 1.500e-04, eta: 0:58:21, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9729, top5_acc: 0.9964, loss_cls: 0.1548, loss: 0.1548, grad_norm: 1.4015 2021-06-05 23:27:57,595 - mmaction - INFO - Epoch [19][60/100] lr: 1.500e-04, eta: 0:56:04, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9672, top5_acc: 0.9953, loss_cls: 0.1639, loss: 0.1639, grad_norm: 1.4710 2021-06-05 23:28:13,431 - mmaction - INFO - Epoch [19][80/100] lr: 1.500e-04, eta: 0:53:49, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9990, loss_cls: 0.1390, loss: 0.1390, grad_norm: 1.4109 2021-06-05 23:28:28,855 - mmaction - INFO - Epoch [19][100/100] lr: 1.500e-04, eta: 0:51:37, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9657, top5_acc: 0.9973, loss_cls: 0.1716, loss: 0.1716, grad_norm: 1.6065 2021-06-05 23:28:33,731 - mmaction - INFO - Saving checkpoint at 19 epochs 2021-06-05 23:36:12,097 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:36:12,166 - mmaction - INFO - top1_acc 0.9426 top5_acc 0.9955 2021-06-05 23:36:12,167 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:36:12,219 - mmaction - INFO - mean_acc 0.9419 2021-06-05 23:36:12,221 - mmaction - INFO - Epoch(val) [19][100] top1_acc: 0.9426, top5_acc: 0.9955, mean_class_accuracy: 0.9419 2021-06-05 23:43:56,233 - mmaction - INFO - Epoch [20][20/100] lr: 1.500e-04, eta: 0:51:43, time: 23.199, data_time: 22.301, memory: 10389, top1_acc: 0.9729, top5_acc: 0.9979, loss_cls: 0.1457, loss: 0.1457, grad_norm: 1.4200 2021-06-05 23:44:12,041 - mmaction - INFO - Epoch [20][40/100] lr: 1.500e-04, eta: 0:49:29, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9734, top5_acc: 0.9979, loss_cls: 0.1507, loss: 0.1507, grad_norm: 1.4053 2021-06-05 23:44:27,853 - mmaction - INFO - Epoch [20][60/100] lr: 1.500e-04, eta: 0:47:18, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9724, top5_acc: 0.9984, loss_cls: 0.1377, loss: 0.1377, grad_norm: 1.3904 2021-06-05 23:44:43,692 - mmaction - INFO - Epoch [20][80/100] lr: 1.500e-04, eta: 0:45:10, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9812, top5_acc: 0.9979, loss_cls: 0.1347, loss: 0.1347, grad_norm: 1.3480 2021-06-05 23:44:59,119 - mmaction - INFO - Epoch [20][100/100] lr: 1.500e-04, eta: 0:43:03, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9700, top5_acc: 0.9984, loss_cls: 0.1562, loss: 0.1562, grad_norm: 1.4678 2021-06-05 23:45:03,929 - mmaction - INFO - Saving checkpoint at 20 epochs 2021-06-05 23:52:37,272 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:52:37,342 - mmaction - INFO - top1_acc 0.9450 top5_acc 0.9955 2021-06-05 23:52:37,342 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:52:37,395 - mmaction - INFO - mean_acc 0.9439 2021-06-05 23:52:37,396 - mmaction - INFO - Epoch(val) [20][100] top1_acc: 0.9450, top5_acc: 0.9955, mean_class_accuracy: 0.9439 2021-06-06 00:00:25,649 - mmaction - INFO - Epoch [21][20/100] lr: 1.500e-05, eta: 0:42:47, time: 23.406, data_time: 22.134, memory: 10389, top1_acc: 0.9724, top5_acc: 0.9995, loss_cls: 0.1540, loss: 0.1540, grad_norm: 1.5375 2021-06-06 00:00:41,505 - mmaction - INFO - Epoch [21][40/100] lr: 1.500e-05, eta: 0:40:39, time: 0.799, data_time: 0.011, memory: 10389, top1_acc: 0.9714, top5_acc: 0.9974, loss_cls: 0.1533, loss: 0.1533, grad_norm: 1.4851 2021-06-06 00:00:57,322 - mmaction - INFO - Epoch [21][60/100] lr: 1.500e-05, eta: 0:38:34, time: 0.791, data_time: 0.001, memory: 10389, top1_acc: 0.9703, top5_acc: 0.9974, loss_cls: 0.1561, loss: 0.1561, grad_norm: 1.4602 2021-06-06 00:01:13,167 - mmaction - INFO - Epoch [21][80/100] lr: 1.500e-05, eta: 0:36:31, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9786, top5_acc: 0.9990, loss_cls: 0.1433, loss: 0.1433, grad_norm: 1.4507 2021-06-06 00:01:28,567 - mmaction - INFO - Epoch [21][100/100] lr: 1.500e-05, eta: 0:34:29, time: 0.770, data_time: 0.001, memory: 10389, top1_acc: 0.9748, top5_acc: 0.9973, loss_cls: 0.1415, loss: 0.1415, grad_norm: 1.4262 2021-06-06 00:01:34,913 - mmaction - INFO - Saving checkpoint at 21 epochs 2021-06-06 00:09:07,461 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:09:07,534 - mmaction - INFO - top1_acc 0.9434 top5_acc 0.9960 2021-06-06 00:09:07,534 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:09:07,589 - mmaction - INFO - mean_acc 0.9422 2021-06-06 00:09:07,591 - mmaction - INFO - Epoch(val) [21][100] top1_acc: 0.9434, top5_acc: 0.9960, mean_class_accuracy: 0.9422 2021-06-06 00:16:49,510 - mmaction - INFO - Epoch [22][20/100] lr: 1.500e-05, eta: 0:33:50, time: 23.094, data_time: 22.181, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9990, loss_cls: 0.1475, loss: 0.1475, grad_norm: 1.4127 2021-06-06 00:17:05,330 - mmaction - INFO - Epoch [22][40/100] lr: 1.500e-05, eta: 0:31:48, time: 0.793, data_time: 0.003, memory: 10389, top1_acc: 0.9693, top5_acc: 0.9974, loss_cls: 0.1523, loss: 0.1523, grad_norm: 1.4670 2021-06-06 00:17:21,182 - mmaction - INFO - Epoch [22][60/100] lr: 1.500e-05, eta: 0:29:48, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9750, top5_acc: 0.9979, loss_cls: 0.1456, loss: 0.1456, grad_norm: 1.4414 2021-06-06 00:17:37,031 - mmaction - INFO - Epoch [22][80/100] lr: 1.500e-05, eta: 0:27:49, time: 0.793, data_time: 0.001, memory: 10389, top1_acc: 0.9698, top5_acc: 0.9984, loss_cls: 0.1547, loss: 0.1547, grad_norm: 1.4548 2021-06-06 00:17:52,462 - mmaction - INFO - Epoch [22][100/100] lr: 1.500e-05, eta: 0:25:53, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9726, top5_acc: 1.0000, loss_cls: 0.1418, loss: 0.1418, grad_norm: 1.4051 2021-06-06 00:17:56,948 - mmaction - INFO - Saving checkpoint at 22 epochs 2021-06-06 00:25:39,011 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:25:39,088 - mmaction - INFO - top1_acc 0.9463 top5_acc 0.9963 2021-06-06 00:25:39,088 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:25:39,148 - mmaction - INFO - mean_acc 0.9455 2021-06-06 00:25:40,085 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_22.pth. 2021-06-06 00:25:40,085 - mmaction - INFO - Best top1_acc is 0.9463 at 22 epoch. 2021-06-06 00:25:40,099 - mmaction - INFO - Epoch(val) [22][100] top1_acc: 0.9463, top5_acc: 0.9963, mean_class_accuracy: 0.9455 2021-06-06 00:33:20,121 - mmaction - INFO - Epoch [23][20/100] lr: 1.500e-05, eta: 0:24:54, time: 22.996, data_time: 22.075, memory: 10389, top1_acc: 0.9771, top5_acc: 0.9974, loss_cls: 0.1462, loss: 0.1462, grad_norm: 1.4315 2021-06-06 00:33:35,944 - mmaction - INFO - Epoch [23][40/100] lr: 1.500e-05, eta: 0:22:57, time: 0.794, data_time: 0.004, memory: 10389, top1_acc: 0.9724, top5_acc: 0.9984, loss_cls: 0.1478, loss: 0.1478, grad_norm: 1.4690 2021-06-06 00:33:51,777 - mmaction - INFO - Epoch [23][60/100] lr: 1.500e-05, eta: 0:21:01, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9953, loss_cls: 0.1492, loss: 0.1492, grad_norm: 1.3848 2021-06-06 00:34:07,623 - mmaction - INFO - Epoch [23][80/100] lr: 1.500e-05, eta: 0:19:08, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9979, loss_cls: 0.1484, loss: 0.1484, grad_norm: 1.4357 2021-06-06 00:34:23,041 - mmaction - INFO - Epoch [23][100/100] lr: 1.500e-05, eta: 0:17:15, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9710, top5_acc: 0.9962, loss_cls: 0.1543, loss: 0.1543, grad_norm: 1.4457 2021-06-06 00:34:27,606 - mmaction - INFO - Saving checkpoint at 23 epochs 2021-06-06 00:42:00,700 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:42:00,782 - mmaction - INFO - top1_acc 0.9416 top5_acc 0.9963 2021-06-06 00:42:00,782 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:42:00,842 - mmaction - INFO - mean_acc 0.9405 2021-06-06 00:42:00,843 - mmaction - INFO - Epoch(val) [23][100] top1_acc: 0.9416, top5_acc: 0.9963, mean_class_accuracy: 0.9405 2021-06-06 00:49:36,293 - mmaction - INFO - Epoch [24][20/100] lr: 1.500e-05, eta: 0:15:59, time: 22.771, data_time: 21.859, memory: 10389, top1_acc: 0.9656, top5_acc: 0.9969, loss_cls: 0.1534, loss: 0.1534, grad_norm: 1.4732 2021-06-06 00:49:52,110 - mmaction - INFO - Epoch [24][40/100] lr: 1.500e-05, eta: 0:14:06, time: 0.792, data_time: 0.003, memory: 10389, top1_acc: 0.9677, top5_acc: 0.9953, loss_cls: 0.1681, loss: 0.1681, grad_norm: 1.4704 2021-06-06 00:50:07,960 - mmaction - INFO - Epoch [24][60/100] lr: 1.500e-05, eta: 0:12:15, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9781, top5_acc: 0.9969, loss_cls: 0.1433, loss: 0.1433, grad_norm: 1.4138 2021-06-06 00:50:23,806 - mmaction - INFO - Epoch [24][80/100] lr: 1.500e-05, eta: 0:10:26, time: 0.793, data_time: 0.002, memory: 10389, top1_acc: 0.9729, top5_acc: 0.9964, loss_cls: 0.1567, loss: 0.1567, grad_norm: 1.4372 2021-06-06 00:50:39,218 - mmaction - INFO - Epoch [24][100/100] lr: 1.500e-05, eta: 0:08:38, time: 0.771, data_time: 0.001, memory: 10389, top1_acc: 0.9710, top5_acc: 0.9973, loss_cls: 0.1545, loss: 0.1545, grad_norm: 1.4728 2021-06-06 00:50:44,295 - mmaction - INFO - Saving checkpoint at 24 epochs 2021-06-06 00:58:16,773 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:58:16,857 - mmaction - INFO - top1_acc 0.9413 top5_acc 0.9950 2021-06-06 00:58:16,857 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:58:16,917 - mmaction - INFO - mean_acc 0.9408 2021-06-06 00:58:16,918 - mmaction - INFO - Epoch(val) [24][100] top1_acc: 0.9413, top5_acc: 0.9950, mean_class_accuracy: 0.9408 2021-06-06 01:06:04,566 - mmaction - INFO - Epoch [25][20/100] lr: 1.500e-05, eta: 0:07:06, time: 23.380, data_time: 22.471, memory: 10389, top1_acc: 0.9714, top5_acc: 0.9974, loss_cls: 0.1492, loss: 0.1492, grad_norm: 1.4459 2021-06-06 01:06:20,388 - mmaction - INFO - Epoch [25][40/100] lr: 1.500e-05, eta: 0:05:17, time: 0.793, data_time: 0.003, memory: 10389, top1_acc: 0.9740, top5_acc: 0.9974, loss_cls: 0.1435, loss: 0.1435, grad_norm: 1.3842 2021-06-06 01:06:36,227 - mmaction - INFO - Epoch [25][60/100] lr: 1.500e-05, eta: 0:03:30, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9745, top5_acc: 0.9979, loss_cls: 0.1462, loss: 0.1462, grad_norm: 1.4095 2021-06-06 01:06:52,072 - mmaction - INFO - Epoch [25][80/100] lr: 1.500e-05, eta: 0:01:44, time: 0.792, data_time: 0.001, memory: 10389, top1_acc: 0.9740, top5_acc: 0.9984, loss_cls: 0.1474, loss: 0.1474, grad_norm: 1.4387 2021-06-06 01:07:07,456 - mmaction - INFO - Epoch [25][100/100] lr: 1.500e-05, eta: 0:00:00, time: 0.769, data_time: 0.001, memory: 10389, top1_acc: 0.9689, top5_acc: 0.9973, loss_cls: 0.1575, loss: 0.1575, grad_norm: 1.5027 2021-06-06 01:07:12,120 - mmaction - INFO - Saving checkpoint at 25 epochs 2021-06-06 01:14:30,324 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 01:14:30,396 - mmaction - INFO - top1_acc 0.9450 top5_acc 0.9958 2021-06-06 01:14:30,396 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 01:14:30,447 - mmaction - INFO - mean_acc 0.9439 2021-06-06 01:14:30,448 - mmaction - INFO - Epoch(val) [25][100] top1_acc: 0.9450, top5_acc: 0.9958, mean_class_accuracy: 0.9439