2021-06-05 18:25:07,650 - 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 18:25:07,651 - mmaction - INFO - Distributed training: True 2021-06-05 18:25:08,686 - mmaction - INFO - Config: model = dict( type='Recognizer2D', backbone=dict( type='ResNetTSM', pretrained='torchvision://resnet50', depth=50, norm_eval=False, shift_div=8, num_segments=16), cls_head=dict( type='TSMHead', num_classes=51, 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=16), train_cfg=None, test_cfg=dict(average_clips='prob')) optimizer = dict( type='SGD', constructor='TSMOptimizerConstructor', paramwise_cfg=dict(fc_lr5=True), lr=0.00075, 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_1x1x16_50e_kinetics400_rgb/tsm_r50_256p_1x1x16_50e_kinetics400_rgb_20201010-85645c2a.pth' resume_from = None workflow = [('train', 1)] split = 1 dataset_type = 'RawframeDataset' data_root = 'data/hmdb51/rawframes' data_root_val = 'data/hmdb51/rawframes' ann_file_train = 'data/hmdb51/hmdb51_rgb_train_split_1.txt' ann_file_val = 'data/hmdb51/hmdb51_rgb_val_split_1.txt' ann_file_test = 'data/hmdb51/hmdb51_rgb_val_split_1.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=16), 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=16, 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=16, 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=6, workers_per_gpu=4, train=dict( type='RawframeDataset', ann_file='data/hmdb51/hmdb51_rgb_train_split_1.txt', data_prefix='data/hmdb51/rawframes', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=16), 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/hmdb51/hmdb51_rgb_val_split_1.txt', data_prefix='data/hmdb51/rawframes', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=16, 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/hmdb51/hmdb51_rgb_val_split_1.txt', data_prefix='data/hmdb51/rawframes', pipeline=[ dict( type='SampleFrames', clip_len=1, frame_interval=1, num_clips=16, 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_1x1x16_25e_hmdb51_rgb/' gpu_ids = range(0, 8) omnisource = False module_hooks = [] 2021-06-05 18:25:30,005 - mmaction - INFO - These parameters in pretrained checkpoint are not loaded: {'fc.bias', 'fc.weight'} 2021-06-05 18:25:31,963 - mmaction - INFO - load checkpoint from https://download.openmmlab.com/mmaction/recognition/tsm/tsm_r50_256p_1x1x16_50e_kinetics400_rgb/tsm_r50_256p_1x1x16_50e_kinetics400_rgb_20201010-85645c2a.pth 2021-06-05 18:25:31,964 - mmaction - INFO - Use load_from_http loader 2021-06-05 18:25:32,543 - 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([51, 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([51]). 2021-06-05 18:25:32,547 - mmaction - INFO - Start running, host: linjintao@SH-IDC1-10-5-30-42, work_dir: /mnt/lustre/linjintao/try/mmaction2_dev/work_dirs/tsm_r50_1x1x16_25e_hmdb51_rgb 2021-06-05 18:25:32,547 - mmaction - INFO - workflow: [('train', 1)], max: 25 epochs 2021-06-05 18:33:44,998 - mmaction - INFO - Epoch [1][20/75] lr: 7.500e-04, eta: 12:41:08, time: 24.619, data_time: 23.190, memory: 10388, top1_acc: 0.0646, top5_acc: 0.1948, loss_cls: 3.8750, loss: 3.8750, grad_norm: 2.0509 2021-06-05 18:34:23,422 - mmaction - INFO - Epoch [1][40/75] lr: 7.500e-04, eta: 6:45:52, time: 1.923, data_time: 0.100, memory: 10388, top1_acc: 0.2271, top5_acc: 0.4521, loss_cls: 3.6502, loss: 3.6502, grad_norm: 2.1891 2021-06-05 18:34:56,292 - mmaction - INFO - Epoch [1][60/75] lr: 7.500e-04, eta: 4:44:12, time: 1.644, data_time: 0.030, memory: 10388, top1_acc: 0.3771, top5_acc: 0.6083, loss_cls: 3.3364, loss: 3.3364, grad_norm: 2.3908 2021-06-05 18:35:22,854 - mmaction - INFO - Saving checkpoint at 1 epochs 2021-06-05 18:44:38,422 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 18:44:38,474 - mmaction - INFO - top1_acc 0.4745 top5_acc 0.7673 2021-06-05 18:44:38,474 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 18:44:38,540 - mmaction - INFO - mean_acc 0.4745 2021-06-05 18:44:39,385 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_1.pth. 2021-06-05 18:44:39,385 - mmaction - INFO - Best top1_acc is 0.4745 at 1 epoch. 2021-06-05 18:44:39,397 - mmaction - INFO - Epoch(val) [1][75] top1_acc: 0.4745, top5_acc: 0.7673, mean_class_accuracy: 0.4745 2021-06-05 18:51:43,230 - mmaction - INFO - Epoch [2][20/75] lr: 7.500e-04, eta: 5:08:23, time: 21.190, data_time: 19.993, memory: 10388, top1_acc: 0.4948, top5_acc: 0.7719, loss_cls: 2.7521, loss: 2.7521, grad_norm: 2.6316 2021-06-05 18:52:04,000 - mmaction - INFO - Epoch [2][40/75] lr: 7.500e-04, eta: 4:17:11, time: 1.040, data_time: 0.004, memory: 10388, top1_acc: 0.4948, top5_acc: 0.7885, loss_cls: 2.5566, loss: 2.5566, grad_norm: 2.6680 2021-06-05 18:52:29,011 - mmaction - INFO - Epoch [2][60/75] lr: 7.500e-04, eta: 3:41:58, time: 1.250, data_time: 0.001, memory: 10388, top1_acc: 0.5417, top5_acc: 0.8104, loss_cls: 2.3024, loss: 2.3024, grad_norm: 2.7269 2021-06-05 18:52:47,496 - mmaction - INFO - Saving checkpoint at 2 epochs 2021-06-05 19:00:08,202 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:00:08,255 - mmaction - INFO - top1_acc 0.6118 top5_acc 0.8928 2021-06-05 19:00:08,255 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:00:08,315 - mmaction - INFO - mean_acc 0.6118 2021-06-05 19:00:09,174 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_2.pth. 2021-06-05 19:00:09,174 - mmaction - INFO - Best top1_acc is 0.6118 at 2 epoch. 2021-06-05 19:00:09,175 - mmaction - INFO - Epoch(val) [2][75] top1_acc: 0.6118, top5_acc: 0.8928, mean_class_accuracy: 0.6118 2021-06-05 19:07:36,312 - mmaction - INFO - Epoch [3][20/75] lr: 7.500e-04, eta: 4:07:27, time: 22.355, data_time: 21.427, memory: 10388, top1_acc: 0.6094, top5_acc: 0.8510, loss_cls: 1.9887, loss: 1.9887, grad_norm: 2.9079 2021-06-05 19:07:52,589 - mmaction - INFO - Epoch [3][40/75] lr: 7.500e-04, eta: 3:41:13, time: 0.815, data_time: 0.003, memory: 10388, top1_acc: 0.6260, top5_acc: 0.8729, loss_cls: 1.8572, loss: 1.8572, grad_norm: 3.0585 2021-06-05 19:08:08,856 - mmaction - INFO - Epoch [3][60/75] lr: 7.500e-04, eta: 3:19:55, time: 0.813, data_time: 0.001, memory: 10388, top1_acc: 0.5979, top5_acc: 0.8375, loss_cls: 1.8385, loss: 1.8385, grad_norm: 3.0224 2021-06-05 19:08:26,317 - mmaction - INFO - Saving checkpoint at 3 epochs 2021-06-05 19:16:03,267 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:16:03,320 - mmaction - INFO - top1_acc 0.6542 top5_acc 0.9163 2021-06-05 19:16:03,320 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:16:03,386 - mmaction - INFO - mean_acc 0.6542 2021-06-05 19:16:04,256 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_3.pth. 2021-06-05 19:16:04,256 - mmaction - INFO - Best top1_acc is 0.6542 at 3 epoch. 2021-06-05 19:16:04,257 - mmaction - INFO - Epoch(val) [3][75] top1_acc: 0.6542, top5_acc: 0.9163, mean_class_accuracy: 0.6542 2021-06-05 19:23:48,833 - mmaction - INFO - Epoch [4][20/75] lr: 7.500e-04, eta: 3:39:15, time: 23.222, data_time: 22.300, memory: 10388, top1_acc: 0.6438, top5_acc: 0.8771, loss_cls: 1.6198, loss: 1.6198, grad_norm: 3.0346 2021-06-05 19:24:05,049 - mmaction - INFO - Epoch [4][40/75] lr: 7.500e-04, eta: 3:21:52, time: 0.812, data_time: 0.004, memory: 10388, top1_acc: 0.6229, top5_acc: 0.8885, loss_cls: 1.5917, loss: 1.5917, grad_norm: 3.2145 2021-06-05 19:24:21,540 - mmaction - INFO - Epoch [4][60/75] lr: 7.500e-04, eta: 3:06:54, time: 0.824, data_time: 0.001, memory: 10388, top1_acc: 0.6500, top5_acc: 0.8583, loss_cls: 1.5597, loss: 1.5597, grad_norm: 3.2244 2021-06-05 19:24:38,141 - mmaction - INFO - Saving checkpoint at 4 epochs 2021-06-05 19:31:58,509 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:31:58,565 - mmaction - INFO - top1_acc 0.6791 top5_acc 0.9170 2021-06-05 19:31:58,565 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:31:58,625 - mmaction - INFO - mean_acc 0.6791 2021-06-05 19:31:59,487 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_4.pth. 2021-06-05 19:31:59,487 - mmaction - INFO - Best top1_acc is 0.6791 at 4 epoch. 2021-06-05 19:31:59,488 - mmaction - INFO - Epoch(val) [4][75] top1_acc: 0.6791, top5_acc: 0.9170, mean_class_accuracy: 0.6791 2021-06-05 19:39:30,301 - mmaction - INFO - Epoch [5][20/75] lr: 7.500e-04, eta: 3:19:18, time: 22.539, data_time: 21.637, memory: 10388, top1_acc: 0.6500, top5_acc: 0.8948, loss_cls: 1.4709, loss: 1.4709, grad_norm: 3.3334 2021-06-05 19:39:46,562 - mmaction - INFO - Epoch [5][40/75] lr: 7.500e-04, eta: 3:06:23, time: 0.815, data_time: 0.003, memory: 10388, top1_acc: 0.6677, top5_acc: 0.9042, loss_cls: 1.3805, loss: 1.3805, grad_norm: 3.2775 2021-06-05 19:40:03,133 - mmaction - INFO - Epoch [5][60/75] lr: 7.500e-04, eta: 2:54:54, time: 0.828, data_time: 0.001, memory: 10388, top1_acc: 0.6885, top5_acc: 0.9021, loss_cls: 1.3434, loss: 1.3434, grad_norm: 3.3469 2021-06-05 19:40:20,237 - mmaction - INFO - Saving checkpoint at 5 epochs 2021-06-05 19:48:10,414 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 19:48:10,462 - mmaction - INFO - top1_acc 0.7026 top5_acc 0.9275 2021-06-05 19:48:10,462 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 19:48:10,528 - mmaction - INFO - mean_acc 0.7026 2021-06-05 19:48:11,476 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_5.pth. 2021-06-05 19:48:11,477 - mmaction - INFO - Best top1_acc is 0.7026 at 5 epoch. 2021-06-05 19:48:11,477 - mmaction - INFO - Epoch(val) [5][75] top1_acc: 0.7026, top5_acc: 0.9275, mean_class_accuracy: 0.7026 2021-06-05 19:55:22,247 - mmaction - INFO - Epoch [6][20/75] lr: 7.500e-04, eta: 3:02:37, time: 21.531, data_time: 20.653, memory: 10388, top1_acc: 0.6677, top5_acc: 0.8938, loss_cls: 1.3134, loss: 1.3134, grad_norm: 3.3723 2021-06-05 19:55:38,558 - mmaction - INFO - Epoch [6][40/75] lr: 7.500e-04, eta: 2:52:25, time: 0.818, data_time: 0.004, memory: 10388, top1_acc: 0.6917, top5_acc: 0.9052, loss_cls: 1.2732, loss: 1.2732, grad_norm: 3.4783 2021-06-05 19:55:54,877 - mmaction - INFO - Epoch [6][60/75] lr: 7.500e-04, eta: 2:43:08, time: 0.815, data_time: 0.001, memory: 10388, top1_acc: 0.6802, top5_acc: 0.8969, loss_cls: 1.2779, loss: 1.2779, grad_norm: 3.5265 2021-06-05 19:56:09,689 - mmaction - INFO - Saving checkpoint at 6 epochs 2021-06-05 20:03:31,386 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:03:31,442 - mmaction - INFO - top1_acc 0.7131 top5_acc 0.9320 2021-06-05 20:03:31,442 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:03:31,507 - mmaction - INFO - mean_acc 0.7131 2021-06-05 20:03:32,371 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_6.pth. 2021-06-05 20:03:32,372 - mmaction - INFO - Best top1_acc is 0.7131 at 6 epoch. 2021-06-05 20:03:32,373 - mmaction - INFO - Epoch(val) [6][75] top1_acc: 0.7131, top5_acc: 0.9320, mean_class_accuracy: 0.7131 2021-06-05 20:10:57,380 - mmaction - INFO - Epoch [7][20/75] lr: 7.500e-04, eta: 2:49:29, time: 22.244, data_time: 21.341, memory: 10388, top1_acc: 0.6844, top5_acc: 0.9125, loss_cls: 1.2235, loss: 1.2235, grad_norm: 3.5897 2021-06-05 20:11:13,642 - mmaction - INFO - Epoch [7][40/75] lr: 7.500e-04, eta: 2:41:01, time: 0.815, data_time: 0.003, memory: 10388, top1_acc: 0.7125, top5_acc: 0.9125, loss_cls: 1.1442, loss: 1.1442, grad_norm: 3.3594 2021-06-05 20:11:30,097 - mmaction - INFO - Epoch [7][60/75] lr: 7.500e-04, eta: 2:33:12, time: 0.822, data_time: 0.002, memory: 10388, top1_acc: 0.7177, top5_acc: 0.9010, loss_cls: 1.1707, loss: 1.1707, grad_norm: 3.4816 2021-06-05 20:11:46,739 - mmaction - INFO - Saving checkpoint at 7 epochs 2021-06-05 20:19:05,274 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:19:05,323 - mmaction - INFO - top1_acc 0.7294 top5_acc 0.9346 2021-06-05 20:19:05,324 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:19:05,384 - mmaction - INFO - mean_acc 0.7294 2021-06-05 20:19:06,271 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_7.pth. 2021-06-05 20:19:06,271 - mmaction - INFO - Best top1_acc is 0.7294 at 7 epoch. 2021-06-05 20:19:06,272 - mmaction - INFO - Epoch(val) [7][75] top1_acc: 0.7294, top5_acc: 0.9346, mean_class_accuracy: 0.7294 2021-06-05 20:26:35,935 - mmaction - INFO - Epoch [8][20/75] lr: 7.500e-04, eta: 2:37:59, time: 22.480, data_time: 21.536, memory: 10388, top1_acc: 0.7104, top5_acc: 0.9208, loss_cls: 1.1045, loss: 1.1045, grad_norm: 3.6540 2021-06-05 20:26:52,276 - mmaction - INFO - Epoch [8][40/75] lr: 7.500e-04, eta: 2:30:44, time: 0.820, data_time: 0.004, memory: 10388, top1_acc: 0.7323, top5_acc: 0.9156, loss_cls: 1.1015, loss: 1.1015, grad_norm: 3.4561 2021-06-05 20:27:08,583 - mmaction - INFO - Epoch [8][60/75] lr: 7.500e-04, eta: 2:23:57, time: 0.815, data_time: 0.001, memory: 10388, top1_acc: 0.7052, top5_acc: 0.9177, loss_cls: 1.1283, loss: 1.1283, grad_norm: 3.7060 2021-06-05 20:27:25,087 - mmaction - INFO - Saving checkpoint at 8 epochs 2021-06-05 20:35:09,489 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:35:09,534 - mmaction - INFO - top1_acc 0.7359 top5_acc 0.9366 2021-06-05 20:35:09,534 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:35:09,595 - mmaction - INFO - mean_acc 0.7359 2021-06-05 20:35:10,455 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_8.pth. 2021-06-05 20:35:10,455 - mmaction - INFO - Best top1_acc is 0.7359 at 8 epoch. 2021-06-05 20:35:10,456 - mmaction - INFO - Epoch(val) [8][75] top1_acc: 0.7359, top5_acc: 0.9366, mean_class_accuracy: 0.7359 2021-06-05 20:42:38,755 - mmaction - INFO - Epoch [9][20/75] lr: 7.500e-04, eta: 2:27:15, time: 22.406, data_time: 21.442, memory: 10388, top1_acc: 0.7208, top5_acc: 0.9156, loss_cls: 1.1037, loss: 1.1037, grad_norm: 3.7787 2021-06-05 20:42:55,101 - mmaction - INFO - Epoch [9][40/75] lr: 7.500e-04, eta: 2:20:54, time: 0.820, data_time: 0.005, memory: 10388, top1_acc: 0.7344, top5_acc: 0.9260, loss_cls: 1.0290, loss: 1.0290, grad_norm: 3.6853 2021-06-05 20:43:11,391 - mmaction - INFO - Epoch [9][60/75] lr: 7.500e-04, eta: 2:14:55, time: 0.815, data_time: 0.001, memory: 10388, top1_acc: 0.7167, top5_acc: 0.9125, loss_cls: 1.0865, loss: 1.0865, grad_norm: 3.8400 2021-06-05 20:43:28,204 - mmaction - INFO - Saving checkpoint at 9 epochs 2021-06-05 20:51:06,507 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 20:51:06,558 - mmaction - INFO - top1_acc 0.7268 top5_acc 0.9346 2021-06-05 20:51:06,558 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 20:51:06,618 - mmaction - INFO - mean_acc 0.7268 2021-06-05 20:51:06,619 - mmaction - INFO - Epoch(val) [9][75] top1_acc: 0.7268, top5_acc: 0.9346, mean_class_accuracy: 0.7268 2021-06-05 20:58:24,932 - mmaction - INFO - Epoch [10][20/75] lr: 7.500e-04, eta: 2:16:50, time: 21.913, data_time: 21.009, memory: 10388, top1_acc: 0.7354, top5_acc: 0.9156, loss_cls: 1.0270, loss: 1.0270, grad_norm: 3.7641 2021-06-05 20:58:41,210 - mmaction - INFO - Epoch [10][40/75] lr: 7.500e-04, eta: 2:11:12, time: 0.815, data_time: 0.003, memory: 10388, top1_acc: 0.7479, top5_acc: 0.9313, loss_cls: 0.9924, loss: 0.9924, grad_norm: 3.7466 2021-06-05 20:58:57,460 - mmaction - INFO - Epoch [10][60/75] lr: 7.500e-04, eta: 2:05:51, time: 0.813, data_time: 0.001, memory: 10388, top1_acc: 0.7354, top5_acc: 0.9198, loss_cls: 1.0109, loss: 1.0109, grad_norm: 3.8310 2021-06-05 20:59:14,049 - mmaction - INFO - Saving checkpoint at 10 epochs 2021-06-05 21:06:33,374 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:06:33,425 - mmaction - INFO - top1_acc 0.7412 top5_acc 0.9386 2021-06-05 21:06:33,425 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:06:33,485 - mmaction - INFO - mean_acc 0.7412 2021-06-05 21:06:34,347 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_10.pth. 2021-06-05 21:06:34,347 - mmaction - INFO - Best top1_acc is 0.7412 at 10 epoch. 2021-06-05 21:06:34,348 - mmaction - INFO - Epoch(val) [10][75] top1_acc: 0.7412, top5_acc: 0.9386, mean_class_accuracy: 0.7412 2021-06-05 21:14:13,402 - mmaction - INFO - Epoch [11][20/75] lr: 7.500e-05, eta: 2:07:25, time: 22.950, data_time: 22.003, memory: 10388, top1_acc: 0.7646, top5_acc: 0.9344, loss_cls: 0.9224, loss: 0.9224, grad_norm: 3.6857 2021-06-05 21:14:29,751 - mmaction - INFO - Epoch [11][40/75] lr: 7.500e-05, eta: 2:02:19, time: 0.819, data_time: 0.005, memory: 10388, top1_acc: 0.7552, top5_acc: 0.9323, loss_cls: 0.9669, loss: 0.9669, grad_norm: 3.8407 2021-06-05 21:14:46,161 - mmaction - INFO - Epoch [11][60/75] lr: 7.500e-05, eta: 1:57:27, time: 0.821, data_time: 0.002, memory: 10388, top1_acc: 0.7573, top5_acc: 0.9396, loss_cls: 0.9421, loss: 0.9421, grad_norm: 3.7128 2021-06-05 21:15:03,222 - mmaction - INFO - Saving checkpoint at 11 epochs 2021-06-05 21:22:47,803 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:22:47,861 - mmaction - INFO - top1_acc 0.7373 top5_acc 0.9405 2021-06-05 21:22:47,861 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:22:47,927 - mmaction - INFO - mean_acc 0.7373 2021-06-05 21:22:47,927 - mmaction - INFO - Epoch(val) [11][75] top1_acc: 0.7373, top5_acc: 0.9405, mean_class_accuracy: 0.7373 2021-06-05 21:30:30,616 - mmaction - INFO - Epoch [12][20/75] lr: 7.500e-05, eta: 1:58:17, time: 23.132, data_time: 22.288, memory: 10388, top1_acc: 0.7417, top5_acc: 0.9219, loss_cls: 0.9708, loss: 0.9708, grad_norm: 3.7930 2021-06-05 21:30:46,866 - mmaction - INFO - Epoch [12][40/75] lr: 7.500e-05, eta: 1:53:38, time: 0.814, data_time: 0.004, memory: 10388, top1_acc: 0.7562, top5_acc: 0.9323, loss_cls: 0.9438, loss: 0.9438, grad_norm: 3.7154 2021-06-05 21:31:03,435 - mmaction - INFO - Epoch [12][60/75] lr: 7.500e-05, eta: 1:49:10, time: 0.829, data_time: 0.001, memory: 10388, top1_acc: 0.7594, top5_acc: 0.9240, loss_cls: 0.9547, loss: 0.9547, grad_norm: 3.7330 2021-06-05 21:31:20,194 - mmaction - INFO - Saving checkpoint at 12 epochs 2021-06-05 21:38:53,300 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:38:53,353 - mmaction - INFO - top1_acc 0.7366 top5_acc 0.9399 2021-06-05 21:38:53,353 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:38:53,418 - mmaction - INFO - mean_acc 0.7366 2021-06-05 21:38:53,419 - mmaction - INFO - Epoch(val) [12][75] top1_acc: 0.7366, top5_acc: 0.9399, mean_class_accuracy: 0.7366 2021-06-05 21:46:41,506 - mmaction - INFO - Epoch [13][20/75] lr: 7.500e-05, eta: 1:49:24, time: 23.403, data_time: 22.462, memory: 10388, top1_acc: 0.7667, top5_acc: 0.9385, loss_cls: 0.9316, loss: 0.9316, grad_norm: 3.8207 2021-06-05 21:46:57,730 - mmaction - INFO - Epoch [13][40/75] lr: 7.500e-05, eta: 1:45:06, time: 0.812, data_time: 0.002, memory: 10388, top1_acc: 0.7583, top5_acc: 0.9365, loss_cls: 0.9338, loss: 0.9338, grad_norm: 3.7665 2021-06-05 21:47:14,065 - mmaction - INFO - Epoch [13][60/75] lr: 7.500e-05, eta: 1:40:58, time: 0.816, data_time: 0.001, memory: 10388, top1_acc: 0.7583, top5_acc: 0.9323, loss_cls: 0.9680, loss: 0.9680, grad_norm: 3.7590 2021-06-05 21:47:30,853 - mmaction - INFO - Saving checkpoint at 13 epochs 2021-06-05 21:55:11,685 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 21:55:11,739 - mmaction - INFO - top1_acc 0.7412 top5_acc 0.9464 2021-06-05 21:55:11,739 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 21:55:11,803 - mmaction - INFO - mean_acc 0.7412 2021-06-05 21:55:11,804 - mmaction - INFO - Epoch(val) [13][75] top1_acc: 0.7412, top5_acc: 0.9464, mean_class_accuracy: 0.7412 2021-06-05 22:02:54,268 - mmaction - INFO - Epoch [14][20/75] lr: 7.500e-05, eta: 1:40:30, time: 23.121, data_time: 22.179, memory: 10388, top1_acc: 0.7729, top5_acc: 0.9281, loss_cls: 0.9395, loss: 0.9395, grad_norm: 3.7462 2021-06-05 22:03:10,539 - mmaction - INFO - Epoch [14][40/75] lr: 7.500e-05, eta: 1:36:31, time: 0.815, data_time: 0.004, memory: 10388, top1_acc: 0.7531, top5_acc: 0.9385, loss_cls: 0.9472, loss: 0.9472, grad_norm: 3.8490 2021-06-05 22:03:26,895 - mmaction - INFO - Epoch [14][60/75] lr: 7.500e-05, eta: 1:32:40, time: 0.817, data_time: 0.001, memory: 10388, top1_acc: 0.7500, top5_acc: 0.9375, loss_cls: 0.9432, loss: 0.9432, grad_norm: 3.7433 2021-06-05 22:03:43,548 - mmaction - INFO - Saving checkpoint at 14 epochs 2021-06-05 22:10:59,043 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:10:59,101 - mmaction - INFO - top1_acc 0.7477 top5_acc 0.9386 2021-06-05 22:10:59,101 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:10:59,166 - mmaction - INFO - mean_acc 0.7477 2021-06-05 22:11:00,025 - mmaction - INFO - Now best checkpoint is saved as best_top1_acc_epoch_14.pth. 2021-06-05 22:11:00,025 - mmaction - INFO - Best top1_acc is 0.7477 at 14 epoch. 2021-06-05 22:11:00,026 - mmaction - INFO - Epoch(val) [14][75] top1_acc: 0.7477, top5_acc: 0.9386, mean_class_accuracy: 0.7477 2021-06-05 22:18:48,770 - mmaction - INFO - Epoch [15][20/75] lr: 7.500e-05, eta: 1:31:47, time: 23.435, data_time: 22.440, memory: 10388, top1_acc: 0.7771, top5_acc: 0.9375, loss_cls: 0.9219, loss: 0.9219, grad_norm: 3.7849 2021-06-05 22:19:04,820 - mmaction - INFO - Epoch [15][40/75] lr: 7.500e-05, eta: 1:28:03, time: 0.804, data_time: 0.004, memory: 10388, top1_acc: 0.7625, top5_acc: 0.9354, loss_cls: 0.9413, loss: 0.9413, grad_norm: 3.8462 2021-06-05 22:19:21,169 - mmaction - INFO - Epoch [15][60/75] lr: 7.500e-05, eta: 1:24:27, time: 0.817, data_time: 0.001, memory: 10388, top1_acc: 0.7542, top5_acc: 0.9260, loss_cls: 0.9347, loss: 0.9347, grad_norm: 3.8097 2021-06-05 22:19:37,958 - mmaction - INFO - Saving checkpoint at 15 epochs 2021-06-05 22:27:25,792 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:27:25,848 - mmaction - INFO - top1_acc 0.7359 top5_acc 0.9373 2021-06-05 22:27:25,848 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:27:25,913 - mmaction - INFO - mean_acc 0.7359 2021-06-05 22:27:25,914 - mmaction - INFO - Epoch(val) [15][75] top1_acc: 0.7359, top5_acc: 0.9373, mean_class_accuracy: 0.7359 2021-06-05 22:35:08,957 - mmaction - INFO - Epoch [16][20/75] lr: 7.500e-05, eta: 1:23:02, time: 23.150, data_time: 22.242, memory: 10388, top1_acc: 0.7438, top5_acc: 0.9292, loss_cls: 0.9577, loss: 0.9577, grad_norm: 3.8367 2021-06-05 22:35:25,234 - mmaction - INFO - Epoch [16][40/75] lr: 7.500e-05, eta: 1:19:33, time: 0.815, data_time: 0.004, memory: 10388, top1_acc: 0.7781, top5_acc: 0.9385, loss_cls: 0.9214, loss: 0.9214, grad_norm: 3.8616 2021-06-05 22:35:41,461 - mmaction - INFO - Epoch [16][60/75] lr: 7.500e-05, eta: 1:16:09, time: 0.812, data_time: 0.001, memory: 10388, top1_acc: 0.7719, top5_acc: 0.9385, loss_cls: 0.9239, loss: 0.9239, grad_norm: 3.7100 2021-06-05 22:35:58,521 - mmaction - INFO - Saving checkpoint at 16 epochs 2021-06-05 22:43:40,543 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:43:40,592 - mmaction - INFO - top1_acc 0.7425 top5_acc 0.9399 2021-06-05 22:43:40,592 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:43:40,658 - mmaction - INFO - mean_acc 0.7425 2021-06-05 22:43:40,658 - mmaction - INFO - Epoch(val) [16][75] top1_acc: 0.7425, top5_acc: 0.9399, mean_class_accuracy: 0.7425 2021-06-05 22:51:25,470 - mmaction - INFO - Epoch [17][20/75] lr: 7.500e-05, eta: 1:14:23, time: 23.239, data_time: 22.317, memory: 10388, top1_acc: 0.7448, top5_acc: 0.9323, loss_cls: 0.9335, loss: 0.9335, grad_norm: 3.8124 2021-06-05 22:51:41,907 - mmaction - INFO - Epoch [17][40/75] lr: 7.500e-05, eta: 1:11:05, time: 0.823, data_time: 0.003, memory: 10388, top1_acc: 0.7448, top5_acc: 0.9167, loss_cls: 1.0069, loss: 1.0069, grad_norm: 3.8349 2021-06-05 22:51:58,213 - mmaction - INFO - Epoch [17][60/75] lr: 7.500e-05, eta: 1:07:53, time: 0.815, data_time: 0.001, memory: 10388, top1_acc: 0.7604, top5_acc: 0.9313, loss_cls: 0.9272, loss: 0.9272, grad_norm: 3.8032 2021-06-05 22:52:14,928 - mmaction - INFO - Saving checkpoint at 17 epochs 2021-06-05 22:59:41,798 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 22:59:41,856 - mmaction - INFO - top1_acc 0.7477 top5_acc 0.9438 2021-06-05 22:59:41,856 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 22:59:41,922 - mmaction - INFO - mean_acc 0.7477 2021-06-05 22:59:41,922 - mmaction - INFO - Epoch(val) [17][75] top1_acc: 0.7477, top5_acc: 0.9438, mean_class_accuracy: 0.7477 2021-06-05 23:07:29,041 - mmaction - INFO - Epoch [18][20/75] lr: 7.500e-05, eta: 1:05:47, time: 23.354, data_time: 22.467, memory: 10388, top1_acc: 0.7708, top5_acc: 0.9271, loss_cls: 0.9612, loss: 0.9612, grad_norm: 3.7515 2021-06-05 23:07:45,294 - mmaction - INFO - Epoch [18][40/75] lr: 7.500e-05, eta: 1:02:39, time: 0.814, data_time: 0.003, memory: 10388, top1_acc: 0.7875, top5_acc: 0.9427, loss_cls: 0.8484, loss: 0.8484, grad_norm: 3.6601 2021-06-05 23:08:01,811 - mmaction - INFO - Epoch [18][60/75] lr: 7.500e-05, eta: 0:59:38, time: 0.826, data_time: 0.001, memory: 10388, top1_acc: 0.7427, top5_acc: 0.9229, loss_cls: 0.9793, loss: 0.9793, grad_norm: 3.8729 2021-06-05 23:08:18,636 - mmaction - INFO - Saving checkpoint at 18 epochs 2021-06-05 23:16:11,224 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:16:11,273 - mmaction - INFO - top1_acc 0.7438 top5_acc 0.9405 2021-06-05 23:16:11,273 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:16:11,333 - mmaction - INFO - mean_acc 0.7438 2021-06-05 23:16:11,334 - mmaction - INFO - Epoch(val) [18][75] top1_acc: 0.7438, top5_acc: 0.9405, mean_class_accuracy: 0.7438 2021-06-05 23:23:49,991 - mmaction - INFO - Epoch [19][20/75] lr: 7.500e-05, eta: 0:57:09, time: 22.929, data_time: 21.962, memory: 10388, top1_acc: 0.7594, top5_acc: 0.9354, loss_cls: 0.8874, loss: 0.8874, grad_norm: 3.7134 2021-06-05 23:24:06,310 - mmaction - INFO - Epoch [19][40/75] lr: 7.500e-05, eta: 0:54:12, time: 0.819, data_time: 0.006, memory: 10388, top1_acc: 0.7521, top5_acc: 0.9229, loss_cls: 0.9500, loss: 0.9500, grad_norm: 3.7913 2021-06-05 23:24:22,541 - mmaction - INFO - Epoch [19][60/75] lr: 7.500e-05, eta: 0:51:19, time: 0.811, data_time: 0.001, memory: 10388, top1_acc: 0.7646, top5_acc: 0.9448, loss_cls: 0.9257, loss: 0.9257, grad_norm: 3.7823 2021-06-05 23:24:39,147 - mmaction - INFO - Saving checkpoint at 19 epochs 2021-06-05 23:31:47,248 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:31:47,291 - mmaction - INFO - top1_acc 0.7412 top5_acc 0.9405 2021-06-05 23:31:47,291 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:31:47,331 - mmaction - INFO - mean_acc 0.7412 2021-06-05 23:31:47,332 - mmaction - INFO - Epoch(val) [19][75] top1_acc: 0.7412, top5_acc: 0.9405, mean_class_accuracy: 0.7412 2021-06-05 23:39:12,968 - mmaction - INFO - Epoch [20][20/75] lr: 7.500e-05, eta: 0:48:31, time: 22.280, data_time: 21.375, memory: 10388, top1_acc: 0.7573, top5_acc: 0.9250, loss_cls: 0.9599, loss: 0.9599, grad_norm: 3.8401 2021-06-05 23:39:29,492 - mmaction - INFO - Epoch [20][40/75] lr: 7.500e-05, eta: 0:45:42, time: 0.827, data_time: 0.003, memory: 10388, top1_acc: 0.7719, top5_acc: 0.9344, loss_cls: 0.9122, loss: 0.9122, grad_norm: 3.8042 2021-06-05 23:39:45,867 - mmaction - INFO - Epoch [20][60/75] lr: 7.500e-05, eta: 0:42:57, time: 0.819, data_time: 0.001, memory: 10388, top1_acc: 0.7802, top5_acc: 0.9281, loss_cls: 0.9214, loss: 0.9214, grad_norm: 3.7239 2021-06-05 23:40:02,710 - mmaction - INFO - Saving checkpoint at 20 epochs 2021-06-05 23:47:43,912 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-05 23:47:43,970 - mmaction - INFO - top1_acc 0.7366 top5_acc 0.9405 2021-06-05 23:47:43,971 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-05 23:47:44,036 - mmaction - INFO - mean_acc 0.7366 2021-06-05 23:47:44,036 - mmaction - INFO - Epoch(val) [20][75] top1_acc: 0.7366, top5_acc: 0.9405, mean_class_accuracy: 0.7366 2021-06-05 23:55:02,668 - mmaction - INFO - Epoch [21][20/75] lr: 7.500e-06, eta: 0:39:54, time: 21.927, data_time: 20.953, memory: 10388, top1_acc: 0.7604, top5_acc: 0.9344, loss_cls: 0.9026, loss: 0.9026, grad_norm: 3.7574 2021-06-05 23:55:19,392 - mmaction - INFO - Epoch [21][40/75] lr: 7.500e-06, eta: 0:37:14, time: 0.840, data_time: 0.006, memory: 10388, top1_acc: 0.7594, top5_acc: 0.9417, loss_cls: 0.9085, loss: 0.9085, grad_norm: 3.8115 2021-06-05 23:55:35,690 - mmaction - INFO - Epoch [21][60/75] lr: 7.500e-06, eta: 0:34:37, time: 0.815, data_time: 0.001, memory: 10388, top1_acc: 0.7896, top5_acc: 0.9323, loss_cls: 0.9042, loss: 0.9042, grad_norm: 3.6575 2021-06-05 23:55:51,199 - mmaction - INFO - Saving checkpoint at 21 epochs 2021-06-06 00:03:35,137 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:03:35,181 - mmaction - INFO - top1_acc 0.7431 top5_acc 0.9425 2021-06-06 00:03:35,181 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:03:35,242 - mmaction - INFO - mean_acc 0.7431 2021-06-06 00:03:35,242 - mmaction - INFO - Epoch(val) [21][75] top1_acc: 0.7431, top5_acc: 0.9425, mean_class_accuracy: 0.7431 2021-06-06 00:11:18,579 - mmaction - INFO - Epoch [22][20/75] lr: 7.500e-06, eta: 0:31:27, time: 23.164, data_time: 22.207, memory: 10388, top1_acc: 0.7656, top5_acc: 0.9323, loss_cls: 0.9448, loss: 0.9448, grad_norm: 3.7806 2021-06-06 00:11:34,917 - mmaction - INFO - Epoch [22][40/75] lr: 7.500e-06, eta: 0:28:53, time: 0.819, data_time: 0.004, memory: 10388, top1_acc: 0.7573, top5_acc: 0.9406, loss_cls: 0.8935, loss: 0.8935, grad_norm: 3.8137 2021-06-06 00:11:51,250 - mmaction - INFO - Epoch [22][60/75] lr: 7.500e-06, eta: 0:26:22, time: 0.816, data_time: 0.001, memory: 10388, top1_acc: 0.7688, top5_acc: 0.9385, loss_cls: 0.8658, loss: 0.8658, grad_norm: 3.7176 2021-06-06 00:12:07,846 - mmaction - INFO - Saving checkpoint at 22 epochs 2021-06-06 00:19:48,008 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:19:48,062 - mmaction - INFO - top1_acc 0.7379 top5_acc 0.9418 2021-06-06 00:19:48,062 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:19:48,128 - mmaction - INFO - mean_acc 0.7379 2021-06-06 00:19:48,128 - mmaction - INFO - Epoch(val) [22][75] top1_acc: 0.7379, top5_acc: 0.9418, mean_class_accuracy: 0.7379 2021-06-06 00:27:37,791 - mmaction - INFO - Epoch [23][20/75] lr: 7.500e-06, eta: 0:23:01, time: 23.481, data_time: 22.551, memory: 10388, top1_acc: 0.7760, top5_acc: 0.9427, loss_cls: 0.8561, loss: 0.8561, grad_norm: 3.6448 2021-06-06 00:27:54,372 - mmaction - INFO - Epoch [23][40/75] lr: 7.500e-06, eta: 0:20:33, time: 0.830, data_time: 0.003, memory: 10388, top1_acc: 0.7458, top5_acc: 0.9271, loss_cls: 0.9503, loss: 0.9503, grad_norm: 3.9455 2021-06-06 00:28:10,684 - mmaction - INFO - Epoch [23][60/75] lr: 7.500e-06, eta: 0:18:08, time: 0.816, data_time: 0.002, memory: 10388, top1_acc: 0.7510, top5_acc: 0.9240, loss_cls: 0.9567, loss: 0.9567, grad_norm: 3.8858 2021-06-06 00:28:27,420 - mmaction - INFO - Saving checkpoint at 23 epochs 2021-06-06 00:35:33,081 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:35:33,129 - mmaction - INFO - top1_acc 0.7412 top5_acc 0.9451 2021-06-06 00:35:33,130 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:35:33,195 - mmaction - INFO - mean_acc 0.7412 2021-06-06 00:35:33,195 - mmaction - INFO - Epoch(val) [23][75] top1_acc: 0.7412, top5_acc: 0.9451, mean_class_accuracy: 0.7412 2021-06-06 00:42:43,397 - mmaction - INFO - Epoch [24][20/75] lr: 7.500e-06, eta: 0:14:32, time: 21.508, data_time: 20.599, memory: 10388, top1_acc: 0.7667, top5_acc: 0.9313, loss_cls: 0.9134, loss: 0.9134, grad_norm: 3.8999 2021-06-06 00:42:59,709 - mmaction - INFO - Epoch [24][40/75] lr: 7.500e-06, eta: 0:12:11, time: 0.817, data_time: 0.003, memory: 10388, top1_acc: 0.7531, top5_acc: 0.9365, loss_cls: 0.9682, loss: 0.9682, grad_norm: 3.9063 2021-06-06 00:43:16,218 - mmaction - INFO - Epoch [24][60/75] lr: 7.500e-06, eta: 0:09:52, time: 0.826, data_time: 0.001, memory: 10388, top1_acc: 0.7708, top5_acc: 0.9406, loss_cls: 0.8658, loss: 0.8658, grad_norm: 3.6090 2021-06-06 00:43:31,080 - mmaction - INFO - Saving checkpoint at 24 epochs 2021-06-06 00:51:02,882 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 00:51:02,940 - mmaction - INFO - top1_acc 0.7405 top5_acc 0.9425 2021-06-06 00:51:02,940 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 00:51:03,005 - mmaction - INFO - mean_acc 0.7405 2021-06-06 00:51:03,006 - mmaction - INFO - Epoch(val) [24][75] top1_acc: 0.7405, top5_acc: 0.9425, mean_class_accuracy: 0.7405 2021-06-06 00:58:47,273 - mmaction - INFO - Epoch [25][20/75] lr: 7.500e-06, eta: 0:06:09, time: 23.212, data_time: 22.286, memory: 10388, top1_acc: 0.7677, top5_acc: 0.9323, loss_cls: 0.9127, loss: 0.9127, grad_norm: 3.7870 2021-06-06 00:59:03,637 - mmaction - INFO - Epoch [25][40/75] lr: 7.500e-06, eta: 0:03:52, time: 0.819, data_time: 0.002, memory: 10388, top1_acc: 0.7448, top5_acc: 0.9500, loss_cls: 0.9353, loss: 0.9353, grad_norm: 3.8136 2021-06-06 00:59:20,046 - mmaction - INFO - Epoch [25][60/75] lr: 7.500e-06, eta: 0:01:38, time: 0.820, data_time: 0.001, memory: 10388, top1_acc: 0.7573, top5_acc: 0.9323, loss_cls: 0.9006, loss: 0.9006, grad_norm: 3.6395 2021-06-06 00:59:36,810 - mmaction - INFO - Saving checkpoint at 25 epochs 2021-06-06 01:07:17,463 - mmaction - INFO - Evaluating top_k_accuracy ... 2021-06-06 01:07:17,512 - mmaction - INFO - top1_acc 0.7359 top5_acc 0.9405 2021-06-06 01:07:17,512 - mmaction - INFO - Evaluating mean_class_accuracy ... 2021-06-06 01:07:17,577 - mmaction - INFO - mean_acc 0.7359 2021-06-06 01:07:17,578 - mmaction - INFO - Epoch(val) [25][75] top1_acc: 0.7359, top5_acc: 0.9405, mean_class_accuracy: 0.7359