------------------------------------ Environment Versions: - Python: 3.7.6 (default, Jan 8 2020, 19:59:22) [GCC 7.3.0] - PyTorch: 1.5.0 - TorchVison: 0.2.1 ------------------------------------ SELFY Configurations: - dataset: something - modality: RGB - train_list: data/something_train.txt - val_list: data/something_val.txt - arch: SELFY - num_segments: 8 - mode: 1 - consensus_type: avg - pretrained_parts: finetune - k: 3 - dropout: 0.5 - loss_type: nll - rep_flow: False - epochs: 50 - batch_size: 64 - iter_size: 1 - lr: 0.01 - lr_steps: [30.0, 40.0] - momentum: 0.9 - weight_decay: 0.0005 - clip_gradient: 20.0 - no_partialbn: True - nesterov: True - print_freq: 20 - eval_freq: 1 - workers: 4 - resume: - evaluate: False - snapshot_pref: net_runs/SELFY_resnet50_something_run1/SELFY_TSM_ResNet - val_output_folder: net_runs/SELFY_resnet50_something_run1/validation - start_epoch: 0 - gpus: None - flow_prefix: img_ - rgb_prefix: img_ ------------------------------------ 1 Initializing TSN with base model: SELFY. TSN Configurations: input_modality: RGB num_segments: 8 new_length: 1 consensus_module: avg dropout_ratio: 0.5 pretrained_parts: finetune ------------------------------------ Model: DataParallel( (module): TSN( (base_model): ResNet( (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False) (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (sigmoid): Sigmoid() (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False) (softmax): Softmax(dim=1) (layer1): Sequential( (0): Bottleneck( (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (layer2): Sequential( (0): Bottleneck( (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (3): Bottleneck( (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (selfy): SELFYBlock( (stss_transformation): STSSTransformation( (downsample): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) ) (stss_extraction): STSSExtraction( (conv0): Sequential( (0): Conv3d(1, 4, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(4, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv1): Sequential( (0): Conv3d(4, 16, kernel_size=(1, 3, 3), stride=(1, 2, 2), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv2): Sequential( (0): Conv3d(16, 64, kernel_size=(1, 3, 3), stride=(1, 2, 2), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv3): Sequential( (0): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), bias=False) (1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) ) (stss_integration): STSSIntegration( (conv0): Sequential( (0): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv1): Sequential( (0): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv2): Sequential( (0): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): ReLU(inplace=True) ) (conv3_fuse): Sequential( (0): Rearrange('(b l) c t h w -> b (l c) t h w', l=5) (1): Conv3d(320, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (2): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (3): ReLU(inplace=True) ) (upsample): Sequential( (0): ConvTranspose3d(64, 512, kernel_size=(1, 1, 1), stride=(1, 2, 2), output_padding=(0, 1, 1), bias=False) (1): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (2): Rearrange('b c t h w -> (b t) c h w') ) ) ) (layer3): Sequential( (0): Bottleneck( (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False) (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (3): Bottleneck( (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (4): Bottleneck( (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (5): Bottleneck( (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (layer4): Sequential( (0): Bottleneck( (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False) (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False) (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (avgpool): AdaptiveAvgPool2d(output_size=(1, 1)) (fc1): Dropout(p=0.5, inplace=False) ) (new_fc): Conv1d(2048, 174, kernel_size=(1,), stride=(1,)) (consensus): ConsensusModule() ) ) ------------------------------------ group: first_conv_weight has 1 params, lr_mult: 1, decay_mult: 1 group: first_conv_bias has 0 params, lr_mult: 2, decay_mult: 0 group: normal_weight has 62 params, lr_mult: 1, decay_mult: 1 group: normal_bias has 0 params, lr_mult: 2, decay_mult: 0 group: BN scale/shift has 126 params, lr_mult: 1, decay_mult: 0 group: custom_ops has 0 params, lr_mult: 1, decay_mult: 1 group: lr5_weight has 1 params, lr_mult: 1, decay_mult: 1 group: lr10_bias has 1 params, lr_mult: 2, decay_mult: 0 No BN layer Freezing. Epoch: [0][0/1345], lr: 0.01000 Time 28.876 (28.876) Data 3.331 (3.331) Loss 5.4941 (5.4941) Prec@1 0.000 (0.000) Prec@5 1.562 (1.562) Epoch: [0][20/1345], lr: 0.01000 Time 0.713 (2.126) Data 0.000 (0.159) Loss 5.2757 (5.2907) Prec@1 0.000 (1.339) Prec@5 1.562 (3.199) Epoch: [0][40/1345], lr: 0.01000 Time 0.901 (1.471) Data 0.001 (0.082) Loss 5.1611 (5.2493) Prec@1 1.562 (1.296) Prec@5 3.125 (4.192) Epoch: [0][60/1345], lr: 0.01000 Time 0.807 (1.239) Data 0.000 (0.055) Loss 5.1402 (5.2094) Prec@1 1.562 (1.332) Prec@5 6.250 (5.174) Epoch: [0][80/1345], lr: 0.01000 Time 0.775 (1.122) Data 0.000 (0.042) Loss 5.0733 (5.1698) Prec@1 3.125 (1.640) Prec@5 7.812 (6.154) Epoch: [0][100/1345], lr: 0.01000 Time 0.709 (1.048) Data 0.001 (0.033) Loss 4.8413 (5.1292) Prec@1 3.125 (2.073) Prec@5 14.062 (7.147) Epoch: [0][120/1345], lr: 0.01000 Time 0.771 (0.996) Data 0.000 (0.028) Loss 4.6747 (5.0911) Prec@1 0.000 (2.105) Prec@5 15.625 (7.851) Epoch: [0][140/1345], lr: 0.01000 Time 0.866 (0.964) Data 0.000 (0.024) Loss 4.6141 (5.0548) Prec@1 6.250 (2.460) Prec@5 17.188 (8.743) Epoch: [0][160/1345], lr: 0.01000 Time 0.789 (0.936) Data 0.000 (0.021) Loss 4.8171 (5.0224) Prec@1 4.688 (2.737) Prec@5 14.062 (9.346) Epoch: [0][180/1345], lr: 0.01000 Time 0.806 (0.916) Data 0.000 (0.019) Loss 4.8163 (4.9961) Prec@1 1.562 (2.935) Prec@5 10.938 (9.893) Epoch: [0][200/1345], lr: 0.01000 Time 0.728 (0.899) Data 0.000 (0.017) Loss 4.5277 (4.9657) Prec@1 4.688 (3.203) Prec@5 25.000 (10.533) Epoch: [0][220/1345], lr: 0.01000 Time 0.792 (0.885) Data 0.000 (0.015) Loss 4.4475 (4.9398) Prec@1 4.688 (3.358) Prec@5 28.125 (11.185) Epoch: [0][240/1345], lr: 0.01000 Time 0.711 (0.875) Data 0.000 (0.014) Loss 4.6880 (4.9129) Prec@1 1.562 (3.533) Prec@5 15.625 (11.741) Epoch: [0][260/1345], lr: 0.01000 Time 0.717 (0.866) Data 0.001 (0.013) Loss 4.5183 (4.8913) Prec@1 3.125 (3.790) Prec@5 17.188 (12.201) Epoch: [0][280/1345], lr: 0.01000 Time 0.711 (0.859) Data 0.000 (0.012) Loss 4.6928 (4.8728) Prec@1 6.250 (3.959) Prec@5 14.062 (12.711) Epoch: [0][300/1345], lr: 0.01000 Time 0.843 (0.852) Data 0.000 (0.011) Loss 4.4414 (4.8497) Prec@1 9.375 (4.215) Prec@5 28.125 (13.382) Epoch: [0][320/1345], lr: 0.01000 Time 0.847 (0.846) Data 0.000 (0.011) Loss 4.6040 (4.8334) Prec@1 10.938 (4.366) Prec@5 23.438 (13.805) Epoch: [0][340/1345], lr: 0.01000 Time 0.707 (0.842) Data 0.000 (0.010) Loss 4.4990 (4.8119) Prec@1 7.812 (4.555) Prec@5 18.750 (14.296) Epoch: [0][360/1345], lr: 0.01000 Time 0.742 (0.836) Data 0.001 (0.010) Loss 4.4180 (4.7950) Prec@1 14.062 (4.692) Prec@5 31.250 (14.733) Epoch: [0][380/1345], lr: 0.01000 Time 0.752 (0.832) Data 0.000 (0.009) Loss 4.3885 (4.7786) Prec@1 6.250 (4.839) Prec@5 20.312 (15.125) Epoch: [0][400/1345], lr: 0.01000 Time 0.708 (0.827) Data 0.000 (0.009) Loss 4.6899 (4.7613) Prec@1 4.688 (5.042) Prec@5 10.938 (15.547) Epoch: [0][420/1345], lr: 0.01000 Time 0.783 (0.824) Data 0.000 (0.008) Loss 4.2045 (4.7449) Prec@1 6.250 (5.151) Prec@5 29.688 (15.937) Epoch: [0][440/1345], lr: 0.01000 Time 0.784 (0.820) Data 0.000 (0.008) Loss 4.3139 (4.7281) Prec@1 10.938 (5.357) Prec@5 26.562 (16.366) Epoch: [0][460/1345], lr: 0.01000 Time 0.711 (0.817) Data 0.000 (0.008) Loss 4.7952 (4.7167) Prec@1 6.250 (5.464) Prec@5 17.188 (16.642) Epoch: [0][480/1345], lr: 0.01000 Time 0.712 (0.815) Data 0.000 (0.007) Loss 4.3366 (4.7005) Prec@1 7.812 (5.610) Prec@5 26.562 (17.035) Epoch: [0][500/1345], lr: 0.01000 Time 0.769 (0.812) Data 0.000 (0.007) Loss 4.4811 (4.6837) Prec@1 7.812 (5.813) Prec@5 21.875 (17.478) Epoch: [0][520/1345], lr: 0.01000 Time 0.710 (0.810) Data 0.000 (0.007) Loss 4.4894 (4.6702) Prec@1 15.625 (5.983) Prec@5 29.688 (17.832) Epoch: [0][540/1345], lr: 0.01000 Time 0.707 (0.808) Data 0.000 (0.007) Loss 4.2341 (4.6553) Prec@1 10.938 (6.140) Prec@5 26.562 (18.239) Epoch: [0][560/1345], lr: 0.01000 Time 0.790 (0.806) Data 0.000 (0.006) Loss 4.2889 (4.6418) Prec@1 10.938 (6.317) Prec@5 32.812 (18.616) Epoch: [0][580/1345], lr: 0.01000 Time 0.733 (0.804) Data 0.000 (0.006) Loss 4.2684 (4.6293) Prec@1 6.250 (6.446) Prec@5 29.688 (18.960) Epoch: [0][600/1345], lr: 0.01000 Time 0.709 (0.802) Data 0.000 (0.006) Loss 4.2806 (4.6163) Prec@1 14.062 (6.562) Prec@5 29.688 (19.280) Epoch: [0][620/1345], lr: 0.01000 Time 0.753 (0.800) Data 0.001 (0.006) Loss 4.2039 (4.6002) Prec@1 9.375 (6.763) Prec@5 29.688 (19.696) Epoch: [0][640/1345], lr: 0.01000 Time 0.722 (0.799) Data 0.000 (0.006) Loss 4.0877 (4.5871) Prec@1 12.500 (6.894) Prec@5 29.688 (20.093) Epoch: [0][660/1345], lr: 0.01000 Time 0.782 (0.797) Data 0.000 (0.005) Loss 4.0313 (4.5754) Prec@1 10.938 (7.035) Prec@5 32.812 (20.417) Epoch: [0][680/1345], lr: 0.01000 Time 0.709 (0.796) Data 0.000 (0.005) Loss 4.2304 (4.5623) Prec@1 12.500 (7.177) Prec@5 25.000 (20.746) Epoch: [0][700/1345], lr: 0.01000 Time 0.824 (0.794) Data 0.000 (0.005) Loss 4.3038 (4.5474) Prec@1 9.375 (7.331) Prec@5 28.125 (21.148) Epoch: [0][720/1345], lr: 0.01000 Time 0.778 (0.793) Data 0.000 (0.005) Loss 3.6216 (4.5324) Prec@1 23.438 (7.548) Prec@5 43.750 (21.489) Epoch: [0][740/1345], lr: 0.01000 Time 0.711 (0.792) Data 0.000 (0.005) Loss 4.0707 (4.5202) Prec@1 17.188 (7.688) Prec@5 34.375 (21.755) Epoch: [0][760/1345], lr: 0.01000 Time 0.709 (0.791) Data 0.000 (0.005) Loss 4.1377 (4.5078) Prec@1 15.625 (7.817) Prec@5 32.812 (22.021) Epoch: [0][780/1345], lr: 0.01000 Time 0.710 (0.790) Data 0.000 (0.005) Loss 4.2503 (4.4955) Prec@1 14.062 (7.959) Prec@5 29.688 (22.331) Epoch: [0][800/1345], lr: 0.01000 Time 0.710 (0.789) Data 0.000 (0.005) Loss 3.8286 (4.4841) Prec@1 21.875 (8.086) Prec@5 35.938 (22.618) Epoch: [0][820/1345], lr: 0.01000 Time 0.745 (0.788) Data 0.000 (0.004) Loss 3.6955 (4.4705) Prec@1 14.062 (8.222) Prec@5 48.438 (22.967) Epoch: [0][840/1345], lr: 0.01000 Time 0.719 (0.787) Data 0.000 (0.004) Loss 3.8157 (4.4567) Prec@1 15.625 (8.383) Prec@5 39.062 (23.285) Epoch: [0][860/1345], lr: 0.01000 Time 0.734 (0.786) Data 0.000 (0.004) Loss 3.7531 (4.4422) Prec@1 20.312 (8.587) Prec@5 40.625 (23.690) Epoch: [0][880/1345], lr: 0.01000 Time 0.708 (0.785) Data 0.000 (0.004) Loss 3.7257 (4.4293) Prec@1 18.750 (8.776) Prec@5 37.500 (24.058) Epoch: [0][900/1345], lr: 0.01000 Time 0.739 (0.785) Data 0.000 (0.004) Loss 3.7975 (4.4173) Prec@1 18.750 (8.966) Prec@5 37.500 (24.391) Epoch: [0][920/1345], lr: 0.01000 Time 0.751 (0.784) Data 0.000 (0.004) Loss 3.9090 (4.4047) Prec@1 10.938 (9.127) Prec@5 25.000 (24.737) Epoch: [0][940/1345], lr: 0.01000 Time 0.779 (0.783) Data 0.000 (0.004) Loss 4.1917 (4.3929) Prec@1 17.188 (9.305) Prec@5 34.375 (25.051) Epoch: [0][960/1345], lr: 0.01000 Time 0.710 (0.783) Data 0.000 (0.004) Loss 3.9469 (4.3830) Prec@1 10.938 (9.404) Prec@5 40.625 (25.304) Epoch: [0][980/1345], lr: 0.01000 Time 0.710 (0.782) Data 0.000 (0.004) Loss 3.8929 (4.3720) Prec@1 9.375 (9.534) Prec@5 34.375 (25.575) Epoch: [0][1000/1345], lr: 0.01000 Time 0.707 (0.782) Data 0.000 (0.004) Loss 3.8066 (4.3604) Prec@1 21.875 (9.683) Prec@5 43.750 (25.887) Epoch: [0][1020/1345], lr: 0.01000 Time 0.738 (0.781) Data 0.001 (0.004) Loss 3.8348 (4.3467) Prec@1 20.312 (9.852) Prec@5 34.375 (26.221) Epoch: [0][1040/1345], lr: 0.01000 Time 0.807 (0.781) Data 0.001 (0.004) Loss 3.8115 (4.3348) Prec@1 9.375 (9.972) Prec@5 34.375 (26.513) Epoch: [0][1060/1345], lr: 0.01000 Time 0.734 (0.781) Data 0.000 (0.004) Loss 3.4355 (4.3228) Prec@1 18.750 (10.133) Prec@5 45.312 (26.795) Epoch: [0][1080/1345], lr: 0.01000 Time 0.780 (0.780) Data 0.000 (0.003) Loss 3.6384 (4.3116) Prec@1 26.562 (10.290) Prec@5 42.188 (27.079) Epoch: [0][1100/1345], lr: 0.01000 Time 0.740 (0.780) Data 0.000 (0.003) Loss 3.7223 (4.3003) Prec@1 17.188 (10.432) Prec@5 46.875 (27.353) Epoch: [0][1120/1345], lr: 0.01000 Time 0.748 (0.779) Data 0.000 (0.003) Loss 3.6529 (4.2878) Prec@1 18.750 (10.599) Prec@5 42.188 (27.661) Epoch: [0][1140/1345], lr: 0.01000 Time 0.710 (0.779) Data 0.000 (0.003) Loss 3.8726 (4.2758) Prec@1 14.062 (10.780) Prec@5 43.750 (27.959) Epoch: [0][1160/1345], lr: 0.01000 Time 0.827 (0.779) Data 0.001 (0.003) Loss 3.9721 (4.2641) Prec@1 12.500 (10.951) Prec@5 32.812 (28.276) Epoch: [0][1180/1345], lr: 0.01000 Time 0.708 (0.778) Data 0.000 (0.003) Loss 3.6024 (4.2529) Prec@1 15.625 (11.088) Prec@5 48.438 (28.555) Epoch: [0][1200/1345], lr: 0.01000 Time 0.740 (0.777) Data 0.000 (0.003) Loss 3.5945 (4.2413) Prec@1 20.312 (11.234) Prec@5 50.000 (28.865) Epoch: [0][1220/1345], lr: 0.01000 Time 0.717 (0.777) Data 0.000 (0.003) Loss 3.8212 (4.2304) Prec@1 12.500 (11.383) Prec@5 42.188 (29.144) Epoch: [0][1240/1345], lr: 0.01000 Time 0.748 (0.776) Data 0.000 (0.003) Loss 3.3440 (4.2187) Prec@1 26.562 (11.542) Prec@5 53.125 (29.438) Epoch: [0][1260/1345], lr: 0.01000 Time 0.710 (0.776) Data 0.000 (0.003) Loss 3.7277 (4.2063) Prec@1 17.188 (11.733) Prec@5 51.562 (29.759) Epoch: [0][1280/1345], lr: 0.01000 Time 0.749 (0.776) Data 0.000 (0.003) Loss 3.5266 (4.1956) Prec@1 25.000 (11.897) Prec@5 54.688 (30.003) Epoch: [0][1300/1345], lr: 0.01000 Time 0.728 (0.775) Data 0.000 (0.003) Loss 3.7503 (4.1847) Prec@1 18.750 (12.048) Prec@5 39.062 (30.274) Epoch: [0][1320/1345], lr: 0.01000 Time 0.779 (0.775) Data 0.000 (0.003) Loss 3.4535 (4.1749) Prec@1 17.188 (12.185) Prec@5 51.562 (30.517) Epoch: [0][1340/1345], lr: 0.01000 Time 0.708 (0.775) Data 0.000 (0.003) Loss 3.3386 (4.1654) Prec@1 26.562 (12.319) Prec@5 53.125 (30.779) No BN layer Freezing. Test: [0/181] Time 3.165 (3.1649) Loss 3.8634 (3.8634) Prec@1 20.312 (20.312) Prec@5 39.062 (39.062) Test: [20/181] Time 0.668 (0.5805) Loss 3.6146 (3.8156) Prec@1 23.438 (18.229) Prec@5 43.750 (41.964) Test: [40/181] Time 0.749 (0.5193) Loss 3.6582 (3.8407) Prec@1 18.750 (18.255) Prec@5 43.750 (42.302) Test: [60/181] Time 0.684 (0.4960) Loss 4.0057 (3.8754) Prec@1 15.625 (17.700) Prec@5 31.250 (41.624) Test: [80/181] Time 0.711 (0.4849) Loss 3.5921 (3.8932) Prec@1 20.312 (17.149) Prec@5 46.875 (41.184) Test: [100/181] Time 0.736 (0.4779) Loss 4.8620 (3.9010) Prec@1 17.188 (17.203) Prec@5 34.375 (40.873) Test: [120/181] Time 0.771 (0.4736) Loss 3.9755 (3.8908) Prec@1 14.062 (17.110) Prec@5 40.625 (40.987) Test: [140/181] Time 0.608 (0.4692) Loss 4.2310 (3.8768) Prec@1 9.375 (17.232) Prec@5 29.688 (41.246) Test: [160/181] Time 0.556 (0.4668) Loss 4.3412 (3.8660) Prec@1 10.938 (17.372) Prec@5 29.688 (41.508) Testing Results: Prec@1 17.604 Prec@5 41.832 Loss 3.85425 Time 0.4640 No BN layer Freezing. Epoch: [1][0/1345], lr: 0.01000 Time 4.483 (4.483) Data 3.736 (3.736) Loss 3.6082 (3.6082) Prec@1 21.875 (21.875) Prec@5 40.625 (40.625) Epoch: [1][20/1345], lr: 0.01000 Time 0.711 (0.918) Data 0.000 (0.178) Loss 3.3795 (3.3912) Prec@1 28.125 (24.926) Prec@5 51.562 (49.033) Epoch: [1][40/1345], lr: 0.01000 Time 0.735 (0.835) Data 0.000 (0.092) Loss 3.0036 (3.3816) Prec@1 28.125 (24.886) Prec@5 62.500 (49.123) Epoch: [1][60/1345], lr: 0.01000 Time 0.785 (0.802) Data 0.000 (0.062) Loss 3.4584 (3.3811) Prec@1 21.875 (23.924) Prec@5 46.875 (49.129) Epoch: [1][80/1345], lr: 0.01000 Time 0.728 (0.787) Data 0.000 (0.047) Loss 3.0594 (3.3691) Prec@1 25.000 (23.881) Prec@5 60.938 (49.460) Epoch: [1][100/1345], lr: 0.01000 Time 0.712 (0.778) Data 0.000 (0.037) Loss 3.1475 (3.3672) Prec@1 26.562 (24.134) Prec@5 48.438 (49.536) Epoch: [1][120/1345], lr: 0.01000 Time 0.711 (0.770) Data 0.000 (0.031) Loss 3.3448 (3.3489) Prec@1 17.188 (24.186) Prec@5 51.562 (50.026) Epoch: [1][140/1345], lr: 0.01000 Time 0.720 (0.766) Data 0.000 (0.027) Loss 2.8559 (3.3324) Prec@1 34.375 (24.313) Prec@5 65.625 (50.399) Epoch: [1][160/1345], lr: 0.01000 Time 0.711 (0.764) Data 0.000 (0.024) Loss 2.9915 (3.3286) Prec@1 29.688 (24.292) Prec@5 60.938 (50.359) Epoch: [1][180/1345], lr: 0.01000 Time 0.708 (0.761) Data 0.000 (0.021) Loss 3.2733 (3.3292) Prec@1 28.125 (24.413) Prec@5 50.000 (50.285) Epoch: [1][200/1345], lr: 0.01000 Time 0.711 (0.758) Data 0.000 (0.019) Loss 3.4678 (3.3322) Prec@1 14.062 (24.347) Prec@5 48.438 (50.389) Epoch: [1][220/1345], lr: 0.01000 Time 0.729 (0.755) Data 0.000 (0.017) Loss 3.6906 (3.3260) Prec@1 23.438 (24.399) Prec@5 42.188 (50.452) Epoch: [1][240/1345], lr: 0.01000 Time 0.712 (0.754) Data 0.000 (0.016) Loss 3.1043 (3.3170) Prec@1 26.562 (24.501) Prec@5 53.125 (50.739) Epoch: [1][260/1345], lr: 0.01000 Time 0.709 (0.753) Data 0.000 (0.015) Loss 3.0566 (3.3081) Prec@1 28.125 (24.689) Prec@5 60.938 (50.988) Epoch: [1][280/1345], lr: 0.01000 Time 0.709 (0.752) Data 0.001 (0.014) Loss 3.3580 (3.2999) Prec@1 21.875 (24.833) Prec@5 46.875 (51.207) Epoch: [1][300/1345], lr: 0.01000 Time 0.733 (0.751) Data 0.000 (0.013) Loss 2.9787 (3.2934) Prec@1 32.812 (25.073) Prec@5 56.250 (51.402) Epoch: [1][320/1345], lr: 0.01000 Time 0.737 (0.752) Data 0.000 (0.012) Loss 3.3411 (3.2896) Prec@1 21.875 (25.156) Prec@5 56.250 (51.548) Epoch: [1][340/1345], lr: 0.01000 Time 0.849 (0.752) Data 0.000 (0.011) Loss 3.2080 (3.2848) Prec@1 28.125 (25.357) Prec@5 57.812 (51.782) Epoch: [1][360/1345], lr: 0.01000 Time 0.845 (0.751) Data 0.000 (0.011) Loss 2.9828 (3.2751) Prec@1 37.500 (25.550) Prec@5 54.688 (52.065) Epoch: [1][380/1345], lr: 0.01000 Time 0.711 (0.751) Data 0.000 (0.010) Loss 3.1836 (3.2679) Prec@1 20.312 (25.693) Prec@5 46.875 (52.215) Epoch: [1][400/1345], lr: 0.01000 Time 0.732 (0.751) Data 0.000 (0.010) Loss 2.9296 (3.2603) Prec@1 29.688 (25.807) Prec@5 57.812 (52.353) Epoch: [1][420/1345], lr: 0.01000 Time 0.730 (0.750) Data 0.000 (0.009) Loss 2.9816 (3.2561) Prec@1 29.688 (25.835) Prec@5 60.938 (52.420) Epoch: [1][440/1345], lr: 0.01000 Time 0.760 (0.750) Data 0.000 (0.009) Loss 2.8733 (3.2528) Prec@1 26.562 (25.932) Prec@5 59.375 (52.537) Epoch: [1][460/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.009) Loss 2.7321 (3.2483) Prec@1 37.500 (25.990) Prec@5 64.062 (52.559) Epoch: [1][480/1345], lr: 0.01000 Time 0.835 (0.749) Data 0.000 (0.008) Loss 3.1540 (3.2419) Prec@1 31.250 (26.111) Prec@5 54.688 (52.680) Epoch: [1][500/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.008) Loss 2.8962 (3.2318) Prec@1 31.250 (26.263) Prec@5 64.062 (52.916) Epoch: [1][520/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.008) Loss 2.9658 (3.2250) Prec@1 34.375 (26.353) Prec@5 62.500 (53.026) Epoch: [1][540/1345], lr: 0.01000 Time 0.737 (0.748) Data 0.000 (0.007) Loss 3.0901 (3.2197) Prec@1 28.125 (26.493) Prec@5 57.812 (53.186) Epoch: [1][560/1345], lr: 0.01000 Time 0.766 (0.748) Data 0.000 (0.007) Loss 3.2358 (3.2138) Prec@1 21.875 (26.554) Prec@5 53.125 (53.314) Epoch: [1][580/1345], lr: 0.01000 Time 0.731 (0.748) Data 0.000 (0.007) Loss 3.2678 (3.2063) Prec@1 25.000 (26.700) Prec@5 51.562 (53.496) Epoch: [1][600/1345], lr: 0.01000 Time 0.763 (0.748) Data 0.000 (0.007) Loss 3.0510 (3.2003) Prec@1 34.375 (26.796) Prec@5 60.938 (53.642) Epoch: [1][620/1345], lr: 0.01000 Time 0.738 (0.748) Data 0.000 (0.006) Loss 3.4683 (3.1970) Prec@1 9.375 (26.857) Prec@5 46.875 (53.734) Epoch: [1][640/1345], lr: 0.01000 Time 0.718 (0.747) Data 0.001 (0.006) Loss 3.0509 (3.1919) Prec@1 29.688 (26.987) Prec@5 60.938 (53.820) Epoch: [1][660/1345], lr: 0.01000 Time 0.754 (0.747) Data 0.000 (0.006) Loss 3.2931 (3.1893) Prec@1 28.125 (26.995) Prec@5 54.688 (53.926) Epoch: [1][680/1345], lr: 0.01000 Time 0.842 (0.747) Data 0.000 (0.006) Loss 2.8233 (3.1836) Prec@1 35.938 (27.104) Prec@5 64.062 (54.079) Epoch: [1][700/1345], lr: 0.01000 Time 0.739 (0.747) Data 0.000 (0.006) Loss 2.5939 (3.1792) Prec@1 35.938 (27.162) Prec@5 68.750 (54.217) Epoch: [1][720/1345], lr: 0.01000 Time 0.713 (0.747) Data 0.000 (0.006) Loss 3.1695 (3.1749) Prec@1 29.688 (27.226) Prec@5 56.250 (54.319) Epoch: [1][740/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.005) Loss 3.4898 (3.1703) Prec@1 23.438 (27.290) Prec@5 48.438 (54.415) Epoch: [1][760/1345], lr: 0.01000 Time 0.711 (0.747) Data 0.000 (0.005) Loss 2.8549 (3.1692) Prec@1 32.812 (27.328) Prec@5 60.938 (54.429) Epoch: [1][780/1345], lr: 0.01000 Time 0.713 (0.747) Data 0.000 (0.005) Loss 2.8617 (3.1640) Prec@1 32.812 (27.431) Prec@5 59.375 (54.539) Epoch: [1][800/1345], lr: 0.01000 Time 0.741 (0.747) Data 0.001 (0.005) Loss 3.1145 (3.1573) Prec@1 34.375 (27.577) Prec@5 62.500 (54.693) Epoch: [1][820/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.005) Loss 3.0978 (3.1522) Prec@1 25.000 (27.621) Prec@5 50.000 (54.847) Epoch: [1][840/1345], lr: 0.01000 Time 0.733 (0.746) Data 0.000 (0.005) Loss 2.9846 (3.1484) Prec@1 28.125 (27.659) Prec@5 56.250 (54.914) Epoch: [1][860/1345], lr: 0.01000 Time 0.711 (0.746) Data 0.000 (0.005) Loss 2.7176 (3.1454) Prec@1 32.812 (27.680) Prec@5 62.500 (55.003) Epoch: [1][880/1345], lr: 0.01000 Time 0.716 (0.746) Data 0.000 (0.005) Loss 3.0275 (3.1415) Prec@1 20.312 (27.772) Prec@5 59.375 (55.101) Epoch: [1][900/1345], lr: 0.01000 Time 0.751 (0.746) Data 0.000 (0.005) Loss 3.0815 (3.1360) Prec@1 29.688 (27.861) Prec@5 59.375 (55.229) Epoch: [1][920/1345], lr: 0.01000 Time 0.750 (0.746) Data 0.001 (0.005) Loss 2.9211 (3.1306) Prec@1 29.688 (27.962) Prec@5 60.938 (55.380) Epoch: [1][940/1345], lr: 0.01000 Time 0.853 (0.746) Data 0.001 (0.004) Loss 3.3928 (3.1275) Prec@1 29.688 (28.050) Prec@5 54.688 (55.456) Epoch: [1][960/1345], lr: 0.01000 Time 0.713 (0.746) Data 0.000 (0.004) Loss 2.7507 (3.1218) Prec@1 34.375 (28.177) Prec@5 59.375 (55.582) Epoch: [1][980/1345], lr: 0.01000 Time 0.773 (0.746) Data 0.000 (0.004) Loss 2.9553 (3.1179) Prec@1 32.812 (28.224) Prec@5 57.812 (55.675) Epoch: [1][1000/1345], lr: 0.01000 Time 0.725 (0.746) Data 0.001 (0.004) Loss 3.0230 (3.1144) Prec@1 26.562 (28.275) Prec@5 60.938 (55.771) Epoch: [1][1020/1345], lr: 0.01000 Time 0.739 (0.746) Data 0.000 (0.004) Loss 3.3155 (3.1107) Prec@1 29.688 (28.353) Prec@5 51.562 (55.858) Epoch: [1][1040/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.001 (0.004) Loss 3.3256 (3.1071) Prec@1 26.562 (28.407) Prec@5 51.562 (55.918) Epoch: [1][1060/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 3.2127 (3.1039) Prec@1 26.562 (28.474) Prec@5 53.125 (55.969) Epoch: [1][1080/1345], lr: 0.01000 Time 0.772 (0.745) Data 0.000 (0.004) Loss 3.0067 (3.1004) Prec@1 28.125 (28.540) Prec@5 59.375 (56.039) Epoch: [1][1100/1345], lr: 0.01000 Time 0.834 (0.745) Data 0.000 (0.004) Loss 2.9375 (3.0954) Prec@1 31.250 (28.623) Prec@5 57.812 (56.162) Epoch: [1][1120/1345], lr: 0.01000 Time 0.751 (0.745) Data 0.000 (0.004) Loss 2.8675 (3.0899) Prec@1 32.812 (28.705) Prec@5 67.188 (56.279) Epoch: [1][1140/1345], lr: 0.01000 Time 0.741 (0.745) Data 0.001 (0.004) Loss 3.0254 (3.0867) Prec@1 28.125 (28.762) Prec@5 51.562 (56.343) Epoch: [1][1160/1345], lr: 0.01000 Time 0.826 (0.745) Data 0.000 (0.004) Loss 2.9002 (3.0842) Prec@1 31.250 (28.809) Prec@5 62.500 (56.401) Epoch: [1][1180/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.004) Loss 2.9505 (3.0792) Prec@1 34.375 (28.906) Prec@5 53.125 (56.476) Epoch: [1][1200/1345], lr: 0.01000 Time 0.715 (0.745) Data 0.000 (0.004) Loss 2.8899 (3.0766) Prec@1 31.250 (28.956) Prec@5 65.625 (56.543) Epoch: [1][1220/1345], lr: 0.01000 Time 0.816 (0.745) Data 0.000 (0.004) Loss 3.0316 (3.0725) Prec@1 32.812 (29.028) Prec@5 56.250 (56.645) Epoch: [1][1240/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.003) Loss 2.9518 (3.0693) Prec@1 31.250 (29.108) Prec@5 59.375 (56.703) Epoch: [1][1260/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.003) Loss 2.4469 (3.0649) Prec@1 40.625 (29.173) Prec@5 73.438 (56.811) Epoch: [1][1280/1345], lr: 0.01000 Time 0.847 (0.744) Data 0.000 (0.003) Loss 2.7987 (3.0610) Prec@1 37.500 (29.222) Prec@5 62.500 (56.889) Epoch: [1][1300/1345], lr: 0.01000 Time 0.730 (0.744) Data 0.000 (0.003) Loss 2.6903 (3.0563) Prec@1 35.938 (29.296) Prec@5 62.500 (56.979) Epoch: [1][1320/1345], lr: 0.01000 Time 0.718 (0.744) Data 0.001 (0.003) Loss 2.7721 (3.0522) Prec@1 31.250 (29.371) Prec@5 59.375 (57.064) Epoch: [1][1340/1345], lr: 0.01000 Time 0.706 (0.744) Data 0.000 (0.003) Loss 2.9693 (3.0477) Prec@1 37.500 (29.454) Prec@5 59.375 (57.160) No BN layer Freezing. Test: [0/181] Time 3.632 (3.6315) Loss 3.3244 (3.3244) Prec@1 23.438 (23.438) Prec@5 53.125 (53.125) Test: [20/181] Time 1.062 (0.5978) Loss 3.1041 (3.2067) Prec@1 28.125 (27.455) Prec@5 54.688 (56.473) Test: [40/181] Time 0.771 (0.5233) Loss 3.1724 (3.2873) Prec@1 28.125 (27.477) Prec@5 51.562 (55.450) Test: [60/181] Time 0.886 (0.5029) Loss 3.6772 (3.3531) Prec@1 26.562 (26.639) Prec@5 45.312 (54.611) Test: [80/181] Time 0.956 (0.4913) Loss 3.0971 (3.3909) Prec@1 21.875 (25.675) Prec@5 57.812 (54.070) Test: [100/181] Time 0.924 (0.4833) Loss 3.6266 (3.3912) Prec@1 23.438 (25.820) Prec@5 50.000 (54.146) Test: [120/181] Time 0.743 (0.4763) Loss 3.7881 (3.3824) Prec@1 29.688 (25.749) Prec@5 51.562 (54.520) Test: [140/181] Time 0.981 (0.4754) Loss 3.9756 (3.3660) Prec@1 17.188 (25.809) Prec@5 45.312 (54.676) Test: [160/181] Time 0.845 (0.4725) Loss 3.6692 (3.3504) Prec@1 17.188 (26.068) Prec@5 48.438 (55.047) Testing Results: Prec@1 26.345 Prec@5 55.278 Loss 3.33540 Time 0.4693 No BN layer Freezing. Epoch: [2][0/1345], lr: 0.01000 Time 4.083 (4.083) Data 3.325 (3.325) Loss 2.7284 (2.7284) Prec@1 32.812 (32.812) Prec@5 65.625 (65.625) Epoch: [2][20/1345], lr: 0.01000 Time 0.710 (0.921) Data 0.000 (0.159) Loss 2.4926 (2.7078) Prec@1 40.625 (35.789) Prec@5 65.625 (63.988) Epoch: [2][40/1345], lr: 0.01000 Time 0.724 (0.834) Data 0.000 (0.082) Loss 2.5201 (2.7350) Prec@1 39.062 (35.023) Prec@5 68.750 (63.415) Epoch: [2][60/1345], lr: 0.01000 Time 0.710 (0.803) Data 0.000 (0.055) Loss 2.7654 (2.7260) Prec@1 40.625 (35.092) Prec@5 68.750 (63.678) Epoch: [2][80/1345], lr: 0.01000 Time 0.732 (0.791) Data 0.000 (0.041) Loss 2.9942 (2.7246) Prec@1 25.000 (35.282) Prec@5 65.625 (63.812) Epoch: [2][100/1345], lr: 0.01000 Time 0.715 (0.777) Data 0.001 (0.033) Loss 2.6561 (2.7183) Prec@1 37.500 (35.613) Prec@5 65.625 (64.217) Epoch: [2][120/1345], lr: 0.01000 Time 0.783 (0.769) Data 0.000 (0.028) Loss 2.6727 (2.7169) Prec@1 29.688 (35.460) Prec@5 68.750 (64.050) Epoch: [2][140/1345], lr: 0.01000 Time 0.732 (0.763) Data 0.000 (0.024) Loss 2.3500 (2.7006) Prec@1 40.625 (35.982) Prec@5 75.000 (64.384) Epoch: [2][160/1345], lr: 0.01000 Time 0.722 (0.759) Data 0.001 (0.021) Loss 2.7864 (2.7035) Prec@1 31.250 (36.044) Prec@5 64.062 (64.237) Epoch: [2][180/1345], lr: 0.01000 Time 0.709 (0.756) Data 0.000 (0.019) Loss 2.3941 (2.7159) Prec@1 50.000 (35.679) Prec@5 73.438 (64.114) Epoch: [2][200/1345], lr: 0.01000 Time 0.710 (0.753) Data 0.000 (0.017) Loss 2.7496 (2.7148) Prec@1 32.812 (35.743) Prec@5 71.875 (64.101) Epoch: [2][220/1345], lr: 0.01000 Time 0.729 (0.752) Data 0.000 (0.015) Loss 3.0493 (2.7136) Prec@1 37.500 (35.633) Prec@5 64.062 (64.197) Epoch: [2][240/1345], lr: 0.01000 Time 0.737 (0.750) Data 0.000 (0.014) Loss 2.7815 (2.7089) Prec@1 31.250 (35.672) Prec@5 51.562 (64.179) Epoch: [2][260/1345], lr: 0.01000 Time 0.830 (0.751) Data 0.000 (0.013) Loss 2.7476 (2.7130) Prec@1 26.562 (35.728) Prec@5 65.625 (64.051) Epoch: [2][280/1345], lr: 0.01000 Time 0.840 (0.750) Data 0.000 (0.012) Loss 2.4655 (2.7079) Prec@1 37.500 (35.876) Prec@5 70.312 (64.157) Epoch: [2][300/1345], lr: 0.01000 Time 0.733 (0.750) Data 0.000 (0.011) Loss 2.5962 (2.7010) Prec@1 40.625 (36.140) Prec@5 67.188 (64.358) Epoch: [2][320/1345], lr: 0.01000 Time 0.731 (0.750) Data 0.000 (0.011) Loss 2.9605 (2.7002) Prec@1 32.812 (36.132) Prec@5 59.375 (64.262) Epoch: [2][340/1345], lr: 0.01000 Time 0.725 (0.751) Data 0.000 (0.010) Loss 2.2858 (2.6971) Prec@1 37.500 (36.135) Prec@5 76.562 (64.347) Epoch: [2][360/1345], lr: 0.01000 Time 0.715 (0.750) Data 0.000 (0.010) Loss 2.4873 (2.6919) Prec@1 39.062 (36.111) Prec@5 71.875 (64.487) Epoch: [2][380/1345], lr: 0.01000 Time 0.768 (0.749) Data 0.000 (0.009) Loss 2.6563 (2.6884) Prec@1 37.500 (36.184) Prec@5 60.938 (64.596) Epoch: [2][400/1345], lr: 0.01000 Time 0.711 (0.750) Data 0.000 (0.009) Loss 2.4493 (2.6870) Prec@1 42.188 (36.175) Prec@5 67.188 (64.565) Epoch: [2][420/1345], lr: 0.01000 Time 0.742 (0.750) Data 0.000 (0.008) Loss 2.6294 (2.6862) Prec@1 37.500 (36.301) Prec@5 67.188 (64.604) Epoch: [2][440/1345], lr: 0.01000 Time 0.773 (0.750) Data 0.000 (0.008) Loss 2.5901 (2.6879) Prec@1 35.938 (36.263) Prec@5 70.312 (64.530) Epoch: [2][460/1345], lr: 0.01000 Time 0.711 (0.750) Data 0.000 (0.008) Loss 2.1962 (2.6857) Prec@1 42.188 (36.293) Prec@5 75.000 (64.530) Epoch: [2][480/1345], lr: 0.01000 Time 0.758 (0.749) Data 0.000 (0.007) Loss 2.1993 (2.6831) Prec@1 50.000 (36.360) Prec@5 73.438 (64.579) Epoch: [2][500/1345], lr: 0.01000 Time 0.727 (0.749) Data 0.000 (0.007) Loss 2.7159 (2.6828) Prec@1 42.188 (36.327) Prec@5 67.188 (64.527) Epoch: [2][520/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.007) Loss 2.7778 (2.6785) Prec@1 29.688 (36.333) Prec@5 67.188 (64.629) Epoch: [2][540/1345], lr: 0.01000 Time 0.828 (0.749) Data 0.000 (0.007) Loss 2.1865 (2.6745) Prec@1 45.312 (36.417) Prec@5 75.000 (64.727) Epoch: [2][560/1345], lr: 0.01000 Time 0.724 (0.749) Data 0.000 (0.006) Loss 2.8593 (2.6716) Prec@1 34.375 (36.472) Prec@5 65.625 (64.781) Epoch: [2][580/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.006) Loss 3.5468 (2.6735) Prec@1 18.750 (36.457) Prec@5 46.875 (64.732) Epoch: [2][600/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.006) Loss 2.3060 (2.6673) Prec@1 40.625 (36.478) Prec@5 68.750 (64.827) Epoch: [2][620/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.006) Loss 1.9231 (2.6654) Prec@1 57.812 (36.491) Prec@5 82.812 (64.885) Epoch: [2][640/1345], lr: 0.01000 Time 0.716 (0.749) Data 0.000 (0.006) Loss 2.4338 (2.6643) Prec@1 43.750 (36.535) Prec@5 68.750 (64.950) Epoch: [2][660/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.001 (0.005) Loss 2.1118 (2.6641) Prec@1 50.000 (36.526) Prec@5 79.688 (65.003) Epoch: [2][680/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.005) Loss 2.8538 (2.6630) Prec@1 32.812 (36.527) Prec@5 57.812 (64.996) Epoch: [2][700/1345], lr: 0.01000 Time 0.803 (0.749) Data 0.000 (0.005) Loss 2.5894 (2.6610) Prec@1 35.938 (36.557) Prec@5 68.750 (65.063) Epoch: [2][720/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.005) Loss 2.7933 (2.6595) Prec@1 29.688 (36.585) Prec@5 67.188 (65.092) Epoch: [2][740/1345], lr: 0.01000 Time 0.737 (0.749) Data 0.000 (0.005) Loss 2.0004 (2.6590) Prec@1 48.438 (36.600) Prec@5 71.875 (65.085) Epoch: [2][760/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.005) Loss 2.8566 (2.6584) Prec@1 32.812 (36.615) Prec@5 59.375 (65.103) Epoch: [2][780/1345], lr: 0.01000 Time 0.722 (0.749) Data 0.000 (0.005) Loss 2.7445 (2.6573) Prec@1 34.375 (36.608) Prec@5 62.500 (65.137) Epoch: [2][800/1345], lr: 0.01000 Time 0.718 (0.748) Data 0.000 (0.005) Loss 2.9520 (2.6569) Prec@1 35.938 (36.620) Prec@5 57.812 (65.159) Epoch: [2][820/1345], lr: 0.01000 Time 0.771 (0.749) Data 0.000 (0.004) Loss 2.8483 (2.6557) Prec@1 34.375 (36.626) Prec@5 64.062 (65.235) Epoch: [2][840/1345], lr: 0.01000 Time 0.871 (0.749) Data 0.000 (0.004) Loss 2.6068 (2.6561) Prec@1 37.500 (36.601) Prec@5 68.750 (65.203) Epoch: [2][860/1345], lr: 0.01000 Time 0.724 (0.748) Data 0.000 (0.004) Loss 2.5616 (2.6555) Prec@1 32.812 (36.629) Prec@5 65.625 (65.211) Epoch: [2][880/1345], lr: 0.01000 Time 0.770 (0.749) Data 0.000 (0.004) Loss 2.3307 (2.6540) Prec@1 45.312 (36.675) Prec@5 73.438 (65.286) Epoch: [2][900/1345], lr: 0.01000 Time 0.857 (0.749) Data 0.000 (0.004) Loss 2.5012 (2.6503) Prec@1 40.625 (36.751) Prec@5 73.438 (65.387) Epoch: [2][920/1345], lr: 0.01000 Time 0.754 (0.748) Data 0.000 (0.004) Loss 2.4062 (2.6493) Prec@1 42.188 (36.765) Prec@5 68.750 (65.421) Epoch: [2][940/1345], lr: 0.01000 Time 0.725 (0.748) Data 0.000 (0.004) Loss 2.6813 (2.6475) Prec@1 32.812 (36.803) Prec@5 67.188 (65.442) Epoch: [2][960/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.004) Loss 2.9884 (2.6459) Prec@1 34.375 (36.845) Prec@5 64.062 (65.487) Epoch: [2][980/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.004) Loss 2.2074 (2.6415) Prec@1 50.000 (36.925) Prec@5 76.562 (65.593) Epoch: [2][1000/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.004) Loss 2.3499 (2.6392) Prec@1 45.312 (36.971) Prec@5 67.188 (65.642) Epoch: [2][1020/1345], lr: 0.01000 Time 0.734 (0.748) Data 0.000 (0.004) Loss 2.4532 (2.6363) Prec@1 43.750 (37.001) Prec@5 70.312 (65.672) Epoch: [2][1040/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.004) Loss 2.4244 (2.6364) Prec@1 40.625 (37.015) Prec@5 70.312 (65.673) Epoch: [2][1060/1345], lr: 0.01000 Time 0.715 (0.748) Data 0.000 (0.004) Loss 2.4741 (2.6335) Prec@1 37.500 (37.083) Prec@5 64.062 (65.725) Epoch: [2][1080/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.004) Loss 2.5536 (2.6313) Prec@1 37.500 (37.085) Prec@5 70.312 (65.761) Epoch: [2][1100/1345], lr: 0.01000 Time 0.775 (0.748) Data 0.000 (0.003) Loss 2.7378 (2.6314) Prec@1 42.188 (37.077) Prec@5 64.062 (65.768) Epoch: [2][1120/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.003) Loss 2.7030 (2.6292) Prec@1 35.938 (37.156) Prec@5 62.500 (65.817) Epoch: [2][1140/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.003) Loss 2.2905 (2.6257) Prec@1 46.875 (37.225) Prec@5 67.188 (65.873) Epoch: [2][1160/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.7800 (2.6229) Prec@1 32.812 (37.242) Prec@5 62.500 (65.935) Epoch: [2][1180/1345], lr: 0.01000 Time 0.774 (0.748) Data 0.001 (0.003) Loss 2.9687 (2.6215) Prec@1 23.438 (37.268) Prec@5 62.500 (65.956) Epoch: [2][1200/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.1163 (2.6204) Prec@1 51.562 (37.306) Prec@5 78.125 (65.998) Epoch: [2][1220/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.4724 (2.6183) Prec@1 39.062 (37.339) Prec@5 65.625 (66.046) Epoch: [2][1240/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.003) Loss 2.2659 (2.6180) Prec@1 29.688 (37.344) Prec@5 78.125 (66.076) Epoch: [2][1260/1345], lr: 0.01000 Time 0.725 (0.747) Data 0.000 (0.003) Loss 2.3529 (2.6166) Prec@1 39.062 (37.361) Prec@5 71.875 (66.074) Epoch: [2][1280/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.003) Loss 2.6682 (2.6151) Prec@1 34.375 (37.376) Prec@5 57.812 (66.106) Epoch: [2][1300/1345], lr: 0.01000 Time 0.764 (0.747) Data 0.000 (0.003) Loss 2.6699 (2.6146) Prec@1 40.625 (37.379) Prec@5 64.062 (66.109) Epoch: [2][1320/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.003) Loss 2.0323 (2.6109) Prec@1 51.562 (37.444) Prec@5 82.812 (66.177) Epoch: [2][1340/1345], lr: 0.01000 Time 0.707 (0.747) Data 0.000 (0.003) Loss 2.7536 (2.6095) Prec@1 32.812 (37.463) Prec@5 62.500 (66.226) No BN layer Freezing. Test: [0/181] Time 3.412 (3.4123) Loss 3.2526 (3.2526) Prec@1 29.688 (29.688) Prec@5 60.938 (60.938) Test: [20/181] Time 0.970 (0.6085) Loss 2.8941 (2.9350) Prec@1 34.375 (32.961) Prec@5 62.500 (61.979) Test: [40/181] Time 0.760 (0.5177) Loss 2.8538 (2.9713) Prec@1 23.438 (32.012) Prec@5 64.062 (61.509) Test: [60/181] Time 0.576 (0.4932) Loss 3.4680 (3.0433) Prec@1 25.000 (30.917) Prec@5 57.812 (60.067) Test: [80/181] Time 0.504 (0.4809) Loss 3.0282 (3.0723) Prec@1 26.562 (30.363) Prec@5 62.500 (59.510) Test: [100/181] Time 0.237 (0.4736) Loss 3.5539 (3.0732) Prec@1 28.125 (30.554) Prec@5 51.562 (59.669) Test: [120/181] Time 0.312 (0.4725) Loss 3.4795 (3.0763) Prec@1 32.812 (30.746) Prec@5 59.375 (59.840) Test: [140/181] Time 0.238 (0.4720) Loss 3.5903 (3.0744) Prec@1 20.312 (30.796) Prec@5 53.125 (59.907) Test: [160/181] Time 0.238 (0.4706) Loss 3.2585 (3.0641) Prec@1 25.000 (30.804) Prec@5 65.625 (60.229) Testing Results: Prec@1 31.207 Prec@5 60.451 Loss 3.05725 Time 0.4700 No BN layer Freezing. Epoch: [3][0/1345], lr: 0.01000 Time 3.834 (3.834) Data 3.086 (3.086) Loss 2.5127 (2.5127) Prec@1 42.188 (42.188) Prec@5 68.750 (68.750) Epoch: [3][20/1345], lr: 0.01000 Time 0.744 (0.890) Data 0.000 (0.147) Loss 2.7157 (2.4722) Prec@1 40.625 (40.104) Prec@5 68.750 (69.792) Epoch: [3][40/1345], lr: 0.01000 Time 0.757 (0.828) Data 0.000 (0.076) Loss 2.3868 (2.3903) Prec@1 45.312 (42.340) Prec@5 70.312 (70.808) Epoch: [3][60/1345], lr: 0.01000 Time 0.770 (0.800) Data 0.001 (0.051) Loss 2.5106 (2.3670) Prec@1 43.750 (42.444) Prec@5 67.188 (71.055) Epoch: [3][80/1345], lr: 0.01000 Time 0.715 (0.786) Data 0.001 (0.039) Loss 2.3861 (2.3889) Prec@1 43.750 (41.647) Prec@5 67.188 (70.409) Epoch: [3][100/1345], lr: 0.01000 Time 0.712 (0.777) Data 0.000 (0.031) Loss 2.2079 (2.3783) Prec@1 40.625 (42.095) Prec@5 76.562 (70.452) Epoch: [3][120/1345], lr: 0.01000 Time 0.854 (0.773) Data 0.000 (0.026) Loss 2.4128 (2.3868) Prec@1 42.188 (41.632) Prec@5 70.312 (70.287) Epoch: [3][140/1345], lr: 0.01000 Time 0.851 (0.770) Data 0.000 (0.022) Loss 2.4259 (2.3828) Prec@1 39.062 (41.656) Prec@5 76.562 (70.567) Epoch: [3][160/1345], lr: 0.01000 Time 0.708 (0.766) Data 0.000 (0.020) Loss 2.3189 (2.3850) Prec@1 39.062 (41.731) Prec@5 75.000 (70.759) Epoch: [3][180/1345], lr: 0.01000 Time 0.713 (0.763) Data 0.000 (0.017) Loss 2.5106 (2.3802) Prec@1 43.750 (41.859) Prec@5 67.188 (70.770) Epoch: [3][200/1345], lr: 0.01000 Time 0.712 (0.761) Data 0.000 (0.016) Loss 1.9929 (2.3774) Prec@1 46.875 (41.752) Prec@5 73.438 (70.896) Epoch: [3][220/1345], lr: 0.01000 Time 0.764 (0.759) Data 0.000 (0.014) Loss 2.1768 (2.3807) Prec@1 43.750 (41.678) Prec@5 75.000 (70.850) Epoch: [3][240/1345], lr: 0.01000 Time 0.710 (0.759) Data 0.000 (0.013) Loss 2.5926 (2.3885) Prec@1 34.375 (41.533) Prec@5 64.062 (70.643) Epoch: [3][260/1345], lr: 0.01000 Time 0.784 (0.758) Data 0.000 (0.012) Loss 2.3316 (2.3912) Prec@1 42.188 (41.571) Prec@5 68.750 (70.528) Epoch: [3][280/1345], lr: 0.01000 Time 0.856 (0.757) Data 0.000 (0.011) Loss 2.4277 (2.3901) Prec@1 42.188 (41.720) Prec@5 68.750 (70.524) Epoch: [3][300/1345], lr: 0.01000 Time 0.810 (0.756) Data 0.000 (0.011) Loss 2.4242 (2.3944) Prec@1 48.438 (41.539) Prec@5 64.062 (70.453) Epoch: [3][320/1345], lr: 0.01000 Time 0.729 (0.756) Data 0.000 (0.010) Loss 2.2662 (2.3937) Prec@1 43.750 (41.545) Prec@5 75.000 (70.444) Epoch: [3][340/1345], lr: 0.01000 Time 0.738 (0.755) Data 0.000 (0.009) Loss 2.2648 (2.3929) Prec@1 48.438 (41.573) Prec@5 76.562 (70.473) Epoch: [3][360/1345], lr: 0.01000 Time 0.737 (0.755) Data 0.000 (0.009) Loss 2.3061 (2.3910) Prec@1 43.750 (41.633) Prec@5 70.312 (70.416) Epoch: [3][380/1345], lr: 0.01000 Time 0.719 (0.754) Data 0.000 (0.009) Loss 2.1409 (2.3894) Prec@1 51.562 (41.646) Prec@5 70.312 (70.403) Epoch: [3][400/1345], lr: 0.01000 Time 0.709 (0.754) Data 0.000 (0.008) Loss 2.3222 (2.3902) Prec@1 45.312 (41.646) Prec@5 65.625 (70.363) Epoch: [3][420/1345], lr: 0.01000 Time 0.710 (0.754) Data 0.000 (0.008) Loss 1.9052 (2.3850) Prec@1 51.562 (41.731) Prec@5 79.688 (70.398) Epoch: [3][440/1345], lr: 0.01000 Time 0.837 (0.753) Data 0.000 (0.007) Loss 1.7738 (2.3865) Prec@1 51.562 (41.638) Prec@5 81.250 (70.387) Epoch: [3][460/1345], lr: 0.01000 Time 0.834 (0.753) Data 0.000 (0.007) Loss 2.5987 (2.3863) Prec@1 35.938 (41.686) Prec@5 71.875 (70.394) Epoch: [3][480/1345], lr: 0.01000 Time 0.782 (0.753) Data 0.000 (0.007) Loss 2.6256 (2.3850) Prec@1 28.125 (41.668) Prec@5 73.438 (70.472) Epoch: [3][500/1345], lr: 0.01000 Time 0.716 (0.753) Data 0.000 (0.007) Loss 2.3177 (2.3878) Prec@1 35.938 (41.586) Prec@5 71.875 (70.428) Epoch: [3][520/1345], lr: 0.01000 Time 0.776 (0.753) Data 0.000 (0.006) Loss 2.7146 (2.3869) Prec@1 32.812 (41.576) Prec@5 64.062 (70.417) Epoch: [3][540/1345], lr: 0.01000 Time 0.709 (0.752) Data 0.000 (0.006) Loss 2.3184 (2.3875) Prec@1 37.500 (41.543) Prec@5 73.438 (70.376) Epoch: [3][560/1345], lr: 0.01000 Time 0.781 (0.752) Data 0.000 (0.006) Loss 2.2167 (2.3849) Prec@1 48.438 (41.603) Prec@5 75.000 (70.432) Epoch: [3][580/1345], lr: 0.01000 Time 0.758 (0.752) Data 0.000 (0.006) Loss 2.2710 (2.3821) Prec@1 42.188 (41.668) Prec@5 81.250 (70.509) Epoch: [3][600/1345], lr: 0.01000 Time 0.851 (0.752) Data 0.000 (0.006) Loss 2.2920 (2.3789) Prec@1 50.000 (41.777) Prec@5 73.438 (70.526) Epoch: [3][620/1345], lr: 0.01000 Time 0.808 (0.752) Data 0.000 (0.005) Loss 1.8798 (2.3810) Prec@1 54.688 (41.752) Prec@5 73.438 (70.451) Epoch: [3][640/1345], lr: 0.01000 Time 0.711 (0.751) Data 0.000 (0.005) Loss 2.5368 (2.3809) Prec@1 32.812 (41.717) Prec@5 67.188 (70.464) Epoch: [3][660/1345], lr: 0.01000 Time 0.711 (0.751) Data 0.000 (0.005) Loss 2.0838 (2.3802) Prec@1 43.750 (41.757) Prec@5 81.250 (70.511) Epoch: [3][680/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.005) Loss 2.0169 (2.3817) Prec@1 42.188 (41.715) Prec@5 71.875 (70.498) Epoch: [3][700/1345], lr: 0.01000 Time 0.709 (0.750) Data 0.000 (0.005) Loss 2.0922 (2.3830) Prec@1 46.875 (41.657) Prec@5 75.000 (70.484) Epoch: [3][720/1345], lr: 0.01000 Time 0.717 (0.750) Data 0.001 (0.005) Loss 2.2337 (2.3839) Prec@1 39.062 (41.652) Prec@5 73.438 (70.479) Epoch: [3][740/1345], lr: 0.01000 Time 0.864 (0.750) Data 0.000 (0.005) Loss 2.3620 (2.3835) Prec@1 45.312 (41.650) Prec@5 70.312 (70.492) Epoch: [3][760/1345], lr: 0.01000 Time 0.742 (0.750) Data 0.000 (0.005) Loss 2.1487 (2.3826) Prec@1 39.062 (41.668) Prec@5 71.875 (70.491) Epoch: [3][780/1345], lr: 0.01000 Time 0.785 (0.750) Data 0.000 (0.004) Loss 2.5913 (2.3835) Prec@1 31.250 (41.645) Prec@5 67.188 (70.457) Epoch: [3][800/1345], lr: 0.01000 Time 0.711 (0.750) Data 0.000 (0.004) Loss 1.5947 (2.3807) Prec@1 57.812 (41.727) Prec@5 89.062 (70.517) Epoch: [3][820/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.004) Loss 2.3645 (2.3794) Prec@1 35.938 (41.790) Prec@5 71.875 (70.554) Epoch: [3][840/1345], lr: 0.01000 Time 0.758 (0.750) Data 0.000 (0.004) Loss 2.1233 (2.3776) Prec@1 48.438 (41.821) Prec@5 70.312 (70.602) Epoch: [3][860/1345], lr: 0.01000 Time 0.751 (0.750) Data 0.000 (0.004) Loss 2.6138 (2.3791) Prec@1 37.500 (41.763) Prec@5 64.062 (70.594) Epoch: [3][880/1345], lr: 0.01000 Time 0.725 (0.750) Data 0.000 (0.004) Loss 2.3186 (2.3787) Prec@1 40.625 (41.753) Prec@5 73.438 (70.612) Epoch: [3][900/1345], lr: 0.01000 Time 0.709 (0.750) Data 0.000 (0.004) Loss 2.2855 (2.3780) Prec@1 46.875 (41.780) Prec@5 73.438 (70.637) Epoch: [3][920/1345], lr: 0.01000 Time 0.734 (0.750) Data 0.000 (0.004) Loss 2.2957 (2.3782) Prec@1 48.438 (41.763) Prec@5 71.875 (70.608) Epoch: [3][940/1345], lr: 0.01000 Time 0.726 (0.750) Data 0.000 (0.004) Loss 2.0607 (2.3790) Prec@1 48.438 (41.777) Prec@5 75.000 (70.567) Epoch: [3][960/1345], lr: 0.01000 Time 0.718 (0.750) Data 0.000 (0.004) Loss 2.5350 (2.3786) Prec@1 45.312 (41.804) Prec@5 68.750 (70.587) Epoch: [3][980/1345], lr: 0.01000 Time 0.730 (0.749) Data 0.000 (0.004) Loss 2.4531 (2.3776) Prec@1 45.312 (41.861) Prec@5 68.750 (70.599) Epoch: [3][1000/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.004) Loss 2.2708 (2.3767) Prec@1 40.625 (41.903) Prec@5 70.312 (70.606) Epoch: [3][1020/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.003) Loss 2.6841 (2.3766) Prec@1 43.750 (41.903) Prec@5 64.062 (70.616) Epoch: [3][1040/1345], lr: 0.01000 Time 0.789 (0.749) Data 0.001 (0.003) Loss 2.2342 (2.3764) Prec@1 48.438 (41.911) Prec@5 79.688 (70.608) Epoch: [3][1060/1345], lr: 0.01000 Time 0.885 (0.749) Data 0.001 (0.003) Loss 2.4283 (2.3752) Prec@1 42.188 (41.930) Prec@5 68.750 (70.606) Epoch: [3][1080/1345], lr: 0.01000 Time 0.756 (0.749) Data 0.000 (0.003) Loss 2.1863 (2.3742) Prec@1 45.312 (41.989) Prec@5 75.000 (70.633) Epoch: [3][1100/1345], lr: 0.01000 Time 0.715 (0.749) Data 0.000 (0.003) Loss 2.2422 (2.3764) Prec@1 39.062 (41.965) Prec@5 75.000 (70.592) Epoch: [3][1120/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.003) Loss 2.8445 (2.3761) Prec@1 43.750 (41.992) Prec@5 67.188 (70.601) Epoch: [3][1140/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.003) Loss 2.2155 (2.3743) Prec@1 42.188 (42.030) Prec@5 78.125 (70.618) Epoch: [3][1160/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.003) Loss 2.3186 (2.3723) Prec@1 40.625 (42.074) Prec@5 73.438 (70.656) Epoch: [3][1180/1345], lr: 0.01000 Time 0.769 (0.748) Data 0.000 (0.003) Loss 2.3707 (2.3705) Prec@1 40.625 (42.105) Prec@5 65.625 (70.707) Epoch: [3][1200/1345], lr: 0.01000 Time 0.781 (0.748) Data 0.000 (0.003) Loss 1.9428 (2.3680) Prec@1 48.438 (42.165) Prec@5 79.688 (70.746) Epoch: [3][1220/1345], lr: 0.01000 Time 0.848 (0.748) Data 0.000 (0.003) Loss 2.4364 (2.3666) Prec@1 35.938 (42.180) Prec@5 70.312 (70.768) Epoch: [3][1240/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.3353 (2.3657) Prec@1 45.312 (42.210) Prec@5 73.438 (70.773) Epoch: [3][1260/1345], lr: 0.01000 Time 0.714 (0.748) Data 0.000 (0.003) Loss 2.1311 (2.3637) Prec@1 46.875 (42.237) Prec@5 73.438 (70.802) Epoch: [3][1280/1345], lr: 0.01000 Time 0.742 (0.748) Data 0.000 (0.003) Loss 2.3779 (2.3630) Prec@1 35.938 (42.257) Prec@5 75.000 (70.826) Epoch: [3][1300/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.003) Loss 2.1997 (2.3607) Prec@1 43.750 (42.318) Prec@5 71.875 (70.866) Epoch: [3][1320/1345], lr: 0.01000 Time 0.771 (0.748) Data 0.000 (0.003) Loss 2.1831 (2.3596) Prec@1 48.438 (42.332) Prec@5 73.438 (70.905) Epoch: [3][1340/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.003) Loss 2.3790 (2.3580) Prec@1 48.438 (42.372) Prec@5 68.750 (70.935) No BN layer Freezing. Test: [0/181] Time 3.262 (3.2623) Loss 3.1673 (3.1673) Prec@1 37.500 (37.500) Prec@5 59.375 (59.375) Test: [20/181] Time 0.908 (0.5932) Loss 2.7362 (2.9168) Prec@1 37.500 (35.491) Prec@5 65.625 (65.402) Test: [40/181] Time 0.740 (0.5223) Loss 3.0119 (2.9546) Prec@1 32.812 (34.832) Prec@5 64.062 (64.863) Test: [60/181] Time 0.544 (0.5007) Loss 2.9501 (2.9959) Prec@1 32.812 (34.298) Prec@5 56.250 (63.371) Test: [80/181] Time 0.718 (0.4918) Loss 2.7136 (3.0306) Prec@1 32.812 (33.603) Prec@5 71.875 (62.886) Test: [100/181] Time 0.885 (0.4852) Loss 3.6728 (3.0273) Prec@1 28.125 (33.586) Prec@5 57.812 (63.103) Test: [120/181] Time 0.814 (0.4809) Loss 3.2822 (3.0347) Prec@1 34.375 (33.574) Prec@5 59.375 (63.068) Test: [140/181] Time 0.767 (0.4770) Loss 3.3566 (3.0252) Prec@1 21.875 (33.666) Prec@5 60.938 (63.132) Test: [160/181] Time 1.027 (0.4763) Loss 2.9515 (3.0137) Prec@1 37.500 (33.948) Prec@5 68.750 (63.373) Testing Results: Prec@1 34.089 Prec@5 63.403 Loss 3.01307 Time 0.4710 No BN layer Freezing. Epoch: [4][0/1345], lr: 0.01000 Time 4.480 (4.480) Data 3.750 (3.750) Loss 2.1435 (2.1435) Prec@1 43.750 (43.750) Prec@5 73.438 (73.438) Epoch: [4][20/1345], lr: 0.01000 Time 0.812 (0.933) Data 0.000 (0.179) Loss 2.4580 (2.1905) Prec@1 43.750 (45.982) Prec@5 65.625 (73.586) Epoch: [4][40/1345], lr: 0.01000 Time 0.713 (0.844) Data 0.000 (0.092) Loss 2.5513 (2.2134) Prec@1 43.750 (45.732) Prec@5 64.062 (72.675) Epoch: [4][60/1345], lr: 0.01000 Time 0.711 (0.810) Data 0.000 (0.062) Loss 2.1665 (2.1875) Prec@1 48.438 (46.055) Prec@5 68.750 (73.028) Epoch: [4][80/1345], lr: 0.01000 Time 0.749 (0.791) Data 0.000 (0.047) Loss 1.9567 (2.1657) Prec@1 46.875 (46.547) Prec@5 81.250 (73.881) Epoch: [4][100/1345], lr: 0.01000 Time 0.737 (0.779) Data 0.001 (0.038) Loss 2.0722 (2.1753) Prec@1 43.750 (46.457) Prec@5 78.125 (73.840) Epoch: [4][120/1345], lr: 0.01000 Time 0.708 (0.772) Data 0.000 (0.031) Loss 2.1824 (2.1695) Prec@1 46.875 (46.307) Prec@5 73.438 (74.199) Epoch: [4][140/1345], lr: 0.01000 Time 0.705 (0.768) Data 0.000 (0.027) Loss 2.2842 (2.1853) Prec@1 45.312 (45.911) Prec@5 76.562 (74.036) Epoch: [4][160/1345], lr: 0.01000 Time 0.885 (0.765) Data 0.001 (0.024) Loss 2.4452 (2.1953) Prec@1 42.188 (45.837) Prec@5 68.750 (73.826) Epoch: [4][180/1345], lr: 0.01000 Time 0.755 (0.763) Data 0.001 (0.021) Loss 2.2158 (2.1952) Prec@1 43.750 (45.718) Prec@5 71.875 (73.817) Epoch: [4][200/1345], lr: 0.01000 Time 0.755 (0.760) Data 0.001 (0.019) Loss 2.4342 (2.1961) Prec@1 39.062 (45.732) Prec@5 73.438 (73.818) Epoch: [4][220/1345], lr: 0.01000 Time 0.720 (0.758) Data 0.000 (0.017) Loss 2.0374 (2.1892) Prec@1 48.438 (45.779) Prec@5 79.688 (73.904) Epoch: [4][240/1345], lr: 0.01000 Time 0.710 (0.757) Data 0.001 (0.016) Loss 2.2777 (2.1895) Prec@1 45.312 (45.656) Prec@5 73.438 (74.015) Epoch: [4][260/1345], lr: 0.01000 Time 0.707 (0.754) Data 0.000 (0.015) Loss 2.2691 (2.1924) Prec@1 35.938 (45.612) Prec@5 73.438 (73.958) Epoch: [4][280/1345], lr: 0.01000 Time 0.708 (0.751) Data 0.000 (0.014) Loss 2.2811 (2.2013) Prec@1 42.188 (45.507) Prec@5 71.875 (73.743) Epoch: [4][300/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.013) Loss 2.2324 (2.2023) Prec@1 51.562 (45.468) Prec@5 76.562 (73.759) Epoch: [4][320/1345], lr: 0.01000 Time 0.713 (0.747) Data 0.000 (0.012) Loss 2.3722 (2.2049) Prec@1 37.500 (45.424) Prec@5 73.438 (73.681) Epoch: [4][340/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.011) Loss 1.9894 (2.2060) Prec@1 51.562 (45.436) Prec@5 76.562 (73.639) Epoch: [4][360/1345], lr: 0.01000 Time 0.803 (0.745) Data 0.000 (0.011) Loss 1.9044 (2.2053) Prec@1 51.562 (45.442) Prec@5 81.250 (73.710) Epoch: [4][380/1345], lr: 0.01000 Time 0.779 (0.746) Data 0.000 (0.010) Loss 2.3630 (2.2087) Prec@1 39.062 (45.292) Prec@5 65.625 (73.643) Epoch: [4][400/1345], lr: 0.01000 Time 0.770 (0.745) Data 0.000 (0.010) Loss 2.2330 (2.2085) Prec@1 50.000 (45.328) Prec@5 73.438 (73.683) Epoch: [4][420/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.009) Loss 2.3669 (2.2069) Prec@1 40.625 (45.476) Prec@5 71.875 (73.675) Epoch: [4][440/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.009) Loss 1.7342 (2.2129) Prec@1 54.688 (45.355) Prec@5 85.938 (73.565) Epoch: [4][460/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.009) Loss 2.1388 (2.2160) Prec@1 39.062 (45.238) Prec@5 75.000 (73.509) Epoch: [4][480/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.008) Loss 2.6797 (2.2210) Prec@1 35.938 (45.173) Prec@5 68.750 (73.425) Epoch: [4][500/1345], lr: 0.01000 Time 0.807 (0.743) Data 0.000 (0.008) Loss 2.1135 (2.2188) Prec@1 45.312 (45.256) Prec@5 79.688 (73.469) Epoch: [4][520/1345], lr: 0.01000 Time 0.802 (0.743) Data 0.001 (0.008) Loss 1.9844 (2.2174) Prec@1 45.312 (45.244) Prec@5 79.688 (73.494) Epoch: [4][540/1345], lr: 0.01000 Time 0.714 (0.742) Data 0.000 (0.007) Loss 1.9162 (2.2192) Prec@1 54.688 (45.197) Prec@5 73.438 (73.492) Epoch: [4][560/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.007) Loss 2.2189 (2.2200) Prec@1 48.438 (45.190) Prec@5 70.312 (73.421) Epoch: [4][580/1345], lr: 0.01000 Time 0.716 (0.741) Data 0.000 (0.007) Loss 2.3724 (2.2221) Prec@1 37.500 (45.124) Prec@5 73.438 (73.373) Epoch: [4][600/1345], lr: 0.01000 Time 0.758 (0.741) Data 0.000 (0.007) Loss 2.3679 (2.2207) Prec@1 37.500 (45.180) Prec@5 70.312 (73.401) Epoch: [4][620/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.006) Loss 1.9484 (2.2165) Prec@1 54.688 (45.297) Prec@5 76.562 (73.495) Epoch: [4][640/1345], lr: 0.01000 Time 0.705 (0.740) Data 0.000 (0.006) Loss 2.1784 (2.2168) Prec@1 46.875 (45.312) Prec@5 75.000 (73.503) Epoch: [4][660/1345], lr: 0.01000 Time 0.806 (0.740) Data 0.000 (0.006) Loss 1.8596 (2.2182) Prec@1 56.250 (45.291) Prec@5 81.250 (73.442) Epoch: [4][680/1345], lr: 0.01000 Time 0.817 (0.740) Data 0.001 (0.006) Loss 2.0650 (2.2201) Prec@1 37.500 (45.257) Prec@5 79.688 (73.401) Epoch: [4][700/1345], lr: 0.01000 Time 0.706 (0.740) Data 0.000 (0.006) Loss 2.3739 (2.2228) Prec@1 37.500 (45.210) Prec@5 65.625 (73.333) Epoch: [4][720/1345], lr: 0.01000 Time 0.755 (0.740) Data 0.000 (0.006) Loss 2.1347 (2.2225) Prec@1 46.875 (45.258) Prec@5 71.875 (73.320) Epoch: [4][740/1345], lr: 0.01000 Time 0.705 (0.740) Data 0.000 (0.006) Loss 2.5296 (2.2236) Prec@1 35.938 (45.230) Prec@5 71.875 (73.334) Epoch: [4][760/1345], lr: 0.01000 Time 0.710 (0.739) Data 0.001 (0.005) Loss 2.2317 (2.2221) Prec@1 46.875 (45.253) Prec@5 71.875 (73.392) Epoch: [4][780/1345], lr: 0.01000 Time 0.707 (0.739) Data 0.000 (0.005) Loss 1.9427 (2.2193) Prec@1 48.438 (45.292) Prec@5 78.125 (73.435) Epoch: [4][800/1345], lr: 0.01000 Time 0.774 (0.739) Data 0.000 (0.005) Loss 2.5725 (2.2181) Prec@1 40.625 (45.297) Prec@5 64.062 (73.430) Epoch: [4][820/1345], lr: 0.01000 Time 0.840 (0.740) Data 0.000 (0.005) Loss 2.1665 (2.2179) Prec@1 51.562 (45.349) Prec@5 73.438 (73.449) Epoch: [4][840/1345], lr: 0.01000 Time 0.782 (0.740) Data 0.000 (0.005) Loss 1.9449 (2.2149) Prec@1 51.562 (45.441) Prec@5 81.250 (73.514) Epoch: [4][860/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.005) Loss 2.2952 (2.2143) Prec@1 43.750 (45.441) Prec@5 73.438 (73.521) Epoch: [4][880/1345], lr: 0.01000 Time 0.707 (0.739) Data 0.000 (0.005) Loss 2.0567 (2.2130) Prec@1 48.438 (45.504) Prec@5 78.125 (73.546) Epoch: [4][900/1345], lr: 0.01000 Time 0.778 (0.739) Data 0.000 (0.005) Loss 1.9254 (2.2112) Prec@1 54.688 (45.508) Prec@5 82.812 (73.569) Epoch: [4][920/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.005) Loss 2.0044 (2.2070) Prec@1 57.812 (45.577) Prec@5 78.125 (73.636) Epoch: [4][940/1345], lr: 0.01000 Time 0.787 (0.739) Data 0.000 (0.004) Loss 2.1403 (2.2083) Prec@1 46.875 (45.557) Prec@5 79.688 (73.635) Epoch: [4][960/1345], lr: 0.01000 Time 0.713 (0.738) Data 0.000 (0.004) Loss 2.4542 (2.2075) Prec@1 45.312 (45.586) Prec@5 62.500 (73.631) Epoch: [4][980/1345], lr: 0.01000 Time 0.777 (0.738) Data 0.001 (0.004) Loss 1.9809 (2.2080) Prec@1 46.875 (45.561) Prec@5 76.562 (73.614) Epoch: [4][1000/1345], lr: 0.01000 Time 0.728 (0.739) Data 0.000 (0.004) Loss 2.2977 (2.2077) Prec@1 45.312 (45.562) Prec@5 70.312 (73.633) Epoch: [4][1020/1345], lr: 0.01000 Time 0.770 (0.739) Data 0.001 (0.004) Loss 2.4961 (2.2089) Prec@1 39.062 (45.534) Prec@5 65.625 (73.603) Epoch: [4][1040/1345], lr: 0.01000 Time 0.785 (0.738) Data 0.000 (0.004) Loss 2.3985 (2.2099) Prec@1 42.188 (45.553) Prec@5 75.000 (73.598) Epoch: [4][1060/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.004) Loss 2.2657 (2.2084) Prec@1 45.312 (45.604) Prec@5 71.875 (73.614) Epoch: [4][1080/1345], lr: 0.01000 Time 0.715 (0.738) Data 0.001 (0.004) Loss 2.2813 (2.2088) Prec@1 40.625 (45.596) Prec@5 70.312 (73.588) Epoch: [4][1100/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.004) Loss 2.0931 (2.2094) Prec@1 48.438 (45.554) Prec@5 73.438 (73.594) Epoch: [4][1120/1345], lr: 0.01000 Time 0.749 (0.737) Data 0.000 (0.004) Loss 2.1151 (2.2087) Prec@1 53.125 (45.580) Prec@5 70.312 (73.592) Epoch: [4][1140/1345], lr: 0.01000 Time 0.707 (0.737) Data 0.000 (0.004) Loss 2.2795 (2.2072) Prec@1 42.188 (45.589) Prec@5 76.562 (73.620) Epoch: [4][1160/1345], lr: 0.01000 Time 0.709 (0.737) Data 0.000 (0.004) Loss 2.6617 (2.2060) Prec@1 39.062 (45.625) Prec@5 65.625 (73.637) Epoch: [4][1180/1345], lr: 0.01000 Time 0.713 (0.737) Data 0.000 (0.004) Loss 2.4224 (2.2052) Prec@1 42.188 (45.635) Prec@5 65.625 (73.632) Epoch: [4][1200/1345], lr: 0.01000 Time 0.713 (0.737) Data 0.001 (0.004) Loss 1.8221 (2.2046) Prec@1 54.688 (45.666) Prec@5 78.125 (73.638) Epoch: [4][1220/1345], lr: 0.01000 Time 0.723 (0.737) Data 0.001 (0.004) Loss 2.1862 (2.2037) Prec@1 46.875 (45.686) Prec@5 71.875 (73.655) Epoch: [4][1240/1345], lr: 0.01000 Time 0.781 (0.738) Data 0.001 (0.003) Loss 2.1455 (2.2030) Prec@1 48.438 (45.715) Prec@5 70.312 (73.662) Epoch: [4][1260/1345], lr: 0.01000 Time 0.853 (0.738) Data 0.000 (0.003) Loss 2.3196 (2.2038) Prec@1 48.438 (45.716) Prec@5 71.875 (73.652) Epoch: [4][1280/1345], lr: 0.01000 Time 0.786 (0.738) Data 0.000 (0.003) Loss 2.4293 (2.2034) Prec@1 48.438 (45.716) Prec@5 68.750 (73.662) Epoch: [4][1300/1345], lr: 0.01000 Time 0.749 (0.738) Data 0.000 (0.003) Loss 1.7183 (2.2032) Prec@1 53.125 (45.699) Prec@5 87.500 (73.685) Epoch: [4][1320/1345], lr: 0.01000 Time 0.705 (0.738) Data 0.000 (0.003) Loss 1.7994 (2.2020) Prec@1 50.000 (45.729) Prec@5 78.125 (73.717) Epoch: [4][1340/1345], lr: 0.01000 Time 0.706 (0.737) Data 0.000 (0.003) Loss 2.3344 (2.2022) Prec@1 43.750 (45.718) Prec@5 73.438 (73.705) No BN layer Freezing. Test: [0/181] Time 3.455 (3.4553) Loss 3.0116 (3.0116) Prec@1 39.062 (39.062) Prec@5 68.750 (68.750) Test: [20/181] Time 0.585 (0.5822) Loss 2.7539 (2.6724) Prec@1 37.500 (37.872) Prec@5 67.188 (67.783) Test: [40/181] Time 0.647 (0.5196) Loss 2.5499 (2.6995) Prec@1 39.062 (37.881) Prec@5 67.188 (67.035) Test: [60/181] Time 0.648 (0.4964) Loss 2.5713 (2.7456) Prec@1 39.062 (37.321) Prec@5 62.500 (66.035) Test: [80/181] Time 0.517 (0.4848) Loss 2.5654 (2.7615) Prec@1 35.938 (36.343) Prec@5 64.062 (65.721) Test: [100/181] Time 0.378 (0.4777) Loss 3.1384 (2.7613) Prec@1 20.312 (36.231) Prec@5 56.250 (65.579) Test: [120/181] Time 0.237 (0.4739) Loss 3.2838 (2.7652) Prec@1 40.625 (36.519) Prec@5 62.500 (65.728) Test: [140/181] Time 0.237 (0.4708) Loss 3.4148 (2.7655) Prec@1 29.688 (36.325) Prec@5 62.500 (65.769) Test: [160/181] Time 0.236 (0.4694) Loss 2.6311 (2.7428) Prec@1 32.812 (36.588) Prec@5 71.875 (66.159) Testing Results: Prec@1 36.797 Prec@5 66.458 Loss 2.73651 Time 0.4686 No BN layer Freezing. Epoch: [5][0/1345], lr: 0.01000 Time 3.891 (3.891) Data 3.146 (3.146) Loss 2.1862 (2.1862) Prec@1 50.000 (50.000) Prec@5 68.750 (68.750) Epoch: [5][20/1345], lr: 0.01000 Time 0.707 (0.890) Data 0.000 (0.150) Loss 2.2175 (2.0425) Prec@1 43.750 (48.289) Prec@5 78.125 (76.339) Epoch: [5][40/1345], lr: 0.01000 Time 0.754 (0.811) Data 0.000 (0.077) Loss 2.3832 (2.1091) Prec@1 42.188 (47.332) Prec@5 68.750 (75.724) Epoch: [5][60/1345], lr: 0.01000 Time 0.708 (0.780) Data 0.000 (0.052) Loss 1.8500 (2.1011) Prec@1 56.250 (48.335) Prec@5 78.125 (75.845) Epoch: [5][80/1345], lr: 0.01000 Time 0.749 (0.770) Data 0.000 (0.039) Loss 1.9657 (2.1006) Prec@1 50.000 (48.148) Prec@5 81.250 (75.829) Epoch: [5][100/1345], lr: 0.01000 Time 0.756 (0.767) Data 0.000 (0.032) Loss 2.4284 (2.0852) Prec@1 42.188 (48.360) Prec@5 68.750 (75.928) Epoch: [5][120/1345], lr: 0.01000 Time 0.707 (0.762) Data 0.000 (0.026) Loss 1.7398 (2.0860) Prec@1 45.312 (48.179) Prec@5 81.250 (75.723) Epoch: [5][140/1345], lr: 0.01000 Time 0.740 (0.759) Data 0.000 (0.023) Loss 1.9819 (2.0846) Prec@1 45.312 (48.271) Prec@5 70.312 (75.643) Epoch: [5][160/1345], lr: 0.01000 Time 0.835 (0.757) Data 0.000 (0.020) Loss 2.1924 (2.0775) Prec@1 48.438 (48.195) Prec@5 78.125 (75.718) Epoch: [5][180/1345], lr: 0.01000 Time 0.754 (0.757) Data 0.000 (0.018) Loss 2.0450 (2.0821) Prec@1 46.875 (47.894) Prec@5 79.688 (75.699) Epoch: [5][200/1345], lr: 0.01000 Time 0.707 (0.753) Data 0.000 (0.016) Loss 2.2356 (2.0765) Prec@1 51.562 (48.010) Prec@5 75.000 (75.902) Epoch: [5][220/1345], lr: 0.01000 Time 0.705 (0.750) Data 0.000 (0.015) Loss 1.9188 (2.0695) Prec@1 51.562 (48.105) Prec@5 79.688 (76.103) Epoch: [5][240/1345], lr: 0.01000 Time 0.773 (0.749) Data 0.000 (0.013) Loss 1.9309 (2.0738) Prec@1 54.688 (48.036) Prec@5 79.688 (75.992) Epoch: [5][260/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.001 (0.012) Loss 2.5277 (2.0738) Prec@1 43.750 (48.066) Prec@5 75.000 (75.904) Epoch: [5][280/1345], lr: 0.01000 Time 0.894 (0.745) Data 0.000 (0.012) Loss 2.0318 (2.0810) Prec@1 43.750 (47.904) Prec@5 68.750 (75.751) Epoch: [5][300/1345], lr: 0.01000 Time 0.771 (0.744) Data 0.001 (0.011) Loss 2.0783 (2.0797) Prec@1 42.188 (47.872) Prec@5 75.000 (75.825) Epoch: [5][320/1345], lr: 0.01000 Time 0.705 (0.743) Data 0.000 (0.010) Loss 2.4257 (2.0795) Prec@1 42.188 (47.819) Prec@5 76.562 (75.827) Epoch: [5][340/1345], lr: 0.01000 Time 0.703 (0.743) Data 0.000 (0.010) Loss 2.1684 (2.0802) Prec@1 53.125 (47.878) Prec@5 76.562 (75.866) Epoch: [5][360/1345], lr: 0.01000 Time 0.718 (0.742) Data 0.000 (0.009) Loss 1.6437 (2.0810) Prec@1 56.250 (47.935) Prec@5 78.125 (75.866) Epoch: [5][380/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.009) Loss 2.2669 (2.0865) Prec@1 48.438 (47.835) Prec@5 76.562 (75.808) Epoch: [5][400/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.008) Loss 2.0383 (2.0893) Prec@1 46.875 (47.771) Prec@5 81.250 (75.810) Epoch: [5][420/1345], lr: 0.01000 Time 0.793 (0.741) Data 0.000 (0.008) Loss 2.3655 (2.0924) Prec@1 45.312 (47.658) Prec@5 68.750 (75.779) Epoch: [5][440/1345], lr: 0.01000 Time 0.797 (0.740) Data 0.000 (0.008) Loss 2.4627 (2.0925) Prec@1 39.062 (47.725) Prec@5 67.188 (75.719) Epoch: [5][460/1345], lr: 0.01000 Time 0.717 (0.739) Data 0.000 (0.007) Loss 2.0020 (2.0913) Prec@1 51.562 (47.682) Prec@5 75.000 (75.742) Epoch: [5][480/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.007) Loss 2.0998 (2.0907) Prec@1 46.875 (47.668) Prec@5 79.688 (75.812) Epoch: [5][500/1345], lr: 0.01000 Time 0.785 (0.738) Data 0.001 (0.007) Loss 2.4081 (2.0890) Prec@1 42.188 (47.680) Prec@5 68.750 (75.826) Epoch: [5][520/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.006) Loss 1.8884 (2.0865) Prec@1 51.562 (47.769) Prec@5 79.688 (75.906) Epoch: [5][540/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.006) Loss 2.0671 (2.0879) Prec@1 46.875 (47.765) Prec@5 76.562 (75.898) Epoch: [5][560/1345], lr: 0.01000 Time 0.708 (0.737) Data 0.000 (0.006) Loss 2.2324 (2.0887) Prec@1 45.312 (47.797) Prec@5 71.875 (75.875) Epoch: [5][580/1345], lr: 0.01000 Time 0.713 (0.737) Data 0.000 (0.006) Loss 1.7618 (2.0887) Prec@1 51.562 (47.787) Prec@5 85.938 (75.879) Epoch: [5][600/1345], lr: 0.01000 Time 0.706 (0.737) Data 0.000 (0.006) Loss 1.5047 (2.0875) Prec@1 64.062 (47.782) Prec@5 87.500 (75.861) Epoch: [5][620/1345], lr: 0.01000 Time 0.709 (0.736) Data 0.000 (0.005) Loss 2.1894 (2.0854) Prec@1 46.875 (47.849) Prec@5 71.875 (75.896) Epoch: [5][640/1345], lr: 0.01000 Time 0.705 (0.736) Data 0.000 (0.005) Loss 2.1194 (2.0863) Prec@1 40.625 (47.774) Prec@5 71.875 (75.880) Epoch: [5][660/1345], lr: 0.01000 Time 0.812 (0.736) Data 0.000 (0.005) Loss 2.5809 (2.0867) Prec@1 45.312 (47.754) Prec@5 68.750 (75.868) Epoch: [5][680/1345], lr: 0.01000 Time 0.706 (0.736) Data 0.000 (0.005) Loss 2.0180 (2.0844) Prec@1 50.000 (47.811) Prec@5 76.562 (75.943) Epoch: [5][700/1345], lr: 0.01000 Time 0.714 (0.736) Data 0.000 (0.005) Loss 2.0192 (2.0864) Prec@1 43.750 (47.715) Prec@5 76.562 (75.950) Epoch: [5][720/1345], lr: 0.01000 Time 0.711 (0.736) Data 0.000 (0.005) Loss 1.8982 (2.0852) Prec@1 53.125 (47.718) Prec@5 76.562 (75.977) Epoch: [5][740/1345], lr: 0.01000 Time 0.713 (0.737) Data 0.001 (0.005) Loss 1.7562 (2.0841) Prec@1 51.562 (47.750) Prec@5 84.375 (75.991) Epoch: [5][760/1345], lr: 0.01000 Time 0.706 (0.737) Data 0.000 (0.005) Loss 2.0214 (2.0871) Prec@1 51.562 (47.653) Prec@5 73.438 (75.940) Epoch: [5][780/1345], lr: 0.01000 Time 0.716 (0.738) Data 0.001 (0.004) Loss 2.1555 (2.0859) Prec@1 48.438 (47.707) Prec@5 78.125 (75.964) Epoch: [5][800/1345], lr: 0.01000 Time 0.714 (0.738) Data 0.000 (0.004) Loss 2.1509 (2.0875) Prec@1 45.312 (47.657) Prec@5 73.438 (75.905) Epoch: [5][820/1345], lr: 0.01000 Time 0.740 (0.738) Data 0.000 (0.004) Loss 2.2025 (2.0874) Prec@1 51.562 (47.686) Prec@5 67.188 (75.887) Epoch: [5][840/1345], lr: 0.01000 Time 0.716 (0.737) Data 0.000 (0.004) Loss 2.2814 (2.0869) Prec@1 37.500 (47.711) Prec@5 73.438 (75.894) Epoch: [5][860/1345], lr: 0.01000 Time 0.713 (0.738) Data 0.000 (0.004) Loss 2.6567 (2.0887) Prec@1 35.938 (47.701) Prec@5 67.188 (75.873) Epoch: [5][880/1345], lr: 0.01000 Time 0.765 (0.738) Data 0.000 (0.004) Loss 1.7350 (2.0883) Prec@1 56.250 (47.710) Prec@5 81.250 (75.910) Epoch: [5][900/1345], lr: 0.01000 Time 0.882 (0.738) Data 0.001 (0.004) Loss 2.3835 (2.0864) Prec@1 34.375 (47.706) Prec@5 67.188 (75.921) Epoch: [5][920/1345], lr: 0.01000 Time 0.778 (0.738) Data 0.000 (0.004) Loss 1.6390 (2.0855) Prec@1 57.812 (47.744) Prec@5 87.500 (75.938) Epoch: [5][940/1345], lr: 0.01000 Time 0.730 (0.738) Data 0.000 (0.004) Loss 1.6792 (2.0840) Prec@1 62.500 (47.772) Prec@5 84.375 (75.978) Epoch: [5][960/1345], lr: 0.01000 Time 0.715 (0.738) Data 0.001 (0.004) Loss 2.4239 (2.0850) Prec@1 42.188 (47.737) Prec@5 67.188 (75.972) Epoch: [5][980/1345], lr: 0.01000 Time 0.709 (0.738) Data 0.000 (0.004) Loss 2.3177 (2.0836) Prec@1 51.562 (47.789) Prec@5 71.875 (75.984) Epoch: [5][1000/1345], lr: 0.01000 Time 0.710 (0.738) Data 0.000 (0.004) Loss 1.9160 (2.0842) Prec@1 51.562 (47.729) Prec@5 81.250 (75.993) Epoch: [5][1020/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.004) Loss 2.8909 (2.0852) Prec@1 34.375 (47.706) Prec@5 57.812 (75.986) Epoch: [5][1040/1345], lr: 0.01000 Time 0.761 (0.738) Data 0.000 (0.003) Loss 2.1512 (2.0849) Prec@1 40.625 (47.686) Prec@5 78.125 (75.982) Epoch: [5][1060/1345], lr: 0.01000 Time 0.712 (0.738) Data 0.000 (0.003) Loss 1.8403 (2.0834) Prec@1 53.125 (47.716) Prec@5 78.125 (75.985) Epoch: [5][1080/1345], lr: 0.01000 Time 0.716 (0.738) Data 0.000 (0.003) Loss 1.8939 (2.0831) Prec@1 53.125 (47.762) Prec@5 81.250 (75.992) Epoch: [5][1100/1345], lr: 0.01000 Time 0.713 (0.738) Data 0.000 (0.003) Loss 2.1789 (2.0813) Prec@1 59.375 (47.800) Prec@5 73.438 (76.019) Epoch: [5][1120/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.7736 (2.0812) Prec@1 46.875 (47.773) Prec@5 84.375 (76.019) Epoch: [5][1140/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.003) Loss 1.9964 (2.0788) Prec@1 54.688 (47.836) Prec@5 78.125 (76.045) Epoch: [5][1160/1345], lr: 0.01000 Time 0.773 (0.738) Data 0.000 (0.003) Loss 2.2951 (2.0788) Prec@1 37.500 (47.836) Prec@5 71.875 (76.026) Epoch: [5][1180/1345], lr: 0.01000 Time 0.711 (0.737) Data 0.000 (0.003) Loss 2.2854 (2.0795) Prec@1 42.188 (47.826) Prec@5 73.438 (76.023) Epoch: [5][1200/1345], lr: 0.01000 Time 0.770 (0.737) Data 0.000 (0.003) Loss 1.8550 (2.0791) Prec@1 57.812 (47.831) Prec@5 78.125 (76.032) Epoch: [5][1220/1345], lr: 0.01000 Time 0.707 (0.737) Data 0.000 (0.003) Loss 2.0410 (2.0775) Prec@1 48.438 (47.835) Prec@5 81.250 (76.069) Epoch: [5][1240/1345], lr: 0.01000 Time 0.788 (0.737) Data 0.000 (0.003) Loss 1.8732 (2.0781) Prec@1 54.688 (47.831) Prec@5 75.000 (76.059) Epoch: [5][1260/1345], lr: 0.01000 Time 0.709 (0.737) Data 0.000 (0.003) Loss 2.1535 (2.0781) Prec@1 43.750 (47.818) Prec@5 78.125 (76.062) Epoch: [5][1280/1345], lr: 0.01000 Time 0.706 (0.737) Data 0.000 (0.003) Loss 1.7913 (2.0783) Prec@1 54.688 (47.835) Prec@5 82.812 (76.047) Epoch: [5][1300/1345], lr: 0.01000 Time 0.707 (0.736) Data 0.000 (0.003) Loss 2.4059 (2.0775) Prec@1 45.312 (47.827) Prec@5 68.750 (76.051) Epoch: [5][1320/1345], lr: 0.01000 Time 0.778 (0.736) Data 0.000 (0.003) Loss 1.7110 (2.0776) Prec@1 57.812 (47.832) Prec@5 78.125 (76.039) Epoch: [5][1340/1345], lr: 0.01000 Time 0.814 (0.736) Data 0.000 (0.003) Loss 1.8447 (2.0774) Prec@1 57.812 (47.843) Prec@5 76.562 (76.054) No BN layer Freezing. Test: [0/181] Time 3.056 (3.0556) Loss 2.9518 (2.9518) Prec@1 37.500 (37.500) Prec@5 68.750 (68.750) Test: [20/181] Time 0.980 (0.5965) Loss 2.6806 (2.7713) Prec@1 40.625 (37.872) Prec@5 65.625 (68.006) Test: [40/181] Time 1.117 (0.5300) Loss 2.9994 (2.8273) Prec@1 40.625 (36.928) Prec@5 68.750 (67.569) Test: [60/181] Time 0.992 (0.5047) Loss 2.8409 (2.8788) Prec@1 32.812 (35.835) Prec@5 59.375 (66.445) Test: [80/181] Time 1.175 (0.4934) Loss 2.8052 (2.9090) Prec@1 32.812 (35.475) Prec@5 76.562 (65.895) Test: [100/181] Time 1.253 (0.4888) Loss 3.7278 (2.9032) Prec@1 28.125 (35.876) Prec@5 53.125 (66.136) Test: [120/181] Time 1.175 (0.4833) Loss 3.1159 (2.9115) Prec@1 35.938 (35.873) Prec@5 60.938 (66.129) Test: [140/181] Time 1.159 (0.4797) Loss 3.3073 (2.9109) Prec@1 31.250 (35.904) Prec@5 60.938 (66.223) Test: [160/181] Time 1.160 (0.4768) Loss 3.0882 (2.8969) Prec@1 32.812 (36.083) Prec@5 70.312 (66.460) Testing Results: Prec@1 36.137 Prec@5 66.649 Loss 2.88313 Time 0.4708 No BN layer Freezing. Epoch: [6][0/1345], lr: 0.01000 Time 4.003 (4.003) Data 3.259 (3.259) Loss 1.8450 (1.8450) Prec@1 57.812 (57.812) Prec@5 79.688 (79.688) Epoch: [6][20/1345], lr: 0.01000 Time 0.706 (0.906) Data 0.000 (0.156) Loss 1.9406 (1.9058) Prec@1 59.375 (51.562) Prec@5 78.125 (79.241) Epoch: [6][40/1345], lr: 0.01000 Time 0.768 (0.828) Data 0.001 (0.080) Loss 2.3804 (1.9470) Prec@1 50.000 (51.258) Prec@5 70.312 (78.201) Epoch: [6][60/1345], lr: 0.01000 Time 0.710 (0.797) Data 0.000 (0.054) Loss 2.0386 (1.9460) Prec@1 50.000 (51.076) Prec@5 76.562 (78.432) Epoch: [6][80/1345], lr: 0.01000 Time 0.707 (0.780) Data 0.000 (0.041) Loss 2.1560 (1.9389) Prec@1 48.438 (51.080) Prec@5 73.438 (78.607) Epoch: [6][100/1345], lr: 0.01000 Time 0.757 (0.772) Data 0.000 (0.033) Loss 1.9525 (1.9548) Prec@1 51.562 (50.572) Prec@5 78.125 (78.373) Epoch: [6][120/1345], lr: 0.01000 Time 0.725 (0.768) Data 0.000 (0.027) Loss 2.2804 (1.9536) Prec@1 43.750 (50.607) Prec@5 70.312 (78.435) Epoch: [6][140/1345], lr: 0.01000 Time 0.719 (0.764) Data 0.001 (0.024) Loss 1.9075 (1.9629) Prec@1 50.000 (50.521) Prec@5 79.688 (78.269) Epoch: [6][160/1345], lr: 0.01000 Time 0.714 (0.762) Data 0.000 (0.021) Loss 1.9019 (1.9704) Prec@1 54.688 (50.311) Prec@5 79.688 (78.144) Epoch: [6][180/1345], lr: 0.01000 Time 0.709 (0.760) Data 0.000 (0.018) Loss 1.8725 (1.9664) Prec@1 56.250 (50.466) Prec@5 76.562 (78.151) Epoch: [6][200/1345], lr: 0.01000 Time 0.716 (0.758) Data 0.001 (0.017) Loss 2.0982 (1.9611) Prec@1 43.750 (50.567) Prec@5 73.438 (78.288) Epoch: [6][220/1345], lr: 0.01000 Time 0.728 (0.756) Data 0.000 (0.015) Loss 1.9916 (1.9590) Prec@1 46.875 (50.537) Prec@5 78.125 (78.387) Epoch: [6][240/1345], lr: 0.01000 Time 0.714 (0.754) Data 0.000 (0.014) Loss 1.7516 (1.9568) Prec@1 51.562 (50.635) Prec@5 85.938 (78.501) Epoch: [6][260/1345], lr: 0.01000 Time 0.714 (0.753) Data 0.000 (0.013) Loss 2.0344 (1.9575) Prec@1 53.125 (50.712) Prec@5 73.438 (78.436) Epoch: [6][280/1345], lr: 0.01000 Time 0.714 (0.753) Data 0.001 (0.012) Loss 2.0097 (1.9602) Prec@1 50.000 (50.639) Prec@5 76.562 (78.397) Epoch: [6][300/1345], lr: 0.01000 Time 0.832 (0.753) Data 0.000 (0.011) Loss 1.8392 (1.9644) Prec@1 54.688 (50.587) Prec@5 82.812 (78.322) Epoch: [6][320/1345], lr: 0.01000 Time 0.713 (0.752) Data 0.000 (0.011) Loss 2.0145 (1.9641) Prec@1 46.875 (50.531) Prec@5 76.562 (78.320) Epoch: [6][340/1345], lr: 0.01000 Time 0.736 (0.752) Data 0.000 (0.010) Loss 2.0871 (1.9665) Prec@1 54.688 (50.490) Prec@5 70.312 (78.249) Epoch: [6][360/1345], lr: 0.01000 Time 0.711 (0.752) Data 0.000 (0.010) Loss 2.3153 (1.9717) Prec@1 40.625 (50.351) Prec@5 73.438 (78.116) Epoch: [6][380/1345], lr: 0.01000 Time 0.719 (0.752) Data 0.000 (0.009) Loss 2.2425 (1.9766) Prec@1 40.625 (50.324) Prec@5 73.438 (77.940) Epoch: [6][400/1345], lr: 0.01000 Time 0.712 (0.752) Data 0.000 (0.009) Loss 2.0941 (1.9786) Prec@1 43.750 (50.261) Prec@5 76.562 (77.911) Epoch: [6][420/1345], lr: 0.01000 Time 0.730 (0.751) Data 0.000 (0.008) Loss 1.7698 (1.9797) Prec@1 56.250 (50.234) Prec@5 89.062 (77.936) Epoch: [6][440/1345], lr: 0.01000 Time 0.728 (0.751) Data 0.000 (0.008) Loss 2.0865 (1.9803) Prec@1 48.438 (50.188) Prec@5 78.125 (77.951) Epoch: [6][460/1345], lr: 0.01000 Time 0.815 (0.751) Data 0.000 (0.008) Loss 1.9663 (1.9838) Prec@1 46.875 (50.105) Prec@5 79.688 (77.915) Epoch: [6][480/1345], lr: 0.01000 Time 0.729 (0.751) Data 0.000 (0.007) Loss 2.2169 (1.9887) Prec@1 53.125 (49.997) Prec@5 73.438 (77.813) Epoch: [6][500/1345], lr: 0.01000 Time 0.773 (0.751) Data 0.000 (0.007) Loss 1.9639 (1.9900) Prec@1 53.125 (49.941) Prec@5 75.000 (77.769) Epoch: [6][520/1345], lr: 0.01000 Time 0.729 (0.751) Data 0.000 (0.007) Loss 1.9359 (1.9909) Prec@1 48.438 (49.937) Prec@5 71.875 (77.717) Epoch: [6][540/1345], lr: 0.01000 Time 0.729 (0.751) Data 0.000 (0.007) Loss 1.8902 (1.9889) Prec@1 54.688 (49.991) Prec@5 85.938 (77.747) Epoch: [6][560/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.000 (0.006) Loss 2.1349 (1.9918) Prec@1 43.750 (49.930) Prec@5 75.000 (77.668) Epoch: [6][580/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.000 (0.006) Loss 1.8948 (1.9896) Prec@1 57.812 (50.011) Prec@5 82.812 (77.724) Epoch: [6][600/1345], lr: 0.01000 Time 0.767 (0.750) Data 0.001 (0.006) Loss 1.8376 (1.9887) Prec@1 57.812 (50.055) Prec@5 79.688 (77.766) Epoch: [6][620/1345], lr: 0.01000 Time 0.843 (0.750) Data 0.001 (0.006) Loss 2.2988 (1.9896) Prec@1 45.312 (50.075) Prec@5 68.750 (77.712) Epoch: [6][640/1345], lr: 0.01000 Time 0.753 (0.750) Data 0.000 (0.006) Loss 1.9564 (1.9884) Prec@1 53.125 (50.124) Prec@5 73.438 (77.730) Epoch: [6][660/1345], lr: 0.01000 Time 0.733 (0.750) Data 0.000 (0.005) Loss 1.8237 (1.9903) Prec@1 50.000 (50.099) Prec@5 82.812 (77.726) Epoch: [6][680/1345], lr: 0.01000 Time 0.731 (0.750) Data 0.000 (0.005) Loss 2.0283 (1.9933) Prec@1 43.750 (50.009) Prec@5 73.438 (77.648) Epoch: [6][700/1345], lr: 0.01000 Time 0.725 (0.750) Data 0.000 (0.005) Loss 1.6217 (1.9921) Prec@1 46.875 (50.033) Prec@5 92.188 (77.646) Epoch: [6][720/1345], lr: 0.01000 Time 0.711 (0.750) Data 0.000 (0.005) Loss 1.6231 (1.9921) Prec@1 56.250 (50.020) Prec@5 85.938 (77.603) Epoch: [6][740/1345], lr: 0.01000 Time 0.710 (0.749) Data 0.000 (0.005) Loss 1.9723 (1.9927) Prec@1 56.250 (50.015) Prec@5 78.125 (77.604) Epoch: [6][760/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.005) Loss 1.9963 (1.9926) Prec@1 53.125 (50.021) Prec@5 78.125 (77.593) Epoch: [6][780/1345], lr: 0.01000 Time 0.841 (0.749) Data 0.000 (0.005) Loss 2.2920 (1.9932) Prec@1 42.188 (50.006) Prec@5 68.750 (77.581) Epoch: [6][800/1345], lr: 0.01000 Time 0.717 (0.749) Data 0.000 (0.005) Loss 2.1773 (1.9931) Prec@1 50.000 (49.994) Prec@5 76.562 (77.602) Epoch: [6][820/1345], lr: 0.01000 Time 0.723 (0.749) Data 0.000 (0.004) Loss 2.2345 (1.9899) Prec@1 45.312 (50.059) Prec@5 73.438 (77.638) Epoch: [6][840/1345], lr: 0.01000 Time 0.726 (0.749) Data 0.000 (0.004) Loss 1.6048 (1.9884) Prec@1 56.250 (50.126) Prec@5 82.812 (77.646) Epoch: [6][860/1345], lr: 0.01000 Time 0.758 (0.748) Data 0.000 (0.004) Loss 2.0721 (1.9920) Prec@1 43.750 (50.058) Prec@5 71.875 (77.604) Epoch: [6][880/1345], lr: 0.01000 Time 0.737 (0.748) Data 0.000 (0.004) Loss 1.8163 (1.9907) Prec@1 53.125 (50.055) Prec@5 78.125 (77.604) Epoch: [6][900/1345], lr: 0.01000 Time 0.836 (0.749) Data 0.000 (0.004) Loss 1.7202 (1.9905) Prec@1 60.938 (50.066) Prec@5 81.250 (77.600) Epoch: [6][920/1345], lr: 0.01000 Time 0.737 (0.749) Data 0.000 (0.004) Loss 1.9355 (1.9908) Prec@1 50.000 (50.046) Prec@5 81.250 (77.602) Epoch: [6][940/1345], lr: 0.01000 Time 0.761 (0.749) Data 0.000 (0.004) Loss 2.0975 (1.9903) Prec@1 45.312 (50.080) Prec@5 75.000 (77.602) Epoch: [6][960/1345], lr: 0.01000 Time 0.821 (0.749) Data 0.000 (0.004) Loss 2.0141 (1.9908) Prec@1 48.438 (50.086) Prec@5 75.000 (77.572) Epoch: [6][980/1345], lr: 0.01000 Time 0.764 (0.749) Data 0.000 (0.004) Loss 2.0809 (1.9907) Prec@1 46.875 (50.083) Prec@5 75.000 (77.572) Epoch: [6][1000/1345], lr: 0.01000 Time 0.715 (0.749) Data 0.000 (0.004) Loss 1.9777 (1.9916) Prec@1 50.000 (50.076) Prec@5 81.250 (77.557) Epoch: [6][1020/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.004) Loss 1.9709 (1.9911) Prec@1 45.312 (50.069) Prec@5 81.250 (77.560) Epoch: [6][1040/1345], lr: 0.01000 Time 0.718 (0.748) Data 0.001 (0.004) Loss 2.0386 (1.9924) Prec@1 45.312 (50.083) Prec@5 73.438 (77.523) Epoch: [6][1060/1345], lr: 0.01000 Time 0.749 (0.748) Data 0.000 (0.004) Loss 1.5686 (1.9911) Prec@1 62.500 (50.119) Prec@5 84.375 (77.548) Epoch: [6][1080/1345], lr: 0.01000 Time 0.762 (0.748) Data 0.000 (0.003) Loss 1.4844 (1.9900) Prec@1 60.938 (50.181) Prec@5 82.812 (77.550) Epoch: [6][1100/1345], lr: 0.01000 Time 0.723 (0.748) Data 0.001 (0.003) Loss 1.8777 (1.9884) Prec@1 53.125 (50.202) Prec@5 76.562 (77.560) Epoch: [6][1120/1345], lr: 0.01000 Time 0.761 (0.748) Data 0.000 (0.003) Loss 1.8258 (1.9877) Prec@1 48.438 (50.209) Prec@5 84.375 (77.580) Epoch: [6][1140/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.003) Loss 1.6349 (1.9893) Prec@1 59.375 (50.185) Prec@5 84.375 (77.558) Epoch: [6][1160/1345], lr: 0.01000 Time 0.842 (0.748) Data 0.000 (0.003) Loss 1.9715 (1.9897) Prec@1 56.250 (50.180) Prec@5 71.875 (77.545) Epoch: [6][1180/1345], lr: 0.01000 Time 0.840 (0.748) Data 0.000 (0.003) Loss 1.5874 (1.9896) Prec@1 64.062 (50.192) Prec@5 81.250 (77.547) Epoch: [6][1200/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.003) Loss 2.1250 (1.9888) Prec@1 50.000 (50.225) Prec@5 78.125 (77.573) Epoch: [6][1220/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.2183 (1.9884) Prec@1 43.750 (50.218) Prec@5 76.562 (77.584) Epoch: [6][1240/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.4640 (1.9883) Prec@1 43.750 (50.213) Prec@5 70.312 (77.570) Epoch: [6][1260/1345], lr: 0.01000 Time 0.735 (0.747) Data 0.000 (0.003) Loss 1.5170 (1.9865) Prec@1 57.812 (50.240) Prec@5 84.375 (77.623) Epoch: [6][1280/1345], lr: 0.01000 Time 0.712 (0.748) Data 0.000 (0.003) Loss 2.2638 (1.9864) Prec@1 43.750 (50.213) Prec@5 73.438 (77.638) Epoch: [6][1300/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.001 (0.003) Loss 2.1134 (1.9852) Prec@1 42.188 (50.228) Prec@5 76.562 (77.655) Epoch: [6][1320/1345], lr: 0.01000 Time 0.845 (0.747) Data 0.000 (0.003) Loss 1.8040 (1.9866) Prec@1 50.000 (50.185) Prec@5 79.688 (77.612) Epoch: [6][1340/1345], lr: 0.01000 Time 0.796 (0.748) Data 0.001 (0.003) Loss 2.0922 (1.9862) Prec@1 48.438 (50.184) Prec@5 73.438 (77.598) No BN layer Freezing. Test: [0/181] Time 3.488 (3.4884) Loss 2.8735 (2.8735) Prec@1 37.500 (37.500) Prec@5 70.312 (70.312) Test: [20/181] Time 1.160 (0.6096) Loss 2.6952 (2.6254) Prec@1 40.625 (41.295) Prec@5 65.625 (69.792) Test: [40/181] Time 1.009 (0.5359) Loss 2.8274 (2.7110) Prec@1 35.938 (39.520) Prec@5 64.062 (68.331) Test: [60/181] Time 0.868 (0.5082) Loss 2.9741 (2.7754) Prec@1 39.062 (38.678) Prec@5 56.250 (67.341) Test: [80/181] Time 1.449 (0.5011) Loss 2.4120 (2.7919) Prec@1 34.375 (37.828) Prec@5 73.438 (67.226) Test: [100/181] Time 0.981 (0.4901) Loss 3.0616 (2.7795) Prec@1 26.562 (37.562) Prec@5 65.625 (67.311) Test: [120/181] Time 1.524 (0.4902) Loss 3.3659 (2.7948) Prec@1 34.375 (37.280) Prec@5 59.375 (66.968) Test: [140/181] Time 0.963 (0.4837) Loss 3.2538 (2.7939) Prec@1 28.125 (37.456) Prec@5 56.250 (67.132) Test: [160/181] Time 1.031 (0.4811) Loss 3.0513 (2.7770) Prec@1 42.188 (37.869) Prec@5 60.938 (67.459) Testing Results: Prec@1 37.951 Prec@5 67.665 Loss 2.77882 Time 0.4766 No BN layer Freezing. Epoch: [7][0/1345], lr: 0.01000 Time 4.427 (4.427) Data 3.461 (3.461) Loss 1.9317 (1.9317) Prec@1 56.250 (56.250) Prec@5 81.250 (81.250) Epoch: [7][20/1345], lr: 0.01000 Time 0.712 (0.932) Data 0.000 (0.165) Loss 1.7197 (1.9019) Prec@1 59.375 (51.265) Prec@5 82.812 (78.795) Epoch: [7][40/1345], lr: 0.01000 Time 0.716 (0.837) Data 0.000 (0.085) Loss 2.0057 (1.8657) Prec@1 43.750 (52.401) Prec@5 78.125 (79.611) Epoch: [7][60/1345], lr: 0.01000 Time 0.726 (0.809) Data 0.001 (0.057) Loss 1.9333 (1.8692) Prec@1 45.312 (52.177) Prec@5 78.125 (79.534) Epoch: [7][80/1345], lr: 0.01000 Time 0.769 (0.797) Data 0.000 (0.043) Loss 1.8010 (1.8625) Prec@1 54.688 (52.373) Prec@5 76.562 (79.745) Epoch: [7][100/1345], lr: 0.01000 Time 0.837 (0.790) Data 0.001 (0.035) Loss 1.9256 (1.8777) Prec@1 59.375 (52.073) Prec@5 78.125 (79.610) Epoch: [7][120/1345], lr: 0.01000 Time 0.714 (0.782) Data 0.000 (0.029) Loss 1.7976 (1.8938) Prec@1 59.375 (51.705) Prec@5 81.250 (79.158) Epoch: [7][140/1345], lr: 0.01000 Time 0.713 (0.776) Data 0.000 (0.025) Loss 2.0277 (1.8913) Prec@1 50.000 (51.862) Prec@5 75.000 (79.056) Epoch: [7][160/1345], lr: 0.01000 Time 0.761 (0.771) Data 0.000 (0.022) Loss 1.8461 (1.8861) Prec@1 56.250 (51.990) Prec@5 82.812 (79.125) Epoch: [7][180/1345], lr: 0.01000 Time 0.776 (0.768) Data 0.001 (0.020) Loss 2.1071 (1.8870) Prec@1 45.312 (52.063) Prec@5 75.000 (79.144) Epoch: [7][200/1345], lr: 0.01000 Time 0.799 (0.765) Data 0.000 (0.018) Loss 1.7702 (1.8848) Prec@1 60.938 (52.239) Prec@5 79.688 (79.198) Epoch: [7][220/1345], lr: 0.01000 Time 0.714 (0.762) Data 0.000 (0.016) Loss 2.0102 (1.8846) Prec@1 48.438 (52.128) Prec@5 78.125 (79.362) Epoch: [7][240/1345], lr: 0.01000 Time 0.712 (0.760) Data 0.000 (0.015) Loss 1.9518 (1.8757) Prec@1 53.125 (52.341) Prec@5 79.688 (79.474) Epoch: [7][260/1345], lr: 0.01000 Time 0.740 (0.759) Data 0.000 (0.014) Loss 2.3810 (1.8739) Prec@1 39.062 (52.311) Prec@5 70.312 (79.544) Epoch: [7][280/1345], lr: 0.01000 Time 0.850 (0.758) Data 0.000 (0.013) Loss 1.7676 (1.8773) Prec@1 51.562 (52.185) Prec@5 79.688 (79.471) Epoch: [7][300/1345], lr: 0.01000 Time 0.721 (0.757) Data 0.000 (0.012) Loss 1.4892 (1.8762) Prec@1 60.938 (52.040) Prec@5 85.938 (79.532) Epoch: [7][320/1345], lr: 0.01000 Time 0.711 (0.755) Data 0.000 (0.011) Loss 1.8924 (1.8765) Prec@1 51.562 (52.015) Prec@5 79.688 (79.498) Epoch: [7][340/1345], lr: 0.01000 Time 0.840 (0.753) Data 0.000 (0.011) Loss 2.0586 (1.8777) Prec@1 51.562 (52.012) Prec@5 71.875 (79.463) Epoch: [7][360/1345], lr: 0.01000 Time 0.713 (0.753) Data 0.000 (0.010) Loss 1.9647 (1.8805) Prec@1 45.312 (51.974) Prec@5 75.000 (79.402) Epoch: [7][380/1345], lr: 0.01000 Time 0.710 (0.752) Data 0.000 (0.010) Loss 1.8360 (1.8876) Prec@1 53.125 (51.854) Prec@5 84.375 (79.323) Epoch: [7][400/1345], lr: 0.01000 Time 0.755 (0.751) Data 0.001 (0.009) Loss 1.4356 (1.8864) Prec@1 56.250 (51.804) Prec@5 90.625 (79.384) Epoch: [7][420/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.001 (0.009) Loss 1.7562 (1.8910) Prec@1 46.875 (51.715) Prec@5 87.500 (79.324) Epoch: [7][440/1345], lr: 0.01000 Time 0.756 (0.750) Data 0.000 (0.008) Loss 1.6921 (1.8935) Prec@1 57.812 (51.708) Prec@5 82.812 (79.273) Epoch: [7][460/1345], lr: 0.01000 Time 0.754 (0.750) Data 0.000 (0.008) Loss 1.8504 (1.8949) Prec@1 51.562 (51.664) Prec@5 85.938 (79.308) Epoch: [7][480/1345], lr: 0.01000 Time 0.729 (0.749) Data 0.000 (0.008) Loss 2.3144 (1.8957) Prec@1 51.562 (51.640) Prec@5 70.312 (79.272) Epoch: [7][500/1345], lr: 0.01000 Time 0.726 (0.750) Data 0.000 (0.007) Loss 2.2706 (1.8959) Prec@1 37.500 (51.656) Prec@5 65.625 (79.266) Epoch: [7][520/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.007) Loss 1.8378 (1.8988) Prec@1 53.125 (51.595) Prec@5 82.812 (79.256) Epoch: [7][540/1345], lr: 0.01000 Time 0.734 (0.749) Data 0.000 (0.007) Loss 2.1202 (1.9014) Prec@1 43.750 (51.580) Prec@5 75.000 (79.223) Epoch: [7][560/1345], lr: 0.01000 Time 0.726 (0.749) Data 0.000 (0.007) Loss 1.8153 (1.9022) Prec@1 53.125 (51.599) Prec@5 81.250 (79.195) Epoch: [7][580/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.006) Loss 1.9258 (1.9020) Prec@1 50.000 (51.600) Prec@5 79.688 (79.206) Epoch: [7][600/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.006) Loss 1.8472 (1.9013) Prec@1 54.688 (51.625) Prec@5 79.688 (79.243) Epoch: [7][620/1345], lr: 0.01000 Time 0.742 (0.748) Data 0.000 (0.006) Loss 2.2099 (1.9028) Prec@1 45.312 (51.600) Prec@5 73.438 (79.230) Epoch: [7][640/1345], lr: 0.01000 Time 0.723 (0.748) Data 0.000 (0.006) Loss 1.6252 (1.9012) Prec@1 62.500 (51.631) Prec@5 81.250 (79.271) Epoch: [7][660/1345], lr: 0.01000 Time 0.741 (0.748) Data 0.000 (0.006) Loss 1.6983 (1.9030) Prec@1 54.688 (51.560) Prec@5 79.688 (79.267) Epoch: [7][680/1345], lr: 0.01000 Time 0.766 (0.748) Data 0.000 (0.006) Loss 1.5618 (1.9033) Prec@1 51.562 (51.595) Prec@5 90.625 (79.242) Epoch: [7][700/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.005) Loss 2.1513 (1.9057) Prec@1 39.062 (51.560) Prec@5 76.562 (79.195) Epoch: [7][720/1345], lr: 0.01000 Time 0.777 (0.748) Data 0.001 (0.005) Loss 2.1564 (1.9071) Prec@1 46.875 (51.541) Prec@5 70.312 (79.157) Epoch: [7][740/1345], lr: 0.01000 Time 0.769 (0.748) Data 0.000 (0.005) Loss 1.8164 (1.9064) Prec@1 59.375 (51.567) Prec@5 82.812 (79.143) Epoch: [7][760/1345], lr: 0.01000 Time 0.725 (0.747) Data 0.001 (0.005) Loss 2.2145 (1.9075) Prec@1 43.750 (51.567) Prec@5 71.875 (79.113) Epoch: [7][780/1345], lr: 0.01000 Time 0.724 (0.747) Data 0.001 (0.005) Loss 1.8611 (1.9072) Prec@1 56.250 (51.595) Prec@5 81.250 (79.145) Epoch: [7][800/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.005) Loss 1.9261 (1.9071) Prec@1 53.125 (51.574) Prec@5 76.562 (79.161) Epoch: [7][820/1345], lr: 0.01000 Time 0.734 (0.747) Data 0.000 (0.005) Loss 1.8207 (1.9061) Prec@1 51.562 (51.557) Prec@5 73.438 (79.206) Epoch: [7][840/1345], lr: 0.01000 Time 0.729 (0.746) Data 0.000 (0.005) Loss 1.8954 (1.9063) Prec@1 53.125 (51.583) Prec@5 78.125 (79.212) Epoch: [7][860/1345], lr: 0.01000 Time 0.731 (0.746) Data 0.000 (0.005) Loss 1.7733 (1.9069) Prec@1 54.688 (51.581) Prec@5 78.125 (79.185) Epoch: [7][880/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.000 (0.004) Loss 1.7882 (1.9059) Prec@1 60.938 (51.612) Prec@5 76.562 (79.189) Epoch: [7][900/1345], lr: 0.01000 Time 0.714 (0.746) Data 0.000 (0.004) Loss 1.9007 (1.9055) Prec@1 53.125 (51.587) Prec@5 78.125 (79.216) Epoch: [7][920/1345], lr: 0.01000 Time 0.736 (0.746) Data 0.000 (0.004) Loss 1.6504 (1.9058) Prec@1 59.375 (51.603) Prec@5 81.250 (79.204) Epoch: [7][940/1345], lr: 0.01000 Time 0.711 (0.746) Data 0.000 (0.004) Loss 2.2238 (1.9086) Prec@1 43.750 (51.546) Prec@5 78.125 (79.136) Epoch: [7][960/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.000 (0.004) Loss 1.9267 (1.9087) Prec@1 53.125 (51.553) Prec@5 81.250 (79.136) Epoch: [7][980/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 1.7206 (1.9081) Prec@1 56.250 (51.596) Prec@5 81.250 (79.133) Epoch: [7][1000/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.001 (0.004) Loss 2.0034 (1.9072) Prec@1 50.000 (51.633) Prec@5 82.812 (79.163) Epoch: [7][1020/1345], lr: 0.01000 Time 0.898 (0.745) Data 0.000 (0.004) Loss 2.2882 (1.9078) Prec@1 50.000 (51.639) Prec@5 75.000 (79.167) Epoch: [7][1040/1345], lr: 0.01000 Time 0.893 (0.745) Data 0.001 (0.004) Loss 1.8623 (1.9076) Prec@1 59.375 (51.632) Prec@5 76.562 (79.155) Epoch: [7][1060/1345], lr: 0.01000 Time 0.724 (0.745) Data 0.000 (0.004) Loss 1.7789 (1.9088) Prec@1 48.438 (51.586) Prec@5 76.562 (79.146) Epoch: [7][1080/1345], lr: 0.01000 Time 0.732 (0.745) Data 0.000 (0.004) Loss 2.1707 (1.9098) Prec@1 43.750 (51.568) Prec@5 79.688 (79.117) Epoch: [7][1100/1345], lr: 0.01000 Time 0.722 (0.745) Data 0.000 (0.004) Loss 1.9402 (1.9097) Prec@1 57.812 (51.589) Prec@5 79.688 (79.113) Epoch: [7][1120/1345], lr: 0.01000 Time 0.764 (0.745) Data 0.001 (0.004) Loss 1.6890 (1.9087) Prec@1 56.250 (51.600) Prec@5 81.250 (79.141) Epoch: [7][1140/1345], lr: 0.01000 Time 0.724 (0.745) Data 0.001 (0.004) Loss 1.2741 (1.9084) Prec@1 67.188 (51.612) Prec@5 87.500 (79.125) Epoch: [7][1160/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.003) Loss 2.1750 (1.9106) Prec@1 43.750 (51.567) Prec@5 71.875 (79.105) Epoch: [7][1180/1345], lr: 0.01000 Time 0.802 (0.745) Data 0.000 (0.003) Loss 1.7983 (1.9115) Prec@1 53.125 (51.562) Prec@5 79.688 (79.086) Epoch: [7][1200/1345], lr: 0.01000 Time 0.845 (0.745) Data 0.000 (0.003) Loss 2.1012 (1.9124) Prec@1 37.500 (51.539) Prec@5 75.000 (79.064) Epoch: [7][1220/1345], lr: 0.01000 Time 0.728 (0.745) Data 0.000 (0.003) Loss 1.9195 (1.9118) Prec@1 56.250 (51.568) Prec@5 82.812 (79.077) Epoch: [7][1240/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.003) Loss 1.8757 (1.9115) Prec@1 51.562 (51.575) Prec@5 81.250 (79.081) Epoch: [7][1260/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.003) Loss 2.1441 (1.9114) Prec@1 43.750 (51.582) Prec@5 79.688 (79.087) Epoch: [7][1280/1345], lr: 0.01000 Time 0.758 (0.745) Data 0.000 (0.003) Loss 2.0755 (1.9115) Prec@1 53.125 (51.597) Prec@5 78.125 (79.075) Epoch: [7][1300/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.003) Loss 1.8840 (1.9112) Prec@1 53.125 (51.603) Prec@5 73.438 (79.080) Epoch: [7][1320/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.003) Loss 2.0262 (1.9118) Prec@1 46.875 (51.587) Prec@5 79.688 (79.057) Epoch: [7][1340/1345], lr: 0.01000 Time 0.784 (0.744) Data 0.000 (0.003) Loss 1.9379 (1.9123) Prec@1 57.812 (51.587) Prec@5 78.125 (79.043) No BN layer Freezing. Test: [0/181] Time 3.115 (3.1145) Loss 2.6894 (2.6894) Prec@1 42.188 (42.188) Prec@5 71.875 (71.875) Test: [20/181] Time 1.096 (0.6102) Loss 2.3327 (2.4927) Prec@1 45.312 (41.890) Prec@5 71.875 (71.949) Test: [40/181] Time 1.111 (0.5356) Loss 2.6484 (2.5373) Prec@1 40.625 (41.273) Prec@5 67.188 (71.265) Test: [60/181] Time 1.028 (0.5062) Loss 2.7216 (2.5854) Prec@1 42.188 (40.856) Prec@5 64.062 (70.236) Test: [80/181] Time 1.081 (0.4949) Loss 2.2938 (2.6195) Prec@1 45.312 (39.950) Prec@5 71.875 (69.387) Test: [100/181] Time 1.082 (0.4872) Loss 3.1721 (2.6188) Prec@1 31.250 (39.944) Prec@5 57.812 (69.725) Test: [120/181] Time 0.952 (0.4826) Loss 2.9547 (2.6220) Prec@1 40.625 (39.747) Prec@5 62.500 (70.041) Test: [140/181] Time 0.703 (0.4787) Loss 2.9738 (2.6212) Prec@1 28.125 (39.661) Prec@5 70.312 (70.146) Test: [160/181] Time 0.794 (0.4774) Loss 2.4770 (2.6036) Prec@1 42.188 (39.887) Prec@5 76.562 (70.303) Testing Results: Prec@1 40.069 Prec@5 70.286 Loss 2.59872 Time 0.4738 No BN layer Freezing. Epoch: [8][0/1345], lr: 0.01000 Time 4.120 (4.120) Data 3.365 (3.365) Loss 1.4196 (1.4196) Prec@1 62.500 (62.500) Prec@5 89.062 (89.062) Epoch: [8][20/1345], lr: 0.01000 Time 0.711 (0.903) Data 0.000 (0.161) Loss 1.7040 (1.8636) Prec@1 54.688 (53.348) Prec@5 87.500 (80.432) Epoch: [8][40/1345], lr: 0.01000 Time 0.712 (0.824) Data 0.000 (0.083) Loss 1.5010 (1.8454) Prec@1 71.875 (53.544) Prec@5 81.250 (80.907) Epoch: [8][60/1345], lr: 0.01000 Time 0.708 (0.799) Data 0.000 (0.056) Loss 2.1354 (1.8347) Prec@1 45.312 (53.765) Prec@5 65.625 (80.405) Epoch: [8][80/1345], lr: 0.01000 Time 0.754 (0.787) Data 0.000 (0.042) Loss 2.0248 (1.8294) Prec@1 46.875 (53.492) Prec@5 81.250 (80.710) Epoch: [8][100/1345], lr: 0.01000 Time 0.827 (0.781) Data 0.000 (0.034) Loss 1.7484 (1.8227) Prec@1 53.125 (53.620) Prec@5 82.812 (80.554) Epoch: [8][120/1345], lr: 0.01000 Time 0.845 (0.774) Data 0.000 (0.028) Loss 1.8506 (1.8248) Prec@1 57.812 (53.603) Prec@5 78.125 (80.540) Epoch: [8][140/1345], lr: 0.01000 Time 0.711 (0.770) Data 0.000 (0.024) Loss 1.8031 (1.8198) Prec@1 51.562 (53.657) Prec@5 79.688 (80.552) Epoch: [8][160/1345], lr: 0.01000 Time 0.752 (0.766) Data 0.000 (0.021) Loss 1.7041 (1.8281) Prec@1 48.438 (53.426) Prec@5 79.688 (80.260) Epoch: [8][180/1345], lr: 0.01000 Time 0.713 (0.765) Data 0.000 (0.019) Loss 1.3146 (1.8158) Prec@1 64.062 (53.703) Prec@5 89.062 (80.447) Epoch: [8][200/1345], lr: 0.01000 Time 0.761 (0.764) Data 0.001 (0.017) Loss 2.1145 (1.8210) Prec@1 51.562 (53.576) Prec@5 70.312 (80.496) Epoch: [8][220/1345], lr: 0.01000 Time 0.713 (0.762) Data 0.000 (0.016) Loss 1.5480 (1.8252) Prec@1 57.812 (53.500) Prec@5 87.500 (80.338) Epoch: [8][240/1345], lr: 0.01000 Time 0.711 (0.761) Data 0.000 (0.014) Loss 1.8628 (1.8244) Prec@1 48.438 (53.508) Prec@5 71.875 (80.342) Epoch: [8][260/1345], lr: 0.01000 Time 0.773 (0.760) Data 0.000 (0.013) Loss 2.1247 (1.8296) Prec@1 59.375 (53.388) Prec@5 71.875 (80.268) Epoch: [8][280/1345], lr: 0.01000 Time 0.740 (0.760) Data 0.000 (0.012) Loss 2.1643 (1.8336) Prec@1 45.312 (53.353) Prec@5 71.875 (80.166) Epoch: [8][300/1345], lr: 0.01000 Time 0.735 (0.760) Data 0.000 (0.012) Loss 1.2996 (1.8301) Prec@1 68.750 (53.411) Prec@5 84.375 (80.233) Epoch: [8][320/1345], lr: 0.01000 Time 0.808 (0.759) Data 0.000 (0.011) Loss 2.0248 (1.8339) Prec@1 45.312 (53.325) Prec@5 75.000 (80.223) Epoch: [8][340/1345], lr: 0.01000 Time 0.713 (0.758) Data 0.000 (0.010) Loss 1.7171 (1.8372) Prec@1 59.375 (53.267) Prec@5 82.812 (80.169) Epoch: [8][360/1345], lr: 0.01000 Time 0.711 (0.756) Data 0.000 (0.010) Loss 2.0873 (1.8377) Prec@1 48.438 (53.203) Prec@5 78.125 (80.107) Epoch: [8][380/1345], lr: 0.01000 Time 0.764 (0.756) Data 0.000 (0.009) Loss 1.6342 (1.8369) Prec@1 50.000 (53.187) Prec@5 85.938 (80.106) Epoch: [8][400/1345], lr: 0.01000 Time 0.731 (0.756) Data 0.000 (0.009) Loss 1.7443 (1.8356) Prec@1 60.938 (53.160) Prec@5 84.375 (80.151) Epoch: [8][420/1345], lr: 0.01000 Time 0.757 (0.754) Data 0.000 (0.008) Loss 1.8235 (1.8335) Prec@1 48.438 (53.192) Prec@5 79.688 (80.196) Epoch: [8][440/1345], lr: 0.01000 Time 0.732 (0.754) Data 0.000 (0.008) Loss 1.5112 (1.8348) Prec@1 57.812 (53.189) Prec@5 85.938 (80.120) Epoch: [8][460/1345], lr: 0.01000 Time 0.787 (0.753) Data 0.001 (0.008) Loss 2.2527 (1.8319) Prec@1 45.312 (53.240) Prec@5 71.875 (80.179) Epoch: [8][480/1345], lr: 0.01000 Time 0.732 (0.752) Data 0.000 (0.007) Loss 2.2075 (1.8331) Prec@1 45.312 (53.200) Prec@5 78.125 (80.175) Epoch: [8][500/1345], lr: 0.01000 Time 0.732 (0.752) Data 0.000 (0.007) Loss 1.7059 (1.8344) Prec@1 50.000 (53.116) Prec@5 76.562 (80.124) Epoch: [8][520/1345], lr: 0.01000 Time 0.761 (0.752) Data 0.000 (0.007) Loss 2.1830 (1.8350) Prec@1 50.000 (53.086) Prec@5 73.438 (80.083) Epoch: [8][540/1345], lr: 0.01000 Time 0.714 (0.751) Data 0.000 (0.007) Loss 1.9811 (1.8371) Prec@1 53.125 (53.050) Prec@5 79.688 (80.075) Epoch: [8][560/1345], lr: 0.01000 Time 0.713 (0.752) Data 0.000 (0.006) Loss 2.2229 (1.8392) Prec@1 45.312 (53.011) Prec@5 75.000 (80.072) Epoch: [8][580/1345], lr: 0.01000 Time 0.764 (0.752) Data 0.000 (0.006) Loss 1.9801 (1.8407) Prec@1 53.125 (52.883) Prec@5 78.125 (80.077) Epoch: [8][600/1345], lr: 0.01000 Time 0.846 (0.751) Data 0.000 (0.006) Loss 2.0251 (1.8410) Prec@1 50.000 (52.875) Prec@5 76.562 (80.072) Epoch: [8][620/1345], lr: 0.01000 Time 0.835 (0.751) Data 0.001 (0.006) Loss 1.9239 (1.8437) Prec@1 40.625 (52.831) Prec@5 84.375 (80.025) Epoch: [8][640/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.006) Loss 2.1217 (1.8444) Prec@1 43.750 (52.820) Prec@5 78.125 (79.990) Epoch: [8][660/1345], lr: 0.01000 Time 0.729 (0.751) Data 0.000 (0.006) Loss 1.8606 (1.8449) Prec@1 53.125 (52.782) Prec@5 75.000 (79.983) Epoch: [8][680/1345], lr: 0.01000 Time 0.715 (0.750) Data 0.000 (0.005) Loss 2.2996 (1.8488) Prec@1 48.438 (52.723) Prec@5 70.312 (79.905) Epoch: [8][700/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.000 (0.005) Loss 1.9010 (1.8474) Prec@1 50.000 (52.759) Prec@5 76.562 (79.913) Epoch: [8][720/1345], lr: 0.01000 Time 0.716 (0.750) Data 0.000 (0.005) Loss 1.7804 (1.8488) Prec@1 54.688 (52.754) Prec@5 85.938 (79.876) Epoch: [8][740/1345], lr: 0.01000 Time 0.746 (0.750) Data 0.000 (0.005) Loss 1.6268 (1.8485) Prec@1 56.250 (52.764) Prec@5 84.375 (79.913) Epoch: [8][760/1345], lr: 0.01000 Time 0.826 (0.750) Data 0.000 (0.005) Loss 2.0220 (1.8487) Prec@1 51.562 (52.772) Prec@5 75.000 (79.917) Epoch: [8][780/1345], lr: 0.01000 Time 0.856 (0.750) Data 0.000 (0.005) Loss 2.1295 (1.8494) Prec@1 56.250 (52.771) Prec@5 67.188 (79.870) Epoch: [8][800/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.000 (0.005) Loss 2.1430 (1.8515) Prec@1 40.625 (52.704) Prec@5 75.000 (79.838) Epoch: [8][820/1345], lr: 0.01000 Time 0.751 (0.750) Data 0.000 (0.005) Loss 1.8781 (1.8530) Prec@1 59.375 (52.630) Prec@5 79.688 (79.845) Epoch: [8][840/1345], lr: 0.01000 Time 0.754 (0.750) Data 0.000 (0.004) Loss 2.1581 (1.8533) Prec@1 42.188 (52.614) Prec@5 76.562 (79.868) Epoch: [8][860/1345], lr: 0.01000 Time 0.721 (0.750) Data 0.000 (0.004) Loss 1.7727 (1.8530) Prec@1 54.688 (52.608) Prec@5 81.250 (79.869) Epoch: [8][880/1345], lr: 0.01000 Time 0.718 (0.750) Data 0.000 (0.004) Loss 1.8580 (1.8522) Prec@1 50.000 (52.639) Prec@5 84.375 (79.856) Epoch: [8][900/1345], lr: 0.01000 Time 0.743 (0.750) Data 0.001 (0.004) Loss 1.8630 (1.8525) Prec@1 51.562 (52.636) Prec@5 76.562 (79.845) Epoch: [8][920/1345], lr: 0.01000 Time 0.841 (0.750) Data 0.000 (0.004) Loss 1.9853 (1.8544) Prec@1 48.438 (52.626) Prec@5 78.125 (79.803) Epoch: [8][940/1345], lr: 0.01000 Time 0.844 (0.750) Data 0.000 (0.004) Loss 1.9634 (1.8534) Prec@1 46.875 (52.663) Prec@5 76.562 (79.827) Epoch: [8][960/1345], lr: 0.01000 Time 0.725 (0.750) Data 0.000 (0.004) Loss 1.4254 (1.8507) Prec@1 67.188 (52.725) Prec@5 82.812 (79.866) Epoch: [8][980/1345], lr: 0.01000 Time 0.731 (0.750) Data 0.000 (0.004) Loss 1.6683 (1.8502) Prec@1 50.000 (52.714) Prec@5 87.500 (79.880) Epoch: [8][1000/1345], lr: 0.01000 Time 0.712 (0.750) Data 0.000 (0.004) Loss 1.6655 (1.8516) Prec@1 54.688 (52.666) Prec@5 82.812 (79.855) Epoch: [8][1020/1345], lr: 0.01000 Time 0.717 (0.749) Data 0.001 (0.004) Loss 1.6401 (1.8506) Prec@1 56.250 (52.712) Prec@5 79.688 (79.844) Epoch: [8][1040/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.004) Loss 1.7983 (1.8501) Prec@1 57.812 (52.735) Prec@5 82.812 (79.848) Epoch: [8][1060/1345], lr: 0.01000 Time 0.719 (0.749) Data 0.000 (0.004) Loss 2.0159 (1.8532) Prec@1 42.188 (52.663) Prec@5 79.688 (79.807) Epoch: [8][1080/1345], lr: 0.01000 Time 0.811 (0.749) Data 0.000 (0.004) Loss 1.8169 (1.8521) Prec@1 53.125 (52.699) Prec@5 79.688 (79.812) Epoch: [8][1100/1345], lr: 0.01000 Time 0.824 (0.749) Data 0.000 (0.003) Loss 1.7447 (1.8517) Prec@1 54.688 (52.696) Prec@5 82.812 (79.825) Epoch: [8][1120/1345], lr: 0.01000 Time 0.743 (0.749) Data 0.000 (0.003) Loss 1.7326 (1.8502) Prec@1 59.375 (52.740) Prec@5 85.938 (79.869) Epoch: [8][1140/1345], lr: 0.01000 Time 0.737 (0.749) Data 0.000 (0.003) Loss 1.7403 (1.8494) Prec@1 56.250 (52.746) Prec@5 82.812 (79.874) Epoch: [8][1160/1345], lr: 0.01000 Time 0.744 (0.750) Data 0.001 (0.003) Loss 1.7254 (1.8494) Prec@1 59.375 (52.755) Prec@5 75.000 (79.873) Epoch: [8][1180/1345], lr: 0.01000 Time 0.739 (0.750) Data 0.000 (0.003) Loss 1.7104 (1.8487) Prec@1 54.688 (52.798) Prec@5 82.812 (79.858) Epoch: [8][1200/1345], lr: 0.01000 Time 0.773 (0.750) Data 0.001 (0.003) Loss 1.7352 (1.8500) Prec@1 50.000 (52.758) Prec@5 81.250 (79.851) Epoch: [8][1220/1345], lr: 0.01000 Time 0.729 (0.749) Data 0.000 (0.003) Loss 1.9278 (1.8519) Prec@1 45.312 (52.723) Prec@5 71.875 (79.826) Epoch: [8][1240/1345], lr: 0.01000 Time 0.715 (0.749) Data 0.000 (0.003) Loss 1.9899 (1.8523) Prec@1 54.688 (52.720) Prec@5 78.125 (79.849) Epoch: [8][1260/1345], lr: 0.01000 Time 0.727 (0.749) Data 0.000 (0.003) Loss 1.9005 (1.8518) Prec@1 50.000 (52.746) Prec@5 84.375 (79.860) Epoch: [8][1280/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.003) Loss 1.4846 (1.8508) Prec@1 59.375 (52.743) Prec@5 84.375 (79.874) Epoch: [8][1300/1345], lr: 0.01000 Time 0.714 (0.749) Data 0.001 (0.003) Loss 1.3765 (1.8501) Prec@1 67.188 (52.765) Prec@5 84.375 (79.886) Epoch: [8][1320/1345], lr: 0.01000 Time 0.722 (0.749) Data 0.000 (0.003) Loss 1.6432 (1.8505) Prec@1 62.500 (52.769) Prec@5 79.688 (79.891) Epoch: [8][1340/1345], lr: 0.01000 Time 0.714 (0.749) Data 0.000 (0.003) Loss 1.8354 (1.8507) Prec@1 54.688 (52.751) Prec@5 73.438 (79.889) No BN layer Freezing. Test: [0/181] Time 3.116 (3.1163) Loss 2.6961 (2.6961) Prec@1 42.188 (42.188) Prec@5 71.875 (71.875) Test: [20/181] Time 1.081 (0.5831) Loss 2.4436 (2.5071) Prec@1 46.875 (42.560) Prec@5 70.312 (72.693) Test: [40/181] Time 0.723 (0.5185) Loss 2.6113 (2.5645) Prec@1 42.188 (41.730) Prec@5 70.312 (71.532) Test: [60/181] Time 0.579 (0.4983) Loss 2.4704 (2.5851) Prec@1 40.625 (41.060) Prec@5 70.312 (70.953) Test: [80/181] Time 1.264 (0.4923) Loss 2.4266 (2.6306) Prec@1 43.750 (40.374) Prec@5 73.438 (70.467) Test: [100/181] Time 0.927 (0.4812) Loss 3.1671 (2.6248) Prec@1 29.688 (40.114) Prec@5 59.375 (70.483) Test: [120/181] Time 0.834 (0.4771) Loss 2.7348 (2.6227) Prec@1 37.500 (40.096) Prec@5 65.625 (70.545) Test: [140/181] Time 0.832 (0.4738) Loss 2.9998 (2.6242) Prec@1 31.250 (40.060) Prec@5 64.062 (70.645) Test: [160/181] Time 1.264 (0.4744) Loss 2.4763 (2.6023) Prec@1 46.875 (40.382) Prec@5 75.000 (70.943) Testing Results: Prec@1 40.382 Prec@5 70.851 Loss 2.60441 Time 0.4706 No BN layer Freezing. Epoch: [9][0/1345], lr: 0.01000 Time 4.061 (4.061) Data 3.315 (3.315) Loss 1.7563 (1.7563) Prec@1 53.125 (53.125) Prec@5 79.688 (79.688) Epoch: [9][20/1345], lr: 0.01000 Time 0.715 (0.900) Data 0.000 (0.158) Loss 1.6343 (1.8608) Prec@1 56.250 (51.190) Prec@5 84.375 (80.357) Epoch: [9][40/1345], lr: 0.01000 Time 0.762 (0.823) Data 0.000 (0.081) Loss 1.6724 (1.8363) Prec@1 56.250 (52.248) Prec@5 84.375 (80.297) Epoch: [9][60/1345], lr: 0.01000 Time 0.746 (0.798) Data 0.001 (0.055) Loss 1.7930 (1.7904) Prec@1 57.812 (53.304) Prec@5 79.688 (81.148) Epoch: [9][80/1345], lr: 0.01000 Time 0.851 (0.786) Data 0.001 (0.041) Loss 1.6580 (1.7824) Prec@1 59.375 (53.897) Prec@5 85.938 (81.327) Epoch: [9][100/1345], lr: 0.01000 Time 0.836 (0.782) Data 0.000 (0.033) Loss 1.7117 (1.7988) Prec@1 56.250 (53.728) Prec@5 82.812 (80.724) Epoch: [9][120/1345], lr: 0.01000 Time 0.733 (0.777) Data 0.001 (0.028) Loss 2.0996 (1.8020) Prec@1 54.688 (53.848) Prec@5 75.000 (80.617) Epoch: [9][140/1345], lr: 0.01000 Time 0.732 (0.771) Data 0.000 (0.024) Loss 1.8798 (1.8037) Prec@1 60.938 (53.779) Prec@5 79.688 (80.618) Epoch: [9][160/1345], lr: 0.01000 Time 0.714 (0.768) Data 0.000 (0.021) Loss 1.9485 (1.7940) Prec@1 53.125 (53.911) Prec@5 79.688 (80.910) Epoch: [9][180/1345], lr: 0.01000 Time 0.757 (0.765) Data 0.000 (0.019) Loss 1.8827 (1.7995) Prec@1 45.312 (53.833) Prec@5 79.688 (80.862) Epoch: [9][200/1345], lr: 0.01000 Time 0.765 (0.764) Data 0.000 (0.017) Loss 1.4796 (1.7896) Prec@1 60.938 (54.136) Prec@5 87.500 (80.993) Epoch: [9][220/1345], lr: 0.01000 Time 0.713 (0.762) Data 0.000 (0.015) Loss 1.6545 (1.7906) Prec@1 54.688 (54.150) Prec@5 84.375 (81.031) Epoch: [9][240/1345], lr: 0.01000 Time 0.739 (0.761) Data 0.001 (0.014) Loss 2.1543 (1.7921) Prec@1 40.625 (54.020) Prec@5 76.562 (80.952) Epoch: [9][260/1345], lr: 0.01000 Time 0.736 (0.761) Data 0.000 (0.013) Loss 1.1991 (1.7904) Prec@1 70.312 (54.185) Prec@5 90.625 (80.933) Epoch: [9][280/1345], lr: 0.01000 Time 0.736 (0.760) Data 0.000 (0.012) Loss 1.8750 (1.7881) Prec@1 59.375 (54.354) Prec@5 76.562 (80.900) Epoch: [9][300/1345], lr: 0.01000 Time 0.714 (0.758) Data 0.000 (0.011) Loss 1.4844 (1.7870) Prec@1 71.875 (54.366) Prec@5 85.938 (80.985) Epoch: [9][320/1345], lr: 0.01000 Time 0.729 (0.758) Data 0.000 (0.011) Loss 1.9405 (1.7860) Prec@1 53.125 (54.376) Prec@5 75.000 (80.938) Epoch: [9][340/1345], lr: 0.01000 Time 0.712 (0.757) Data 0.000 (0.010) Loss 1.9973 (1.7902) Prec@1 54.688 (54.298) Prec@5 78.125 (80.847) Epoch: [9][360/1345], lr: 0.01000 Time 0.713 (0.756) Data 0.000 (0.010) Loss 1.7735 (1.7929) Prec@1 54.688 (54.211) Prec@5 79.688 (80.774) Epoch: [9][380/1345], lr: 0.01000 Time 0.712 (0.755) Data 0.000 (0.009) Loss 1.9660 (1.7948) Prec@1 53.125 (54.245) Prec@5 78.125 (80.717) Epoch: [9][400/1345], lr: 0.01000 Time 0.841 (0.754) Data 0.000 (0.009) Loss 1.7671 (1.7916) Prec@1 53.125 (54.271) Prec@5 76.562 (80.767) Epoch: [9][420/1345], lr: 0.01000 Time 0.744 (0.754) Data 0.000 (0.008) Loss 2.0549 (1.7906) Prec@1 40.625 (54.209) Prec@5 76.562 (80.782) Epoch: [9][440/1345], lr: 0.01000 Time 0.719 (0.753) Data 0.000 (0.008) Loss 2.0530 (1.7895) Prec@1 50.000 (54.220) Prec@5 75.000 (80.804) Epoch: [9][460/1345], lr: 0.01000 Time 0.771 (0.752) Data 0.000 (0.008) Loss 1.7566 (1.7855) Prec@1 56.250 (54.335) Prec@5 78.125 (80.826) Epoch: [9][480/1345], lr: 0.01000 Time 0.711 (0.752) Data 0.000 (0.007) Loss 1.6136 (1.7850) Prec@1 64.062 (54.372) Prec@5 87.500 (80.847) Epoch: [9][500/1345], lr: 0.01000 Time 0.712 (0.752) Data 0.000 (0.007) Loss 1.6890 (1.7867) Prec@1 60.938 (54.329) Prec@5 85.938 (80.810) Epoch: [9][520/1345], lr: 0.01000 Time 0.714 (0.752) Data 0.000 (0.007) Loss 1.6477 (1.7865) Prec@1 53.125 (54.307) Prec@5 84.375 (80.812) Epoch: [9][540/1345], lr: 0.01000 Time 0.709 (0.751) Data 0.000 (0.007) Loss 2.1781 (1.7862) Prec@1 46.875 (54.283) Prec@5 75.000 (80.854) Epoch: [9][560/1345], lr: 0.01000 Time 0.721 (0.752) Data 0.001 (0.006) Loss 1.7766 (1.7875) Prec@1 56.250 (54.264) Prec@5 78.125 (80.821) Epoch: [9][580/1345], lr: 0.01000 Time 0.856 (0.751) Data 0.001 (0.006) Loss 1.7926 (1.7899) Prec@1 45.312 (54.198) Prec@5 84.375 (80.790) Epoch: [9][600/1345], lr: 0.01000 Time 0.730 (0.751) Data 0.000 (0.006) Loss 1.9980 (1.7918) Prec@1 50.000 (54.121) Prec@5 81.250 (80.761) Epoch: [9][620/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.006) Loss 2.1014 (1.7924) Prec@1 39.062 (54.051) Prec@5 67.188 (80.717) Epoch: [9][640/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.006) Loss 1.8401 (1.7936) Prec@1 51.562 (54.012) Prec@5 79.688 (80.699) Epoch: [9][660/1345], lr: 0.01000 Time 0.776 (0.751) Data 0.001 (0.005) Loss 2.0921 (1.7936) Prec@1 48.438 (54.019) Prec@5 75.000 (80.711) Epoch: [9][680/1345], lr: 0.01000 Time 0.735 (0.751) Data 0.000 (0.005) Loss 1.6831 (1.7936) Prec@1 51.562 (54.027) Prec@5 85.938 (80.695) Epoch: [9][700/1345], lr: 0.01000 Time 0.734 (0.751) Data 0.001 (0.005) Loss 1.5506 (1.7925) Prec@1 60.938 (54.048) Prec@5 85.938 (80.717) Epoch: [9][720/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.005) Loss 1.9159 (1.7936) Prec@1 51.562 (54.044) Prec@5 82.812 (80.695) Epoch: [9][740/1345], lr: 0.01000 Time 0.721 (0.751) Data 0.000 (0.005) Loss 2.0791 (1.7937) Prec@1 50.000 (54.080) Prec@5 76.562 (80.681) Epoch: [9][760/1345], lr: 0.01000 Time 0.769 (0.751) Data 0.000 (0.005) Loss 1.8124 (1.7940) Prec@1 57.812 (54.106) Prec@5 79.688 (80.683) Epoch: [9][780/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.005) Loss 1.6113 (1.7939) Prec@1 56.250 (54.119) Prec@5 81.250 (80.702) Epoch: [9][800/1345], lr: 0.01000 Time 0.733 (0.751) Data 0.000 (0.005) Loss 1.8611 (1.7936) Prec@1 54.688 (54.151) Prec@5 76.562 (80.723) Epoch: [9][820/1345], lr: 0.01000 Time 0.734 (0.751) Data 0.000 (0.004) Loss 1.8801 (1.7934) Prec@1 56.250 (54.149) Prec@5 79.688 (80.734) Epoch: [9][840/1345], lr: 0.01000 Time 0.712 (0.751) Data 0.000 (0.004) Loss 1.9665 (1.7961) Prec@1 48.438 (54.063) Prec@5 75.000 (80.691) Epoch: [9][860/1345], lr: 0.01000 Time 0.744 (0.751) Data 0.000 (0.004) Loss 1.7689 (1.7955) Prec@1 43.750 (54.045) Prec@5 82.812 (80.733) Epoch: [9][880/1345], lr: 0.01000 Time 0.732 (0.751) Data 0.001 (0.004) Loss 1.7775 (1.7958) Prec@1 56.250 (54.033) Prec@5 79.688 (80.734) Epoch: [9][900/1345], lr: 0.01000 Time 0.723 (0.751) Data 0.000 (0.004) Loss 1.5026 (1.7961) Prec@1 67.188 (54.029) Prec@5 81.250 (80.730) Epoch: [9][920/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.004) Loss 1.9037 (1.7966) Prec@1 51.562 (54.029) Prec@5 81.250 (80.712) Epoch: [9][940/1345], lr: 0.01000 Time 0.769 (0.751) Data 0.001 (0.004) Loss 1.4968 (1.7967) Prec@1 59.375 (54.045) Prec@5 90.625 (80.715) Epoch: [9][960/1345], lr: 0.01000 Time 0.712 (0.751) Data 0.000 (0.004) Loss 2.3560 (1.7974) Prec@1 40.625 (54.055) Prec@5 68.750 (80.686) Epoch: [9][980/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.004) Loss 2.4355 (1.7987) Prec@1 39.062 (54.027) Prec@5 73.438 (80.654) Epoch: [9][1000/1345], lr: 0.01000 Time 0.722 (0.751) Data 0.000 (0.004) Loss 1.9534 (1.7978) Prec@1 59.375 (54.085) Prec@5 76.562 (80.665) Epoch: [9][1020/1345], lr: 0.01000 Time 0.733 (0.751) Data 0.000 (0.004) Loss 1.6124 (1.7976) Prec@1 64.062 (54.106) Prec@5 84.375 (80.667) Epoch: [9][1040/1345], lr: 0.01000 Time 0.764 (0.751) Data 0.000 (0.004) Loss 1.9788 (1.7990) Prec@1 46.875 (54.111) Prec@5 81.250 (80.632) Epoch: [9][1060/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.004) Loss 1.5541 (1.7998) Prec@1 59.375 (54.085) Prec@5 82.812 (80.609) Epoch: [9][1080/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.004) Loss 1.4634 (1.7995) Prec@1 59.375 (54.096) Prec@5 89.062 (80.640) Epoch: [9][1100/1345], lr: 0.01000 Time 0.712 (0.750) Data 0.000 (0.003) Loss 2.0357 (1.7991) Prec@1 48.438 (54.135) Prec@5 76.562 (80.661) Epoch: [9][1120/1345], lr: 0.01000 Time 0.714 (0.750) Data 0.000 (0.003) Loss 1.8875 (1.7998) Prec@1 51.562 (54.127) Prec@5 81.250 (80.658) Epoch: [9][1140/1345], lr: 0.01000 Time 0.762 (0.750) Data 0.000 (0.003) Loss 1.8754 (1.7991) Prec@1 51.562 (54.127) Prec@5 81.250 (80.687) Epoch: [9][1160/1345], lr: 0.01000 Time 0.715 (0.750) Data 0.000 (0.003) Loss 1.5115 (1.7983) Prec@1 59.375 (54.144) Prec@5 81.250 (80.704) Epoch: [9][1180/1345], lr: 0.01000 Time 0.714 (0.750) Data 0.000 (0.003) Loss 1.8453 (1.7987) Prec@1 48.438 (54.136) Prec@5 84.375 (80.701) Epoch: [9][1200/1345], lr: 0.01000 Time 0.852 (0.750) Data 0.000 (0.003) Loss 1.8514 (1.7981) Prec@1 57.812 (54.154) Prec@5 79.688 (80.710) Epoch: [9][1220/1345], lr: 0.01000 Time 0.842 (0.749) Data 0.000 (0.003) Loss 1.7922 (1.7998) Prec@1 50.000 (54.089) Prec@5 79.688 (80.693) Epoch: [9][1240/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.003) Loss 1.6234 (1.7999) Prec@1 56.250 (54.086) Prec@5 84.375 (80.697) Epoch: [9][1260/1345], lr: 0.01000 Time 0.736 (0.749) Data 0.000 (0.003) Loss 1.7952 (1.7995) Prec@1 56.250 (54.088) Prec@5 81.250 (80.701) Epoch: [9][1280/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.003) Loss 1.4541 (1.7996) Prec@1 65.625 (54.120) Prec@5 79.688 (80.696) Epoch: [9][1300/1345], lr: 0.01000 Time 0.764 (0.749) Data 0.000 (0.003) Loss 1.6883 (1.8010) Prec@1 56.250 (54.081) Prec@5 82.812 (80.686) Epoch: [9][1320/1345], lr: 0.01000 Time 0.752 (0.749) Data 0.000 (0.003) Loss 2.2198 (1.8009) Prec@1 45.312 (54.091) Prec@5 76.562 (80.680) Epoch: [9][1340/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.003) Loss 1.7152 (1.8008) Prec@1 50.000 (54.089) Prec@5 81.250 (80.680) No BN layer Freezing. Test: [0/181] Time 2.829 (2.8294) Loss 2.9519 (2.9519) Prec@1 42.188 (42.188) Prec@5 70.312 (70.312) Test: [20/181] Time 0.922 (0.5727) Loss 2.5034 (2.5037) Prec@1 46.875 (45.312) Prec@5 73.438 (73.438) Test: [40/181] Time 0.471 (0.5168) Loss 2.5889 (2.6106) Prec@1 42.188 (42.912) Prec@5 73.438 (71.723) Test: [60/181] Time 0.239 (0.4976) Loss 2.3618 (2.6556) Prec@1 45.312 (42.264) Prec@5 70.312 (70.697) Test: [80/181] Time 0.239 (0.4886) Loss 2.4123 (2.6973) Prec@1 42.188 (41.184) Prec@5 79.688 (69.830) Test: [100/181] Time 0.313 (0.4865) Loss 2.8786 (2.7031) Prec@1 31.250 (40.919) Prec@5 62.500 (69.910) Test: [120/181] Time 0.299 (0.4807) Loss 2.9139 (2.7048) Prec@1 42.188 (40.754) Prec@5 60.938 (69.990) Test: [140/181] Time 0.303 (0.4779) Loss 3.2590 (2.7107) Prec@1 32.812 (40.669) Prec@5 62.500 (70.024) Test: [160/181] Time 0.366 (0.4757) Loss 2.5713 (2.7000) Prec@1 43.750 (40.839) Prec@5 71.875 (70.186) Testing Results: Prec@1 41.042 Prec@5 70.365 Loss 2.69019 Time 0.4738 No BN layer Freezing. Epoch: [10][0/1345], lr: 0.01000 Time 4.234 (4.234) Data 3.477 (3.477) Loss 1.6606 (1.6606) Prec@1 57.812 (57.812) Prec@5 85.938 (85.938) Epoch: [10][20/1345], lr: 0.01000 Time 0.714 (0.914) Data 0.000 (0.166) Loss 1.8188 (1.7334) Prec@1 45.312 (55.952) Prec@5 79.688 (81.771) Epoch: [10][40/1345], lr: 0.01000 Time 0.710 (0.827) Data 0.000 (0.085) Loss 1.7690 (1.7293) Prec@1 54.688 (55.145) Prec@5 78.125 (82.050) Epoch: [10][60/1345], lr: 0.01000 Time 0.713 (0.804) Data 0.001 (0.057) Loss 1.8752 (1.7544) Prec@1 46.875 (54.764) Prec@5 82.812 (81.967) Epoch: [10][80/1345], lr: 0.01000 Time 0.755 (0.791) Data 0.001 (0.043) Loss 2.2654 (1.7591) Prec@1 42.188 (54.533) Prec@5 76.562 (82.137) Epoch: [10][100/1345], lr: 0.01000 Time 0.839 (0.783) Data 0.001 (0.035) Loss 1.9115 (1.7487) Prec@1 51.562 (54.950) Prec@5 81.250 (82.287) Epoch: [10][120/1345], lr: 0.01000 Time 0.723 (0.776) Data 0.000 (0.029) Loss 1.7792 (1.7467) Prec@1 57.812 (54.985) Prec@5 79.688 (82.335) Epoch: [10][140/1345], lr: 0.01000 Time 0.740 (0.771) Data 0.000 (0.025) Loss 1.7255 (1.7409) Prec@1 48.438 (55.064) Prec@5 85.938 (82.270) Epoch: [10][160/1345], lr: 0.01000 Time 0.742 (0.768) Data 0.000 (0.022) Loss 1.5277 (1.7379) Prec@1 53.125 (55.231) Prec@5 84.375 (82.279) Epoch: [10][180/1345], lr: 0.01000 Time 0.714 (0.766) Data 0.000 (0.020) Loss 1.9322 (1.7504) Prec@1 48.438 (54.938) Prec@5 76.562 (81.992) Epoch: [10][200/1345], lr: 0.01000 Time 0.847 (0.764) Data 0.001 (0.018) Loss 1.7867 (1.7497) Prec@1 59.375 (55.084) Prec@5 81.250 (81.965) Epoch: [10][220/1345], lr: 0.01000 Time 0.711 (0.761) Data 0.000 (0.016) Loss 1.8737 (1.7362) Prec@1 54.688 (55.423) Prec@5 76.562 (82.204) Epoch: [10][240/1345], lr: 0.01000 Time 0.712 (0.760) Data 0.000 (0.015) Loss 1.6202 (1.7354) Prec@1 60.938 (55.485) Prec@5 82.812 (82.132) Epoch: [10][260/1345], lr: 0.01000 Time 0.840 (0.758) Data 0.000 (0.014) Loss 1.6927 (1.7364) Prec@1 56.250 (55.358) Prec@5 87.500 (82.184) Epoch: [10][280/1345], lr: 0.01000 Time 0.711 (0.757) Data 0.000 (0.013) Loss 1.6725 (1.7371) Prec@1 51.562 (55.349) Prec@5 84.375 (82.112) Epoch: [10][300/1345], lr: 0.01000 Time 0.731 (0.756) Data 0.000 (0.012) Loss 1.3549 (1.7401) Prec@1 59.375 (55.238) Prec@5 90.625 (82.029) Epoch: [10][320/1345], lr: 0.01000 Time 0.710 (0.755) Data 0.000 (0.011) Loss 1.9911 (1.7380) Prec@1 48.438 (55.296) Prec@5 75.000 (81.985) Epoch: [10][340/1345], lr: 0.01000 Time 0.714 (0.754) Data 0.000 (0.011) Loss 1.5348 (1.7304) Prec@1 53.125 (55.476) Prec@5 90.625 (82.134) Epoch: [10][360/1345], lr: 0.01000 Time 0.718 (0.753) Data 0.001 (0.010) Loss 1.7711 (1.7310) Prec@1 56.250 (55.484) Prec@5 76.562 (82.146) Epoch: [10][380/1345], lr: 0.01000 Time 0.727 (0.753) Data 0.000 (0.010) Loss 1.8164 (1.7310) Prec@1 53.125 (55.516) Prec@5 78.125 (82.050) Epoch: [10][400/1345], lr: 0.01000 Time 0.711 (0.753) Data 0.000 (0.009) Loss 1.8668 (1.7347) Prec@1 48.438 (55.451) Prec@5 84.375 (81.963) Epoch: [10][420/1345], lr: 0.01000 Time 0.713 (0.752) Data 0.000 (0.009) Loss 1.6079 (1.7355) Prec@1 56.250 (55.419) Prec@5 81.250 (81.992) Epoch: [10][440/1345], lr: 0.01000 Time 0.768 (0.752) Data 0.001 (0.008) Loss 2.1153 (1.7354) Prec@1 45.312 (55.414) Prec@5 70.312 (81.998) Epoch: [10][460/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.008) Loss 1.5133 (1.7356) Prec@1 64.062 (55.460) Prec@5 87.500 (81.955) Epoch: [10][480/1345], lr: 0.01000 Time 0.763 (0.751) Data 0.000 (0.008) Loss 1.6485 (1.7364) Prec@1 53.125 (55.412) Prec@5 85.938 (81.987) Epoch: [10][500/1345], lr: 0.01000 Time 0.776 (0.751) Data 0.000 (0.007) Loss 1.6302 (1.7356) Prec@1 59.375 (55.424) Prec@5 81.250 (82.027) Epoch: [10][520/1345], lr: 0.01000 Time 0.719 (0.750) Data 0.000 (0.007) Loss 1.5720 (1.7338) Prec@1 59.375 (55.452) Prec@5 87.500 (82.099) Epoch: [10][540/1345], lr: 0.01000 Time 0.716 (0.750) Data 0.001 (0.007) Loss 1.8602 (1.7343) Prec@1 54.688 (55.493) Prec@5 75.000 (82.062) Epoch: [10][560/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.001 (0.007) Loss 1.6715 (1.7325) Prec@1 57.812 (55.543) Prec@5 85.938 (82.130) Epoch: [10][580/1345], lr: 0.01000 Time 0.891 (0.749) Data 0.001 (0.006) Loss 1.6427 (1.7308) Prec@1 59.375 (55.567) Prec@5 81.250 (82.162) Epoch: [10][600/1345], lr: 0.01000 Time 0.752 (0.749) Data 0.000 (0.006) Loss 2.0167 (1.7304) Prec@1 56.250 (55.608) Prec@5 76.562 (82.124) Epoch: [10][620/1345], lr: 0.01000 Time 0.771 (0.749) Data 0.000 (0.006) Loss 1.8024 (1.7328) Prec@1 51.562 (55.563) Prec@5 82.812 (82.095) Epoch: [10][640/1345], lr: 0.01000 Time 0.717 (0.749) Data 0.000 (0.006) Loss 2.0062 (1.7355) Prec@1 42.188 (55.507) Prec@5 79.688 (82.040) Epoch: [10][660/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.006) Loss 1.7026 (1.7367) Prec@1 57.812 (55.477) Prec@5 85.938 (82.021) Epoch: [10][680/1345], lr: 0.01000 Time 0.741 (0.748) Data 0.001 (0.006) Loss 1.4980 (1.7387) Prec@1 54.688 (55.449) Prec@5 85.938 (82.035) Epoch: [10][700/1345], lr: 0.01000 Time 0.763 (0.747) Data 0.001 (0.005) Loss 1.6732 (1.7392) Prec@1 62.500 (55.481) Prec@5 82.812 (82.010) Epoch: [10][720/1345], lr: 0.01000 Time 0.716 (0.747) Data 0.001 (0.005) Loss 1.8526 (1.7408) Prec@1 53.125 (55.431) Prec@5 82.812 (81.989) Epoch: [10][740/1345], lr: 0.01000 Time 0.743 (0.748) Data 0.001 (0.005) Loss 2.1524 (1.7428) Prec@1 46.875 (55.440) Prec@5 78.125 (81.942) Epoch: [10][760/1345], lr: 0.01000 Time 0.737 (0.748) Data 0.000 (0.005) Loss 1.9766 (1.7405) Prec@1 51.562 (55.525) Prec@5 76.562 (81.979) Epoch: [10][780/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.000 (0.005) Loss 1.9876 (1.7434) Prec@1 54.688 (55.458) Prec@5 78.125 (81.904) Epoch: [10][800/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.005) Loss 1.3172 (1.7417) Prec@1 62.500 (55.454) Prec@5 85.938 (81.913) Epoch: [10][820/1345], lr: 0.01000 Time 0.713 (0.747) Data 0.001 (0.005) Loss 1.9809 (1.7427) Prec@1 51.562 (55.447) Prec@5 79.688 (81.884) Epoch: [10][840/1345], lr: 0.01000 Time 0.714 (0.746) Data 0.000 (0.005) Loss 1.7658 (1.7440) Prec@1 46.875 (55.420) Prec@5 82.812 (81.846) Epoch: [10][860/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.001 (0.005) Loss 1.8937 (1.7447) Prec@1 51.562 (55.421) Prec@5 79.688 (81.842) Epoch: [10][880/1345], lr: 0.01000 Time 0.716 (0.746) Data 0.000 (0.004) Loss 1.4677 (1.7425) Prec@1 59.375 (55.425) Prec@5 89.062 (81.874) Epoch: [10][900/1345], lr: 0.01000 Time 0.732 (0.746) Data 0.000 (0.004) Loss 1.9695 (1.7420) Prec@1 50.000 (55.412) Prec@5 75.000 (81.860) Epoch: [10][920/1345], lr: 0.01000 Time 0.756 (0.746) Data 0.001 (0.004) Loss 2.1210 (1.7418) Prec@1 51.562 (55.425) Prec@5 76.562 (81.864) Epoch: [10][940/1345], lr: 0.01000 Time 0.714 (0.746) Data 0.001 (0.004) Loss 1.9074 (1.7409) Prec@1 53.125 (55.460) Prec@5 82.812 (81.888) Epoch: [10][960/1345], lr: 0.01000 Time 0.750 (0.746) Data 0.001 (0.004) Loss 1.8888 (1.7431) Prec@1 50.000 (55.385) Prec@5 84.375 (81.865) Epoch: [10][980/1345], lr: 0.01000 Time 0.822 (0.746) Data 0.000 (0.004) Loss 1.8279 (1.7448) Prec@1 48.438 (55.339) Prec@5 79.688 (81.846) Epoch: [10][1000/1345], lr: 0.01000 Time 0.714 (0.746) Data 0.000 (0.004) Loss 1.8393 (1.7457) Prec@1 54.688 (55.301) Prec@5 81.250 (81.846) Epoch: [10][1020/1345], lr: 0.01000 Time 0.715 (0.745) Data 0.000 (0.004) Loss 2.0373 (1.7478) Prec@1 40.625 (55.234) Prec@5 81.250 (81.819) Epoch: [10][1040/1345], lr: 0.01000 Time 0.714 (0.745) Data 0.000 (0.004) Loss 1.8084 (1.7485) Prec@1 56.250 (55.220) Prec@5 82.812 (81.835) Epoch: [10][1060/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.000 (0.004) Loss 1.8230 (1.7500) Prec@1 54.688 (55.181) Prec@5 79.688 (81.779) Epoch: [10][1080/1345], lr: 0.01000 Time 0.745 (0.745) Data 0.001 (0.004) Loss 1.9780 (1.7500) Prec@1 46.875 (55.191) Prec@5 76.562 (81.749) Epoch: [10][1100/1345], lr: 0.01000 Time 0.732 (0.745) Data 0.000 (0.004) Loss 1.9705 (1.7515) Prec@1 54.688 (55.153) Prec@5 81.250 (81.717) Epoch: [10][1120/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 1.6616 (1.7501) Prec@1 53.125 (55.177) Prec@5 81.250 (81.736) Epoch: [10][1140/1345], lr: 0.01000 Time 0.714 (0.745) Data 0.000 (0.004) Loss 1.5455 (1.7500) Prec@1 60.938 (55.170) Prec@5 81.250 (81.755) Epoch: [10][1160/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.003) Loss 1.9395 (1.7502) Prec@1 51.562 (55.163) Prec@5 79.688 (81.757) Epoch: [10][1180/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.003) Loss 1.5009 (1.7505) Prec@1 60.938 (55.158) Prec@5 85.938 (81.741) Epoch: [10][1200/1345], lr: 0.01000 Time 0.735 (0.744) Data 0.001 (0.003) Loss 1.9780 (1.7513) Prec@1 48.438 (55.143) Prec@5 79.688 (81.722) Epoch: [10][1220/1345], lr: 0.01000 Time 0.713 (0.744) Data 0.000 (0.003) Loss 1.7437 (1.7512) Prec@1 51.562 (55.155) Prec@5 82.812 (81.723) Epoch: [10][1240/1345], lr: 0.01000 Time 0.716 (0.744) Data 0.000 (0.003) Loss 2.0037 (1.7503) Prec@1 48.438 (55.155) Prec@5 78.125 (81.747) Epoch: [10][1260/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.003) Loss 1.6054 (1.7507) Prec@1 51.562 (55.152) Prec@5 85.938 (81.726) Epoch: [10][1280/1345], lr: 0.01000 Time 0.833 (0.744) Data 0.000 (0.003) Loss 1.7668 (1.7500) Prec@1 54.688 (55.161) Prec@5 79.688 (81.735) Epoch: [10][1300/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.003) Loss 1.8413 (1.7505) Prec@1 46.875 (55.156) Prec@5 79.688 (81.721) Epoch: [10][1320/1345], lr: 0.01000 Time 0.752 (0.744) Data 0.000 (0.003) Loss 1.3842 (1.7515) Prec@1 65.625 (55.117) Prec@5 84.375 (81.696) Epoch: [10][1340/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.003) Loss 1.6896 (1.7512) Prec@1 57.812 (55.131) Prec@5 84.375 (81.692) No BN layer Freezing. Test: [0/181] Time 3.693 (3.6934) Loss 3.0506 (3.0506) Prec@1 32.812 (32.812) Prec@5 71.875 (71.875) Test: [20/181] Time 0.934 (0.6048) Loss 2.3123 (2.4559) Prec@1 43.750 (41.592) Prec@5 73.438 (73.438) Test: [40/181] Time 1.250 (0.5333) Loss 2.7102 (2.5030) Prec@1 42.188 (41.387) Prec@5 71.875 (72.713) Test: [60/181] Time 1.346 (0.5112) Loss 2.5895 (2.5519) Prec@1 40.625 (40.523) Prec@5 65.625 (71.542) Test: [80/181] Time 1.162 (0.4963) Loss 2.2049 (2.5904) Prec@1 42.188 (39.931) Prec@5 75.000 (70.756) Test: [100/181] Time 1.106 (0.4861) Loss 2.6855 (2.5979) Prec@1 37.500 (39.774) Prec@5 70.312 (70.869) Test: [120/181] Time 0.900 (0.4827) Loss 2.7298 (2.6000) Prec@1 45.312 (40.134) Prec@5 68.750 (70.881) Test: [140/181] Time 1.031 (0.4808) Loss 3.0142 (2.6054) Prec@1 34.375 (40.016) Prec@5 62.500 (70.811) Test: [160/181] Time 0.893 (0.4780) Loss 2.9563 (2.5929) Prec@1 39.062 (40.431) Prec@5 71.875 (71.060) Testing Results: Prec@1 40.608 Prec@5 71.033 Loss 2.59062 Time 0.4721 No BN layer Freezing. Epoch: [11][0/1345], lr: 0.01000 Time 4.688 (4.688) Data 3.820 (3.820) Loss 1.5919 (1.5919) Prec@1 57.812 (57.812) Prec@5 79.688 (79.688) Epoch: [11][20/1345], lr: 0.01000 Time 0.741 (0.924) Data 0.000 (0.182) Loss 1.7009 (1.6807) Prec@1 53.125 (56.250) Prec@5 84.375 (82.961) Epoch: [11][40/1345], lr: 0.01000 Time 0.728 (0.833) Data 0.000 (0.094) Loss 1.8100 (1.6564) Prec@1 50.000 (57.127) Prec@5 75.000 (83.232) Epoch: [11][60/1345], lr: 0.01000 Time 0.733 (0.804) Data 0.000 (0.063) Loss 1.5983 (1.6934) Prec@1 57.812 (56.506) Prec@5 81.250 (82.044) Epoch: [11][80/1345], lr: 0.01000 Time 0.733 (0.790) Data 0.000 (0.048) Loss 1.6187 (1.6801) Prec@1 56.250 (56.925) Prec@5 82.812 (82.079) Epoch: [11][100/1345], lr: 0.01000 Time 0.712 (0.780) Data 0.000 (0.038) Loss 1.8296 (1.6822) Prec@1 56.250 (56.590) Prec@5 81.250 (82.240) Epoch: [11][120/1345], lr: 0.01000 Time 0.728 (0.773) Data 0.001 (0.032) Loss 1.6364 (1.6866) Prec@1 59.375 (56.224) Prec@5 81.250 (82.038) Epoch: [11][140/1345], lr: 0.01000 Time 0.713 (0.769) Data 0.000 (0.028) Loss 1.6902 (1.6858) Prec@1 54.688 (56.161) Prec@5 84.375 (82.203) Epoch: [11][160/1345], lr: 0.01000 Time 0.851 (0.767) Data 0.000 (0.024) Loss 1.8685 (1.6808) Prec@1 54.688 (56.240) Prec@5 76.562 (82.405) Epoch: [11][180/1345], lr: 0.01000 Time 0.736 (0.764) Data 0.000 (0.022) Loss 1.6769 (1.6764) Prec@1 60.938 (56.362) Prec@5 81.250 (82.528) Epoch: [11][200/1345], lr: 0.01000 Time 0.717 (0.762) Data 0.000 (0.019) Loss 1.6997 (1.6809) Prec@1 56.250 (56.133) Prec@5 87.500 (82.494) Epoch: [11][220/1345], lr: 0.01000 Time 0.819 (0.759) Data 0.000 (0.018) Loss 1.8501 (1.6824) Prec@1 50.000 (56.144) Prec@5 82.812 (82.544) Epoch: [11][240/1345], lr: 0.01000 Time 0.713 (0.756) Data 0.000 (0.016) Loss 1.5806 (1.6792) Prec@1 59.375 (56.237) Prec@5 85.938 (82.709) Epoch: [11][260/1345], lr: 0.01000 Time 0.711 (0.755) Data 0.000 (0.015) Loss 1.5522 (1.6915) Prec@1 59.375 (55.969) Prec@5 84.375 (82.537) Epoch: [11][280/1345], lr: 0.01000 Time 0.840 (0.753) Data 0.000 (0.014) Loss 1.9449 (1.6956) Prec@1 46.875 (55.900) Prec@5 81.250 (82.512) Epoch: [11][300/1345], lr: 0.01000 Time 0.737 (0.752) Data 0.000 (0.013) Loss 1.6590 (1.6968) Prec@1 54.688 (56.001) Prec@5 85.938 (82.428) Epoch: [11][320/1345], lr: 0.01000 Time 0.713 (0.751) Data 0.000 (0.012) Loss 1.7013 (1.6993) Prec@1 50.000 (55.934) Prec@5 79.688 (82.374) Epoch: [11][340/1345], lr: 0.01000 Time 0.841 (0.751) Data 0.000 (0.012) Loss 1.7479 (1.7014) Prec@1 51.562 (55.989) Prec@5 82.812 (82.341) Epoch: [11][360/1345], lr: 0.01000 Time 0.724 (0.750) Data 0.000 (0.011) Loss 1.5286 (1.7007) Prec@1 65.625 (56.038) Prec@5 81.250 (82.267) Epoch: [11][380/1345], lr: 0.01000 Time 0.720 (0.749) Data 0.001 (0.010) Loss 1.9571 (1.7011) Prec@1 48.438 (56.020) Prec@5 78.125 (82.206) Epoch: [11][400/1345], lr: 0.01000 Time 0.842 (0.748) Data 0.001 (0.010) Loss 1.8439 (1.6976) Prec@1 48.438 (56.145) Prec@5 84.375 (82.240) Epoch: [11][420/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.010) Loss 2.0007 (1.6961) Prec@1 50.000 (56.168) Prec@5 73.438 (82.271) Epoch: [11][440/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.001 (0.009) Loss 1.7889 (1.6990) Prec@1 46.875 (56.140) Prec@5 84.375 (82.231) Epoch: [11][460/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.000 (0.009) Loss 1.2859 (1.6992) Prec@1 64.062 (56.074) Prec@5 96.875 (82.246) Epoch: [11][480/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.000 (0.008) Loss 1.7163 (1.6997) Prec@1 56.250 (56.130) Prec@5 79.688 (82.273) Epoch: [11][500/1345], lr: 0.01000 Time 0.772 (0.747) Data 0.000 (0.008) Loss 1.5734 (1.7009) Prec@1 54.688 (56.088) Prec@5 87.500 (82.320) Epoch: [11][520/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.000 (0.008) Loss 1.5388 (1.7007) Prec@1 59.375 (56.103) Prec@5 87.500 (82.330) Epoch: [11][540/1345], lr: 0.01000 Time 0.713 (0.747) Data 0.000 (0.008) Loss 1.4782 (1.6998) Prec@1 59.375 (56.172) Prec@5 81.250 (82.365) Epoch: [11][560/1345], lr: 0.01000 Time 0.777 (0.747) Data 0.000 (0.007) Loss 1.8117 (1.6978) Prec@1 51.562 (56.222) Prec@5 79.688 (82.414) Epoch: [11][580/1345], lr: 0.01000 Time 0.757 (0.746) Data 0.000 (0.007) Loss 1.8213 (1.6989) Prec@1 56.250 (56.253) Prec@5 79.688 (82.414) Epoch: [11][600/1345], lr: 0.01000 Time 0.719 (0.746) Data 0.000 (0.007) Loss 1.7966 (1.6997) Prec@1 56.250 (56.221) Prec@5 81.250 (82.397) Epoch: [11][620/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.007) Loss 1.3669 (1.6993) Prec@1 60.938 (56.240) Prec@5 90.625 (82.415) Epoch: [11][640/1345], lr: 0.01000 Time 0.760 (0.746) Data 0.001 (0.006) Loss 1.9295 (1.6997) Prec@1 45.312 (56.243) Prec@5 81.250 (82.393) Epoch: [11][660/1345], lr: 0.01000 Time 0.748 (0.746) Data 0.000 (0.006) Loss 1.5238 (1.7002) Prec@1 67.188 (56.222) Prec@5 79.688 (82.370) Epoch: [11][680/1345], lr: 0.01000 Time 0.713 (0.746) Data 0.000 (0.006) Loss 1.4248 (1.7006) Prec@1 59.375 (56.229) Prec@5 85.938 (82.333) Epoch: [11][700/1345], lr: 0.01000 Time 0.719 (0.746) Data 0.000 (0.006) Loss 1.9313 (1.7027) Prec@1 50.000 (56.154) Prec@5 76.562 (82.315) Epoch: [11][720/1345], lr: 0.01000 Time 0.713 (0.746) Data 0.000 (0.006) Loss 1.9642 (1.7080) Prec@1 60.938 (56.025) Prec@5 75.000 (82.214) Epoch: [11][740/1345], lr: 0.01000 Time 0.744 (0.746) Data 0.001 (0.006) Loss 1.7606 (1.7105) Prec@1 54.688 (56.005) Prec@5 82.812 (82.176) Epoch: [11][760/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.000 (0.005) Loss 1.4224 (1.7087) Prec@1 64.062 (56.055) Prec@5 89.062 (82.229) Epoch: [11][780/1345], lr: 0.01000 Time 0.761 (0.746) Data 0.000 (0.005) Loss 1.6696 (1.7099) Prec@1 54.688 (56.058) Prec@5 87.500 (82.208) Epoch: [11][800/1345], lr: 0.01000 Time 0.713 (0.746) Data 0.001 (0.005) Loss 1.3803 (1.7104) Prec@1 60.938 (56.039) Prec@5 90.625 (82.184) Epoch: [11][820/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.005) Loss 2.0509 (1.7112) Prec@1 46.875 (56.020) Prec@5 79.688 (82.177) Epoch: [11][840/1345], lr: 0.01000 Time 0.817 (0.745) Data 0.000 (0.005) Loss 1.7936 (1.7126) Prec@1 50.000 (55.994) Prec@5 76.562 (82.149) Epoch: [11][860/1345], lr: 0.01000 Time 0.832 (0.745) Data 0.000 (0.005) Loss 1.6242 (1.7120) Prec@1 57.812 (55.980) Prec@5 81.250 (82.177) Epoch: [11][880/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.005) Loss 1.5821 (1.7106) Prec@1 57.812 (55.995) Prec@5 87.500 (82.208) Epoch: [11][900/1345], lr: 0.01000 Time 0.741 (0.745) Data 0.001 (0.005) Loss 1.9662 (1.7116) Prec@1 45.312 (55.950) Prec@5 78.125 (82.190) Epoch: [11][920/1345], lr: 0.01000 Time 0.730 (0.745) Data 0.000 (0.005) Loss 1.5020 (1.7116) Prec@1 56.250 (55.946) Prec@5 85.938 (82.186) Epoch: [11][940/1345], lr: 0.01000 Time 0.715 (0.745) Data 0.001 (0.005) Loss 1.8344 (1.7119) Prec@1 51.562 (55.953) Prec@5 79.688 (82.172) Epoch: [11][960/1345], lr: 0.01000 Time 0.759 (0.745) Data 0.000 (0.004) Loss 1.9415 (1.7118) Prec@1 53.125 (55.952) Prec@5 75.000 (82.183) Epoch: [11][980/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 1.9130 (1.7130) Prec@1 51.562 (55.908) Prec@5 81.250 (82.171) Epoch: [11][1000/1345], lr: 0.01000 Time 0.831 (0.745) Data 0.000 (0.004) Loss 1.6477 (1.7123) Prec@1 64.062 (55.950) Prec@5 82.812 (82.158) Epoch: [11][1020/1345], lr: 0.01000 Time 0.840 (0.745) Data 0.001 (0.004) Loss 1.5549 (1.7117) Prec@1 53.125 (55.944) Prec@5 89.062 (82.185) Epoch: [11][1040/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 1.6176 (1.7098) Prec@1 54.688 (55.971) Prec@5 82.812 (82.221) Epoch: [11][1060/1345], lr: 0.01000 Time 0.737 (0.745) Data 0.000 (0.004) Loss 1.5955 (1.7104) Prec@1 62.500 (55.976) Prec@5 87.500 (82.210) Epoch: [11][1080/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.004) Loss 2.1250 (1.7113) Prec@1 50.000 (55.967) Prec@5 78.125 (82.213) Epoch: [11][1100/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.004) Loss 1.7722 (1.7119) Prec@1 59.375 (55.960) Prec@5 84.375 (82.218) Epoch: [11][1120/1345], lr: 0.01000 Time 0.711 (0.745) Data 0.000 (0.004) Loss 1.5712 (1.7144) Prec@1 62.500 (55.906) Prec@5 84.375 (82.145) Epoch: [11][1140/1345], lr: 0.01000 Time 0.817 (0.745) Data 0.000 (0.004) Loss 1.8679 (1.7174) Prec@1 53.125 (55.834) Prec@5 81.250 (82.100) Epoch: [11][1160/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.000 (0.004) Loss 2.1222 (1.7187) Prec@1 48.438 (55.803) Prec@5 79.688 (82.078) Epoch: [11][1180/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.004) Loss 1.3120 (1.7181) Prec@1 67.188 (55.811) Prec@5 92.188 (82.105) Epoch: [11][1200/1345], lr: 0.01000 Time 0.787 (0.745) Data 0.001 (0.004) Loss 1.9167 (1.7198) Prec@1 53.125 (55.770) Prec@5 76.562 (82.072) Epoch: [11][1220/1345], lr: 0.01000 Time 0.726 (0.745) Data 0.000 (0.004) Loss 1.6915 (1.7206) Prec@1 59.375 (55.748) Prec@5 81.250 (82.057) Epoch: [11][1240/1345], lr: 0.01000 Time 0.721 (0.745) Data 0.000 (0.004) Loss 1.9248 (1.7209) Prec@1 51.562 (55.749) Prec@5 73.438 (82.042) Epoch: [11][1260/1345], lr: 0.01000 Time 0.753 (0.745) Data 0.000 (0.004) Loss 1.2638 (1.7196) Prec@1 68.750 (55.787) Prec@5 89.062 (82.055) Epoch: [11][1280/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.003) Loss 1.3348 (1.7211) Prec@1 68.750 (55.769) Prec@5 92.188 (82.021) Epoch: [11][1300/1345], lr: 0.01000 Time 0.774 (0.744) Data 0.001 (0.003) Loss 1.8056 (1.7216) Prec@1 50.000 (55.758) Prec@5 81.250 (82.017) Epoch: [11][1320/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.003) Loss 1.4668 (1.7218) Prec@1 62.500 (55.739) Prec@5 82.812 (81.998) Epoch: [11][1340/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.003) Loss 2.2542 (1.7218) Prec@1 46.875 (55.754) Prec@5 70.312 (82.016) No BN layer Freezing. Test: [0/181] Time 3.583 (3.5829) Loss 2.8489 (2.8489) Prec@1 39.062 (39.062) Prec@5 67.188 (67.188) Test: [20/181] Time 1.259 (0.6058) Loss 2.3539 (2.5554) Prec@1 50.000 (42.113) Prec@5 78.125 (72.545) Test: [40/181] Time 1.003 (0.5318) Loss 2.2583 (2.6108) Prec@1 46.875 (40.396) Prec@5 76.562 (71.418) Test: [60/181] Time 1.083 (0.5039) Loss 2.8089 (2.6610) Prec@1 40.625 (39.677) Prec@5 62.500 (70.389) Test: [80/181] Time 0.959 (0.4902) Loss 2.1932 (2.6849) Prec@1 42.188 (39.603) Prec@5 78.125 (70.004) Test: [100/181] Time 0.799 (0.4825) Loss 3.1240 (2.6865) Prec@1 32.812 (39.821) Prec@5 65.625 (70.158) Test: [120/181] Time 0.520 (0.4786) Loss 2.8921 (2.6855) Prec@1 40.625 (39.824) Prec@5 67.188 (70.287) Test: [140/181] Time 0.514 (0.4753) Loss 3.2941 (2.6883) Prec@1 25.000 (39.550) Prec@5 64.062 (70.346) Test: [160/181] Time 0.516 (0.4733) Loss 2.8129 (2.6698) Prec@1 39.062 (39.839) Prec@5 76.562 (70.613) Testing Results: Prec@1 39.922 Prec@5 70.547 Loss 2.66317 Time 0.4723 No BN layer Freezing. Epoch: [12][0/1345], lr: 0.01000 Time 4.322 (4.322) Data 3.578 (3.578) Loss 1.5962 (1.5962) Prec@1 56.250 (56.250) Prec@5 84.375 (84.375) Epoch: [12][20/1345], lr: 0.01000 Time 0.715 (0.923) Data 0.001 (0.171) Loss 1.9611 (1.5493) Prec@1 45.312 (59.077) Prec@5 82.812 (86.086) Epoch: [12][40/1345], lr: 0.01000 Time 0.738 (0.836) Data 0.000 (0.088) Loss 2.1198 (1.5708) Prec@1 50.000 (59.032) Prec@5 71.875 (84.985) Epoch: [12][60/1345], lr: 0.01000 Time 0.712 (0.802) Data 0.000 (0.059) Loss 1.7270 (1.5833) Prec@1 57.812 (59.016) Prec@5 85.938 (84.913) Epoch: [12][80/1345], lr: 0.01000 Time 0.711 (0.785) Data 0.000 (0.045) Loss 1.6543 (1.5962) Prec@1 51.562 (58.738) Prec@5 85.938 (84.549) Epoch: [12][100/1345], lr: 0.01000 Time 0.843 (0.774) Data 0.000 (0.036) Loss 1.5135 (1.6037) Prec@1 57.812 (58.679) Prec@5 85.938 (84.112) Epoch: [12][120/1345], lr: 0.01000 Time 0.836 (0.770) Data 0.000 (0.030) Loss 1.6861 (1.6282) Prec@1 62.500 (58.252) Prec@5 78.125 (83.613) Epoch: [12][140/1345], lr: 0.01000 Time 0.716 (0.767) Data 0.000 (0.026) Loss 1.4942 (1.6308) Prec@1 56.250 (58.045) Prec@5 89.062 (83.477) Epoch: [12][160/1345], lr: 0.01000 Time 0.714 (0.765) Data 0.001 (0.023) Loss 1.4071 (1.6368) Prec@1 65.625 (57.812) Prec@5 85.938 (83.414) Epoch: [12][180/1345], lr: 0.01000 Time 0.740 (0.763) Data 0.000 (0.020) Loss 1.6498 (1.6514) Prec@1 53.125 (57.562) Prec@5 81.250 (83.175) Epoch: [12][200/1345], lr: 0.01000 Time 0.713 (0.761) Data 0.000 (0.018) Loss 1.9133 (1.6623) Prec@1 51.562 (57.432) Prec@5 75.000 (82.898) Epoch: [12][220/1345], lr: 0.01000 Time 0.776 (0.760) Data 0.000 (0.017) Loss 1.6177 (1.6568) Prec@1 60.938 (57.445) Prec@5 82.812 (83.010) Epoch: [12][240/1345], lr: 0.01000 Time 0.733 (0.759) Data 0.000 (0.015) Loss 1.7824 (1.6666) Prec@1 57.812 (57.158) Prec@5 85.938 (82.981) Epoch: [12][260/1345], lr: 0.01000 Time 0.774 (0.758) Data 0.000 (0.014) Loss 1.3196 (1.6644) Prec@1 64.062 (57.196) Prec@5 90.625 (83.010) Epoch: [12][280/1345], lr: 0.01000 Time 0.712 (0.756) Data 0.000 (0.013) Loss 1.8048 (1.6696) Prec@1 54.688 (57.056) Prec@5 82.812 (82.907) Epoch: [12][300/1345], lr: 0.01000 Time 0.714 (0.755) Data 0.001 (0.012) Loss 1.2367 (1.6695) Prec@1 73.438 (57.055) Prec@5 87.500 (82.890) Epoch: [12][320/1345], lr: 0.01000 Time 0.714 (0.755) Data 0.000 (0.012) Loss 1.3724 (1.6679) Prec@1 60.938 (57.131) Prec@5 87.500 (82.832) Epoch: [12][340/1345], lr: 0.01000 Time 0.713 (0.754) Data 0.000 (0.011) Loss 1.5554 (1.6747) Prec@1 57.812 (56.960) Prec@5 87.500 (82.698) Epoch: [12][360/1345], lr: 0.01000 Time 0.843 (0.754) Data 0.000 (0.010) Loss 1.7836 (1.6780) Prec@1 56.250 (56.938) Prec@5 82.812 (82.678) Epoch: [12][380/1345], lr: 0.01000 Time 0.878 (0.754) Data 0.001 (0.010) Loss 1.5243 (1.6769) Prec@1 64.062 (56.943) Prec@5 84.375 (82.681) Epoch: [12][400/1345], lr: 0.01000 Time 0.713 (0.754) Data 0.000 (0.009) Loss 1.6851 (1.6803) Prec@1 56.250 (56.799) Prec@5 84.375 (82.625) Epoch: [12][420/1345], lr: 0.01000 Time 0.761 (0.753) Data 0.000 (0.009) Loss 1.3793 (1.6830) Prec@1 60.938 (56.781) Prec@5 89.062 (82.586) Epoch: [12][440/1345], lr: 0.01000 Time 0.718 (0.753) Data 0.000 (0.009) Loss 1.7538 (1.6827) Prec@1 54.688 (56.764) Prec@5 79.688 (82.639) Epoch: [12][460/1345], lr: 0.01000 Time 0.712 (0.752) Data 0.000 (0.008) Loss 1.7929 (1.6816) Prec@1 50.000 (56.748) Prec@5 78.125 (82.599) Epoch: [12][480/1345], lr: 0.01000 Time 0.722 (0.752) Data 0.000 (0.008) Loss 1.6466 (1.6777) Prec@1 56.250 (56.783) Prec@5 79.688 (82.683) Epoch: [12][500/1345], lr: 0.01000 Time 0.739 (0.752) Data 0.000 (0.008) Loss 2.1070 (1.6764) Prec@1 48.438 (56.814) Prec@5 76.562 (82.685) Epoch: [12][520/1345], lr: 0.01000 Time 0.828 (0.751) Data 0.000 (0.007) Loss 1.8138 (1.6741) Prec@1 51.562 (56.853) Prec@5 90.625 (82.768) Epoch: [12][540/1345], lr: 0.01000 Time 0.830 (0.751) Data 0.000 (0.007) Loss 1.4397 (1.6724) Prec@1 59.375 (56.897) Prec@5 85.938 (82.804) Epoch: [12][560/1345], lr: 0.01000 Time 0.714 (0.751) Data 0.000 (0.007) Loss 1.7909 (1.6718) Prec@1 62.500 (56.918) Prec@5 76.562 (82.779) Epoch: [12][580/1345], lr: 0.01000 Time 0.723 (0.751) Data 0.000 (0.007) Loss 1.1588 (1.6703) Prec@1 70.312 (56.906) Prec@5 89.062 (82.831) Epoch: [12][600/1345], lr: 0.01000 Time 0.750 (0.751) Data 0.000 (0.006) Loss 1.6581 (1.6722) Prec@1 57.812 (56.832) Prec@5 84.375 (82.815) Epoch: [12][620/1345], lr: 0.01000 Time 0.714 (0.750) Data 0.000 (0.006) Loss 1.6838 (1.6751) Prec@1 50.000 (56.806) Prec@5 87.500 (82.765) Epoch: [12][640/1345], lr: 0.01000 Time 0.718 (0.750) Data 0.000 (0.006) Loss 1.4939 (1.6741) Prec@1 60.938 (56.857) Prec@5 84.375 (82.727) Epoch: [12][660/1345], lr: 0.01000 Time 0.718 (0.750) Data 0.000 (0.006) Loss 1.6821 (1.6728) Prec@1 57.812 (56.874) Prec@5 82.812 (82.737) Epoch: [12][680/1345], lr: 0.01000 Time 0.716 (0.750) Data 0.000 (0.006) Loss 1.9425 (1.6724) Prec@1 48.438 (56.844) Prec@5 79.688 (82.732) Epoch: [12][700/1345], lr: 0.01000 Time 0.716 (0.750) Data 0.000 (0.006) Loss 1.4918 (1.6717) Prec@1 60.938 (56.845) Prec@5 81.250 (82.757) Epoch: [12][720/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.005) Loss 2.0756 (1.6717) Prec@1 43.750 (56.855) Prec@5 78.125 (82.743) Epoch: [12][740/1345], lr: 0.01000 Time 0.710 (0.749) Data 0.000 (0.005) Loss 1.7906 (1.6713) Prec@1 56.250 (56.864) Prec@5 81.250 (82.749) Epoch: [12][760/1345], lr: 0.01000 Time 0.855 (0.749) Data 0.001 (0.005) Loss 1.6870 (1.6725) Prec@1 56.250 (56.841) Prec@5 85.938 (82.757) Epoch: [12][780/1345], lr: 0.01000 Time 0.733 (0.749) Data 0.000 (0.005) Loss 1.3741 (1.6717) Prec@1 60.938 (56.852) Prec@5 85.938 (82.800) Epoch: [12][800/1345], lr: 0.01000 Time 0.737 (0.749) Data 0.000 (0.005) Loss 1.9082 (1.6735) Prec@1 53.125 (56.843) Prec@5 84.375 (82.799) Epoch: [12][820/1345], lr: 0.01000 Time 0.742 (0.749) Data 0.000 (0.005) Loss 1.8014 (1.6719) Prec@1 51.562 (56.870) Prec@5 75.000 (82.822) Epoch: [12][840/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.005) Loss 1.8063 (1.6720) Prec@1 54.688 (56.871) Prec@5 81.250 (82.812) Epoch: [12][860/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.005) Loss 1.7033 (1.6723) Prec@1 54.688 (56.836) Prec@5 81.250 (82.791) Epoch: [12][880/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.005) Loss 2.1284 (1.6725) Prec@1 50.000 (56.849) Prec@5 75.000 (82.773) Epoch: [12][900/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.004) Loss 1.8881 (1.6761) Prec@1 53.125 (56.782) Prec@5 81.250 (82.724) Epoch: [12][920/1345], lr: 0.01000 Time 0.820 (0.748) Data 0.000 (0.004) Loss 1.2193 (1.6755) Prec@1 70.312 (56.805) Prec@5 87.500 (82.733) Epoch: [12][940/1345], lr: 0.01000 Time 0.730 (0.748) Data 0.001 (0.004) Loss 1.2859 (1.6759) Prec@1 60.938 (56.791) Prec@5 90.625 (82.723) Epoch: [12][960/1345], lr: 0.01000 Time 0.715 (0.749) Data 0.000 (0.004) Loss 1.5850 (1.6752) Prec@1 54.688 (56.803) Prec@5 87.500 (82.752) Epoch: [12][980/1345], lr: 0.01000 Time 0.746 (0.749) Data 0.001 (0.004) Loss 1.8000 (1.6758) Prec@1 53.125 (56.776) Prec@5 85.938 (82.744) Epoch: [12][1000/1345], lr: 0.01000 Time 0.734 (0.748) Data 0.000 (0.004) Loss 1.8961 (1.6767) Prec@1 54.688 (56.759) Prec@5 79.688 (82.752) Epoch: [12][1020/1345], lr: 0.01000 Time 0.729 (0.748) Data 0.000 (0.004) Loss 1.6935 (1.6774) Prec@1 67.188 (56.764) Prec@5 82.812 (82.727) Epoch: [12][1040/1345], lr: 0.01000 Time 0.769 (0.748) Data 0.000 (0.004) Loss 1.3664 (1.6770) Prec@1 65.625 (56.750) Prec@5 82.812 (82.713) Epoch: [12][1060/1345], lr: 0.01000 Time 0.722 (0.748) Data 0.001 (0.004) Loss 1.7106 (1.6767) Prec@1 54.688 (56.782) Prec@5 84.375 (82.701) Epoch: [12][1080/1345], lr: 0.01000 Time 0.858 (0.748) Data 0.000 (0.004) Loss 2.3351 (1.6768) Prec@1 40.625 (56.763) Prec@5 71.875 (82.704) Epoch: [12][1100/1345], lr: 0.01000 Time 0.752 (0.748) Data 0.000 (0.004) Loss 1.6734 (1.6771) Prec@1 57.812 (56.764) Prec@5 79.688 (82.679) Epoch: [12][1120/1345], lr: 0.01000 Time 0.714 (0.748) Data 0.000 (0.004) Loss 1.8322 (1.6765) Prec@1 60.938 (56.803) Prec@5 82.812 (82.677) Epoch: [12][1140/1345], lr: 0.01000 Time 0.728 (0.748) Data 0.001 (0.004) Loss 1.6398 (1.6765) Prec@1 53.125 (56.805) Prec@5 82.812 (82.680) Epoch: [12][1160/1345], lr: 0.01000 Time 0.729 (0.748) Data 0.000 (0.004) Loss 1.2786 (1.6758) Prec@1 62.500 (56.846) Prec@5 89.062 (82.709) Epoch: [12][1180/1345], lr: 0.01000 Time 0.765 (0.748) Data 0.001 (0.004) Loss 1.6666 (1.6774) Prec@1 60.938 (56.845) Prec@5 81.250 (82.687) Epoch: [12][1200/1345], lr: 0.01000 Time 0.716 (0.748) Data 0.001 (0.003) Loss 1.9489 (1.6781) Prec@1 48.438 (56.816) Prec@5 76.562 (82.668) Epoch: [12][1220/1345], lr: 0.01000 Time 0.726 (0.748) Data 0.001 (0.003) Loss 1.6221 (1.6798) Prec@1 57.812 (56.773) Prec@5 82.812 (82.641) Epoch: [12][1240/1345], lr: 0.01000 Time 0.832 (0.748) Data 0.000 (0.003) Loss 2.0654 (1.6783) Prec@1 43.750 (56.786) Prec@5 76.562 (82.675) Epoch: [12][1260/1345], lr: 0.01000 Time 0.774 (0.748) Data 0.001 (0.003) Loss 1.7259 (1.6778) Prec@1 57.812 (56.790) Prec@5 85.938 (82.708) Epoch: [12][1280/1345], lr: 0.01000 Time 0.716 (0.748) Data 0.000 (0.003) Loss 1.8756 (1.6769) Prec@1 53.125 (56.790) Prec@5 76.562 (82.715) Epoch: [12][1300/1345], lr: 0.01000 Time 0.727 (0.748) Data 0.001 (0.003) Loss 1.7715 (1.6768) Prec@1 53.125 (56.780) Prec@5 82.812 (82.715) Epoch: [12][1320/1345], lr: 0.01000 Time 0.780 (0.748) Data 0.001 (0.003) Loss 1.6352 (1.6777) Prec@1 57.812 (56.780) Prec@5 85.938 (82.713) Epoch: [12][1340/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.003) Loss 1.5305 (1.6791) Prec@1 60.938 (56.764) Prec@5 81.250 (82.679) No BN layer Freezing. Test: [0/181] Time 3.395 (3.3947) Loss 2.9171 (2.9171) Prec@1 40.625 (40.625) Prec@5 65.625 (65.625) Test: [20/181] Time 1.249 (0.6072) Loss 2.3522 (2.5577) Prec@1 43.750 (44.271) Prec@5 76.562 (72.619) Test: [40/181] Time 0.891 (0.5340) Loss 2.6308 (2.6263) Prec@1 43.750 (42.492) Prec@5 71.875 (71.418) Test: [60/181] Time 0.988 (0.5054) Loss 2.8174 (2.6547) Prec@1 46.875 (41.368) Prec@5 62.500 (70.927) Test: [80/181] Time 1.107 (0.4932) Loss 2.2811 (2.6916) Prec@1 45.312 (40.702) Prec@5 76.562 (70.332) Test: [100/181] Time 1.169 (0.4864) Loss 3.3384 (2.6899) Prec@1 29.688 (40.656) Prec@5 60.938 (70.560) Test: [120/181] Time 1.215 (0.4825) Loss 3.1892 (2.6841) Prec@1 39.062 (40.819) Prec@5 59.375 (70.338) Test: [140/181] Time 1.293 (0.4776) Loss 3.0497 (2.6806) Prec@1 32.812 (40.813) Prec@5 71.875 (70.423) Test: [160/181] Time 1.041 (0.4740) Loss 2.9410 (2.6641) Prec@1 34.375 (41.003) Prec@5 68.750 (70.749) Testing Results: Prec@1 41.250 Prec@5 70.807 Loss 2.65452 Time 0.4681 No BN layer Freezing. Epoch: [13][0/1345], lr: 0.01000 Time 4.494 (4.494) Data 3.751 (3.751) Loss 1.7346 (1.7346) Prec@1 50.000 (50.000) Prec@5 82.812 (82.812) Epoch: [13][20/1345], lr: 0.01000 Time 0.712 (0.928) Data 0.000 (0.179) Loss 1.5970 (1.5749) Prec@1 56.250 (58.780) Prec@5 81.250 (84.524) Epoch: [13][40/1345], lr: 0.01000 Time 0.729 (0.842) Data 0.001 (0.092) Loss 1.7320 (1.5931) Prec@1 50.000 (58.727) Prec@5 82.812 (83.994) Epoch: [13][60/1345], lr: 0.01000 Time 0.712 (0.808) Data 0.000 (0.062) Loss 1.4779 (1.5755) Prec@1 59.375 (59.247) Prec@5 85.938 (84.529) Epoch: [13][80/1345], lr: 0.01000 Time 0.716 (0.792) Data 0.000 (0.047) Loss 1.4597 (1.5989) Prec@1 56.250 (58.642) Prec@5 87.500 (84.066) Epoch: [13][100/1345], lr: 0.01000 Time 0.769 (0.784) Data 0.000 (0.038) Loss 1.6303 (1.6101) Prec@1 56.250 (58.648) Prec@5 82.812 (83.741) Epoch: [13][120/1345], lr: 0.01000 Time 0.860 (0.778) Data 0.001 (0.031) Loss 1.8373 (1.6018) Prec@1 51.562 (58.910) Prec@5 81.250 (83.884) Epoch: [13][140/1345], lr: 0.01000 Time 0.778 (0.774) Data 0.000 (0.027) Loss 1.4555 (1.6016) Prec@1 54.688 (58.666) Prec@5 84.375 (83.965) Epoch: [13][160/1345], lr: 0.01000 Time 0.748 (0.770) Data 0.000 (0.024) Loss 1.8969 (1.6026) Prec@1 54.688 (58.540) Prec@5 84.375 (84.064) Epoch: [13][180/1345], lr: 0.01000 Time 0.749 (0.766) Data 0.000 (0.021) Loss 1.6873 (1.6121) Prec@1 53.125 (58.244) Prec@5 78.125 (83.814) Epoch: [13][200/1345], lr: 0.01000 Time 0.715 (0.764) Data 0.000 (0.019) Loss 1.4905 (1.6124) Prec@1 57.812 (58.116) Prec@5 89.062 (83.815) Epoch: [13][220/1345], lr: 0.01000 Time 0.712 (0.762) Data 0.000 (0.017) Loss 1.5755 (1.6184) Prec@1 53.125 (58.018) Prec@5 82.812 (83.661) Epoch: [13][240/1345], lr: 0.01000 Time 0.713 (0.761) Data 0.000 (0.016) Loss 1.5529 (1.6140) Prec@1 56.250 (58.182) Prec@5 79.688 (83.681) Epoch: [13][260/1345], lr: 0.01000 Time 0.739 (0.760) Data 0.000 (0.015) Loss 1.5051 (1.6077) Prec@1 56.250 (58.363) Prec@5 84.375 (83.806) Epoch: [13][280/1345], lr: 0.01000 Time 0.714 (0.759) Data 0.000 (0.014) Loss 1.4661 (1.6053) Prec@1 60.938 (58.491) Prec@5 89.062 (83.852) Epoch: [13][300/1345], lr: 0.01000 Time 0.736 (0.759) Data 0.000 (0.013) Loss 2.3116 (1.6082) Prec@1 42.188 (58.389) Prec@5 64.062 (83.737) Epoch: [13][320/1345], lr: 0.01000 Time 0.769 (0.758) Data 0.000 (0.012) Loss 1.5013 (1.6081) Prec@1 67.188 (58.445) Prec@5 85.938 (83.728) Epoch: [13][340/1345], lr: 0.01000 Time 0.755 (0.757) Data 0.000 (0.011) Loss 1.5470 (1.6111) Prec@1 60.938 (58.385) Prec@5 82.812 (83.637) Epoch: [13][360/1345], lr: 0.01000 Time 0.714 (0.756) Data 0.000 (0.011) Loss 1.9918 (1.6127) Prec@1 50.000 (58.237) Prec@5 78.125 (83.605) Epoch: [13][380/1345], lr: 0.01000 Time 0.713 (0.756) Data 0.000 (0.010) Loss 1.7645 (1.6152) Prec@1 46.875 (58.173) Prec@5 81.250 (83.518) Epoch: [13][400/1345], lr: 0.01000 Time 0.773 (0.755) Data 0.000 (0.010) Loss 1.3460 (1.6172) Prec@1 71.875 (58.159) Prec@5 85.938 (83.502) Epoch: [13][420/1345], lr: 0.01000 Time 0.722 (0.755) Data 0.001 (0.009) Loss 1.4198 (1.6149) Prec@1 68.750 (58.228) Prec@5 85.938 (83.525) Epoch: [13][440/1345], lr: 0.01000 Time 0.741 (0.755) Data 0.000 (0.009) Loss 1.9323 (1.6208) Prec@1 57.812 (58.114) Prec@5 79.688 (83.447) Epoch: [13][460/1345], lr: 0.01000 Time 0.712 (0.754) Data 0.000 (0.009) Loss 2.0059 (1.6197) Prec@1 54.688 (58.162) Prec@5 71.875 (83.443) Epoch: [13][480/1345], lr: 0.01000 Time 0.760 (0.754) Data 0.000 (0.008) Loss 1.5275 (1.6216) Prec@1 60.938 (58.082) Prec@5 84.375 (83.456) Epoch: [13][500/1345], lr: 0.01000 Time 0.741 (0.754) Data 0.000 (0.008) Loss 1.9124 (1.6278) Prec@1 51.562 (57.912) Prec@5 70.312 (83.333) Epoch: [13][520/1345], lr: 0.01000 Time 0.714 (0.753) Data 0.000 (0.008) Loss 1.3899 (1.6281) Prec@1 65.625 (57.923) Prec@5 85.938 (83.292) Epoch: [13][540/1345], lr: 0.01000 Time 0.713 (0.753) Data 0.001 (0.007) Loss 2.1804 (1.6260) Prec@1 42.188 (57.992) Prec@5 73.438 (83.312) Epoch: [13][560/1345], lr: 0.01000 Time 0.723 (0.753) Data 0.001 (0.007) Loss 1.6227 (1.6289) Prec@1 59.375 (57.921) Prec@5 81.250 (83.253) Epoch: [13][580/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.007) Loss 1.5782 (1.6279) Prec@1 59.375 (57.936) Prec@5 84.375 (83.275) Epoch: [13][600/1345], lr: 0.01000 Time 0.746 (0.752) Data 0.001 (0.007) Loss 1.6668 (1.6297) Prec@1 54.688 (57.864) Prec@5 84.375 (83.236) Epoch: [13][620/1345], lr: 0.01000 Time 0.710 (0.752) Data 0.000 (0.007) Loss 1.6080 (1.6327) Prec@1 51.562 (57.750) Prec@5 87.500 (83.228) Epoch: [13][640/1345], lr: 0.01000 Time 0.716 (0.753) Data 0.000 (0.006) Loss 1.2211 (1.6329) Prec@1 64.062 (57.720) Prec@5 92.188 (83.237) Epoch: [13][660/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.006) Loss 1.6364 (1.6350) Prec@1 54.688 (57.656) Prec@5 84.375 (83.217) Epoch: [13][680/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.006) Loss 1.5047 (1.6355) Prec@1 60.938 (57.634) Prec@5 85.938 (83.216) Epoch: [13][700/1345], lr: 0.01000 Time 0.775 (0.752) Data 0.001 (0.006) Loss 1.4634 (1.6387) Prec@1 57.812 (57.541) Prec@5 87.500 (83.174) Epoch: [13][720/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.006) Loss 1.7685 (1.6390) Prec@1 56.250 (57.498) Prec@5 82.812 (83.183) Epoch: [13][740/1345], lr: 0.01000 Time 0.708 (0.751) Data 0.000 (0.006) Loss 1.6288 (1.6382) Prec@1 50.000 (57.509) Prec@5 82.812 (83.222) Epoch: [13][760/1345], lr: 0.01000 Time 0.814 (0.751) Data 0.000 (0.005) Loss 1.4076 (1.6397) Prec@1 65.625 (57.494) Prec@5 87.500 (83.178) Epoch: [13][780/1345], lr: 0.01000 Time 0.854 (0.751) Data 0.000 (0.005) Loss 1.9079 (1.6423) Prec@1 50.000 (57.456) Prec@5 81.250 (83.141) Epoch: [13][800/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.005) Loss 1.4589 (1.6422) Prec@1 62.500 (57.471) Prec@5 82.812 (83.136) Epoch: [13][820/1345], lr: 0.01000 Time 0.711 (0.750) Data 0.000 (0.005) Loss 1.9026 (1.6416) Prec@1 59.375 (57.500) Prec@5 81.250 (83.178) Epoch: [13][840/1345], lr: 0.01000 Time 0.709 (0.750) Data 0.001 (0.005) Loss 1.8066 (1.6417) Prec@1 59.375 (57.515) Prec@5 79.688 (83.182) Epoch: [13][860/1345], lr: 0.01000 Time 0.706 (0.749) Data 0.000 (0.005) Loss 1.6956 (1.6410) Prec@1 54.688 (57.508) Prec@5 79.688 (83.192) Epoch: [13][880/1345], lr: 0.01000 Time 0.794 (0.749) Data 0.000 (0.005) Loss 1.2696 (1.6395) Prec@1 64.062 (57.541) Prec@5 89.062 (83.208) Epoch: [13][900/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.005) Loss 1.5921 (1.6392) Prec@1 60.938 (57.552) Prec@5 82.812 (83.218) Epoch: [13][920/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.005) Loss 1.6470 (1.6387) Prec@1 50.000 (57.575) Prec@5 87.500 (83.240) Epoch: [13][940/1345], lr: 0.01000 Time 0.705 (0.748) Data 0.000 (0.004) Loss 1.7602 (1.6407) Prec@1 56.250 (57.524) Prec@5 82.812 (83.203) Epoch: [13][960/1345], lr: 0.01000 Time 0.714 (0.748) Data 0.000 (0.004) Loss 1.9600 (1.6418) Prec@1 50.000 (57.513) Prec@5 82.812 (83.208) Epoch: [13][980/1345], lr: 0.01000 Time 0.706 (0.747) Data 0.000 (0.004) Loss 1.3920 (1.6430) Prec@1 67.188 (57.507) Prec@5 89.062 (83.204) Epoch: [13][1000/1345], lr: 0.01000 Time 0.751 (0.747) Data 0.000 (0.004) Loss 1.5886 (1.6426) Prec@1 53.125 (57.508) Prec@5 87.500 (83.201) Epoch: [13][1020/1345], lr: 0.01000 Time 0.717 (0.746) Data 0.000 (0.004) Loss 1.6025 (1.6430) Prec@1 59.375 (57.505) Prec@5 81.250 (83.198) Epoch: [13][1040/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.004) Loss 1.7007 (1.6447) Prec@1 54.688 (57.473) Prec@5 84.375 (83.159) Epoch: [13][1060/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.004) Loss 1.6121 (1.6449) Prec@1 59.375 (57.458) Prec@5 84.375 (83.153) Epoch: [13][1080/1345], lr: 0.01000 Time 0.788 (0.745) Data 0.000 (0.004) Loss 1.6557 (1.6460) Prec@1 62.500 (57.467) Prec@5 84.375 (83.158) Epoch: [13][1100/1345], lr: 0.01000 Time 0.742 (0.745) Data 0.000 (0.004) Loss 1.6479 (1.6476) Prec@1 57.812 (57.412) Prec@5 81.250 (83.156) Epoch: [13][1120/1345], lr: 0.01000 Time 0.725 (0.745) Data 0.000 (0.004) Loss 1.7234 (1.6484) Prec@1 57.812 (57.407) Prec@5 82.812 (83.143) Epoch: [13][1140/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.004) Loss 1.5807 (1.6493) Prec@1 60.938 (57.369) Prec@5 87.500 (83.133) Epoch: [13][1160/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.001 (0.004) Loss 1.1647 (1.6488) Prec@1 71.875 (57.386) Prec@5 93.750 (83.142) Epoch: [13][1180/1345], lr: 0.01000 Time 0.721 (0.745) Data 0.001 (0.004) Loss 1.6963 (1.6504) Prec@1 57.812 (57.348) Prec@5 84.375 (83.096) Epoch: [13][1200/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.004) Loss 1.7581 (1.6512) Prec@1 53.125 (57.332) Prec@5 82.812 (83.087) Epoch: [13][1220/1345], lr: 0.01000 Time 0.835 (0.745) Data 0.000 (0.004) Loss 1.5091 (1.6518) Prec@1 57.812 (57.333) Prec@5 87.500 (83.071) Epoch: [13][1240/1345], lr: 0.01000 Time 0.851 (0.745) Data 0.000 (0.003) Loss 1.2631 (1.6517) Prec@1 70.312 (57.349) Prec@5 92.188 (83.082) Epoch: [13][1260/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.003) Loss 1.5599 (1.6520) Prec@1 62.500 (57.364) Prec@5 85.938 (83.084) Epoch: [13][1280/1345], lr: 0.01000 Time 0.717 (0.745) Data 0.001 (0.003) Loss 1.6305 (1.6527) Prec@1 64.062 (57.356) Prec@5 81.250 (83.075) Epoch: [13][1300/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.000 (0.003) Loss 1.5883 (1.6532) Prec@1 59.375 (57.351) Prec@5 82.812 (83.055) Epoch: [13][1320/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.001 (0.003) Loss 2.1401 (1.6536) Prec@1 45.312 (57.364) Prec@5 82.812 (83.036) Epoch: [13][1340/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.003) Loss 1.6696 (1.6528) Prec@1 48.438 (57.381) Prec@5 85.938 (83.053) No BN layer Freezing. Test: [0/181] Time 3.360 (3.3602) Loss 2.5555 (2.5555) Prec@1 46.875 (46.875) Prec@5 73.438 (73.438) Test: [20/181] Time 0.802 (0.5898) Loss 2.4433 (2.3888) Prec@1 48.438 (45.387) Prec@5 75.000 (75.000) Test: [40/181] Time 0.975 (0.5272) Loss 2.4070 (2.4735) Prec@1 43.750 (43.674) Prec@5 75.000 (73.895) Test: [60/181] Time 1.049 (0.5068) Loss 2.5121 (2.5026) Prec@1 43.750 (42.879) Prec@5 65.625 (72.848) Test: [80/181] Time 1.013 (0.4941) Loss 2.1869 (2.5397) Prec@1 46.875 (41.782) Prec@5 78.125 (72.415) Test: [100/181] Time 1.166 (0.4896) Loss 2.8274 (2.5364) Prec@1 39.062 (41.878) Prec@5 67.188 (72.447) Test: [120/181] Time 0.978 (0.4846) Loss 3.0133 (2.5426) Prec@1 37.500 (42.123) Prec@5 64.062 (72.521) Test: [140/181] Time 1.032 (0.4820) Loss 2.9067 (2.5438) Prec@1 32.812 (42.010) Prec@5 67.188 (72.651) Test: [160/181] Time 0.863 (0.4810) Loss 2.8258 (2.5301) Prec@1 35.938 (42.382) Prec@5 73.438 (72.836) Testing Results: Prec@1 42.726 Prec@5 72.873 Loss 2.52113 Time 0.4767 No BN layer Freezing. Epoch: [14][0/1345], lr: 0.01000 Time 4.490 (4.490) Data 3.757 (3.757) Loss 1.3095 (1.3095) Prec@1 57.812 (57.812) Prec@5 92.188 (92.188) Epoch: [14][20/1345], lr: 0.01000 Time 0.752 (0.927) Data 0.000 (0.179) Loss 1.5433 (1.5238) Prec@1 65.625 (60.565) Prec@5 84.375 (84.747) Epoch: [14][40/1345], lr: 0.01000 Time 0.708 (0.844) Data 0.000 (0.092) Loss 1.5091 (1.5604) Prec@1 64.062 (59.337) Prec@5 84.375 (84.261) Epoch: [14][60/1345], lr: 0.01000 Time 0.776 (0.811) Data 0.000 (0.062) Loss 1.2158 (1.5529) Prec@1 67.188 (59.349) Prec@5 90.625 (84.529) Epoch: [14][80/1345], lr: 0.01000 Time 0.710 (0.794) Data 0.000 (0.047) Loss 1.3919 (1.5862) Prec@1 65.625 (59.008) Prec@5 89.062 (84.298) Epoch: [14][100/1345], lr: 0.01000 Time 0.840 (0.784) Data 0.000 (0.038) Loss 1.7946 (1.5902) Prec@1 45.312 (58.632) Prec@5 79.688 (84.205) Epoch: [14][120/1345], lr: 0.01000 Time 0.845 (0.778) Data 0.000 (0.031) Loss 1.1805 (1.5925) Prec@1 70.312 (58.445) Prec@5 89.062 (84.207) Epoch: [14][140/1345], lr: 0.01000 Time 0.707 (0.773) Data 0.000 (0.027) Loss 1.7971 (1.5929) Prec@1 48.438 (58.156) Prec@5 81.250 (84.164) Epoch: [14][160/1345], lr: 0.01000 Time 0.759 (0.770) Data 0.000 (0.024) Loss 1.6250 (1.5880) Prec@1 51.562 (58.191) Prec@5 84.375 (84.268) Epoch: [14][180/1345], lr: 0.01000 Time 0.712 (0.767) Data 0.000 (0.021) Loss 1.8424 (1.5857) Prec@1 54.688 (58.348) Prec@5 82.812 (84.401) Epoch: [14][200/1345], lr: 0.01000 Time 0.783 (0.765) Data 0.001 (0.019) Loss 1.8235 (1.5929) Prec@1 54.688 (58.162) Prec@5 82.812 (84.282) Epoch: [14][220/1345], lr: 0.01000 Time 0.708 (0.762) Data 0.000 (0.017) Loss 1.3253 (1.5910) Prec@1 60.938 (58.145) Prec@5 95.312 (84.410) Epoch: [14][240/1345], lr: 0.01000 Time 0.892 (0.762) Data 0.000 (0.016) Loss 1.3910 (1.5923) Prec@1 65.625 (58.208) Prec@5 85.938 (84.284) Epoch: [14][260/1345], lr: 0.01000 Time 0.753 (0.762) Data 0.000 (0.015) Loss 1.7335 (1.5911) Prec@1 56.250 (58.226) Prec@5 78.125 (84.315) Epoch: [14][280/1345], lr: 0.01000 Time 0.710 (0.759) Data 0.000 (0.014) Loss 1.6029 (1.5870) Prec@1 53.125 (58.274) Prec@5 84.375 (84.347) Epoch: [14][300/1345], lr: 0.01000 Time 0.743 (0.759) Data 0.000 (0.013) Loss 1.6571 (1.5839) Prec@1 56.250 (58.290) Prec@5 84.375 (84.391) Epoch: [14][320/1345], lr: 0.01000 Time 0.715 (0.759) Data 0.000 (0.012) Loss 1.7664 (1.5838) Prec@1 57.812 (58.353) Prec@5 76.562 (84.346) Epoch: [14][340/1345], lr: 0.01000 Time 0.709 (0.758) Data 0.000 (0.011) Loss 1.2761 (1.5868) Prec@1 64.062 (58.381) Prec@5 87.500 (84.228) Epoch: [14][360/1345], lr: 0.01000 Time 0.730 (0.757) Data 0.000 (0.011) Loss 1.4734 (1.5886) Prec@1 64.062 (58.345) Prec@5 79.688 (84.185) Epoch: [14][380/1345], lr: 0.01000 Time 0.833 (0.757) Data 0.000 (0.010) Loss 1.4155 (1.5900) Prec@1 64.062 (58.313) Prec@5 87.500 (84.137) Epoch: [14][400/1345], lr: 0.01000 Time 0.839 (0.756) Data 0.000 (0.010) Loss 1.5105 (1.5926) Prec@1 64.062 (58.245) Prec@5 87.500 (84.079) Epoch: [14][420/1345], lr: 0.01000 Time 0.767 (0.756) Data 0.000 (0.009) Loss 1.7610 (1.5913) Prec@1 50.000 (58.317) Prec@5 82.812 (84.134) Epoch: [14][440/1345], lr: 0.01000 Time 0.739 (0.755) Data 0.000 (0.009) Loss 1.6186 (1.5916) Prec@1 53.125 (58.291) Prec@5 84.375 (84.095) Epoch: [14][460/1345], lr: 0.01000 Time 0.715 (0.754) Data 0.000 (0.009) Loss 1.3205 (1.5923) Prec@1 64.062 (58.277) Prec@5 89.062 (84.070) Epoch: [14][480/1345], lr: 0.01000 Time 0.709 (0.753) Data 0.000 (0.008) Loss 2.0390 (1.5948) Prec@1 57.812 (58.280) Prec@5 78.125 (84.037) Epoch: [14][500/1345], lr: 0.01000 Time 0.717 (0.753) Data 0.001 (0.008) Loss 1.6357 (1.5925) Prec@1 56.250 (58.302) Prec@5 85.938 (84.044) Epoch: [14][520/1345], lr: 0.01000 Time 0.711 (0.753) Data 0.000 (0.008) Loss 1.5725 (1.5939) Prec@1 60.938 (58.256) Prec@5 82.812 (84.030) Epoch: [14][540/1345], lr: 0.01000 Time 0.837 (0.753) Data 0.000 (0.007) Loss 1.7265 (1.5924) Prec@1 59.375 (58.353) Prec@5 84.375 (84.008) Epoch: [14][560/1345], lr: 0.01000 Time 0.805 (0.752) Data 0.000 (0.007) Loss 1.4825 (1.5960) Prec@1 60.938 (58.283) Prec@5 82.812 (83.954) Epoch: [14][580/1345], lr: 0.01000 Time 0.750 (0.752) Data 0.000 (0.007) Loss 1.7813 (1.5933) Prec@1 50.000 (58.426) Prec@5 81.250 (83.998) Epoch: [14][600/1345], lr: 0.01000 Time 0.752 (0.752) Data 0.000 (0.007) Loss 1.8225 (1.5946) Prec@1 59.375 (58.403) Prec@5 76.562 (83.951) Epoch: [14][620/1345], lr: 0.01000 Time 0.712 (0.752) Data 0.000 (0.006) Loss 1.4626 (1.5944) Prec@1 60.938 (58.406) Prec@5 85.938 (84.000) Epoch: [14][640/1345], lr: 0.01000 Time 0.712 (0.751) Data 0.000 (0.006) Loss 1.7526 (1.5970) Prec@1 60.938 (58.368) Prec@5 79.688 (83.956) Epoch: [14][660/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.006) Loss 1.4909 (1.6004) Prec@1 54.688 (58.290) Prec@5 87.500 (83.886) Epoch: [14][680/1345], lr: 0.01000 Time 0.709 (0.751) Data 0.000 (0.006) Loss 1.6120 (1.6011) Prec@1 56.250 (58.304) Prec@5 84.375 (83.863) Epoch: [14][700/1345], lr: 0.01000 Time 0.857 (0.751) Data 0.000 (0.006) Loss 1.6919 (1.6006) Prec@1 53.125 (58.310) Prec@5 78.125 (83.882) Epoch: [14][720/1345], lr: 0.01000 Time 0.856 (0.750) Data 0.000 (0.006) Loss 1.8067 (1.6026) Prec@1 56.250 (58.300) Prec@5 81.250 (83.855) Epoch: [14][740/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.000 (0.005) Loss 1.4367 (1.6039) Prec@1 59.375 (58.272) Prec@5 82.812 (83.865) Epoch: [14][760/1345], lr: 0.01000 Time 0.712 (0.750) Data 0.000 (0.005) Loss 1.7865 (1.6065) Prec@1 54.688 (58.252) Prec@5 79.688 (83.814) Epoch: [14][780/1345], lr: 0.01000 Time 0.743 (0.750) Data 0.000 (0.005) Loss 1.4582 (1.6086) Prec@1 59.375 (58.225) Prec@5 89.062 (83.779) Epoch: [14][800/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.005) Loss 1.7502 (1.6095) Prec@1 48.438 (58.173) Prec@5 79.688 (83.776) Epoch: [14][820/1345], lr: 0.01000 Time 0.726 (0.749) Data 0.001 (0.005) Loss 1.3468 (1.6090) Prec@1 56.250 (58.191) Prec@5 87.500 (83.777) Epoch: [14][840/1345], lr: 0.01000 Time 0.848 (0.750) Data 0.000 (0.005) Loss 1.3868 (1.6086) Prec@1 59.375 (58.225) Prec@5 87.500 (83.801) Epoch: [14][860/1345], lr: 0.01000 Time 0.750 (0.750) Data 0.000 (0.005) Loss 1.6217 (1.6092) Prec@1 50.000 (58.201) Prec@5 84.375 (83.789) Epoch: [14][880/1345], lr: 0.01000 Time 0.712 (0.750) Data 0.000 (0.005) Loss 1.2634 (1.6074) Prec@1 65.625 (58.229) Prec@5 90.625 (83.836) Epoch: [14][900/1345], lr: 0.01000 Time 0.752 (0.750) Data 0.000 (0.005) Loss 2.0237 (1.6104) Prec@1 56.250 (58.180) Prec@5 78.125 (83.804) Epoch: [14][920/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.004) Loss 1.4557 (1.6112) Prec@1 59.375 (58.147) Prec@5 84.375 (83.795) Epoch: [14][940/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.004) Loss 1.8414 (1.6122) Prec@1 56.250 (58.126) Prec@5 76.562 (83.752) Epoch: [14][960/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.004) Loss 1.8475 (1.6144) Prec@1 56.250 (58.128) Prec@5 78.125 (83.741) Epoch: [14][980/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.004) Loss 1.8513 (1.6150) Prec@1 54.688 (58.112) Prec@5 78.125 (83.744) Epoch: [14][1000/1345], lr: 0.01000 Time 0.767 (0.749) Data 0.001 (0.004) Loss 1.4982 (1.6144) Prec@1 57.812 (58.122) Prec@5 85.938 (83.765) Epoch: [14][1020/1345], lr: 0.01000 Time 0.713 (0.749) Data 0.000 (0.004) Loss 2.1576 (1.6127) Prec@1 53.125 (58.183) Prec@5 71.875 (83.769) Epoch: [14][1040/1345], lr: 0.01000 Time 0.714 (0.749) Data 0.001 (0.004) Loss 1.1195 (1.6129) Prec@1 67.188 (58.185) Prec@5 95.312 (83.776) Epoch: [14][1060/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.004) Loss 1.8056 (1.6140) Prec@1 51.562 (58.207) Prec@5 82.812 (83.742) Epoch: [14][1080/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.004) Loss 1.4255 (1.6146) Prec@1 65.625 (58.183) Prec@5 89.062 (83.755) Epoch: [14][1100/1345], lr: 0.01000 Time 0.855 (0.749) Data 0.000 (0.004) Loss 1.4759 (1.6143) Prec@1 64.062 (58.179) Prec@5 85.938 (83.755) Epoch: [14][1120/1345], lr: 0.01000 Time 0.710 (0.749) Data 0.000 (0.004) Loss 1.5335 (1.6142) Prec@1 59.375 (58.185) Prec@5 84.375 (83.737) Epoch: [14][1140/1345], lr: 0.01000 Time 0.715 (0.749) Data 0.000 (0.004) Loss 1.8661 (1.6163) Prec@1 57.812 (58.148) Prec@5 85.938 (83.692) Epoch: [14][1160/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.004) Loss 1.9103 (1.6171) Prec@1 54.688 (58.162) Prec@5 75.000 (83.671) Epoch: [14][1180/1345], lr: 0.01000 Time 0.829 (0.748) Data 0.000 (0.004) Loss 1.5812 (1.6176) Prec@1 57.812 (58.151) Prec@5 82.812 (83.658) Epoch: [14][1200/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.004) Loss 1.6776 (1.6184) Prec@1 59.375 (58.151) Prec@5 82.812 (83.637) Epoch: [14][1220/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.003) Loss 1.8404 (1.6181) Prec@1 56.250 (58.167) Prec@5 76.562 (83.639) Epoch: [14][1240/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.003) Loss 1.7109 (1.6182) Prec@1 48.438 (58.152) Prec@5 84.375 (83.649) Epoch: [14][1260/1345], lr: 0.01000 Time 0.716 (0.748) Data 0.000 (0.003) Loss 1.2687 (1.6185) Prec@1 71.875 (58.140) Prec@5 85.938 (83.645) Epoch: [14][1280/1345], lr: 0.01000 Time 0.777 (0.748) Data 0.000 (0.003) Loss 1.5726 (1.6174) Prec@1 56.250 (58.156) Prec@5 82.812 (83.657) Epoch: [14][1300/1345], lr: 0.01000 Time 0.717 (0.748) Data 0.000 (0.003) Loss 1.4144 (1.6180) Prec@1 65.625 (58.142) Prec@5 85.938 (83.658) Epoch: [14][1320/1345], lr: 0.01000 Time 0.709 (0.748) Data 0.000 (0.003) Loss 1.5132 (1.6188) Prec@1 57.812 (58.138) Prec@5 85.938 (83.661) Epoch: [14][1340/1345], lr: 0.01000 Time 0.818 (0.748) Data 0.001 (0.003) Loss 1.2584 (1.6201) Prec@1 68.750 (58.104) Prec@5 90.625 (83.661) No BN layer Freezing. Test: [0/181] Time 2.891 (2.8909) Loss 2.5526 (2.5526) Prec@1 48.438 (48.438) Prec@5 76.562 (76.562) Test: [20/181] Time 0.923 (0.5792) Loss 2.3209 (2.4198) Prec@1 46.875 (44.048) Prec@5 70.312 (73.586) Test: [40/181] Time 0.733 (0.5201) Loss 2.3194 (2.5289) Prec@1 48.438 (41.806) Prec@5 71.875 (71.799) Test: [60/181] Time 0.833 (0.5001) Loss 2.7550 (2.5655) Prec@1 37.500 (41.445) Prec@5 65.625 (71.644) Test: [80/181] Time 0.649 (0.4877) Loss 2.2128 (2.5846) Prec@1 46.875 (40.895) Prec@5 75.000 (71.489) Test: [100/181] Time 0.900 (0.4817) Loss 2.8516 (2.5869) Prec@1 34.375 (41.027) Prec@5 65.625 (71.612) Test: [120/181] Time 0.969 (0.4772) Loss 2.8199 (2.5778) Prec@1 40.625 (41.464) Prec@5 65.625 (71.681) Test: [140/181] Time 1.165 (0.4760) Loss 3.1564 (2.5826) Prec@1 29.688 (41.312) Prec@5 75.000 (71.953) Test: [160/181] Time 1.017 (0.4751) Loss 2.8194 (2.5709) Prec@1 40.625 (41.722) Prec@5 78.125 (72.273) Testing Results: Prec@1 42.075 Prec@5 72.335 Loss 2.56577 Time 0.4694 No BN layer Freezing. Epoch: [15][0/1345], lr: 0.01000 Time 4.034 (4.034) Data 3.293 (3.293) Loss 1.8485 (1.8485) Prec@1 54.688 (54.688) Prec@5 81.250 (81.250) Epoch: [15][20/1345], lr: 0.01000 Time 0.707 (0.900) Data 0.000 (0.157) Loss 1.5109 (1.5611) Prec@1 62.500 (60.640) Prec@5 82.812 (83.482) Epoch: [15][40/1345], lr: 0.01000 Time 0.720 (0.825) Data 0.000 (0.081) Loss 1.5658 (1.5898) Prec@1 62.500 (60.099) Prec@5 89.062 (83.803) Epoch: [15][60/1345], lr: 0.01000 Time 0.708 (0.797) Data 0.000 (0.054) Loss 1.7319 (1.5578) Prec@1 57.812 (60.886) Prec@5 75.000 (84.144) Epoch: [15][80/1345], lr: 0.01000 Time 0.757 (0.782) Data 0.000 (0.041) Loss 1.2496 (1.5558) Prec@1 62.500 (60.590) Prec@5 92.188 (84.298) Epoch: [15][100/1345], lr: 0.01000 Time 0.709 (0.771) Data 0.000 (0.033) Loss 1.6277 (1.5581) Prec@1 68.750 (60.659) Prec@5 81.250 (84.143) Epoch: [15][120/1345], lr: 0.01000 Time 0.857 (0.764) Data 0.000 (0.028) Loss 1.5019 (1.5563) Prec@1 57.812 (60.318) Prec@5 84.375 (84.452) Epoch: [15][140/1345], lr: 0.01000 Time 0.766 (0.762) Data 0.000 (0.024) Loss 1.5523 (1.5624) Prec@1 54.688 (59.940) Prec@5 84.375 (84.375) Epoch: [15][160/1345], lr: 0.01000 Time 0.751 (0.758) Data 0.000 (0.021) Loss 1.9998 (1.5669) Prec@1 50.000 (59.618) Prec@5 81.250 (84.462) Epoch: [15][180/1345], lr: 0.01000 Time 0.742 (0.757) Data 0.000 (0.019) Loss 1.4927 (1.5744) Prec@1 57.812 (59.608) Prec@5 82.812 (84.418) Epoch: [15][200/1345], lr: 0.01000 Time 0.710 (0.754) Data 0.000 (0.017) Loss 1.3385 (1.5694) Prec@1 56.250 (59.554) Prec@5 90.625 (84.569) Epoch: [15][220/1345], lr: 0.01000 Time 0.707 (0.752) Data 0.000 (0.015) Loss 1.3580 (1.5655) Prec@1 65.625 (59.630) Prec@5 85.938 (84.580) Epoch: [15][240/1345], lr: 0.01000 Time 0.707 (0.750) Data 0.000 (0.014) Loss 1.5530 (1.5620) Prec@1 65.625 (59.829) Prec@5 84.375 (84.608) Epoch: [15][260/1345], lr: 0.01000 Time 0.707 (0.748) Data 0.000 (0.013) Loss 1.4509 (1.5664) Prec@1 67.188 (59.686) Prec@5 85.938 (84.537) Epoch: [15][280/1345], lr: 0.01000 Time 0.855 (0.748) Data 0.000 (0.012) Loss 1.6993 (1.5694) Prec@1 51.562 (59.597) Prec@5 85.938 (84.453) Epoch: [15][300/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.011) Loss 1.5264 (1.5743) Prec@1 59.375 (59.463) Prec@5 85.938 (84.354) Epoch: [15][320/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.011) Loss 1.6036 (1.5750) Prec@1 60.938 (59.283) Prec@5 85.938 (84.375) Epoch: [15][340/1345], lr: 0.01000 Time 0.804 (0.744) Data 0.000 (0.010) Loss 1.8009 (1.5762) Prec@1 48.438 (59.219) Prec@5 76.562 (84.306) Epoch: [15][360/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.010) Loss 1.3579 (1.5744) Prec@1 62.500 (59.349) Prec@5 82.812 (84.267) Epoch: [15][380/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.001 (0.009) Loss 1.2676 (1.5716) Prec@1 65.625 (59.391) Prec@5 89.062 (84.281) Epoch: [15][400/1345], lr: 0.01000 Time 0.757 (0.742) Data 0.000 (0.009) Loss 1.2413 (1.5681) Prec@1 65.625 (59.426) Prec@5 90.625 (84.305) Epoch: [15][420/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.008) Loss 1.5338 (1.5694) Prec@1 57.812 (59.416) Prec@5 85.938 (84.316) Epoch: [15][440/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.008) Loss 1.5404 (1.5697) Prec@1 60.938 (59.403) Prec@5 82.812 (84.336) Epoch: [15][460/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.000 (0.008) Loss 1.5378 (1.5700) Prec@1 62.500 (59.324) Prec@5 84.375 (84.344) Epoch: [15][480/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.007) Loss 1.7192 (1.5686) Prec@1 57.812 (59.369) Prec@5 78.125 (84.339) Epoch: [15][500/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.007) Loss 1.4428 (1.5708) Prec@1 60.938 (59.316) Prec@5 87.500 (84.350) Epoch: [15][520/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.007) Loss 1.9987 (1.5697) Prec@1 53.125 (59.321) Prec@5 75.000 (84.393) Epoch: [15][540/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.007) Loss 1.3282 (1.5709) Prec@1 64.062 (59.254) Prec@5 92.188 (84.375) Epoch: [15][560/1345], lr: 0.01000 Time 0.780 (0.741) Data 0.000 (0.006) Loss 1.6148 (1.5722) Prec@1 56.250 (59.211) Prec@5 89.062 (84.361) Epoch: [15][580/1345], lr: 0.01000 Time 0.782 (0.740) Data 0.000 (0.006) Loss 1.7375 (1.5733) Prec@1 56.250 (59.219) Prec@5 76.562 (84.335) Epoch: [15][600/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.006) Loss 1.7468 (1.5730) Prec@1 51.562 (59.240) Prec@5 84.375 (84.328) Epoch: [15][620/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.006) Loss 1.7715 (1.5753) Prec@1 54.688 (59.211) Prec@5 82.812 (84.282) Epoch: [15][640/1345], lr: 0.01000 Time 0.710 (0.740) Data 0.000 (0.006) Loss 1.6427 (1.5766) Prec@1 56.250 (59.170) Prec@5 78.125 (84.234) Epoch: [15][660/1345], lr: 0.01000 Time 0.748 (0.739) Data 0.000 (0.005) Loss 1.1492 (1.5748) Prec@1 71.875 (59.224) Prec@5 90.625 (84.276) Epoch: [15][680/1345], lr: 0.01000 Time 0.749 (0.739) Data 0.000 (0.005) Loss 1.3453 (1.5753) Prec@1 64.062 (59.230) Prec@5 82.812 (84.286) Epoch: [15][700/1345], lr: 0.01000 Time 0.797 (0.739) Data 0.000 (0.005) Loss 1.4406 (1.5745) Prec@1 51.562 (59.277) Prec@5 84.375 (84.288) Epoch: [15][720/1345], lr: 0.01000 Time 0.709 (0.739) Data 0.001 (0.005) Loss 1.1547 (1.5735) Prec@1 68.750 (59.308) Prec@5 92.188 (84.297) Epoch: [15][740/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.005) Loss 1.4225 (1.5728) Prec@1 60.938 (59.329) Prec@5 89.062 (84.284) Epoch: [15][760/1345], lr: 0.01000 Time 0.772 (0.738) Data 0.001 (0.005) Loss 1.4944 (1.5743) Prec@1 56.250 (59.276) Prec@5 90.625 (84.272) Epoch: [15][780/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.005) Loss 2.1538 (1.5747) Prec@1 51.562 (59.249) Prec@5 71.875 (84.291) Epoch: [15][800/1345], lr: 0.01000 Time 0.717 (0.737) Data 0.001 (0.005) Loss 1.6531 (1.5776) Prec@1 53.125 (59.158) Prec@5 84.375 (84.229) Epoch: [15][820/1345], lr: 0.01000 Time 0.709 (0.737) Data 0.000 (0.004) Loss 1.3375 (1.5782) Prec@1 67.188 (59.116) Prec@5 89.062 (84.213) Epoch: [15][840/1345], lr: 0.01000 Time 0.847 (0.738) Data 0.000 (0.004) Loss 1.4381 (1.5779) Prec@1 57.812 (59.100) Prec@5 87.500 (84.221) Epoch: [15][860/1345], lr: 0.01000 Time 0.714 (0.738) Data 0.001 (0.004) Loss 1.2220 (1.5764) Prec@1 62.500 (59.077) Prec@5 92.188 (84.241) Epoch: [15][880/1345], lr: 0.01000 Time 0.769 (0.738) Data 0.000 (0.004) Loss 1.7880 (1.5770) Prec@1 57.812 (59.072) Prec@5 84.375 (84.228) Epoch: [15][900/1345], lr: 0.01000 Time 0.743 (0.738) Data 0.000 (0.004) Loss 1.9060 (1.5782) Prec@1 51.562 (59.047) Prec@5 79.688 (84.202) Epoch: [15][920/1345], lr: 0.01000 Time 0.706 (0.737) Data 0.000 (0.004) Loss 1.4652 (1.5795) Prec@1 62.500 (59.003) Prec@5 85.938 (84.199) Epoch: [15][940/1345], lr: 0.01000 Time 0.739 (0.738) Data 0.000 (0.004) Loss 2.1316 (1.5815) Prec@1 50.000 (58.962) Prec@5 75.000 (84.169) Epoch: [15][960/1345], lr: 0.01000 Time 0.771 (0.738) Data 0.000 (0.004) Loss 1.7482 (1.5812) Prec@1 60.938 (58.957) Prec@5 79.688 (84.185) Epoch: [15][980/1345], lr: 0.01000 Time 0.793 (0.738) Data 0.000 (0.004) Loss 1.6439 (1.5822) Prec@1 56.250 (58.940) Prec@5 82.812 (84.163) Epoch: [15][1000/1345], lr: 0.01000 Time 0.831 (0.738) Data 0.000 (0.004) Loss 1.5766 (1.5823) Prec@1 56.250 (58.968) Prec@5 84.375 (84.177) Epoch: [15][1020/1345], lr: 0.01000 Time 0.718 (0.738) Data 0.000 (0.004) Loss 1.8652 (1.5833) Prec@1 53.125 (58.922) Prec@5 78.125 (84.139) Epoch: [15][1040/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.004) Loss 1.4100 (1.5860) Prec@1 68.750 (58.881) Prec@5 87.500 (84.109) Epoch: [15][1060/1345], lr: 0.01000 Time 0.713 (0.738) Data 0.001 (0.004) Loss 1.8047 (1.5866) Prec@1 56.250 (58.833) Prec@5 78.125 (84.101) Epoch: [15][1080/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.003) Loss 1.3521 (1.5862) Prec@1 62.500 (58.843) Prec@5 82.812 (84.108) Epoch: [15][1100/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 2.0167 (1.5881) Prec@1 53.125 (58.809) Prec@5 76.562 (84.080) Epoch: [15][1120/1345], lr: 0.01000 Time 0.710 (0.738) Data 0.000 (0.003) Loss 1.7160 (1.5901) Prec@1 56.250 (58.732) Prec@5 84.375 (84.066) Epoch: [15][1140/1345], lr: 0.01000 Time 0.767 (0.737) Data 0.000 (0.003) Loss 1.1249 (1.5901) Prec@1 73.438 (58.731) Prec@5 90.625 (84.082) Epoch: [15][1160/1345], lr: 0.01000 Time 0.709 (0.737) Data 0.000 (0.003) Loss 1.6703 (1.5923) Prec@1 60.938 (58.707) Prec@5 81.250 (84.064) Epoch: [15][1180/1345], lr: 0.01000 Time 0.752 (0.737) Data 0.000 (0.003) Loss 1.5681 (1.5934) Prec@1 59.375 (58.679) Prec@5 81.250 (84.054) Epoch: [15][1200/1345], lr: 0.01000 Time 0.735 (0.737) Data 0.000 (0.003) Loss 1.6747 (1.5950) Prec@1 48.438 (58.633) Prec@5 85.938 (84.015) Epoch: [15][1220/1345], lr: 0.01000 Time 0.717 (0.737) Data 0.000 (0.003) Loss 1.3667 (1.5949) Prec@1 62.500 (58.633) Prec@5 85.938 (84.027) Epoch: [15][1240/1345], lr: 0.01000 Time 0.851 (0.737) Data 0.000 (0.003) Loss 1.7035 (1.5957) Prec@1 59.375 (58.604) Prec@5 81.250 (84.006) Epoch: [15][1260/1345], lr: 0.01000 Time 0.714 (0.737) Data 0.000 (0.003) Loss 1.2934 (1.5956) Prec@1 64.062 (58.606) Prec@5 89.062 (83.987) Epoch: [15][1280/1345], lr: 0.01000 Time 0.717 (0.737) Data 0.000 (0.003) Loss 2.1287 (1.5967) Prec@1 48.438 (58.589) Prec@5 76.562 (83.981) Epoch: [15][1300/1345], lr: 0.01000 Time 0.718 (0.737) Data 0.000 (0.003) Loss 1.6705 (1.5970) Prec@1 62.500 (58.584) Prec@5 85.938 (83.971) Epoch: [15][1320/1345], lr: 0.01000 Time 0.710 (0.737) Data 0.000 (0.003) Loss 1.9496 (1.5967) Prec@1 53.125 (58.585) Prec@5 78.125 (83.972) Epoch: [15][1340/1345], lr: 0.01000 Time 0.770 (0.737) Data 0.000 (0.003) Loss 1.4105 (1.5970) Prec@1 60.938 (58.558) Prec@5 85.938 (83.964) No BN layer Freezing. Test: [0/181] Time 3.374 (3.3742) Loss 2.6749 (2.6749) Prec@1 50.000 (50.000) Prec@5 71.875 (71.875) Test: [20/181] Time 1.107 (0.6005) Loss 2.4483 (2.5753) Prec@1 45.312 (42.857) Prec@5 75.000 (72.470) Test: [40/181] Time 0.987 (0.5282) Loss 2.6386 (2.6436) Prec@1 42.188 (40.892) Prec@5 70.312 (71.151) Test: [60/181] Time 0.240 (0.4967) Loss 3.1234 (2.6966) Prec@1 42.188 (39.985) Prec@5 60.938 (70.108) Test: [80/181] Time 0.236 (0.4896) Loss 2.2232 (2.7124) Prec@1 50.000 (39.622) Prec@5 81.250 (70.042) Test: [100/181] Time 0.237 (0.4848) Loss 3.0178 (2.7075) Prec@1 35.938 (39.573) Prec@5 70.312 (70.374) Test: [120/181] Time 0.380 (0.4798) Loss 2.9668 (2.7075) Prec@1 46.875 (39.824) Prec@5 62.500 (70.300) Test: [140/181] Time 0.332 (0.4760) Loss 3.4155 (2.7058) Prec@1 23.438 (40.016) Prec@5 64.062 (70.390) Test: [160/181] Time 0.323 (0.4735) Loss 3.0545 (2.6914) Prec@1 35.938 (40.276) Prec@5 70.312 (70.613) Testing Results: Prec@1 40.547 Prec@5 70.703 Loss 2.67985 Time 0.4711 No BN layer Freezing. Epoch: [16][0/1345], lr: 0.01000 Time 4.092 (4.092) Data 3.350 (3.350) Loss 1.7120 (1.7120) Prec@1 53.125 (53.125) Prec@5 84.375 (84.375) Epoch: [16][20/1345], lr: 0.01000 Time 0.850 (0.913) Data 0.000 (0.160) Loss 1.3422 (1.6404) Prec@1 65.625 (58.185) Prec@5 90.625 (82.589) Epoch: [16][40/1345], lr: 0.01000 Time 0.709 (0.827) Data 0.000 (0.082) Loss 1.6839 (1.5871) Prec@1 56.250 (58.918) Prec@5 78.125 (83.765) Epoch: [16][60/1345], lr: 0.01000 Time 0.776 (0.799) Data 0.000 (0.055) Loss 1.2917 (1.5573) Prec@1 59.375 (59.477) Prec@5 92.188 (84.426) Epoch: [16][80/1345], lr: 0.01000 Time 0.771 (0.784) Data 0.000 (0.042) Loss 1.7932 (1.5442) Prec@1 51.562 (59.568) Prec@5 85.938 (84.780) Epoch: [16][100/1345], lr: 0.01000 Time 0.708 (0.775) Data 0.000 (0.034) Loss 1.4877 (1.5405) Prec@1 56.250 (59.684) Prec@5 87.500 (84.886) Epoch: [16][120/1345], lr: 0.01000 Time 0.708 (0.769) Data 0.000 (0.028) Loss 1.8549 (1.5374) Prec@1 56.250 (59.969) Prec@5 73.438 (84.995) Epoch: [16][140/1345], lr: 0.01000 Time 0.751 (0.766) Data 0.001 (0.024) Loss 1.4384 (1.5264) Prec@1 65.625 (60.217) Prec@5 85.938 (85.195) Epoch: [16][160/1345], lr: 0.01000 Time 0.748 (0.763) Data 0.000 (0.021) Loss 1.6112 (1.5284) Prec@1 54.688 (60.336) Prec@5 82.812 (85.083) Epoch: [16][180/1345], lr: 0.01000 Time 0.716 (0.760) Data 0.000 (0.019) Loss 1.4662 (1.5251) Prec@1 59.375 (60.402) Prec@5 89.062 (85.195) Epoch: [16][200/1345], lr: 0.01000 Time 0.708 (0.759) Data 0.000 (0.017) Loss 1.3187 (1.5294) Prec@1 59.375 (60.137) Prec@5 87.500 (85.082) Epoch: [16][220/1345], lr: 0.01000 Time 0.714 (0.757) Data 0.001 (0.016) Loss 1.4956 (1.5338) Prec@1 57.812 (60.075) Prec@5 87.500 (84.962) Epoch: [16][240/1345], lr: 0.01000 Time 0.757 (0.756) Data 0.000 (0.014) Loss 1.3411 (1.5310) Prec@1 68.750 (60.198) Prec@5 89.062 (84.991) Epoch: [16][260/1345], lr: 0.01000 Time 0.708 (0.755) Data 0.000 (0.013) Loss 2.0513 (1.5323) Prec@1 46.875 (60.105) Prec@5 71.875 (84.926) Epoch: [16][280/1345], lr: 0.01000 Time 0.712 (0.754) Data 0.000 (0.012) Loss 1.6420 (1.5344) Prec@1 62.500 (60.142) Prec@5 79.688 (84.859) Epoch: [16][300/1345], lr: 0.01000 Time 0.716 (0.755) Data 0.000 (0.012) Loss 1.7601 (1.5361) Prec@1 51.562 (60.143) Prec@5 82.812 (84.744) Epoch: [16][320/1345], lr: 0.01000 Time 0.714 (0.754) Data 0.001 (0.011) Loss 1.1524 (1.5406) Prec@1 75.000 (60.042) Prec@5 93.750 (84.711) Epoch: [16][340/1345], lr: 0.01000 Time 0.780 (0.753) Data 0.000 (0.010) Loss 1.7604 (1.5439) Prec@1 53.125 (59.998) Prec@5 82.812 (84.682) Epoch: [16][360/1345], lr: 0.01000 Time 0.776 (0.753) Data 0.001 (0.010) Loss 1.5597 (1.5460) Prec@1 62.500 (59.998) Prec@5 85.938 (84.596) Epoch: [16][380/1345], lr: 0.01000 Time 0.718 (0.753) Data 0.001 (0.009) Loss 1.7257 (1.5479) Prec@1 53.125 (59.961) Prec@5 79.688 (84.568) Epoch: [16][400/1345], lr: 0.01000 Time 0.854 (0.753) Data 0.000 (0.009) Loss 2.1398 (1.5514) Prec@1 53.125 (59.963) Prec@5 70.312 (84.543) Epoch: [16][420/1345], lr: 0.01000 Time 0.844 (0.753) Data 0.000 (0.008) Loss 1.3566 (1.5518) Prec@1 59.375 (59.876) Prec@5 82.812 (84.535) Epoch: [16][440/1345], lr: 0.01000 Time 0.715 (0.753) Data 0.000 (0.008) Loss 1.2798 (1.5504) Prec@1 62.500 (59.867) Prec@5 89.062 (84.584) Epoch: [16][460/1345], lr: 0.01000 Time 0.752 (0.752) Data 0.000 (0.008) Loss 1.6824 (1.5517) Prec@1 57.812 (59.846) Prec@5 81.250 (84.548) Epoch: [16][480/1345], lr: 0.01000 Time 0.756 (0.753) Data 0.000 (0.007) Loss 1.3209 (1.5510) Prec@1 67.188 (59.853) Prec@5 90.625 (84.557) Epoch: [16][500/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.007) Loss 2.0171 (1.5512) Prec@1 50.000 (59.790) Prec@5 75.000 (84.590) Epoch: [16][520/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.007) Loss 1.5506 (1.5511) Prec@1 48.438 (59.714) Prec@5 85.938 (84.597) Epoch: [16][540/1345], lr: 0.01000 Time 0.846 (0.752) Data 0.000 (0.007) Loss 1.5840 (1.5527) Prec@1 56.250 (59.615) Prec@5 85.938 (84.609) Epoch: [16][560/1345], lr: 0.01000 Time 0.749 (0.752) Data 0.000 (0.006) Loss 1.5120 (1.5505) Prec@1 64.062 (59.673) Prec@5 85.938 (84.665) Epoch: [16][580/1345], lr: 0.01000 Time 0.709 (0.751) Data 0.000 (0.006) Loss 1.3208 (1.5540) Prec@1 64.062 (59.622) Prec@5 90.625 (84.579) Epoch: [16][600/1345], lr: 0.01000 Time 0.705 (0.750) Data 0.000 (0.006) Loss 1.2952 (1.5540) Prec@1 65.625 (59.586) Prec@5 90.625 (84.599) Epoch: [16][620/1345], lr: 0.01000 Time 0.712 (0.749) Data 0.000 (0.006) Loss 1.3839 (1.5541) Prec@1 62.500 (59.528) Prec@5 85.938 (84.599) Epoch: [16][640/1345], lr: 0.01000 Time 0.707 (0.748) Data 0.000 (0.006) Loss 1.6521 (1.5559) Prec@1 53.125 (59.477) Prec@5 82.812 (84.585) Epoch: [16][660/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.005) Loss 1.5239 (1.5567) Prec@1 57.812 (59.470) Prec@5 87.500 (84.566) Epoch: [16][680/1345], lr: 0.01000 Time 0.829 (0.748) Data 0.000 (0.005) Loss 1.7233 (1.5558) Prec@1 56.250 (59.464) Prec@5 81.250 (84.602) Epoch: [16][700/1345], lr: 0.01000 Time 0.838 (0.748) Data 0.000 (0.005) Loss 1.2726 (1.5547) Prec@1 71.875 (59.513) Prec@5 89.062 (84.629) Epoch: [16][720/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.005) Loss 1.5492 (1.5547) Prec@1 59.375 (59.518) Prec@5 85.938 (84.607) Epoch: [16][740/1345], lr: 0.01000 Time 0.752 (0.748) Data 0.000 (0.005) Loss 1.6656 (1.5553) Prec@1 56.250 (59.512) Prec@5 85.938 (84.599) Epoch: [16][760/1345], lr: 0.01000 Time 0.766 (0.748) Data 0.000 (0.005) Loss 1.4779 (1.5561) Prec@1 60.938 (59.474) Prec@5 81.250 (84.603) Epoch: [16][780/1345], lr: 0.01000 Time 0.726 (0.748) Data 0.000 (0.005) Loss 1.7394 (1.5565) Prec@1 56.250 (59.449) Prec@5 78.125 (84.599) Epoch: [16][800/1345], lr: 0.01000 Time 0.708 (0.748) Data 0.000 (0.005) Loss 1.7292 (1.5580) Prec@1 48.438 (59.402) Prec@5 87.500 (84.586) Epoch: [16][820/1345], lr: 0.01000 Time 0.717 (0.748) Data 0.001 (0.005) Loss 1.2061 (1.5580) Prec@1 65.625 (59.337) Prec@5 92.188 (84.624) Epoch: [16][840/1345], lr: 0.01000 Time 0.743 (0.748) Data 0.001 (0.004) Loss 1.6609 (1.5571) Prec@1 56.250 (59.310) Prec@5 82.812 (84.650) Epoch: [16][860/1345], lr: 0.01000 Time 0.722 (0.748) Data 0.000 (0.004) Loss 1.3595 (1.5583) Prec@1 60.938 (59.288) Prec@5 87.500 (84.644) Epoch: [16][880/1345], lr: 0.01000 Time 0.742 (0.747) Data 0.000 (0.004) Loss 1.6381 (1.5591) Prec@1 56.250 (59.295) Prec@5 87.500 (84.636) Epoch: [16][900/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.000 (0.004) Loss 1.8718 (1.5615) Prec@1 53.125 (59.229) Prec@5 81.250 (84.594) Epoch: [16][920/1345], lr: 0.01000 Time 0.708 (0.747) Data 0.000 (0.004) Loss 1.6358 (1.5620) Prec@1 50.000 (59.251) Prec@5 84.375 (84.589) Epoch: [16][940/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.004) Loss 1.7030 (1.5625) Prec@1 53.125 (59.250) Prec@5 82.812 (84.586) Epoch: [16][960/1345], lr: 0.01000 Time 0.849 (0.747) Data 0.000 (0.004) Loss 1.2951 (1.5615) Prec@1 71.875 (59.281) Prec@5 87.500 (84.621) Epoch: [16][980/1345], lr: 0.01000 Time 0.734 (0.747) Data 0.000 (0.004) Loss 1.2501 (1.5611) Prec@1 67.188 (59.287) Prec@5 92.188 (84.630) Epoch: [16][1000/1345], lr: 0.01000 Time 0.711 (0.747) Data 0.000 (0.004) Loss 1.5252 (1.5601) Prec@1 56.250 (59.292) Prec@5 89.062 (84.639) Epoch: [16][1020/1345], lr: 0.01000 Time 0.711 (0.747) Data 0.000 (0.004) Loss 1.5627 (1.5609) Prec@1 59.375 (59.300) Prec@5 85.938 (84.606) Epoch: [16][1040/1345], lr: 0.01000 Time 0.779 (0.747) Data 0.000 (0.004) Loss 1.9623 (1.5614) Prec@1 40.625 (59.258) Prec@5 81.250 (84.602) Epoch: [16][1060/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.004) Loss 1.4215 (1.5624) Prec@1 56.250 (59.226) Prec@5 89.062 (84.587) Epoch: [16][1080/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.004) Loss 1.7253 (1.5617) Prec@1 51.562 (59.233) Prec@5 82.812 (84.596) Epoch: [16][1100/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.003) Loss 1.7792 (1.5623) Prec@1 60.938 (59.237) Prec@5 78.125 (84.582) Epoch: [16][1120/1345], lr: 0.01000 Time 0.769 (0.746) Data 0.000 (0.003) Loss 1.3403 (1.5634) Prec@1 67.188 (59.244) Prec@5 89.062 (84.562) Epoch: [16][1140/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.000 (0.003) Loss 1.5147 (1.5643) Prec@1 53.125 (59.218) Prec@5 87.500 (84.541) Epoch: [16][1160/1345], lr: 0.01000 Time 0.755 (0.746) Data 0.000 (0.003) Loss 1.6562 (1.5652) Prec@1 57.812 (59.200) Prec@5 84.375 (84.526) Epoch: [16][1180/1345], lr: 0.01000 Time 0.739 (0.746) Data 0.000 (0.003) Loss 1.6394 (1.5657) Prec@1 56.250 (59.214) Prec@5 78.125 (84.506) Epoch: [16][1200/1345], lr: 0.01000 Time 0.709 (0.746) Data 0.000 (0.003) Loss 1.2198 (1.5648) Prec@1 65.625 (59.241) Prec@5 89.062 (84.505) Epoch: [16][1220/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.003) Loss 1.4522 (1.5642) Prec@1 62.500 (59.243) Prec@5 85.938 (84.517) Epoch: [16][1240/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.003) Loss 1.5251 (1.5649) Prec@1 59.375 (59.239) Prec@5 82.812 (84.524) Epoch: [16][1260/1345], lr: 0.01000 Time 0.711 (0.746) Data 0.000 (0.003) Loss 1.7049 (1.5671) Prec@1 60.938 (59.211) Prec@5 79.688 (84.482) Epoch: [16][1280/1345], lr: 0.01000 Time 0.752 (0.746) Data 0.000 (0.003) Loss 1.5146 (1.5681) Prec@1 54.688 (59.182) Prec@5 84.375 (84.471) Epoch: [16][1300/1345], lr: 0.01000 Time 0.779 (0.746) Data 0.000 (0.003) Loss 1.8948 (1.5682) Prec@1 51.562 (59.183) Prec@5 82.812 (84.481) Epoch: [16][1320/1345], lr: 0.01000 Time 0.752 (0.746) Data 0.000 (0.003) Loss 1.8376 (1.5695) Prec@1 56.250 (59.156) Prec@5 85.938 (84.468) Epoch: [16][1340/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.003) Loss 1.5196 (1.5707) Prec@1 64.062 (59.123) Prec@5 84.375 (84.448) No BN layer Freezing. Test: [0/181] Time 3.066 (3.0661) Loss 2.8857 (2.8857) Prec@1 46.875 (46.875) Prec@5 73.438 (73.438) Test: [20/181] Time 0.808 (0.5883) Loss 2.4555 (2.4568) Prec@1 40.625 (44.643) Prec@5 73.438 (75.149) Test: [40/181] Time 1.117 (0.5251) Loss 2.2853 (2.4913) Prec@1 48.438 (43.979) Prec@5 76.562 (74.276) Test: [60/181] Time 0.516 (0.4994) Loss 2.4429 (2.5230) Prec@1 45.312 (43.212) Prec@5 75.000 (73.540) Test: [80/181] Time 0.303 (0.4872) Loss 2.0443 (2.5631) Prec@1 48.438 (42.477) Prec@5 81.250 (72.878) Test: [100/181] Time 0.238 (0.4829) Loss 2.8132 (2.5553) Prec@1 34.375 (42.389) Prec@5 71.875 (73.159) Test: [120/181] Time 0.238 (0.4818) Loss 2.8858 (2.5575) Prec@1 42.188 (42.588) Prec@5 65.625 (73.037) Test: [140/181] Time 0.237 (0.4783) Loss 2.9686 (2.5538) Prec@1 28.125 (42.465) Prec@5 71.875 (73.061) Test: [160/181] Time 0.235 (0.4758) Loss 3.0221 (2.5308) Prec@1 37.500 (42.789) Prec@5 70.312 (73.273) Testing Results: Prec@1 43.073 Prec@5 73.194 Loss 2.51932 Time 0.4767 No BN layer Freezing. Epoch: [17][0/1345], lr: 0.01000 Time 4.328 (4.328) Data 3.575 (3.575) Loss 1.5312 (1.5312) Prec@1 60.938 (60.938) Prec@5 89.062 (89.062) Epoch: [17][20/1345], lr: 0.01000 Time 0.735 (0.929) Data 0.000 (0.171) Loss 1.6477 (1.4214) Prec@1 54.688 (60.714) Prec@5 84.375 (87.500) Epoch: [17][40/1345], lr: 0.01000 Time 0.710 (0.836) Data 0.000 (0.088) Loss 1.6585 (1.4643) Prec@1 60.938 (60.823) Prec@5 78.125 (86.319) Epoch: [17][60/1345], lr: 0.01000 Time 0.737 (0.805) Data 0.000 (0.059) Loss 1.4883 (1.4734) Prec@1 62.500 (60.707) Prec@5 84.375 (85.989) Epoch: [17][80/1345], lr: 0.01000 Time 0.748 (0.788) Data 0.000 (0.045) Loss 1.7081 (1.4718) Prec@1 59.375 (61.111) Prec@5 82.812 (85.918) Epoch: [17][100/1345], lr: 0.01000 Time 0.766 (0.777) Data 0.000 (0.036) Loss 1.2390 (1.4856) Prec@1 62.500 (60.613) Prec@5 93.750 (85.659) Epoch: [17][120/1345], lr: 0.01000 Time 0.730 (0.770) Data 0.000 (0.030) Loss 1.6364 (1.4859) Prec@1 48.438 (60.615) Prec@5 82.812 (85.666) Epoch: [17][140/1345], lr: 0.01000 Time 0.710 (0.766) Data 0.000 (0.026) Loss 1.4177 (1.4872) Prec@1 65.625 (60.771) Prec@5 85.938 (85.738) Epoch: [17][160/1345], lr: 0.01000 Time 0.712 (0.762) Data 0.000 (0.023) Loss 1.3415 (1.4864) Prec@1 60.938 (60.724) Prec@5 87.500 (85.734) Epoch: [17][180/1345], lr: 0.01000 Time 0.743 (0.758) Data 0.000 (0.020) Loss 1.5850 (1.4881) Prec@1 57.812 (60.799) Prec@5 85.938 (85.834) Epoch: [17][200/1345], lr: 0.01000 Time 0.714 (0.755) Data 0.001 (0.018) Loss 1.8416 (1.4923) Prec@1 54.688 (60.728) Prec@5 84.375 (85.790) Epoch: [17][220/1345], lr: 0.01000 Time 0.746 (0.754) Data 0.000 (0.017) Loss 1.2373 (1.4880) Prec@1 65.625 (60.874) Prec@5 90.625 (85.853) Epoch: [17][240/1345], lr: 0.01000 Time 0.743 (0.752) Data 0.000 (0.015) Loss 1.2576 (1.4833) Prec@1 65.625 (61.015) Prec@5 89.062 (85.931) Epoch: [17][260/1345], lr: 0.01000 Time 0.712 (0.751) Data 0.000 (0.014) Loss 1.3005 (1.4889) Prec@1 67.188 (60.896) Prec@5 87.500 (85.812) Epoch: [17][280/1345], lr: 0.01000 Time 0.743 (0.751) Data 0.000 (0.013) Loss 1.5222 (1.4935) Prec@1 62.500 (60.882) Prec@5 82.812 (85.704) Epoch: [17][300/1345], lr: 0.01000 Time 0.749 (0.751) Data 0.000 (0.012) Loss 1.6426 (1.4895) Prec@1 48.438 (60.969) Prec@5 85.938 (85.756) Epoch: [17][320/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.012) Loss 1.4629 (1.4923) Prec@1 60.938 (60.869) Prec@5 84.375 (85.704) Epoch: [17][340/1345], lr: 0.01000 Time 0.708 (0.750) Data 0.000 (0.011) Loss 1.6282 (1.4945) Prec@1 68.750 (60.841) Prec@5 81.250 (85.690) Epoch: [17][360/1345], lr: 0.01000 Time 0.813 (0.750) Data 0.000 (0.010) Loss 1.8535 (1.4999) Prec@1 50.000 (60.648) Prec@5 89.062 (85.587) Epoch: [17][380/1345], lr: 0.01000 Time 0.733 (0.749) Data 0.000 (0.010) Loss 1.6416 (1.4991) Prec@1 54.688 (60.671) Prec@5 85.938 (85.626) Epoch: [17][400/1345], lr: 0.01000 Time 0.741 (0.749) Data 0.000 (0.009) Loss 1.7353 (1.5029) Prec@1 59.375 (60.626) Prec@5 82.812 (85.595) Epoch: [17][420/1345], lr: 0.01000 Time 0.816 (0.749) Data 0.000 (0.009) Loss 1.3209 (1.5005) Prec@1 65.625 (60.659) Prec@5 85.938 (85.633) Epoch: [17][440/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.009) Loss 1.8178 (1.5002) Prec@1 56.250 (60.651) Prec@5 85.938 (85.626) Epoch: [17][460/1345], lr: 0.01000 Time 0.727 (0.749) Data 0.000 (0.008) Loss 1.7864 (1.5077) Prec@1 53.125 (60.558) Prec@5 79.688 (85.521) Epoch: [17][480/1345], lr: 0.01000 Time 0.719 (0.748) Data 0.000 (0.008) Loss 1.8809 (1.5097) Prec@1 53.125 (60.496) Prec@5 79.688 (85.476) Epoch: [17][500/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.001 (0.008) Loss 1.6499 (1.5077) Prec@1 48.438 (60.516) Prec@5 87.500 (85.520) Epoch: [17][520/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.007) Loss 1.5102 (1.5082) Prec@1 59.375 (60.479) Prec@5 84.375 (85.524) Epoch: [17][540/1345], lr: 0.01000 Time 0.761 (0.748) Data 0.000 (0.007) Loss 1.9604 (1.5104) Prec@1 56.250 (60.447) Prec@5 81.250 (85.461) Epoch: [17][560/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.007) Loss 1.5268 (1.5130) Prec@1 57.812 (60.417) Prec@5 87.500 (85.414) Epoch: [17][580/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.000 (0.007) Loss 1.5968 (1.5155) Prec@1 62.500 (60.373) Prec@5 85.938 (85.386) Epoch: [17][600/1345], lr: 0.01000 Time 0.749 (0.748) Data 0.000 (0.006) Loss 1.8747 (1.5173) Prec@1 45.312 (60.337) Prec@5 78.125 (85.347) Epoch: [17][620/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.006) Loss 1.7802 (1.5191) Prec@1 57.812 (60.306) Prec@5 81.250 (85.331) Epoch: [17][640/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.006) Loss 1.5033 (1.5195) Prec@1 62.500 (60.301) Prec@5 82.812 (85.331) Epoch: [17][660/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.006) Loss 0.9976 (1.5166) Prec@1 73.438 (60.387) Prec@5 96.875 (85.387) Epoch: [17][680/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.006) Loss 1.5176 (1.5151) Prec@1 59.375 (60.401) Prec@5 84.375 (85.405) Epoch: [17][700/1345], lr: 0.01000 Time 0.706 (0.745) Data 0.000 (0.006) Loss 1.3973 (1.5152) Prec@1 67.188 (60.416) Prec@5 82.812 (85.389) Epoch: [17][720/1345], lr: 0.01000 Time 0.772 (0.745) Data 0.001 (0.005) Loss 1.3840 (1.5148) Prec@1 65.625 (60.433) Prec@5 84.375 (85.396) Epoch: [17][740/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.005) Loss 1.9132 (1.5176) Prec@1 53.125 (60.360) Prec@5 81.250 (85.358) Epoch: [17][760/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.005) Loss 2.0501 (1.5185) Prec@1 50.000 (60.322) Prec@5 76.562 (85.354) Epoch: [17][780/1345], lr: 0.01000 Time 0.706 (0.744) Data 0.000 (0.005) Loss 1.2985 (1.5199) Prec@1 64.062 (60.279) Prec@5 89.062 (85.351) Epoch: [17][800/1345], lr: 0.01000 Time 0.783 (0.744) Data 0.000 (0.005) Loss 1.8056 (1.5213) Prec@1 48.438 (60.288) Prec@5 81.250 (85.313) Epoch: [17][820/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.005) Loss 1.7459 (1.5212) Prec@1 54.688 (60.306) Prec@5 79.688 (85.298) Epoch: [17][840/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.005) Loss 1.7079 (1.5231) Prec@1 54.688 (60.274) Prec@5 81.250 (85.241) Epoch: [17][860/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.005) Loss 1.8458 (1.5252) Prec@1 57.812 (60.217) Prec@5 76.562 (85.210) Epoch: [17][880/1345], lr: 0.01000 Time 0.789 (0.743) Data 0.000 (0.004) Loss 1.6420 (1.5248) Prec@1 57.812 (60.225) Prec@5 84.375 (85.203) Epoch: [17][900/1345], lr: 0.01000 Time 0.749 (0.743) Data 0.000 (0.004) Loss 1.4043 (1.5259) Prec@1 59.375 (60.207) Prec@5 87.500 (85.199) Epoch: [17][920/1345], lr: 0.01000 Time 0.762 (0.743) Data 0.000 (0.004) Loss 1.5656 (1.5250) Prec@1 54.688 (60.203) Prec@5 85.938 (85.213) Epoch: [17][940/1345], lr: 0.01000 Time 0.735 (0.743) Data 0.000 (0.004) Loss 1.3690 (1.5271) Prec@1 62.500 (60.164) Prec@5 85.938 (85.175) Epoch: [17][960/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.4536 (1.5281) Prec@1 54.688 (60.139) Prec@5 89.062 (85.164) Epoch: [17][980/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.5967 (1.5305) Prec@1 57.812 (60.097) Prec@5 84.375 (85.140) Epoch: [17][1000/1345], lr: 0.01000 Time 0.751 (0.743) Data 0.000 (0.004) Loss 1.4843 (1.5314) Prec@1 57.812 (60.102) Prec@5 82.812 (85.102) Epoch: [17][1020/1345], lr: 0.01000 Time 0.855 (0.743) Data 0.000 (0.004) Loss 1.8820 (1.5329) Prec@1 54.688 (60.076) Prec@5 78.125 (85.064) Epoch: [17][1040/1345], lr: 0.01000 Time 0.825 (0.743) Data 0.000 (0.004) Loss 1.7949 (1.5346) Prec@1 59.375 (60.068) Prec@5 78.125 (85.046) Epoch: [17][1060/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.000 (0.004) Loss 1.7487 (1.5357) Prec@1 51.562 (60.051) Prec@5 82.812 (85.014) Epoch: [17][1080/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.000 (0.004) Loss 1.3537 (1.5366) Prec@1 60.938 (60.018) Prec@5 89.062 (85.008) Epoch: [17][1100/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.6192 (1.5374) Prec@1 51.562 (59.991) Prec@5 81.250 (84.998) Epoch: [17][1120/1345], lr: 0.01000 Time 0.757 (0.742) Data 0.000 (0.004) Loss 1.6040 (1.5395) Prec@1 54.688 (59.906) Prec@5 82.812 (84.951) Epoch: [17][1140/1345], lr: 0.01000 Time 0.734 (0.742) Data 0.000 (0.004) Loss 1.3209 (1.5409) Prec@1 67.188 (59.906) Prec@5 87.500 (84.924) Epoch: [17][1160/1345], lr: 0.01000 Time 0.721 (0.742) Data 0.000 (0.004) Loss 1.5613 (1.5403) Prec@1 59.375 (59.943) Prec@5 82.812 (84.921) Epoch: [17][1180/1345], lr: 0.01000 Time 0.751 (0.742) Data 0.001 (0.003) Loss 1.4450 (1.5408) Prec@1 64.062 (59.913) Prec@5 85.938 (84.915) Epoch: [17][1200/1345], lr: 0.01000 Time 0.713 (0.742) Data 0.001 (0.003) Loss 1.4216 (1.5420) Prec@1 62.500 (59.882) Prec@5 90.625 (84.906) Epoch: [17][1220/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.001 (0.003) Loss 1.6972 (1.5421) Prec@1 59.375 (59.851) Prec@5 82.812 (84.891) Epoch: [17][1240/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.003) Loss 1.9429 (1.5427) Prec@1 50.000 (59.855) Prec@5 75.000 (84.882) Epoch: [17][1260/1345], lr: 0.01000 Time 0.837 (0.742) Data 0.000 (0.003) Loss 1.3705 (1.5430) Prec@1 64.062 (59.850) Prec@5 89.062 (84.867) Epoch: [17][1280/1345], lr: 0.01000 Time 0.706 (0.742) Data 0.000 (0.003) Loss 1.6994 (1.5444) Prec@1 56.250 (59.814) Prec@5 84.375 (84.853) Epoch: [17][1300/1345], lr: 0.01000 Time 0.757 (0.742) Data 0.000 (0.003) Loss 1.3801 (1.5448) Prec@1 64.062 (59.804) Prec@5 87.500 (84.849) Epoch: [17][1320/1345], lr: 0.01000 Time 0.750 (0.742) Data 0.000 (0.003) Loss 1.3159 (1.5457) Prec@1 64.062 (59.777) Prec@5 92.188 (84.834) Epoch: [17][1340/1345], lr: 0.01000 Time 0.716 (0.742) Data 0.001 (0.003) Loss 1.6529 (1.5468) Prec@1 53.125 (59.763) Prec@5 85.938 (84.817) No BN layer Freezing. Test: [0/181] Time 3.764 (3.7639) Loss 2.8725 (2.8725) Prec@1 35.938 (35.938) Prec@5 76.562 (76.562) Test: [20/181] Time 1.025 (0.6085) Loss 2.5912 (2.4854) Prec@1 46.875 (42.634) Prec@5 76.562 (73.735) Test: [40/181] Time 1.006 (0.5363) Loss 2.4400 (2.4919) Prec@1 45.312 (41.883) Prec@5 78.125 (73.476) Test: [60/181] Time 0.915 (0.5108) Loss 2.2582 (2.5263) Prec@1 46.875 (41.726) Prec@5 68.750 (72.951) Test: [80/181] Time 1.144 (0.5004) Loss 1.9228 (2.5387) Prec@1 53.125 (41.667) Prec@5 82.812 (72.704) Test: [100/181] Time 0.974 (0.4918) Loss 2.8988 (2.5307) Prec@1 34.375 (41.739) Prec@5 64.062 (72.803) Test: [120/181] Time 1.205 (0.4890) Loss 3.0697 (2.5460) Prec@1 40.625 (42.110) Prec@5 70.312 (72.676) Test: [140/181] Time 1.124 (0.4876) Loss 3.2776 (2.5555) Prec@1 32.812 (42.032) Prec@5 68.750 (72.551) Test: [160/181] Time 0.975 (0.4861) Loss 2.3456 (2.5380) Prec@1 48.438 (42.420) Prec@5 75.000 (72.807) Testing Results: Prec@1 42.569 Prec@5 72.865 Loss 2.53729 Time 0.4806 No BN layer Freezing. Epoch: [18][0/1345], lr: 0.01000 Time 3.978 (3.978) Data 3.218 (3.218) Loss 1.6079 (1.6079) Prec@1 60.938 (60.938) Prec@5 87.500 (87.500) Epoch: [18][20/1345], lr: 0.01000 Time 0.773 (0.907) Data 0.000 (0.154) Loss 1.1947 (1.4763) Prec@1 65.625 (61.905) Prec@5 90.625 (85.863) Epoch: [18][40/1345], lr: 0.01000 Time 0.750 (0.822) Data 0.000 (0.079) Loss 1.6818 (1.5402) Prec@1 57.812 (59.909) Prec@5 85.938 (85.366) Epoch: [18][60/1345], lr: 0.01000 Time 0.710 (0.792) Data 0.000 (0.053) Loss 1.5446 (1.5326) Prec@1 60.938 (60.220) Prec@5 81.250 (84.887) Epoch: [18][80/1345], lr: 0.01000 Time 0.712 (0.779) Data 0.000 (0.040) Loss 1.6999 (1.5351) Prec@1 60.938 (60.185) Prec@5 78.125 (84.896) Epoch: [18][100/1345], lr: 0.01000 Time 0.714 (0.770) Data 0.001 (0.032) Loss 1.8439 (1.5274) Prec@1 51.562 (60.118) Prec@5 81.250 (85.087) Epoch: [18][120/1345], lr: 0.01000 Time 0.708 (0.762) Data 0.000 (0.027) Loss 1.6987 (1.5055) Prec@1 62.500 (60.886) Prec@5 78.125 (85.421) Epoch: [18][140/1345], lr: 0.01000 Time 0.711 (0.759) Data 0.001 (0.023) Loss 1.8095 (1.5096) Prec@1 53.125 (60.749) Prec@5 84.375 (85.273) Epoch: [18][160/1345], lr: 0.01000 Time 0.714 (0.756) Data 0.000 (0.020) Loss 1.3136 (1.5103) Prec@1 60.938 (60.724) Prec@5 92.188 (85.229) Epoch: [18][180/1345], lr: 0.01000 Time 0.711 (0.753) Data 0.001 (0.018) Loss 1.3295 (1.5073) Prec@1 59.375 (60.739) Prec@5 90.625 (85.264) Epoch: [18][200/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.016) Loss 1.4495 (1.5111) Prec@1 59.375 (60.627) Prec@5 79.688 (85.238) Epoch: [18][220/1345], lr: 0.01000 Time 0.708 (0.750) Data 0.000 (0.015) Loss 1.6797 (1.5183) Prec@1 54.688 (60.372) Prec@5 81.250 (85.068) Epoch: [18][240/1345], lr: 0.01000 Time 0.780 (0.748) Data 0.000 (0.014) Loss 1.3219 (1.5131) Prec@1 62.500 (60.490) Prec@5 84.375 (85.153) Epoch: [18][260/1345], lr: 0.01000 Time 0.890 (0.747) Data 0.001 (0.013) Loss 1.1785 (1.5085) Prec@1 67.188 (60.554) Prec@5 90.625 (85.267) Epoch: [18][280/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.012) Loss 1.2904 (1.5098) Prec@1 60.938 (60.415) Prec@5 90.625 (85.326) Epoch: [18][300/1345], lr: 0.01000 Time 0.752 (0.747) Data 0.000 (0.011) Loss 1.5234 (1.5057) Prec@1 57.812 (60.564) Prec@5 89.062 (85.460) Epoch: [18][320/1345], lr: 0.01000 Time 0.755 (0.746) Data 0.000 (0.010) Loss 1.7521 (1.5058) Prec@1 57.812 (60.558) Prec@5 81.250 (85.460) Epoch: [18][340/1345], lr: 0.01000 Time 0.706 (0.745) Data 0.000 (0.010) Loss 1.5556 (1.5061) Prec@1 64.062 (60.557) Prec@5 85.938 (85.479) Epoch: [18][360/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.009) Loss 1.7341 (1.5074) Prec@1 59.375 (60.509) Prec@5 84.375 (85.487) Epoch: [18][380/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.009) Loss 1.8723 (1.5107) Prec@1 56.250 (60.429) Prec@5 78.125 (85.429) Epoch: [18][400/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.008) Loss 1.2074 (1.5105) Prec@1 64.062 (60.450) Prec@5 93.750 (85.380) Epoch: [18][420/1345], lr: 0.01000 Time 0.830 (0.742) Data 0.000 (0.008) Loss 1.8851 (1.5155) Prec@1 56.250 (60.362) Prec@5 84.375 (85.318) Epoch: [18][440/1345], lr: 0.01000 Time 0.828 (0.742) Data 0.000 (0.008) Loss 1.8168 (1.5155) Prec@1 53.125 (60.402) Prec@5 79.688 (85.328) Epoch: [18][460/1345], lr: 0.01000 Time 0.754 (0.742) Data 0.000 (0.007) Loss 1.4613 (1.5185) Prec@1 57.812 (60.365) Prec@5 92.188 (85.249) Epoch: [18][480/1345], lr: 0.01000 Time 0.746 (0.742) Data 0.000 (0.007) Loss 2.0439 (1.5233) Prec@1 50.000 (60.252) Prec@5 76.562 (85.181) Epoch: [18][500/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.007) Loss 1.9186 (1.5256) Prec@1 50.000 (60.205) Prec@5 81.250 (85.086) Epoch: [18][520/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.007) Loss 1.7718 (1.5287) Prec@1 57.812 (60.149) Prec@5 79.688 (85.014) Epoch: [18][540/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.006) Loss 1.3599 (1.5292) Prec@1 65.625 (60.129) Prec@5 89.062 (85.019) Epoch: [18][560/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.006) Loss 1.4873 (1.5311) Prec@1 56.250 (60.066) Prec@5 90.625 (85.007) Epoch: [18][580/1345], lr: 0.01000 Time 0.756 (0.739) Data 0.000 (0.006) Loss 1.1462 (1.5320) Prec@1 70.312 (60.029) Prec@5 89.062 (84.996) Epoch: [18][600/1345], lr: 0.01000 Time 0.749 (0.739) Data 0.000 (0.006) Loss 1.8187 (1.5352) Prec@1 46.875 (59.952) Prec@5 79.688 (84.934) Epoch: [18][620/1345], lr: 0.01000 Time 0.719 (0.739) Data 0.000 (0.006) Loss 1.8361 (1.5348) Prec@1 54.688 (59.959) Prec@5 82.812 (84.946) Epoch: [18][640/1345], lr: 0.01000 Time 0.722 (0.740) Data 0.001 (0.005) Loss 1.5873 (1.5359) Prec@1 56.250 (59.928) Prec@5 89.062 (84.943) Epoch: [18][660/1345], lr: 0.01000 Time 0.772 (0.740) Data 0.000 (0.005) Loss 1.4743 (1.5360) Prec@1 60.938 (59.968) Prec@5 84.375 (84.961) Epoch: [18][680/1345], lr: 0.01000 Time 0.706 (0.741) Data 0.000 (0.005) Loss 1.4713 (1.5369) Prec@1 60.938 (59.928) Prec@5 84.375 (84.951) Epoch: [18][700/1345], lr: 0.01000 Time 0.814 (0.741) Data 0.001 (0.005) Loss 1.4503 (1.5372) Prec@1 56.250 (59.930) Prec@5 89.062 (84.939) Epoch: [18][720/1345], lr: 0.01000 Time 0.716 (0.742) Data 0.001 (0.005) Loss 1.5988 (1.5354) Prec@1 57.812 (59.993) Prec@5 84.375 (84.975) Epoch: [18][740/1345], lr: 0.01000 Time 0.740 (0.742) Data 0.000 (0.005) Loss 1.7545 (1.5379) Prec@1 56.250 (59.951) Prec@5 76.562 (84.930) Epoch: [18][760/1345], lr: 0.01000 Time 0.744 (0.742) Data 0.001 (0.005) Loss 1.6495 (1.5363) Prec@1 46.875 (59.962) Prec@5 85.938 (84.960) Epoch: [18][780/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.001 (0.005) Loss 1.5120 (1.5364) Prec@1 64.062 (59.963) Prec@5 84.375 (84.969) Epoch: [18][800/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.004) Loss 1.2220 (1.5361) Prec@1 70.312 (60.007) Prec@5 93.750 (84.966) Epoch: [18][820/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.004) Loss 1.6289 (1.5373) Prec@1 53.125 (60.011) Prec@5 89.062 (84.973) Epoch: [18][840/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.004) Loss 1.4937 (1.5401) Prec@1 57.812 (59.960) Prec@5 82.812 (84.908) Epoch: [18][860/1345], lr: 0.01000 Time 0.783 (0.740) Data 0.000 (0.004) Loss 1.6833 (1.5396) Prec@1 51.562 (59.977) Prec@5 84.375 (84.887) Epoch: [18][880/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.004) Loss 1.5361 (1.5413) Prec@1 56.250 (59.928) Prec@5 92.188 (84.872) Epoch: [18][900/1345], lr: 0.01000 Time 0.777 (0.739) Data 0.000 (0.004) Loss 1.5078 (1.5412) Prec@1 62.500 (59.918) Prec@5 84.375 (84.876) Epoch: [18][920/1345], lr: 0.01000 Time 0.710 (0.739) Data 0.000 (0.004) Loss 1.5060 (1.5406) Prec@1 70.312 (59.937) Prec@5 84.375 (84.898) Epoch: [18][940/1345], lr: 0.01000 Time 0.728 (0.739) Data 0.000 (0.004) Loss 1.7026 (1.5412) Prec@1 54.688 (59.963) Prec@5 82.812 (84.886) Epoch: [18][960/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.004) Loss 1.6264 (1.5414) Prec@1 56.250 (59.959) Prec@5 79.688 (84.890) Epoch: [18][980/1345], lr: 0.01000 Time 0.790 (0.739) Data 0.000 (0.004) Loss 1.0566 (1.5404) Prec@1 75.000 (59.987) Prec@5 95.312 (84.929) Epoch: [18][1000/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.004) Loss 1.1575 (1.5393) Prec@1 60.938 (59.998) Prec@5 92.188 (84.962) Epoch: [18][1020/1345], lr: 0.01000 Time 0.751 (0.739) Data 0.000 (0.004) Loss 1.5410 (1.5392) Prec@1 57.812 (59.993) Prec@5 87.500 (84.990) Epoch: [18][1040/1345], lr: 0.01000 Time 0.711 (0.739) Data 0.000 (0.004) Loss 1.3318 (1.5391) Prec@1 60.938 (59.990) Prec@5 87.500 (85.010) Epoch: [18][1060/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.5773 (1.5398) Prec@1 54.688 (59.979) Prec@5 82.812 (84.991) Epoch: [18][1080/1345], lr: 0.01000 Time 0.766 (0.738) Data 0.000 (0.003) Loss 1.2884 (1.5388) Prec@1 68.750 (60.020) Prec@5 89.062 (85.007) Epoch: [18][1100/1345], lr: 0.01000 Time 0.711 (0.738) Data 0.000 (0.003) Loss 1.5766 (1.5379) Prec@1 57.812 (60.028) Prec@5 79.688 (85.014) Epoch: [18][1120/1345], lr: 0.01000 Time 0.713 (0.738) Data 0.000 (0.003) Loss 1.7896 (1.5378) Prec@1 50.000 (60.001) Prec@5 79.688 (85.011) Epoch: [18][1140/1345], lr: 0.01000 Time 0.711 (0.738) Data 0.000 (0.003) Loss 1.7574 (1.5386) Prec@1 54.688 (59.958) Prec@5 85.938 (85.001) Epoch: [18][1160/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.2015 (1.5374) Prec@1 70.312 (59.981) Prec@5 84.375 (85.010) Epoch: [18][1180/1345], lr: 0.01000 Time 0.732 (0.738) Data 0.000 (0.003) Loss 1.6578 (1.5386) Prec@1 56.250 (59.940) Prec@5 84.375 (84.969) Epoch: [18][1200/1345], lr: 0.01000 Time 0.709 (0.738) Data 0.000 (0.003) Loss 1.3696 (1.5389) Prec@1 65.625 (59.938) Prec@5 85.938 (84.970) Epoch: [18][1220/1345], lr: 0.01000 Time 0.709 (0.738) Data 0.000 (0.003) Loss 1.4404 (1.5388) Prec@1 67.188 (59.943) Prec@5 89.062 (84.953) Epoch: [18][1240/1345], lr: 0.01000 Time 0.709 (0.737) Data 0.000 (0.003) Loss 1.4952 (1.5403) Prec@1 57.812 (59.918) Prec@5 84.375 (84.924) Epoch: [18][1260/1345], lr: 0.01000 Time 0.836 (0.737) Data 0.001 (0.003) Loss 1.5755 (1.5392) Prec@1 60.938 (59.931) Prec@5 82.812 (84.914) Epoch: [18][1280/1345], lr: 0.01000 Time 0.716 (0.737) Data 0.000 (0.003) Loss 1.6652 (1.5394) Prec@1 57.812 (59.947) Prec@5 79.688 (84.908) Epoch: [18][1300/1345], lr: 0.01000 Time 0.718 (0.737) Data 0.001 (0.003) Loss 1.2685 (1.5398) Prec@1 67.188 (59.936) Prec@5 85.938 (84.890) Epoch: [18][1320/1345], lr: 0.01000 Time 0.754 (0.738) Data 0.000 (0.003) Loss 1.8731 (1.5408) Prec@1 46.875 (59.925) Prec@5 81.250 (84.888) Epoch: [18][1340/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.003) Loss 1.4099 (1.5417) Prec@1 68.750 (59.889) Prec@5 90.625 (84.880) No BN layer Freezing. Test: [0/181] Time 3.154 (3.1544) Loss 2.7917 (2.7917) Prec@1 32.812 (32.812) Prec@5 71.875 (71.875) Test: [20/181] Time 0.540 (0.5840) Loss 2.4050 (2.4838) Prec@1 48.438 (45.610) Prec@5 76.562 (75.000) Test: [40/181] Time 0.668 (0.5236) Loss 2.6651 (2.5675) Prec@1 39.062 (44.855) Prec@5 73.438 (73.780) Test: [60/181] Time 0.756 (0.5021) Loss 3.1053 (2.6402) Prec@1 37.500 (43.366) Prec@5 67.188 (73.079) Test: [80/181] Time 0.861 (0.4912) Loss 2.2405 (2.6630) Prec@1 53.125 (42.785) Prec@5 78.125 (72.704) Test: [100/181] Time 0.722 (0.4846) Loss 3.1108 (2.6513) Prec@1 29.688 (42.853) Prec@5 68.750 (73.376) Test: [120/181] Time 0.674 (0.4791) Loss 3.4083 (2.6495) Prec@1 42.188 (43.104) Prec@5 59.375 (73.102) Test: [140/181] Time 0.471 (0.4757) Loss 3.0219 (2.6440) Prec@1 35.938 (43.118) Prec@5 71.875 (72.939) Test: [160/181] Time 0.472 (0.4742) Loss 2.9855 (2.6288) Prec@1 35.938 (43.478) Prec@5 68.750 (73.166) Testing Results: Prec@1 43.750 Prec@5 73.325 Loss 2.61096 Time 0.4727 No BN layer Freezing. Epoch: [19][0/1345], lr: 0.01000 Time 4.087 (4.087) Data 3.337 (3.337) Loss 1.5282 (1.5282) Prec@1 64.062 (64.062) Prec@5 81.250 (81.250) Epoch: [19][20/1345], lr: 0.01000 Time 0.726 (0.903) Data 0.001 (0.159) Loss 1.3657 (1.4465) Prec@1 68.750 (62.054) Prec@5 87.500 (86.682) Epoch: [19][40/1345], lr: 0.01000 Time 0.710 (0.815) Data 0.000 (0.082) Loss 1.3240 (1.4136) Prec@1 70.312 (61.623) Prec@5 90.625 (86.814) Epoch: [19][60/1345], lr: 0.01000 Time 0.791 (0.790) Data 0.000 (0.055) Loss 1.1737 (1.4070) Prec@1 67.188 (61.911) Prec@5 92.188 (87.269) Epoch: [19][80/1345], lr: 0.01000 Time 0.709 (0.774) Data 0.000 (0.042) Loss 1.3705 (1.4358) Prec@1 59.375 (61.728) Prec@5 89.062 (86.728) Epoch: [19][100/1345], lr: 0.01000 Time 0.708 (0.763) Data 0.000 (0.033) Loss 1.3939 (1.4274) Prec@1 65.625 (62.175) Prec@5 85.938 (86.819) Epoch: [19][120/1345], lr: 0.01000 Time 0.708 (0.760) Data 0.000 (0.028) Loss 1.1523 (1.4318) Prec@1 73.438 (62.126) Prec@5 89.062 (86.557) Epoch: [19][140/1345], lr: 0.01000 Time 0.784 (0.757) Data 0.000 (0.024) Loss 1.2978 (1.4426) Prec@1 60.938 (61.990) Prec@5 92.188 (86.336) Epoch: [19][160/1345], lr: 0.01000 Time 0.708 (0.754) Data 0.000 (0.021) Loss 2.0485 (1.4497) Prec@1 56.250 (61.908) Prec@5 76.562 (86.277) Epoch: [19][180/1345], lr: 0.01000 Time 0.709 (0.751) Data 0.000 (0.019) Loss 1.5470 (1.4475) Prec@1 60.938 (61.853) Prec@5 82.812 (86.335) Epoch: [19][200/1345], lr: 0.01000 Time 0.737 (0.751) Data 0.000 (0.017) Loss 1.4870 (1.4538) Prec@1 71.875 (61.715) Prec@5 82.812 (86.186) Epoch: [19][220/1345], lr: 0.01000 Time 0.748 (0.749) Data 0.001 (0.016) Loss 1.5627 (1.4616) Prec@1 64.062 (61.517) Prec@5 82.812 (86.079) Epoch: [19][240/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.014) Loss 1.9104 (1.4679) Prec@1 45.312 (61.327) Prec@5 78.125 (86.015) Epoch: [19][260/1345], lr: 0.01000 Time 0.721 (0.748) Data 0.001 (0.013) Loss 1.4449 (1.4697) Prec@1 59.375 (61.237) Prec@5 82.812 (85.961) Epoch: [19][280/1345], lr: 0.01000 Time 0.787 (0.748) Data 0.000 (0.012) Loss 1.7409 (1.4699) Prec@1 54.688 (61.282) Prec@5 78.125 (85.887) Epoch: [19][300/1345], lr: 0.01000 Time 0.787 (0.748) Data 0.000 (0.012) Loss 1.7375 (1.4698) Prec@1 53.125 (61.244) Prec@5 89.062 (85.906) Epoch: [19][320/1345], lr: 0.01000 Time 0.714 (0.747) Data 0.001 (0.011) Loss 1.4952 (1.4723) Prec@1 67.188 (61.303) Prec@5 81.250 (85.879) Epoch: [19][340/1345], lr: 0.01000 Time 0.711 (0.746) Data 0.000 (0.010) Loss 1.2508 (1.4766) Prec@1 62.500 (61.231) Prec@5 92.188 (85.791) Epoch: [19][360/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.010) Loss 1.4698 (1.4753) Prec@1 62.500 (61.184) Prec@5 81.250 (85.795) Epoch: [19][380/1345], lr: 0.01000 Time 0.747 (0.746) Data 0.000 (0.009) Loss 1.6369 (1.4756) Prec@1 56.250 (61.126) Prec@5 84.375 (85.847) Epoch: [19][400/1345], lr: 0.01000 Time 0.763 (0.747) Data 0.000 (0.009) Loss 1.4329 (1.4825) Prec@1 60.938 (60.992) Prec@5 85.938 (85.680) Epoch: [19][420/1345], lr: 0.01000 Time 0.709 (0.746) Data 0.000 (0.008) Loss 1.3106 (1.4822) Prec@1 68.750 (60.982) Prec@5 92.188 (85.733) Epoch: [19][440/1345], lr: 0.01000 Time 0.725 (0.747) Data 0.000 (0.008) Loss 1.6570 (1.4837) Prec@1 53.125 (61.008) Prec@5 84.375 (85.743) Epoch: [19][460/1345], lr: 0.01000 Time 0.758 (0.747) Data 0.000 (0.008) Loss 1.7326 (1.4807) Prec@1 67.188 (61.097) Prec@5 81.250 (85.754) Epoch: [19][480/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.007) Loss 1.1861 (1.4787) Prec@1 71.875 (61.168) Prec@5 92.188 (85.811) Epoch: [19][500/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.007) Loss 1.9037 (1.4790) Prec@1 48.438 (61.171) Prec@5 79.688 (85.782) Epoch: [19][520/1345], lr: 0.01000 Time 0.711 (0.745) Data 0.001 (0.007) Loss 1.2466 (1.4787) Prec@1 70.312 (61.147) Prec@5 87.500 (85.782) Epoch: [19][540/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.000 (0.007) Loss 1.5598 (1.4809) Prec@1 67.188 (61.154) Prec@5 81.250 (85.692) Epoch: [19][560/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.006) Loss 1.7531 (1.4821) Prec@1 56.250 (61.132) Prec@5 78.125 (85.704) Epoch: [19][580/1345], lr: 0.01000 Time 0.744 (0.744) Data 0.000 (0.006) Loss 1.6840 (1.4844) Prec@1 57.812 (61.099) Prec@5 78.125 (85.658) Epoch: [19][600/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.006) Loss 1.3891 (1.4871) Prec@1 64.062 (61.005) Prec@5 87.500 (85.628) Epoch: [19][620/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.006) Loss 1.4622 (1.4879) Prec@1 60.938 (60.990) Prec@5 82.812 (85.638) Epoch: [19][640/1345], lr: 0.01000 Time 0.772 (0.743) Data 0.001 (0.006) Loss 2.2238 (1.4889) Prec@1 54.688 (60.979) Prec@5 67.188 (85.596) Epoch: [19][660/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.005) Loss 1.5274 (1.4906) Prec@1 64.062 (60.966) Prec@5 84.375 (85.559) Epoch: [19][680/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.005) Loss 1.6378 (1.4881) Prec@1 56.250 (61.052) Prec@5 85.938 (85.605) Epoch: [19][700/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.005) Loss 1.4616 (1.4888) Prec@1 57.812 (61.049) Prec@5 85.938 (85.596) Epoch: [19][720/1345], lr: 0.01000 Time 0.718 (0.741) Data 0.001 (0.005) Loss 1.7369 (1.4895) Prec@1 51.562 (60.996) Prec@5 84.375 (85.610) Epoch: [19][740/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.005) Loss 1.5447 (1.4905) Prec@1 60.938 (60.992) Prec@5 84.375 (85.585) Epoch: [19][760/1345], lr: 0.01000 Time 0.787 (0.741) Data 0.000 (0.005) Loss 1.8422 (1.4938) Prec@1 45.312 (60.917) Prec@5 84.375 (85.533) Epoch: [19][780/1345], lr: 0.01000 Time 0.844 (0.741) Data 0.000 (0.005) Loss 1.2917 (1.4948) Prec@1 60.938 (60.901) Prec@5 89.062 (85.507) Epoch: [19][800/1345], lr: 0.01000 Time 0.711 (0.741) Data 0.000 (0.005) Loss 1.4217 (1.4946) Prec@1 64.062 (60.875) Prec@5 89.062 (85.502) Epoch: [19][820/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.2906 (1.4941) Prec@1 62.500 (60.865) Prec@5 87.500 (85.528) Epoch: [19][840/1345], lr: 0.01000 Time 0.874 (0.741) Data 0.000 (0.004) Loss 1.8222 (1.4953) Prec@1 59.375 (60.854) Prec@5 79.688 (85.531) Epoch: [19][860/1345], lr: 0.01000 Time 0.710 (0.741) Data 0.000 (0.004) Loss 1.2674 (1.4962) Prec@1 62.500 (60.803) Prec@5 89.062 (85.529) Epoch: [19][880/1345], lr: 0.01000 Time 0.741 (0.741) Data 0.000 (0.004) Loss 1.4197 (1.4953) Prec@1 68.750 (60.801) Prec@5 85.938 (85.547) Epoch: [19][900/1345], lr: 0.01000 Time 0.786 (0.741) Data 0.001 (0.004) Loss 1.5587 (1.4966) Prec@1 54.688 (60.766) Prec@5 82.812 (85.528) Epoch: [19][920/1345], lr: 0.01000 Time 0.710 (0.741) Data 0.000 (0.004) Loss 1.7270 (1.4982) Prec@1 51.562 (60.731) Prec@5 81.250 (85.508) Epoch: [19][940/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.004) Loss 1.6131 (1.4997) Prec@1 59.375 (60.705) Prec@5 85.938 (85.488) Epoch: [19][960/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.004) Loss 1.4792 (1.5013) Prec@1 59.375 (60.651) Prec@5 90.625 (85.461) Epoch: [19][980/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.004) Loss 1.8746 (1.5038) Prec@1 45.312 (60.608) Prec@5 84.375 (85.423) Epoch: [19][1000/1345], lr: 0.01000 Time 0.725 (0.740) Data 0.000 (0.004) Loss 1.6133 (1.5044) Prec@1 62.500 (60.599) Prec@5 81.250 (85.411) Epoch: [19][1020/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.004) Loss 1.6332 (1.5054) Prec@1 57.812 (60.550) Prec@5 79.688 (85.402) Epoch: [19][1040/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.004) Loss 1.5865 (1.5065) Prec@1 56.250 (60.517) Prec@5 82.812 (85.385) Epoch: [19][1060/1345], lr: 0.01000 Time 0.746 (0.740) Data 0.000 (0.004) Loss 1.6214 (1.5076) Prec@1 60.938 (60.465) Prec@5 81.250 (85.360) Epoch: [19][1080/1345], lr: 0.01000 Time 0.854 (0.740) Data 0.001 (0.004) Loss 1.7983 (1.5073) Prec@1 54.688 (60.446) Prec@5 78.125 (85.367) Epoch: [19][1100/1345], lr: 0.01000 Time 0.874 (0.740) Data 0.000 (0.003) Loss 1.6251 (1.5079) Prec@1 54.688 (60.438) Prec@5 85.938 (85.370) Epoch: [19][1120/1345], lr: 0.01000 Time 0.710 (0.740) Data 0.000 (0.003) Loss 1.7531 (1.5073) Prec@1 59.375 (60.482) Prec@5 78.125 (85.373) Epoch: [19][1140/1345], lr: 0.01000 Time 0.712 (0.740) Data 0.000 (0.003) Loss 1.6811 (1.5085) Prec@1 56.250 (60.446) Prec@5 81.250 (85.364) Epoch: [19][1160/1345], lr: 0.01000 Time 0.739 (0.740) Data 0.000 (0.003) Loss 1.4210 (1.5090) Prec@1 57.812 (60.434) Prec@5 95.312 (85.352) Epoch: [19][1180/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.003) Loss 1.5236 (1.5112) Prec@1 54.688 (60.392) Prec@5 82.812 (85.308) Epoch: [19][1200/1345], lr: 0.01000 Time 0.717 (0.740) Data 0.000 (0.003) Loss 1.7753 (1.5113) Prec@1 57.812 (60.391) Prec@5 82.812 (85.303) Epoch: [19][1220/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.003) Loss 1.6314 (1.5106) Prec@1 53.125 (60.404) Prec@5 81.250 (85.312) Epoch: [19][1240/1345], lr: 0.01000 Time 0.824 (0.740) Data 0.000 (0.003) Loss 1.4875 (1.5116) Prec@1 64.062 (60.392) Prec@5 85.938 (85.289) Epoch: [19][1260/1345], lr: 0.01000 Time 0.782 (0.740) Data 0.000 (0.003) Loss 1.7213 (1.5117) Prec@1 50.000 (60.426) Prec@5 82.812 (85.273) Epoch: [19][1280/1345], lr: 0.01000 Time 0.771 (0.740) Data 0.000 (0.003) Loss 1.8612 (1.5121) Prec@1 57.812 (60.408) Prec@5 78.125 (85.270) Epoch: [19][1300/1345], lr: 0.01000 Time 0.717 (0.740) Data 0.000 (0.003) Loss 1.4579 (1.5122) Prec@1 59.375 (60.419) Prec@5 81.250 (85.252) Epoch: [19][1320/1345], lr: 0.01000 Time 0.719 (0.740) Data 0.000 (0.003) Loss 1.7375 (1.5118) Prec@1 50.000 (60.419) Prec@5 79.688 (85.274) Epoch: [19][1340/1345], lr: 0.01000 Time 0.706 (0.740) Data 0.000 (0.003) Loss 1.4789 (1.5109) Prec@1 65.625 (60.442) Prec@5 84.375 (85.284) No BN layer Freezing. Test: [0/181] Time 3.388 (3.3881) Loss 2.7116 (2.7116) Prec@1 46.875 (46.875) Prec@5 78.125 (78.125) Test: [20/181] Time 1.155 (0.6060) Loss 2.4075 (2.3549) Prec@1 46.875 (46.726) Prec@5 73.438 (76.265) Test: [40/181] Time 0.972 (0.5339) Loss 2.3964 (2.4399) Prec@1 46.875 (45.312) Prec@5 78.125 (75.152) Test: [60/181] Time 0.477 (0.5080) Loss 2.5970 (2.4904) Prec@1 39.062 (44.800) Prec@5 73.438 (74.565) Test: [80/181] Time 0.236 (0.5009) Loss 2.0820 (2.5105) Prec@1 51.562 (44.715) Prec@5 79.688 (74.074) Test: [100/181] Time 0.238 (0.4940) Loss 3.1369 (2.4966) Prec@1 34.375 (44.632) Prec@5 65.625 (74.319) Test: [120/181] Time 0.237 (0.4914) Loss 3.1602 (2.5152) Prec@1 34.375 (44.602) Prec@5 68.750 (74.109) Test: [140/181] Time 0.240 (0.4902) Loss 3.0730 (2.5105) Prec@1 32.812 (44.470) Prec@5 73.438 (74.324) Test: [160/181] Time 0.298 (0.4903) Loss 2.6724 (2.4949) Prec@1 45.312 (44.662) Prec@5 73.438 (74.563) Testing Results: Prec@1 44.922 Prec@5 74.583 Loss 2.48269 Time 0.4892 No BN layer Freezing. Epoch: [20][0/1345], lr: 0.01000 Time 3.990 (3.990) Data 3.228 (3.228) Loss 1.6094 (1.6094) Prec@1 64.062 (64.062) Prec@5 87.500 (87.500) Epoch: [20][20/1345], lr: 0.01000 Time 0.709 (0.905) Data 0.000 (0.154) Loss 1.3416 (1.5366) Prec@1 71.875 (61.384) Prec@5 89.062 (83.854) Epoch: [20][40/1345], lr: 0.01000 Time 0.709 (0.824) Data 0.000 (0.079) Loss 1.2256 (1.4971) Prec@1 68.750 (61.966) Prec@5 92.188 (84.642) Epoch: [20][60/1345], lr: 0.01000 Time 0.760 (0.797) Data 0.000 (0.053) Loss 1.7044 (1.4855) Prec@1 53.125 (61.552) Prec@5 85.938 (84.810) Epoch: [20][80/1345], lr: 0.01000 Time 0.710 (0.779) Data 0.000 (0.040) Loss 1.5796 (1.4794) Prec@1 59.375 (61.593) Prec@5 81.250 (85.069) Epoch: [20][100/1345], lr: 0.01000 Time 0.711 (0.773) Data 0.000 (0.032) Loss 1.3097 (1.4724) Prec@1 64.062 (61.402) Prec@5 92.188 (85.288) Epoch: [20][120/1345], lr: 0.01000 Time 0.712 (0.768) Data 0.000 (0.027) Loss 1.5167 (1.4785) Prec@1 64.062 (61.609) Prec@5 79.688 (85.072) Epoch: [20][140/1345], lr: 0.01000 Time 0.708 (0.765) Data 0.000 (0.023) Loss 1.2538 (1.4735) Prec@1 68.750 (61.669) Prec@5 85.938 (85.151) Epoch: [20][160/1345], lr: 0.01000 Time 0.709 (0.760) Data 0.000 (0.020) Loss 1.5873 (1.4646) Prec@1 57.812 (61.976) Prec@5 85.938 (85.316) Epoch: [20][180/1345], lr: 0.01000 Time 0.705 (0.758) Data 0.000 (0.018) Loss 1.3279 (1.4648) Prec@1 68.750 (62.051) Prec@5 87.500 (85.281) Epoch: [20][200/1345], lr: 0.01000 Time 0.789 (0.756) Data 0.000 (0.016) Loss 1.6104 (1.4663) Prec@1 50.000 (62.010) Prec@5 82.812 (85.300) Epoch: [20][220/1345], lr: 0.01000 Time 0.787 (0.755) Data 0.000 (0.015) Loss 1.6184 (1.4642) Prec@1 60.938 (61.998) Prec@5 81.250 (85.443) Epoch: [20][240/1345], lr: 0.01000 Time 0.715 (0.753) Data 0.000 (0.014) Loss 1.2416 (1.4631) Prec@1 65.625 (61.975) Prec@5 92.188 (85.419) Epoch: [20][260/1345], lr: 0.01000 Time 0.740 (0.751) Data 0.000 (0.013) Loss 1.6521 (1.4617) Prec@1 57.812 (61.961) Prec@5 81.250 (85.459) Epoch: [20][280/1345], lr: 0.01000 Time 0.754 (0.751) Data 0.000 (0.012) Loss 1.9005 (1.4649) Prec@1 54.688 (61.849) Prec@5 75.000 (85.354) Epoch: [20][300/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.011) Loss 1.5594 (1.4686) Prec@1 62.500 (61.768) Prec@5 85.938 (85.356) Epoch: [20][320/1345], lr: 0.01000 Time 0.709 (0.748) Data 0.000 (0.010) Loss 1.4749 (1.4667) Prec@1 57.812 (61.702) Prec@5 85.938 (85.460) Epoch: [20][340/1345], lr: 0.01000 Time 0.814 (0.748) Data 0.000 (0.010) Loss 1.5093 (1.4655) Prec@1 64.062 (61.712) Prec@5 82.812 (85.534) Epoch: [20][360/1345], lr: 0.01000 Time 0.771 (0.746) Data 0.000 (0.009) Loss 1.7349 (1.4652) Prec@1 53.125 (61.730) Prec@5 82.812 (85.518) Epoch: [20][380/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.009) Loss 1.3188 (1.4660) Prec@1 65.625 (61.684) Prec@5 92.188 (85.540) Epoch: [20][400/1345], lr: 0.01000 Time 0.775 (0.744) Data 0.000 (0.008) Loss 1.3422 (1.4668) Prec@1 60.938 (61.670) Prec@5 84.375 (85.524) Epoch: [20][420/1345], lr: 0.01000 Time 0.722 (0.744) Data 0.001 (0.008) Loss 1.3666 (1.4672) Prec@1 65.625 (61.658) Prec@5 81.250 (85.589) Epoch: [20][440/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.008) Loss 1.3248 (1.4668) Prec@1 64.062 (61.657) Prec@5 85.938 (85.622) Epoch: [20][460/1345], lr: 0.01000 Time 0.779 (0.744) Data 0.001 (0.007) Loss 1.5064 (1.4678) Prec@1 59.375 (61.588) Prec@5 89.062 (85.649) Epoch: [20][480/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.007) Loss 1.7937 (1.4722) Prec@1 56.250 (61.506) Prec@5 81.250 (85.616) Epoch: [20][500/1345], lr: 0.01000 Time 0.822 (0.743) Data 0.000 (0.007) Loss 1.3190 (1.4689) Prec@1 67.188 (61.608) Prec@5 87.500 (85.682) Epoch: [20][520/1345], lr: 0.01000 Time 0.711 (0.743) Data 0.000 (0.007) Loss 1.2515 (1.4696) Prec@1 68.750 (61.582) Prec@5 85.938 (85.695) Epoch: [20][540/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.006) Loss 1.5986 (1.4685) Prec@1 57.812 (61.541) Prec@5 82.812 (85.724) Epoch: [20][560/1345], lr: 0.01000 Time 0.751 (0.742) Data 0.000 (0.006) Loss 1.8206 (1.4720) Prec@1 50.000 (61.408) Prec@5 81.250 (85.715) Epoch: [20][580/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.000 (0.006) Loss 1.6376 (1.4722) Prec@1 59.375 (61.430) Prec@5 81.250 (85.738) Epoch: [20][600/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.006) Loss 1.2965 (1.4713) Prec@1 68.750 (61.447) Prec@5 85.938 (85.763) Epoch: [20][620/1345], lr: 0.01000 Time 0.763 (0.742) Data 0.000 (0.006) Loss 1.4368 (1.4664) Prec@1 64.062 (61.617) Prec@5 84.375 (85.819) Epoch: [20][640/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.005) Loss 1.1390 (1.4688) Prec@1 68.750 (61.610) Prec@5 92.188 (85.784) Epoch: [20][660/1345], lr: 0.01000 Time 0.846 (0.741) Data 0.000 (0.005) Loss 1.6717 (1.4697) Prec@1 60.938 (61.614) Prec@5 79.688 (85.765) Epoch: [20][680/1345], lr: 0.01000 Time 0.747 (0.741) Data 0.000 (0.005) Loss 1.5830 (1.4703) Prec@1 57.812 (61.619) Prec@5 82.812 (85.749) Epoch: [20][700/1345], lr: 0.01000 Time 0.711 (0.741) Data 0.000 (0.005) Loss 1.3084 (1.4727) Prec@1 64.062 (61.575) Prec@5 85.938 (85.686) Epoch: [20][720/1345], lr: 0.01000 Time 0.778 (0.741) Data 0.000 (0.005) Loss 1.3080 (1.4741) Prec@1 60.938 (61.516) Prec@5 87.500 (85.671) Epoch: [20][740/1345], lr: 0.01000 Time 0.716 (0.741) Data 0.000 (0.005) Loss 1.7626 (1.4771) Prec@1 51.562 (61.435) Prec@5 87.500 (85.628) Epoch: [20][760/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.005) Loss 1.6910 (1.4783) Prec@1 59.375 (61.369) Prec@5 79.688 (85.615) Epoch: [20][780/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.005) Loss 1.9602 (1.4812) Prec@1 50.000 (61.292) Prec@5 82.812 (85.557) Epoch: [20][800/1345], lr: 0.01000 Time 0.709 (0.740) Data 0.000 (0.004) Loss 1.3395 (1.4799) Prec@1 68.750 (61.324) Prec@5 84.375 (85.577) Epoch: [20][820/1345], lr: 0.01000 Time 0.849 (0.740) Data 0.000 (0.004) Loss 1.3109 (1.4803) Prec@1 64.062 (61.282) Prec@5 89.062 (85.574) Epoch: [20][840/1345], lr: 0.01000 Time 0.710 (0.740) Data 0.000 (0.004) Loss 1.3785 (1.4829) Prec@1 70.312 (61.240) Prec@5 81.250 (85.532) Epoch: [20][860/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.004) Loss 1.5505 (1.4841) Prec@1 64.062 (61.253) Prec@5 81.250 (85.535) Epoch: [20][880/1345], lr: 0.01000 Time 0.763 (0.740) Data 0.000 (0.004) Loss 1.8847 (1.4864) Prec@1 50.000 (61.219) Prec@5 84.375 (85.537) Epoch: [20][900/1345], lr: 0.01000 Time 0.707 (0.739) Data 0.000 (0.004) Loss 1.3668 (1.4869) Prec@1 65.625 (61.212) Prec@5 82.812 (85.532) Epoch: [20][920/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.004) Loss 1.6768 (1.4896) Prec@1 60.938 (61.167) Prec@5 78.125 (85.501) Epoch: [20][940/1345], lr: 0.01000 Time 0.709 (0.739) Data 0.000 (0.004) Loss 1.5831 (1.4922) Prec@1 64.062 (61.090) Prec@5 81.250 (85.461) Epoch: [20][960/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.004) Loss 1.1851 (1.4914) Prec@1 65.625 (61.079) Prec@5 90.625 (85.490) Epoch: [20][980/1345], lr: 0.01000 Time 0.786 (0.739) Data 0.000 (0.004) Loss 1.6306 (1.4926) Prec@1 54.688 (61.062) Prec@5 89.062 (85.485) Epoch: [20][1000/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.004) Loss 1.2946 (1.4935) Prec@1 65.625 (61.044) Prec@5 85.938 (85.482) Epoch: [20][1020/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.004) Loss 1.8086 (1.4956) Prec@1 53.125 (61.016) Prec@5 78.125 (85.465) Epoch: [20][1040/1345], lr: 0.01000 Time 0.778 (0.738) Data 0.000 (0.004) Loss 1.5619 (1.4950) Prec@1 54.688 (61.029) Prec@5 84.375 (85.475) Epoch: [20][1060/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.003) Loss 1.5689 (1.4930) Prec@1 57.812 (61.069) Prec@5 85.938 (85.518) Epoch: [20][1080/1345], lr: 0.01000 Time 0.706 (0.738) Data 0.000 (0.003) Loss 1.1818 (1.4924) Prec@1 70.312 (61.094) Prec@5 89.062 (85.527) Epoch: [20][1100/1345], lr: 0.01000 Time 0.773 (0.738) Data 0.000 (0.003) Loss 1.6320 (1.4930) Prec@1 62.500 (61.067) Prec@5 81.250 (85.522) Epoch: [20][1120/1345], lr: 0.01000 Time 0.800 (0.738) Data 0.000 (0.003) Loss 1.4096 (1.4933) Prec@1 65.625 (61.041) Prec@5 84.375 (85.540) Epoch: [20][1140/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.5121 (1.4951) Prec@1 54.688 (60.987) Prec@5 82.812 (85.531) Epoch: [20][1160/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.2849 (1.4951) Prec@1 64.062 (61.010) Prec@5 90.625 (85.527) Epoch: [20][1180/1345], lr: 0.01000 Time 0.778 (0.738) Data 0.000 (0.003) Loss 1.3162 (1.4953) Prec@1 65.625 (60.997) Prec@5 87.500 (85.510) Epoch: [20][1200/1345], lr: 0.01000 Time 0.708 (0.738) Data 0.000 (0.003) Loss 1.7866 (1.4948) Prec@1 62.500 (61.004) Prec@5 76.562 (85.519) Epoch: [20][1220/1345], lr: 0.01000 Time 0.708 (0.737) Data 0.000 (0.003) Loss 1.4618 (1.4948) Prec@1 64.062 (61.000) Prec@5 90.625 (85.532) Epoch: [20][1240/1345], lr: 0.01000 Time 0.793 (0.738) Data 0.000 (0.003) Loss 1.4008 (1.4949) Prec@1 64.062 (61.005) Prec@5 87.500 (85.523) Epoch: [20][1260/1345], lr: 0.01000 Time 0.712 (0.738) Data 0.000 (0.003) Loss 1.4260 (1.4954) Prec@1 62.500 (60.998) Prec@5 87.500 (85.509) Epoch: [20][1280/1345], lr: 0.01000 Time 0.707 (0.738) Data 0.000 (0.003) Loss 1.5110 (1.4953) Prec@1 59.375 (60.998) Prec@5 84.375 (85.507) Epoch: [20][1300/1345], lr: 0.01000 Time 0.794 (0.738) Data 0.000 (0.003) Loss 1.3985 (1.4960) Prec@1 60.938 (60.977) Prec@5 92.188 (85.492) Epoch: [20][1320/1345], lr: 0.01000 Time 0.707 (0.737) Data 0.000 (0.003) Loss 1.5742 (1.4968) Prec@1 57.812 (60.965) Prec@5 87.500 (85.494) Epoch: [20][1340/1345], lr: 0.01000 Time 0.708 (0.737) Data 0.000 (0.003) Loss 1.5688 (1.4971) Prec@1 67.188 (60.955) Prec@5 85.938 (85.494) No BN layer Freezing. Test: [0/181] Time 3.533 (3.5328) Loss 2.6510 (2.6510) Prec@1 43.750 (43.750) Prec@5 73.438 (73.438) Test: [20/181] Time 1.317 (0.6198) Loss 2.3142 (2.4511) Prec@1 42.188 (44.345) Prec@5 79.688 (73.810) Test: [40/181] Time 1.184 (0.5365) Loss 2.2184 (2.5033) Prec@1 51.562 (43.331) Prec@5 76.562 (73.857) Test: [60/181] Time 1.217 (0.5124) Loss 2.5887 (2.5545) Prec@1 42.188 (42.495) Prec@5 71.875 (73.463) Test: [80/181] Time 1.125 (0.4970) Loss 2.1492 (2.5875) Prec@1 50.000 (41.686) Prec@5 81.250 (73.245) Test: [100/181] Time 0.942 (0.4893) Loss 3.1689 (2.5798) Prec@1 29.688 (41.553) Prec@5 62.500 (73.438) Test: [120/181] Time 0.956 (0.4829) Loss 2.6799 (2.5809) Prec@1 43.750 (41.684) Prec@5 65.625 (73.334) Test: [140/181] Time 0.880 (0.4788) Loss 3.2912 (2.5809) Prec@1 31.250 (41.556) Prec@5 67.188 (73.271) Test: [160/181] Time 0.751 (0.4758) Loss 3.1064 (2.5680) Prec@1 37.500 (41.906) Prec@5 65.625 (73.360) Testing Results: Prec@1 42.196 Prec@5 73.316 Loss 2.55989 Time 0.4728 No BN layer Freezing. Epoch: [21][0/1345], lr: 0.01000 Time 4.150 (4.150) Data 3.178 (3.178) Loss 1.3223 (1.3223) Prec@1 65.625 (65.625) Prec@5 90.625 (90.625) Epoch: [21][20/1345], lr: 0.01000 Time 0.707 (0.895) Data 0.000 (0.152) Loss 1.4809 (1.3985) Prec@1 59.375 (64.137) Prec@5 79.688 (86.607) Epoch: [21][40/1345], lr: 0.01000 Time 0.714 (0.821) Data 0.000 (0.078) Loss 1.3737 (1.3681) Prec@1 67.188 (64.177) Prec@5 87.500 (87.233) Epoch: [21][60/1345], lr: 0.01000 Time 0.708 (0.788) Data 0.000 (0.052) Loss 1.1644 (1.3926) Prec@1 68.750 (64.088) Prec@5 89.062 (86.860) Epoch: [21][80/1345], lr: 0.01000 Time 0.709 (0.776) Data 0.000 (0.040) Loss 1.5842 (1.4256) Prec@1 59.375 (63.310) Prec@5 82.812 (86.111) Epoch: [21][100/1345], lr: 0.01000 Time 0.712 (0.767) Data 0.000 (0.032) Loss 1.7147 (1.4303) Prec@1 53.125 (63.026) Prec@5 87.500 (86.200) Epoch: [21][120/1345], lr: 0.01000 Time 0.710 (0.760) Data 0.000 (0.027) Loss 1.7001 (1.4389) Prec@1 53.125 (62.603) Prec@5 79.688 (85.989) Epoch: [21][140/1345], lr: 0.01000 Time 0.846 (0.756) Data 0.000 (0.023) Loss 1.3210 (1.4401) Prec@1 70.312 (62.489) Prec@5 87.500 (86.082) Epoch: [21][160/1345], lr: 0.01000 Time 0.747 (0.756) Data 0.000 (0.020) Loss 1.5375 (1.4427) Prec@1 64.062 (62.471) Prec@5 84.375 (86.073) Epoch: [21][180/1345], lr: 0.01000 Time 0.745 (0.753) Data 0.000 (0.018) Loss 1.3191 (1.4328) Prec@1 62.500 (62.759) Prec@5 90.625 (86.257) Epoch: [21][200/1345], lr: 0.01000 Time 0.708 (0.751) Data 0.000 (0.016) Loss 1.5131 (1.4319) Prec@1 56.250 (62.904) Prec@5 89.062 (86.256) Epoch: [21][220/1345], lr: 0.01000 Time 0.754 (0.750) Data 0.000 (0.015) Loss 1.3148 (1.4384) Prec@1 62.500 (62.811) Prec@5 90.625 (86.206) Epoch: [21][240/1345], lr: 0.01000 Time 0.713 (0.750) Data 0.001 (0.014) Loss 1.8193 (1.4382) Prec@1 60.938 (62.876) Prec@5 73.438 (86.249) Epoch: [21][260/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.013) Loss 1.5991 (1.4419) Prec@1 53.125 (62.710) Prec@5 84.375 (86.213) Epoch: [21][280/1345], lr: 0.01000 Time 0.819 (0.749) Data 0.000 (0.012) Loss 1.4513 (1.4465) Prec@1 60.938 (62.489) Prec@5 84.375 (86.121) Epoch: [21][300/1345], lr: 0.01000 Time 0.737 (0.749) Data 0.000 (0.011) Loss 1.3182 (1.4433) Prec@1 64.062 (62.458) Prec@5 89.062 (86.213) Epoch: [21][320/1345], lr: 0.01000 Time 0.746 (0.749) Data 0.000 (0.010) Loss 1.7542 (1.4471) Prec@1 54.688 (62.388) Prec@5 79.688 (86.093) Epoch: [21][340/1345], lr: 0.01000 Time 0.728 (0.749) Data 0.000 (0.010) Loss 1.5503 (1.4473) Prec@1 60.938 (62.372) Prec@5 85.938 (86.116) Epoch: [21][360/1345], lr: 0.01000 Time 0.711 (0.748) Data 0.000 (0.009) Loss 1.5664 (1.4439) Prec@1 57.812 (62.361) Prec@5 82.812 (86.202) Epoch: [21][380/1345], lr: 0.01000 Time 0.753 (0.747) Data 0.001 (0.009) Loss 1.7783 (1.4435) Prec@1 53.125 (62.283) Prec@5 81.250 (86.184) Epoch: [21][400/1345], lr: 0.01000 Time 0.785 (0.747) Data 0.000 (0.008) Loss 1.9371 (1.4452) Prec@1 51.562 (62.239) Prec@5 79.688 (86.167) Epoch: [21][420/1345], lr: 0.01000 Time 0.724 (0.747) Data 0.001 (0.008) Loss 1.4407 (1.4476) Prec@1 60.938 (62.192) Prec@5 81.250 (86.145) Epoch: [21][440/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.000 (0.008) Loss 1.6923 (1.4494) Prec@1 57.812 (62.174) Prec@5 79.688 (86.108) Epoch: [21][460/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.007) Loss 1.9024 (1.4571) Prec@1 51.562 (61.992) Prec@5 75.000 (86.002) Epoch: [21][480/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.007) Loss 1.2796 (1.4609) Prec@1 59.375 (61.909) Prec@5 89.062 (85.980) Epoch: [21][500/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.007) Loss 1.8246 (1.4660) Prec@1 54.688 (61.767) Prec@5 78.125 (85.928) Epoch: [21][520/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.007) Loss 1.4941 (1.4669) Prec@1 60.938 (61.717) Prec@5 85.938 (85.935) Epoch: [21][540/1345], lr: 0.01000 Time 0.715 (0.744) Data 0.000 (0.006) Loss 1.6277 (1.4720) Prec@1 62.500 (61.665) Prec@5 79.688 (85.839) Epoch: [21][560/1345], lr: 0.01000 Time 0.706 (0.743) Data 0.000 (0.006) Loss 1.5094 (1.4695) Prec@1 62.500 (61.684) Prec@5 82.812 (85.885) Epoch: [21][580/1345], lr: 0.01000 Time 0.740 (0.744) Data 0.000 (0.006) Loss 1.4741 (1.4702) Prec@1 64.062 (61.734) Prec@5 90.625 (85.892) Epoch: [21][600/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.006) Loss 1.2239 (1.4706) Prec@1 65.625 (61.798) Prec@5 89.062 (85.880) Epoch: [21][620/1345], lr: 0.01000 Time 0.770 (0.743) Data 0.000 (0.006) Loss 1.4883 (1.4726) Prec@1 59.375 (61.755) Prec@5 82.812 (85.860) Epoch: [21][640/1345], lr: 0.01000 Time 0.713 (0.742) Data 0.000 (0.005) Loss 1.7060 (1.4729) Prec@1 56.250 (61.710) Prec@5 82.812 (85.847) Epoch: [21][660/1345], lr: 0.01000 Time 0.755 (0.742) Data 0.000 (0.005) Loss 1.2494 (1.4686) Prec@1 65.625 (61.805) Prec@5 89.062 (85.935) Epoch: [21][680/1345], lr: 0.01000 Time 0.753 (0.742) Data 0.001 (0.005) Loss 1.4914 (1.4721) Prec@1 60.938 (61.674) Prec@5 89.062 (85.869) Epoch: [21][700/1345], lr: 0.01000 Time 0.706 (0.742) Data 0.000 (0.005) Loss 1.0637 (1.4707) Prec@1 68.750 (61.686) Prec@5 93.750 (85.884) Epoch: [21][720/1345], lr: 0.01000 Time 0.706 (0.742) Data 0.000 (0.005) Loss 1.6250 (1.4710) Prec@1 57.812 (61.694) Prec@5 82.812 (85.890) Epoch: [21][740/1345], lr: 0.01000 Time 0.785 (0.742) Data 0.000 (0.005) Loss 1.5749 (1.4722) Prec@1 62.500 (61.642) Prec@5 85.938 (85.887) Epoch: [21][760/1345], lr: 0.01000 Time 0.872 (0.742) Data 0.000 (0.005) Loss 1.5563 (1.4727) Prec@1 62.500 (61.644) Prec@5 87.500 (85.878) Epoch: [21][780/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.004) Loss 1.6034 (1.4734) Prec@1 57.812 (61.596) Prec@5 82.812 (85.861) Epoch: [21][800/1345], lr: 0.01000 Time 0.749 (0.741) Data 0.000 (0.004) Loss 1.5138 (1.4753) Prec@1 54.688 (61.534) Prec@5 84.375 (85.842) Epoch: [21][820/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.4256 (1.4758) Prec@1 57.812 (61.545) Prec@5 87.500 (85.825) Epoch: [21][840/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.004) Loss 1.4888 (1.4744) Prec@1 59.375 (61.543) Prec@5 85.938 (85.850) Epoch: [21][860/1345], lr: 0.01000 Time 0.706 (0.741) Data 0.000 (0.004) Loss 1.2314 (1.4767) Prec@1 68.750 (61.486) Prec@5 89.062 (85.820) Epoch: [21][880/1345], lr: 0.01000 Time 0.711 (0.741) Data 0.000 (0.004) Loss 1.4441 (1.4764) Prec@1 60.938 (61.464) Prec@5 82.812 (85.831) Epoch: [21][900/1345], lr: 0.01000 Time 0.783 (0.742) Data 0.000 (0.004) Loss 1.6159 (1.4779) Prec@1 60.938 (61.414) Prec@5 87.500 (85.804) Epoch: [21][920/1345], lr: 0.01000 Time 0.717 (0.742) Data 0.000 (0.004) Loss 1.5362 (1.4772) Prec@1 56.250 (61.436) Prec@5 95.312 (85.827) Epoch: [21][940/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.004) Loss 1.5090 (1.4764) Prec@1 57.812 (61.464) Prec@5 84.375 (85.826) Epoch: [21][960/1345], lr: 0.01000 Time 0.714 (0.742) Data 0.000 (0.004) Loss 1.6954 (1.4752) Prec@1 60.938 (61.468) Prec@5 82.812 (85.832) Epoch: [21][980/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.004) Loss 1.8563 (1.4761) Prec@1 45.312 (61.441) Prec@5 81.250 (85.812) Epoch: [21][1000/1345], lr: 0.01000 Time 0.756 (0.741) Data 0.000 (0.004) Loss 1.3490 (1.4786) Prec@1 67.188 (61.400) Prec@5 85.938 (85.749) Epoch: [21][1020/1345], lr: 0.01000 Time 0.723 (0.742) Data 0.000 (0.004) Loss 1.8004 (1.4814) Prec@1 53.125 (61.346) Prec@5 82.812 (85.714) Epoch: [21][1040/1345], lr: 0.01000 Time 0.860 (0.742) Data 0.000 (0.003) Loss 1.1682 (1.4800) Prec@1 67.188 (61.395) Prec@5 93.750 (85.711) Epoch: [21][1060/1345], lr: 0.01000 Time 0.830 (0.741) Data 0.001 (0.003) Loss 1.7505 (1.4802) Prec@1 53.125 (61.379) Prec@5 78.125 (85.737) Epoch: [21][1080/1345], lr: 0.01000 Time 0.773 (0.741) Data 0.000 (0.003) Loss 1.4683 (1.4809) Prec@1 65.625 (61.342) Prec@5 85.938 (85.739) Epoch: [21][1100/1345], lr: 0.01000 Time 0.727 (0.741) Data 0.000 (0.003) Loss 1.6354 (1.4818) Prec@1 62.500 (61.321) Prec@5 79.688 (85.720) Epoch: [21][1120/1345], lr: 0.01000 Time 0.706 (0.741) Data 0.000 (0.003) Loss 1.4823 (1.4805) Prec@1 59.375 (61.336) Prec@5 87.500 (85.755) Epoch: [21][1140/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.003) Loss 1.5228 (1.4805) Prec@1 68.750 (61.351) Prec@5 84.375 (85.749) Epoch: [21][1160/1345], lr: 0.01000 Time 0.709 (0.741) Data 0.000 (0.003) Loss 1.4003 (1.4809) Prec@1 56.250 (61.337) Prec@5 87.500 (85.746) Epoch: [21][1180/1345], lr: 0.01000 Time 0.739 (0.741) Data 0.000 (0.003) Loss 1.4716 (1.4810) Prec@1 62.500 (61.330) Prec@5 87.500 (85.754) Epoch: [21][1200/1345], lr: 0.01000 Time 0.798 (0.741) Data 0.000 (0.003) Loss 1.4686 (1.4814) Prec@1 59.375 (61.297) Prec@5 85.938 (85.761) Epoch: [21][1220/1345], lr: 0.01000 Time 0.787 (0.741) Data 0.000 (0.003) Loss 1.4144 (1.4820) Prec@1 60.938 (61.286) Prec@5 87.500 (85.751) Epoch: [21][1240/1345], lr: 0.01000 Time 0.782 (0.741) Data 0.000 (0.003) Loss 1.6712 (1.4839) Prec@1 57.812 (61.243) Prec@5 81.250 (85.716) Epoch: [21][1260/1345], lr: 0.01000 Time 0.714 (0.741) Data 0.000 (0.003) Loss 1.4767 (1.4843) Prec@1 60.938 (61.246) Prec@5 87.500 (85.716) Epoch: [21][1280/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.003) Loss 1.1977 (1.4856) Prec@1 71.875 (61.234) Prec@5 90.625 (85.687) Epoch: [21][1300/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.003) Loss 1.4556 (1.4857) Prec@1 59.375 (61.219) Prec@5 87.500 (85.684) Epoch: [21][1320/1345], lr: 0.01000 Time 0.751 (0.741) Data 0.000 (0.003) Loss 1.6399 (1.4858) Prec@1 51.562 (61.221) Prec@5 85.938 (85.700) Epoch: [21][1340/1345], lr: 0.01000 Time 0.719 (0.741) Data 0.000 (0.003) Loss 1.6686 (1.4867) Prec@1 56.250 (61.211) Prec@5 82.812 (85.679) No BN layer Freezing. Test: [0/181] Time 3.173 (3.1734) Loss 2.4220 (2.4220) Prec@1 45.312 (45.312) Prec@5 76.562 (76.562) Test: [20/181] Time 0.931 (0.6126) Loss 2.5411 (2.4014) Prec@1 42.188 (45.312) Prec@5 71.875 (73.810) Test: [40/181] Time 0.934 (0.5331) Loss 2.3180 (2.5080) Prec@1 35.938 (43.636) Prec@5 76.562 (73.476) Test: [60/181] Time 0.595 (0.5000) Loss 2.6920 (2.5500) Prec@1 35.938 (43.289) Prec@5 67.188 (72.823) Test: [80/181] Time 0.878 (0.4907) Loss 2.1714 (2.5579) Prec@1 48.438 (43.113) Prec@5 76.562 (72.647) Test: [100/181] Time 0.828 (0.4839) Loss 2.9961 (2.5647) Prec@1 37.500 (43.116) Prec@5 65.625 (72.587) Test: [120/181] Time 0.753 (0.4791) Loss 2.8338 (2.5676) Prec@1 40.625 (43.066) Prec@5 65.625 (72.585) Test: [140/181] Time 1.186 (0.4807) Loss 2.9814 (2.5719) Prec@1 40.625 (43.129) Prec@5 68.750 (72.750) Test: [160/181] Time 1.085 (0.4771) Loss 3.0059 (2.5518) Prec@1 37.500 (43.333) Prec@5 67.188 (73.195) Testing Results: Prec@1 43.533 Prec@5 73.168 Loss 2.54962 Time 0.4715 No BN layer Freezing. Epoch: [22][0/1345], lr: 0.01000 Time 3.853 (3.853) Data 3.088 (3.088) Loss 1.6043 (1.6043) Prec@1 62.500 (62.500) Prec@5 85.938 (85.938) Epoch: [22][20/1345], lr: 0.01000 Time 0.707 (0.890) Data 0.000 (0.147) Loss 1.4986 (1.4583) Prec@1 65.625 (61.607) Prec@5 82.812 (85.342) Epoch: [22][40/1345], lr: 0.01000 Time 0.807 (0.819) Data 0.001 (0.076) Loss 1.3712 (1.4200) Prec@1 59.375 (62.386) Prec@5 85.938 (86.090) Epoch: [22][60/1345], lr: 0.01000 Time 0.797 (0.787) Data 0.000 (0.051) Loss 1.2177 (1.4078) Prec@1 73.438 (62.987) Prec@5 90.625 (86.219) Epoch: [22][80/1345], lr: 0.01000 Time 0.709 (0.779) Data 0.000 (0.039) Loss 1.6664 (1.4035) Prec@1 60.938 (63.291) Prec@5 79.688 (86.265) Epoch: [22][100/1345], lr: 0.01000 Time 0.737 (0.772) Data 0.000 (0.031) Loss 1.5494 (1.4001) Prec@1 68.750 (63.397) Prec@5 85.938 (86.278) Epoch: [22][120/1345], lr: 0.01000 Time 0.726 (0.767) Data 0.000 (0.026) Loss 1.1546 (1.4043) Prec@1 65.625 (63.107) Prec@5 95.312 (86.389) Epoch: [22][140/1345], lr: 0.01000 Time 0.714 (0.764) Data 0.000 (0.022) Loss 1.2599 (1.4043) Prec@1 64.062 (63.121) Prec@5 87.500 (86.414) Epoch: [22][160/1345], lr: 0.01000 Time 0.715 (0.762) Data 0.001 (0.020) Loss 1.2274 (1.4066) Prec@1 67.188 (63.063) Prec@5 92.188 (86.452) Epoch: [22][180/1345], lr: 0.01000 Time 0.744 (0.761) Data 0.000 (0.018) Loss 1.2244 (1.4067) Prec@1 65.625 (62.966) Prec@5 89.062 (86.447) Epoch: [22][200/1345], lr: 0.01000 Time 0.827 (0.759) Data 0.001 (0.016) Loss 1.8093 (1.4033) Prec@1 57.812 (62.928) Prec@5 78.125 (86.645) Epoch: [22][220/1345], lr: 0.01000 Time 0.795 (0.758) Data 0.000 (0.014) Loss 1.2830 (1.4060) Prec@1 65.625 (62.938) Prec@5 89.062 (86.715) Epoch: [22][240/1345], lr: 0.01000 Time 0.716 (0.758) Data 0.000 (0.013) Loss 1.4734 (1.4221) Prec@1 59.375 (62.494) Prec@5 87.500 (86.443) Epoch: [22][260/1345], lr: 0.01000 Time 0.751 (0.757) Data 0.000 (0.012) Loss 1.1065 (1.4254) Prec@1 68.750 (62.422) Prec@5 90.625 (86.357) Epoch: [22][280/1345], lr: 0.01000 Time 0.710 (0.755) Data 0.000 (0.011) Loss 1.2868 (1.4293) Prec@1 65.625 (62.361) Prec@5 92.188 (86.332) Epoch: [22][300/1345], lr: 0.01000 Time 0.708 (0.753) Data 0.000 (0.011) Loss 1.5894 (1.4337) Prec@1 57.812 (62.220) Prec@5 82.812 (86.254) Epoch: [22][320/1345], lr: 0.01000 Time 0.708 (0.752) Data 0.000 (0.010) Loss 0.9807 (1.4336) Prec@1 73.438 (62.247) Prec@5 92.188 (86.259) Epoch: [22][340/1345], lr: 0.01000 Time 0.708 (0.750) Data 0.000 (0.009) Loss 1.7631 (1.4374) Prec@1 43.750 (62.097) Prec@5 84.375 (86.222) Epoch: [22][360/1345], lr: 0.01000 Time 0.847 (0.749) Data 0.000 (0.009) Loss 1.3626 (1.4409) Prec@1 67.188 (61.972) Prec@5 90.625 (86.176) Epoch: [22][380/1345], lr: 0.01000 Time 0.778 (0.749) Data 0.000 (0.009) Loss 1.1934 (1.4435) Prec@1 67.188 (61.975) Prec@5 92.188 (86.147) Epoch: [22][400/1345], lr: 0.01000 Time 0.707 (0.747) Data 0.000 (0.008) Loss 1.7124 (1.4422) Prec@1 51.562 (62.025) Prec@5 87.500 (86.132) Epoch: [22][420/1345], lr: 0.01000 Time 0.707 (0.747) Data 0.000 (0.008) Loss 1.3530 (1.4449) Prec@1 65.625 (61.932) Prec@5 89.062 (86.127) Epoch: [22][440/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.007) Loss 1.2631 (1.4488) Prec@1 59.375 (61.880) Prec@5 87.500 (86.065) Epoch: [22][460/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.007) Loss 1.4090 (1.4521) Prec@1 65.625 (61.870) Prec@5 85.938 (86.043) Epoch: [22][480/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.007) Loss 1.1013 (1.4530) Prec@1 71.875 (61.883) Prec@5 90.625 (86.006) Epoch: [22][500/1345], lr: 0.01000 Time 0.786 (0.745) Data 0.000 (0.007) Loss 1.2728 (1.4532) Prec@1 64.062 (61.957) Prec@5 87.500 (86.006) Epoch: [22][520/1345], lr: 0.01000 Time 0.754 (0.744) Data 0.000 (0.006) Loss 1.2447 (1.4541) Prec@1 70.312 (61.963) Prec@5 89.062 (86.012) Epoch: [22][540/1345], lr: 0.01000 Time 0.746 (0.744) Data 0.000 (0.006) Loss 1.5070 (1.4522) Prec@1 57.812 (61.943) Prec@5 85.938 (86.105) Epoch: [22][560/1345], lr: 0.01000 Time 0.767 (0.744) Data 0.000 (0.006) Loss 1.4662 (1.4513) Prec@1 64.062 (61.921) Prec@5 81.250 (86.124) Epoch: [22][580/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.006) Loss 1.6503 (1.4500) Prec@1 56.250 (61.992) Prec@5 84.375 (86.131) Epoch: [22][600/1345], lr: 0.01000 Time 0.713 (0.744) Data 0.001 (0.006) Loss 1.6659 (1.4499) Prec@1 53.125 (62.032) Prec@5 82.812 (86.119) Epoch: [22][620/1345], lr: 0.01000 Time 0.790 (0.744) Data 0.000 (0.005) Loss 1.7422 (1.4532) Prec@1 56.250 (61.964) Prec@5 81.250 (86.063) Epoch: [22][640/1345], lr: 0.01000 Time 0.790 (0.743) Data 0.000 (0.005) Loss 1.2451 (1.4534) Prec@1 65.625 (61.969) Prec@5 87.500 (86.064) Epoch: [22][660/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.005) Loss 1.3118 (1.4523) Prec@1 60.938 (61.978) Prec@5 87.500 (86.084) Epoch: [22][680/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.000 (0.005) Loss 1.1481 (1.4527) Prec@1 71.875 (61.998) Prec@5 93.750 (86.089) Epoch: [22][700/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.000 (0.005) Loss 1.4864 (1.4516) Prec@1 54.688 (62.045) Prec@5 87.500 (86.123) Epoch: [22][720/1345], lr: 0.01000 Time 0.712 (0.743) Data 0.000 (0.005) Loss 1.7241 (1.4530) Prec@1 59.375 (62.032) Prec@5 78.125 (86.109) Epoch: [22][740/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.005) Loss 1.6469 (1.4538) Prec@1 56.250 (62.004) Prec@5 82.812 (86.091) Epoch: [22][760/1345], lr: 0.01000 Time 0.765 (0.743) Data 0.000 (0.004) Loss 1.6064 (1.4564) Prec@1 60.938 (61.939) Prec@5 82.812 (86.044) Epoch: [22][780/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.9523 (1.4585) Prec@1 48.438 (61.840) Prec@5 81.250 (86.030) Epoch: [22][800/1345], lr: 0.01000 Time 0.742 (0.742) Data 0.000 (0.004) Loss 1.3124 (1.4589) Prec@1 54.688 (61.780) Prec@5 93.750 (86.058) Epoch: [22][820/1345], lr: 0.01000 Time 0.745 (0.742) Data 0.000 (0.004) Loss 1.3849 (1.4604) Prec@1 67.188 (61.743) Prec@5 85.938 (86.057) Epoch: [22][840/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.3565 (1.4620) Prec@1 67.188 (61.692) Prec@5 87.500 (86.062) Epoch: [22][860/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.7543 (1.4607) Prec@1 54.688 (61.709) Prec@5 79.688 (86.090) Epoch: [22][880/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.3528 (1.4593) Prec@1 65.625 (61.743) Prec@5 87.500 (86.111) Epoch: [22][900/1345], lr: 0.01000 Time 0.861 (0.742) Data 0.001 (0.004) Loss 1.0360 (1.4600) Prec@1 73.438 (61.735) Prec@5 87.500 (86.085) Epoch: [22][920/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.004) Loss 1.5243 (1.4618) Prec@1 64.062 (61.692) Prec@5 89.062 (86.078) Epoch: [22][940/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.1453 (1.4606) Prec@1 64.062 (61.695) Prec@5 92.188 (86.089) Epoch: [22][960/1345], lr: 0.01000 Time 0.795 (0.741) Data 0.000 (0.004) Loss 1.5224 (1.4612) Prec@1 64.062 (61.666) Prec@5 87.500 (86.098) Epoch: [22][980/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.004) Loss 1.7175 (1.4618) Prec@1 51.562 (61.673) Prec@5 85.938 (86.081) Epoch: [22][1000/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.004) Loss 1.6112 (1.4622) Prec@1 54.688 (61.649) Prec@5 81.250 (86.076) Epoch: [22][1020/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.003) Loss 1.4241 (1.4634) Prec@1 68.750 (61.638) Prec@5 87.500 (86.046) Epoch: [22][1040/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.003) Loss 1.6710 (1.4656) Prec@1 54.688 (61.593) Prec@5 81.250 (86.019) Epoch: [22][1060/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.003) Loss 1.0903 (1.4662) Prec@1 70.312 (61.580) Prec@5 93.750 (86.023) Epoch: [22][1080/1345], lr: 0.01000 Time 0.763 (0.741) Data 0.001 (0.003) Loss 1.2487 (1.4665) Prec@1 65.625 (61.565) Prec@5 85.938 (86.008) Epoch: [22][1100/1345], lr: 0.01000 Time 0.749 (0.741) Data 0.000 (0.003) Loss 1.2712 (1.4665) Prec@1 64.062 (61.585) Prec@5 89.062 (86.006) Epoch: [22][1120/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.003) Loss 1.1000 (1.4660) Prec@1 75.000 (61.618) Prec@5 90.625 (86.017) Epoch: [22][1140/1345], lr: 0.01000 Time 0.793 (0.741) Data 0.000 (0.003) Loss 1.1162 (1.4666) Prec@1 68.750 (61.619) Prec@5 95.312 (85.998) Epoch: [22][1160/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.003) Loss 1.5812 (1.4664) Prec@1 64.062 (61.614) Prec@5 79.688 (85.991) Epoch: [22][1180/1345], lr: 0.01000 Time 0.715 (0.740) Data 0.000 (0.003) Loss 1.3883 (1.4670) Prec@1 65.625 (61.582) Prec@5 87.500 (86.001) Epoch: [22][1200/1345], lr: 0.01000 Time 0.707 (0.740) Data 0.000 (0.003) Loss 1.4541 (1.4678) Prec@1 56.250 (61.536) Prec@5 92.188 (85.979) Epoch: [22][1220/1345], lr: 0.01000 Time 0.715 (0.740) Data 0.001 (0.003) Loss 1.6781 (1.4682) Prec@1 57.812 (61.511) Prec@5 84.375 (85.981) Epoch: [22][1240/1345], lr: 0.01000 Time 0.734 (0.740) Data 0.000 (0.003) Loss 1.5821 (1.4688) Prec@1 56.250 (61.459) Prec@5 81.250 (85.974) Epoch: [22][1260/1345], lr: 0.01000 Time 0.708 (0.740) Data 0.000 (0.003) Loss 1.2816 (1.4691) Prec@1 68.750 (61.469) Prec@5 89.062 (85.955) Epoch: [22][1280/1345], lr: 0.01000 Time 0.772 (0.739) Data 0.000 (0.003) Loss 1.1061 (1.4694) Prec@1 70.312 (61.458) Prec@5 93.750 (85.947) Epoch: [22][1300/1345], lr: 0.01000 Time 0.708 (0.739) Data 0.000 (0.003) Loss 1.4080 (1.4688) Prec@1 59.375 (61.480) Prec@5 84.375 (85.938) Epoch: [22][1320/1345], lr: 0.01000 Time 0.705 (0.739) Data 0.000 (0.003) Loss 1.8740 (1.4690) Prec@1 45.312 (61.473) Prec@5 79.688 (85.934) Epoch: [22][1340/1345], lr: 0.01000 Time 0.707 (0.739) Data 0.000 (0.003) Loss 1.2469 (1.4689) Prec@1 73.438 (61.503) Prec@5 90.625 (85.946) No BN layer Freezing. Test: [0/181] Time 3.566 (3.5663) Loss 2.3776 (2.3776) Prec@1 50.000 (50.000) Prec@5 76.562 (76.562) Test: [20/181] Time 0.902 (0.6001) Loss 2.3147 (2.3619) Prec@1 39.062 (45.164) Prec@5 76.562 (74.777) Test: [40/181] Time 0.946 (0.5290) Loss 2.5560 (2.4691) Prec@1 42.188 (43.941) Prec@5 73.438 (73.666) Test: [60/181] Time 0.834 (0.5056) Loss 2.5716 (2.5089) Prec@1 43.750 (43.315) Prec@5 65.625 (72.900) Test: [80/181] Time 0.592 (0.4934) Loss 2.2136 (2.5455) Prec@1 45.312 (42.805) Prec@5 81.250 (72.492) Test: [100/181] Time 0.611 (0.4874) Loss 3.2653 (2.5372) Prec@1 32.812 (42.698) Prec@5 57.812 (72.850) Test: [120/181] Time 0.523 (0.4824) Loss 2.7627 (2.5471) Prec@1 40.625 (42.717) Prec@5 64.062 (72.947) Test: [140/181] Time 0.480 (0.4794) Loss 3.0409 (2.5547) Prec@1 37.500 (42.897) Prec@5 68.750 (73.061) Test: [160/181] Time 0.550 (0.4784) Loss 2.3141 (2.5345) Prec@1 48.438 (43.274) Prec@5 79.688 (73.263) Testing Results: Prec@1 43.455 Prec@5 73.299 Loss 2.51825 Time 0.4759 No BN layer Freezing. Epoch: [23][0/1345], lr: 0.01000 Time 4.623 (4.623) Data 3.894 (3.894) Loss 1.4264 (1.4264) Prec@1 62.500 (62.500) Prec@5 85.938 (85.938) Epoch: [23][20/1345], lr: 0.01000 Time 0.705 (0.921) Data 0.000 (0.186) Loss 1.1267 (1.4247) Prec@1 65.625 (62.574) Prec@5 92.188 (85.863) Epoch: [23][40/1345], lr: 0.01000 Time 0.770 (0.839) Data 0.001 (0.095) Loss 1.4601 (1.3852) Prec@1 67.188 (63.605) Prec@5 85.938 (86.623) Epoch: [23][60/1345], lr: 0.01000 Time 0.709 (0.804) Data 0.000 (0.064) Loss 1.7045 (1.3791) Prec@1 51.562 (63.243) Prec@5 82.812 (86.834) Epoch: [23][80/1345], lr: 0.01000 Time 0.708 (0.786) Data 0.000 (0.048) Loss 1.3900 (1.3801) Prec@1 70.312 (63.561) Prec@5 90.625 (87.076) Epoch: [23][100/1345], lr: 0.01000 Time 0.717 (0.772) Data 0.001 (0.039) Loss 1.2864 (1.3867) Prec@1 60.938 (63.444) Prec@5 89.062 (86.928) Epoch: [23][120/1345], lr: 0.01000 Time 0.708 (0.765) Data 0.000 (0.033) Loss 1.4730 (1.3968) Prec@1 60.938 (63.004) Prec@5 84.375 (86.803) Epoch: [23][140/1345], lr: 0.01000 Time 0.706 (0.761) Data 0.000 (0.028) Loss 1.2418 (1.3946) Prec@1 70.312 (63.010) Prec@5 89.062 (86.924) Epoch: [23][160/1345], lr: 0.01000 Time 0.709 (0.757) Data 0.000 (0.025) Loss 1.0806 (1.3993) Prec@1 70.312 (63.024) Prec@5 92.188 (86.995) Epoch: [23][180/1345], lr: 0.01000 Time 0.769 (0.755) Data 0.000 (0.022) Loss 1.2835 (1.4050) Prec@1 67.188 (62.966) Prec@5 90.625 (86.948) Epoch: [23][200/1345], lr: 0.01000 Time 0.868 (0.755) Data 0.000 (0.020) Loss 1.7689 (1.4054) Prec@1 56.250 (63.083) Prec@5 78.125 (86.964) Epoch: [23][220/1345], lr: 0.01000 Time 0.707 (0.754) Data 0.000 (0.018) Loss 1.6367 (1.4067) Prec@1 51.562 (63.094) Prec@5 87.500 (86.871) Epoch: [23][240/1345], lr: 0.01000 Time 0.707 (0.752) Data 0.000 (0.017) Loss 1.6061 (1.4114) Prec@1 60.938 (62.947) Prec@5 82.812 (86.793) Epoch: [23][260/1345], lr: 0.01000 Time 0.767 (0.751) Data 0.000 (0.015) Loss 1.8821 (1.4107) Prec@1 64.062 (62.907) Prec@5 79.688 (86.800) Epoch: [23][280/1345], lr: 0.01000 Time 0.707 (0.749) Data 0.000 (0.014) Loss 1.1623 (1.4101) Prec@1 75.000 (62.861) Prec@5 92.188 (86.799) Epoch: [23][300/1345], lr: 0.01000 Time 0.762 (0.750) Data 0.000 (0.013) Loss 1.1176 (1.4081) Prec@1 64.062 (62.895) Prec@5 95.312 (86.841) Epoch: [23][320/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.000 (0.013) Loss 1.1620 (1.4080) Prec@1 62.500 (62.938) Prec@5 93.750 (86.853) Epoch: [23][340/1345], lr: 0.01000 Time 0.747 (0.748) Data 0.000 (0.012) Loss 1.1843 (1.4022) Prec@1 64.062 (62.986) Prec@5 87.500 (86.987) Epoch: [23][360/1345], lr: 0.01000 Time 0.711 (0.747) Data 0.000 (0.011) Loss 1.5482 (1.4072) Prec@1 60.938 (62.894) Prec@5 82.812 (86.942) Epoch: [23][380/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.011) Loss 1.4025 (1.4129) Prec@1 62.500 (62.705) Prec@5 87.500 (86.889) Epoch: [23][400/1345], lr: 0.01000 Time 0.706 (0.745) Data 0.000 (0.010) Loss 1.3542 (1.4114) Prec@1 62.500 (62.664) Prec@5 89.062 (86.931) Epoch: [23][420/1345], lr: 0.01000 Time 0.706 (0.745) Data 0.000 (0.010) Loss 1.6025 (1.4120) Prec@1 57.812 (62.656) Prec@5 79.688 (86.895) Epoch: [23][440/1345], lr: 0.01000 Time 0.746 (0.745) Data 0.000 (0.009) Loss 1.1233 (1.4135) Prec@1 67.188 (62.635) Prec@5 92.188 (86.855) Epoch: [23][460/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.009) Loss 1.2568 (1.4151) Prec@1 67.188 (62.575) Prec@5 85.938 (86.832) Epoch: [23][480/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.009) Loss 1.8496 (1.4159) Prec@1 59.375 (62.558) Prec@5 73.438 (86.815) Epoch: [23][500/1345], lr: 0.01000 Time 0.727 (0.744) Data 0.000 (0.008) Loss 1.3646 (1.4165) Prec@1 60.938 (62.506) Prec@5 89.062 (86.858) Epoch: [23][520/1345], lr: 0.01000 Time 0.756 (0.744) Data 0.000 (0.008) Loss 1.9630 (1.4203) Prec@1 54.688 (62.371) Prec@5 76.562 (86.831) Epoch: [23][540/1345], lr: 0.01000 Time 0.727 (0.745) Data 0.000 (0.008) Loss 1.4456 (1.4240) Prec@1 62.500 (62.286) Prec@5 82.812 (86.790) Epoch: [23][560/1345], lr: 0.01000 Time 0.710 (0.745) Data 0.001 (0.007) Loss 1.6197 (1.4238) Prec@1 59.375 (62.280) Prec@5 79.688 (86.759) Epoch: [23][580/1345], lr: 0.01000 Time 0.751 (0.746) Data 0.001 (0.007) Loss 1.7901 (1.4284) Prec@1 51.562 (62.175) Prec@5 84.375 (86.699) Epoch: [23][600/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.007) Loss 1.3424 (1.4292) Prec@1 60.938 (62.131) Prec@5 90.625 (86.689) Epoch: [23][620/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.001 (0.007) Loss 1.5510 (1.4274) Prec@1 60.938 (62.165) Prec@5 81.250 (86.697) Epoch: [23][640/1345], lr: 0.01000 Time 0.765 (0.745) Data 0.001 (0.007) Loss 1.3613 (1.4290) Prec@1 60.938 (62.139) Prec@5 87.500 (86.671) Epoch: [23][660/1345], lr: 0.01000 Time 0.713 (0.745) Data 0.000 (0.006) Loss 1.5390 (1.4313) Prec@1 56.250 (62.034) Prec@5 82.812 (86.647) Epoch: [23][680/1345], lr: 0.01000 Time 0.863 (0.745) Data 0.001 (0.006) Loss 1.1268 (1.4307) Prec@1 67.188 (62.064) Prec@5 92.188 (86.674) Epoch: [23][700/1345], lr: 0.01000 Time 0.897 (0.745) Data 0.000 (0.006) Loss 1.3125 (1.4334) Prec@1 64.062 (61.976) Prec@5 85.938 (86.642) Epoch: [23][720/1345], lr: 0.01000 Time 0.750 (0.746) Data 0.000 (0.006) Loss 1.5154 (1.4361) Prec@1 56.250 (61.934) Prec@5 85.938 (86.575) Epoch: [23][740/1345], lr: 0.01000 Time 0.731 (0.746) Data 0.000 (0.006) Loss 1.2415 (1.4359) Prec@1 68.750 (61.935) Prec@5 84.375 (86.568) Epoch: [23][760/1345], lr: 0.01000 Time 0.728 (0.746) Data 0.000 (0.006) Loss 1.2652 (1.4345) Prec@1 59.375 (61.950) Prec@5 89.062 (86.558) Epoch: [23][780/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.001 (0.005) Loss 1.4594 (1.4337) Prec@1 59.375 (61.970) Prec@5 89.062 (86.550) Epoch: [23][800/1345], lr: 0.01000 Time 0.714 (0.746) Data 0.001 (0.005) Loss 1.3075 (1.4345) Prec@1 64.062 (61.979) Prec@5 89.062 (86.521) Epoch: [23][820/1345], lr: 0.01000 Time 0.868 (0.746) Data 0.001 (0.005) Loss 1.7366 (1.4349) Prec@1 54.688 (61.973) Prec@5 78.125 (86.503) Epoch: [23][840/1345], lr: 0.01000 Time 0.761 (0.747) Data 0.001 (0.005) Loss 1.4522 (1.4344) Prec@1 65.625 (62.008) Prec@5 87.500 (86.499) Epoch: [23][860/1345], lr: 0.01000 Time 0.708 (0.747) Data 0.000 (0.005) Loss 1.5071 (1.4351) Prec@1 59.375 (61.992) Prec@5 87.500 (86.476) Epoch: [23][880/1345], lr: 0.01000 Time 0.709 (0.747) Data 0.000 (0.005) Loss 1.5862 (1.4342) Prec@1 53.125 (62.005) Prec@5 89.062 (86.487) Epoch: [23][900/1345], lr: 0.01000 Time 0.709 (0.747) Data 0.000 (0.005) Loss 1.5504 (1.4353) Prec@1 56.250 (62.013) Prec@5 85.938 (86.458) Epoch: [23][920/1345], lr: 0.01000 Time 0.796 (0.746) Data 0.001 (0.005) Loss 1.2875 (1.4370) Prec@1 62.500 (61.966) Prec@5 89.062 (86.450) Epoch: [23][940/1345], lr: 0.01000 Time 0.709 (0.746) Data 0.000 (0.005) Loss 1.5661 (1.4382) Prec@1 67.188 (61.949) Prec@5 82.812 (86.459) Epoch: [23][960/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.005) Loss 1.7439 (1.4402) Prec@1 60.938 (61.903) Prec@5 84.375 (86.440) Epoch: [23][980/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.004) Loss 1.0826 (1.4420) Prec@1 65.625 (61.869) Prec@5 89.062 (86.423) Epoch: [23][1000/1345], lr: 0.01000 Time 0.846 (0.745) Data 0.000 (0.004) Loss 1.8691 (1.4430) Prec@1 54.688 (61.838) Prec@5 79.688 (86.398) Epoch: [23][1020/1345], lr: 0.01000 Time 0.739 (0.745) Data 0.000 (0.004) Loss 1.2459 (1.4444) Prec@1 76.562 (61.828) Prec@5 87.500 (86.363) Epoch: [23][1040/1345], lr: 0.01000 Time 0.773 (0.745) Data 0.000 (0.004) Loss 1.2468 (1.4439) Prec@1 59.375 (61.838) Prec@5 90.625 (86.362) Epoch: [23][1060/1345], lr: 0.01000 Time 0.775 (0.745) Data 0.001 (0.004) Loss 1.6626 (1.4438) Prec@1 57.812 (61.855) Prec@5 85.938 (86.370) Epoch: [23][1080/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.004) Loss 1.8283 (1.4445) Prec@1 56.250 (61.847) Prec@5 79.688 (86.357) Epoch: [23][1100/1345], lr: 0.01000 Time 0.760 (0.745) Data 0.000 (0.004) Loss 1.5918 (1.4452) Prec@1 64.062 (61.836) Prec@5 84.375 (86.369) Epoch: [23][1120/1345], lr: 0.01000 Time 0.815 (0.745) Data 0.000 (0.004) Loss 1.1097 (1.4450) Prec@1 73.438 (61.832) Prec@5 89.062 (86.384) Epoch: [23][1140/1345], lr: 0.01000 Time 0.816 (0.745) Data 0.000 (0.004) Loss 1.5050 (1.4457) Prec@1 64.062 (61.787) Prec@5 81.250 (86.373) Epoch: [23][1160/1345], lr: 0.01000 Time 0.724 (0.745) Data 0.000 (0.004) Loss 1.3353 (1.4466) Prec@1 67.188 (61.767) Prec@5 90.625 (86.368) Epoch: [23][1180/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.004) Loss 1.3256 (1.4463) Prec@1 65.625 (61.749) Prec@5 90.625 (86.370) Epoch: [23][1200/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.004) Loss 1.4827 (1.4460) Prec@1 62.500 (61.773) Prec@5 84.375 (86.366) Epoch: [23][1220/1345], lr: 0.01000 Time 0.775 (0.744) Data 0.000 (0.004) Loss 1.2434 (1.4461) Prec@1 67.188 (61.765) Prec@5 90.625 (86.376) Epoch: [23][1240/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.004) Loss 1.2480 (1.4468) Prec@1 67.188 (61.751) Prec@5 85.938 (86.377) Epoch: [23][1260/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.004) Loss 1.7878 (1.4479) Prec@1 56.250 (61.716) Prec@5 81.250 (86.350) Epoch: [23][1280/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.003) Loss 1.3214 (1.4485) Prec@1 64.062 (61.688) Prec@5 92.188 (86.349) Epoch: [23][1300/1345], lr: 0.01000 Time 0.745 (0.744) Data 0.000 (0.003) Loss 1.7353 (1.4493) Prec@1 67.188 (61.662) Prec@5 81.250 (86.359) Epoch: [23][1320/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.003) Loss 1.5393 (1.4499) Prec@1 56.250 (61.693) Prec@5 84.375 (86.359) Epoch: [23][1340/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.003) Loss 1.5208 (1.4490) Prec@1 57.812 (61.718) Prec@5 89.062 (86.359) No BN layer Freezing. Test: [0/181] Time 3.405 (3.4052) Loss 2.7610 (2.7610) Prec@1 42.188 (42.188) Prec@5 70.312 (70.312) Test: [20/181] Time 1.009 (0.6119) Loss 2.5349 (2.5163) Prec@1 46.875 (45.089) Prec@5 67.188 (74.405) Test: [40/181] Time 1.274 (0.5425) Loss 3.1866 (2.6655) Prec@1 40.625 (43.521) Prec@5 62.500 (72.409) Test: [60/181] Time 1.056 (0.5136) Loss 2.5387 (2.6923) Prec@1 45.312 (42.777) Prec@5 73.438 (71.311) Test: [80/181] Time 1.172 (0.5018) Loss 2.2288 (2.7068) Prec@1 45.312 (42.361) Prec@5 84.375 (71.258) Test: [100/181] Time 1.127 (0.4939) Loss 2.9113 (2.6887) Prec@1 37.500 (42.358) Prec@5 68.750 (71.674) Test: [120/181] Time 1.094 (0.4890) Loss 3.0784 (2.6973) Prec@1 35.938 (42.291) Prec@5 67.188 (71.539) Test: [140/181] Time 0.975 (0.4839) Loss 3.0479 (2.7089) Prec@1 37.500 (42.132) Prec@5 76.562 (71.609) Test: [160/181] Time 0.892 (0.4794) Loss 2.7680 (2.6998) Prec@1 39.062 (42.158) Prec@5 68.750 (71.749) Testing Results: Prec@1 42.344 Prec@5 71.745 Loss 2.68613 Time 0.4744 No BN layer Freezing. Epoch: [24][0/1345], lr: 0.01000 Time 4.556 (4.556) Data 3.820 (3.820) Loss 1.5127 (1.5127) Prec@1 67.188 (67.188) Prec@5 85.938 (85.938) Epoch: [24][20/1345], lr: 0.01000 Time 0.782 (0.919) Data 0.000 (0.182) Loss 1.4010 (1.3955) Prec@1 65.625 (62.128) Prec@5 89.062 (87.500) Epoch: [24][40/1345], lr: 0.01000 Time 0.715 (0.831) Data 0.000 (0.094) Loss 1.3522 (1.3930) Prec@1 60.938 (62.576) Prec@5 92.188 (87.500) Epoch: [24][60/1345], lr: 0.01000 Time 0.766 (0.801) Data 0.000 (0.063) Loss 1.3628 (1.3655) Prec@1 59.375 (63.038) Prec@5 92.188 (87.782) Epoch: [24][80/1345], lr: 0.01000 Time 0.752 (0.786) Data 0.000 (0.048) Loss 1.0751 (1.3538) Prec@1 73.438 (63.735) Prec@5 93.750 (88.079) Epoch: [24][100/1345], lr: 0.01000 Time 0.710 (0.777) Data 0.000 (0.038) Loss 1.2069 (1.3468) Prec@1 62.500 (63.985) Prec@5 92.188 (88.196) Epoch: [24][120/1345], lr: 0.01000 Time 0.709 (0.771) Data 0.000 (0.032) Loss 1.2750 (1.3658) Prec@1 64.062 (63.649) Prec@5 90.625 (87.810) Epoch: [24][140/1345], lr: 0.01000 Time 0.710 (0.767) Data 0.000 (0.028) Loss 1.7352 (1.3704) Prec@1 50.000 (63.497) Prec@5 81.250 (87.777) Epoch: [24][160/1345], lr: 0.01000 Time 0.711 (0.764) Data 0.000 (0.024) Loss 1.4076 (1.3759) Prec@1 62.500 (63.276) Prec@5 84.375 (87.704) Epoch: [24][180/1345], lr: 0.01000 Time 0.747 (0.760) Data 0.000 (0.022) Loss 1.4208 (1.3769) Prec@1 60.938 (63.260) Prec@5 90.625 (87.707) Epoch: [24][200/1345], lr: 0.01000 Time 0.722 (0.758) Data 0.001 (0.019) Loss 1.3928 (1.3803) Prec@1 60.938 (63.153) Prec@5 87.500 (87.749) Epoch: [24][220/1345], lr: 0.01000 Time 0.707 (0.755) Data 0.000 (0.018) Loss 1.3490 (1.3822) Prec@1 62.500 (63.143) Prec@5 85.938 (87.684) Epoch: [24][240/1345], lr: 0.01000 Time 0.706 (0.753) Data 0.000 (0.016) Loss 0.9613 (1.3872) Prec@1 76.562 (63.064) Prec@5 92.188 (87.532) Epoch: [24][260/1345], lr: 0.01000 Time 0.711 (0.751) Data 0.000 (0.015) Loss 1.2313 (1.3888) Prec@1 67.188 (62.985) Prec@5 90.625 (87.494) Epoch: [24][280/1345], lr: 0.01000 Time 0.706 (0.750) Data 0.000 (0.014) Loss 1.6738 (1.3917) Prec@1 53.125 (63.089) Prec@5 84.375 (87.411) Epoch: [24][300/1345], lr: 0.01000 Time 0.709 (0.749) Data 0.000 (0.013) Loss 1.8323 (1.3968) Prec@1 46.875 (62.874) Prec@5 81.250 (87.318) Epoch: [24][320/1345], lr: 0.01000 Time 0.708 (0.749) Data 0.000 (0.012) Loss 1.4320 (1.3929) Prec@1 65.625 (63.089) Prec@5 85.938 (87.369) Epoch: [24][340/1345], lr: 0.01000 Time 0.713 (0.748) Data 0.001 (0.012) Loss 1.3986 (1.3986) Prec@1 57.812 (62.894) Prec@5 89.062 (87.271) Epoch: [24][360/1345], lr: 0.01000 Time 0.750 (0.747) Data 0.000 (0.011) Loss 1.1216 (1.3972) Prec@1 70.312 (62.950) Prec@5 89.062 (87.275) Epoch: [24][380/1345], lr: 0.01000 Time 0.706 (0.746) Data 0.000 (0.010) Loss 1.2143 (1.3987) Prec@1 68.750 (62.980) Prec@5 85.938 (87.229) Epoch: [24][400/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.010) Loss 1.5021 (1.3969) Prec@1 56.250 (63.081) Prec@5 90.625 (87.255) Epoch: [24][420/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.009) Loss 1.6024 (1.3963) Prec@1 64.062 (63.075) Prec@5 79.688 (87.225) Epoch: [24][440/1345], lr: 0.01000 Time 0.706 (0.744) Data 0.000 (0.009) Loss 0.9680 (1.3998) Prec@1 73.438 (62.968) Prec@5 96.875 (87.220) Epoch: [24][460/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.009) Loss 1.7360 (1.4048) Prec@1 53.125 (62.883) Prec@5 81.250 (87.103) Epoch: [24][480/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.008) Loss 1.6155 (1.4058) Prec@1 53.125 (62.786) Prec@5 79.688 (87.091) Epoch: [24][500/1345], lr: 0.01000 Time 0.713 (0.744) Data 0.000 (0.008) Loss 1.3734 (1.4060) Prec@1 60.938 (62.790) Prec@5 84.375 (87.116) Epoch: [24][520/1345], lr: 0.01000 Time 0.706 (0.744) Data 0.000 (0.008) Loss 1.1255 (1.4076) Prec@1 78.125 (62.740) Prec@5 90.625 (87.086) Epoch: [24][540/1345], lr: 0.01000 Time 0.740 (0.744) Data 0.000 (0.007) Loss 1.2597 (1.4060) Prec@1 70.312 (62.803) Prec@5 87.500 (87.087) Epoch: [24][560/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.007) Loss 1.6131 (1.4098) Prec@1 51.562 (62.684) Prec@5 87.500 (87.015) Epoch: [24][580/1345], lr: 0.01000 Time 0.862 (0.744) Data 0.000 (0.007) Loss 1.7270 (1.4117) Prec@1 50.000 (62.640) Prec@5 87.500 (86.978) Epoch: [24][600/1345], lr: 0.01000 Time 0.786 (0.744) Data 0.000 (0.007) Loss 1.4693 (1.4128) Prec@1 70.312 (62.659) Prec@5 89.062 (86.964) Epoch: [24][620/1345], lr: 0.01000 Time 0.752 (0.744) Data 0.000 (0.007) Loss 1.3331 (1.4153) Prec@1 67.188 (62.570) Prec@5 90.625 (86.929) Epoch: [24][640/1345], lr: 0.01000 Time 0.717 (0.743) Data 0.001 (0.006) Loss 1.7557 (1.4154) Prec@1 56.250 (62.605) Prec@5 81.250 (86.915) Epoch: [24][660/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.006) Loss 1.1870 (1.4154) Prec@1 71.875 (62.651) Prec@5 89.062 (86.914) Epoch: [24][680/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.006) Loss 1.2815 (1.4168) Prec@1 64.062 (62.624) Prec@5 82.812 (86.878) Epoch: [24][700/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.006) Loss 1.2631 (1.4165) Prec@1 67.188 (62.647) Prec@5 82.812 (86.860) Epoch: [24][720/1345], lr: 0.01000 Time 0.791 (0.744) Data 0.000 (0.006) Loss 1.4816 (1.4179) Prec@1 60.938 (62.604) Prec@5 85.938 (86.824) Epoch: [24][740/1345], lr: 0.01000 Time 0.780 (0.743) Data 0.000 (0.006) Loss 1.4176 (1.4184) Prec@1 64.062 (62.580) Prec@5 85.938 (86.796) Epoch: [24][760/1345], lr: 0.01000 Time 0.706 (0.743) Data 0.000 (0.005) Loss 1.0837 (1.4202) Prec@1 71.875 (62.529) Prec@5 87.500 (86.761) Epoch: [24][780/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.005) Loss 1.0901 (1.4216) Prec@1 71.875 (62.506) Prec@5 92.188 (86.726) Epoch: [24][800/1345], lr: 0.01000 Time 0.710 (0.743) Data 0.000 (0.005) Loss 1.4487 (1.4198) Prec@1 57.812 (62.518) Prec@5 87.500 (86.761) Epoch: [24][820/1345], lr: 0.01000 Time 0.775 (0.743) Data 0.001 (0.005) Loss 1.6163 (1.4196) Prec@1 59.375 (62.506) Prec@5 82.812 (86.756) Epoch: [24][840/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.005) Loss 1.2221 (1.4208) Prec@1 65.625 (62.457) Prec@5 93.750 (86.725) Epoch: [24][860/1345], lr: 0.01000 Time 0.715 (0.744) Data 0.001 (0.005) Loss 1.3984 (1.4221) Prec@1 50.000 (62.431) Prec@5 90.625 (86.705) Epoch: [24][880/1345], lr: 0.01000 Time 0.845 (0.744) Data 0.000 (0.005) Loss 1.3649 (1.4253) Prec@1 65.625 (62.340) Prec@5 89.062 (86.642) Epoch: [24][900/1345], lr: 0.01000 Time 0.795 (0.744) Data 0.000 (0.005) Loss 1.8143 (1.4268) Prec@1 56.250 (62.271) Prec@5 81.250 (86.635) Epoch: [24][920/1345], lr: 0.01000 Time 0.772 (0.743) Data 0.000 (0.005) Loss 1.5161 (1.4263) Prec@1 59.375 (62.285) Prec@5 84.375 (86.631) Epoch: [24][940/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.004) Loss 1.0570 (1.4282) Prec@1 75.000 (62.236) Prec@5 90.625 (86.622) Epoch: [24][960/1345], lr: 0.01000 Time 0.706 (0.743) Data 0.000 (0.004) Loss 1.3301 (1.4293) Prec@1 68.750 (62.224) Prec@5 90.625 (86.585) Epoch: [24][980/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.1254 (1.4286) Prec@1 70.312 (62.252) Prec@5 95.312 (86.584) Epoch: [24][1000/1345], lr: 0.01000 Time 0.713 (0.742) Data 0.000 (0.004) Loss 1.4744 (1.4300) Prec@1 67.188 (62.278) Prec@5 87.500 (86.538) Epoch: [24][1020/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.1873 (1.4311) Prec@1 67.188 (62.275) Prec@5 85.938 (86.504) Epoch: [24][1040/1345], lr: 0.01000 Time 0.875 (0.742) Data 0.000 (0.004) Loss 1.3860 (1.4312) Prec@1 67.188 (62.269) Prec@5 85.938 (86.485) Epoch: [24][1060/1345], lr: 0.01000 Time 0.779 (0.742) Data 0.000 (0.004) Loss 1.3428 (1.4319) Prec@1 64.062 (62.247) Prec@5 90.625 (86.503) Epoch: [24][1080/1345], lr: 0.01000 Time 0.706 (0.742) Data 0.000 (0.004) Loss 1.6574 (1.4322) Prec@1 54.688 (62.233) Prec@5 84.375 (86.504) Epoch: [24][1100/1345], lr: 0.01000 Time 0.749 (0.742) Data 0.000 (0.004) Loss 1.6668 (1.4343) Prec@1 54.688 (62.164) Prec@5 79.688 (86.473) Epoch: [24][1120/1345], lr: 0.01000 Time 0.745 (0.742) Data 0.000 (0.004) Loss 1.6973 (1.4357) Prec@1 50.000 (62.150) Prec@5 87.500 (86.443) Epoch: [24][1140/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.0721 (1.4351) Prec@1 78.125 (62.178) Prec@5 90.625 (86.432) Epoch: [24][1160/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.4182 (1.4361) Prec@1 59.375 (62.150) Prec@5 89.062 (86.417) Epoch: [24][1180/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.0708 (1.4355) Prec@1 70.312 (62.168) Prec@5 92.188 (86.420) Epoch: [24][1200/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.6857 (1.4358) Prec@1 56.250 (62.172) Prec@5 84.375 (86.419) Epoch: [24][1220/1345], lr: 0.01000 Time 0.708 (0.741) Data 0.000 (0.004) Loss 1.5545 (1.4373) Prec@1 60.938 (62.156) Prec@5 79.688 (86.383) Epoch: [24][1240/1345], lr: 0.01000 Time 0.715 (0.741) Data 0.000 (0.004) Loss 1.1638 (1.4371) Prec@1 71.875 (62.156) Prec@5 90.625 (86.391) Epoch: [24][1260/1345], lr: 0.01000 Time 0.807 (0.741) Data 0.000 (0.003) Loss 1.4137 (1.4376) Prec@1 64.062 (62.160) Prec@5 87.500 (86.391) Epoch: [24][1280/1345], lr: 0.01000 Time 0.820 (0.741) Data 0.001 (0.003) Loss 1.6127 (1.4378) Prec@1 59.375 (62.166) Prec@5 78.125 (86.392) Epoch: [24][1300/1345], lr: 0.01000 Time 0.755 (0.741) Data 0.000 (0.003) Loss 1.3956 (1.4394) Prec@1 54.688 (62.112) Prec@5 89.062 (86.379) Epoch: [24][1320/1345], lr: 0.01000 Time 0.712 (0.741) Data 0.000 (0.003) Loss 1.0801 (1.4397) Prec@1 73.438 (62.062) Prec@5 93.750 (86.370) Epoch: [24][1340/1345], lr: 0.01000 Time 0.707 (0.741) Data 0.000 (0.003) Loss 1.0447 (1.4395) Prec@1 71.875 (62.055) Prec@5 95.312 (86.376) No BN layer Freezing. Test: [0/181] Time 3.156 (3.1561) Loss 2.9424 (2.9424) Prec@1 39.062 (39.062) Prec@5 76.562 (76.562) Test: [20/181] Time 0.705 (0.5716) Loss 2.4607 (2.5248) Prec@1 42.188 (44.792) Prec@5 75.000 (74.405) Test: [40/181] Time 0.917 (0.5219) Loss 2.9796 (2.6080) Prec@1 46.875 (43.598) Prec@5 68.750 (73.247) Test: [60/181] Time 0.889 (0.4976) Loss 3.0874 (2.6360) Prec@1 37.500 (43.648) Prec@5 60.938 (72.643) Test: [80/181] Time 0.950 (0.4873) Loss 2.6910 (2.6576) Prec@1 42.188 (43.036) Prec@5 79.688 (72.569) Test: [100/181] Time 0.583 (0.4774) Loss 3.4092 (2.6598) Prec@1 37.500 (42.899) Prec@5 65.625 (72.509) Test: [120/181] Time 0.553 (0.4752) Loss 3.0786 (2.6656) Prec@1 43.750 (42.924) Prec@5 62.500 (72.534) Test: [140/181] Time 0.474 (0.4722) Loss 2.9296 (2.6672) Prec@1 29.688 (42.764) Prec@5 68.750 (72.584) Test: [160/181] Time 0.472 (0.4703) Loss 2.8927 (2.6514) Prec@1 40.625 (43.042) Prec@5 71.875 (72.700) Testing Results: Prec@1 43.125 Prec@5 72.613 Loss 2.65840 Time 0.4672 No BN layer Freezing. Epoch: [25][0/1345], lr: 0.01000 Time 4.258 (4.258) Data 3.513 (3.513) Loss 1.8902 (1.8902) Prec@1 54.688 (54.688) Prec@5 79.688 (79.688) Epoch: [25][20/1345], lr: 0.01000 Time 0.827 (0.906) Data 0.000 (0.168) Loss 1.5988 (1.3526) Prec@1 64.062 (62.798) Prec@5 79.688 (87.798) Epoch: [25][40/1345], lr: 0.01000 Time 0.775 (0.822) Data 0.000 (0.086) Loss 1.0796 (1.3652) Prec@1 73.438 (63.072) Prec@5 90.625 (87.729) Epoch: [25][60/1345], lr: 0.01000 Time 0.709 (0.791) Data 0.000 (0.058) Loss 1.5346 (1.3755) Prec@1 59.375 (62.756) Prec@5 85.938 (87.551) Epoch: [25][80/1345], lr: 0.01000 Time 0.710 (0.775) Data 0.000 (0.044) Loss 1.3567 (1.3930) Prec@1 67.188 (62.635) Prec@5 82.812 (86.998) Epoch: [25][100/1345], lr: 0.01000 Time 0.711 (0.768) Data 0.000 (0.035) Loss 1.6952 (1.3927) Prec@1 57.812 (62.949) Prec@5 78.125 (87.005) Epoch: [25][120/1345], lr: 0.01000 Time 0.709 (0.765) Data 0.000 (0.029) Loss 1.4677 (1.3962) Prec@1 60.938 (63.004) Prec@5 87.500 (87.022) Epoch: [25][140/1345], lr: 0.01000 Time 0.732 (0.762) Data 0.001 (0.025) Loss 1.1786 (1.3984) Prec@1 65.625 (63.176) Prec@5 89.062 (86.857) Epoch: [25][160/1345], lr: 0.01000 Time 0.727 (0.760) Data 0.000 (0.022) Loss 1.4868 (1.3964) Prec@1 59.375 (63.043) Prec@5 87.500 (87.024) Epoch: [25][180/1345], lr: 0.01000 Time 0.713 (0.758) Data 0.001 (0.020) Loss 1.3526 (1.3926) Prec@1 67.188 (62.966) Prec@5 85.938 (87.068) Epoch: [25][200/1345], lr: 0.01000 Time 0.747 (0.755) Data 0.000 (0.018) Loss 1.6086 (1.4008) Prec@1 59.375 (62.920) Prec@5 85.938 (86.964) Epoch: [25][220/1345], lr: 0.01000 Time 0.707 (0.754) Data 0.000 (0.016) Loss 1.3443 (1.4009) Prec@1 67.188 (62.938) Prec@5 87.500 (86.977) Epoch: [25][240/1345], lr: 0.01000 Time 0.710 (0.751) Data 0.000 (0.015) Loss 1.1565 (1.4008) Prec@1 64.062 (62.928) Prec@5 90.625 (86.968) Epoch: [25][260/1345], lr: 0.01000 Time 0.782 (0.749) Data 0.000 (0.014) Loss 1.4928 (1.4021) Prec@1 59.375 (62.805) Prec@5 85.938 (86.967) Epoch: [25][280/1345], lr: 0.01000 Time 0.782 (0.749) Data 0.000 (0.013) Loss 1.2104 (1.4008) Prec@1 75.000 (62.895) Prec@5 89.062 (86.961) Epoch: [25][300/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.012) Loss 1.5128 (1.3943) Prec@1 62.500 (63.004) Prec@5 84.375 (86.996) Epoch: [25][320/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.011) Loss 0.9877 (1.3950) Prec@1 67.188 (62.943) Prec@5 92.188 (86.916) Epoch: [25][340/1345], lr: 0.01000 Time 0.711 (0.745) Data 0.000 (0.011) Loss 1.4495 (1.3935) Prec@1 62.500 (63.027) Prec@5 84.375 (86.936) Epoch: [25][360/1345], lr: 0.01000 Time 0.718 (0.745) Data 0.001 (0.010) Loss 1.3341 (1.3958) Prec@1 68.750 (63.063) Prec@5 89.062 (86.959) Epoch: [25][380/1345], lr: 0.01000 Time 0.770 (0.746) Data 0.000 (0.010) Loss 1.4956 (1.3929) Prec@1 59.375 (63.181) Prec@5 84.375 (86.983) Epoch: [25][400/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.009) Loss 1.7258 (1.3966) Prec@1 56.250 (63.096) Prec@5 84.375 (86.982) Epoch: [25][420/1345], lr: 0.01000 Time 0.788 (0.745) Data 0.000 (0.009) Loss 1.5413 (1.4003) Prec@1 53.125 (63.005) Prec@5 90.625 (86.925) Epoch: [25][440/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.008) Loss 1.6613 (1.4019) Prec@1 53.125 (62.975) Prec@5 87.500 (86.891) Epoch: [25][460/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.008) Loss 1.2592 (1.4018) Prec@1 67.188 (62.961) Prec@5 89.062 (86.890) Epoch: [25][480/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.008) Loss 1.2826 (1.4023) Prec@1 68.750 (62.961) Prec@5 89.062 (86.883) Epoch: [25][500/1345], lr: 0.01000 Time 0.737 (0.744) Data 0.000 (0.007) Loss 1.3155 (1.4035) Prec@1 60.938 (62.921) Prec@5 89.062 (86.867) Epoch: [25][520/1345], lr: 0.01000 Time 0.711 (0.743) Data 0.000 (0.007) Loss 1.7277 (1.4007) Prec@1 57.812 (63.031) Prec@5 82.812 (86.900) Epoch: [25][540/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.000 (0.007) Loss 1.4562 (1.4012) Prec@1 57.812 (63.005) Prec@5 82.812 (86.879) Epoch: [25][560/1345], lr: 0.01000 Time 0.710 (0.743) Data 0.000 (0.007) Loss 1.4676 (1.4036) Prec@1 68.750 (62.954) Prec@5 82.812 (86.859) Epoch: [25][580/1345], lr: 0.01000 Time 0.774 (0.743) Data 0.001 (0.006) Loss 1.2510 (1.4025) Prec@1 62.500 (63.011) Prec@5 90.625 (86.884) Epoch: [25][600/1345], lr: 0.01000 Time 0.807 (0.744) Data 0.000 (0.006) Loss 1.4023 (1.4028) Prec@1 67.188 (63.062) Prec@5 85.938 (86.910) Epoch: [25][620/1345], lr: 0.01000 Time 0.793 (0.743) Data 0.000 (0.006) Loss 1.3350 (1.4030) Prec@1 65.625 (63.076) Prec@5 87.500 (86.896) Epoch: [25][640/1345], lr: 0.01000 Time 0.753 (0.744) Data 0.000 (0.006) Loss 1.5123 (1.4061) Prec@1 56.250 (63.039) Prec@5 84.375 (86.832) Epoch: [25][660/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.006) Loss 1.4474 (1.4078) Prec@1 65.625 (63.034) Prec@5 90.625 (86.826) Epoch: [25][680/1345], lr: 0.01000 Time 0.753 (0.744) Data 0.001 (0.006) Loss 1.4454 (1.4099) Prec@1 53.125 (62.947) Prec@5 90.625 (86.784) Epoch: [25][700/1345], lr: 0.01000 Time 0.766 (0.744) Data 0.001 (0.005) Loss 1.9319 (1.4108) Prec@1 46.875 (62.906) Prec@5 78.125 (86.773) Epoch: [25][720/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.005) Loss 1.4575 (1.4115) Prec@1 64.062 (62.875) Prec@5 85.938 (86.776) Epoch: [25][740/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.005) Loss 1.4017 (1.4105) Prec@1 67.188 (62.901) Prec@5 79.688 (86.804) Epoch: [25][760/1345], lr: 0.01000 Time 0.792 (0.744) Data 0.000 (0.005) Loss 1.1267 (1.4105) Prec@1 73.438 (62.925) Prec@5 93.750 (86.827) Epoch: [25][780/1345], lr: 0.01000 Time 0.835 (0.744) Data 0.000 (0.005) Loss 1.8516 (1.4122) Prec@1 57.812 (62.876) Prec@5 78.125 (86.782) Epoch: [25][800/1345], lr: 0.01000 Time 0.716 (0.744) Data 0.000 (0.005) Loss 1.1891 (1.4117) Prec@1 70.312 (62.888) Prec@5 84.375 (86.792) Epoch: [25][820/1345], lr: 0.01000 Time 0.745 (0.744) Data 0.000 (0.005) Loss 1.9089 (1.4125) Prec@1 56.250 (62.839) Prec@5 73.438 (86.796) Epoch: [25][840/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.005) Loss 1.8692 (1.4118) Prec@1 53.125 (62.853) Prec@5 78.125 (86.796) Epoch: [25][860/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.005) Loss 1.3167 (1.4109) Prec@1 62.500 (62.881) Prec@5 89.062 (86.816) Epoch: [25][880/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.004) Loss 1.6543 (1.4118) Prec@1 56.250 (62.890) Prec@5 89.062 (86.794) Epoch: [25][900/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.004) Loss 1.3644 (1.4154) Prec@1 60.938 (62.783) Prec@5 84.375 (86.716) Epoch: [25][920/1345], lr: 0.01000 Time 0.712 (0.744) Data 0.000 (0.004) Loss 1.2563 (1.4141) Prec@1 67.188 (62.810) Prec@5 89.062 (86.738) Epoch: [25][940/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.004) Loss 1.3352 (1.4136) Prec@1 68.750 (62.837) Prec@5 89.062 (86.743) Epoch: [25][960/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.004) Loss 1.8005 (1.4144) Prec@1 57.812 (62.817) Prec@5 82.812 (86.723) Epoch: [25][980/1345], lr: 0.01000 Time 0.746 (0.744) Data 0.000 (0.004) Loss 1.4543 (1.4143) Prec@1 67.188 (62.839) Prec@5 85.938 (86.732) Epoch: [25][1000/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.004) Loss 1.5032 (1.4153) Prec@1 59.375 (62.790) Prec@5 87.500 (86.715) Epoch: [25][1020/1345], lr: 0.01000 Time 0.717 (0.744) Data 0.000 (0.004) Loss 1.8598 (1.4161) Prec@1 53.125 (62.774) Prec@5 81.250 (86.713) Epoch: [25][1040/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.004) Loss 1.6016 (1.4173) Prec@1 51.562 (62.749) Prec@5 87.500 (86.694) Epoch: [25][1060/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.004) Loss 1.4627 (1.4175) Prec@1 60.938 (62.744) Prec@5 89.062 (86.693) Epoch: [25][1080/1345], lr: 0.01000 Time 0.726 (0.744) Data 0.001 (0.004) Loss 1.4501 (1.4181) Prec@1 59.375 (62.753) Prec@5 89.062 (86.698) Epoch: [25][1100/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.004) Loss 1.7274 (1.4192) Prec@1 57.812 (62.760) Prec@5 85.938 (86.685) Epoch: [25][1120/1345], lr: 0.01000 Time 0.844 (0.744) Data 0.001 (0.004) Loss 1.2388 (1.4187) Prec@1 64.062 (62.743) Prec@5 92.188 (86.690) Epoch: [25][1140/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.004) Loss 1.3907 (1.4168) Prec@1 70.312 (62.805) Prec@5 85.938 (86.726) Epoch: [25][1160/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.003) Loss 0.9688 (1.4162) Prec@1 71.875 (62.804) Prec@5 93.750 (86.727) Epoch: [25][1180/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.001 (0.003) Loss 1.6762 (1.4152) Prec@1 50.000 (62.831) Prec@5 85.938 (86.758) Epoch: [25][1200/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.003) Loss 1.3966 (1.4152) Prec@1 59.375 (62.815) Prec@5 84.375 (86.748) Epoch: [25][1220/1345], lr: 0.01000 Time 0.779 (0.744) Data 0.000 (0.003) Loss 1.0262 (1.4158) Prec@1 78.125 (62.803) Prec@5 92.188 (86.736) Epoch: [25][1240/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.003) Loss 1.5126 (1.4166) Prec@1 59.375 (62.772) Prec@5 85.938 (86.738) Epoch: [25][1260/1345], lr: 0.01000 Time 0.729 (0.744) Data 0.001 (0.003) Loss 1.5310 (1.4175) Prec@1 53.125 (62.742) Prec@5 90.625 (86.729) Epoch: [25][1280/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.003) Loss 1.5550 (1.4178) Prec@1 60.938 (62.745) Prec@5 82.812 (86.728) Epoch: [25][1300/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.003) Loss 1.2807 (1.4189) Prec@1 65.625 (62.703) Prec@5 90.625 (86.713) Epoch: [25][1320/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.000 (0.003) Loss 1.4884 (1.4204) Prec@1 64.062 (62.663) Prec@5 85.938 (86.706) Epoch: [25][1340/1345], lr: 0.01000 Time 0.711 (0.744) Data 0.000 (0.003) Loss 1.6302 (1.4208) Prec@1 57.812 (62.664) Prec@5 78.125 (86.686) No BN layer Freezing. Test: [0/181] Time 2.865 (2.8648) Loss 2.6727 (2.6727) Prec@1 46.875 (46.875) Prec@5 76.562 (76.562) Test: [20/181] Time 0.843 (0.5928) Loss 2.6661 (2.4991) Prec@1 46.875 (44.866) Prec@5 73.438 (74.554) Test: [40/181] Time 1.240 (0.5318) Loss 2.8129 (2.5680) Prec@1 51.562 (43.674) Prec@5 68.750 (73.209) Test: [60/181] Time 1.197 (0.5107) Loss 2.4181 (2.5990) Prec@1 46.875 (42.905) Prec@5 71.875 (72.362) Test: [80/181] Time 1.207 (0.4962) Loss 2.2905 (2.6552) Prec@1 48.438 (42.091) Prec@5 81.250 (71.759) Test: [100/181] Time 1.126 (0.4885) Loss 3.0657 (2.6548) Prec@1 39.062 (41.863) Prec@5 64.062 (71.813) Test: [120/181] Time 1.055 (0.4828) Loss 2.8806 (2.6532) Prec@1 42.188 (41.878) Prec@5 60.938 (71.759) Test: [140/181] Time 1.063 (0.4792) Loss 3.1114 (2.6604) Prec@1 37.500 (41.689) Prec@5 68.750 (71.897) Test: [160/181] Time 1.112 (0.4776) Loss 2.7648 (2.6404) Prec@1 43.750 (42.032) Prec@5 71.875 (72.156) Testing Results: Prec@1 42.405 Prec@5 72.387 Loss 2.62581 Time 0.4737 No BN layer Freezing. Epoch: [26][0/1345], lr: 0.01000 Time 4.240 (4.240) Data 3.487 (3.487) Loss 1.4469 (1.4469) Prec@1 60.938 (60.938) Prec@5 87.500 (87.500) Epoch: [26][20/1345], lr: 0.01000 Time 0.709 (0.904) Data 0.000 (0.166) Loss 1.2176 (1.2730) Prec@1 71.875 (66.369) Prec@5 89.062 (89.211) Epoch: [26][40/1345], lr: 0.01000 Time 0.727 (0.820) Data 0.001 (0.085) Loss 1.2698 (1.3189) Prec@1 65.625 (64.939) Prec@5 89.062 (87.462) Epoch: [26][60/1345], lr: 0.01000 Time 0.711 (0.796) Data 0.000 (0.058) Loss 1.4915 (1.3302) Prec@1 62.500 (64.703) Prec@5 82.812 (87.577) Epoch: [26][80/1345], lr: 0.01000 Time 0.707 (0.779) Data 0.000 (0.043) Loss 1.3369 (1.3191) Prec@1 65.625 (64.873) Prec@5 87.500 (87.924) Epoch: [26][100/1345], lr: 0.01000 Time 0.714 (0.771) Data 0.001 (0.035) Loss 1.3260 (1.3284) Prec@1 67.188 (64.728) Prec@5 89.062 (87.918) Epoch: [26][120/1345], lr: 0.01000 Time 0.706 (0.767) Data 0.000 (0.029) Loss 1.3136 (1.3386) Prec@1 67.188 (64.347) Prec@5 85.938 (87.900) Epoch: [26][140/1345], lr: 0.01000 Time 0.775 (0.763) Data 0.000 (0.025) Loss 1.3384 (1.3453) Prec@1 59.375 (64.173) Prec@5 89.062 (87.766) Epoch: [26][160/1345], lr: 0.01000 Time 0.716 (0.760) Data 0.001 (0.022) Loss 1.4283 (1.3484) Prec@1 65.625 (64.053) Prec@5 84.375 (87.762) Epoch: [26][180/1345], lr: 0.01000 Time 0.860 (0.759) Data 0.000 (0.020) Loss 1.1001 (1.3485) Prec@1 68.750 (64.080) Prec@5 90.625 (87.828) Epoch: [26][200/1345], lr: 0.01000 Time 0.836 (0.757) Data 0.000 (0.018) Loss 1.3729 (1.3489) Prec@1 59.375 (63.907) Prec@5 92.188 (87.935) Epoch: [26][220/1345], lr: 0.01000 Time 0.758 (0.756) Data 0.000 (0.016) Loss 1.2312 (1.3444) Prec@1 70.312 (64.077) Prec@5 89.062 (87.960) Epoch: [26][240/1345], lr: 0.01000 Time 0.710 (0.755) Data 0.000 (0.015) Loss 1.6042 (1.3543) Prec@1 59.375 (63.836) Prec@5 87.500 (87.934) Epoch: [26][260/1345], lr: 0.01000 Time 0.712 (0.754) Data 0.000 (0.014) Loss 1.5371 (1.3564) Prec@1 60.938 (63.823) Prec@5 84.375 (87.847) Epoch: [26][280/1345], lr: 0.01000 Time 0.707 (0.752) Data 0.000 (0.013) Loss 1.3744 (1.3587) Prec@1 62.500 (63.768) Prec@5 89.062 (87.772) Epoch: [26][300/1345], lr: 0.01000 Time 0.706 (0.751) Data 0.000 (0.012) Loss 1.0488 (1.3542) Prec@1 73.438 (63.959) Prec@5 92.188 (87.832) Epoch: [26][320/1345], lr: 0.01000 Time 0.707 (0.751) Data 0.000 (0.011) Loss 1.1225 (1.3573) Prec@1 70.312 (63.834) Prec@5 93.750 (87.768) Epoch: [26][340/1345], lr: 0.01000 Time 0.860 (0.751) Data 0.000 (0.011) Loss 1.2544 (1.3602) Prec@1 62.500 (63.742) Prec@5 89.062 (87.711) Epoch: [26][360/1345], lr: 0.01000 Time 0.846 (0.751) Data 0.000 (0.010) Loss 0.9820 (1.3601) Prec@1 68.750 (63.794) Prec@5 92.188 (87.699) Epoch: [26][380/1345], lr: 0.01000 Time 0.752 (0.750) Data 0.000 (0.010) Loss 1.3284 (1.3659) Prec@1 54.688 (63.632) Prec@5 92.188 (87.652) Epoch: [26][400/1345], lr: 0.01000 Time 0.743 (0.750) Data 0.000 (0.009) Loss 1.0903 (1.3647) Prec@1 62.500 (63.708) Prec@5 93.750 (87.652) Epoch: [26][420/1345], lr: 0.01000 Time 0.774 (0.751) Data 0.000 (0.009) Loss 1.4627 (1.3673) Prec@1 70.312 (63.651) Prec@5 82.812 (87.619) Epoch: [26][440/1345], lr: 0.01000 Time 0.709 (0.750) Data 0.000 (0.008) Loss 1.2374 (1.3698) Prec@1 65.625 (63.595) Prec@5 85.938 (87.560) Epoch: [26][460/1345], lr: 0.01000 Time 0.711 (0.749) Data 0.000 (0.008) Loss 1.6255 (1.3728) Prec@1 56.250 (63.585) Prec@5 85.938 (87.510) Epoch: [26][480/1345], lr: 0.01000 Time 0.798 (0.749) Data 0.001 (0.008) Loss 1.1481 (1.3773) Prec@1 65.625 (63.432) Prec@5 93.750 (87.455) Epoch: [26][500/1345], lr: 0.01000 Time 0.710 (0.749) Data 0.000 (0.007) Loss 1.4765 (1.3795) Prec@1 67.188 (63.323) Prec@5 85.938 (87.425) Epoch: [26][520/1345], lr: 0.01000 Time 0.709 (0.748) Data 0.000 (0.007) Loss 1.1500 (1.3764) Prec@1 67.188 (63.424) Prec@5 92.188 (87.449) Epoch: [26][540/1345], lr: 0.01000 Time 0.751 (0.748) Data 0.000 (0.007) Loss 1.4613 (1.3750) Prec@1 56.250 (63.450) Prec@5 85.938 (87.491) Epoch: [26][560/1345], lr: 0.01000 Time 0.709 (0.747) Data 0.000 (0.007) Loss 1.5176 (1.3742) Prec@1 59.375 (63.464) Prec@5 81.250 (87.483) Epoch: [26][580/1345], lr: 0.01000 Time 0.711 (0.746) Data 0.000 (0.006) Loss 1.9459 (1.3747) Prec@1 62.500 (63.468) Prec@5 78.125 (87.452) Epoch: [26][600/1345], lr: 0.01000 Time 0.769 (0.747) Data 0.001 (0.006) Loss 1.6139 (1.3756) Prec@1 56.250 (63.491) Prec@5 84.375 (87.464) Epoch: [26][620/1345], lr: 0.01000 Time 0.843 (0.747) Data 0.000 (0.006) Loss 1.2404 (1.3777) Prec@1 70.312 (63.476) Prec@5 87.500 (87.445) Epoch: [26][640/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.006) Loss 1.5885 (1.3802) Prec@1 57.812 (63.392) Prec@5 84.375 (87.407) Epoch: [26][660/1345], lr: 0.01000 Time 0.710 (0.747) Data 0.000 (0.006) Loss 1.4840 (1.3787) Prec@1 57.812 (63.438) Prec@5 79.688 (87.417) Epoch: [26][680/1345], lr: 0.01000 Time 0.716 (0.746) Data 0.001 (0.006) Loss 1.1629 (1.3780) Prec@1 64.062 (63.411) Prec@5 89.062 (87.431) Epoch: [26][700/1345], lr: 0.01000 Time 0.708 (0.746) Data 0.000 (0.005) Loss 1.6445 (1.3809) Prec@1 57.812 (63.351) Prec@5 87.500 (87.386) Epoch: [26][720/1345], lr: 0.01000 Time 0.776 (0.746) Data 0.000 (0.005) Loss 1.4945 (1.3817) Prec@1 62.500 (63.278) Prec@5 85.938 (87.392) Epoch: [26][740/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.005) Loss 1.5435 (1.3846) Prec@1 62.500 (63.209) Prec@5 79.688 (87.329) Epoch: [26][760/1345], lr: 0.01000 Time 0.739 (0.745) Data 0.000 (0.005) Loss 1.7203 (1.3852) Prec@1 54.688 (63.190) Prec@5 82.812 (87.336) Epoch: [26][780/1345], lr: 0.01000 Time 0.763 (0.745) Data 0.000 (0.005) Loss 1.5987 (1.3891) Prec@1 57.812 (63.106) Prec@5 84.375 (87.270) Epoch: [26][800/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.005) Loss 1.4334 (1.3921) Prec@1 64.062 (63.068) Prec@5 85.938 (87.202) Epoch: [26][820/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.005) Loss 1.7168 (1.3930) Prec@1 57.812 (63.054) Prec@5 81.250 (87.173) Epoch: [26][840/1345], lr: 0.01000 Time 0.829 (0.745) Data 0.000 (0.005) Loss 1.5758 (1.3949) Prec@1 62.500 (62.979) Prec@5 81.250 (87.158) Epoch: [26][860/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.004) Loss 1.2554 (1.3953) Prec@1 68.750 (62.972) Prec@5 89.062 (87.148) Epoch: [26][880/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.000 (0.004) Loss 1.4978 (1.3954) Prec@1 60.938 (62.965) Prec@5 92.188 (87.136) Epoch: [26][900/1345], lr: 0.01000 Time 0.715 (0.745) Data 0.000 (0.004) Loss 1.3820 (1.3952) Prec@1 59.375 (62.968) Prec@5 89.062 (87.122) Epoch: [26][920/1345], lr: 0.01000 Time 0.743 (0.745) Data 0.000 (0.004) Loss 1.5157 (1.3976) Prec@1 62.500 (62.931) Prec@5 87.500 (87.079) Epoch: [26][940/1345], lr: 0.01000 Time 0.780 (0.745) Data 0.000 (0.004) Loss 1.6333 (1.3977) Prec@1 53.125 (62.937) Prec@5 79.688 (87.077) Epoch: [26][960/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.004) Loss 1.6770 (1.4003) Prec@1 59.375 (62.910) Prec@5 84.375 (87.032) Epoch: [26][980/1345], lr: 0.01000 Time 0.745 (0.745) Data 0.000 (0.004) Loss 1.2524 (1.4016) Prec@1 65.625 (62.884) Prec@5 89.062 (87.016) Epoch: [26][1000/1345], lr: 0.01000 Time 0.781 (0.745) Data 0.000 (0.004) Loss 2.1060 (1.4030) Prec@1 45.312 (62.856) Prec@5 68.750 (86.988) Epoch: [26][1020/1345], lr: 0.01000 Time 0.730 (0.745) Data 0.000 (0.004) Loss 1.4315 (1.4036) Prec@1 65.625 (62.850) Prec@5 82.812 (86.980) Epoch: [26][1040/1345], lr: 0.01000 Time 0.712 (0.745) Data 0.000 (0.004) Loss 1.5625 (1.4039) Prec@1 54.688 (62.845) Prec@5 85.938 (86.969) Epoch: [26][1060/1345], lr: 0.01000 Time 0.709 (0.745) Data 0.000 (0.004) Loss 1.8093 (1.4051) Prec@1 46.875 (62.824) Prec@5 82.812 (86.940) Epoch: [26][1080/1345], lr: 0.01000 Time 0.838 (0.745) Data 0.000 (0.004) Loss 1.6549 (1.4054) Prec@1 51.562 (62.788) Prec@5 82.812 (86.941) Epoch: [26][1100/1345], lr: 0.01000 Time 0.776 (0.745) Data 0.000 (0.004) Loss 1.2904 (1.4057) Prec@1 62.500 (62.794) Prec@5 89.062 (86.929) Epoch: [26][1120/1345], lr: 0.01000 Time 0.757 (0.744) Data 0.000 (0.004) Loss 1.2163 (1.4060) Prec@1 65.625 (62.804) Prec@5 90.625 (86.910) Epoch: [26][1140/1345], lr: 0.01000 Time 0.724 (0.744) Data 0.001 (0.003) Loss 1.4098 (1.4058) Prec@1 56.250 (62.811) Prec@5 90.625 (86.889) Epoch: [26][1160/1345], lr: 0.01000 Time 0.739 (0.744) Data 0.000 (0.003) Loss 1.4734 (1.4068) Prec@1 59.375 (62.787) Prec@5 84.375 (86.880) Epoch: [26][1180/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.001 (0.003) Loss 1.6535 (1.4081) Prec@1 59.375 (62.777) Prec@5 82.812 (86.865) Epoch: [26][1200/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.003) Loss 1.2793 (1.4089) Prec@1 60.938 (62.754) Prec@5 87.500 (86.842) Epoch: [26][1220/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.003) Loss 1.1882 (1.4090) Prec@1 64.062 (62.721) Prec@5 89.062 (86.842) Epoch: [26][1240/1345], lr: 0.01000 Time 0.755 (0.743) Data 0.000 (0.003) Loss 1.1174 (1.4098) Prec@1 64.062 (62.708) Prec@5 90.625 (86.830) Epoch: [26][1260/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.003) Loss 1.6456 (1.4100) Prec@1 59.375 (62.716) Prec@5 84.375 (86.821) Epoch: [26][1280/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.003) Loss 1.3108 (1.4113) Prec@1 60.938 (62.710) Prec@5 92.188 (86.799) Epoch: [26][1300/1345], lr: 0.01000 Time 0.758 (0.744) Data 0.000 (0.003) Loss 1.6040 (1.4124) Prec@1 57.812 (62.698) Prec@5 85.938 (86.777) Epoch: [26][1320/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.003) Loss 1.2692 (1.4139) Prec@1 64.062 (62.687) Prec@5 89.062 (86.757) Epoch: [26][1340/1345], lr: 0.01000 Time 0.710 (0.743) Data 0.000 (0.003) Loss 1.6446 (1.4148) Prec@1 51.562 (62.667) Prec@5 84.375 (86.744) No BN layer Freezing. Test: [0/181] Time 3.399 (3.3987) Loss 2.7500 (2.7500) Prec@1 40.625 (40.625) Prec@5 73.438 (73.438) Test: [20/181] Time 1.024 (0.5890) Loss 2.3098 (2.4715) Prec@1 46.875 (44.940) Prec@5 79.688 (75.223) Test: [40/181] Time 1.097 (0.5304) Loss 2.4889 (2.5656) Prec@1 46.875 (43.331) Prec@5 78.125 (73.590) Test: [60/181] Time 1.051 (0.5057) Loss 2.6219 (2.5767) Prec@1 42.188 (42.905) Prec@5 68.750 (72.439) Test: [80/181] Time 1.019 (0.4924) Loss 2.5374 (2.6025) Prec@1 43.750 (42.168) Prec@5 82.812 (72.454) Test: [100/181] Time 0.949 (0.4851) Loss 3.1034 (2.5871) Prec@1 34.375 (42.064) Prec@5 64.062 (72.494) Test: [120/181] Time 0.932 (0.4798) Loss 3.0825 (2.5989) Prec@1 37.500 (42.265) Prec@5 65.625 (72.133) Test: [140/181] Time 0.709 (0.4764) Loss 3.1926 (2.6008) Prec@1 32.812 (42.188) Prec@5 65.625 (72.263) Test: [160/181] Time 0.577 (0.4745) Loss 2.4264 (2.5794) Prec@1 46.875 (42.469) Prec@5 76.562 (72.632) Testing Results: Prec@1 42.639 Prec@5 72.804 Loss 2.57005 Time 0.4723 No BN layer Freezing. Epoch: [27][0/1345], lr: 0.01000 Time 4.467 (4.467) Data 3.733 (3.733) Loss 1.3983 (1.3983) Prec@1 56.250 (56.250) Prec@5 87.500 (87.500) Epoch: [27][20/1345], lr: 0.01000 Time 0.773 (0.926) Data 0.000 (0.178) Loss 1.0160 (1.3294) Prec@1 79.688 (64.435) Prec@5 90.625 (87.723) Epoch: [27][40/1345], lr: 0.01000 Time 0.728 (0.835) Data 0.001 (0.091) Loss 0.8535 (1.2860) Prec@1 71.875 (64.748) Prec@5 95.312 (88.681) Epoch: [27][60/1345], lr: 0.01000 Time 0.770 (0.807) Data 0.000 (0.062) Loss 1.1843 (1.2788) Prec@1 78.125 (65.932) Prec@5 82.812 (88.525) Epoch: [27][80/1345], lr: 0.01000 Time 0.845 (0.795) Data 0.001 (0.046) Loss 1.2744 (1.2829) Prec@1 62.500 (65.837) Prec@5 85.938 (88.580) Epoch: [27][100/1345], lr: 0.01000 Time 0.710 (0.783) Data 0.000 (0.037) Loss 1.4862 (1.3064) Prec@1 64.062 (65.486) Prec@5 81.250 (88.165) Epoch: [27][120/1345], lr: 0.01000 Time 0.750 (0.774) Data 0.000 (0.031) Loss 1.4585 (1.3164) Prec@1 57.812 (65.496) Prec@5 87.500 (88.004) Epoch: [27][140/1345], lr: 0.01000 Time 0.706 (0.770) Data 0.000 (0.027) Loss 1.2814 (1.3184) Prec@1 67.188 (65.437) Prec@5 85.938 (87.832) Epoch: [27][160/1345], lr: 0.01000 Time 0.710 (0.764) Data 0.000 (0.024) Loss 1.6865 (1.3276) Prec@1 53.125 (65.140) Prec@5 85.938 (87.626) Epoch: [27][180/1345], lr: 0.01000 Time 0.810 (0.761) Data 0.000 (0.021) Loss 1.4633 (1.3296) Prec@1 67.188 (65.116) Prec@5 82.812 (87.569) Epoch: [27][200/1345], lr: 0.01000 Time 0.784 (0.758) Data 0.000 (0.019) Loss 1.2760 (1.3391) Prec@1 62.500 (64.995) Prec@5 87.500 (87.500) Epoch: [27][220/1345], lr: 0.01000 Time 0.710 (0.755) Data 0.000 (0.017) Loss 1.6436 (1.3374) Prec@1 59.375 (64.953) Prec@5 84.375 (87.592) Epoch: [27][240/1345], lr: 0.01000 Time 0.732 (0.754) Data 0.000 (0.016) Loss 1.5719 (1.3448) Prec@1 60.938 (64.659) Prec@5 87.500 (87.552) Epoch: [27][260/1345], lr: 0.01000 Time 0.722 (0.753) Data 0.001 (0.015) Loss 1.2846 (1.3503) Prec@1 65.625 (64.494) Prec@5 92.188 (87.518) Epoch: [27][280/1345], lr: 0.01000 Time 0.707 (0.751) Data 0.000 (0.014) Loss 1.3836 (1.3476) Prec@1 65.625 (64.468) Prec@5 89.062 (87.633) Epoch: [27][300/1345], lr: 0.01000 Time 0.758 (0.751) Data 0.000 (0.013) Loss 1.1136 (1.3436) Prec@1 70.312 (64.576) Prec@5 90.625 (87.656) Epoch: [27][320/1345], lr: 0.01000 Time 0.707 (0.750) Data 0.000 (0.012) Loss 1.0912 (1.3449) Prec@1 67.188 (64.564) Prec@5 90.625 (87.714) Epoch: [27][340/1345], lr: 0.01000 Time 0.828 (0.750) Data 0.000 (0.011) Loss 1.6723 (1.3439) Prec@1 57.812 (64.539) Prec@5 82.812 (87.757) Epoch: [27][360/1345], lr: 0.01000 Time 0.864 (0.749) Data 0.001 (0.011) Loss 1.6426 (1.3394) Prec@1 60.938 (64.504) Prec@5 79.688 (87.868) Epoch: [27][380/1345], lr: 0.01000 Time 0.744 (0.749) Data 0.001 (0.010) Loss 1.2743 (1.3434) Prec@1 70.312 (64.399) Prec@5 84.375 (87.762) Epoch: [27][400/1345], lr: 0.01000 Time 0.706 (0.749) Data 0.000 (0.010) Loss 1.4295 (1.3469) Prec@1 65.625 (64.331) Prec@5 84.375 (87.742) Epoch: [27][420/1345], lr: 0.01000 Time 0.707 (0.748) Data 0.000 (0.009) Loss 1.4897 (1.3454) Prec@1 65.625 (64.434) Prec@5 84.375 (87.789) Epoch: [27][440/1345], lr: 0.01000 Time 0.771 (0.748) Data 0.000 (0.009) Loss 1.4756 (1.3452) Prec@1 62.500 (64.477) Prec@5 87.500 (87.791) Epoch: [27][460/1345], lr: 0.01000 Time 0.804 (0.748) Data 0.000 (0.009) Loss 1.4598 (1.3502) Prec@1 60.938 (64.354) Prec@5 84.375 (87.714) Epoch: [27][480/1345], lr: 0.01000 Time 0.712 (0.747) Data 0.000 (0.008) Loss 1.2251 (1.3498) Prec@1 67.188 (64.397) Prec@5 92.188 (87.744) Epoch: [27][500/1345], lr: 0.01000 Time 0.729 (0.747) Data 0.000 (0.008) Loss 1.3671 (1.3517) Prec@1 71.875 (64.384) Prec@5 87.500 (87.656) Epoch: [27][520/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.008) Loss 1.0172 (1.3497) Prec@1 64.062 (64.392) Prec@5 100.000 (87.701) Epoch: [27][540/1345], lr: 0.01000 Time 0.707 (0.746) Data 0.000 (0.007) Loss 1.1402 (1.3499) Prec@1 67.188 (64.386) Prec@5 87.500 (87.705) Epoch: [27][560/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.007) Loss 1.5451 (1.3511) Prec@1 62.500 (64.344) Prec@5 76.562 (87.675) Epoch: [27][580/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.001 (0.007) Loss 1.1943 (1.3561) Prec@1 68.750 (64.253) Prec@5 90.625 (87.586) Epoch: [27][600/1345], lr: 0.01000 Time 0.739 (0.745) Data 0.000 (0.007) Loss 1.4278 (1.3600) Prec@1 60.938 (64.122) Prec@5 87.500 (87.531) Epoch: [27][620/1345], lr: 0.01000 Time 0.722 (0.745) Data 0.001 (0.006) Loss 1.7983 (1.3630) Prec@1 56.250 (64.075) Prec@5 79.688 (87.490) Epoch: [27][640/1345], lr: 0.01000 Time 0.714 (0.744) Data 0.001 (0.006) Loss 1.2388 (1.3652) Prec@1 70.312 (64.021) Prec@5 92.188 (87.463) Epoch: [27][660/1345], lr: 0.01000 Time 0.742 (0.744) Data 0.000 (0.006) Loss 1.0830 (1.3658) Prec@1 71.875 (63.968) Prec@5 92.188 (87.460) Epoch: [27][680/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.006) Loss 1.0419 (1.3659) Prec@1 73.438 (63.984) Prec@5 85.938 (87.440) Epoch: [27][700/1345], lr: 0.01000 Time 0.787 (0.744) Data 0.000 (0.006) Loss 1.6335 (1.3662) Prec@1 57.812 (64.009) Prec@5 84.375 (87.400) Epoch: [27][720/1345], lr: 0.01000 Time 0.783 (0.744) Data 0.000 (0.006) Loss 1.1681 (1.3662) Prec@1 67.188 (63.984) Prec@5 92.188 (87.428) Epoch: [27][740/1345], lr: 0.01000 Time 0.748 (0.744) Data 0.000 (0.005) Loss 1.6518 (1.3673) Prec@1 54.688 (63.949) Prec@5 82.812 (87.376) Epoch: [27][760/1345], lr: 0.01000 Time 0.776 (0.744) Data 0.000 (0.005) Loss 1.2140 (1.3691) Prec@1 67.188 (63.900) Prec@5 90.625 (87.360) Epoch: [27][780/1345], lr: 0.01000 Time 0.751 (0.744) Data 0.000 (0.005) Loss 1.8771 (1.3707) Prec@1 54.688 (63.864) Prec@5 81.250 (87.334) Epoch: [27][800/1345], lr: 0.01000 Time 0.709 (0.744) Data 0.000 (0.005) Loss 1.3879 (1.3722) Prec@1 60.938 (63.811) Prec@5 85.938 (87.303) Epoch: [27][820/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.001 (0.005) Loss 1.1883 (1.3720) Prec@1 65.625 (63.802) Prec@5 92.188 (87.321) Epoch: [27][840/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.005) Loss 1.7084 (1.3717) Prec@1 54.688 (63.823) Prec@5 79.688 (87.335) Epoch: [27][860/1345], lr: 0.01000 Time 0.855 (0.744) Data 0.000 (0.005) Loss 1.8955 (1.3739) Prec@1 48.438 (63.769) Prec@5 78.125 (87.306) Epoch: [27][880/1345], lr: 0.01000 Time 0.845 (0.744) Data 0.000 (0.005) Loss 1.6348 (1.3757) Prec@1 50.000 (63.678) Prec@5 81.250 (87.298) Epoch: [27][900/1345], lr: 0.01000 Time 0.742 (0.744) Data 0.000 (0.005) Loss 1.1586 (1.3775) Prec@1 65.625 (63.650) Prec@5 90.625 (87.275) Epoch: [27][920/1345], lr: 0.01000 Time 0.753 (0.744) Data 0.001 (0.004) Loss 1.7445 (1.3794) Prec@1 53.125 (63.613) Prec@5 82.812 (87.273) Epoch: [27][940/1345], lr: 0.01000 Time 0.720 (0.744) Data 0.000 (0.004) Loss 1.7538 (1.3808) Prec@1 57.812 (63.566) Prec@5 81.250 (87.254) Epoch: [27][960/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.004) Loss 1.3853 (1.3817) Prec@1 62.500 (63.521) Prec@5 89.062 (87.271) Epoch: [27][980/1345], lr: 0.01000 Time 0.836 (0.743) Data 0.000 (0.004) Loss 0.9843 (1.3827) Prec@1 68.750 (63.495) Prec@5 92.188 (87.248) Epoch: [27][1000/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.004) Loss 1.1755 (1.3846) Prec@1 62.500 (63.480) Prec@5 93.750 (87.241) Epoch: [27][1020/1345], lr: 0.01000 Time 0.707 (0.743) Data 0.000 (0.004) Loss 1.0292 (1.3863) Prec@1 67.188 (63.417) Prec@5 95.312 (87.226) Epoch: [27][1040/1345], lr: 0.01000 Time 0.710 (0.743) Data 0.000 (0.004) Loss 1.8521 (1.3877) Prec@1 53.125 (63.396) Prec@5 82.812 (87.213) Epoch: [27][1060/1345], lr: 0.01000 Time 0.706 (0.743) Data 0.000 (0.004) Loss 1.3699 (1.3877) Prec@1 59.375 (63.394) Prec@5 89.062 (87.219) Epoch: [27][1080/1345], lr: 0.01000 Time 0.713 (0.743) Data 0.001 (0.004) Loss 1.1477 (1.3905) Prec@1 70.312 (63.343) Prec@5 89.062 (87.169) Epoch: [27][1100/1345], lr: 0.01000 Time 0.785 (0.743) Data 0.000 (0.004) Loss 1.6313 (1.3931) Prec@1 57.812 (63.258) Prec@5 78.125 (87.127) Epoch: [27][1120/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.004) Loss 1.1981 (1.3933) Prec@1 71.875 (63.265) Prec@5 90.625 (87.129) Epoch: [27][1140/1345], lr: 0.01000 Time 0.718 (0.743) Data 0.001 (0.004) Loss 1.3063 (1.3940) Prec@1 65.625 (63.268) Prec@5 92.188 (87.106) Epoch: [27][1160/1345], lr: 0.01000 Time 0.751 (0.743) Data 0.000 (0.004) Loss 1.5875 (1.3953) Prec@1 54.688 (63.256) Prec@5 78.125 (87.068) Epoch: [27][1180/1345], lr: 0.01000 Time 0.711 (0.743) Data 0.000 (0.004) Loss 1.3283 (1.3963) Prec@1 59.375 (63.203) Prec@5 89.062 (87.063) Epoch: [27][1200/1345], lr: 0.01000 Time 0.761 (0.743) Data 0.000 (0.004) Loss 1.3402 (1.3959) Prec@1 67.188 (63.201) Prec@5 82.812 (87.080) Epoch: [27][1220/1345], lr: 0.01000 Time 0.754 (0.743) Data 0.000 (0.003) Loss 1.1642 (1.3960) Prec@1 75.000 (63.210) Prec@5 89.062 (87.062) Epoch: [27][1240/1345], lr: 0.01000 Time 0.849 (0.743) Data 0.000 (0.003) Loss 1.7319 (1.3950) Prec@1 56.250 (63.233) Prec@5 84.375 (87.082) Epoch: [27][1260/1345], lr: 0.01000 Time 0.840 (0.743) Data 0.000 (0.003) Loss 1.3669 (1.3948) Prec@1 67.188 (63.235) Prec@5 81.250 (87.086) Epoch: [27][1280/1345], lr: 0.01000 Time 0.718 (0.743) Data 0.000 (0.003) Loss 1.7003 (1.3950) Prec@1 50.000 (63.220) Prec@5 81.250 (87.079) Epoch: [27][1300/1345], lr: 0.01000 Time 0.711 (0.743) Data 0.000 (0.003) Loss 1.5874 (1.3956) Prec@1 57.812 (63.200) Prec@5 82.812 (87.059) Epoch: [27][1320/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.003) Loss 1.5168 (1.3964) Prec@1 60.938 (63.180) Prec@5 79.688 (87.049) Epoch: [27][1340/1345], lr: 0.01000 Time 0.710 (0.743) Data 0.000 (0.003) Loss 1.2533 (1.3966) Prec@1 59.375 (63.187) Prec@5 89.062 (87.050) No BN layer Freezing. Test: [0/181] Time 3.214 (3.2143) Loss 2.4411 (2.4411) Prec@1 45.312 (45.312) Prec@5 78.125 (78.125) Test: [20/181] Time 1.236 (0.6077) Loss 2.5808 (2.3452) Prec@1 42.188 (46.057) Prec@5 71.875 (75.521) Test: [40/181] Time 1.024 (0.5362) Loss 2.3657 (2.3996) Prec@1 48.438 (44.855) Prec@5 70.312 (74.581) Test: [60/181] Time 0.913 (0.5108) Loss 2.5820 (2.4312) Prec@1 50.000 (44.314) Prec@5 67.188 (74.155) Test: [80/181] Time 1.045 (0.4957) Loss 2.2478 (2.4633) Prec@1 40.625 (43.692) Prec@5 82.812 (73.785) Test: [100/181] Time 1.008 (0.4866) Loss 2.6268 (2.4603) Prec@1 40.625 (44.044) Prec@5 71.875 (74.087) Test: [120/181] Time 0.899 (0.4820) Loss 2.9337 (2.4679) Prec@1 39.062 (43.776) Prec@5 70.312 (74.006) Test: [140/181] Time 1.190 (0.4805) Loss 2.5883 (2.4681) Prec@1 40.625 (43.783) Prec@5 79.688 (73.947) Test: [160/181] Time 1.013 (0.4780) Loss 2.6940 (2.4570) Prec@1 45.312 (43.896) Prec@5 70.312 (73.971) Testing Results: Prec@1 44.106 Prec@5 73.993 Loss 2.45435 Time 0.4730 No BN layer Freezing. Epoch: [28][0/1345], lr: 0.01000 Time 4.350 (4.350) Data 3.601 (3.601) Loss 1.1829 (1.1829) Prec@1 67.188 (67.188) Prec@5 89.062 (89.062) Epoch: [28][20/1345], lr: 0.01000 Time 0.714 (0.920) Data 0.000 (0.172) Loss 1.4715 (1.3448) Prec@1 57.812 (64.286) Prec@5 85.938 (88.318) Epoch: [28][40/1345], lr: 0.01000 Time 0.738 (0.839) Data 0.001 (0.088) Loss 1.2128 (1.3291) Prec@1 73.438 (64.748) Prec@5 92.188 (88.491) Epoch: [28][60/1345], lr: 0.01000 Time 0.710 (0.808) Data 0.000 (0.059) Loss 1.2245 (1.3246) Prec@1 71.875 (64.831) Prec@5 90.625 (88.550) Epoch: [28][80/1345], lr: 0.01000 Time 0.711 (0.790) Data 0.000 (0.045) Loss 1.3308 (1.3137) Prec@1 73.438 (65.258) Prec@5 87.500 (88.850) Epoch: [28][100/1345], lr: 0.01000 Time 0.715 (0.777) Data 0.001 (0.036) Loss 1.4831 (1.3228) Prec@1 57.812 (65.099) Prec@5 78.125 (88.397) Epoch: [28][120/1345], lr: 0.01000 Time 0.708 (0.769) Data 0.000 (0.030) Loss 1.3309 (1.3272) Prec@1 64.062 (65.070) Prec@5 84.375 (88.159) Epoch: [28][140/1345], lr: 0.01000 Time 0.707 (0.762) Data 0.000 (0.026) Loss 1.5882 (1.3299) Prec@1 62.500 (65.082) Prec@5 84.375 (88.165) Epoch: [28][160/1345], lr: 0.01000 Time 0.876 (0.760) Data 0.001 (0.023) Loss 1.4085 (1.3368) Prec@1 59.375 (64.965) Prec@5 87.500 (87.888) Epoch: [28][180/1345], lr: 0.01000 Time 0.829 (0.758) Data 0.000 (0.020) Loss 1.6827 (1.3278) Prec@1 57.812 (65.159) Prec@5 85.938 (88.061) Epoch: [28][200/1345], lr: 0.01000 Time 0.746 (0.757) Data 0.000 (0.018) Loss 1.4989 (1.3374) Prec@1 59.375 (64.871) Prec@5 85.938 (87.904) Epoch: [28][220/1345], lr: 0.01000 Time 0.762 (0.755) Data 0.001 (0.017) Loss 1.3126 (1.3379) Prec@1 65.625 (64.868) Prec@5 89.062 (87.861) Epoch: [28][240/1345], lr: 0.01000 Time 0.716 (0.755) Data 0.001 (0.015) Loss 1.3149 (1.3416) Prec@1 60.938 (64.704) Prec@5 92.188 (87.798) Epoch: [28][260/1345], lr: 0.01000 Time 0.707 (0.753) Data 0.000 (0.014) Loss 1.3311 (1.3522) Prec@1 70.312 (64.326) Prec@5 85.938 (87.722) Epoch: [28][280/1345], lr: 0.01000 Time 0.815 (0.752) Data 0.000 (0.013) Loss 1.4966 (1.3482) Prec@1 60.938 (64.368) Prec@5 85.938 (87.795) Epoch: [28][300/1345], lr: 0.01000 Time 0.742 (0.752) Data 0.000 (0.012) Loss 1.3079 (1.3489) Prec@1 65.625 (64.384) Prec@5 90.625 (87.760) Epoch: [28][320/1345], lr: 0.01000 Time 0.707 (0.751) Data 0.000 (0.012) Loss 1.2205 (1.3520) Prec@1 65.625 (64.374) Prec@5 93.750 (87.709) Epoch: [28][340/1345], lr: 0.01000 Time 0.705 (0.749) Data 0.000 (0.011) Loss 1.2290 (1.3542) Prec@1 67.188 (64.241) Prec@5 87.500 (87.697) Epoch: [28][360/1345], lr: 0.01000 Time 0.722 (0.748) Data 0.000 (0.010) Loss 1.4820 (1.3526) Prec@1 65.625 (64.244) Prec@5 87.500 (87.747) Epoch: [28][380/1345], lr: 0.01000 Time 0.767 (0.747) Data 0.001 (0.010) Loss 1.7187 (1.3526) Prec@1 62.500 (64.288) Prec@5 73.438 (87.693) Epoch: [28][400/1345], lr: 0.01000 Time 0.709 (0.746) Data 0.000 (0.009) Loss 1.2539 (1.3600) Prec@1 65.625 (64.148) Prec@5 90.625 (87.558) Epoch: [28][420/1345], lr: 0.01000 Time 0.706 (0.746) Data 0.000 (0.009) Loss 1.0700 (1.3612) Prec@1 68.750 (64.092) Prec@5 92.188 (87.582) Epoch: [28][440/1345], lr: 0.01000 Time 0.742 (0.745) Data 0.000 (0.009) Loss 1.1252 (1.3600) Prec@1 73.438 (64.087) Prec@5 89.062 (87.606) Epoch: [28][460/1345], lr: 0.01000 Time 0.838 (0.745) Data 0.000 (0.008) Loss 1.6942 (1.3575) Prec@1 54.688 (64.144) Prec@5 84.375 (87.642) Epoch: [28][480/1345], lr: 0.01000 Time 0.856 (0.745) Data 0.000 (0.008) Loss 1.3107 (1.3593) Prec@1 65.625 (64.088) Prec@5 92.188 (87.610) Epoch: [28][500/1345], lr: 0.01000 Time 0.765 (0.745) Data 0.000 (0.008) Loss 1.5412 (1.3598) Prec@1 62.500 (64.034) Prec@5 82.812 (87.578) Epoch: [28][520/1345], lr: 0.01000 Time 0.781 (0.744) Data 0.001 (0.007) Loss 1.4933 (1.3604) Prec@1 57.812 (64.030) Prec@5 84.375 (87.560) Epoch: [28][540/1345], lr: 0.01000 Time 0.710 (0.744) Data 0.000 (0.007) Loss 1.1758 (1.3609) Prec@1 71.875 (64.013) Prec@5 87.500 (87.523) Epoch: [28][560/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.007) Loss 1.6449 (1.3630) Prec@1 67.188 (63.912) Prec@5 78.125 (87.517) Epoch: [28][580/1345], lr: 0.01000 Time 0.727 (0.743) Data 0.001 (0.007) Loss 1.3022 (1.3646) Prec@1 67.188 (63.888) Prec@5 84.375 (87.497) Epoch: [28][600/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.006) Loss 1.3256 (1.3644) Prec@1 59.375 (63.904) Prec@5 87.500 (87.461) Epoch: [28][620/1345], lr: 0.01000 Time 0.836 (0.744) Data 0.000 (0.006) Loss 1.1479 (1.3632) Prec@1 64.062 (63.924) Prec@5 92.188 (87.467) Epoch: [28][640/1345], lr: 0.01000 Time 0.849 (0.744) Data 0.001 (0.006) Loss 1.1265 (1.3638) Prec@1 71.875 (63.914) Prec@5 92.188 (87.478) Epoch: [28][660/1345], lr: 0.01000 Time 0.728 (0.744) Data 0.000 (0.006) Loss 1.2032 (1.3619) Prec@1 64.062 (63.930) Prec@5 90.625 (87.500) Epoch: [28][680/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.006) Loss 0.9669 (1.3617) Prec@1 73.438 (63.936) Prec@5 92.188 (87.493) Epoch: [28][700/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.006) Loss 1.8631 (1.3604) Prec@1 53.125 (63.967) Prec@5 82.812 (87.547) Epoch: [28][720/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.005) Loss 1.1926 (1.3627) Prec@1 67.188 (63.928) Prec@5 92.188 (87.496) Epoch: [28][740/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.005) Loss 1.5961 (1.3641) Prec@1 59.375 (63.915) Prec@5 87.500 (87.483) Epoch: [28][760/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.005) Loss 1.3365 (1.3646) Prec@1 57.812 (63.878) Prec@5 85.938 (87.467) Epoch: [28][780/1345], lr: 0.01000 Time 0.740 (0.742) Data 0.000 (0.005) Loss 1.0695 (1.3648) Prec@1 75.000 (63.856) Prec@5 92.188 (87.444) Epoch: [28][800/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.005) Loss 1.2322 (1.3669) Prec@1 65.625 (63.805) Prec@5 85.938 (87.404) Epoch: [28][820/1345], lr: 0.01000 Time 0.714 (0.742) Data 0.000 (0.005) Loss 1.3666 (1.3679) Prec@1 60.938 (63.779) Prec@5 84.375 (87.374) Epoch: [28][840/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.005) Loss 1.7588 (1.3690) Prec@1 62.500 (63.767) Prec@5 84.375 (87.359) Epoch: [28][860/1345], lr: 0.01000 Time 0.816 (0.742) Data 0.000 (0.005) Loss 1.2179 (1.3697) Prec@1 70.312 (63.774) Prec@5 89.062 (87.349) Epoch: [28][880/1345], lr: 0.01000 Time 0.779 (0.742) Data 0.000 (0.005) Loss 1.7607 (1.3709) Prec@1 56.250 (63.789) Prec@5 76.562 (87.307) Epoch: [28][900/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.000 (0.004) Loss 1.0634 (1.3698) Prec@1 70.312 (63.815) Prec@5 92.188 (87.328) Epoch: [28][920/1345], lr: 0.01000 Time 0.742 (0.742) Data 0.000 (0.004) Loss 1.4474 (1.3709) Prec@1 65.625 (63.791) Prec@5 84.375 (87.290) Epoch: [28][940/1345], lr: 0.01000 Time 0.712 (0.742) Data 0.000 (0.004) Loss 1.2405 (1.3730) Prec@1 62.500 (63.737) Prec@5 90.625 (87.274) Epoch: [28][960/1345], lr: 0.01000 Time 0.752 (0.742) Data 0.000 (0.004) Loss 1.9083 (1.3743) Prec@1 51.562 (63.713) Prec@5 84.375 (87.258) Epoch: [28][980/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.000 (0.004) Loss 1.4926 (1.3762) Prec@1 54.688 (63.658) Prec@5 87.500 (87.226) Epoch: [28][1000/1345], lr: 0.01000 Time 0.767 (0.742) Data 0.000 (0.004) Loss 1.4530 (1.3767) Prec@1 59.375 (63.654) Prec@5 89.062 (87.235) Epoch: [28][1020/1345], lr: 0.01000 Time 0.839 (0.742) Data 0.000 (0.004) Loss 1.3994 (1.3787) Prec@1 67.188 (63.610) Prec@5 87.500 (87.199) Epoch: [28][1040/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.004) Loss 1.6355 (1.3794) Prec@1 59.375 (63.582) Prec@5 82.812 (87.195) Epoch: [28][1060/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.004) Loss 1.3633 (1.3792) Prec@1 65.625 (63.566) Prec@5 89.062 (87.228) Epoch: [28][1080/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.004) Loss 1.2386 (1.3796) Prec@1 67.188 (63.557) Prec@5 89.062 (87.214) Epoch: [28][1100/1345], lr: 0.01000 Time 0.705 (0.742) Data 0.000 (0.004) Loss 1.1118 (1.3809) Prec@1 60.938 (63.522) Prec@5 93.750 (87.199) Epoch: [28][1120/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.8722 (1.3815) Prec@1 54.688 (63.491) Prec@5 76.562 (87.198) Epoch: [28][1140/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.004) Loss 1.4827 (1.3817) Prec@1 57.812 (63.476) Prec@5 85.938 (87.184) Epoch: [28][1160/1345], lr: 0.01000 Time 0.758 (0.742) Data 0.000 (0.004) Loss 1.6815 (1.3830) Prec@1 60.938 (63.427) Prec@5 78.125 (87.170) Epoch: [28][1180/1345], lr: 0.01000 Time 0.743 (0.742) Data 0.000 (0.003) Loss 1.3619 (1.3821) Prec@1 64.062 (63.461) Prec@5 87.500 (87.197) Epoch: [28][1200/1345], lr: 0.01000 Time 0.707 (0.742) Data 0.000 (0.003) Loss 1.0783 (1.3822) Prec@1 71.875 (63.469) Prec@5 90.625 (87.220) Epoch: [28][1220/1345], lr: 0.01000 Time 0.847 (0.742) Data 0.001 (0.003) Loss 1.5167 (1.3829) Prec@1 62.500 (63.448) Prec@5 84.375 (87.210) Epoch: [28][1240/1345], lr: 0.01000 Time 0.832 (0.742) Data 0.000 (0.003) Loss 1.2816 (1.3849) Prec@1 68.750 (63.404) Prec@5 85.938 (87.160) Epoch: [28][1260/1345], lr: 0.01000 Time 0.766 (0.742) Data 0.001 (0.003) Loss 1.2533 (1.3844) Prec@1 64.062 (63.426) Prec@5 90.625 (87.175) Epoch: [28][1280/1345], lr: 0.01000 Time 0.763 (0.742) Data 0.000 (0.003) Loss 1.2094 (1.3856) Prec@1 67.188 (63.390) Prec@5 93.750 (87.168) Epoch: [28][1300/1345], lr: 0.01000 Time 0.751 (0.742) Data 0.000 (0.003) Loss 1.5792 (1.3866) Prec@1 67.188 (63.379) Prec@5 81.250 (87.136) Epoch: [28][1320/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.003) Loss 1.4802 (1.3868) Prec@1 59.375 (63.337) Prec@5 87.500 (87.139) Epoch: [28][1340/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.003) Loss 0.9885 (1.3877) Prec@1 75.000 (63.328) Prec@5 92.188 (87.122) No BN layer Freezing. Test: [0/181] Time 3.540 (3.5403) Loss 2.4737 (2.4737) Prec@1 39.062 (39.062) Prec@5 73.438 (73.438) Test: [20/181] Time 1.023 (0.5927) Loss 2.8623 (2.5957) Prec@1 37.500 (42.485) Prec@5 75.000 (72.098) Test: [40/181] Time 1.002 (0.5266) Loss 2.7599 (2.6752) Prec@1 42.188 (41.273) Prec@5 65.625 (70.732) Test: [60/181] Time 0.795 (0.5074) Loss 3.1309 (2.7201) Prec@1 39.062 (40.727) Prec@5 59.375 (69.698) Test: [80/181] Time 0.956 (0.4953) Loss 2.5266 (2.7399) Prec@1 42.188 (40.104) Prec@5 79.688 (69.502) Test: [100/181] Time 0.768 (0.4882) Loss 2.7814 (2.7169) Prec@1 35.938 (40.439) Prec@5 62.500 (69.817) Test: [120/181] Time 0.696 (0.4821) Loss 3.4383 (2.7245) Prec@1 34.375 (40.548) Prec@5 50.000 (70.015) Test: [140/181] Time 0.890 (0.4798) Loss 3.3554 (2.7260) Prec@1 31.250 (40.348) Prec@5 60.938 (70.246) Test: [160/181] Time 0.643 (0.4766) Loss 3.2088 (2.7150) Prec@1 45.312 (40.596) Prec@5 67.188 (70.516) Testing Results: Prec@1 40.738 Prec@5 70.486 Loss 2.70428 Time 0.4726 No BN layer Freezing. Epoch: [29][0/1345], lr: 0.01000 Time 4.398 (4.398) Data 3.633 (3.633) Loss 1.2925 (1.2925) Prec@1 70.312 (70.312) Prec@5 82.812 (82.812) Epoch: [29][20/1345], lr: 0.01000 Time 0.735 (0.912) Data 0.001 (0.173) Loss 1.2052 (1.2893) Prec@1 71.875 (65.179) Prec@5 89.062 (88.542) Epoch: [29][40/1345], lr: 0.01000 Time 0.708 (0.829) Data 0.000 (0.089) Loss 1.1527 (1.2552) Prec@1 68.750 (65.968) Prec@5 89.062 (88.567) Epoch: [29][60/1345], lr: 0.01000 Time 0.710 (0.795) Data 0.000 (0.060) Loss 1.2977 (1.2764) Prec@1 70.312 (66.112) Prec@5 87.500 (88.473) Epoch: [29][80/1345], lr: 0.01000 Time 0.708 (0.778) Data 0.000 (0.045) Loss 1.3276 (1.2993) Prec@1 65.625 (65.664) Prec@5 85.938 (88.137) Epoch: [29][100/1345], lr: 0.01000 Time 0.721 (0.771) Data 0.001 (0.036) Loss 1.3359 (1.2922) Prec@1 70.312 (65.873) Prec@5 85.938 (88.150) Epoch: [29][120/1345], lr: 0.01000 Time 0.815 (0.764) Data 0.000 (0.030) Loss 1.4902 (1.3105) Prec@1 62.500 (65.160) Prec@5 89.062 (88.055) Epoch: [29][140/1345], lr: 0.01000 Time 0.845 (0.760) Data 0.000 (0.026) Loss 1.2977 (1.3077) Prec@1 60.938 (65.016) Prec@5 87.500 (88.387) Epoch: [29][160/1345], lr: 0.01000 Time 0.752 (0.758) Data 0.000 (0.023) Loss 1.3165 (1.3124) Prec@1 67.188 (64.946) Prec@5 87.500 (88.267) Epoch: [29][180/1345], lr: 0.01000 Time 0.718 (0.757) Data 0.000 (0.021) Loss 1.2219 (1.3077) Prec@1 71.875 (65.064) Prec@5 87.500 (88.406) Epoch: [29][200/1345], lr: 0.01000 Time 0.706 (0.754) Data 0.000 (0.019) Loss 1.3915 (1.3113) Prec@1 60.938 (65.135) Prec@5 85.938 (88.355) Epoch: [29][220/1345], lr: 0.01000 Time 0.709 (0.753) Data 0.000 (0.017) Loss 1.4958 (1.3160) Prec@1 65.625 (64.989) Prec@5 85.938 (88.221) Epoch: [29][240/1345], lr: 0.01000 Time 0.707 (0.753) Data 0.000 (0.016) Loss 1.9334 (1.3206) Prec@1 56.250 (64.873) Prec@5 79.688 (88.200) Epoch: [29][260/1345], lr: 0.01000 Time 0.785 (0.753) Data 0.001 (0.014) Loss 1.2185 (1.3196) Prec@1 67.188 (65.026) Prec@5 90.625 (88.188) Epoch: [29][280/1345], lr: 0.01000 Time 0.859 (0.753) Data 0.000 (0.013) Loss 1.4620 (1.3188) Prec@1 57.812 (65.080) Prec@5 85.938 (88.195) Epoch: [29][300/1345], lr: 0.01000 Time 0.846 (0.752) Data 0.000 (0.012) Loss 1.3231 (1.3228) Prec@1 62.500 (65.059) Prec@5 89.062 (88.040) Epoch: [29][320/1345], lr: 0.01000 Time 0.708 (0.751) Data 0.000 (0.012) Loss 1.1190 (1.3296) Prec@1 68.750 (64.963) Prec@5 92.188 (87.948) Epoch: [29][340/1345], lr: 0.01000 Time 0.716 (0.751) Data 0.000 (0.011) Loss 1.6081 (1.3284) Prec@1 65.625 (64.942) Prec@5 81.250 (87.917) Epoch: [29][360/1345], lr: 0.01000 Time 0.731 (0.751) Data 0.000 (0.010) Loss 1.4988 (1.3368) Prec@1 57.812 (64.772) Prec@5 85.938 (87.768) Epoch: [29][380/1345], lr: 0.01000 Time 0.709 (0.751) Data 0.000 (0.010) Loss 1.6423 (1.3369) Prec@1 57.812 (64.792) Prec@5 81.250 (87.775) Epoch: [29][400/1345], lr: 0.01000 Time 0.710 (0.750) Data 0.001 (0.009) Loss 1.1939 (1.3414) Prec@1 71.875 (64.748) Prec@5 89.062 (87.656) Epoch: [29][420/1345], lr: 0.01000 Time 0.750 (0.749) Data 0.000 (0.009) Loss 1.4121 (1.3441) Prec@1 60.938 (64.641) Prec@5 87.500 (87.667) Epoch: [29][440/1345], lr: 0.01000 Time 0.707 (0.749) Data 0.000 (0.009) Loss 1.3295 (1.3466) Prec@1 65.625 (64.590) Prec@5 87.500 (87.620) Epoch: [29][460/1345], lr: 0.01000 Time 0.710 (0.748) Data 0.000 (0.008) Loss 1.5273 (1.3495) Prec@1 65.625 (64.540) Prec@5 82.812 (87.581) Epoch: [29][480/1345], lr: 0.01000 Time 0.707 (0.747) Data 0.000 (0.008) Loss 1.2019 (1.3535) Prec@1 68.750 (64.371) Prec@5 90.625 (87.529) Epoch: [29][500/1345], lr: 0.01000 Time 0.710 (0.746) Data 0.000 (0.008) Loss 1.5452 (1.3579) Prec@1 59.375 (64.312) Prec@5 82.812 (87.456) Epoch: [29][520/1345], lr: 0.01000 Time 0.712 (0.746) Data 0.000 (0.007) Loss 1.6170 (1.3605) Prec@1 56.250 (64.233) Prec@5 84.375 (87.437) Epoch: [29][540/1345], lr: 0.01000 Time 0.708 (0.745) Data 0.000 (0.007) Loss 1.5257 (1.3585) Prec@1 65.625 (64.273) Prec@5 87.500 (87.500) Epoch: [29][560/1345], lr: 0.01000 Time 0.785 (0.745) Data 0.000 (0.007) Loss 1.3031 (1.3601) Prec@1 60.938 (64.252) Prec@5 84.375 (87.436) Epoch: [29][580/1345], lr: 0.01000 Time 0.743 (0.745) Data 0.000 (0.007) Loss 1.2060 (1.3602) Prec@1 71.875 (64.221) Prec@5 89.062 (87.468) Epoch: [29][600/1345], lr: 0.01000 Time 0.716 (0.745) Data 0.000 (0.006) Loss 1.2994 (1.3603) Prec@1 65.625 (64.213) Prec@5 89.062 (87.495) Epoch: [29][620/1345], lr: 0.01000 Time 0.753 (0.745) Data 0.000 (0.006) Loss 1.5360 (1.3627) Prec@1 60.938 (64.130) Prec@5 85.938 (87.472) Epoch: [29][640/1345], lr: 0.01000 Time 0.707 (0.745) Data 0.000 (0.006) Loss 1.3053 (1.3645) Prec@1 67.188 (64.062) Prec@5 87.500 (87.427) Epoch: [29][660/1345], lr: 0.01000 Time 0.707 (0.744) Data 0.000 (0.006) Loss 1.5484 (1.3630) Prec@1 64.062 (64.098) Prec@5 85.938 (87.457) Epoch: [29][680/1345], lr: 0.01000 Time 0.708 (0.744) Data 0.000 (0.006) Loss 0.9448 (1.3635) Prec@1 75.000 (64.108) Prec@5 95.312 (87.440) Epoch: [29][700/1345], lr: 0.01000 Time 0.713 (0.744) Data 0.000 (0.006) Loss 1.7585 (1.3666) Prec@1 50.000 (64.045) Prec@5 76.562 (87.391) Epoch: [29][720/1345], lr: 0.01000 Time 0.708 (0.743) Data 0.000 (0.005) Loss 1.3351 (1.3657) Prec@1 59.375 (64.043) Prec@5 90.625 (87.396) Epoch: [29][740/1345], lr: 0.01000 Time 0.711 (0.743) Data 0.000 (0.005) Loss 1.4055 (1.3671) Prec@1 60.938 (64.014) Prec@5 90.625 (87.405) Epoch: [29][760/1345], lr: 0.01000 Time 0.714 (0.742) Data 0.000 (0.005) Loss 1.4168 (1.3690) Prec@1 64.062 (63.974) Prec@5 85.938 (87.354) Epoch: [29][780/1345], lr: 0.01000 Time 0.720 (0.743) Data 0.000 (0.005) Loss 1.4025 (1.3708) Prec@1 60.938 (63.916) Prec@5 89.062 (87.340) Epoch: [29][800/1345], lr: 0.01000 Time 0.709 (0.743) Data 0.000 (0.005) Loss 1.0890 (1.3709) Prec@1 73.438 (63.885) Prec@5 93.750 (87.346) Epoch: [29][820/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.005) Loss 1.1911 (1.3720) Prec@1 67.188 (63.880) Prec@5 92.188 (87.348) Epoch: [29][840/1345], lr: 0.01000 Time 0.769 (0.742) Data 0.001 (0.005) Loss 1.9096 (1.3726) Prec@1 54.688 (63.897) Prec@5 76.562 (87.338) Epoch: [29][860/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.005) Loss 1.4293 (1.3724) Prec@1 67.188 (63.917) Prec@5 87.500 (87.351) Epoch: [29][880/1345], lr: 0.01000 Time 0.847 (0.742) Data 0.000 (0.005) Loss 1.4640 (1.3725) Prec@1 57.812 (63.926) Prec@5 89.062 (87.367) Epoch: [29][900/1345], lr: 0.01000 Time 0.730 (0.742) Data 0.000 (0.004) Loss 1.5138 (1.3702) Prec@1 53.125 (63.957) Prec@5 87.500 (87.420) Epoch: [29][920/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.0986 (1.3704) Prec@1 64.062 (63.923) Prec@5 93.750 (87.427) Epoch: [29][940/1345], lr: 0.01000 Time 0.708 (0.742) Data 0.000 (0.004) Loss 1.3620 (1.3713) Prec@1 65.625 (63.868) Prec@5 90.625 (87.437) Epoch: [29][960/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.000 (0.004) Loss 1.5847 (1.3720) Prec@1 60.938 (63.848) Prec@5 84.375 (87.428) Epoch: [29][980/1345], lr: 0.01000 Time 0.772 (0.742) Data 0.000 (0.004) Loss 1.1581 (1.3743) Prec@1 70.312 (63.814) Prec@5 85.938 (87.392) Epoch: [29][1000/1345], lr: 0.01000 Time 0.706 (0.742) Data 0.000 (0.004) Loss 1.2935 (1.3746) Prec@1 65.625 (63.822) Prec@5 85.938 (87.367) Epoch: [29][1020/1345], lr: 0.01000 Time 0.851 (0.742) Data 0.000 (0.004) Loss 1.3358 (1.3744) Prec@1 67.188 (63.821) Prec@5 89.062 (87.368) Epoch: [29][1040/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.004) Loss 0.8823 (1.3742) Prec@1 75.000 (63.830) Prec@5 92.188 (87.366) Epoch: [29][1060/1345], lr: 0.01000 Time 0.772 (0.742) Data 0.000 (0.004) Loss 1.3652 (1.3739) Prec@1 60.938 (63.840) Prec@5 87.500 (87.382) Epoch: [29][1080/1345], lr: 0.01000 Time 0.720 (0.742) Data 0.001 (0.004) Loss 1.4244 (1.3746) Prec@1 64.062 (63.810) Prec@5 84.375 (87.380) Epoch: [29][1100/1345], lr: 0.01000 Time 0.753 (0.741) Data 0.000 (0.004) Loss 1.3804 (1.3751) Prec@1 68.750 (63.780) Prec@5 85.938 (87.377) Epoch: [29][1120/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.004) Loss 1.5194 (1.3758) Prec@1 54.688 (63.747) Prec@5 85.938 (87.370) Epoch: [29][1140/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.004) Loss 1.7062 (1.3760) Prec@1 54.688 (63.735) Prec@5 81.250 (87.378) Epoch: [29][1160/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.000 (0.004) Loss 1.6586 (1.3776) Prec@1 57.812 (63.696) Prec@5 84.375 (87.353) Epoch: [29][1180/1345], lr: 0.01000 Time 0.751 (0.742) Data 0.000 (0.003) Loss 1.2038 (1.3769) Prec@1 67.188 (63.712) Prec@5 90.625 (87.349) Epoch: [29][1200/1345], lr: 0.01000 Time 0.711 (0.742) Data 0.000 (0.003) Loss 1.3124 (1.3760) Prec@1 70.312 (63.726) Prec@5 87.500 (87.353) Epoch: [29][1220/1345], lr: 0.01000 Time 0.763 (0.742) Data 0.000 (0.003) Loss 1.5409 (1.3778) Prec@1 64.062 (63.682) Prec@5 85.938 (87.329) Epoch: [29][1240/1345], lr: 0.01000 Time 0.747 (0.742) Data 0.000 (0.003) Loss 1.7075 (1.3784) Prec@1 64.062 (63.661) Prec@5 82.812 (87.328) Epoch: [29][1260/1345], lr: 0.01000 Time 0.751 (0.742) Data 0.000 (0.003) Loss 1.0808 (1.3784) Prec@1 71.875 (63.651) Prec@5 85.938 (87.320) Epoch: [29][1280/1345], lr: 0.01000 Time 0.710 (0.742) Data 0.000 (0.003) Loss 1.5190 (1.3807) Prec@1 59.375 (63.615) Prec@5 85.938 (87.279) Epoch: [29][1300/1345], lr: 0.01000 Time 0.717 (0.742) Data 0.001 (0.003) Loss 1.3834 (1.3813) Prec@1 60.938 (63.593) Prec@5 84.375 (87.260) Epoch: [29][1320/1345], lr: 0.01000 Time 0.714 (0.742) Data 0.000 (0.003) Loss 1.2111 (1.3809) Prec@1 65.625 (63.596) Prec@5 92.188 (87.265) Epoch: [29][1340/1345], lr: 0.01000 Time 0.709 (0.742) Data 0.000 (0.003) Loss 1.3052 (1.3808) Prec@1 70.312 (63.616) Prec@5 85.938 (87.256) No BN layer Freezing. Test: [0/181] Time 3.638 (3.6382) Loss 2.5935 (2.5935) Prec@1 50.000 (50.000) Prec@5 76.562 (76.562) Test: [20/181] Time 1.247 (0.6153) Loss 2.4268 (2.4152) Prec@1 43.750 (46.205) Prec@5 78.125 (76.116) Test: [40/181] Time 1.144 (0.5392) Loss 2.5203 (2.5374) Prec@1 40.625 (43.941) Prec@5 73.438 (74.619) Test: [60/181] Time 1.106 (0.5123) Loss 2.6859 (2.5653) Prec@1 45.312 (43.161) Prec@5 64.062 (73.796) Test: [80/181] Time 1.193 (0.4996) Loss 2.2653 (2.5904) Prec@1 40.625 (42.535) Prec@5 81.250 (73.418) Test: [100/181] Time 0.938 (0.4912) Loss 2.9654 (2.5817) Prec@1 37.500 (42.590) Prec@5 68.750 (73.592) Test: [120/181] Time 1.015 (0.4860) Loss 3.0372 (2.5909) Prec@1 37.500 (42.381) Prec@5 68.750 (73.722) Test: [140/181] Time 1.099 (0.4820) Loss 3.1235 (2.5905) Prec@1 39.062 (42.509) Prec@5 70.312 (73.836) Test: [160/181] Time 1.110 (0.4798) Loss 3.0147 (2.5823) Prec@1 37.500 (42.741) Prec@5 68.750 (73.767) Testing Results: Prec@1 42.943 Prec@5 73.611 Loss 2.58130 Time 0.4749 No BN layer Freezing. Epoch: [30][0/1345], lr: 0.00100 Time 3.874 (3.874) Data 3.121 (3.121) Loss 1.3342 (1.3342) Prec@1 60.938 (60.938) Prec@5 89.062 (89.062) Epoch: [30][20/1345], lr: 0.00100 Time 0.778 (0.884) Data 0.000 (0.149) Loss 1.1148 (1.3626) Prec@1 62.500 (62.351) Prec@5 93.750 (86.830) Epoch: [30][40/1345], lr: 0.00100 Time 0.751 (0.809) Data 0.000 (0.077) Loss 1.1568 (1.3599) Prec@1 68.750 (63.720) Prec@5 89.062 (86.547) Epoch: [30][60/1345], lr: 0.00100 Time 0.781 (0.783) Data 0.001 (0.052) Loss 0.9867 (1.2857) Prec@1 71.875 (65.548) Prec@5 95.312 (87.961) Epoch: [30][80/1345], lr: 0.00100 Time 0.762 (0.775) Data 0.000 (0.039) Loss 1.0582 (1.2624) Prec@1 71.875 (66.011) Prec@5 89.062 (88.310) Epoch: [30][100/1345], lr: 0.00100 Time 0.740 (0.766) Data 0.001 (0.031) Loss 0.9869 (1.2434) Prec@1 70.312 (66.584) Prec@5 92.188 (88.738) Epoch: [30][120/1345], lr: 0.00100 Time 0.729 (0.762) Data 0.000 (0.026) Loss 1.1188 (1.2255) Prec@1 67.188 (67.136) Prec@5 89.062 (88.778) Epoch: [30][140/1345], lr: 0.00100 Time 0.715 (0.760) Data 0.000 (0.023) Loss 1.1213 (1.2190) Prec@1 70.312 (67.287) Prec@5 90.625 (88.852) Epoch: [30][160/1345], lr: 0.00100 Time 0.768 (0.757) Data 0.000 (0.020) Loss 1.2116 (1.1996) Prec@1 62.500 (67.634) Prec@5 90.625 (89.218) Epoch: [30][180/1345], lr: 0.00100 Time 0.713 (0.756) Data 0.000 (0.018) Loss 1.1447 (1.1859) Prec@1 73.438 (67.939) Prec@5 90.625 (89.555) Epoch: [30][200/1345], lr: 0.00100 Time 0.707 (0.752) Data 0.000 (0.016) Loss 1.1822 (1.1812) Prec@1 73.438 (68.081) Prec@5 89.062 (89.599) Epoch: [30][220/1345], lr: 0.00100 Time 0.708 (0.751) Data 0.000 (0.015) Loss 1.1822 (1.1770) Prec@1 70.312 (68.347) Prec@5 84.375 (89.614) Epoch: [30][240/1345], lr: 0.00100 Time 0.717 (0.748) Data 0.000 (0.013) Loss 0.7887 (1.1716) Prec@1 73.438 (68.471) Prec@5 95.312 (89.737) Epoch: [30][260/1345], lr: 0.00100 Time 0.711 (0.746) Data 0.000 (0.012) Loss 0.8767 (1.1624) Prec@1 81.250 (68.654) Prec@5 95.312 (89.871) Epoch: [30][280/1345], lr: 0.00100 Time 0.708 (0.745) Data 0.000 (0.012) Loss 1.1657 (1.1592) Prec@1 71.875 (68.733) Prec@5 90.625 (89.908) Epoch: [30][300/1345], lr: 0.00100 Time 0.713 (0.744) Data 0.000 (0.011) Loss 1.1792 (1.1529) Prec@1 76.562 (68.864) Prec@5 90.625 (89.987) Epoch: [30][320/1345], lr: 0.00100 Time 0.707 (0.744) Data 0.000 (0.010) Loss 0.8633 (1.1422) Prec@1 79.688 (69.188) Prec@5 93.750 (90.114) Epoch: [30][340/1345], lr: 0.00100 Time 0.709 (0.744) Data 0.000 (0.010) Loss 1.5005 (1.1388) Prec@1 60.938 (69.300) Prec@5 85.938 (90.208) Epoch: [30][360/1345], lr: 0.00100 Time 0.721 (0.743) Data 0.000 (0.009) Loss 0.8828 (1.1294) Prec@1 81.250 (69.559) Prec@5 89.062 (90.378) Epoch: [30][380/1345], lr: 0.00100 Time 0.784 (0.742) Data 0.000 (0.009) Loss 0.9004 (1.1266) Prec@1 78.125 (69.628) Prec@5 96.875 (90.432) Epoch: [30][400/1345], lr: 0.00100 Time 0.763 (0.741) Data 0.001 (0.008) Loss 1.0513 (1.1222) Prec@1 78.125 (69.810) Prec@5 89.062 (90.500) Epoch: [30][420/1345], lr: 0.00100 Time 0.708 (0.740) Data 0.000 (0.008) Loss 1.0330 (1.1179) Prec@1 67.188 (69.927) Prec@5 89.062 (90.536) Epoch: [30][440/1345], lr: 0.00100 Time 0.708 (0.740) Data 0.000 (0.008) Loss 1.2975 (1.1141) Prec@1 68.750 (70.072) Prec@5 89.062 (90.575) Epoch: [30][460/1345], lr: 0.00100 Time 0.733 (0.740) Data 0.000 (0.007) Loss 1.3259 (1.1081) Prec@1 70.312 (70.272) Prec@5 84.375 (90.639) Epoch: [30][480/1345], lr: 0.00100 Time 0.746 (0.740) Data 0.000 (0.007) Loss 1.1369 (1.1075) Prec@1 75.000 (70.293) Prec@5 87.500 (90.657) Epoch: [30][500/1345], lr: 0.00100 Time 0.732 (0.741) Data 0.000 (0.007) Loss 1.0293 (1.1037) Prec@1 67.188 (70.328) Prec@5 92.188 (90.709) Epoch: [30][520/1345], lr: 0.00100 Time 0.823 (0.741) Data 0.001 (0.006) Loss 0.9225 (1.1026) Prec@1 75.000 (70.357) Prec@5 90.625 (90.733) Epoch: [30][540/1345], lr: 0.00100 Time 0.708 (0.741) Data 0.000 (0.006) Loss 1.3232 (1.1012) Prec@1 71.875 (70.411) Prec@5 85.938 (90.720) Epoch: [30][560/1345], lr: 0.00100 Time 0.779 (0.740) Data 0.000 (0.006) Loss 0.8686 (1.0949) Prec@1 76.562 (70.535) Prec@5 90.625 (90.800) Epoch: [30][580/1345], lr: 0.00100 Time 0.709 (0.741) Data 0.000 (0.006) Loss 0.9282 (1.0910) Prec@1 78.125 (70.662) Prec@5 89.062 (90.864) Epoch: [30][600/1345], lr: 0.00100 Time 0.710 (0.741) Data 0.000 (0.006) Loss 0.6603 (1.0865) Prec@1 84.375 (70.767) Prec@5 96.875 (90.911) Epoch: [30][620/1345], lr: 0.00100 Time 0.810 (0.741) Data 0.001 (0.005) Loss 0.9923 (1.0831) Prec@1 71.875 (70.838) Prec@5 92.188 (90.947) Epoch: [30][640/1345], lr: 0.00100 Time 0.709 (0.740) Data 0.000 (0.005) Loss 1.1774 (1.0798) Prec@1 68.750 (70.927) Prec@5 89.062 (90.991) Epoch: [30][660/1345], lr: 0.00100 Time 0.783 (0.740) Data 0.000 (0.005) Loss 1.1362 (1.0801) Prec@1 73.438 (70.918) Prec@5 89.062 (90.982) Epoch: [30][680/1345], lr: 0.00100 Time 0.722 (0.740) Data 0.000 (0.005) Loss 0.8797 (1.0762) Prec@1 70.312 (71.008) Prec@5 98.438 (91.047) Epoch: [30][700/1345], lr: 0.00100 Time 0.709 (0.741) Data 0.000 (0.005) Loss 0.9729 (1.0732) Prec@1 73.438 (71.077) Prec@5 93.750 (91.077) Epoch: [30][720/1345], lr: 0.00100 Time 0.784 (0.740) Data 0.000 (0.005) Loss 0.9118 (1.0690) Prec@1 76.562 (71.175) Prec@5 90.625 (91.132) Epoch: [30][740/1345], lr: 0.00100 Time 0.782 (0.740) Data 0.000 (0.005) Loss 0.8623 (1.0664) Prec@1 81.250 (71.280) Prec@5 92.188 (91.154) Epoch: [30][760/1345], lr: 0.00100 Time 0.713 (0.740) Data 0.000 (0.005) Loss 0.9291 (1.0644) Prec@1 71.875 (71.339) Prec@5 92.188 (91.179) Epoch: [30][780/1345], lr: 0.00100 Time 0.709 (0.740) Data 0.000 (0.004) Loss 0.9096 (1.0617) Prec@1 73.438 (71.381) Prec@5 93.750 (91.243) Epoch: [30][800/1345], lr: 0.00100 Time 0.709 (0.739) Data 0.000 (0.004) Loss 0.9023 (1.0576) Prec@1 76.562 (71.485) Prec@5 95.312 (91.294) Epoch: [30][820/1345], lr: 0.00100 Time 0.819 (0.740) Data 0.000 (0.004) Loss 0.9524 (1.0579) Prec@1 73.438 (71.481) Prec@5 90.625 (91.297) Epoch: [30][840/1345], lr: 0.00100 Time 0.854 (0.740) Data 0.000 (0.004) Loss 0.8682 (1.0550) Prec@1 73.438 (71.541) Prec@5 95.312 (91.333) Epoch: [30][860/1345], lr: 0.00100 Time 0.714 (0.740) Data 0.000 (0.004) Loss 1.0671 (1.0522) Prec@1 68.750 (71.621) Prec@5 92.188 (91.373) Epoch: [30][880/1345], lr: 0.00100 Time 0.710 (0.740) Data 0.001 (0.004) Loss 0.5138 (1.0499) Prec@1 82.812 (71.683) Prec@5 96.875 (91.395) Epoch: [30][900/1345], lr: 0.00100 Time 0.711 (0.739) Data 0.001 (0.004) Loss 0.8753 (1.0471) Prec@1 73.438 (71.767) Prec@5 93.750 (91.418) Epoch: [30][920/1345], lr: 0.00100 Time 0.732 (0.740) Data 0.000 (0.004) Loss 1.0785 (1.0462) Prec@1 73.438 (71.778) Prec@5 90.625 (91.427) Epoch: [30][940/1345], lr: 0.00100 Time 0.748 (0.740) Data 0.000 (0.004) Loss 0.9212 (1.0436) Prec@1 71.875 (71.833) Prec@5 92.188 (91.447) Epoch: [30][960/1345], lr: 0.00100 Time 0.709 (0.740) Data 0.000 (0.004) Loss 1.2114 (1.0434) Prec@1 68.750 (71.878) Prec@5 85.938 (91.444) Epoch: [30][980/1345], lr: 0.00100 Time 0.708 (0.740) Data 0.000 (0.004) Loss 1.2492 (1.0416) Prec@1 71.875 (71.932) Prec@5 85.938 (91.456) Epoch: [30][1000/1345], lr: 0.00100 Time 0.716 (0.740) Data 0.001 (0.004) Loss 1.2153 (1.0403) Prec@1 65.625 (71.967) Prec@5 89.062 (91.463) Epoch: [30][1020/1345], lr: 0.00100 Time 0.835 (0.740) Data 0.000 (0.003) Loss 1.1114 (1.0377) Prec@1 65.625 (72.016) Prec@5 92.188 (91.497) Epoch: [30][1040/1345], lr: 0.00100 Time 0.710 (0.740) Data 0.000 (0.003) Loss 0.8826 (1.0372) Prec@1 76.562 (72.042) Prec@5 92.188 (91.497) Epoch: [30][1060/1345], lr: 0.00100 Time 0.707 (0.739) Data 0.000 (0.003) Loss 0.7968 (1.0353) Prec@1 76.562 (72.080) Prec@5 95.312 (91.523) Epoch: [30][1080/1345], lr: 0.00100 Time 0.707 (0.739) Data 0.000 (0.003) Loss 0.8862 (1.0340) Prec@1 79.688 (72.131) Prec@5 93.750 (91.534) Epoch: [30][1100/1345], lr: 0.00100 Time 0.710 (0.739) Data 0.000 (0.003) Loss 1.2020 (1.0331) Prec@1 67.188 (72.170) Prec@5 90.625 (91.536) Epoch: [30][1120/1345], lr: 0.00100 Time 0.791 (0.738) Data 0.000 (0.003) Loss 0.7899 (1.0315) Prec@1 81.250 (72.229) Prec@5 93.750 (91.544) Epoch: [30][1140/1345], lr: 0.00100 Time 0.707 (0.738) Data 0.000 (0.003) Loss 1.0104 (1.0301) Prec@1 65.625 (72.282) Prec@5 89.062 (91.555) Epoch: [30][1160/1345], lr: 0.00100 Time 0.707 (0.738) Data 0.000 (0.003) Loss 1.0168 (1.0287) Prec@1 68.750 (72.311) Prec@5 90.625 (91.567) Epoch: [30][1180/1345], lr: 0.00100 Time 0.707 (0.738) Data 0.000 (0.003) Loss 0.7019 (1.0269) Prec@1 76.562 (72.361) Prec@5 93.750 (91.578) Epoch: [30][1200/1345], lr: 0.00100 Time 0.708 (0.738) Data 0.000 (0.003) Loss 0.8223 (1.0250) Prec@1 79.688 (72.395) Prec@5 95.312 (91.597) Epoch: [30][1220/1345], lr: 0.00100 Time 0.708 (0.738) Data 0.000 (0.003) Loss 0.9641 (1.0238) Prec@1 76.562 (72.437) Prec@5 90.625 (91.598) Epoch: [30][1240/1345], lr: 0.00100 Time 0.760 (0.738) Data 0.001 (0.003) Loss 1.2392 (1.0232) Prec@1 71.875 (72.458) Prec@5 89.062 (91.612) Epoch: [30][1260/1345], lr: 0.00100 Time 0.707 (0.738) Data 0.000 (0.003) Loss 0.7352 (1.0218) Prec@1 79.688 (72.486) Prec@5 95.312 (91.636) Epoch: [30][1280/1345], lr: 0.00100 Time 0.709 (0.738) Data 0.000 (0.003) Loss 0.9488 (1.0208) Prec@1 70.312 (72.502) Prec@5 95.312 (91.643) Epoch: [30][1300/1345], lr: 0.00100 Time 0.705 (0.738) Data 0.000 (0.003) Loss 0.8063 (1.0190) Prec@1 82.812 (72.544) Prec@5 93.750 (91.673) Epoch: [30][1320/1345], lr: 0.00100 Time 0.708 (0.738) Data 0.000 (0.003) Loss 0.8991 (1.0180) Prec@1 73.438 (72.569) Prec@5 90.625 (91.691) Epoch: [30][1340/1345], lr: 0.00100 Time 0.782 (0.738) Data 0.000 (0.003) Loss 1.0062 (1.0176) Prec@1 73.438 (72.588) Prec@5 89.062 (91.689) No BN layer Freezing. Test: [0/181] Time 3.144 (3.1437) Loss 2.1666 (2.1666) Prec@1 57.812 (57.812) Prec@5 79.688 (79.688) Test: [20/181] Time 1.275 (0.6042) Loss 2.1959 (2.0321) Prec@1 45.312 (53.646) Prec@5 82.812 (80.580) Test: [40/181] Time 1.183 (0.5349) Loss 2.2278 (2.1182) Prec@1 51.562 (51.791) Prec@5 73.438 (79.383) Test: [60/181] Time 1.088 (0.5093) Loss 2.3456 (2.1607) Prec@1 45.312 (50.564) Prec@5 75.000 (78.919) Test: [80/181] Time 1.185 (0.4991) Loss 1.8803 (2.1835) Prec@1 64.062 (50.058) Prec@5 84.375 (78.877) Test: [100/181] Time 1.153 (0.4919) Loss 2.7960 (2.1674) Prec@1 45.312 (50.371) Prec@5 71.875 (79.131) Test: [120/181] Time 1.159 (0.4873) Loss 2.6985 (2.1766) Prec@1 46.875 (50.271) Prec@5 70.312 (79.171) Test: [140/181] Time 0.917 (0.4828) Loss 2.7701 (2.1734) Prec@1 43.750 (50.233) Prec@5 76.562 (79.322) Test: [160/181] Time 1.238 (0.4815) Loss 2.6103 (2.1641) Prec@1 46.875 (50.388) Prec@5 81.250 (79.435) Testing Results: Prec@1 50.625 Prec@5 79.462 Loss 2.15758 Time 0.4753 No BN layer Freezing. Epoch: [31][0/1345], lr: 0.00100 Time 4.565 (4.565) Data 3.826 (3.826) Loss 0.8942 (0.8942) Prec@1 81.250 (81.250) Prec@5 92.188 (92.188) Epoch: [31][20/1345], lr: 0.00100 Time 0.758 (0.928) Data 0.000 (0.183) Loss 1.2173 (0.9391) Prec@1 65.625 (75.000) Prec@5 90.625 (93.304) Epoch: [31][40/1345], lr: 0.00100 Time 0.712 (0.833) Data 0.000 (0.094) Loss 0.5889 (0.9231) Prec@1 85.938 (75.076) Prec@5 98.438 (92.988) Epoch: [31][60/1345], lr: 0.00100 Time 0.795 (0.802) Data 0.000 (0.063) Loss 1.0421 (0.9191) Prec@1 67.188 (74.872) Prec@5 92.188 (93.186) Epoch: [31][80/1345], lr: 0.00100 Time 0.713 (0.787) Data 0.000 (0.048) Loss 0.6196 (0.9229) Prec@1 84.375 (74.865) Prec@5 95.312 (93.056) Epoch: [31][100/1345], lr: 0.00100 Time 0.709 (0.777) Data 0.000 (0.038) Loss 0.9144 (0.9151) Prec@1 79.688 (75.278) Prec@5 93.750 (93.069) Epoch: [31][120/1345], lr: 0.00100 Time 0.705 (0.771) Data 0.000 (0.032) Loss 1.1141 (0.9241) Prec@1 71.875 (74.793) Prec@5 90.625 (93.079) Epoch: [31][140/1345], lr: 0.00100 Time 0.748 (0.767) Data 0.001 (0.028) Loss 0.5687 (0.9236) Prec@1 84.375 (74.634) Prec@5 98.438 (93.074) Epoch: [31][160/1345], lr: 0.00100 Time 0.738 (0.766) Data 0.000 (0.024) Loss 0.9516 (0.9228) Prec@1 71.875 (74.622) Prec@5 93.750 (93.003) Epoch: [31][180/1345], lr: 0.00100 Time 0.709 (0.763) Data 0.000 (0.022) Loss 0.5736 (0.9181) Prec@1 78.125 (74.741) Prec@5 98.438 (93.085) Epoch: [31][200/1345], lr: 0.00100 Time 0.711 (0.761) Data 0.000 (0.019) Loss 0.8801 (0.9165) Prec@1 70.312 (74.720) Prec@5 95.312 (93.050) Epoch: [31][220/1345], lr: 0.00100 Time 0.712 (0.760) Data 0.001 (0.018) Loss 0.7715 (0.9052) Prec@1 84.375 (75.007) Prec@5 90.625 (93.191) Epoch: [31][240/1345], lr: 0.00100 Time 0.870 (0.759) Data 0.000 (0.016) Loss 0.8087 (0.9018) Prec@1 78.125 (75.058) Prec@5 96.875 (93.316) Epoch: [31][260/1345], lr: 0.00100 Time 0.856 (0.758) Data 0.001 (0.015) Loss 0.7771 (0.9014) Prec@1 78.125 (75.036) Prec@5 93.750 (93.355) Epoch: [31][280/1345], lr: 0.00100 Time 0.765 (0.757) Data 0.001 (0.014) Loss 0.9389 (0.9009) Prec@1 71.875 (75.050) Prec@5 92.188 (93.333) Epoch: [31][300/1345], lr: 0.00100 Time 0.745 (0.757) Data 0.000 (0.013) Loss 0.9517 (0.9081) Prec@1 76.562 (74.964) Prec@5 90.625 (93.210) Epoch: [31][320/1345], lr: 0.00100 Time 0.749 (0.756) Data 0.000 (0.012) Loss 0.7975 (0.9062) Prec@1 79.688 (75.054) Prec@5 93.750 (93.210) Epoch: [31][340/1345], lr: 0.00100 Time 0.709 (0.755) Data 0.000 (0.012) Loss 1.1210 (0.9062) Prec@1 67.188 (75.110) Prec@5 92.188 (93.177) Epoch: [31][360/1345], lr: 0.00100 Time 0.735 (0.754) Data 0.000 (0.011) Loss 0.7472 (0.9092) Prec@1 84.375 (75.022) Prec@5 93.750 (93.148) Epoch: [31][380/1345], lr: 0.00100 Time 0.728 (0.754) Data 0.001 (0.010) Loss 1.1114 (0.9084) Prec@1 71.875 (75.045) Prec@5 85.938 (93.164) Epoch: [31][400/1345], lr: 0.00100 Time 0.953 (0.754) Data 0.000 (0.010) Loss 1.1304 (0.9077) Prec@1 67.188 (75.016) Prec@5 87.500 (93.189) Epoch: [31][420/1345], lr: 0.00100 Time 0.873 (0.754) Data 0.001 (0.010) Loss 1.1414 (0.9068) Prec@1 76.562 (75.078) Prec@5 89.062 (93.149) Epoch: [31][440/1345], lr: 0.00100 Time 0.741 (0.754) Data 0.000 (0.009) Loss 0.6754 (0.9066) Prec@1 76.562 (75.057) Prec@5 92.188 (93.134) Epoch: [31][460/1345], lr: 0.00100 Time 0.732 (0.754) Data 0.000 (0.009) Loss 0.5593 (0.9082) Prec@1 85.938 (75.037) Prec@5 98.438 (93.120) Epoch: [31][480/1345], lr: 0.00100 Time 0.737 (0.753) Data 0.001 (0.008) Loss 0.6251 (0.9077) Prec@1 84.375 (75.039) Prec@5 96.875 (93.126) Epoch: [31][500/1345], lr: 0.00100 Time 0.710 (0.753) Data 0.000 (0.008) Loss 0.8006 (0.9070) Prec@1 76.562 (75.044) Prec@5 93.750 (93.120) Epoch: [31][520/1345], lr: 0.00100 Time 0.708 (0.752) Data 0.000 (0.008) Loss 0.7947 (0.9067) Prec@1 75.000 (75.039) Prec@5 96.875 (93.123) Epoch: [31][540/1345], lr: 0.00100 Time 0.833 (0.752) Data 0.001 (0.008) Loss 0.6517 (0.9060) Prec@1 85.938 (75.098) Prec@5 96.875 (93.135) Epoch: [31][560/1345], lr: 0.00100 Time 0.840 (0.752) Data 0.000 (0.007) Loss 1.0590 (0.9039) Prec@1 70.312 (75.156) Prec@5 93.750 (93.176) Epoch: [31][580/1345], lr: 0.00100 Time 0.747 (0.751) Data 0.000 (0.007) Loss 0.7248 (0.9041) Prec@1 75.000 (75.143) Prec@5 98.438 (93.169) Epoch: [31][600/1345], lr: 0.00100 Time 0.769 (0.751) Data 0.001 (0.007) Loss 1.1032 (0.9048) Prec@1 70.312 (75.117) Prec@5 90.625 (93.152) Epoch: [31][620/1345], lr: 0.00100 Time 0.713 (0.751) Data 0.000 (0.007) Loss 1.1934 (0.9060) Prec@1 71.875 (75.108) Prec@5 89.062 (93.149) Epoch: [31][640/1345], lr: 0.00100 Time 0.754 (0.751) Data 0.000 (0.006) Loss 0.6455 (0.9055) Prec@1 81.250 (75.112) Prec@5 95.312 (93.158) Epoch: [31][660/1345], lr: 0.00100 Time 0.714 (0.751) Data 0.001 (0.006) Loss 0.8892 (0.9072) Prec@1 75.000 (75.059) Prec@5 90.625 (93.128) Epoch: [31][680/1345], lr: 0.00100 Time 0.720 (0.751) Data 0.000 (0.006) Loss 0.9064 (0.9069) Prec@1 73.438 (75.071) Prec@5 93.750 (93.131) Epoch: [31][700/1345], lr: 0.00100 Time 0.710 (0.750) Data 0.000 (0.006) Loss 1.0729 (0.9074) Prec@1 73.438 (75.047) Prec@5 92.188 (93.124) Epoch: [31][720/1345], lr: 0.00100 Time 0.769 (0.750) Data 0.000 (0.006) Loss 0.6758 (0.9054) Prec@1 81.250 (75.095) Prec@5 96.875 (93.130) Epoch: [31][740/1345], lr: 0.00100 Time 0.738 (0.749) Data 0.001 (0.006) Loss 0.9821 (0.9052) Prec@1 78.125 (75.143) Prec@5 89.062 (93.122) Epoch: [31][760/1345], lr: 0.00100 Time 0.710 (0.749) Data 0.000 (0.005) Loss 1.0224 (0.9049) Prec@1 70.312 (75.131) Prec@5 89.062 (93.140) Epoch: [31][780/1345], lr: 0.00100 Time 0.707 (0.748) Data 0.000 (0.005) Loss 1.0736 (0.9055) Prec@1 71.875 (75.102) Prec@5 90.625 (93.138) Epoch: [31][800/1345], lr: 0.00100 Time 0.856 (0.748) Data 0.001 (0.005) Loss 1.1682 (0.9062) Prec@1 68.750 (75.094) Prec@5 90.625 (93.122) Epoch: [31][820/1345], lr: 0.00100 Time 0.711 (0.748) Data 0.000 (0.005) Loss 0.4949 (0.9060) Prec@1 82.812 (75.076) Prec@5 98.438 (93.149) Epoch: [31][840/1345], lr: 0.00100 Time 0.724 (0.748) Data 0.000 (0.005) Loss 0.6627 (0.9057) Prec@1 84.375 (75.080) Prec@5 93.750 (93.141) Epoch: [31][860/1345], lr: 0.00100 Time 0.712 (0.747) Data 0.000 (0.005) Loss 1.0357 (0.9051) Prec@1 75.000 (75.122) Prec@5 92.188 (93.140) Epoch: [31][880/1345], lr: 0.00100 Time 0.707 (0.747) Data 0.000 (0.005) Loss 0.8768 (0.9025) Prec@1 73.438 (75.184) Prec@5 95.312 (93.163) Epoch: [31][900/1345], lr: 0.00100 Time 0.779 (0.748) Data 0.000 (0.005) Loss 0.9247 (0.9026) Prec@1 81.250 (75.222) Prec@5 89.062 (93.141) Epoch: [31][920/1345], lr: 0.00100 Time 0.708 (0.748) Data 0.000 (0.005) Loss 1.0952 (0.9036) Prec@1 73.438 (75.198) Prec@5 90.625 (93.119) Epoch: [31][940/1345], lr: 0.00100 Time 0.859 (0.748) Data 0.000 (0.005) Loss 0.8908 (0.9043) Prec@1 71.875 (75.153) Prec@5 92.188 (93.126) Epoch: [31][960/1345], lr: 0.00100 Time 0.829 (0.748) Data 0.000 (0.004) Loss 0.6482 (0.9046) Prec@1 82.812 (75.172) Prec@5 95.312 (93.129) Epoch: [31][980/1345], lr: 0.00100 Time 0.752 (0.748) Data 0.001 (0.004) Loss 0.7217 (0.9044) Prec@1 78.125 (75.177) Prec@5 98.438 (93.137) Epoch: [31][1000/1345], lr: 0.00100 Time 0.746 (0.748) Data 0.000 (0.004) Loss 1.3052 (0.9047) Prec@1 70.312 (75.186) Prec@5 87.500 (93.110) Epoch: [31][1020/1345], lr: 0.00100 Time 0.747 (0.747) Data 0.001 (0.004) Loss 0.9016 (0.9054) Prec@1 75.000 (75.167) Prec@5 90.625 (93.089) Epoch: [31][1040/1345], lr: 0.00100 Time 0.710 (0.747) Data 0.000 (0.004) Loss 0.9148 (0.9050) Prec@1 78.125 (75.188) Prec@5 93.750 (93.100) Epoch: [31][1060/1345], lr: 0.00100 Time 0.863 (0.747) Data 0.001 (0.004) Loss 1.1912 (0.9058) Prec@1 62.500 (75.147) Prec@5 87.500 (93.071) Epoch: [31][1080/1345], lr: 0.00100 Time 0.707 (0.747) Data 0.000 (0.004) Loss 1.0507 (0.9067) Prec@1 73.438 (75.133) Prec@5 93.750 (93.050) Epoch: [31][1100/1345], lr: 0.00100 Time 0.735 (0.747) Data 0.000 (0.004) Loss 0.9320 (0.9070) Prec@1 75.000 (75.145) Prec@5 93.750 (93.039) Epoch: [31][1120/1345], lr: 0.00100 Time 0.709 (0.747) Data 0.000 (0.004) Loss 0.9789 (0.9063) Prec@1 73.438 (75.167) Prec@5 92.188 (93.045) Epoch: [31][1140/1345], lr: 0.00100 Time 0.751 (0.747) Data 0.000 (0.004) Loss 0.8469 (0.9056) Prec@1 81.250 (75.201) Prec@5 93.750 (93.057) Epoch: [31][1160/1345], lr: 0.00100 Time 0.813 (0.747) Data 0.000 (0.004) Loss 0.6416 (0.9045) Prec@1 82.812 (75.257) Prec@5 93.750 (93.069) Epoch: [31][1180/1345], lr: 0.00100 Time 0.708 (0.747) Data 0.000 (0.004) Loss 1.0377 (0.9049) Prec@1 79.688 (75.279) Prec@5 90.625 (93.074) Epoch: [31][1200/1345], lr: 0.00100 Time 0.707 (0.746) Data 0.000 (0.004) Loss 1.0051 (0.9051) Prec@1 65.625 (75.269) Prec@5 95.312 (93.076) Epoch: [31][1220/1345], lr: 0.00100 Time 0.813 (0.746) Data 0.000 (0.004) Loss 0.6845 (0.9055) Prec@1 79.688 (75.251) Prec@5 96.875 (93.079) Epoch: [31][1240/1345], lr: 0.00100 Time 0.709 (0.746) Data 0.000 (0.004) Loss 0.9716 (0.9042) Prec@1 70.312 (75.274) Prec@5 93.750 (93.100) Epoch: [31][1260/1345], lr: 0.00100 Time 0.710 (0.745) Data 0.000 (0.003) Loss 0.7735 (0.9050) Prec@1 78.125 (75.265) Prec@5 96.875 (93.095) Epoch: [31][1280/1345], lr: 0.00100 Time 0.851 (0.745) Data 0.000 (0.003) Loss 0.8211 (0.9048) Prec@1 81.250 (75.271) Prec@5 96.875 (93.101) Epoch: [31][1300/1345], lr: 0.00100 Time 0.712 (0.745) Data 0.001 (0.003) Loss 0.8586 (0.9043) Prec@1 75.000 (75.281) Prec@5 93.750 (93.112) Epoch: [31][1320/1345], lr: 0.00100 Time 0.709 (0.745) Data 0.000 (0.003) Loss 0.8825 (0.9048) Prec@1 73.438 (75.268) Prec@5 95.312 (93.111) Epoch: [31][1340/1345], lr: 0.00100 Time 0.708 (0.745) Data 0.000 (0.003) Loss 1.5016 (0.9046) Prec@1 59.375 (75.283) Prec@5 89.062 (93.113) No BN layer Freezing. Test: [0/181] Time 3.622 (3.6218) Loss 2.1258 (2.1258) Prec@1 54.688 (54.688) Prec@5 79.688 (79.688) Test: [20/181] Time 0.777 (0.6123) Loss 2.2762 (2.0312) Prec@1 45.312 (53.497) Prec@5 79.688 (81.548) Test: [40/181] Time 1.018 (0.5359) Loss 2.2020 (2.0952) Prec@1 57.812 (52.134) Prec@5 71.875 (80.297) Test: [60/181] Time 1.014 (0.5126) Loss 2.3843 (2.1369) Prec@1 54.688 (51.076) Prec@5 75.000 (79.688) Test: [80/181] Time 0.813 (0.4977) Loss 1.8269 (2.1528) Prec@1 59.375 (50.463) Prec@5 84.375 (79.360) Test: [100/181] Time 0.525 (0.4881) Loss 2.7750 (2.1399) Prec@1 46.875 (50.541) Prec@5 68.750 (79.486) Test: [120/181] Time 0.541 (0.4838) Loss 2.6582 (2.1534) Prec@1 50.000 (50.400) Prec@5 75.000 (79.339) Test: [140/181] Time 0.236 (0.4800) Loss 3.0130 (2.1543) Prec@1 35.938 (50.144) Prec@5 75.000 (79.577) Test: [160/181] Time 0.599 (0.4792) Loss 2.4604 (2.1426) Prec@1 53.125 (50.359) Prec@5 79.688 (79.678) Testing Results: Prec@1 50.616 Prec@5 79.774 Loss 2.13539 Time 0.4764 No BN layer Freezing. Epoch: [32][0/1345], lr: 0.00100 Time 4.389 (4.389) Data 3.366 (3.366) Loss 0.8061 (0.8061) Prec@1 71.875 (71.875) Prec@5 93.750 (93.750) Epoch: [32][20/1345], lr: 0.00100 Time 0.784 (0.937) Data 0.001 (0.161) Loss 0.5036 (0.8144) Prec@1 92.188 (78.274) Prec@5 98.438 (94.048) Epoch: [32][40/1345], lr: 0.00100 Time 0.709 (0.849) Data 0.000 (0.083) Loss 0.9059 (0.8459) Prec@1 78.125 (77.515) Prec@5 93.750 (93.674) Epoch: [32][60/1345], lr: 0.00100 Time 0.709 (0.811) Data 0.000 (0.056) Loss 0.8744 (0.8495) Prec@1 82.812 (76.947) Prec@5 92.188 (93.776) Epoch: [32][80/1345], lr: 0.00100 Time 0.732 (0.794) Data 0.000 (0.042) Loss 0.9587 (0.8493) Prec@1 73.438 (76.910) Prec@5 92.188 (93.596) Epoch: [32][100/1345], lr: 0.00100 Time 0.728 (0.783) Data 0.000 (0.034) Loss 0.8374 (0.8473) Prec@1 76.562 (77.058) Prec@5 98.438 (93.657) Epoch: [32][120/1345], lr: 0.00100 Time 0.724 (0.778) Data 0.000 (0.028) Loss 0.8527 (0.8484) Prec@1 76.562 (76.976) Prec@5 92.188 (93.685) Epoch: [32][140/1345], lr: 0.00100 Time 0.708 (0.774) Data 0.000 (0.024) Loss 0.7782 (0.8462) Prec@1 79.688 (76.928) Prec@5 96.875 (93.717) Epoch: [32][160/1345], lr: 0.00100 Time 0.752 (0.770) Data 0.000 (0.021) Loss 0.7921 (0.8441) Prec@1 78.125 (76.922) Prec@5 95.312 (93.740) Epoch: [32][180/1345], lr: 0.00100 Time 0.707 (0.769) Data 0.001 (0.019) Loss 0.8264 (0.8535) Prec@1 73.438 (76.796) Prec@5 95.312 (93.603) Epoch: [32][200/1345], lr: 0.00100 Time 0.721 (0.767) Data 0.000 (0.017) Loss 0.6317 (0.8496) Prec@1 85.938 (77.006) Prec@5 93.750 (93.626) Epoch: [32][220/1345], lr: 0.00100 Time 0.855 (0.765) Data 0.001 (0.016) Loss 1.2424 (0.8493) Prec@1 65.625 (77.100) Prec@5 89.062 (93.623) Epoch: [32][240/1345], lr: 0.00100 Time 0.850 (0.764) Data 0.001 (0.014) Loss 1.1275 (0.8561) Prec@1 67.188 (76.900) Prec@5 89.062 (93.549) Epoch: [32][260/1345], lr: 0.00100 Time 0.730 (0.764) Data 0.001 (0.013) Loss 0.9283 (0.8574) Prec@1 68.750 (76.874) Prec@5 95.312 (93.534) Epoch: [32][280/1345], lr: 0.00100 Time 0.714 (0.763) Data 0.000 (0.012) Loss 0.7086 (0.8552) Prec@1 81.250 (76.980) Prec@5 98.438 (93.505) Epoch: [32][300/1345], lr: 0.00100 Time 0.713 (0.762) Data 0.000 (0.012) Loss 0.8018 (0.8548) Prec@1 76.562 (76.915) Prec@5 98.438 (93.548) Epoch: [32][320/1345], lr: 0.00100 Time 0.724 (0.761) Data 0.001 (0.011) Loss 0.9072 (0.8585) Prec@1 75.000 (76.757) Prec@5 90.625 (93.473) Epoch: [32][340/1345], lr: 0.00100 Time 0.711 (0.760) Data 0.001 (0.010) Loss 0.6492 (0.8582) Prec@1 79.688 (76.815) Prec@5 93.750 (93.438) Epoch: [32][360/1345], lr: 0.00100 Time 0.709 (0.760) Data 0.000 (0.010) Loss 0.7836 (0.8566) Prec@1 78.125 (76.887) Prec@5 96.875 (93.503) Epoch: [32][380/1345], lr: 0.00100 Time 0.928 (0.760) Data 0.000 (0.009) Loss 1.3641 (0.8581) Prec@1 57.812 (76.784) Prec@5 85.938 (93.475) Epoch: [32][400/1345], lr: 0.00100 Time 0.866 (0.760) Data 0.000 (0.009) Loss 0.7465 (0.8560) Prec@1 79.688 (76.785) Prec@5 93.750 (93.516) Epoch: [32][420/1345], lr: 0.00100 Time 0.750 (0.760) Data 0.000 (0.008) Loss 0.6996 (0.8556) Prec@1 81.250 (76.804) Prec@5 96.875 (93.509) Epoch: [32][440/1345], lr: 0.00100 Time 0.718 (0.760) Data 0.001 (0.008) Loss 0.7686 (0.8568) Prec@1 78.125 (76.803) Prec@5 95.312 (93.495) Epoch: [32][460/1345], lr: 0.00100 Time 0.712 (0.759) Data 0.000 (0.008) Loss 0.7446 (0.8602) Prec@1 84.375 (76.746) Prec@5 92.188 (93.448) Epoch: [32][480/1345], lr: 0.00100 Time 0.713 (0.760) Data 0.001 (0.007) Loss 0.8660 (0.8580) Prec@1 79.688 (76.790) Prec@5 96.875 (93.487) Epoch: [32][500/1345], lr: 0.00100 Time 0.722 (0.760) Data 0.001 (0.007) Loss 1.2957 (0.8589) Prec@1 57.812 (76.784) Prec@5 89.062 (93.454) Epoch: [32][520/1345], lr: 0.00100 Time 0.820 (0.759) Data 0.001 (0.007) Loss 0.9028 (0.8596) Prec@1 78.125 (76.724) Prec@5 92.188 (93.441) Epoch: [32][540/1345], lr: 0.00100 Time 0.713 (0.759) Data 0.001 (0.007) Loss 0.5305 (0.8588) Prec@1 85.938 (76.750) Prec@5 96.875 (93.432) Epoch: [32][560/1345], lr: 0.00100 Time 0.709 (0.758) Data 0.000 (0.006) Loss 0.7663 (0.8598) Prec@1 73.438 (76.699) Prec@5 93.750 (93.430) Epoch: [32][580/1345], lr: 0.00100 Time 0.814 (0.757) Data 0.000 (0.006) Loss 0.7769 (0.8566) Prec@1 76.562 (76.805) Prec@5 96.875 (93.470) Epoch: [32][600/1345], lr: 0.00100 Time 0.714 (0.757) Data 0.000 (0.006) Loss 0.6661 (0.8563) Prec@1 79.688 (76.802) Prec@5 95.312 (93.461) Epoch: [32][620/1345], lr: 0.00100 Time 0.710 (0.756) Data 0.000 (0.006) Loss 0.9491 (0.8577) Prec@1 73.438 (76.744) Prec@5 85.938 (93.446) Epoch: [32][640/1345], lr: 0.00100 Time 0.705 (0.756) Data 0.000 (0.006) Loss 1.0367 (0.8596) Prec@1 73.438 (76.711) Prec@5 90.625 (93.394) Epoch: [32][660/1345], lr: 0.00100 Time 0.721 (0.756) Data 0.001 (0.006) Loss 1.0964 (0.8578) Prec@1 73.438 (76.735) Prec@5 90.625 (93.419) Epoch: [32][680/1345], lr: 0.00100 Time 0.728 (0.755) Data 0.001 (0.005) Loss 0.6516 (0.8577) Prec@1 85.938 (76.762) Prec@5 95.312 (93.449) Epoch: [32][700/1345], lr: 0.00100 Time 0.832 (0.755) Data 0.001 (0.005) Loss 0.7204 (0.8565) Prec@1 78.125 (76.792) Prec@5 96.875 (93.469) Epoch: [32][720/1345], lr: 0.00100 Time 0.716 (0.755) Data 0.000 (0.005) Loss 0.9266 (0.8570) Prec@1 76.562 (76.790) Prec@5 89.062 (93.473) Epoch: [32][740/1345], lr: 0.00100 Time 0.718 (0.754) Data 0.001 (0.005) Loss 1.0623 (0.8572) Prec@1 70.312 (76.792) Prec@5 89.062 (93.472) Epoch: [32][760/1345], lr: 0.00100 Time 0.723 (0.754) Data 0.001 (0.005) Loss 1.0748 (0.8597) Prec@1 71.875 (76.735) Prec@5 89.062 (93.440) Epoch: [32][780/1345], lr: 0.00100 Time 0.715 (0.754) Data 0.000 (0.005) Loss 0.8123 (0.8607) Prec@1 75.000 (76.711) Prec@5 93.750 (93.418) Epoch: [32][800/1345], lr: 0.00100 Time 0.712 (0.754) Data 0.001 (0.005) Loss 0.6601 (0.8595) Prec@1 82.812 (76.742) Prec@5 95.312 (93.444) Epoch: [32][820/1345], lr: 0.00100 Time 0.853 (0.754) Data 0.000 (0.005) Loss 0.9540 (0.8593) Prec@1 75.000 (76.751) Prec@5 89.062 (93.440) Epoch: [32][840/1345], lr: 0.00100 Time 0.743 (0.754) Data 0.001 (0.004) Loss 0.8991 (0.8592) Prec@1 75.000 (76.739) Prec@5 89.062 (93.449) Epoch: [32][860/1345], lr: 0.00100 Time 0.711 (0.754) Data 0.000 (0.004) Loss 0.9079 (0.8579) Prec@1 79.688 (76.764) Prec@5 90.625 (93.487) Epoch: [32][880/1345], lr: 0.00100 Time 0.712 (0.753) Data 0.001 (0.004) Loss 0.5855 (0.8564) Prec@1 82.812 (76.800) Prec@5 98.438 (93.516) Epoch: [32][900/1345], lr: 0.00100 Time 0.725 (0.753) Data 0.000 (0.004) Loss 0.9040 (0.8562) Prec@1 79.688 (76.816) Prec@5 93.750 (93.519) Epoch: [32][920/1345], lr: 0.00100 Time 0.709 (0.753) Data 0.000 (0.004) Loss 0.9942 (0.8553) Prec@1 75.000 (76.831) Prec@5 89.062 (93.529) Epoch: [32][940/1345], lr: 0.00100 Time 0.720 (0.752) Data 0.000 (0.004) Loss 0.8577 (0.8548) Prec@1 75.000 (76.828) Prec@5 93.750 (93.527) Epoch: [32][960/1345], lr: 0.00100 Time 0.726 (0.752) Data 0.001 (0.004) Loss 0.9411 (0.8553) Prec@1 70.312 (76.808) Prec@5 90.625 (93.526) Epoch: [32][980/1345], lr: 0.00100 Time 0.717 (0.752) Data 0.001 (0.004) Loss 0.6389 (0.8545) Prec@1 79.688 (76.816) Prec@5 95.312 (93.527) Epoch: [32][1000/1345], lr: 0.00100 Time 0.713 (0.753) Data 0.000 (0.004) Loss 0.8771 (0.8534) Prec@1 76.562 (76.839) Prec@5 92.188 (93.546) Epoch: [32][1020/1345], lr: 0.00100 Time 0.709 (0.752) Data 0.000 (0.004) Loss 0.7302 (0.8538) Prec@1 79.688 (76.810) Prec@5 95.312 (93.553) Epoch: [32][1040/1345], lr: 0.00100 Time 0.830 (0.752) Data 0.000 (0.004) Loss 0.7653 (0.8554) Prec@1 73.438 (76.770) Prec@5 95.312 (93.543) Epoch: [32][1060/1345], lr: 0.00100 Time 0.737 (0.752) Data 0.000 (0.004) Loss 1.0222 (0.8549) Prec@1 70.312 (76.789) Prec@5 89.062 (93.551) Epoch: [32][1080/1345], lr: 0.00100 Time 0.742 (0.752) Data 0.001 (0.004) Loss 1.0540 (0.8558) Prec@1 68.750 (76.745) Prec@5 93.750 (93.552) Epoch: [32][1100/1345], lr: 0.00100 Time 0.721 (0.753) Data 0.000 (0.004) Loss 0.9901 (0.8566) Prec@1 78.125 (76.738) Prec@5 85.938 (93.539) Epoch: [32][1120/1345], lr: 0.00100 Time 0.735 (0.753) Data 0.001 (0.003) Loss 0.8262 (0.8563) Prec@1 75.000 (76.720) Prec@5 92.188 (93.546) Epoch: [32][1140/1345], lr: 0.00100 Time 0.762 (0.753) Data 0.000 (0.003) Loss 1.0378 (0.8574) Prec@1 73.438 (76.690) Prec@5 89.062 (93.530) Epoch: [32][1160/1345], lr: 0.00100 Time 0.711 (0.752) Data 0.000 (0.003) Loss 1.0715 (0.8567) Prec@1 71.875 (76.720) Prec@5 87.500 (93.533) Epoch: [32][1180/1345], lr: 0.00100 Time 0.726 (0.752) Data 0.001 (0.003) Loss 0.9203 (0.8563) Prec@1 75.000 (76.729) Prec@5 90.625 (93.544) Epoch: [32][1200/1345], lr: 0.00100 Time 0.861 (0.752) Data 0.000 (0.003) Loss 1.2511 (0.8563) Prec@1 64.062 (76.716) Prec@5 85.938 (93.551) Epoch: [32][1220/1345], lr: 0.00100 Time 0.714 (0.753) Data 0.000 (0.003) Loss 0.8393 (0.8567) Prec@1 79.688 (76.721) Prec@5 90.625 (93.544) Epoch: [32][1240/1345], lr: 0.00100 Time 0.779 (0.752) Data 0.001 (0.003) Loss 0.6899 (0.8565) Prec@1 81.250 (76.740) Prec@5 98.438 (93.551) Epoch: [32][1260/1345], lr: 0.00100 Time 0.710 (0.753) Data 0.000 (0.003) Loss 0.7414 (0.8557) Prec@1 81.250 (76.758) Prec@5 93.750 (93.557) Epoch: [32][1280/1345], lr: 0.00100 Time 0.837 (0.753) Data 0.001 (0.003) Loss 0.5199 (0.8553) Prec@1 87.500 (76.771) Prec@5 95.312 (93.572) Epoch: [32][1300/1345], lr: 0.00100 Time 0.712 (0.753) Data 0.000 (0.003) Loss 0.7407 (0.8548) Prec@1 79.688 (76.776) Prec@5 96.875 (93.571) Epoch: [32][1320/1345], lr: 0.00100 Time 0.716 (0.753) Data 0.001 (0.003) Loss 0.6739 (0.8547) Prec@1 84.375 (76.771) Prec@5 93.750 (93.569) Epoch: [32][1340/1345], lr: 0.00100 Time 0.721 (0.753) Data 0.001 (0.003) Loss 0.8896 (0.8552) Prec@1 78.125 (76.754) Prec@5 90.625 (93.569) No BN layer Freezing. Test: [0/181] Time 3.191 (3.1906) Loss 2.1427 (2.1427) Prec@1 54.688 (54.688) Prec@5 81.250 (81.250) Test: [20/181] Time 0.952 (0.6191) Loss 2.2612 (2.0453) Prec@1 48.438 (52.455) Prec@5 78.125 (80.580) Test: [40/181] Time 1.360 (0.5563) Loss 2.3001 (2.1051) Prec@1 56.250 (51.258) Prec@5 73.438 (79.688) Test: [60/181] Time 0.978 (0.5394) Loss 2.3355 (2.1447) Prec@1 51.562 (50.461) Prec@5 75.000 (79.559) Test: [80/181] Time 1.042 (0.5273) Loss 1.7967 (2.1585) Prec@1 62.500 (50.289) Prec@5 84.375 (79.398) Test: [100/181] Time 0.921 (0.5202) Loss 2.8262 (2.1486) Prec@1 43.750 (50.402) Prec@5 68.750 (79.858) Test: [120/181] Time 1.126 (0.5147) Loss 2.7355 (2.1622) Prec@1 48.438 (50.336) Prec@5 73.438 (79.907) Test: [140/181] Time 1.079 (0.5112) Loss 2.9659 (2.1602) Prec@1 35.938 (50.188) Prec@5 76.562 (80.031) Test: [160/181] Time 1.020 (0.5065) Loss 2.5368 (2.1490) Prec@1 50.000 (50.446) Prec@5 79.688 (80.085) Testing Results: Prec@1 50.634 Prec@5 80.165 Loss 2.14101 Time 0.5001 No BN layer Freezing. Epoch: [33][0/1345], lr: 0.00100 Time 4.393 (4.393) Data 3.623 (3.623) Loss 0.7501 (0.7501) Prec@1 84.375 (84.375) Prec@5 93.750 (93.750) Epoch: [33][20/1345], lr: 0.00100 Time 0.746 (0.922) Data 0.001 (0.173) Loss 0.5252 (0.7585) Prec@1 85.938 (78.720) Prec@5 96.875 (94.792) Epoch: [33][40/1345], lr: 0.00100 Time 0.713 (0.839) Data 0.000 (0.089) Loss 0.8047 (0.7974) Prec@1 78.125 (77.858) Prec@5 95.312 (94.169) Epoch: [33][60/1345], lr: 0.00100 Time 0.747 (0.817) Data 0.001 (0.060) Loss 0.9989 (0.8017) Prec@1 71.875 (78.074) Prec@5 93.750 (94.211) Epoch: [33][80/1345], lr: 0.00100 Time 0.742 (0.802) Data 0.001 (0.045) Loss 0.6461 (0.8143) Prec@1 84.375 (77.739) Prec@5 93.750 (94.078) Epoch: [33][100/1345], lr: 0.00100 Time 0.716 (0.789) Data 0.001 (0.036) Loss 0.6788 (0.8198) Prec@1 82.812 (77.568) Prec@5 95.312 (93.812) Epoch: [33][120/1345], lr: 0.00100 Time 0.711 (0.781) Data 0.000 (0.030) Loss 1.0982 (0.8162) Prec@1 71.875 (77.712) Prec@5 89.062 (93.853) Epoch: [33][140/1345], lr: 0.00100 Time 0.712 (0.774) Data 0.000 (0.026) Loss 1.1789 (0.8193) Prec@1 68.750 (77.593) Prec@5 87.500 (93.872) Epoch: [33][160/1345], lr: 0.00100 Time 0.707 (0.771) Data 0.000 (0.023) Loss 0.7738 (0.8180) Prec@1 81.250 (77.669) Prec@5 95.312 (93.944) Epoch: [33][180/1345], lr: 0.00100 Time 0.734 (0.766) Data 0.000 (0.020) Loss 1.0547 (0.8213) Prec@1 73.438 (77.607) Prec@5 90.625 (93.966) Epoch: [33][200/1345], lr: 0.00100 Time 0.711 (0.763) Data 0.000 (0.018) Loss 0.7340 (0.8163) Prec@1 78.125 (77.721) Prec@5 98.438 (93.999) Epoch: [33][220/1345], lr: 0.00100 Time 0.738 (0.761) Data 0.000 (0.017) Loss 0.8600 (0.8159) Prec@1 76.562 (77.687) Prec@5 92.188 (93.983) Epoch: [33][240/1345], lr: 0.00100 Time 0.712 (0.760) Data 0.000 (0.015) Loss 0.7568 (0.8162) Prec@1 75.000 (77.574) Prec@5 95.312 (94.003) Epoch: [33][260/1345], lr: 0.00100 Time 0.715 (0.758) Data 0.000 (0.014) Loss 0.9640 (0.8178) Prec@1 76.562 (77.532) Prec@5 90.625 (93.966) Epoch: [33][280/1345], lr: 0.00100 Time 0.706 (0.756) Data 0.000 (0.013) Loss 0.7851 (0.8152) Prec@1 81.250 (77.630) Prec@5 96.875 (94.039) Epoch: [33][300/1345], lr: 0.00100 Time 0.709 (0.754) Data 0.000 (0.012) Loss 0.7467 (0.8160) Prec@1 79.688 (77.575) Prec@5 95.312 (94.103) Epoch: [33][320/1345], lr: 0.00100 Time 0.854 (0.753) Data 0.001 (0.012) Loss 0.6562 (0.8129) Prec@1 81.250 (77.726) Prec@5 95.312 (94.120) Epoch: [33][340/1345], lr: 0.00100 Time 0.830 (0.753) Data 0.000 (0.011) Loss 0.8011 (0.8120) Prec@1 85.938 (77.758) Prec@5 92.188 (94.153) Epoch: [33][360/1345], lr: 0.00100 Time 0.744 (0.752) Data 0.000 (0.010) Loss 0.8372 (0.8087) Prec@1 76.562 (77.800) Prec@5 93.750 (94.209) Epoch: [33][380/1345], lr: 0.00100 Time 0.709 (0.752) Data 0.000 (0.010) Loss 0.4611 (0.8086) Prec@1 82.812 (77.772) Prec@5 100.000 (94.205) Epoch: [33][400/1345], lr: 0.00100 Time 0.749 (0.752) Data 0.000 (0.009) Loss 0.7406 (0.8091) Prec@1 81.250 (77.809) Prec@5 95.312 (94.210) Epoch: [33][420/1345], lr: 0.00100 Time 0.775 (0.751) Data 0.001 (0.009) Loss 0.6796 (0.8074) Prec@1 78.125 (77.791) Prec@5 96.875 (94.221) Epoch: [33][440/1345], lr: 0.00100 Time 0.706 (0.750) Data 0.000 (0.009) Loss 0.8915 (0.8063) Prec@1 78.125 (77.856) Prec@5 92.188 (94.221) Epoch: [33][460/1345], lr: 0.00100 Time 0.875 (0.749) Data 0.000 (0.008) Loss 0.8165 (0.8073) Prec@1 82.812 (77.830) Prec@5 95.312 (94.225) Epoch: [33][480/1345], lr: 0.00100 Time 0.711 (0.748) Data 0.000 (0.008) Loss 0.8987 (0.8073) Prec@1 75.000 (77.849) Prec@5 90.625 (94.224) Epoch: [33][500/1345], lr: 0.00100 Time 0.708 (0.747) Data 0.000 (0.008) Loss 1.0049 (0.8102) Prec@1 78.125 (77.801) Prec@5 92.188 (94.177) Epoch: [33][520/1345], lr: 0.00100 Time 0.710 (0.747) Data 0.000 (0.007) Loss 0.8659 (0.8092) Prec@1 76.562 (77.813) Prec@5 95.312 (94.179) Epoch: [33][540/1345], lr: 0.00100 Time 0.716 (0.746) Data 0.001 (0.007) Loss 0.9619 (0.8114) Prec@1 67.188 (77.709) Prec@5 98.438 (94.169) Epoch: [33][560/1345], lr: 0.00100 Time 0.766 (0.746) Data 0.000 (0.007) Loss 1.2804 (0.8134) Prec@1 62.500 (77.635) Prec@5 92.188 (94.165) Epoch: [33][580/1345], lr: 0.00100 Time 0.709 (0.746) Data 0.000 (0.007) Loss 0.7323 (0.8150) Prec@1 84.375 (77.576) Prec@5 96.875 (94.170) Epoch: [33][600/1345], lr: 0.00100 Time 0.706 (0.745) Data 0.000 (0.006) Loss 0.7007 (0.8174) Prec@1 79.688 (77.532) Prec@5 95.312 (94.104) Epoch: [33][620/1345], lr: 0.00100 Time 0.708 (0.744) Data 0.000 (0.006) Loss 0.8165 (0.8186) Prec@1 79.688 (77.511) Prec@5 90.625 (94.085) Epoch: [33][640/1345], lr: 0.00100 Time 0.707 (0.745) Data 0.000 (0.006) Loss 0.7975 (0.8181) Prec@1 70.312 (77.491) Prec@5 96.875 (94.089) Epoch: [33][660/1345], lr: 0.00100 Time 0.830 (0.745) Data 0.000 (0.006) Loss 0.9100 (0.8176) Prec@1 79.688 (77.515) Prec@5 89.062 (94.076) Epoch: [33][680/1345], lr: 0.00100 Time 0.707 (0.744) Data 0.000 (0.006) Loss 0.7329 (0.8173) Prec@1 75.000 (77.496) Prec@5 95.312 (94.094) Epoch: [33][700/1345], lr: 0.00100 Time 0.708 (0.744) Data 0.000 (0.006) Loss 0.5698 (0.8160) Prec@1 85.938 (77.510) Prec@5 96.875 (94.111) Epoch: [33][720/1345], lr: 0.00100 Time 0.802 (0.743) Data 0.000 (0.005) Loss 0.7979 (0.8159) Prec@1 75.000 (77.551) Prec@5 96.875 (94.116) Epoch: [33][740/1345], lr: 0.00100 Time 0.706 (0.743) Data 0.000 (0.005) Loss 0.7624 (0.8152) Prec@1 81.250 (77.564) Prec@5 96.875 (94.142) Epoch: [33][760/1345], lr: 0.00100 Time 0.707 (0.742) Data 0.000 (0.005) Loss 0.4778 (0.8149) Prec@1 89.062 (77.564) Prec@5 98.438 (94.144) Epoch: [33][780/1345], lr: 0.00100 Time 0.847 (0.742) Data 0.001 (0.005) Loss 0.9304 (0.8166) Prec@1 71.875 (77.517) Prec@5 93.750 (94.118) Epoch: [33][800/1345], lr: 0.00100 Time 0.707 (0.742) Data 0.000 (0.005) Loss 1.0018 (0.8175) Prec@1 73.438 (77.475) Prec@5 89.062 (94.105) Epoch: [33][820/1345], lr: 0.00100 Time 0.718 (0.742) Data 0.001 (0.005) Loss 0.8418 (0.8169) Prec@1 75.000 (77.516) Prec@5 96.875 (94.110) Epoch: [33][840/1345], lr: 0.00100 Time 0.710 (0.742) Data 0.000 (0.005) Loss 0.3814 (0.8154) Prec@1 87.500 (77.553) Prec@5 100.000 (94.135) Epoch: [33][860/1345], lr: 0.00100 Time 0.712 (0.742) Data 0.000 (0.005) Loss 0.6879 (0.8142) Prec@1 78.125 (77.597) Prec@5 95.312 (94.164) Epoch: [33][880/1345], lr: 0.00100 Time 0.708 (0.742) Data 0.000 (0.005) Loss 1.0217 (0.8150) Prec@1 75.000 (77.579) Prec@5 90.625 (94.146) Epoch: [33][900/1345], lr: 0.00100 Time 0.707 (0.742) Data 0.000 (0.004) Loss 0.8250 (0.8151) Prec@1 78.125 (77.575) Prec@5 93.750 (94.164) Epoch: [33][920/1345], lr: 0.00100 Time 0.776 (0.742) Data 0.000 (0.004) Loss 1.0238 (0.8159) Prec@1 73.438 (77.541) Prec@5 89.062 (94.130) Epoch: [33][940/1345], lr: 0.00100 Time 0.744 (0.742) Data 0.000 (0.004) Loss 0.5774 (0.8165) Prec@1 82.812 (77.527) Prec@5 98.438 (94.120) Epoch: [33][960/1345], lr: 0.00100 Time 0.708 (0.742) Data 0.000 (0.004) Loss 0.8509 (0.8173) Prec@1 76.562 (77.491) Prec@5 90.625 (94.126) Epoch: [33][980/1345], lr: 0.00100 Time 0.708 (0.741) Data 0.000 (0.004) Loss 0.8767 (0.8187) Prec@1 79.688 (77.456) Prec@5 95.312 (94.107) Epoch: [33][1000/1345], lr: 0.00100 Time 0.714 (0.741) Data 0.000 (0.004) Loss 0.5504 (0.8176) Prec@1 82.812 (77.449) Prec@5 98.438 (94.131) Epoch: [33][1020/1345], lr: 0.00100 Time 0.720 (0.741) Data 0.001 (0.004) Loss 0.4964 (0.8172) Prec@1 82.812 (77.458) Prec@5 98.438 (94.126) Epoch: [33][1040/1345], lr: 0.00100 Time 0.707 (0.741) Data 0.000 (0.004) Loss 0.7511 (0.8164) Prec@1 79.688 (77.477) Prec@5 93.750 (94.143) Epoch: [33][1060/1345], lr: 0.00100 Time 0.751 (0.741) Data 0.000 (0.004) Loss 0.9231 (0.8180) Prec@1 81.250 (77.455) Prec@5 92.188 (94.118) Epoch: [33][1080/1345], lr: 0.00100 Time 0.768 (0.741) Data 0.000 (0.004) Loss 0.9901 (0.8173) Prec@1 68.750 (77.459) Prec@5 90.625 (94.121) Epoch: [33][1100/1345], lr: 0.00100 Time 0.711 (0.741) Data 0.000 (0.004) Loss 0.7550 (0.8167) Prec@1 79.688 (77.464) Prec@5 92.188 (94.135) Epoch: [33][1120/1345], lr: 0.00100 Time 0.761 (0.741) Data 0.000 (0.004) Loss 0.8591 (0.8172) Prec@1 75.000 (77.446) Prec@5 95.312 (94.140) Epoch: [33][1140/1345], lr: 0.00100 Time 0.712 (0.741) Data 0.000 (0.004) Loss 0.7402 (0.8180) Prec@1 76.562 (77.439) Prec@5 95.312 (94.127) Epoch: [33][1160/1345], lr: 0.00100 Time 0.709 (0.741) Data 0.000 (0.004) Loss 1.3123 (0.8191) Prec@1 59.375 (77.406) Prec@5 85.938 (94.109) Epoch: [33][1180/1345], lr: 0.00100 Time 0.733 (0.741) Data 0.000 (0.003) Loss 0.7185 (0.8181) Prec@1 79.688 (77.430) Prec@5 96.875 (94.115) Epoch: [33][1200/1345], lr: 0.00100 Time 0.743 (0.741) Data 0.000 (0.003) Loss 0.7396 (0.8172) Prec@1 81.250 (77.435) Prec@5 95.312 (94.126) Epoch: [33][1220/1345], lr: 0.00100 Time 0.711 (0.741) Data 0.000 (0.003) Loss 0.7920 (0.8164) Prec@1 78.125 (77.474) Prec@5 96.875 (94.142) Epoch: [33][1240/1345], lr: 0.00100 Time 0.708 (0.741) Data 0.000 (0.003) Loss 0.6644 (0.8159) Prec@1 79.688 (77.480) Prec@5 95.312 (94.150) Epoch: [33][1260/1345], lr: 0.00100 Time 0.751 (0.741) Data 0.000 (0.003) Loss 0.8592 (0.8162) Prec@1 78.125 (77.489) Prec@5 93.750 (94.139) Epoch: [33][1280/1345], lr: 0.00100 Time 0.707 (0.741) Data 0.000 (0.003) Loss 0.6603 (0.8165) Prec@1 84.375 (77.483) Prec@5 95.312 (94.137) Epoch: [33][1300/1345], lr: 0.00100 Time 0.715 (0.741) Data 0.001 (0.003) Loss 0.8314 (0.8165) Prec@1 76.562 (77.488) Prec@5 96.875 (94.130) Epoch: [33][1320/1345], lr: 0.00100 Time 0.857 (0.741) Data 0.000 (0.003) Loss 0.5613 (0.8163) Prec@1 82.812 (77.495) Prec@5 98.438 (94.137) Epoch: [33][1340/1345], lr: 0.00100 Time 0.802 (0.741) Data 0.000 (0.003) Loss 0.8374 (0.8157) Prec@1 75.000 (77.509) Prec@5 95.312 (94.145) No BN layer Freezing. Test: [0/181] Time 3.401 (3.4006) Loss 2.1513 (2.1513) Prec@1 53.125 (53.125) Prec@5 84.375 (84.375) Test: [20/181] Time 1.140 (0.6145) Loss 2.3089 (2.0364) Prec@1 51.562 (52.976) Prec@5 75.000 (81.845) Test: [40/181] Time 1.299 (0.5388) Loss 2.3124 (2.1152) Prec@1 51.562 (51.258) Prec@5 75.000 (80.755) Test: [60/181] Time 0.874 (0.5067) Loss 2.3636 (2.1510) Prec@1 51.562 (50.410) Prec@5 75.000 (80.200) Test: [80/181] Time 0.679 (0.4913) Loss 1.8865 (2.1604) Prec@1 57.812 (50.212) Prec@5 84.375 (79.861) Test: [100/181] Time 0.459 (0.4839) Loss 2.8505 (2.1548) Prec@1 40.625 (50.402) Prec@5 71.875 (80.074) Test: [120/181] Time 0.691 (0.4812) Loss 2.6649 (2.1679) Prec@1 46.875 (50.504) Prec@5 68.750 (79.739) Test: [140/181] Time 0.802 (0.4790) Loss 2.9501 (2.1680) Prec@1 37.500 (50.321) Prec@5 79.688 (79.898) Test: [160/181] Time 0.804 (0.4768) Loss 2.4982 (2.1526) Prec@1 43.750 (50.631) Prec@5 81.250 (80.017) Testing Results: Prec@1 50.920 Prec@5 80.104 Loss 2.14278 Time 0.4718 No BN layer Freezing. Epoch: [34][0/1345], lr: 0.00100 Time 3.868 (3.868) Data 3.113 (3.113) Loss 0.7108 (0.7108) Prec@1 82.812 (82.812) Prec@5 95.312 (95.312) Epoch: [34][20/1345], lr: 0.00100 Time 0.834 (0.886) Data 0.001 (0.149) Loss 0.7826 (0.7479) Prec@1 75.000 (78.646) Prec@5 93.750 (95.238) Epoch: [34][40/1345], lr: 0.00100 Time 0.718 (0.824) Data 0.001 (0.076) Loss 0.6114 (0.7912) Prec@1 81.250 (78.011) Prec@5 98.438 (94.627) Epoch: [34][60/1345], lr: 0.00100 Time 0.714 (0.800) Data 0.001 (0.051) Loss 0.6059 (0.7884) Prec@1 89.062 (78.227) Prec@5 96.875 (94.390) Epoch: [34][80/1345], lr: 0.00100 Time 0.779 (0.787) Data 0.000 (0.039) Loss 0.8239 (0.7798) Prec@1 81.250 (78.453) Prec@5 92.188 (94.522) Epoch: [34][100/1345], lr: 0.00100 Time 0.711 (0.778) Data 0.000 (0.031) Loss 0.4848 (0.7758) Prec@1 87.500 (78.899) Prec@5 98.438 (94.462) Epoch: [34][120/1345], lr: 0.00100 Time 0.769 (0.773) Data 0.000 (0.026) Loss 0.6165 (0.7764) Prec@1 89.062 (78.822) Prec@5 93.750 (94.409) Epoch: [34][140/1345], lr: 0.00100 Time 0.888 (0.769) Data 0.000 (0.023) Loss 0.8678 (0.7735) Prec@1 75.000 (78.978) Prec@5 92.188 (94.393) Epoch: [34][160/1345], lr: 0.00100 Time 0.711 (0.765) Data 0.000 (0.020) Loss 0.7295 (0.7752) Prec@1 75.000 (78.911) Prec@5 96.875 (94.391) Epoch: [34][180/1345], lr: 0.00100 Time 0.736 (0.763) Data 0.000 (0.018) Loss 1.0220 (0.7789) Prec@1 73.438 (78.798) Prec@5 89.062 (94.372) Epoch: [34][200/1345], lr: 0.00100 Time 0.745 (0.761) Data 0.001 (0.016) Loss 0.7055 (0.7834) Prec@1 82.812 (78.840) Prec@5 95.312 (94.271) Epoch: [34][220/1345], lr: 0.00100 Time 0.710 (0.760) Data 0.000 (0.015) Loss 0.6216 (0.7786) Prec@1 82.812 (78.917) Prec@5 95.312 (94.415) Epoch: [34][240/1345], lr: 0.00100 Time 0.709 (0.757) Data 0.000 (0.013) Loss 0.8789 (0.7798) Prec@1 79.688 (78.916) Prec@5 93.750 (94.392) Epoch: [34][260/1345], lr: 0.00100 Time 0.773 (0.757) Data 0.001 (0.012) Loss 0.5563 (0.7763) Prec@1 85.938 (79.173) Prec@5 98.438 (94.415) Epoch: [34][280/1345], lr: 0.00100 Time 0.715 (0.756) Data 0.000 (0.012) Loss 0.7184 (0.7767) Prec@1 82.812 (79.131) Prec@5 93.750 (94.417) Epoch: [34][300/1345], lr: 0.00100 Time 0.750 (0.756) Data 0.000 (0.011) Loss 0.9559 (0.7773) Prec@1 76.562 (79.070) Prec@5 90.625 (94.425) Epoch: [34][320/1345], lr: 0.00100 Time 0.711 (0.756) Data 0.000 (0.010) Loss 0.5631 (0.7783) Prec@1 81.250 (79.040) Prec@5 98.438 (94.402) Epoch: [34][340/1345], lr: 0.00100 Time 0.709 (0.755) Data 0.001 (0.010) Loss 0.7296 (0.7784) Prec@1 81.250 (79.055) Prec@5 93.750 (94.401) Epoch: [34][360/1345], lr: 0.00100 Time 0.710 (0.754) Data 0.000 (0.009) Loss 0.7690 (0.7762) Prec@1 79.688 (79.133) Prec@5 95.312 (94.404) Epoch: [34][380/1345], lr: 0.00100 Time 0.708 (0.754) Data 0.000 (0.009) Loss 0.6782 (0.7771) Prec@1 81.250 (79.052) Prec@5 96.875 (94.402) Epoch: [34][400/1345], lr: 0.00100 Time 0.875 (0.754) Data 0.000 (0.008) Loss 0.5652 (0.7773) Prec@1 82.812 (79.006) Prec@5 96.875 (94.416) Epoch: [34][420/1345], lr: 0.00100 Time 0.712 (0.753) Data 0.000 (0.008) Loss 0.8777 (0.7763) Prec@1 73.438 (78.953) Prec@5 90.625 (94.451) Epoch: [34][440/1345], lr: 0.00100 Time 0.750 (0.752) Data 0.000 (0.007) Loss 0.5939 (0.7762) Prec@1 89.062 (79.000) Prec@5 92.188 (94.427) Epoch: [34][460/1345], lr: 0.00100 Time 0.716 (0.752) Data 0.000 (0.007) Loss 0.6085 (0.7768) Prec@1 89.062 (78.979) Prec@5 95.312 (94.435) Epoch: [34][480/1345], lr: 0.00100 Time 0.708 (0.752) Data 0.000 (0.007) Loss 0.9471 (0.7776) Prec@1 71.875 (78.937) Prec@5 95.312 (94.484) Epoch: [34][500/1345], lr: 0.00100 Time 0.855 (0.751) Data 0.001 (0.007) Loss 0.8367 (0.7798) Prec@1 73.438 (78.877) Prec@5 89.062 (94.452) Epoch: [34][520/1345], lr: 0.00100 Time 0.709 (0.751) Data 0.000 (0.006) Loss 0.6718 (0.7789) Prec@1 85.938 (78.884) Prec@5 95.312 (94.449) Epoch: [34][540/1345], lr: 0.00100 Time 0.707 (0.750) Data 0.000 (0.006) Loss 0.9166 (0.7786) Prec@1 76.562 (78.899) Prec@5 95.312 (94.443) Epoch: [34][560/1345], lr: 0.00100 Time 0.841 (0.750) Data 0.000 (0.006) Loss 0.8232 (0.7804) Prec@1 76.562 (78.846) Prec@5 90.625 (94.416) Epoch: [34][580/1345], lr: 0.00100 Time 0.741 (0.749) Data 0.000 (0.006) Loss 0.6707 (0.7795) Prec@1 76.562 (78.832) Prec@5 95.312 (94.430) Epoch: [34][600/1345], lr: 0.00100 Time 0.708 (0.749) Data 0.000 (0.006) Loss 0.7752 (0.7805) Prec@1 73.438 (78.788) Prec@5 96.875 (94.408) Epoch: [34][620/1345], lr: 0.00100 Time 0.775 (0.750) Data 0.000 (0.005) Loss 0.7563 (0.7825) Prec@1 78.125 (78.709) Prec@5 95.312 (94.394) Epoch: [34][640/1345], lr: 0.00100 Time 0.752 (0.750) Data 0.000 (0.005) Loss 0.4930 (0.7832) Prec@1 82.812 (78.666) Prec@5 98.438 (94.391) Epoch: [34][660/1345], lr: 0.00100 Time 0.715 (0.750) Data 0.000 (0.005) Loss 0.6570 (0.7831) Prec@1 82.812 (78.652) Prec@5 95.312 (94.386) Epoch: [34][680/1345], lr: 0.00100 Time 0.709 (0.749) Data 0.000 (0.005) Loss 1.1545 (0.7866) Prec@1 67.188 (78.545) Prec@5 90.625 (94.347) Epoch: [34][700/1345], lr: 0.00100 Time 0.711 (0.749) Data 0.000 (0.005) Loss 0.7765 (0.7866) Prec@1 79.688 (78.564) Prec@5 93.750 (94.350) Epoch: [34][720/1345], lr: 0.00100 Time 0.709 (0.749) Data 0.000 (0.005) Loss 0.6711 (0.7865) Prec@1 79.688 (78.595) Prec@5 98.438 (94.342) Epoch: [34][740/1345], lr: 0.00100 Time 0.754 (0.749) Data 0.000 (0.005) Loss 1.1545 (0.7884) Prec@1 71.875 (78.509) Prec@5 87.500 (94.309) Epoch: [34][760/1345], lr: 0.00100 Time 0.718 (0.749) Data 0.000 (0.005) Loss 0.6932 (0.7885) Prec@1 81.250 (78.497) Prec@5 93.750 (94.313) Epoch: [34][780/1345], lr: 0.00100 Time 0.755 (0.749) Data 0.000 (0.004) Loss 0.8499 (0.7872) Prec@1 78.125 (78.561) Prec@5 92.188 (94.328) Epoch: [34][800/1345], lr: 0.00100 Time 0.740 (0.748) Data 0.000 (0.004) Loss 0.5499 (0.7873) Prec@1 87.500 (78.574) Prec@5 98.438 (94.331) Epoch: [34][820/1345], lr: 0.00100 Time 0.720 (0.749) Data 0.000 (0.004) Loss 0.8307 (0.7874) Prec@1 76.562 (78.568) Prec@5 93.750 (94.336) Epoch: [34][840/1345], lr: 0.00100 Time 0.759 (0.748) Data 0.001 (0.004) Loss 0.7515 (0.7879) Prec@1 82.812 (78.560) Prec@5 95.312 (94.315) Epoch: [34][860/1345], lr: 0.00100 Time 0.709 (0.748) Data 0.000 (0.004) Loss 0.8065 (0.7878) Prec@1 76.562 (78.577) Prec@5 96.875 (94.302) Epoch: [34][880/1345], lr: 0.00100 Time 0.710 (0.748) Data 0.000 (0.004) Loss 0.8130 (0.7879) Prec@1 71.875 (78.563) Prec@5 93.750 (94.305) Epoch: [34][900/1345], lr: 0.00100 Time 0.713 (0.748) Data 0.000 (0.004) Loss 0.6742 (0.7868) Prec@1 79.688 (78.595) Prec@5 96.875 (94.308) Epoch: [34][920/1345], lr: 0.00100 Time 0.729 (0.748) Data 0.000 (0.004) Loss 0.6616 (0.7871) Prec@1 85.938 (78.593) Prec@5 93.750 (94.305) Epoch: [34][940/1345], lr: 0.00100 Time 0.714 (0.748) Data 0.000 (0.004) Loss 0.8216 (0.7870) Prec@1 78.125 (78.621) Prec@5 90.625 (94.295) Epoch: [34][960/1345], lr: 0.00100 Time 0.712 (0.748) Data 0.000 (0.004) Loss 0.5769 (0.7880) Prec@1 79.688 (78.575) Prec@5 96.875 (94.277) Epoch: [34][980/1345], lr: 0.00100 Time 0.707 (0.748) Data 0.000 (0.004) Loss 1.2888 (0.7887) Prec@1 71.875 (78.533) Prec@5 84.375 (94.261) Epoch: [34][1000/1345], lr: 0.00100 Time 0.753 (0.748) Data 0.000 (0.004) Loss 1.0088 (0.7891) Prec@1 73.438 (78.525) Prec@5 90.625 (94.257) Epoch: [34][1020/1345], lr: 0.00100 Time 0.718 (0.748) Data 0.001 (0.003) Loss 0.7587 (0.7897) Prec@1 79.688 (78.538) Prec@5 95.312 (94.252) Epoch: [34][1040/1345], lr: 0.00100 Time 0.737 (0.748) Data 0.000 (0.003) Loss 0.8502 (0.7886) Prec@1 71.875 (78.539) Prec@5 95.312 (94.253) Epoch: [34][1060/1345], lr: 0.00100 Time 0.717 (0.748) Data 0.000 (0.003) Loss 0.7597 (0.7888) Prec@1 79.688 (78.533) Prec@5 96.875 (94.262) Epoch: [34][1080/1345], lr: 0.00100 Time 0.709 (0.747) Data 0.000 (0.003) Loss 0.6581 (0.7900) Prec@1 81.250 (78.499) Prec@5 98.438 (94.270) Epoch: [34][1100/1345], lr: 0.00100 Time 0.754 (0.747) Data 0.000 (0.003) Loss 0.9023 (0.7895) Prec@1 82.812 (78.545) Prec@5 93.750 (94.267) Epoch: [34][1120/1345], lr: 0.00100 Time 0.778 (0.747) Data 0.000 (0.003) Loss 1.0051 (0.7897) Prec@1 68.750 (78.529) Prec@5 90.625 (94.275) Epoch: [34][1140/1345], lr: 0.00100 Time 0.713 (0.747) Data 0.000 (0.003) Loss 0.9669 (0.7897) Prec@1 71.875 (78.536) Prec@5 95.312 (94.269) Epoch: [34][1160/1345], lr: 0.00100 Time 0.715 (0.747) Data 0.000 (0.003) Loss 0.7377 (0.7896) Prec@1 81.250 (78.540) Prec@5 92.188 (94.279) Epoch: [34][1180/1345], lr: 0.00100 Time 0.708 (0.747) Data 0.000 (0.003) Loss 0.7410 (0.7897) Prec@1 79.688 (78.550) Prec@5 95.312 (94.281) Epoch: [34][1200/1345], lr: 0.00100 Time 0.745 (0.747) Data 0.000 (0.003) Loss 0.8051 (0.7897) Prec@1 79.688 (78.560) Prec@5 93.750 (94.290) Epoch: [34][1220/1345], lr: 0.00100 Time 0.710 (0.747) Data 0.000 (0.003) Loss 0.6410 (0.7894) Prec@1 79.688 (78.563) Prec@5 95.312 (94.290) Epoch: [34][1240/1345], lr: 0.00100 Time 0.849 (0.747) Data 0.000 (0.003) Loss 0.8605 (0.7894) Prec@1 73.438 (78.547) Prec@5 95.312 (94.301) Epoch: [34][1260/1345], lr: 0.00100 Time 0.841 (0.747) Data 0.000 (0.003) Loss 0.7161 (0.7889) Prec@1 82.812 (78.561) Prec@5 93.750 (94.314) Epoch: [34][1280/1345], lr: 0.00100 Time 0.764 (0.747) Data 0.001 (0.003) Loss 1.0774 (0.7890) Prec@1 73.438 (78.557) Prec@5 85.938 (94.293) Epoch: [34][1300/1345], lr: 0.00100 Time 0.716 (0.747) Data 0.000 (0.003) Loss 0.8823 (0.7886) Prec@1 71.875 (78.574) Prec@5 93.750 (94.299) Epoch: [34][1320/1345], lr: 0.00100 Time 0.736 (0.747) Data 0.000 (0.003) Loss 0.7461 (0.7881) Prec@1 79.688 (78.592) Prec@5 95.312 (94.309) Epoch: [34][1340/1345], lr: 0.00100 Time 0.709 (0.747) Data 0.000 (0.003) Loss 0.7836 (0.7887) Prec@1 71.875 (78.560) Prec@5 93.750 (94.303) No BN layer Freezing. Test: [0/181] Time 2.868 (2.8678) Loss 2.2242 (2.2242) Prec@1 51.562 (51.562) Prec@5 82.812 (82.812) Test: [20/181] Time 1.407 (0.6043) Loss 2.2562 (2.1008) Prec@1 50.000 (52.232) Prec@5 78.125 (80.506) Test: [40/181] Time 0.986 (0.5391) Loss 2.5995 (2.1807) Prec@1 53.125 (51.334) Prec@5 71.875 (79.878) Test: [60/181] Time 0.886 (0.5103) Loss 2.5111 (2.2183) Prec@1 53.125 (50.640) Prec@5 75.000 (79.636) Test: [80/181] Time 0.780 (0.4990) Loss 1.8788 (2.2301) Prec@1 59.375 (50.367) Prec@5 85.938 (79.147) Test: [100/181] Time 1.024 (0.4924) Loss 2.8320 (2.2190) Prec@1 48.438 (50.464) Prec@5 73.438 (79.440) Test: [120/181] Time 0.953 (0.4873) Loss 2.7704 (2.2324) Prec@1 45.312 (50.323) Prec@5 75.000 (79.390) Test: [140/181] Time 0.747 (0.4808) Loss 3.0962 (2.2294) Prec@1 40.625 (50.188) Prec@5 73.438 (79.488) Test: [160/181] Time 0.689 (0.4784) Loss 2.5489 (2.2164) Prec@1 48.438 (50.446) Prec@5 78.125 (79.581) Testing Results: Prec@1 50.747 Prec@5 79.566 Loss 2.21007 Time 0.4744 No BN layer Freezing. Epoch: [35][0/1345], lr: 0.00100 Time 4.542 (4.542) Data 3.567 (3.567) Loss 0.8292 (0.8292) Prec@1 78.125 (78.125) Prec@5 95.312 (95.312) Epoch: [35][20/1345], lr: 0.00100 Time 0.733 (0.940) Data 0.000 (0.170) Loss 0.7640 (0.7693) Prec@1 82.812 (79.762) Prec@5 95.312 (95.015) Epoch: [35][40/1345], lr: 0.00100 Time 0.776 (0.853) Data 0.001 (0.088) Loss 0.7443 (0.7555) Prec@1 79.688 (79.345) Prec@5 95.312 (94.931) Epoch: [35][60/1345], lr: 0.00100 Time 0.711 (0.819) Data 0.000 (0.059) Loss 0.6397 (0.7483) Prec@1 76.562 (79.278) Prec@5 100.000 (95.005) Epoch: [35][80/1345], lr: 0.00100 Time 0.717 (0.802) Data 0.001 (0.045) Loss 0.7069 (0.7283) Prec@1 75.000 (79.630) Prec@5 98.438 (95.158) Epoch: [35][100/1345], lr: 0.00100 Time 0.714 (0.791) Data 0.000 (0.036) Loss 0.8031 (0.7497) Prec@1 75.000 (78.945) Prec@5 93.750 (94.817) Epoch: [35][120/1345], lr: 0.00100 Time 0.715 (0.786) Data 0.000 (0.030) Loss 0.7167 (0.7490) Prec@1 78.125 (79.132) Prec@5 95.312 (94.757) Epoch: [35][140/1345], lr: 0.00100 Time 0.730 (0.783) Data 0.000 (0.026) Loss 1.0015 (0.7635) Prec@1 70.312 (78.845) Prec@5 90.625 (94.592) Epoch: [35][160/1345], lr: 0.00100 Time 0.871 (0.780) Data 0.000 (0.023) Loss 0.8652 (0.7601) Prec@1 71.875 (78.950) Prec@5 95.312 (94.691) Epoch: [35][180/1345], lr: 0.00100 Time 0.722 (0.776) Data 0.001 (0.020) Loss 0.6421 (0.7552) Prec@1 84.375 (79.049) Prec@5 93.750 (94.700) Epoch: [35][200/1345], lr: 0.00100 Time 0.712 (0.773) Data 0.000 (0.018) Loss 0.7318 (0.7620) Prec@1 85.938 (78.895) Prec@5 95.312 (94.667) Epoch: [35][220/1345], lr: 0.00100 Time 0.887 (0.769) Data 0.000 (0.017) Loss 0.7809 (0.7638) Prec@1 78.125 (78.853) Prec@5 98.438 (94.627) Epoch: [35][240/1345], lr: 0.00100 Time 0.745 (0.767) Data 0.000 (0.015) Loss 0.8350 (0.7600) Prec@1 75.000 (79.046) Prec@5 95.312 (94.645) Epoch: [35][260/1345], lr: 0.00100 Time 0.712 (0.765) Data 0.000 (0.014) Loss 0.6459 (0.7646) Prec@1 82.812 (78.999) Prec@5 95.312 (94.594) Epoch: [35][280/1345], lr: 0.00100 Time 0.714 (0.764) Data 0.001 (0.013) Loss 0.8574 (0.7666) Prec@1 76.562 (78.876) Prec@5 93.750 (94.645) Epoch: [35][300/1345], lr: 0.00100 Time 0.714 (0.763) Data 0.000 (0.012) Loss 0.7636 (0.7656) Prec@1 79.688 (78.966) Prec@5 92.188 (94.586) Epoch: [35][320/1345], lr: 0.00100 Time 0.710 (0.762) Data 0.000 (0.012) Loss 0.8771 (0.7666) Prec@1 82.812 (78.957) Prec@5 90.625 (94.592) Epoch: [35][340/1345], lr: 0.00100 Time 0.758 (0.762) Data 0.001 (0.011) Loss 0.6517 (0.7675) Prec@1 82.812 (78.973) Prec@5 96.875 (94.534) Epoch: [35][360/1345], lr: 0.00100 Time 0.725 (0.762) Data 0.000 (0.010) Loss 0.4107 (0.7653) Prec@1 87.500 (79.021) Prec@5 100.000 (94.546) Epoch: [35][380/1345], lr: 0.00100 Time 0.758 (0.761) Data 0.001 (0.010) Loss 0.6857 (0.7658) Prec@1 79.688 (78.917) Prec@5 96.875 (94.570) Epoch: [35][400/1345], lr: 0.00100 Time 0.712 (0.761) Data 0.000 (0.009) Loss 1.0030 (0.7642) Prec@1 78.125 (78.943) Prec@5 87.500 (94.572) Epoch: [35][420/1345], lr: 0.00100 Time 0.714 (0.760) Data 0.000 (0.009) Loss 0.5574 (0.7664) Prec@1 85.938 (78.890) Prec@5 93.750 (94.518) Epoch: [35][440/1345], lr: 0.00100 Time 0.732 (0.760) Data 0.001 (0.009) Loss 0.9051 (0.7688) Prec@1 73.438 (78.809) Prec@5 95.312 (94.494) Epoch: [35][460/1345], lr: 0.00100 Time 0.712 (0.759) Data 0.000 (0.008) Loss 0.6762 (0.7695) Prec@1 82.812 (78.752) Prec@5 95.312 (94.519) Epoch: [35][480/1345], lr: 0.00100 Time 0.751 (0.759) Data 0.001 (0.008) Loss 0.7379 (0.7666) Prec@1 76.562 (78.788) Prec@5 93.750 (94.562) Epoch: [35][500/1345], lr: 0.00100 Time 0.719 (0.758) Data 0.000 (0.008) Loss 0.5792 (0.7659) Prec@1 85.938 (78.836) Prec@5 96.875 (94.552) Epoch: [35][520/1345], lr: 0.00100 Time 0.717 (0.758) Data 0.001 (0.007) Loss 0.7208 (0.7649) Prec@1 78.125 (78.848) Prec@5 95.312 (94.560) Epoch: [35][540/1345], lr: 0.00100 Time 0.714 (0.758) Data 0.000 (0.007) Loss 0.8533 (0.7660) Prec@1 71.875 (78.815) Prec@5 93.750 (94.547) Epoch: [35][560/1345], lr: 0.00100 Time 0.762 (0.758) Data 0.001 (0.007) Loss 1.1295 (0.7689) Prec@1 68.750 (78.707) Prec@5 87.500 (94.513) Epoch: [35][580/1345], lr: 0.00100 Time 0.713 (0.758) Data 0.000 (0.007) Loss 0.7988 (0.7723) Prec@1 84.375 (78.639) Prec@5 98.438 (94.508) Epoch: [35][600/1345], lr: 0.00100 Time 0.734 (0.757) Data 0.000 (0.006) Loss 0.9116 (0.7709) Prec@1 75.000 (78.668) Prec@5 93.750 (94.540) Epoch: [35][620/1345], lr: 0.00100 Time 0.712 (0.757) Data 0.001 (0.006) Loss 0.5657 (0.7680) Prec@1 84.375 (78.729) Prec@5 98.438 (94.588) Epoch: [35][640/1345], lr: 0.00100 Time 0.752 (0.757) Data 0.001 (0.006) Loss 0.6380 (0.7697) Prec@1 82.812 (78.688) Prec@5 98.438 (94.562) Epoch: [35][660/1345], lr: 0.00100 Time 0.715 (0.757) Data 0.000 (0.006) Loss 0.7481 (0.7707) Prec@1 73.438 (78.690) Prec@5 96.875 (94.525) Epoch: [35][680/1345], lr: 0.00100 Time 0.741 (0.757) Data 0.000 (0.006) Loss 0.5441 (0.7698) Prec@1 84.375 (78.710) Prec@5 95.312 (94.551) Epoch: [35][700/1345], lr: 0.00100 Time 0.714 (0.756) Data 0.000 (0.006) Loss 0.7984 (0.7701) Prec@1 71.875 (78.682) Prec@5 95.312 (94.539) Epoch: [35][720/1345], lr: 0.00100 Time 0.714 (0.756) Data 0.000 (0.005) Loss 0.8600 (0.7698) Prec@1 75.000 (78.706) Prec@5 95.312 (94.548) Epoch: [35][740/1345], lr: 0.00100 Time 0.715 (0.756) Data 0.000 (0.005) Loss 0.7702 (0.7701) Prec@1 81.250 (78.713) Prec@5 98.438 (94.539) Epoch: [35][760/1345], lr: 0.00100 Time 0.710 (0.755) Data 0.000 (0.005) Loss 0.6601 (0.7714) Prec@1 85.938 (78.706) Prec@5 95.312 (94.520) Epoch: [35][780/1345], lr: 0.00100 Time 0.713 (0.755) Data 0.001 (0.005) Loss 0.8253 (0.7710) Prec@1 73.438 (78.705) Prec@5 96.875 (94.524) Epoch: [35][800/1345], lr: 0.00100 Time 0.712 (0.755) Data 0.001 (0.005) Loss 0.5277 (0.7706) Prec@1 81.250 (78.714) Prec@5 96.875 (94.536) Epoch: [35][820/1345], lr: 0.00100 Time 0.714 (0.755) Data 0.000 (0.005) Loss 0.7646 (0.7718) Prec@1 81.250 (78.696) Prec@5 92.188 (94.519) Epoch: [35][840/1345], lr: 0.00100 Time 0.767 (0.755) Data 0.001 (0.005) Loss 0.8052 (0.7708) Prec@1 79.688 (78.708) Prec@5 90.625 (94.523) Epoch: [35][860/1345], lr: 0.00100 Time 0.730 (0.755) Data 0.000 (0.005) Loss 0.5312 (0.7699) Prec@1 84.375 (78.738) Prec@5 96.875 (94.538) Epoch: [35][880/1345], lr: 0.00100 Time 0.713 (0.755) Data 0.000 (0.005) Loss 0.6286 (0.7698) Prec@1 81.250 (78.714) Prec@5 98.438 (94.537) Epoch: [35][900/1345], lr: 0.00100 Time 0.745 (0.755) Data 0.001 (0.004) Loss 0.5832 (0.7708) Prec@1 78.125 (78.685) Prec@5 98.438 (94.530) Epoch: [35][920/1345], lr: 0.00100 Time 0.722 (0.755) Data 0.000 (0.004) Loss 0.8595 (0.7718) Prec@1 76.562 (78.639) Prec@5 95.312 (94.530) Epoch: [35][940/1345], lr: 0.00100 Time 0.779 (0.755) Data 0.000 (0.004) Loss 0.4763 (0.7719) Prec@1 84.375 (78.638) Prec@5 96.875 (94.530) Epoch: [35][960/1345], lr: 0.00100 Time 0.711 (0.755) Data 0.000 (0.004) Loss 0.8568 (0.7710) Prec@1 76.562 (78.686) Prec@5 92.188 (94.537) Epoch: [35][980/1345], lr: 0.00100 Time 0.739 (0.755) Data 0.001 (0.004) Loss 0.7949 (0.7708) Prec@1 81.250 (78.689) Prec@5 96.875 (94.551) Epoch: [35][1000/1345], lr: 0.00100 Time 0.715 (0.755) Data 0.000 (0.004) Loss 0.7479 (0.7715) Prec@1 81.250 (78.671) Prec@5 96.875 (94.555) Epoch: [35][1020/1345], lr: 0.00100 Time 0.751 (0.755) Data 0.000 (0.004) Loss 0.7621 (0.7715) Prec@1 81.250 (78.673) Prec@5 93.750 (94.567) Epoch: [35][1040/1345], lr: 0.00100 Time 0.754 (0.755) Data 0.000 (0.004) Loss 0.8392 (0.7708) Prec@1 79.688 (78.691) Prec@5 93.750 (94.568) Epoch: [35][1060/1345], lr: 0.00100 Time 0.708 (0.754) Data 0.000 (0.004) Loss 0.9304 (0.7719) Prec@1 79.688 (78.661) Prec@5 90.625 (94.557) Epoch: [35][1080/1345], lr: 0.00100 Time 0.714 (0.754) Data 0.000 (0.004) Loss 0.9639 (0.7730) Prec@1 76.562 (78.641) Prec@5 93.750 (94.554) Epoch: [35][1100/1345], lr: 0.00100 Time 0.719 (0.754) Data 0.000 (0.004) Loss 0.6388 (0.7734) Prec@1 82.812 (78.612) Prec@5 95.312 (94.569) Epoch: [35][1120/1345], lr: 0.00100 Time 0.712 (0.754) Data 0.000 (0.004) Loss 0.7191 (0.7730) Prec@1 78.125 (78.596) Prec@5 93.750 (94.574) Epoch: [35][1140/1345], lr: 0.00100 Time 0.725 (0.754) Data 0.001 (0.004) Loss 0.8831 (0.7726) Prec@1 68.750 (78.630) Prec@5 93.750 (94.570) Epoch: [35][1160/1345], lr: 0.00100 Time 0.711 (0.754) Data 0.000 (0.004) Loss 0.7734 (0.7721) Prec@1 79.688 (78.647) Prec@5 93.750 (94.584) Epoch: [35][1180/1345], lr: 0.00100 Time 0.744 (0.753) Data 0.001 (0.004) Loss 0.5970 (0.7722) Prec@1 85.938 (78.657) Prec@5 96.875 (94.582) Epoch: [35][1200/1345], lr: 0.00100 Time 0.715 (0.753) Data 0.000 (0.003) Loss 0.8459 (0.7718) Prec@1 79.688 (78.664) Prec@5 90.625 (94.590) Epoch: [35][1220/1345], lr: 0.00100 Time 0.858 (0.753) Data 0.000 (0.003) Loss 0.5675 (0.7709) Prec@1 82.812 (78.706) Prec@5 98.438 (94.605) Epoch: [35][1240/1345], lr: 0.00100 Time 0.850 (0.753) Data 0.001 (0.003) Loss 0.8131 (0.7701) Prec@1 76.562 (78.728) Prec@5 96.875 (94.612) Epoch: [35][1260/1345], lr: 0.00100 Time 0.716 (0.753) Data 0.001 (0.003) Loss 0.9713 (0.7700) Prec@1 73.438 (78.720) Prec@5 93.750 (94.611) Epoch: [35][1280/1345], lr: 0.00100 Time 0.746 (0.753) Data 0.001 (0.003) Loss 0.9985 (0.7700) Prec@1 70.312 (78.702) Prec@5 92.188 (94.609) Epoch: [35][1300/1345], lr: 0.00100 Time 0.711 (0.753) Data 0.000 (0.003) Loss 0.9069 (0.7696) Prec@1 78.125 (78.705) Prec@5 92.188 (94.624) Epoch: [35][1320/1345], lr: 0.00100 Time 0.715 (0.753) Data 0.000 (0.003) Loss 0.9189 (0.7698) Prec@1 75.000 (78.708) Prec@5 95.312 (94.624) Epoch: [35][1340/1345], lr: 0.00100 Time 0.709 (0.753) Data 0.000 (0.003) Loss 0.9388 (0.7701) Prec@1 78.125 (78.696) Prec@5 92.188 (94.622) No BN layer Freezing. Test: [0/181] Time 3.383 (3.3828) Loss 2.1687 (2.1687) Prec@1 53.125 (53.125) Prec@5 82.812 (82.812) Test: [20/181] Time 0.649 (0.5779) Loss 2.1989 (2.0480) Prec@1 50.000 (52.902) Prec@5 79.688 (80.432) Test: [40/181] Time 0.640 (0.5192) Loss 2.3170 (2.1171) Prec@1 54.688 (51.562) Prec@5 75.000 (79.878) Test: [60/181] Time 0.476 (0.4985) Loss 2.5443 (2.1556) Prec@1 53.125 (51.153) Prec@5 71.875 (79.688) Test: [80/181] Time 0.409 (0.4885) Loss 1.8704 (2.1703) Prec@1 57.812 (50.907) Prec@5 85.938 (79.225) Test: [100/181] Time 0.346 (0.4843) Loss 2.9603 (2.1615) Prec@1 46.875 (51.052) Prec@5 67.188 (79.378) Test: [120/181] Time 0.239 (0.4792) Loss 2.8906 (2.1664) Prec@1 45.312 (51.072) Prec@5 67.188 (79.455) Test: [140/181] Time 0.299 (0.4765) Loss 3.0108 (2.1624) Prec@1 37.500 (50.909) Prec@5 78.125 (79.665) Test: [160/181] Time 0.462 (0.4759) Loss 2.3902 (2.1477) Prec@1 51.562 (51.184) Prec@5 79.688 (79.882) Testing Results: Prec@1 51.389 Prec@5 79.878 Loss 2.14176 Time 0.4744 No BN layer Freezing. Epoch: [36][0/1345], lr: 0.00100 Time 3.901 (3.901) Data 3.129 (3.129) Loss 0.7067 (0.7067) Prec@1 84.375 (84.375) Prec@5 93.750 (93.750) Epoch: [36][20/1345], lr: 0.00100 Time 0.763 (0.913) Data 0.000 (0.149) Loss 0.5171 (0.6993) Prec@1 82.812 (80.804) Prec@5 96.875 (95.461) Epoch: [36][40/1345], lr: 0.00100 Time 0.713 (0.831) Data 0.000 (0.077) Loss 0.7053 (0.7027) Prec@1 81.250 (80.297) Prec@5 93.750 (95.427) Epoch: [36][60/1345], lr: 0.00100 Time 0.719 (0.807) Data 0.000 (0.052) Loss 0.5388 (0.7240) Prec@1 85.938 (79.739) Prec@5 98.438 (95.082) Epoch: [36][80/1345], lr: 0.00100 Time 0.836 (0.793) Data 0.001 (0.039) Loss 0.7530 (0.7292) Prec@1 75.000 (79.591) Prec@5 98.438 (95.081) Epoch: [36][100/1345], lr: 0.00100 Time 0.779 (0.784) Data 0.000 (0.031) Loss 0.5554 (0.7290) Prec@1 90.625 (79.703) Prec@5 98.438 (95.065) Epoch: [36][120/1345], lr: 0.00100 Time 0.737 (0.778) Data 0.000 (0.026) Loss 0.7901 (0.7400) Prec@1 79.688 (79.442) Prec@5 93.750 (95.015) Epoch: [36][140/1345], lr: 0.00100 Time 0.746 (0.774) Data 0.000 (0.023) Loss 0.6080 (0.7363) Prec@1 90.625 (79.477) Prec@5 95.312 (95.024) Epoch: [36][160/1345], lr: 0.00100 Time 0.735 (0.772) Data 0.000 (0.020) Loss 0.5279 (0.7378) Prec@1 87.500 (79.493) Prec@5 96.875 (95.031) Epoch: [36][180/1345], lr: 0.00100 Time 0.778 (0.769) Data 0.001 (0.018) Loss 0.9570 (0.7355) Prec@1 71.875 (79.610) Prec@5 90.625 (95.002) Epoch: [36][200/1345], lr: 0.00100 Time 0.713 (0.767) Data 0.000 (0.016) Loss 0.9841 (0.7397) Prec@1 75.000 (79.579) Prec@5 92.188 (94.893) Epoch: [36][220/1345], lr: 0.00100 Time 0.746 (0.765) Data 0.000 (0.015) Loss 0.7930 (0.7413) Prec@1 82.812 (79.461) Prec@5 93.750 (94.931) Epoch: [36][240/1345], lr: 0.00100 Time 0.851 (0.764) Data 0.000 (0.013) Loss 0.6326 (0.7349) Prec@1 79.688 (79.649) Prec@5 96.875 (95.027) Epoch: [36][260/1345], lr: 0.00100 Time 0.732 (0.762) Data 0.000 (0.012) Loss 1.0413 (0.7310) Prec@1 75.000 (79.735) Prec@5 92.188 (95.097) Epoch: [36][280/1345], lr: 0.00100 Time 0.731 (0.761) Data 0.000 (0.012) Loss 1.0789 (0.7333) Prec@1 73.438 (79.704) Prec@5 89.062 (95.062) Epoch: [36][300/1345], lr: 0.00100 Time 0.730 (0.760) Data 0.001 (0.011) Loss 1.3246 (0.7344) Prec@1 65.625 (79.708) Prec@5 82.812 (95.006) Epoch: [36][320/1345], lr: 0.00100 Time 0.712 (0.759) Data 0.000 (0.010) Loss 1.0398 (0.7333) Prec@1 76.562 (79.668) Prec@5 85.938 (95.040) Epoch: [36][340/1345], lr: 0.00100 Time 0.712 (0.758) Data 0.000 (0.010) Loss 0.7107 (0.7352) Prec@1 78.125 (79.582) Prec@5 98.438 (95.015) Epoch: [36][360/1345], lr: 0.00100 Time 0.712 (0.758) Data 0.000 (0.009) Loss 0.7627 (0.7361) Prec@1 79.688 (79.575) Prec@5 93.750 (94.975) Epoch: [36][380/1345], lr: 0.00100 Time 0.725 (0.757) Data 0.001 (0.009) Loss 0.6133 (0.7364) Prec@1 79.688 (79.532) Prec@5 93.750 (94.939) Epoch: [36][400/1345], lr: 0.00100 Time 0.842 (0.757) Data 0.000 (0.008) Loss 1.0107 (0.7417) Prec@1 75.000 (79.430) Prec@5 90.625 (94.814) Epoch: [36][420/1345], lr: 0.00100 Time 0.817 (0.756) Data 0.000 (0.008) Loss 0.8895 (0.7416) Prec@1 79.688 (79.472) Prec@5 90.625 (94.785) Epoch: [36][440/1345], lr: 0.00100 Time 0.750 (0.756) Data 0.001 (0.008) Loss 0.6895 (0.7426) Prec@1 79.688 (79.464) Prec@5 95.312 (94.770) Epoch: [36][460/1345], lr: 0.00100 Time 0.712 (0.755) Data 0.000 (0.007) Loss 0.5732 (0.7439) Prec@1 82.812 (79.440) Prec@5 96.875 (94.770) Epoch: [36][480/1345], lr: 0.00100 Time 0.832 (0.754) Data 0.000 (0.007) Loss 0.6397 (0.7444) Prec@1 84.375 (79.415) Prec@5 95.312 (94.770) Epoch: [36][500/1345], lr: 0.00100 Time 0.710 (0.754) Data 0.000 (0.007) Loss 0.5228 (0.7432) Prec@1 85.938 (79.422) Prec@5 98.438 (94.798) Epoch: [36][520/1345], lr: 0.00100 Time 0.712 (0.753) Data 0.000 (0.006) Loss 0.6881 (0.7440) Prec@1 81.250 (79.442) Prec@5 96.875 (94.806) Epoch: [36][540/1345], lr: 0.00100 Time 0.863 (0.752) Data 0.000 (0.006) Loss 0.4176 (0.7430) Prec@1 92.188 (79.491) Prec@5 100.000 (94.816) Epoch: [36][560/1345], lr: 0.00100 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.7593 (0.7409) Prec@1 81.250 (79.543) Prec@5 96.875 (94.850) Epoch: [36][580/1345], lr: 0.00100 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.8046 (0.7411) Prec@1 75.000 (79.532) Prec@5 93.750 (94.871) Epoch: [36][600/1345], lr: 0.00100 Time 0.717 (0.751) Data 0.000 (0.006) Loss 0.8449 (0.7416) Prec@1 75.000 (79.534) Prec@5 96.875 (94.858) Epoch: [36][620/1345], lr: 0.00100 Time 0.765 (0.751) Data 0.001 (0.006) Loss 0.6204 (0.7416) Prec@1 87.500 (79.547) Prec@5 95.312 (94.882) Epoch: [36][640/1345], lr: 0.00100 Time 0.713 (0.752) Data 0.000 (0.005) Loss 0.4025 (0.7398) Prec@1 87.500 (79.590) Prec@5 98.438 (94.915) Epoch: [36][660/1345], lr: 0.00100 Time 0.734 (0.752) Data 0.000 (0.005) Loss 0.8815 (0.7417) Prec@1 82.812 (79.534) Prec@5 93.750 (94.911) Epoch: [36][680/1345], lr: 0.00100 Time 0.735 (0.752) Data 0.000 (0.005) Loss 0.8466 (0.7416) Prec@1 71.875 (79.550) Prec@5 92.188 (94.925) Epoch: [36][700/1345], lr: 0.00100 Time 0.713 (0.751) Data 0.000 (0.005) Loss 0.6829 (0.7412) Prec@1 82.812 (79.547) Prec@5 98.438 (94.942) Epoch: [36][720/1345], lr: 0.00100 Time 0.715 (0.751) Data 0.000 (0.005) Loss 0.6312 (0.7405) Prec@1 82.812 (79.562) Prec@5 95.312 (94.961) Epoch: [36][740/1345], lr: 0.00100 Time 0.713 (0.751) Data 0.000 (0.005) Loss 0.7131 (0.7406) Prec@1 84.375 (79.595) Prec@5 95.312 (94.952) Epoch: [36][760/1345], lr: 0.00100 Time 0.713 (0.751) Data 0.000 (0.005) Loss 0.6707 (0.7414) Prec@1 76.562 (79.579) Prec@5 95.312 (94.937) Epoch: [36][780/1345], lr: 0.00100 Time 0.740 (0.751) Data 0.000 (0.004) Loss 0.8436 (0.7400) Prec@1 78.125 (79.603) Prec@5 92.188 (94.948) Epoch: [36][800/1345], lr: 0.00100 Time 0.749 (0.750) Data 0.000 (0.004) Loss 0.6192 (0.7397) Prec@1 81.250 (79.598) Prec@5 96.875 (94.961) Epoch: [36][820/1345], lr: 0.00100 Time 0.765 (0.750) Data 0.000 (0.004) Loss 0.7340 (0.7400) Prec@1 81.250 (79.579) Prec@5 92.188 (94.959) Epoch: [36][840/1345], lr: 0.00100 Time 0.733 (0.750) Data 0.000 (0.004) Loss 0.9241 (0.7399) Prec@1 78.125 (79.585) Prec@5 92.188 (94.969) Epoch: [36][860/1345], lr: 0.00100 Time 0.710 (0.750) Data 0.000 (0.004) Loss 0.5915 (0.7398) Prec@1 82.812 (79.586) Prec@5 96.875 (94.964) Epoch: [36][880/1345], lr: 0.00100 Time 0.771 (0.750) Data 0.000 (0.004) Loss 0.7910 (0.7394) Prec@1 78.125 (79.602) Prec@5 93.750 (94.968) Epoch: [36][900/1345], lr: 0.00100 Time 0.713 (0.750) Data 0.000 (0.004) Loss 0.5992 (0.7392) Prec@1 82.812 (79.629) Prec@5 95.312 (94.973) Epoch: [36][920/1345], lr: 0.00100 Time 0.713 (0.749) Data 0.000 (0.004) Loss 0.8385 (0.7397) Prec@1 76.562 (79.596) Prec@5 95.312 (94.951) Epoch: [36][940/1345], lr: 0.00100 Time 0.732 (0.749) Data 0.000 (0.004) Loss 0.9437 (0.7389) Prec@1 76.562 (79.598) Prec@5 92.188 (94.967) Epoch: [36][960/1345], lr: 0.00100 Time 0.712 (0.749) Data 0.000 (0.004) Loss 0.7750 (0.7388) Prec@1 75.000 (79.587) Prec@5 92.188 (94.966) Epoch: [36][980/1345], lr: 0.00100 Time 0.710 (0.749) Data 0.000 (0.004) Loss 0.6143 (0.7390) Prec@1 84.375 (79.590) Prec@5 96.875 (94.965) Epoch: [36][1000/1345], lr: 0.00100 Time 0.722 (0.748) Data 0.000 (0.004) Loss 0.8241 (0.7396) Prec@1 78.125 (79.583) Prec@5 96.875 (94.963) Epoch: [36][1020/1345], lr: 0.00100 Time 0.839 (0.748) Data 0.000 (0.004) Loss 0.9524 (0.7391) Prec@1 71.875 (79.596) Prec@5 92.188 (94.970) Epoch: [36][1040/1345], lr: 0.00100 Time 0.845 (0.748) Data 0.001 (0.003) Loss 0.9518 (0.7391) Prec@1 70.312 (79.590) Prec@5 92.188 (94.964) Epoch: [36][1060/1345], lr: 0.00100 Time 0.716 (0.748) Data 0.001 (0.003) Loss 0.7924 (0.7390) Prec@1 71.875 (79.611) Prec@5 96.875 (94.952) Epoch: [36][1080/1345], lr: 0.00100 Time 0.733 (0.748) Data 0.000 (0.003) Loss 0.8230 (0.7397) Prec@1 79.688 (79.581) Prec@5 96.875 (94.941) Epoch: [36][1100/1345], lr: 0.00100 Time 0.735 (0.748) Data 0.000 (0.003) Loss 1.0763 (0.7398) Prec@1 73.438 (79.574) Prec@5 87.500 (94.935) Epoch: [36][1120/1345], lr: 0.00100 Time 0.714 (0.748) Data 0.000 (0.003) Loss 0.5985 (0.7404) Prec@1 82.812 (79.559) Prec@5 98.438 (94.929) Epoch: [36][1140/1345], lr: 0.00100 Time 0.742 (0.748) Data 0.000 (0.003) Loss 0.8401 (0.7399) Prec@1 75.000 (79.582) Prec@5 95.312 (94.937) Epoch: [36][1160/1345], lr: 0.00100 Time 0.713 (0.748) Data 0.000 (0.003) Loss 0.7112 (0.7401) Prec@1 73.438 (79.578) Prec@5 95.312 (94.921) Epoch: [36][1180/1345], lr: 0.00100 Time 0.790 (0.748) Data 0.000 (0.003) Loss 0.5294 (0.7407) Prec@1 82.812 (79.558) Prec@5 95.312 (94.916) Epoch: [36][1200/1345], lr: 0.00100 Time 0.725 (0.748) Data 0.001 (0.003) Loss 0.9072 (0.7413) Prec@1 75.000 (79.550) Prec@5 92.188 (94.899) Epoch: [36][1220/1345], lr: 0.00100 Time 0.716 (0.748) Data 0.000 (0.003) Loss 0.7851 (0.7410) Prec@1 81.250 (79.566) Prec@5 93.750 (94.904) Epoch: [36][1240/1345], lr: 0.00100 Time 0.714 (0.747) Data 0.000 (0.003) Loss 0.5769 (0.7406) Prec@1 84.375 (79.594) Prec@5 96.875 (94.903) Epoch: [36][1260/1345], lr: 0.00100 Time 0.843 (0.747) Data 0.001 (0.003) Loss 0.9299 (0.7415) Prec@1 73.438 (79.564) Prec@5 92.188 (94.887) Epoch: [36][1280/1345], lr: 0.00100 Time 0.737 (0.747) Data 0.000 (0.003) Loss 0.9606 (0.7416) Prec@1 71.875 (79.562) Prec@5 90.625 (94.887) Epoch: [36][1300/1345], lr: 0.00100 Time 0.715 (0.747) Data 0.000 (0.003) Loss 0.7683 (0.7405) Prec@1 82.812 (79.593) Prec@5 93.750 (94.901) Epoch: [36][1320/1345], lr: 0.00100 Time 0.726 (0.747) Data 0.000 (0.003) Loss 0.7637 (0.7413) Prec@1 79.688 (79.577) Prec@5 96.875 (94.896) Epoch: [36][1340/1345], lr: 0.00100 Time 0.710 (0.747) Data 0.000 (0.003) Loss 0.8028 (0.7414) Prec@1 71.875 (79.566) Prec@5 95.312 (94.894) No BN layer Freezing. Test: [0/181] Time 2.933 (2.9330) Loss 2.2251 (2.2251) Prec@1 54.688 (54.688) Prec@5 85.938 (85.938) Test: [20/181] Time 0.805 (0.5912) Loss 2.1815 (2.0682) Prec@1 50.000 (52.827) Prec@5 78.125 (81.027) Test: [40/181] Time 1.116 (0.5366) Loss 2.4071 (2.1624) Prec@1 53.125 (51.524) Prec@5 73.438 (79.535) Test: [60/181] Time 1.174 (0.5103) Loss 2.6143 (2.2217) Prec@1 50.000 (50.410) Prec@5 75.000 (79.073) Test: [80/181] Time 1.299 (0.5000) Loss 1.9406 (2.2285) Prec@1 60.938 (50.231) Prec@5 85.938 (78.954) Test: [100/181] Time 1.194 (0.4928) Loss 2.8626 (2.2222) Prec@1 48.438 (50.186) Prec@5 71.875 (79.332) Test: [120/181] Time 1.281 (0.4878) Loss 2.8499 (2.2279) Prec@1 45.312 (50.323) Prec@5 73.438 (79.313) Test: [140/181] Time 0.976 (0.4826) Loss 3.0574 (2.2275) Prec@1 39.062 (50.277) Prec@5 79.688 (79.610) Test: [160/181] Time 0.910 (0.4786) Loss 2.5598 (2.2138) Prec@1 48.438 (50.553) Prec@5 79.688 (79.785) Testing Results: Prec@1 50.842 Prec@5 79.800 Loss 2.20583 Time 0.4745 No BN layer Freezing. Epoch: [37][0/1345], lr: 0.00100 Time 3.874 (3.874) Data 3.111 (3.111) Loss 0.8125 (0.8125) Prec@1 73.438 (73.438) Prec@5 93.750 (93.750) Epoch: [37][20/1345], lr: 0.00100 Time 0.714 (0.895) Data 0.000 (0.149) Loss 0.7272 (0.7033) Prec@1 75.000 (81.101) Prec@5 96.875 (94.494) Epoch: [37][40/1345], lr: 0.00100 Time 0.789 (0.818) Data 0.000 (0.076) Loss 0.8236 (0.7029) Prec@1 76.562 (80.564) Prec@5 95.312 (95.122) Epoch: [37][60/1345], lr: 0.00100 Time 0.738 (0.792) Data 0.000 (0.051) Loss 0.8473 (0.7271) Prec@1 81.250 (80.302) Prec@5 93.750 (94.980) Epoch: [37][80/1345], lr: 0.00100 Time 0.715 (0.776) Data 0.000 (0.039) Loss 0.7269 (0.7256) Prec@1 84.375 (80.382) Prec@5 93.750 (94.830) Epoch: [37][100/1345], lr: 0.00100 Time 0.742 (0.773) Data 0.000 (0.031) Loss 0.5561 (0.7157) Prec@1 82.812 (80.600) Prec@5 98.438 (94.910) Epoch: [37][120/1345], lr: 0.00100 Time 0.834 (0.769) Data 0.000 (0.026) Loss 0.8284 (0.7232) Prec@1 68.750 (80.294) Prec@5 96.875 (94.886) Epoch: [37][140/1345], lr: 0.00100 Time 0.716 (0.766) Data 0.000 (0.023) Loss 1.0473 (0.7254) Prec@1 73.438 (80.330) Prec@5 89.062 (94.880) Epoch: [37][160/1345], lr: 0.00100 Time 0.716 (0.764) Data 0.000 (0.020) Loss 0.8075 (0.7262) Prec@1 78.125 (80.396) Prec@5 90.625 (94.905) Epoch: [37][180/1345], lr: 0.00100 Time 0.714 (0.763) Data 0.001 (0.018) Loss 0.6095 (0.7227) Prec@1 82.812 (80.439) Prec@5 98.438 (94.941) Epoch: [37][200/1345], lr: 0.00100 Time 0.713 (0.761) Data 0.001 (0.016) Loss 0.9566 (0.7239) Prec@1 76.562 (80.247) Prec@5 92.188 (94.916) Epoch: [37][220/1345], lr: 0.00100 Time 0.713 (0.760) Data 0.000 (0.015) Loss 0.6858 (0.7208) Prec@1 81.250 (80.373) Prec@5 92.188 (94.888) Epoch: [37][240/1345], lr: 0.00100 Time 0.716 (0.758) Data 0.000 (0.013) Loss 0.8001 (0.7212) Prec@1 79.688 (80.355) Prec@5 95.312 (94.923) Epoch: [37][260/1345], lr: 0.00100 Time 0.830 (0.758) Data 0.000 (0.012) Loss 0.8054 (0.7231) Prec@1 71.875 (80.244) Prec@5 93.750 (94.840) Epoch: [37][280/1345], lr: 0.00100 Time 0.719 (0.757) Data 0.000 (0.012) Loss 0.6041 (0.7243) Prec@1 81.250 (80.216) Prec@5 98.438 (94.806) Epoch: [37][300/1345], lr: 0.00100 Time 0.761 (0.756) Data 0.001 (0.011) Loss 0.7058 (0.7237) Prec@1 79.688 (80.233) Prec@5 93.750 (94.799) Epoch: [37][320/1345], lr: 0.00100 Time 0.714 (0.755) Data 0.000 (0.010) Loss 0.8071 (0.7206) Prec@1 76.562 (80.296) Prec@5 92.188 (94.831) Epoch: [37][340/1345], lr: 0.00100 Time 0.711 (0.754) Data 0.000 (0.010) Loss 0.5625 (0.7196) Prec@1 84.375 (80.311) Prec@5 98.438 (94.877) Epoch: [37][360/1345], lr: 0.00100 Time 0.857 (0.754) Data 0.001 (0.009) Loss 0.6729 (0.7225) Prec@1 79.688 (80.185) Prec@5 98.438 (94.867) Epoch: [37][380/1345], lr: 0.00100 Time 0.865 (0.754) Data 0.001 (0.009) Loss 0.6326 (0.7253) Prec@1 84.375 (80.163) Prec@5 96.875 (94.833) Epoch: [37][400/1345], lr: 0.00100 Time 0.732 (0.753) Data 0.001 (0.008) Loss 0.4469 (0.7244) Prec@1 89.062 (80.163) Prec@5 98.438 (94.872) Epoch: [37][420/1345], lr: 0.00100 Time 0.759 (0.754) Data 0.000 (0.008) Loss 0.9727 (0.7243) Prec@1 78.125 (80.155) Prec@5 93.750 (94.878) Epoch: [37][440/1345], lr: 0.00100 Time 0.726 (0.753) Data 0.000 (0.008) Loss 0.5786 (0.7228) Prec@1 84.375 (80.223) Prec@5 93.750 (94.870) Epoch: [37][460/1345], lr: 0.00100 Time 0.732 (0.753) Data 0.000 (0.007) Loss 0.5968 (0.7212) Prec@1 82.812 (80.220) Prec@5 98.438 (94.930) Epoch: [37][480/1345], lr: 0.00100 Time 0.714 (0.753) Data 0.000 (0.007) Loss 0.6707 (0.7205) Prec@1 79.688 (80.246) Prec@5 98.438 (94.942) Epoch: [37][500/1345], lr: 0.00100 Time 0.713 (0.752) Data 0.000 (0.007) Loss 0.7820 (0.7200) Prec@1 78.125 (80.215) Prec@5 96.875 (94.954) Epoch: [37][520/1345], lr: 0.00100 Time 0.869 (0.752) Data 0.000 (0.006) Loss 0.7584 (0.7212) Prec@1 82.812 (80.131) Prec@5 89.062 (94.947) Epoch: [37][540/1345], lr: 0.00100 Time 0.857 (0.752) Data 0.001 (0.006) Loss 0.8739 (0.7228) Prec@1 75.000 (80.135) Prec@5 96.875 (94.917) Epoch: [37][560/1345], lr: 0.00100 Time 0.730 (0.752) Data 0.000 (0.006) Loss 0.6747 (0.7208) Prec@1 78.125 (80.197) Prec@5 95.312 (94.936) Epoch: [37][580/1345], lr: 0.00100 Time 0.715 (0.751) Data 0.001 (0.006) Loss 0.8984 (0.7214) Prec@1 76.562 (80.217) Prec@5 92.188 (94.893) Epoch: [37][600/1345], lr: 0.00100 Time 0.712 (0.751) Data 0.000 (0.006) Loss 0.6338 (0.7207) Prec@1 84.375 (80.218) Prec@5 95.312 (94.891) Epoch: [37][620/1345], lr: 0.00100 Time 0.722 (0.751) Data 0.001 (0.005) Loss 0.4952 (0.7230) Prec@1 85.938 (80.120) Prec@5 96.875 (94.867) Epoch: [37][640/1345], lr: 0.00100 Time 0.813 (0.751) Data 0.000 (0.005) Loss 0.7335 (0.7244) Prec@1 79.688 (80.078) Prec@5 96.875 (94.881) Epoch: [37][660/1345], lr: 0.00100 Time 0.715 (0.751) Data 0.000 (0.005) Loss 0.6696 (0.7242) Prec@1 79.688 (80.092) Prec@5 96.875 (94.894) Epoch: [37][680/1345], lr: 0.00100 Time 0.726 (0.751) Data 0.000 (0.005) Loss 0.5817 (0.7247) Prec@1 84.375 (80.078) Prec@5 95.312 (94.883) Epoch: [37][700/1345], lr: 0.00100 Time 0.732 (0.751) Data 0.000 (0.005) Loss 0.6750 (0.7264) Prec@1 81.250 (80.037) Prec@5 95.312 (94.889) Epoch: [37][720/1345], lr: 0.00100 Time 0.740 (0.750) Data 0.000 (0.005) Loss 0.9054 (0.7267) Prec@1 75.000 (80.034) Prec@5 92.188 (94.879) Epoch: [37][740/1345], lr: 0.00100 Time 0.715 (0.750) Data 0.000 (0.005) Loss 1.0105 (0.7284) Prec@1 76.562 (79.978) Prec@5 90.625 (94.865) Epoch: [37][760/1345], lr: 0.00100 Time 0.755 (0.750) Data 0.000 (0.005) Loss 0.9856 (0.7291) Prec@1 76.562 (79.965) Prec@5 93.750 (94.851) Epoch: [37][780/1345], lr: 0.00100 Time 0.714 (0.750) Data 0.001 (0.004) Loss 0.5984 (0.7282) Prec@1 85.938 (79.996) Prec@5 96.875 (94.866) Epoch: [37][800/1345], lr: 0.00100 Time 0.714 (0.750) Data 0.000 (0.004) Loss 0.8119 (0.7276) Prec@1 73.438 (80.011) Prec@5 98.438 (94.874) Epoch: [37][820/1345], lr: 0.00100 Time 0.833 (0.750) Data 0.000 (0.004) Loss 0.7330 (0.7278) Prec@1 79.688 (80.011) Prec@5 96.875 (94.879) Epoch: [37][840/1345], lr: 0.00100 Time 0.712 (0.749) Data 0.000 (0.004) Loss 0.6109 (0.7283) Prec@1 84.375 (80.011) Prec@5 92.188 (94.865) Epoch: [37][860/1345], lr: 0.00100 Time 0.715 (0.749) Data 0.000 (0.004) Loss 0.7102 (0.7286) Prec@1 81.250 (79.989) Prec@5 93.750 (94.870) Epoch: [37][880/1345], lr: 0.00100 Time 0.711 (0.749) Data 0.000 (0.004) Loss 0.5161 (0.7287) Prec@1 84.375 (79.969) Prec@5 96.875 (94.873) Epoch: [37][900/1345], lr: 0.00100 Time 0.776 (0.749) Data 0.000 (0.004) Loss 0.4863 (0.7268) Prec@1 85.938 (80.012) Prec@5 100.000 (94.898) Epoch: [37][920/1345], lr: 0.00100 Time 0.756 (0.749) Data 0.001 (0.004) Loss 0.6352 (0.7263) Prec@1 82.812 (80.022) Prec@5 95.312 (94.895) Epoch: [37][940/1345], lr: 0.00100 Time 0.811 (0.749) Data 0.000 (0.004) Loss 0.6188 (0.7276) Prec@1 79.688 (79.996) Prec@5 100.000 (94.872) Epoch: [37][960/1345], lr: 0.00100 Time 0.726 (0.748) Data 0.000 (0.004) Loss 0.7530 (0.7290) Prec@1 85.938 (79.948) Prec@5 93.750 (94.880) Epoch: [37][980/1345], lr: 0.00100 Time 0.714 (0.748) Data 0.000 (0.004) Loss 0.8985 (0.7285) Prec@1 81.250 (79.952) Prec@5 89.062 (94.887) Epoch: [37][1000/1345], lr: 0.00100 Time 0.717 (0.748) Data 0.000 (0.004) Loss 0.7954 (0.7277) Prec@1 78.125 (79.990) Prec@5 93.750 (94.899) Epoch: [37][1020/1345], lr: 0.00100 Time 0.713 (0.748) Data 0.001 (0.004) Loss 0.6842 (0.7288) Prec@1 84.375 (79.946) Prec@5 95.312 (94.882) Epoch: [37][1040/1345], lr: 0.00100 Time 0.717 (0.748) Data 0.000 (0.003) Loss 0.6470 (0.7284) Prec@1 82.812 (79.961) Prec@5 93.750 (94.889) Epoch: [37][1060/1345], lr: 0.00100 Time 0.714 (0.747) Data 0.000 (0.003) Loss 0.5931 (0.7278) Prec@1 84.375 (80.003) Prec@5 95.312 (94.902) Epoch: [37][1080/1345], lr: 0.00100 Time 0.849 (0.748) Data 0.001 (0.003) Loss 0.6935 (0.7280) Prec@1 79.688 (79.992) Prec@5 95.312 (94.896) Epoch: [37][1100/1345], lr: 0.00100 Time 0.842 (0.748) Data 0.000 (0.003) Loss 0.8170 (0.7284) Prec@1 82.812 (79.993) Prec@5 93.750 (94.873) Epoch: [37][1120/1345], lr: 0.00100 Time 0.748 (0.748) Data 0.000 (0.003) Loss 0.5898 (0.7281) Prec@1 84.375 (80.000) Prec@5 93.750 (94.885) Epoch: [37][1140/1345], lr: 0.00100 Time 0.777 (0.748) Data 0.000 (0.003) Loss 0.7218 (0.7277) Prec@1 85.938 (80.035) Prec@5 92.188 (94.885) Epoch: [37][1160/1345], lr: 0.00100 Time 0.732 (0.748) Data 0.000 (0.003) Loss 1.0053 (0.7282) Prec@1 75.000 (80.008) Prec@5 92.188 (94.893) Epoch: [37][1180/1345], lr: 0.00100 Time 0.714 (0.748) Data 0.000 (0.003) Loss 0.5986 (0.7285) Prec@1 81.250 (79.990) Prec@5 95.312 (94.892) Epoch: [37][1200/1345], lr: 0.00100 Time 0.726 (0.748) Data 0.000 (0.003) Loss 0.5666 (0.7282) Prec@1 84.375 (80.001) Prec@5 98.438 (94.892) Epoch: [37][1220/1345], lr: 0.00100 Time 0.714 (0.748) Data 0.000 (0.003) Loss 0.6605 (0.7288) Prec@1 78.125 (80.007) Prec@5 96.875 (94.881) Epoch: [37][1240/1345], lr: 0.00100 Time 0.843 (0.748) Data 0.000 (0.003) Loss 0.6736 (0.7288) Prec@1 85.938 (80.019) Prec@5 95.312 (94.878) Epoch: [37][1260/1345], lr: 0.00100 Time 0.842 (0.748) Data 0.000 (0.003) Loss 0.7676 (0.7286) Prec@1 82.812 (80.023) Prec@5 92.188 (94.870) Epoch: [37][1280/1345], lr: 0.00100 Time 0.739 (0.748) Data 0.000 (0.003) Loss 0.6013 (0.7281) Prec@1 78.125 (80.028) Prec@5 95.312 (94.878) Epoch: [37][1300/1345], lr: 0.00100 Time 0.715 (0.748) Data 0.001 (0.003) Loss 0.8167 (0.7285) Prec@1 73.438 (80.023) Prec@5 95.312 (94.874) Epoch: [37][1320/1345], lr: 0.00100 Time 0.718 (0.748) Data 0.000 (0.003) Loss 0.6671 (0.7280) Prec@1 81.250 (80.022) Prec@5 96.875 (94.889) Epoch: [37][1340/1345], lr: 0.00100 Time 0.712 (0.748) Data 0.001 (0.003) Loss 1.1436 (0.7279) Prec@1 68.750 (80.014) Prec@5 89.062 (94.894) No BN layer Freezing. Test: [0/181] Time 3.327 (3.3270) Loss 2.2876 (2.2876) Prec@1 53.125 (53.125) Prec@5 82.812 (82.812) Test: [20/181] Time 1.015 (0.5930) Loss 2.3286 (2.0984) Prec@1 50.000 (52.530) Prec@5 79.688 (81.696) Test: [40/181] Time 0.697 (0.5178) Loss 2.5222 (2.1733) Prec@1 53.125 (50.915) Prec@5 73.438 (80.297) Test: [60/181] Time 0.560 (0.4972) Loss 2.6061 (2.2161) Prec@1 53.125 (50.102) Prec@5 75.000 (79.739) Test: [80/181] Time 0.570 (0.4876) Loss 1.8421 (2.2260) Prec@1 59.375 (49.942) Prec@5 84.375 (79.533) Test: [100/181] Time 0.673 (0.4826) Loss 2.8737 (2.2164) Prec@1 45.312 (50.201) Prec@5 70.312 (79.641) Test: [120/181] Time 0.999 (0.4809) Loss 2.8798 (2.2287) Prec@1 43.750 (50.220) Prec@5 67.188 (79.378) Test: [140/181] Time 0.740 (0.4768) Loss 3.0989 (2.2243) Prec@1 35.938 (50.133) Prec@5 75.000 (79.499) Test: [160/181] Time 0.654 (0.4747) Loss 2.4064 (2.2062) Prec@1 46.875 (50.495) Prec@5 81.250 (79.678) Testing Results: Prec@1 50.755 Prec@5 79.705 Loss 2.19556 Time 0.4723 No BN layer Freezing. Epoch: [38][0/1345], lr: 0.00100 Time 4.083 (4.083) Data 3.290 (3.290) Loss 0.9568 (0.9568) Prec@1 68.750 (68.750) Prec@5 93.750 (93.750) Epoch: [38][20/1345], lr: 0.00100 Time 0.716 (0.912) Data 0.000 (0.157) Loss 0.5701 (0.7324) Prec@1 85.938 (79.539) Prec@5 96.875 (95.387) Epoch: [38][40/1345], lr: 0.00100 Time 0.714 (0.824) Data 0.001 (0.081) Loss 0.6396 (0.7213) Prec@1 84.375 (80.069) Prec@5 95.312 (95.465) Epoch: [38][60/1345], lr: 0.00100 Time 0.783 (0.799) Data 0.001 (0.054) Loss 0.6583 (0.6945) Prec@1 81.250 (80.533) Prec@5 92.188 (95.671) Epoch: [38][80/1345], lr: 0.00100 Time 0.714 (0.786) Data 0.000 (0.041) Loss 0.5481 (0.6910) Prec@1 90.625 (80.652) Prec@5 95.312 (95.698) Epoch: [38][100/1345], lr: 0.00100 Time 0.723 (0.776) Data 0.001 (0.033) Loss 0.6701 (0.6934) Prec@1 81.250 (80.523) Prec@5 93.750 (95.668) Epoch: [38][120/1345], lr: 0.00100 Time 0.855 (0.771) Data 0.000 (0.028) Loss 0.4170 (0.7018) Prec@1 89.062 (80.385) Prec@5 96.875 (95.403) Epoch: [38][140/1345], lr: 0.00100 Time 0.712 (0.769) Data 0.000 (0.024) Loss 0.5159 (0.7005) Prec@1 81.250 (80.408) Prec@5 98.438 (95.457) Epoch: [38][160/1345], lr: 0.00100 Time 0.713 (0.765) Data 0.000 (0.021) Loss 0.7572 (0.6997) Prec@1 81.250 (80.600) Prec@5 98.438 (95.400) Epoch: [38][180/1345], lr: 0.00100 Time 0.713 (0.762) Data 0.001 (0.019) Loss 0.5960 (0.7010) Prec@1 81.250 (80.585) Prec@5 96.875 (95.312) Epoch: [38][200/1345], lr: 0.00100 Time 0.732 (0.760) Data 0.000 (0.017) Loss 1.0424 (0.6996) Prec@1 67.188 (80.574) Prec@5 93.750 (95.336) Epoch: [38][220/1345], lr: 0.00100 Time 0.742 (0.758) Data 0.000 (0.015) Loss 0.5338 (0.6998) Prec@1 87.500 (80.536) Prec@5 100.000 (95.305) Epoch: [38][240/1345], lr: 0.00100 Time 0.776 (0.757) Data 0.000 (0.014) Loss 0.6758 (0.7032) Prec@1 79.688 (80.440) Prec@5 98.438 (95.248) Epoch: [38][260/1345], lr: 0.00100 Time 0.845 (0.756) Data 0.001 (0.013) Loss 0.5535 (0.7032) Prec@1 85.938 (80.460) Prec@5 96.875 (95.247) Epoch: [38][280/1345], lr: 0.00100 Time 0.772 (0.756) Data 0.001 (0.012) Loss 0.6633 (0.7034) Prec@1 81.250 (80.466) Prec@5 93.750 (95.285) Epoch: [38][300/1345], lr: 0.00100 Time 0.726 (0.756) Data 0.001 (0.011) Loss 0.9667 (0.7057) Prec@1 67.188 (80.326) Prec@5 93.750 (95.281) Epoch: [38][320/1345], lr: 0.00100 Time 0.738 (0.755) Data 0.001 (0.011) Loss 0.6035 (0.7040) Prec@1 82.812 (80.398) Prec@5 98.438 (95.322) Epoch: [38][340/1345], lr: 0.00100 Time 0.714 (0.756) Data 0.000 (0.010) Loss 1.0035 (0.7016) Prec@1 81.250 (80.558) Prec@5 95.312 (95.345) Epoch: [38][360/1345], lr: 0.00100 Time 0.712 (0.756) Data 0.000 (0.010) Loss 1.0753 (0.7000) Prec@1 71.875 (80.609) Prec@5 90.625 (95.356) Epoch: [38][380/1345], lr: 0.00100 Time 0.740 (0.756) Data 0.000 (0.009) Loss 0.6161 (0.7005) Prec@1 87.500 (80.553) Prec@5 96.875 (95.362) Epoch: [38][400/1345], lr: 0.00100 Time 0.843 (0.756) Data 0.000 (0.009) Loss 0.5645 (0.6979) Prec@1 87.500 (80.638) Prec@5 95.312 (95.363) Epoch: [38][420/1345], lr: 0.00100 Time 0.848 (0.755) Data 0.001 (0.008) Loss 0.8530 (0.6969) Prec@1 81.250 (80.671) Prec@5 93.750 (95.350) Epoch: [38][440/1345], lr: 0.00100 Time 0.714 (0.755) Data 0.000 (0.008) Loss 0.6973 (0.6964) Prec@1 81.250 (80.690) Prec@5 95.312 (95.334) Epoch: [38][460/1345], lr: 0.00100 Time 0.717 (0.755) Data 0.001 (0.008) Loss 0.6142 (0.6958) Prec@1 78.125 (80.660) Prec@5 98.438 (95.374) Epoch: [38][480/1345], lr: 0.00100 Time 0.713 (0.754) Data 0.000 (0.007) Loss 0.8007 (0.6968) Prec@1 79.688 (80.623) Prec@5 93.750 (95.345) Epoch: [38][500/1345], lr: 0.00100 Time 0.738 (0.754) Data 0.000 (0.007) Loss 0.6704 (0.6956) Prec@1 82.812 (80.586) Prec@5 95.312 (95.381) Epoch: [38][520/1345], lr: 0.00100 Time 0.714 (0.753) Data 0.000 (0.007) Loss 0.5358 (0.6946) Prec@1 87.500 (80.650) Prec@5 96.875 (95.381) Epoch: [38][540/1345], lr: 0.00100 Time 0.753 (0.753) Data 0.001 (0.007) Loss 0.9204 (0.6952) Prec@1 71.875 (80.675) Prec@5 92.188 (95.370) Epoch: [38][560/1345], lr: 0.00100 Time 0.725 (0.753) Data 0.000 (0.006) Loss 0.5041 (0.6965) Prec@1 89.062 (80.609) Prec@5 96.875 (95.360) Epoch: [38][580/1345], lr: 0.00100 Time 0.732 (0.752) Data 0.000 (0.006) Loss 0.8103 (0.6988) Prec@1 75.000 (80.529) Prec@5 98.438 (95.350) Epoch: [38][600/1345], lr: 0.00100 Time 0.730 (0.752) Data 0.000 (0.006) Loss 0.7383 (0.6998) Prec@1 76.562 (80.493) Prec@5 93.750 (95.325) Epoch: [38][620/1345], lr: 0.00100 Time 0.845 (0.751) Data 0.000 (0.006) Loss 0.4209 (0.7000) Prec@1 89.062 (80.503) Prec@5 98.438 (95.305) Epoch: [38][640/1345], lr: 0.00100 Time 0.854 (0.751) Data 0.001 (0.006) Loss 0.6863 (0.7000) Prec@1 81.250 (80.489) Prec@5 93.750 (95.298) Epoch: [38][660/1345], lr: 0.00100 Time 0.717 (0.751) Data 0.000 (0.005) Loss 0.7652 (0.7028) Prec@1 81.250 (80.439) Prec@5 92.188 (95.256) Epoch: [38][680/1345], lr: 0.00100 Time 0.714 (0.751) Data 0.001 (0.005) Loss 0.5871 (0.7026) Prec@1 84.375 (80.465) Prec@5 96.875 (95.262) Epoch: [38][700/1345], lr: 0.00100 Time 0.886 (0.751) Data 0.001 (0.005) Loss 0.8998 (0.7034) Prec@1 75.000 (80.472) Prec@5 90.625 (95.250) Epoch: [38][720/1345], lr: 0.00100 Time 0.729 (0.750) Data 0.000 (0.005) Loss 0.8998 (0.7050) Prec@1 76.562 (80.418) Prec@5 90.625 (95.239) Epoch: [38][740/1345], lr: 0.00100 Time 0.714 (0.750) Data 0.000 (0.005) Loss 0.6601 (0.7045) Prec@1 81.250 (80.419) Prec@5 95.312 (95.247) Epoch: [38][760/1345], lr: 0.00100 Time 0.710 (0.749) Data 0.000 (0.005) Loss 0.6347 (0.7042) Prec@1 81.250 (80.414) Prec@5 96.875 (95.271) Epoch: [38][780/1345], lr: 0.00100 Time 0.713 (0.749) Data 0.000 (0.005) Loss 0.7059 (0.7043) Prec@1 76.562 (80.428) Prec@5 98.438 (95.276) Epoch: [38][800/1345], lr: 0.00100 Time 0.768 (0.748) Data 0.000 (0.005) Loss 0.5493 (0.7039) Prec@1 81.250 (80.433) Prec@5 95.312 (95.275) Epoch: [38][820/1345], lr: 0.00100 Time 0.863 (0.748) Data 0.001 (0.004) Loss 0.6881 (0.7038) Prec@1 81.250 (80.432) Prec@5 96.875 (95.286) Epoch: [38][840/1345], lr: 0.00100 Time 0.743 (0.748) Data 0.000 (0.004) Loss 0.3100 (0.7038) Prec@1 92.188 (80.446) Prec@5 98.438 (95.288) Epoch: [38][860/1345], lr: 0.00100 Time 0.716 (0.748) Data 0.000 (0.004) Loss 0.6468 (0.7044) Prec@1 84.375 (80.430) Prec@5 95.312 (95.293) Epoch: [38][880/1345], lr: 0.00100 Time 0.716 (0.748) Data 0.001 (0.004) Loss 0.6723 (0.7031) Prec@1 79.688 (80.450) Prec@5 93.750 (95.309) Epoch: [38][900/1345], lr: 0.00100 Time 0.723 (0.748) Data 0.000 (0.004) Loss 0.6687 (0.7025) Prec@1 81.250 (80.468) Prec@5 95.312 (95.311) Epoch: [38][920/1345], lr: 0.00100 Time 0.745 (0.747) Data 0.000 (0.004) Loss 0.7754 (0.7034) Prec@1 78.125 (80.459) Prec@5 95.312 (95.301) Epoch: [38][940/1345], lr: 0.00100 Time 0.742 (0.747) Data 0.001 (0.004) Loss 0.7427 (0.7042) Prec@1 82.812 (80.443) Prec@5 89.062 (95.286) Epoch: [38][960/1345], lr: 0.00100 Time 0.841 (0.747) Data 0.001 (0.004) Loss 0.7875 (0.7038) Prec@1 78.125 (80.494) Prec@5 95.312 (95.282) Epoch: [38][980/1345], lr: 0.00100 Time 0.875 (0.747) Data 0.000 (0.004) Loss 0.6297 (0.7041) Prec@1 76.562 (80.468) Prec@5 96.875 (95.281) Epoch: [38][1000/1345], lr: 0.00100 Time 0.726 (0.747) Data 0.000 (0.004) Loss 0.8684 (0.7036) Prec@1 82.812 (80.487) Prec@5 93.750 (95.286) Epoch: [38][1020/1345], lr: 0.00100 Time 0.715 (0.747) Data 0.000 (0.004) Loss 0.7416 (0.7031) Prec@1 82.812 (80.517) Prec@5 90.625 (95.294) Epoch: [38][1040/1345], lr: 0.00100 Time 0.726 (0.747) Data 0.001 (0.004) Loss 0.5132 (0.7031) Prec@1 87.500 (80.512) Prec@5 98.438 (95.305) Epoch: [38][1060/1345], lr: 0.00100 Time 0.715 (0.747) Data 0.001 (0.004) Loss 0.7162 (0.7016) Prec@1 79.688 (80.546) Prec@5 95.312 (95.312) Epoch: [38][1080/1345], lr: 0.00100 Time 0.880 (0.747) Data 0.000 (0.004) Loss 0.5757 (0.7013) Prec@1 84.375 (80.559) Prec@5 95.312 (95.327) Epoch: [38][1100/1345], lr: 0.00100 Time 0.749 (0.747) Data 0.000 (0.003) Loss 0.6792 (0.7010) Prec@1 84.375 (80.584) Prec@5 96.875 (95.332) Epoch: [38][1120/1345], lr: 0.00100 Time 0.718 (0.747) Data 0.001 (0.003) Loss 0.8915 (0.6994) Prec@1 79.688 (80.634) Prec@5 89.062 (95.339) Epoch: [38][1140/1345], lr: 0.00100 Time 0.712 (0.747) Data 0.000 (0.003) Loss 0.3511 (0.6991) Prec@1 92.188 (80.646) Prec@5 100.000 (95.344) Epoch: [38][1160/1345], lr: 0.00100 Time 0.715 (0.748) Data 0.000 (0.003) Loss 0.7337 (0.6987) Prec@1 81.250 (80.678) Prec@5 96.875 (95.347) Epoch: [38][1180/1345], lr: 0.00100 Time 0.714 (0.747) Data 0.000 (0.003) Loss 0.7042 (0.6985) Prec@1 78.125 (80.677) Prec@5 92.188 (95.342) Epoch: [38][1200/1345], lr: 0.00100 Time 0.715 (0.747) Data 0.000 (0.003) Loss 0.5833 (0.6985) Prec@1 87.500 (80.685) Prec@5 95.312 (95.337) Epoch: [38][1220/1345], lr: 0.00100 Time 0.809 (0.747) Data 0.000 (0.003) Loss 1.1398 (0.6988) Prec@1 75.000 (80.666) Prec@5 89.062 (95.325) Epoch: [38][1240/1345], lr: 0.00100 Time 0.820 (0.747) Data 0.000 (0.003) Loss 0.7972 (0.6994) Prec@1 76.562 (80.651) Prec@5 93.750 (95.312) Epoch: [38][1260/1345], lr: 0.00100 Time 0.733 (0.747) Data 0.000 (0.003) Loss 0.6731 (0.6984) Prec@1 85.938 (80.680) Prec@5 92.188 (95.326) Epoch: [38][1280/1345], lr: 0.00100 Time 0.711 (0.747) Data 0.000 (0.003) Loss 0.7142 (0.6991) Prec@1 75.000 (80.660) Prec@5 96.875 (95.323) Epoch: [38][1300/1345], lr: 0.00100 Time 0.733 (0.747) Data 0.000 (0.003) Loss 0.5308 (0.7001) Prec@1 87.500 (80.629) Prec@5 98.438 (95.308) Epoch: [38][1320/1345], lr: 0.00100 Time 0.714 (0.747) Data 0.000 (0.003) Loss 0.7474 (0.7003) Prec@1 81.250 (80.649) Prec@5 93.750 (95.301) Epoch: [38][1340/1345], lr: 0.00100 Time 0.711 (0.747) Data 0.000 (0.003) Loss 0.8750 (0.6998) Prec@1 79.688 (80.678) Prec@5 92.188 (95.310) No BN layer Freezing. Test: [0/181] Time 3.332 (3.3322) Loss 2.1811 (2.1811) Prec@1 53.125 (53.125) Prec@5 82.812 (82.812) Test: [20/181] Time 0.926 (0.5927) Loss 2.3165 (2.0763) Prec@1 51.562 (52.679) Prec@5 76.562 (81.473) Test: [40/181] Time 1.165 (0.5354) Loss 2.4529 (2.1655) Prec@1 54.688 (51.486) Prec@5 73.438 (80.450) Test: [60/181] Time 0.896 (0.5029) Loss 2.6139 (2.2118) Prec@1 50.000 (50.794) Prec@5 71.875 (79.764) Test: [80/181] Time 1.043 (0.4926) Loss 1.8490 (2.2187) Prec@1 57.812 (50.482) Prec@5 87.500 (79.610) Test: [100/181] Time 0.941 (0.4845) Loss 2.8848 (2.2063) Prec@1 45.312 (50.495) Prec@5 71.875 (79.827) Test: [120/181] Time 0.931 (0.4826) Loss 2.9323 (2.2145) Prec@1 42.188 (50.452) Prec@5 68.750 (79.688) Test: [140/181] Time 1.008 (0.4807) Loss 3.0334 (2.2161) Prec@1 31.250 (50.255) Prec@5 75.000 (79.832) Test: [160/181] Time 1.037 (0.4769) Loss 2.4988 (2.2030) Prec@1 46.875 (50.621) Prec@5 82.812 (79.891) Testing Results: Prec@1 50.885 Prec@5 79.939 Loss 2.19334 Time 0.4743 No BN layer Freezing. Epoch: [39][0/1345], lr: 0.00100 Time 4.315 (4.315) Data 3.572 (3.572) Loss 0.7676 (0.7676) Prec@1 82.812 (82.812) Prec@5 90.625 (90.625) Epoch: [39][20/1345], lr: 0.00100 Time 0.716 (0.918) Data 0.000 (0.171) Loss 0.7419 (0.6697) Prec@1 79.688 (82.440) Prec@5 98.438 (95.610) Epoch: [39][40/1345], lr: 0.00100 Time 0.730 (0.840) Data 0.001 (0.088) Loss 0.7083 (0.6605) Prec@1 75.000 (82.050) Prec@5 95.312 (95.884) Epoch: [39][60/1345], lr: 0.00100 Time 0.713 (0.809) Data 0.000 (0.059) Loss 0.6740 (0.6753) Prec@1 79.688 (81.814) Prec@5 95.312 (95.671) Epoch: [39][80/1345], lr: 0.00100 Time 0.769 (0.793) Data 0.001 (0.045) Loss 0.7701 (0.6632) Prec@1 78.125 (82.176) Prec@5 93.750 (95.775) Epoch: [39][100/1345], lr: 0.00100 Time 0.714 (0.784) Data 0.000 (0.036) Loss 0.8337 (0.6650) Prec@1 81.250 (82.101) Prec@5 95.312 (95.792) Epoch: [39][120/1345], lr: 0.00100 Time 0.731 (0.778) Data 0.000 (0.030) Loss 0.6715 (0.6788) Prec@1 78.125 (81.612) Prec@5 98.438 (95.674) Epoch: [39][140/1345], lr: 0.00100 Time 0.714 (0.774) Data 0.000 (0.026) Loss 0.5252 (0.6821) Prec@1 89.062 (81.494) Prec@5 96.875 (95.501) Epoch: [39][160/1345], lr: 0.00100 Time 0.734 (0.771) Data 0.000 (0.023) Loss 0.5414 (0.6828) Prec@1 85.938 (81.396) Prec@5 95.312 (95.477) Epoch: [39][180/1345], lr: 0.00100 Time 0.739 (0.769) Data 0.000 (0.020) Loss 0.6019 (0.6822) Prec@1 84.375 (81.371) Prec@5 93.750 (95.459) Epoch: [39][200/1345], lr: 0.00100 Time 0.750 (0.767) Data 0.000 (0.018) Loss 0.6151 (0.6842) Prec@1 82.812 (81.234) Prec@5 95.312 (95.421) Epoch: [39][220/1345], lr: 0.00100 Time 0.714 (0.765) Data 0.000 (0.017) Loss 0.6517 (0.6823) Prec@1 81.250 (81.292) Prec@5 92.188 (95.447) Epoch: [39][240/1345], lr: 0.00100 Time 0.777 (0.764) Data 0.000 (0.015) Loss 0.6431 (0.6826) Prec@1 84.375 (81.269) Prec@5 93.750 (95.449) Epoch: [39][260/1345], lr: 0.00100 Time 0.712 (0.763) Data 0.000 (0.014) Loss 0.7864 (0.6822) Prec@1 81.250 (81.256) Prec@5 93.750 (95.474) Epoch: [39][280/1345], lr: 0.00100 Time 0.759 (0.761) Data 0.000 (0.013) Loss 0.7781 (0.6830) Prec@1 79.688 (81.256) Prec@5 92.188 (95.440) Epoch: [39][300/1345], lr: 0.00100 Time 0.771 (0.761) Data 0.000 (0.012) Loss 0.6061 (0.6826) Prec@1 89.062 (81.203) Prec@5 95.312 (95.458) Epoch: [39][320/1345], lr: 0.00100 Time 0.712 (0.760) Data 0.000 (0.012) Loss 0.6249 (0.6853) Prec@1 85.938 (81.094) Prec@5 95.312 (95.424) Epoch: [39][340/1345], lr: 0.00100 Time 0.731 (0.759) Data 0.000 (0.011) Loss 0.5827 (0.6846) Prec@1 85.938 (81.080) Prec@5 96.875 (95.413) Epoch: [39][360/1345], lr: 0.00100 Time 0.714 (0.759) Data 0.000 (0.010) Loss 0.5544 (0.6838) Prec@1 79.688 (81.120) Prec@5 100.000 (95.447) Epoch: [39][380/1345], lr: 0.00100 Time 0.714 (0.758) Data 0.000 (0.010) Loss 0.6561 (0.6861) Prec@1 75.000 (81.033) Prec@5 95.312 (95.436) Epoch: [39][400/1345], lr: 0.00100 Time 0.750 (0.758) Data 0.000 (0.009) Loss 0.7083 (0.6833) Prec@1 82.812 (81.172) Prec@5 92.188 (95.464) Epoch: [39][420/1345], lr: 0.00100 Time 0.714 (0.757) Data 0.000 (0.009) Loss 0.6017 (0.6825) Prec@1 82.812 (81.209) Prec@5 95.312 (95.461) Epoch: [39][440/1345], lr: 0.00100 Time 0.717 (0.757) Data 0.000 (0.009) Loss 0.5920 (0.6817) Prec@1 84.375 (81.239) Prec@5 96.875 (95.472) Epoch: [39][460/1345], lr: 0.00100 Time 0.713 (0.757) Data 0.000 (0.008) Loss 0.3498 (0.6799) Prec@1 92.188 (81.281) Prec@5 100.000 (95.489) Epoch: [39][480/1345], lr: 0.00100 Time 0.710 (0.756) Data 0.000 (0.008) Loss 0.8889 (0.6807) Prec@1 75.000 (81.263) Prec@5 90.625 (95.446) Epoch: [39][500/1345], lr: 0.00100 Time 0.725 (0.755) Data 0.000 (0.008) Loss 0.5740 (0.6826) Prec@1 82.812 (81.188) Prec@5 96.875 (95.434) Epoch: [39][520/1345], lr: 0.00100 Time 0.773 (0.755) Data 0.000 (0.007) Loss 0.7352 (0.6831) Prec@1 75.000 (81.166) Prec@5 95.312 (95.459) Epoch: [39][540/1345], lr: 0.00100 Time 0.844 (0.755) Data 0.000 (0.007) Loss 0.8958 (0.6861) Prec@1 85.938 (81.103) Prec@5 90.625 (95.422) Epoch: [39][560/1345], lr: 0.00100 Time 0.890 (0.754) Data 0.001 (0.007) Loss 0.7153 (0.6872) Prec@1 78.125 (81.033) Prec@5 95.312 (95.427) Epoch: [39][580/1345], lr: 0.00100 Time 0.713 (0.754) Data 0.000 (0.007) Loss 0.8325 (0.6861) Prec@1 76.562 (81.067) Prec@5 93.750 (95.442) Epoch: [39][600/1345], lr: 0.00100 Time 0.732 (0.754) Data 0.000 (0.006) Loss 0.6589 (0.6854) Prec@1 75.000 (81.034) Prec@5 100.000 (95.481) Epoch: [39][620/1345], lr: 0.00100 Time 0.713 (0.753) Data 0.000 (0.006) Loss 0.6426 (0.6830) Prec@1 84.375 (81.076) Prec@5 96.875 (95.521) Epoch: [39][640/1345], lr: 0.00100 Time 0.717 (0.753) Data 0.000 (0.006) Loss 0.4534 (0.6825) Prec@1 87.500 (81.101) Prec@5 100.000 (95.529) Epoch: [39][660/1345], lr: 0.00100 Time 0.733 (0.752) Data 0.001 (0.006) Loss 0.8128 (0.6811) Prec@1 81.250 (81.155) Prec@5 90.625 (95.539) Epoch: [39][680/1345], lr: 0.00100 Time 0.714 (0.752) Data 0.000 (0.006) Loss 0.7383 (0.6815) Prec@1 82.812 (81.167) Prec@5 96.875 (95.530) Epoch: [39][700/1345], lr: 0.00100 Time 0.713 (0.752) Data 0.000 (0.006) Loss 0.6600 (0.6805) Prec@1 89.062 (81.185) Prec@5 93.750 (95.542) Epoch: [39][720/1345], lr: 0.00100 Time 0.713 (0.752) Data 0.000 (0.005) Loss 0.7291 (0.6813) Prec@1 81.250 (81.150) Prec@5 95.312 (95.544) Epoch: [39][740/1345], lr: 0.00100 Time 0.713 (0.751) Data 0.000 (0.005) Loss 0.8057 (0.6810) Prec@1 70.312 (81.132) Prec@5 95.312 (95.547) Epoch: [39][760/1345], lr: 0.00100 Time 0.712 (0.751) Data 0.000 (0.005) Loss 0.7739 (0.6816) Prec@1 79.688 (81.088) Prec@5 96.875 (95.555) Epoch: [39][780/1345], lr: 0.00100 Time 0.850 (0.752) Data 0.000 (0.005) Loss 0.7650 (0.6826) Prec@1 79.688 (81.054) Prec@5 96.875 (95.545) Epoch: [39][800/1345], lr: 0.00100 Time 0.828 (0.752) Data 0.000 (0.005) Loss 1.0228 (0.6831) Prec@1 68.750 (81.012) Prec@5 89.062 (95.539) Epoch: [39][820/1345], lr: 0.00100 Time 0.778 (0.752) Data 0.001 (0.005) Loss 0.6356 (0.6828) Prec@1 84.375 (81.029) Prec@5 98.438 (95.528) Epoch: [39][840/1345], lr: 0.00100 Time 0.733 (0.752) Data 0.000 (0.005) Loss 0.6124 (0.6841) Prec@1 81.250 (80.979) Prec@5 95.312 (95.511) Epoch: [39][860/1345], lr: 0.00100 Time 0.733 (0.752) Data 0.000 (0.005) Loss 0.6101 (0.6843) Prec@1 84.375 (80.976) Prec@5 95.312 (95.516) Epoch: [39][880/1345], lr: 0.00100 Time 0.714 (0.751) Data 0.000 (0.005) Loss 0.4201 (0.6840) Prec@1 89.062 (81.003) Prec@5 98.438 (95.509) Epoch: [39][900/1345], lr: 0.00100 Time 0.712 (0.751) Data 0.000 (0.004) Loss 0.5331 (0.6839) Prec@1 82.812 (81.021) Prec@5 98.438 (95.498) Epoch: [39][920/1345], lr: 0.00100 Time 0.717 (0.751) Data 0.001 (0.004) Loss 0.7336 (0.6839) Prec@1 79.688 (81.035) Prec@5 95.312 (95.504) Epoch: [39][940/1345], lr: 0.00100 Time 0.840 (0.751) Data 0.000 (0.004) Loss 0.9171 (0.6840) Prec@1 78.125 (81.029) Prec@5 93.750 (95.507) Epoch: [39][960/1345], lr: 0.00100 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.8648 (0.6852) Prec@1 71.875 (81.001) Prec@5 96.875 (95.504) Epoch: [39][980/1345], lr: 0.00100 Time 0.719 (0.751) Data 0.000 (0.004) Loss 0.6311 (0.6859) Prec@1 84.375 (80.978) Prec@5 96.875 (95.507) Epoch: [39][1000/1345], lr: 0.00100 Time 0.713 (0.750) Data 0.000 (0.004) Loss 0.5579 (0.6859) Prec@1 84.375 (80.999) Prec@5 95.312 (95.501) Epoch: [39][1020/1345], lr: 0.00100 Time 0.715 (0.750) Data 0.001 (0.004) Loss 0.7106 (0.6872) Prec@1 81.250 (80.990) Prec@5 98.438 (95.485) Epoch: [39][1040/1345], lr: 0.00100 Time 0.712 (0.750) Data 0.000 (0.004) Loss 0.7503 (0.6868) Prec@1 76.562 (80.999) Prec@5 96.875 (95.496) Epoch: [39][1060/1345], lr: 0.00100 Time 0.714 (0.750) Data 0.000 (0.004) Loss 0.5960 (0.6869) Prec@1 85.938 (81.000) Prec@5 96.875 (95.488) Epoch: [39][1080/1345], lr: 0.00100 Time 0.715 (0.749) Data 0.000 (0.004) Loss 0.7490 (0.6868) Prec@1 75.000 (80.996) Prec@5 96.875 (95.496) Epoch: [39][1100/1345], lr: 0.00100 Time 0.734 (0.749) Data 0.000 (0.004) Loss 0.8006 (0.6877) Prec@1 75.000 (80.982) Prec@5 93.750 (95.474) Epoch: [39][1120/1345], lr: 0.00100 Time 0.760 (0.749) Data 0.000 (0.004) Loss 0.6725 (0.6882) Prec@1 82.812 (80.974) Prec@5 92.188 (95.463) Epoch: [39][1140/1345], lr: 0.00100 Time 0.719 (0.749) Data 0.000 (0.004) Loss 0.6754 (0.6876) Prec@1 81.250 (80.990) Prec@5 93.750 (95.467) Epoch: [39][1160/1345], lr: 0.00100 Time 0.749 (0.749) Data 0.000 (0.004) Loss 0.6714 (0.6883) Prec@1 79.688 (80.979) Prec@5 96.875 (95.458) Epoch: [39][1180/1345], lr: 0.00100 Time 0.730 (0.749) Data 0.000 (0.003) Loss 0.5501 (0.6884) Prec@1 84.375 (80.980) Prec@5 98.438 (95.466) Epoch: [39][1200/1345], lr: 0.00100 Time 0.725 (0.749) Data 0.000 (0.003) Loss 0.7757 (0.6883) Prec@1 73.438 (80.972) Prec@5 95.312 (95.470) Epoch: [39][1220/1345], lr: 0.00100 Time 0.714 (0.749) Data 0.000 (0.003) Loss 0.7032 (0.6890) Prec@1 82.812 (80.948) Prec@5 95.312 (95.466) Epoch: [39][1240/1345], lr: 0.00100 Time 0.715 (0.749) Data 0.000 (0.003) Loss 0.5641 (0.6894) Prec@1 84.375 (80.949) Prec@5 96.875 (95.464) Epoch: [39][1260/1345], lr: 0.00100 Time 0.714 (0.749) Data 0.000 (0.003) Loss 0.9340 (0.6892) Prec@1 76.562 (80.945) Prec@5 90.625 (95.461) Epoch: [39][1280/1345], lr: 0.00100 Time 0.827 (0.749) Data 0.000 (0.003) Loss 0.8370 (0.6896) Prec@1 76.562 (80.927) Prec@5 92.188 (95.458) Epoch: [39][1300/1345], lr: 0.00100 Time 0.757 (0.749) Data 0.000 (0.003) Loss 0.8003 (0.6898) Prec@1 78.125 (80.933) Prec@5 93.750 (95.452) Epoch: [39][1320/1345], lr: 0.00100 Time 0.734 (0.749) Data 0.000 (0.003) Loss 0.8173 (0.6904) Prec@1 79.688 (80.919) Prec@5 90.625 (95.439) Epoch: [39][1340/1345], lr: 0.00100 Time 0.720 (0.749) Data 0.000 (0.003) Loss 0.9474 (0.6916) Prec@1 76.562 (80.882) Prec@5 93.750 (95.423) No BN layer Freezing. Test: [0/181] Time 3.506 (3.5063) Loss 2.3151 (2.3151) Prec@1 51.562 (51.562) Prec@5 81.250 (81.250) Test: [20/181] Time 1.199 (0.6047) Loss 2.3575 (2.1072) Prec@1 50.000 (52.827) Prec@5 75.000 (80.432) Test: [40/181] Time 0.950 (0.5279) Loss 2.4793 (2.1724) Prec@1 54.688 (51.982) Prec@5 75.000 (79.726) Test: [60/181] Time 1.099 (0.5047) Loss 2.7000 (2.2216) Prec@1 51.562 (50.666) Prec@5 73.438 (79.329) Test: [80/181] Time 0.727 (0.4922) Loss 1.7657 (2.2241) Prec@1 59.375 (50.424) Prec@5 89.062 (79.437) Test: [100/181] Time 0.731 (0.4871) Loss 3.0268 (2.2150) Prec@1 43.750 (50.619) Prec@5 71.875 (79.765) Test: [120/181] Time 0.760 (0.4830) Loss 2.7876 (2.2245) Prec@1 42.188 (50.633) Prec@5 71.875 (79.649) Test: [140/181] Time 0.878 (0.4800) Loss 2.9759 (2.2188) Prec@1 35.938 (50.499) Prec@5 71.875 (79.898) Test: [160/181] Time 0.973 (0.4778) Loss 2.4908 (2.2052) Prec@1 51.562 (50.728) Prec@5 82.812 (80.066) Testing Results: Prec@1 50.790 Prec@5 80.174 Loss 2.19708 Time 0.4723 No BN layer Freezing. Epoch: [40][0/1345], lr: 0.00010 Time 3.897 (3.897) Data 3.080 (3.080) Loss 0.5813 (0.5813) Prec@1 84.375 (84.375) Prec@5 93.750 (93.750) Epoch: [40][20/1345], lr: 0.00010 Time 0.718 (0.899) Data 0.000 (0.147) Loss 0.8005 (0.6607) Prec@1 76.562 (81.548) Prec@5 93.750 (95.387) Epoch: [40][40/1345], lr: 0.00010 Time 0.715 (0.821) Data 0.000 (0.076) Loss 0.8439 (0.6897) Prec@1 79.688 (81.364) Prec@5 89.062 (94.550) Epoch: [40][60/1345], lr: 0.00010 Time 0.721 (0.797) Data 0.001 (0.051) Loss 0.5718 (0.6899) Prec@1 82.812 (81.224) Prec@5 96.875 (94.903) Epoch: [40][80/1345], lr: 0.00010 Time 0.841 (0.783) Data 0.000 (0.038) Loss 0.6782 (0.6951) Prec@1 85.938 (81.443) Prec@5 96.875 (94.869) Epoch: [40][100/1345], lr: 0.00010 Time 0.717 (0.775) Data 0.000 (0.031) Loss 0.5708 (0.6829) Prec@1 84.375 (82.039) Prec@5 93.750 (95.080) Epoch: [40][120/1345], lr: 0.00010 Time 0.715 (0.770) Data 0.000 (0.026) Loss 0.6111 (0.6813) Prec@1 76.562 (82.012) Prec@5 96.875 (95.158) Epoch: [40][140/1345], lr: 0.00010 Time 0.715 (0.767) Data 0.000 (0.022) Loss 0.7056 (0.6835) Prec@1 78.125 (81.948) Prec@5 96.875 (95.035) Epoch: [40][160/1345], lr: 0.00010 Time 0.714 (0.763) Data 0.000 (0.020) Loss 0.6861 (0.6757) Prec@1 79.688 (81.978) Prec@5 96.875 (95.283) Epoch: [40][180/1345], lr: 0.00010 Time 0.717 (0.760) Data 0.000 (0.017) Loss 0.7879 (0.6739) Prec@1 75.000 (82.010) Prec@5 96.875 (95.347) Epoch: [40][200/1345], lr: 0.00010 Time 0.761 (0.759) Data 0.000 (0.016) Loss 0.6669 (0.6738) Prec@1 79.688 (81.903) Prec@5 96.875 (95.445) Epoch: [40][220/1345], lr: 0.00010 Time 0.825 (0.757) Data 0.000 (0.014) Loss 0.7092 (0.6742) Prec@1 82.812 (81.837) Prec@5 93.750 (95.510) Epoch: [40][240/1345], lr: 0.00010 Time 0.850 (0.757) Data 0.000 (0.013) Loss 0.3834 (0.6669) Prec@1 90.625 (82.086) Prec@5 100.000 (95.598) Epoch: [40][260/1345], lr: 0.00010 Time 0.724 (0.756) Data 0.000 (0.012) Loss 0.6867 (0.6666) Prec@1 81.250 (82.040) Prec@5 93.750 (95.618) Epoch: [40][280/1345], lr: 0.00010 Time 0.727 (0.756) Data 0.000 (0.011) Loss 0.5548 (0.6656) Prec@1 85.938 (82.073) Prec@5 96.875 (95.618) Epoch: [40][300/1345], lr: 0.00010 Time 0.713 (0.754) Data 0.000 (0.011) Loss 0.7354 (0.6659) Prec@1 79.688 (82.070) Prec@5 95.312 (95.624) Epoch: [40][320/1345], lr: 0.00010 Time 0.711 (0.753) Data 0.000 (0.010) Loss 0.7509 (0.6632) Prec@1 79.688 (82.097) Prec@5 93.750 (95.663) Epoch: [40][340/1345], lr: 0.00010 Time 0.761 (0.753) Data 0.000 (0.009) Loss 0.4724 (0.6611) Prec@1 87.500 (82.166) Prec@5 96.875 (95.679) Epoch: [40][360/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.009) Loss 0.5414 (0.6615) Prec@1 84.375 (82.159) Prec@5 98.438 (95.663) Epoch: [40][380/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.009) Loss 0.5926 (0.6620) Prec@1 82.812 (82.136) Prec@5 95.312 (95.653) Epoch: [40][400/1345], lr: 0.00010 Time 0.771 (0.751) Data 0.000 (0.008) Loss 0.6324 (0.6635) Prec@1 87.500 (82.029) Prec@5 92.188 (95.613) Epoch: [40][420/1345], lr: 0.00010 Time 0.772 (0.751) Data 0.000 (0.008) Loss 0.8113 (0.6633) Prec@1 82.812 (81.992) Prec@5 96.875 (95.632) Epoch: [40][440/1345], lr: 0.00010 Time 0.726 (0.751) Data 0.000 (0.007) Loss 0.5004 (0.6598) Prec@1 85.938 (82.090) Prec@5 98.438 (95.674) Epoch: [40][460/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.007) Loss 0.9223 (0.6627) Prec@1 70.312 (81.955) Prec@5 92.188 (95.634) Epoch: [40][480/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.007) Loss 0.5895 (0.6605) Prec@1 84.375 (82.017) Prec@5 95.312 (95.670) Epoch: [40][500/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.007) Loss 0.6399 (0.6585) Prec@1 81.250 (82.042) Prec@5 95.312 (95.718) Epoch: [40][520/1345], lr: 0.00010 Time 0.839 (0.750) Data 0.000 (0.006) Loss 0.5735 (0.6577) Prec@1 87.500 (82.084) Prec@5 96.875 (95.699) Epoch: [40][540/1345], lr: 0.00010 Time 0.742 (0.750) Data 0.001 (0.006) Loss 0.6084 (0.6579) Prec@1 85.938 (82.116) Prec@5 93.750 (95.659) Epoch: [40][560/1345], lr: 0.00010 Time 0.728 (0.749) Data 0.001 (0.006) Loss 0.5684 (0.6574) Prec@1 79.688 (82.125) Prec@5 96.875 (95.658) Epoch: [40][580/1345], lr: 0.00010 Time 0.714 (0.749) Data 0.001 (0.006) Loss 0.5329 (0.6574) Prec@1 84.375 (82.113) Prec@5 96.875 (95.649) Epoch: [40][600/1345], lr: 0.00010 Time 0.714 (0.749) Data 0.000 (0.006) Loss 0.6556 (0.6566) Prec@1 87.500 (82.126) Prec@5 95.312 (95.674) Epoch: [40][620/1345], lr: 0.00010 Time 0.748 (0.749) Data 0.001 (0.005) Loss 0.5107 (0.6553) Prec@1 85.938 (82.141) Prec@5 96.875 (95.657) Epoch: [40][640/1345], lr: 0.00010 Time 0.713 (0.748) Data 0.000 (0.005) Loss 0.5723 (0.6553) Prec@1 79.688 (82.135) Prec@5 98.438 (95.681) Epoch: [40][660/1345], lr: 0.00010 Time 0.730 (0.748) Data 0.000 (0.005) Loss 0.5444 (0.6553) Prec@1 84.375 (82.136) Prec@5 98.438 (95.684) Epoch: [40][680/1345], lr: 0.00010 Time 0.740 (0.748) Data 0.000 (0.005) Loss 0.7453 (0.6542) Prec@1 78.125 (82.147) Prec@5 93.750 (95.705) Epoch: [40][700/1345], lr: 0.00010 Time 0.733 (0.748) Data 0.001 (0.005) Loss 0.3796 (0.6524) Prec@1 92.188 (82.206) Prec@5 100.000 (95.738) Epoch: [40][720/1345], lr: 0.00010 Time 0.723 (0.747) Data 0.000 (0.005) Loss 0.6222 (0.6528) Prec@1 84.375 (82.219) Prec@5 93.750 (95.722) Epoch: [40][740/1345], lr: 0.00010 Time 0.816 (0.748) Data 0.000 (0.005) Loss 1.0668 (0.6523) Prec@1 71.875 (82.220) Prec@5 84.375 (95.726) Epoch: [40][760/1345], lr: 0.00010 Time 0.755 (0.748) Data 0.000 (0.005) Loss 0.4186 (0.6519) Prec@1 92.188 (82.258) Prec@5 98.438 (95.735) Epoch: [40][780/1345], lr: 0.00010 Time 0.714 (0.748) Data 0.000 (0.004) Loss 0.9710 (0.6532) Prec@1 67.188 (82.212) Prec@5 93.750 (95.721) Epoch: [40][800/1345], lr: 0.00010 Time 0.746 (0.748) Data 0.000 (0.004) Loss 0.4472 (0.6517) Prec@1 87.500 (82.237) Prec@5 100.000 (95.736) Epoch: [40][820/1345], lr: 0.00010 Time 0.715 (0.747) Data 0.000 (0.004) Loss 0.4440 (0.6502) Prec@1 90.625 (82.245) Prec@5 96.875 (95.756) Epoch: [40][840/1345], lr: 0.00010 Time 0.715 (0.747) Data 0.001 (0.004) Loss 0.5652 (0.6502) Prec@1 82.812 (82.251) Prec@5 96.875 (95.745) Epoch: [40][860/1345], lr: 0.00010 Time 0.714 (0.747) Data 0.000 (0.004) Loss 0.3414 (0.6493) Prec@1 93.750 (82.268) Prec@5 98.438 (95.746) Epoch: [40][880/1345], lr: 0.00010 Time 0.715 (0.747) Data 0.000 (0.004) Loss 0.6138 (0.6481) Prec@1 85.938 (82.288) Prec@5 96.875 (95.761) Epoch: [40][900/1345], lr: 0.00010 Time 0.837 (0.747) Data 0.000 (0.004) Loss 0.5344 (0.6476) Prec@1 79.688 (82.323) Prec@5 96.875 (95.751) Epoch: [40][920/1345], lr: 0.00010 Time 0.736 (0.747) Data 0.000 (0.004) Loss 0.6838 (0.6469) Prec@1 75.000 (82.344) Prec@5 98.438 (95.762) Epoch: [40][940/1345], lr: 0.00010 Time 0.745 (0.747) Data 0.000 (0.004) Loss 0.5145 (0.6461) Prec@1 82.812 (82.343) Prec@5 98.438 (95.767) Epoch: [40][960/1345], lr: 0.00010 Time 0.766 (0.747) Data 0.000 (0.004) Loss 0.5871 (0.6458) Prec@1 81.250 (82.360) Prec@5 95.312 (95.765) Epoch: [40][980/1345], lr: 0.00010 Time 0.712 (0.747) Data 0.000 (0.004) Loss 0.4569 (0.6454) Prec@1 84.375 (82.370) Prec@5 100.000 (95.762) Epoch: [40][1000/1345], lr: 0.00010 Time 0.757 (0.747) Data 0.000 (0.004) Loss 0.5737 (0.6449) Prec@1 79.688 (82.380) Prec@5 98.438 (95.768) Epoch: [40][1020/1345], lr: 0.00010 Time 0.711 (0.747) Data 0.001 (0.003) Loss 0.6018 (0.6451) Prec@1 89.062 (82.364) Prec@5 95.312 (95.773) Epoch: [40][1040/1345], lr: 0.00010 Time 0.720 (0.747) Data 0.000 (0.003) Loss 0.5742 (0.6441) Prec@1 82.812 (82.389) Prec@5 96.875 (95.781) Epoch: [40][1060/1345], lr: 0.00010 Time 0.817 (0.747) Data 0.000 (0.003) Loss 0.5060 (0.6439) Prec@1 79.688 (82.406) Prec@5 98.438 (95.781) Epoch: [40][1080/1345], lr: 0.00010 Time 0.732 (0.747) Data 0.000 (0.003) Loss 0.5737 (0.6441) Prec@1 89.062 (82.408) Prec@5 96.875 (95.785) Epoch: [40][1100/1345], lr: 0.00010 Time 0.726 (0.747) Data 0.000 (0.003) Loss 0.5589 (0.6440) Prec@1 84.375 (82.394) Prec@5 98.438 (95.787) Epoch: [40][1120/1345], lr: 0.00010 Time 0.716 (0.747) Data 0.000 (0.003) Loss 0.6194 (0.6437) Prec@1 81.250 (82.385) Prec@5 96.875 (95.795) Epoch: [40][1140/1345], lr: 0.00010 Time 0.734 (0.747) Data 0.000 (0.003) Loss 0.6150 (0.6435) Prec@1 89.062 (82.403) Prec@5 95.312 (95.799) Epoch: [40][1160/1345], lr: 0.00010 Time 0.713 (0.747) Data 0.000 (0.003) Loss 0.8381 (0.6439) Prec@1 78.125 (82.378) Prec@5 95.312 (95.798) Epoch: [40][1180/1345], lr: 0.00010 Time 0.775 (0.747) Data 0.001 (0.003) Loss 0.5377 (0.6437) Prec@1 85.938 (82.372) Prec@5 96.875 (95.805) Epoch: [40][1200/1345], lr: 0.00010 Time 0.748 (0.747) Data 0.000 (0.003) Loss 0.8279 (0.6445) Prec@1 76.562 (82.342) Prec@5 92.188 (95.796) Epoch: [40][1220/1345], lr: 0.00010 Time 0.866 (0.747) Data 0.001 (0.003) Loss 0.7462 (0.6445) Prec@1 81.250 (82.336) Prec@5 95.312 (95.799) Epoch: [40][1240/1345], lr: 0.00010 Time 0.715 (0.747) Data 0.000 (0.003) Loss 0.6718 (0.6444) Prec@1 87.500 (82.350) Prec@5 95.312 (95.793) Epoch: [40][1260/1345], lr: 0.00010 Time 0.742 (0.747) Data 0.000 (0.003) Loss 0.6447 (0.6441) Prec@1 82.812 (82.353) Prec@5 92.188 (95.806) Epoch: [40][1280/1345], lr: 0.00010 Time 0.785 (0.747) Data 0.000 (0.003) Loss 0.6061 (0.6438) Prec@1 82.812 (82.358) Prec@5 96.875 (95.803) Epoch: [40][1300/1345], lr: 0.00010 Time 0.717 (0.747) Data 0.000 (0.003) Loss 0.7829 (0.6436) Prec@1 71.875 (82.349) Prec@5 93.750 (95.801) Epoch: [40][1320/1345], lr: 0.00010 Time 0.841 (0.747) Data 0.000 (0.003) Loss 0.5573 (0.6436) Prec@1 89.062 (82.332) Prec@5 96.875 (95.812) Epoch: [40][1340/1345], lr: 0.00010 Time 0.709 (0.747) Data 0.000 (0.003) Loss 0.5313 (0.6438) Prec@1 82.812 (82.332) Prec@5 98.438 (95.810) No BN layer Freezing. Test: [0/181] Time 3.593 (3.5932) Loss 2.2470 (2.2470) Prec@1 51.562 (51.562) Prec@5 82.812 (82.812) Test: [20/181] Time 1.103 (0.6087) Loss 2.3196 (2.0849) Prec@1 50.000 (52.604) Prec@5 78.125 (81.399) Test: [40/181] Time 1.004 (0.5331) Loss 2.4525 (2.1556) Prec@1 54.688 (52.096) Prec@5 75.000 (80.259) Test: [60/181] Time 1.216 (0.5089) Loss 2.6458 (2.2102) Prec@1 53.125 (50.820) Prec@5 70.312 (79.739) Test: [80/181] Time 1.096 (0.4964) Loss 1.7574 (2.2117) Prec@1 60.938 (50.424) Prec@5 87.500 (79.572) Test: [100/181] Time 1.137 (0.4880) Loss 2.8899 (2.2031) Prec@1 45.312 (50.572) Prec@5 70.312 (79.688) Test: [120/181] Time 1.128 (0.4844) Loss 2.8037 (2.2126) Prec@1 43.750 (50.646) Prec@5 71.875 (79.584) Test: [140/181] Time 1.159 (0.4818) Loss 3.0564 (2.2093) Prec@1 31.250 (50.477) Prec@5 73.438 (79.754) Test: [160/181] Time 1.127 (0.4803) Loss 2.4597 (2.1957) Prec@1 45.312 (50.660) Prec@5 82.812 (79.901) Testing Results: Prec@1 50.964 Prec@5 79.939 Loss 2.18491 Time 0.4741 No BN layer Freezing. Epoch: [41][0/1345], lr: 0.00010 Time 4.054 (4.054) Data 3.301 (3.301) Loss 0.6094 (0.6094) Prec@1 87.500 (87.500) Prec@5 96.875 (96.875) Epoch: [41][20/1345], lr: 0.00010 Time 0.728 (0.913) Data 0.001 (0.158) Loss 0.5368 (0.6003) Prec@1 79.688 (82.664) Prec@5 95.312 (96.503) Epoch: [41][40/1345], lr: 0.00010 Time 0.739 (0.829) Data 0.000 (0.081) Loss 0.6722 (0.6239) Prec@1 82.812 (82.622) Prec@5 95.312 (96.113) Epoch: [41][60/1345], lr: 0.00010 Time 0.717 (0.800) Data 0.000 (0.055) Loss 0.4275 (0.6310) Prec@1 85.938 (82.710) Prec@5 96.875 (96.055) Epoch: [41][80/1345], lr: 0.00010 Time 0.716 (0.787) Data 0.000 (0.041) Loss 0.5818 (0.6312) Prec@1 85.938 (82.542) Prec@5 95.312 (96.161) Epoch: [41][100/1345], lr: 0.00010 Time 0.717 (0.778) Data 0.000 (0.033) Loss 0.5902 (0.6392) Prec@1 84.375 (82.287) Prec@5 93.750 (95.947) Epoch: [41][120/1345], lr: 0.00010 Time 0.836 (0.774) Data 0.000 (0.028) Loss 0.5683 (0.6395) Prec@1 81.250 (82.244) Prec@5 98.438 (95.984) Epoch: [41][140/1345], lr: 0.00010 Time 0.756 (0.770) Data 0.000 (0.024) Loss 0.6344 (0.6379) Prec@1 84.375 (82.347) Prec@5 95.312 (95.944) Epoch: [41][160/1345], lr: 0.00010 Time 0.747 (0.767) Data 0.000 (0.021) Loss 0.7238 (0.6380) Prec@1 79.688 (82.366) Prec@5 95.312 (95.914) Epoch: [41][180/1345], lr: 0.00010 Time 0.772 (0.766) Data 0.000 (0.019) Loss 0.9036 (0.6364) Prec@1 79.688 (82.381) Prec@5 87.500 (95.865) Epoch: [41][200/1345], lr: 0.00010 Time 0.755 (0.763) Data 0.001 (0.017) Loss 0.8044 (0.6338) Prec@1 78.125 (82.463) Prec@5 93.750 (95.896) Epoch: [41][220/1345], lr: 0.00010 Time 0.714 (0.762) Data 0.000 (0.015) Loss 0.5064 (0.6278) Prec@1 85.938 (82.600) Prec@5 96.875 (95.956) Epoch: [41][240/1345], lr: 0.00010 Time 0.729 (0.761) Data 0.000 (0.014) Loss 0.8164 (0.6285) Prec@1 78.125 (82.612) Prec@5 92.188 (95.980) Epoch: [41][260/1345], lr: 0.00010 Time 0.714 (0.759) Data 0.001 (0.013) Loss 0.5217 (0.6273) Prec@1 82.812 (82.645) Prec@5 95.312 (95.947) Epoch: [41][280/1345], lr: 0.00010 Time 0.839 (0.759) Data 0.000 (0.012) Loss 0.5416 (0.6278) Prec@1 84.375 (82.562) Prec@5 96.875 (95.952) Epoch: [41][300/1345], lr: 0.00010 Time 0.714 (0.758) Data 0.000 (0.011) Loss 0.6050 (0.6274) Prec@1 82.812 (82.569) Prec@5 96.875 (95.956) Epoch: [41][320/1345], lr: 0.00010 Time 0.752 (0.757) Data 0.000 (0.011) Loss 0.5220 (0.6315) Prec@1 82.812 (82.399) Prec@5 96.875 (95.906) Epoch: [41][340/1345], lr: 0.00010 Time 0.769 (0.757) Data 0.000 (0.010) Loss 0.5312 (0.6344) Prec@1 81.250 (82.308) Prec@5 95.312 (95.858) Epoch: [41][360/1345], lr: 0.00010 Time 0.714 (0.756) Data 0.000 (0.010) Loss 0.5076 (0.6325) Prec@1 90.625 (82.349) Prec@5 96.875 (95.892) Epoch: [41][380/1345], lr: 0.00010 Time 0.853 (0.756) Data 0.000 (0.009) Loss 0.5118 (0.6363) Prec@1 89.062 (82.300) Prec@5 98.438 (95.813) Epoch: [41][400/1345], lr: 0.00010 Time 0.713 (0.755) Data 0.000 (0.009) Loss 0.6240 (0.6360) Prec@1 84.375 (82.333) Prec@5 95.312 (95.807) Epoch: [41][420/1345], lr: 0.00010 Time 0.714 (0.755) Data 0.000 (0.008) Loss 0.5402 (0.6369) Prec@1 84.375 (82.401) Prec@5 98.438 (95.814) Epoch: [41][440/1345], lr: 0.00010 Time 0.872 (0.754) Data 0.001 (0.008) Loss 0.6365 (0.6354) Prec@1 82.812 (82.458) Prec@5 93.750 (95.855) Epoch: [41][460/1345], lr: 0.00010 Time 0.738 (0.754) Data 0.000 (0.008) Loss 0.6832 (0.6379) Prec@1 78.125 (82.402) Prec@5 93.750 (95.811) Epoch: [41][480/1345], lr: 0.00010 Time 0.733 (0.754) Data 0.000 (0.007) Loss 0.3955 (0.6377) Prec@1 87.500 (82.397) Prec@5 100.000 (95.835) Epoch: [41][500/1345], lr: 0.00010 Time 0.715 (0.754) Data 0.000 (0.007) Loss 0.3713 (0.6353) Prec@1 92.188 (82.504) Prec@5 98.438 (95.865) Epoch: [41][520/1345], lr: 0.00010 Time 0.717 (0.753) Data 0.000 (0.007) Loss 0.8445 (0.6353) Prec@1 79.688 (82.492) Prec@5 95.312 (95.861) Epoch: [41][540/1345], lr: 0.00010 Time 0.716 (0.753) Data 0.000 (0.007) Loss 0.7858 (0.6346) Prec@1 78.125 (82.535) Prec@5 93.750 (95.858) Epoch: [41][560/1345], lr: 0.00010 Time 0.716 (0.753) Data 0.000 (0.006) Loss 0.7210 (0.6336) Prec@1 87.500 (82.598) Prec@5 95.312 (95.872) Epoch: [41][580/1345], lr: 0.00010 Time 0.714 (0.753) Data 0.000 (0.006) Loss 0.6527 (0.6358) Prec@1 78.125 (82.546) Prec@5 98.438 (95.856) Epoch: [41][600/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.7891 (0.6353) Prec@1 82.812 (82.607) Prec@5 93.750 (95.851) Epoch: [41][620/1345], lr: 0.00010 Time 0.757 (0.752) Data 0.001 (0.006) Loss 0.8127 (0.6339) Prec@1 79.688 (82.654) Prec@5 95.312 (95.884) Epoch: [41][640/1345], lr: 0.00010 Time 0.777 (0.753) Data 0.000 (0.006) Loss 0.5119 (0.6348) Prec@1 87.500 (82.620) Prec@5 96.875 (95.861) Epoch: [41][660/1345], lr: 0.00010 Time 0.714 (0.753) Data 0.000 (0.005) Loss 0.6590 (0.6370) Prec@1 79.688 (82.557) Prec@5 92.188 (95.840) Epoch: [41][680/1345], lr: 0.00010 Time 0.757 (0.753) Data 0.001 (0.005) Loss 0.6344 (0.6359) Prec@1 82.812 (82.558) Prec@5 93.750 (95.845) Epoch: [41][700/1345], lr: 0.00010 Time 0.734 (0.753) Data 0.001 (0.005) Loss 0.5188 (0.6357) Prec@1 87.500 (82.563) Prec@5 95.312 (95.854) Epoch: [41][720/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.005) Loss 0.4652 (0.6364) Prec@1 82.812 (82.529) Prec@5 98.438 (95.846) Epoch: [41][740/1345], lr: 0.00010 Time 0.725 (0.752) Data 0.000 (0.005) Loss 0.6505 (0.6353) Prec@1 78.125 (82.564) Prec@5 98.438 (95.865) Epoch: [41][760/1345], lr: 0.00010 Time 0.770 (0.752) Data 0.000 (0.005) Loss 0.5036 (0.6354) Prec@1 89.062 (82.578) Prec@5 96.875 (95.855) Epoch: [41][780/1345], lr: 0.00010 Time 0.753 (0.752) Data 0.000 (0.005) Loss 0.5392 (0.6349) Prec@1 79.688 (82.584) Prec@5 98.438 (95.849) Epoch: [41][800/1345], lr: 0.00010 Time 0.744 (0.752) Data 0.000 (0.005) Loss 0.7731 (0.6354) Prec@1 81.250 (82.578) Prec@5 93.750 (95.839) Epoch: [41][820/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.004) Loss 0.5321 (0.6354) Prec@1 92.188 (82.557) Prec@5 98.438 (95.845) Epoch: [41][840/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.004) Loss 0.8285 (0.6356) Prec@1 81.250 (82.578) Prec@5 95.312 (95.846) Epoch: [41][860/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.001 (0.004) Loss 0.7280 (0.6360) Prec@1 76.562 (82.560) Prec@5 95.312 (95.841) Epoch: [41][880/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.6991 (0.6358) Prec@1 78.125 (82.539) Prec@5 95.312 (95.834) Epoch: [41][900/1345], lr: 0.00010 Time 0.717 (0.751) Data 0.001 (0.004) Loss 0.6457 (0.6352) Prec@1 82.812 (82.565) Prec@5 95.312 (95.852) Epoch: [41][920/1345], lr: 0.00010 Time 0.738 (0.751) Data 0.000 (0.004) Loss 0.4494 (0.6350) Prec@1 89.062 (82.567) Prec@5 96.875 (95.845) Epoch: [41][940/1345], lr: 0.00010 Time 0.733 (0.751) Data 0.000 (0.004) Loss 0.4646 (0.6348) Prec@1 85.938 (82.577) Prec@5 100.000 (95.839) Epoch: [41][960/1345], lr: 0.00010 Time 0.747 (0.751) Data 0.000 (0.004) Loss 0.5384 (0.6356) Prec@1 87.500 (82.559) Prec@5 96.875 (95.830) Epoch: [41][980/1345], lr: 0.00010 Time 0.713 (0.751) Data 0.000 (0.004) Loss 0.5392 (0.6358) Prec@1 95.312 (82.553) Prec@5 95.312 (95.830) Epoch: [41][1000/1345], lr: 0.00010 Time 0.883 (0.751) Data 0.000 (0.004) Loss 0.9758 (0.6360) Prec@1 73.438 (82.536) Prec@5 95.312 (95.834) Epoch: [41][1020/1345], lr: 0.00010 Time 0.834 (0.751) Data 0.000 (0.004) Loss 0.8810 (0.6357) Prec@1 79.688 (82.565) Prec@5 89.062 (95.833) Epoch: [41][1040/1345], lr: 0.00010 Time 0.758 (0.751) Data 0.000 (0.004) Loss 0.5365 (0.6344) Prec@1 82.812 (82.604) Prec@5 96.875 (95.856) Epoch: [41][1060/1345], lr: 0.00010 Time 0.749 (0.751) Data 0.000 (0.004) Loss 0.4959 (0.6348) Prec@1 84.375 (82.558) Prec@5 96.875 (95.850) Epoch: [41][1080/1345], lr: 0.00010 Time 0.774 (0.751) Data 0.001 (0.004) Loss 0.6946 (0.6344) Prec@1 84.375 (82.578) Prec@5 95.312 (95.856) Epoch: [41][1100/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.003) Loss 0.5477 (0.6342) Prec@1 81.250 (82.571) Prec@5 98.438 (95.862) Epoch: [41][1120/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.003) Loss 0.5708 (0.6341) Prec@1 84.375 (82.580) Prec@5 96.875 (95.860) Epoch: [41][1140/1345], lr: 0.00010 Time 0.713 (0.751) Data 0.000 (0.003) Loss 0.9477 (0.6349) Prec@1 67.188 (82.541) Prec@5 95.312 (95.847) Epoch: [41][1160/1345], lr: 0.00010 Time 0.865 (0.751) Data 0.000 (0.003) Loss 0.4298 (0.6348) Prec@1 87.500 (82.543) Prec@5 98.438 (95.845) Epoch: [41][1180/1345], lr: 0.00010 Time 0.839 (0.751) Data 0.000 (0.003) Loss 0.5341 (0.6344) Prec@1 85.938 (82.528) Prec@5 96.875 (95.863) Epoch: [41][1200/1345], lr: 0.00010 Time 0.783 (0.751) Data 0.000 (0.003) Loss 0.4955 (0.6348) Prec@1 90.625 (82.512) Prec@5 98.438 (95.855) Epoch: [41][1220/1345], lr: 0.00010 Time 0.740 (0.751) Data 0.000 (0.003) Loss 0.6087 (0.6346) Prec@1 79.688 (82.508) Prec@5 98.438 (95.856) Epoch: [41][1240/1345], lr: 0.00010 Time 0.715 (0.751) Data 0.000 (0.003) Loss 0.6420 (0.6345) Prec@1 84.375 (82.518) Prec@5 96.875 (95.861) Epoch: [41][1260/1345], lr: 0.00010 Time 0.712 (0.751) Data 0.000 (0.003) Loss 0.4682 (0.6338) Prec@1 81.250 (82.519) Prec@5 98.438 (95.874) Epoch: [41][1280/1345], lr: 0.00010 Time 0.713 (0.751) Data 0.000 (0.003) Loss 0.7331 (0.6342) Prec@1 85.938 (82.531) Prec@5 92.188 (95.866) Epoch: [41][1300/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.003) Loss 0.5188 (0.6341) Prec@1 87.500 (82.531) Prec@5 96.875 (95.879) Epoch: [41][1320/1345], lr: 0.00010 Time 0.826 (0.751) Data 0.000 (0.003) Loss 0.6347 (0.6342) Prec@1 79.688 (82.542) Prec@5 95.312 (95.880) Epoch: [41][1340/1345], lr: 0.00010 Time 0.802 (0.751) Data 0.000 (0.003) Loss 0.7929 (0.6344) Prec@1 81.250 (82.527) Prec@5 93.750 (95.878) No BN layer Freezing. Test: [0/181] Time 2.929 (2.9294) Loss 2.2760 (2.2760) Prec@1 51.562 (51.562) Prec@5 82.812 (82.812) Test: [20/181] Time 1.269 (0.5944) Loss 2.3308 (2.1014) Prec@1 51.562 (53.274) Prec@5 75.000 (81.399) Test: [40/181] Time 0.994 (0.5332) Loss 2.5141 (2.1754) Prec@1 54.688 (52.287) Prec@5 75.000 (80.069) Test: [60/181] Time 1.050 (0.5099) Loss 2.7212 (2.2258) Prec@1 50.000 (51.537) Prec@5 70.312 (79.611) Test: [80/181] Time 1.042 (0.4949) Loss 1.7404 (2.2226) Prec@1 60.938 (51.215) Prec@5 85.938 (79.688) Test: [100/181] Time 1.162 (0.4869) Loss 3.0671 (2.2150) Prec@1 42.188 (51.114) Prec@5 70.312 (79.904) Test: [120/181] Time 1.160 (0.4818) Loss 2.9022 (2.2257) Prec@1 43.750 (51.175) Prec@5 71.875 (79.868) Test: [140/181] Time 0.986 (0.4771) Loss 3.0624 (2.2225) Prec@1 35.938 (50.898) Prec@5 71.875 (80.042) Test: [160/181] Time 1.257 (0.4785) Loss 2.5245 (2.2097) Prec@1 46.875 (51.126) Prec@5 82.812 (80.202) Testing Results: Prec@1 51.372 Prec@5 80.243 Loss 2.20038 Time 0.4719 No BN layer Freezing. Epoch: [42][0/1345], lr: 0.00010 Time 4.517 (4.517) Data 3.562 (3.562) Loss 0.6182 (0.6182) Prec@1 84.375 (84.375) Prec@5 95.312 (95.312) Epoch: [42][20/1345], lr: 0.00010 Time 0.713 (0.910) Data 0.000 (0.170) Loss 0.4874 (0.6287) Prec@1 85.938 (82.589) Prec@5 96.875 (95.982) Epoch: [42][40/1345], lr: 0.00010 Time 0.716 (0.831) Data 0.001 (0.087) Loss 0.6174 (0.6390) Prec@1 85.938 (82.431) Prec@5 93.750 (95.922) Epoch: [42][60/1345], lr: 0.00010 Time 0.713 (0.803) Data 0.000 (0.059) Loss 0.5694 (0.6145) Prec@1 85.938 (83.248) Prec@5 96.875 (96.132) Epoch: [42][80/1345], lr: 0.00010 Time 0.750 (0.791) Data 0.000 (0.044) Loss 0.8354 (0.6246) Prec@1 76.562 (82.986) Prec@5 92.188 (95.949) Epoch: [42][100/1345], lr: 0.00010 Time 0.846 (0.782) Data 0.001 (0.036) Loss 0.7691 (0.6306) Prec@1 78.125 (82.751) Prec@5 96.875 (95.900) Epoch: [42][120/1345], lr: 0.00010 Time 0.844 (0.777) Data 0.000 (0.030) Loss 0.5814 (0.6256) Prec@1 82.812 (82.787) Prec@5 95.312 (95.919) Epoch: [42][140/1345], lr: 0.00010 Time 0.746 (0.773) Data 0.001 (0.026) Loss 0.6737 (0.6318) Prec@1 82.812 (82.702) Prec@5 95.312 (95.789) Epoch: [42][160/1345], lr: 0.00010 Time 0.715 (0.769) Data 0.000 (0.023) Loss 0.4713 (0.6314) Prec@1 87.500 (82.725) Prec@5 96.875 (95.837) Epoch: [42][180/1345], lr: 0.00010 Time 0.715 (0.765) Data 0.000 (0.020) Loss 0.6996 (0.6281) Prec@1 81.250 (82.657) Prec@5 95.312 (95.865) Epoch: [42][200/1345], lr: 0.00010 Time 0.840 (0.764) Data 0.000 (0.018) Loss 0.7147 (0.6272) Prec@1 82.812 (82.696) Prec@5 95.312 (95.942) Epoch: [42][220/1345], lr: 0.00010 Time 0.827 (0.762) Data 0.000 (0.017) Loss 0.6400 (0.6279) Prec@1 82.812 (82.664) Prec@5 93.750 (95.913) Epoch: [42][240/1345], lr: 0.00010 Time 0.715 (0.761) Data 0.000 (0.015) Loss 0.8108 (0.6309) Prec@1 81.250 (82.618) Prec@5 92.188 (95.890) Epoch: [42][260/1345], lr: 0.00010 Time 0.714 (0.759) Data 0.000 (0.014) Loss 0.3840 (0.6316) Prec@1 90.625 (82.555) Prec@5 98.438 (95.929) Epoch: [42][280/1345], lr: 0.00010 Time 0.730 (0.758) Data 0.000 (0.013) Loss 0.5879 (0.6314) Prec@1 82.812 (82.601) Prec@5 96.875 (95.891) Epoch: [42][300/1345], lr: 0.00010 Time 0.761 (0.757) Data 0.001 (0.012) Loss 0.4989 (0.6280) Prec@1 87.500 (82.615) Prec@5 96.875 (95.977) Epoch: [42][320/1345], lr: 0.00010 Time 0.713 (0.757) Data 0.000 (0.012) Loss 0.4810 (0.6255) Prec@1 87.500 (82.637) Prec@5 96.875 (95.984) Epoch: [42][340/1345], lr: 0.00010 Time 0.845 (0.757) Data 0.000 (0.011) Loss 0.9897 (0.6271) Prec@1 75.000 (82.620) Prec@5 90.625 (96.000) Epoch: [42][360/1345], lr: 0.00010 Time 0.833 (0.756) Data 0.000 (0.010) Loss 0.5179 (0.6235) Prec@1 87.500 (82.787) Prec@5 100.000 (96.040) Epoch: [42][380/1345], lr: 0.00010 Time 0.731 (0.755) Data 0.000 (0.010) Loss 0.7659 (0.6243) Prec@1 79.688 (82.804) Prec@5 96.875 (96.022) Epoch: [42][400/1345], lr: 0.00010 Time 0.712 (0.754) Data 0.000 (0.009) Loss 0.8166 (0.6277) Prec@1 81.250 (82.711) Prec@5 93.750 (95.990) Epoch: [42][420/1345], lr: 0.00010 Time 0.712 (0.754) Data 0.000 (0.009) Loss 0.5940 (0.6275) Prec@1 85.938 (82.705) Prec@5 95.312 (96.007) Epoch: [42][440/1345], lr: 0.00010 Time 0.859 (0.755) Data 0.000 (0.009) Loss 0.9775 (0.6285) Prec@1 76.562 (82.699) Prec@5 89.062 (95.975) Epoch: [42][460/1345], lr: 0.00010 Time 0.873 (0.754) Data 0.000 (0.008) Loss 0.8031 (0.6298) Prec@1 82.812 (82.670) Prec@5 95.312 (95.994) Epoch: [42][480/1345], lr: 0.00010 Time 0.716 (0.754) Data 0.000 (0.008) Loss 0.4477 (0.6298) Prec@1 90.625 (82.676) Prec@5 100.000 (96.011) Epoch: [42][500/1345], lr: 0.00010 Time 0.715 (0.754) Data 0.000 (0.008) Loss 0.5934 (0.6277) Prec@1 84.375 (82.728) Prec@5 95.312 (96.030) Epoch: [42][520/1345], lr: 0.00010 Time 0.771 (0.754) Data 0.000 (0.007) Loss 0.5890 (0.6279) Prec@1 85.938 (82.708) Prec@5 96.875 (96.026) Epoch: [42][540/1345], lr: 0.00010 Time 0.715 (0.754) Data 0.001 (0.007) Loss 0.6779 (0.6298) Prec@1 84.375 (82.659) Prec@5 96.875 (96.009) Epoch: [42][560/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.007) Loss 0.4322 (0.6301) Prec@1 89.062 (82.645) Prec@5 96.875 (96.000) Epoch: [42][580/1345], lr: 0.00010 Time 0.714 (0.753) Data 0.000 (0.007) Loss 0.7462 (0.6280) Prec@1 81.250 (82.700) Prec@5 96.875 (96.044) Epoch: [42][600/1345], lr: 0.00010 Time 0.845 (0.753) Data 0.000 (0.006) Loss 0.5406 (0.6272) Prec@1 82.812 (82.698) Prec@5 98.438 (96.043) Epoch: [42][620/1345], lr: 0.00010 Time 0.861 (0.753) Data 0.000 (0.006) Loss 0.5406 (0.6271) Prec@1 82.812 (82.692) Prec@5 96.875 (96.047) Epoch: [42][640/1345], lr: 0.00010 Time 0.729 (0.753) Data 0.000 (0.006) Loss 0.5768 (0.6273) Prec@1 82.812 (82.695) Prec@5 96.875 (96.029) Epoch: [42][660/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.006) Loss 0.5649 (0.6287) Prec@1 87.500 (82.682) Prec@5 96.875 (96.000) Epoch: [42][680/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.006) Loss 0.4289 (0.6290) Prec@1 87.500 (82.679) Prec@5 95.312 (95.999) Epoch: [42][700/1345], lr: 0.00010 Time 0.771 (0.752) Data 0.000 (0.006) Loss 0.5930 (0.6322) Prec@1 85.938 (82.603) Prec@5 96.875 (95.961) Epoch: [42][720/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.005) Loss 0.6557 (0.6320) Prec@1 85.938 (82.643) Prec@5 93.750 (95.958) Epoch: [42][740/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.005) Loss 0.3754 (0.6325) Prec@1 90.625 (82.627) Prec@5 100.000 (95.958) Epoch: [42][760/1345], lr: 0.00010 Time 0.883 (0.752) Data 0.000 (0.005) Loss 0.6793 (0.6326) Prec@1 82.812 (82.632) Prec@5 90.625 (95.945) Epoch: [42][780/1345], lr: 0.00010 Time 0.858 (0.751) Data 0.001 (0.005) Loss 0.7776 (0.6324) Prec@1 73.438 (82.638) Prec@5 93.750 (95.941) Epoch: [42][800/1345], lr: 0.00010 Time 0.722 (0.751) Data 0.000 (0.005) Loss 0.6097 (0.6329) Prec@1 84.375 (82.625) Prec@5 96.875 (95.952) Epoch: [42][820/1345], lr: 0.00010 Time 0.766 (0.751) Data 0.000 (0.005) Loss 0.5538 (0.6322) Prec@1 87.500 (82.632) Prec@5 96.875 (95.956) Epoch: [42][840/1345], lr: 0.00010 Time 0.732 (0.751) Data 0.000 (0.005) Loss 0.5582 (0.6319) Prec@1 81.250 (82.645) Prec@5 95.312 (95.957) Epoch: [42][860/1345], lr: 0.00010 Time 0.756 (0.751) Data 0.000 (0.005) Loss 0.4295 (0.6322) Prec@1 92.188 (82.638) Prec@5 95.312 (95.937) Epoch: [42][880/1345], lr: 0.00010 Time 0.747 (0.751) Data 0.000 (0.005) Loss 0.6212 (0.6321) Prec@1 79.688 (82.642) Prec@5 96.875 (95.944) Epoch: [42][900/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.4410 (0.6320) Prec@1 85.938 (82.641) Prec@5 98.438 (95.945) Epoch: [42][920/1345], lr: 0.00010 Time 0.841 (0.750) Data 0.000 (0.004) Loss 0.5004 (0.6310) Prec@1 85.938 (82.667) Prec@5 96.875 (95.966) Epoch: [42][940/1345], lr: 0.00010 Time 0.843 (0.750) Data 0.000 (0.004) Loss 0.3265 (0.6290) Prec@1 95.312 (82.728) Prec@5 100.000 (95.998) Epoch: [42][960/1345], lr: 0.00010 Time 0.750 (0.750) Data 0.000 (0.004) Loss 0.7381 (0.6286) Prec@1 84.375 (82.751) Prec@5 93.750 (96.002) Epoch: [42][980/1345], lr: 0.00010 Time 0.760 (0.750) Data 0.000 (0.004) Loss 0.4736 (0.6278) Prec@1 79.688 (82.768) Prec@5 100.000 (96.017) Epoch: [42][1000/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.004) Loss 0.6665 (0.6277) Prec@1 79.688 (82.759) Prec@5 96.875 (96.015) Epoch: [42][1020/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.001 (0.004) Loss 0.8983 (0.6273) Prec@1 70.312 (82.756) Prec@5 92.188 (96.013) Epoch: [42][1040/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.004) Loss 0.7586 (0.6277) Prec@1 81.250 (82.755) Prec@5 90.625 (95.991) Epoch: [42][1060/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.000 (0.004) Loss 0.6111 (0.6281) Prec@1 84.375 (82.721) Prec@5 95.312 (95.994) Epoch: [42][1080/1345], lr: 0.00010 Time 0.857 (0.750) Data 0.000 (0.004) Loss 0.4925 (0.6271) Prec@1 81.250 (82.749) Prec@5 100.000 (96.002) Epoch: [42][1100/1345], lr: 0.00010 Time 0.856 (0.750) Data 0.000 (0.004) Loss 0.4482 (0.6267) Prec@1 85.938 (82.744) Prec@5 96.875 (96.005) Epoch: [42][1120/1345], lr: 0.00010 Time 0.745 (0.750) Data 0.001 (0.004) Loss 0.5774 (0.6262) Prec@1 81.250 (82.754) Prec@5 96.875 (96.007) Epoch: [42][1140/1345], lr: 0.00010 Time 0.717 (0.749) Data 0.000 (0.004) Loss 0.8323 (0.6256) Prec@1 75.000 (82.785) Prec@5 93.750 (96.011) Epoch: [42][1160/1345], lr: 0.00010 Time 0.736 (0.750) Data 0.000 (0.004) Loss 0.5841 (0.6273) Prec@1 84.375 (82.741) Prec@5 93.750 (95.998) Epoch: [42][1180/1345], lr: 0.00010 Time 0.769 (0.750) Data 0.000 (0.003) Loss 0.5779 (0.6274) Prec@1 82.812 (82.737) Prec@5 95.312 (95.990) Epoch: [42][1200/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.003) Loss 0.7209 (0.6281) Prec@1 78.125 (82.708) Prec@5 96.875 (95.980) Epoch: [42][1220/1345], lr: 0.00010 Time 0.720 (0.750) Data 0.000 (0.003) Loss 0.5360 (0.6280) Prec@1 89.062 (82.738) Prec@5 93.750 (95.973) Epoch: [42][1240/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.003) Loss 0.7018 (0.6280) Prec@1 76.562 (82.726) Prec@5 98.438 (95.981) Epoch: [42][1260/1345], lr: 0.00010 Time 0.735 (0.750) Data 0.000 (0.003) Loss 0.5946 (0.6279) Prec@1 85.938 (82.742) Prec@5 95.312 (95.983) Epoch: [42][1280/1345], lr: 0.00010 Time 0.841 (0.750) Data 0.000 (0.003) Loss 0.7993 (0.6280) Prec@1 75.000 (82.730) Prec@5 93.750 (95.981) Epoch: [42][1300/1345], lr: 0.00010 Time 0.766 (0.750) Data 0.000 (0.003) Loss 0.5585 (0.6266) Prec@1 81.250 (82.768) Prec@5 96.875 (96.002) Epoch: [42][1320/1345], lr: 0.00010 Time 0.735 (0.750) Data 0.000 (0.003) Loss 0.5675 (0.6268) Prec@1 81.250 (82.763) Prec@5 100.000 (96.002) Epoch: [42][1340/1345], lr: 0.00010 Time 0.710 (0.750) Data 0.000 (0.003) Loss 0.5354 (0.6264) Prec@1 81.250 (82.769) Prec@5 100.000 (96.013) No BN layer Freezing. Test: [0/181] Time 3.147 (3.1468) Loss 2.2199 (2.2199) Prec@1 51.562 (51.562) Prec@5 81.250 (81.250) Test: [20/181] Time 1.006 (0.5832) Loss 2.2880 (2.0926) Prec@1 54.688 (52.530) Prec@5 75.000 (80.804) Test: [40/181] Time 0.811 (0.5224) Loss 2.5250 (2.1656) Prec@1 54.688 (52.020) Prec@5 71.875 (80.297) Test: [60/181] Time 0.868 (0.5049) Loss 2.6890 (2.2194) Prec@1 53.125 (51.025) Prec@5 71.875 (79.559) Test: [80/181] Time 1.047 (0.4957) Loss 1.7235 (2.2202) Prec@1 59.375 (50.637) Prec@5 87.500 (79.533) Test: [100/181] Time 1.251 (0.4921) Loss 2.9391 (2.2088) Prec@1 40.625 (50.634) Prec@5 71.875 (79.842) Test: [120/181] Time 0.821 (0.4837) Loss 2.8854 (2.2192) Prec@1 43.750 (50.659) Prec@5 75.000 (79.985) Test: [140/181] Time 0.628 (0.4787) Loss 3.1268 (2.2171) Prec@1 31.250 (50.410) Prec@5 73.438 (80.098) Test: [160/181] Time 0.865 (0.4782) Loss 2.5349 (2.2035) Prec@1 50.000 (50.796) Prec@5 84.375 (80.250) Testing Results: Prec@1 51.042 Prec@5 80.208 Loss 2.19484 Time 0.4746 No BN layer Freezing. Epoch: [43][0/1345], lr: 0.00010 Time 4.026 (4.026) Data 3.276 (3.276) Loss 0.4212 (0.4212) Prec@1 90.625 (90.625) Prec@5 98.438 (98.438) Epoch: [43][20/1345], lr: 0.00010 Time 0.757 (0.919) Data 0.000 (0.156) Loss 0.8177 (0.5903) Prec@1 82.812 (84.821) Prec@5 93.750 (96.429) Epoch: [43][40/1345], lr: 0.00010 Time 0.714 (0.839) Data 0.000 (0.080) Loss 0.6877 (0.6114) Prec@1 78.125 (83.651) Prec@5 95.312 (96.265) Epoch: [43][60/1345], lr: 0.00010 Time 0.750 (0.811) Data 0.000 (0.054) Loss 0.9337 (0.6125) Prec@1 81.250 (83.453) Prec@5 92.188 (96.055) Epoch: [43][80/1345], lr: 0.00010 Time 0.716 (0.793) Data 0.000 (0.041) Loss 0.6892 (0.6162) Prec@1 84.375 (83.218) Prec@5 98.438 (96.007) Epoch: [43][100/1345], lr: 0.00010 Time 0.715 (0.782) Data 0.000 (0.033) Loss 0.7239 (0.6195) Prec@1 79.688 (82.983) Prec@5 95.312 (95.947) Epoch: [43][120/1345], lr: 0.00010 Time 0.769 (0.774) Data 0.000 (0.028) Loss 0.6596 (0.6155) Prec@1 78.125 (83.110) Prec@5 98.438 (96.036) Epoch: [43][140/1345], lr: 0.00010 Time 0.713 (0.772) Data 0.000 (0.024) Loss 0.8233 (0.6175) Prec@1 78.125 (82.990) Prec@5 90.625 (95.944) Epoch: [43][160/1345], lr: 0.00010 Time 0.746 (0.768) Data 0.000 (0.021) Loss 0.7015 (0.6157) Prec@1 81.250 (83.084) Prec@5 93.750 (95.934) Epoch: [43][180/1345], lr: 0.00010 Time 0.778 (0.766) Data 0.000 (0.019) Loss 0.6291 (0.6180) Prec@1 82.812 (82.968) Prec@5 95.312 (96.003) Epoch: [43][200/1345], lr: 0.00010 Time 0.712 (0.764) Data 0.001 (0.017) Loss 0.3940 (0.6128) Prec@1 89.062 (83.053) Prec@5 98.438 (96.004) Epoch: [43][220/1345], lr: 0.00010 Time 0.881 (0.762) Data 0.001 (0.015) Loss 0.6353 (0.6100) Prec@1 85.938 (83.117) Prec@5 96.875 (96.041) Epoch: [43][240/1345], lr: 0.00010 Time 0.853 (0.761) Data 0.001 (0.014) Loss 0.5737 (0.6117) Prec@1 84.375 (83.020) Prec@5 96.875 (96.065) Epoch: [43][260/1345], lr: 0.00010 Time 0.717 (0.760) Data 0.000 (0.013) Loss 0.6344 (0.6128) Prec@1 82.812 (83.076) Prec@5 98.438 (96.049) Epoch: [43][280/1345], lr: 0.00010 Time 0.714 (0.758) Data 0.000 (0.012) Loss 0.5786 (0.6108) Prec@1 82.812 (83.152) Prec@5 96.875 (96.102) Epoch: [43][300/1345], lr: 0.00010 Time 0.813 (0.757) Data 0.000 (0.011) Loss 0.5973 (0.6131) Prec@1 81.250 (83.093) Prec@5 96.875 (96.096) Epoch: [43][320/1345], lr: 0.00010 Time 0.723 (0.756) Data 0.000 (0.011) Loss 0.6167 (0.6130) Prec@1 82.812 (83.158) Prec@5 95.312 (96.072) Epoch: [43][340/1345], lr: 0.00010 Time 0.714 (0.755) Data 0.000 (0.010) Loss 0.7704 (0.6151) Prec@1 78.125 (83.060) Prec@5 96.875 (96.059) Epoch: [43][360/1345], lr: 0.00010 Time 0.756 (0.755) Data 0.001 (0.010) Loss 0.6714 (0.6187) Prec@1 82.812 (82.929) Prec@5 96.875 (96.001) Epoch: [43][380/1345], lr: 0.00010 Time 0.713 (0.755) Data 0.000 (0.009) Loss 0.7875 (0.6141) Prec@1 78.125 (83.013) Prec@5 95.312 (96.059) Epoch: [43][400/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.009) Loss 0.8255 (0.6133) Prec@1 78.125 (83.066) Prec@5 93.750 (96.068) Epoch: [43][420/1345], lr: 0.00010 Time 0.804 (0.753) Data 0.001 (0.008) Loss 0.5307 (0.6119) Prec@1 87.500 (83.076) Prec@5 96.875 (96.092) Epoch: [43][440/1345], lr: 0.00010 Time 0.736 (0.753) Data 0.000 (0.008) Loss 0.3758 (0.6132) Prec@1 90.625 (83.057) Prec@5 98.438 (96.078) Epoch: [43][460/1345], lr: 0.00010 Time 0.716 (0.753) Data 0.000 (0.008) Loss 0.6874 (0.6122) Prec@1 84.375 (83.155) Prec@5 96.875 (96.062) Epoch: [43][480/1345], lr: 0.00010 Time 0.719 (0.753) Data 0.000 (0.007) Loss 0.5897 (0.6131) Prec@1 76.562 (83.095) Prec@5 93.750 (96.037) Epoch: [43][500/1345], lr: 0.00010 Time 0.717 (0.752) Data 0.000 (0.007) Loss 0.8630 (0.6151) Prec@1 79.688 (83.081) Prec@5 90.625 (96.008) Epoch: [43][520/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.007) Loss 0.7577 (0.6163) Prec@1 84.375 (83.082) Prec@5 93.750 (95.969) Epoch: [43][540/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.007) Loss 0.5491 (0.6166) Prec@1 79.688 (83.067) Prec@5 100.000 (95.985) Epoch: [43][560/1345], lr: 0.00010 Time 0.831 (0.751) Data 0.000 (0.006) Loss 0.8112 (0.6186) Prec@1 75.000 (82.991) Prec@5 93.750 (95.984) Epoch: [43][580/1345], lr: 0.00010 Time 0.875 (0.751) Data 0.000 (0.006) Loss 0.6758 (0.6204) Prec@1 76.562 (82.909) Prec@5 95.312 (95.939) Epoch: [43][600/1345], lr: 0.00010 Time 0.745 (0.751) Data 0.000 (0.006) Loss 0.5778 (0.6209) Prec@1 84.375 (82.909) Prec@5 98.438 (95.926) Epoch: [43][620/1345], lr: 0.00010 Time 0.755 (0.751) Data 0.000 (0.006) Loss 0.4902 (0.6221) Prec@1 84.375 (82.875) Prec@5 96.875 (95.921) Epoch: [43][640/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.006) Loss 0.5751 (0.6227) Prec@1 76.562 (82.849) Prec@5 98.438 (95.919) Epoch: [43][660/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.005) Loss 0.8092 (0.6239) Prec@1 78.125 (82.791) Prec@5 90.625 (95.906) Epoch: [43][680/1345], lr: 0.00010 Time 0.775 (0.751) Data 0.001 (0.005) Loss 0.8906 (0.6250) Prec@1 71.875 (82.748) Prec@5 93.750 (95.893) Epoch: [43][700/1345], lr: 0.00010 Time 0.711 (0.751) Data 0.001 (0.005) Loss 0.7882 (0.6260) Prec@1 76.562 (82.717) Prec@5 93.750 (95.899) Epoch: [43][720/1345], lr: 0.00010 Time 0.853 (0.750) Data 0.000 (0.005) Loss 0.6737 (0.6262) Prec@1 78.125 (82.704) Prec@5 95.312 (95.895) Epoch: [43][740/1345], lr: 0.00010 Time 0.738 (0.750) Data 0.000 (0.005) Loss 0.6962 (0.6245) Prec@1 79.688 (82.764) Prec@5 93.750 (95.901) Epoch: [43][760/1345], lr: 0.00010 Time 0.748 (0.750) Data 0.001 (0.005) Loss 0.5849 (0.6248) Prec@1 79.688 (82.749) Prec@5 95.312 (95.908) Epoch: [43][780/1345], lr: 0.00010 Time 0.761 (0.750) Data 0.001 (0.005) Loss 0.5806 (0.6244) Prec@1 81.250 (82.770) Prec@5 98.438 (95.913) Epoch: [43][800/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.005) Loss 0.6418 (0.6239) Prec@1 81.250 (82.820) Prec@5 96.875 (95.909) Epoch: [43][820/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.004) Loss 0.5978 (0.6237) Prec@1 82.812 (82.830) Prec@5 96.875 (95.901) Epoch: [43][840/1345], lr: 0.00010 Time 0.821 (0.750) Data 0.000 (0.004) Loss 0.4848 (0.6245) Prec@1 89.062 (82.805) Prec@5 100.000 (95.903) Epoch: [43][860/1345], lr: 0.00010 Time 0.832 (0.750) Data 0.000 (0.004) Loss 0.6621 (0.6256) Prec@1 79.688 (82.769) Prec@5 95.312 (95.899) Epoch: [43][880/1345], lr: 0.00010 Time 0.735 (0.750) Data 0.001 (0.004) Loss 0.5796 (0.6249) Prec@1 84.375 (82.802) Prec@5 96.875 (95.907) Epoch: [43][900/1345], lr: 0.00010 Time 0.734 (0.750) Data 0.000 (0.004) Loss 0.6155 (0.6250) Prec@1 87.500 (82.781) Prec@5 98.438 (95.911) Epoch: [43][920/1345], lr: 0.00010 Time 0.739 (0.750) Data 0.000 (0.004) Loss 0.4892 (0.6251) Prec@1 82.812 (82.780) Prec@5 98.438 (95.889) Epoch: [43][940/1345], lr: 0.00010 Time 0.760 (0.750) Data 0.000 (0.004) Loss 0.7096 (0.6247) Prec@1 79.688 (82.788) Prec@5 93.750 (95.902) Epoch: [43][960/1345], lr: 0.00010 Time 0.722 (0.749) Data 0.000 (0.004) Loss 0.3565 (0.6252) Prec@1 92.188 (82.772) Prec@5 98.438 (95.888) Epoch: [43][980/1345], lr: 0.00010 Time 0.714 (0.749) Data 0.000 (0.004) Loss 0.5179 (0.6245) Prec@1 81.250 (82.797) Prec@5 100.000 (95.900) Epoch: [43][1000/1345], lr: 0.00010 Time 0.844 (0.749) Data 0.000 (0.004) Loss 0.4939 (0.6249) Prec@1 90.625 (82.800) Prec@5 98.438 (95.907) Epoch: [43][1020/1345], lr: 0.00010 Time 0.808 (0.749) Data 0.000 (0.004) Loss 0.4479 (0.6236) Prec@1 87.500 (82.817) Prec@5 98.438 (95.934) Epoch: [43][1040/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.004) Loss 0.3472 (0.6229) Prec@1 92.188 (82.828) Prec@5 98.438 (95.944) Epoch: [43][1060/1345], lr: 0.00010 Time 0.712 (0.749) Data 0.000 (0.004) Loss 0.4383 (0.6227) Prec@1 87.500 (82.830) Prec@5 95.312 (95.953) Epoch: [43][1080/1345], lr: 0.00010 Time 0.712 (0.749) Data 0.000 (0.004) Loss 0.6479 (0.6226) Prec@1 78.125 (82.821) Prec@5 98.438 (95.951) Epoch: [43][1100/1345], lr: 0.00010 Time 0.714 (0.749) Data 0.000 (0.003) Loss 0.5950 (0.6231) Prec@1 82.812 (82.805) Prec@5 98.438 (95.957) Epoch: [43][1120/1345], lr: 0.00010 Time 0.765 (0.749) Data 0.000 (0.003) Loss 0.5086 (0.6228) Prec@1 87.500 (82.832) Prec@5 96.875 (95.961) Epoch: [43][1140/1345], lr: 0.00010 Time 0.741 (0.749) Data 0.000 (0.003) Loss 0.7000 (0.6226) Prec@1 78.125 (82.843) Prec@5 95.312 (95.959) Epoch: [43][1160/1345], lr: 0.00010 Time 0.882 (0.749) Data 0.000 (0.003) Loss 0.6359 (0.6223) Prec@1 81.250 (82.841) Prec@5 96.875 (95.971) Epoch: [43][1180/1345], lr: 0.00010 Time 0.898 (0.749) Data 0.001 (0.003) Loss 0.8548 (0.6235) Prec@1 79.688 (82.798) Prec@5 95.312 (95.970) Epoch: [43][1200/1345], lr: 0.00010 Time 0.716 (0.749) Data 0.000 (0.003) Loss 0.8231 (0.6241) Prec@1 76.562 (82.790) Prec@5 90.625 (95.964) Epoch: [43][1220/1345], lr: 0.00010 Time 0.775 (0.748) Data 0.000 (0.003) Loss 0.5846 (0.6238) Prec@1 84.375 (82.812) Prec@5 92.188 (95.955) Epoch: [43][1240/1345], lr: 0.00010 Time 0.744 (0.748) Data 0.000 (0.003) Loss 0.5155 (0.6235) Prec@1 87.500 (82.843) Prec@5 96.875 (95.961) Epoch: [43][1260/1345], lr: 0.00010 Time 0.754 (0.748) Data 0.000 (0.003) Loss 0.6642 (0.6245) Prec@1 84.375 (82.806) Prec@5 95.312 (95.953) Epoch: [43][1280/1345], lr: 0.00010 Time 0.750 (0.748) Data 0.000 (0.003) Loss 0.5095 (0.6250) Prec@1 87.500 (82.805) Prec@5 98.438 (95.942) Epoch: [43][1300/1345], lr: 0.00010 Time 0.715 (0.748) Data 0.000 (0.003) Loss 0.5250 (0.6255) Prec@1 85.938 (82.782) Prec@5 95.312 (95.939) Epoch: [43][1320/1345], lr: 0.00010 Time 0.823 (0.748) Data 0.000 (0.003) Loss 0.5091 (0.6248) Prec@1 84.375 (82.801) Prec@5 100.000 (95.957) Epoch: [43][1340/1345], lr: 0.00010 Time 0.817 (0.748) Data 0.000 (0.003) Loss 0.5910 (0.6252) Prec@1 84.375 (82.800) Prec@5 96.875 (95.956) No BN layer Freezing. Test: [0/181] Time 3.147 (3.1465) Loss 2.2923 (2.2923) Prec@1 50.000 (50.000) Prec@5 82.812 (82.812) Test: [20/181] Time 0.945 (0.5915) Loss 2.2800 (2.0940) Prec@1 53.125 (52.902) Prec@5 75.000 (81.101) Test: [40/181] Time 1.273 (0.5425) Loss 2.5110 (2.1664) Prec@1 54.688 (52.058) Prec@5 75.000 (80.259) Test: [60/181] Time 1.062 (0.5135) Loss 2.6722 (2.2188) Prec@1 51.562 (50.973) Prec@5 73.438 (79.662) Test: [80/181] Time 1.363 (0.4999) Loss 1.7383 (2.2189) Prec@1 59.375 (50.791) Prec@5 89.062 (79.437) Test: [100/181] Time 1.275 (0.4941) Loss 2.9552 (2.2070) Prec@1 45.312 (50.866) Prec@5 68.750 (79.595) Test: [120/181] Time 1.311 (0.4892) Loss 2.8817 (2.2181) Prec@1 43.750 (50.814) Prec@5 73.438 (79.649) Test: [140/181] Time 0.996 (0.4833) Loss 3.0776 (2.2148) Prec@1 37.500 (50.554) Prec@5 75.000 (79.832) Test: [160/181] Time 1.140 (0.4806) Loss 2.5114 (2.2007) Prec@1 46.875 (50.932) Prec@5 82.812 (79.950) Testing Results: Prec@1 51.163 Prec@5 80.000 Loss 2.19210 Time 0.4746 No BN layer Freezing. Epoch: [44][0/1345], lr: 0.00010 Time 4.390 (4.390) Data 3.629 (3.629) Loss 0.8985 (0.8985) Prec@1 76.562 (76.562) Prec@5 92.188 (92.188) Epoch: [44][20/1345], lr: 0.00010 Time 0.773 (0.921) Data 0.000 (0.173) Loss 0.6358 (0.6656) Prec@1 87.500 (81.771) Prec@5 93.750 (95.387) Epoch: [44][40/1345], lr: 0.00010 Time 0.715 (0.829) Data 0.000 (0.089) Loss 0.8876 (0.6421) Prec@1 73.438 (82.279) Prec@5 92.188 (95.389) Epoch: [44][60/1345], lr: 0.00010 Time 0.713 (0.799) Data 0.000 (0.060) Loss 0.5733 (0.6351) Prec@1 85.938 (82.505) Prec@5 95.312 (95.492) Epoch: [44][80/1345], lr: 0.00010 Time 0.712 (0.789) Data 0.000 (0.045) Loss 0.6308 (0.6277) Prec@1 84.375 (82.851) Prec@5 96.875 (95.698) Epoch: [44][100/1345], lr: 0.00010 Time 0.715 (0.781) Data 0.000 (0.036) Loss 0.6158 (0.6255) Prec@1 82.812 (83.045) Prec@5 93.750 (95.668) Epoch: [44][120/1345], lr: 0.00010 Time 0.722 (0.777) Data 0.000 (0.030) Loss 0.4459 (0.6247) Prec@1 90.625 (82.967) Prec@5 96.875 (95.687) Epoch: [44][140/1345], lr: 0.00010 Time 0.832 (0.776) Data 0.001 (0.026) Loss 0.7787 (0.6285) Prec@1 79.688 (82.879) Prec@5 93.750 (95.656) Epoch: [44][160/1345], lr: 0.00010 Time 0.832 (0.774) Data 0.000 (0.023) Loss 0.5304 (0.6261) Prec@1 84.375 (82.725) Prec@5 96.875 (95.730) Epoch: [44][180/1345], lr: 0.00010 Time 0.756 (0.772) Data 0.000 (0.021) Loss 0.5404 (0.6215) Prec@1 82.812 (82.795) Prec@5 96.875 (95.848) Epoch: [44][200/1345], lr: 0.00010 Time 0.713 (0.770) Data 0.000 (0.019) Loss 0.4674 (0.6202) Prec@1 89.062 (82.851) Prec@5 98.438 (95.911) Epoch: [44][220/1345], lr: 0.00010 Time 0.724 (0.769) Data 0.000 (0.017) Loss 0.5231 (0.6200) Prec@1 85.938 (82.834) Prec@5 96.875 (95.935) Epoch: [44][240/1345], lr: 0.00010 Time 0.712 (0.766) Data 0.000 (0.016) Loss 0.6345 (0.6173) Prec@1 84.375 (82.936) Prec@5 96.875 (95.993) Epoch: [44][260/1345], lr: 0.00010 Time 0.728 (0.765) Data 0.001 (0.014) Loss 0.6961 (0.6200) Prec@1 82.812 (82.992) Prec@5 92.188 (95.923) Epoch: [44][280/1345], lr: 0.00010 Time 0.713 (0.764) Data 0.000 (0.013) Loss 0.5449 (0.6209) Prec@1 84.375 (82.929) Prec@5 93.750 (95.919) Epoch: [44][300/1345], lr: 0.00010 Time 0.817 (0.764) Data 0.000 (0.013) Loss 0.5938 (0.6196) Prec@1 89.062 (83.010) Prec@5 93.750 (95.930) Epoch: [44][320/1345], lr: 0.00010 Time 0.860 (0.762) Data 0.000 (0.012) Loss 0.5619 (0.6167) Prec@1 78.125 (83.017) Prec@5 96.875 (95.989) Epoch: [44][340/1345], lr: 0.00010 Time 0.727 (0.762) Data 0.000 (0.011) Loss 0.8557 (0.6179) Prec@1 75.000 (83.005) Prec@5 95.312 (95.981) Epoch: [44][360/1345], lr: 0.00010 Time 0.734 (0.761) Data 0.000 (0.011) Loss 0.7677 (0.6191) Prec@1 78.125 (82.981) Prec@5 96.875 (95.953) Epoch: [44][380/1345], lr: 0.00010 Time 0.760 (0.761) Data 0.001 (0.010) Loss 0.4613 (0.6203) Prec@1 87.500 (82.919) Prec@5 98.438 (95.944) Epoch: [44][400/1345], lr: 0.00010 Time 0.711 (0.759) Data 0.000 (0.010) Loss 0.7339 (0.6197) Prec@1 79.688 (82.922) Prec@5 93.750 (95.963) Epoch: [44][420/1345], lr: 0.00010 Time 0.741 (0.758) Data 0.000 (0.009) Loss 0.9678 (0.6195) Prec@1 73.438 (82.909) Prec@5 90.625 (95.966) Epoch: [44][440/1345], lr: 0.00010 Time 0.766 (0.757) Data 0.001 (0.009) Loss 0.4966 (0.6188) Prec@1 87.500 (82.887) Prec@5 98.438 (95.979) Epoch: [44][460/1345], lr: 0.00010 Time 0.781 (0.757) Data 0.000 (0.008) Loss 0.8825 (0.6179) Prec@1 81.250 (82.921) Prec@5 93.750 (96.028) Epoch: [44][480/1345], lr: 0.00010 Time 0.719 (0.756) Data 0.000 (0.008) Loss 0.6034 (0.6169) Prec@1 82.812 (82.936) Prec@5 96.875 (96.043) Epoch: [44][500/1345], lr: 0.00010 Time 0.714 (0.755) Data 0.000 (0.008) Loss 0.5573 (0.6173) Prec@1 81.250 (82.931) Prec@5 96.875 (96.049) Epoch: [44][520/1345], lr: 0.00010 Time 0.735 (0.755) Data 0.000 (0.007) Loss 0.5396 (0.6181) Prec@1 85.938 (82.920) Prec@5 98.438 (96.023) Epoch: [44][540/1345], lr: 0.00010 Time 0.843 (0.755) Data 0.000 (0.007) Loss 0.5396 (0.6165) Prec@1 81.250 (82.951) Prec@5 96.875 (96.040) Epoch: [44][560/1345], lr: 0.00010 Time 0.728 (0.754) Data 0.000 (0.007) Loss 0.4063 (0.6170) Prec@1 89.062 (82.927) Prec@5 98.438 (96.034) Epoch: [44][580/1345], lr: 0.00010 Time 0.716 (0.754) Data 0.000 (0.007) Loss 0.5564 (0.6167) Prec@1 79.688 (82.920) Prec@5 96.875 (96.041) Epoch: [44][600/1345], lr: 0.00010 Time 0.716 (0.754) Data 0.000 (0.007) Loss 0.5529 (0.6157) Prec@1 87.500 (82.950) Prec@5 95.312 (96.051) Epoch: [44][620/1345], lr: 0.00010 Time 0.714 (0.754) Data 0.000 (0.006) Loss 0.6983 (0.6177) Prec@1 78.125 (82.885) Prec@5 96.875 (96.032) Epoch: [44][640/1345], lr: 0.00010 Time 0.714 (0.754) Data 0.000 (0.006) Loss 0.6045 (0.6182) Prec@1 81.250 (82.861) Prec@5 95.312 (96.022) Epoch: [44][660/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.006) Loss 0.4581 (0.6181) Prec@1 87.500 (82.848) Prec@5 98.438 (96.026) Epoch: [44][680/1345], lr: 0.00010 Time 0.795 (0.753) Data 0.000 (0.006) Loss 0.7734 (0.6198) Prec@1 84.375 (82.803) Prec@5 92.188 (96.010) Epoch: [44][700/1345], lr: 0.00010 Time 0.744 (0.753) Data 0.000 (0.006) Loss 0.4686 (0.6204) Prec@1 85.938 (82.790) Prec@5 98.438 (96.008) Epoch: [44][720/1345], lr: 0.00010 Time 0.711 (0.753) Data 0.000 (0.006) Loss 0.5827 (0.6196) Prec@1 82.812 (82.793) Prec@5 95.312 (96.023) Epoch: [44][740/1345], lr: 0.00010 Time 0.716 (0.752) Data 0.000 (0.005) Loss 0.6720 (0.6198) Prec@1 79.688 (82.794) Prec@5 96.875 (96.013) Epoch: [44][760/1345], lr: 0.00010 Time 0.768 (0.752) Data 0.000 (0.005) Loss 0.5739 (0.6198) Prec@1 81.250 (82.776) Prec@5 98.438 (96.009) Epoch: [44][780/1345], lr: 0.00010 Time 0.840 (0.752) Data 0.000 (0.005) Loss 0.4871 (0.6184) Prec@1 81.250 (82.827) Prec@5 98.438 (96.035) Epoch: [44][800/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.005) Loss 0.7452 (0.6198) Prec@1 82.812 (82.793) Prec@5 92.188 (96.009) Epoch: [44][820/1345], lr: 0.00010 Time 0.734 (0.752) Data 0.000 (0.005) Loss 0.6521 (0.6196) Prec@1 84.375 (82.778) Prec@5 98.438 (96.019) Epoch: [44][840/1345], lr: 0.00010 Time 0.847 (0.751) Data 0.000 (0.005) Loss 0.5739 (0.6199) Prec@1 81.250 (82.766) Prec@5 98.438 (96.017) Epoch: [44][860/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.005) Loss 0.4447 (0.6210) Prec@1 84.375 (82.734) Prec@5 100.000 (96.000) Epoch: [44][880/1345], lr: 0.00010 Time 0.757 (0.751) Data 0.000 (0.005) Loss 0.5241 (0.6202) Prec@1 84.375 (82.747) Prec@5 98.438 (96.010) Epoch: [44][900/1345], lr: 0.00010 Time 0.760 (0.751) Data 0.000 (0.004) Loss 0.7022 (0.6200) Prec@1 76.562 (82.752) Prec@5 95.312 (96.011) Epoch: [44][920/1345], lr: 0.00010 Time 0.734 (0.751) Data 0.000 (0.004) Loss 0.6363 (0.6200) Prec@1 84.375 (82.765) Prec@5 95.312 (96.003) Epoch: [44][940/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.7354 (0.6204) Prec@1 79.688 (82.771) Prec@5 93.750 (95.992) Epoch: [44][960/1345], lr: 0.00010 Time 0.747 (0.751) Data 0.000 (0.004) Loss 0.5272 (0.6208) Prec@1 84.375 (82.760) Prec@5 96.875 (95.994) Epoch: [44][980/1345], lr: 0.00010 Time 0.713 (0.751) Data 0.000 (0.004) Loss 0.5623 (0.6218) Prec@1 82.812 (82.750) Prec@5 96.875 (95.978) Epoch: [44][1000/1345], lr: 0.00010 Time 0.729 (0.751) Data 0.000 (0.004) Loss 0.6398 (0.6217) Prec@1 79.688 (82.741) Prec@5 96.875 (95.988) Epoch: [44][1020/1345], lr: 0.00010 Time 0.712 (0.751) Data 0.000 (0.004) Loss 0.6322 (0.6216) Prec@1 84.375 (82.724) Prec@5 95.312 (95.997) Epoch: [44][1040/1345], lr: 0.00010 Time 0.715 (0.751) Data 0.000 (0.004) Loss 0.5434 (0.6209) Prec@1 85.938 (82.718) Prec@5 98.438 (96.013) Epoch: [44][1060/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.001 (0.004) Loss 0.6156 (0.6214) Prec@1 89.062 (82.724) Prec@5 98.438 (96.011) Epoch: [44][1080/1345], lr: 0.00010 Time 0.711 (0.750) Data 0.000 (0.004) Loss 0.6318 (0.6214) Prec@1 79.688 (82.719) Prec@5 98.438 (96.014) Epoch: [44][1100/1345], lr: 0.00010 Time 0.772 (0.750) Data 0.001 (0.004) Loss 0.5064 (0.6206) Prec@1 89.062 (82.734) Prec@5 96.875 (96.032) Epoch: [44][1120/1345], lr: 0.00010 Time 0.733 (0.750) Data 0.000 (0.004) Loss 0.5860 (0.6212) Prec@1 81.250 (82.702) Prec@5 98.438 (96.023) Epoch: [44][1140/1345], lr: 0.00010 Time 0.735 (0.750) Data 0.001 (0.004) Loss 0.7379 (0.6216) Prec@1 75.000 (82.699) Prec@5 96.875 (96.029) Epoch: [44][1160/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.004) Loss 0.5263 (0.6215) Prec@1 87.500 (82.706) Prec@5 96.875 (96.030) Epoch: [44][1180/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.004) Loss 0.7059 (0.6222) Prec@1 79.688 (82.709) Prec@5 93.750 (96.012) Epoch: [44][1200/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.003) Loss 0.8665 (0.6211) Prec@1 84.375 (82.738) Prec@5 95.312 (96.023) Epoch: [44][1220/1345], lr: 0.00010 Time 0.725 (0.750) Data 0.000 (0.003) Loss 0.6617 (0.6209) Prec@1 79.688 (82.747) Prec@5 96.875 (96.032) Epoch: [44][1240/1345], lr: 0.00010 Time 0.728 (0.750) Data 0.001 (0.003) Loss 0.6510 (0.6209) Prec@1 85.938 (82.750) Prec@5 98.438 (96.038) Epoch: [44][1260/1345], lr: 0.00010 Time 0.731 (0.750) Data 0.001 (0.003) Loss 0.5046 (0.6214) Prec@1 87.500 (82.727) Prec@5 95.312 (96.024) Epoch: [44][1280/1345], lr: 0.00010 Time 0.719 (0.750) Data 0.001 (0.003) Loss 0.4664 (0.6209) Prec@1 87.500 (82.750) Prec@5 100.000 (96.046) Epoch: [44][1300/1345], lr: 0.00010 Time 0.725 (0.750) Data 0.001 (0.003) Loss 0.6293 (0.6205) Prec@1 82.812 (82.751) Prec@5 96.875 (96.055) Epoch: [44][1320/1345], lr: 0.00010 Time 0.722 (0.750) Data 0.000 (0.003) Loss 0.5509 (0.6208) Prec@1 82.812 (82.757) Prec@5 96.875 (96.055) Epoch: [44][1340/1345], lr: 0.00010 Time 0.708 (0.749) Data 0.000 (0.003) Loss 0.5120 (0.6202) Prec@1 87.500 (82.775) Prec@5 96.875 (96.061) No BN layer Freezing. Test: [0/181] Time 3.357 (3.3571) Loss 2.1994 (2.1994) Prec@1 53.125 (53.125) Prec@5 85.938 (85.938) Test: [20/181] Time 1.130 (0.6146) Loss 2.2989 (2.0689) Prec@1 50.000 (53.274) Prec@5 79.688 (81.771) Test: [40/181] Time 0.974 (0.5359) Loss 2.5635 (2.1442) Prec@1 53.125 (52.553) Prec@5 73.438 (80.907) Test: [60/181] Time 1.166 (0.5124) Loss 2.6966 (2.1980) Prec@1 53.125 (51.434) Prec@5 70.312 (80.328) Test: [80/181] Time 1.018 (0.4973) Loss 1.7187 (2.1990) Prec@1 64.062 (51.273) Prec@5 85.938 (80.073) Test: [100/181] Time 0.897 (0.4904) Loss 2.9227 (2.1912) Prec@1 45.312 (51.222) Prec@5 70.312 (80.183) Test: [120/181] Time 0.941 (0.4848) Loss 2.9359 (2.2034) Prec@1 43.750 (51.240) Prec@5 71.875 (80.165) Test: [140/181] Time 1.053 (0.4819) Loss 3.0690 (2.2003) Prec@1 34.375 (50.909) Prec@5 73.438 (80.308) Test: [160/181] Time 0.931 (0.4792) Loss 2.5343 (2.1890) Prec@1 45.312 (51.068) Prec@5 82.812 (80.406) Testing Results: Prec@1 51.276 Prec@5 80.477 Loss 2.17921 Time 0.4739 No BN layer Freezing. Epoch: [45][0/1345], lr: 0.00010 Time 4.481 (4.481) Data 3.744 (3.744) Loss 0.8198 (0.8198) Prec@1 79.688 (79.688) Prec@5 92.188 (92.188) Epoch: [45][20/1345], lr: 0.00010 Time 0.727 (0.924) Data 0.000 (0.179) Loss 0.7890 (0.6487) Prec@1 71.875 (81.696) Prec@5 93.750 (95.685) Epoch: [45][40/1345], lr: 0.00010 Time 0.721 (0.833) Data 0.000 (0.092) Loss 0.8044 (0.6306) Prec@1 79.688 (82.355) Prec@5 93.750 (96.113) Epoch: [45][60/1345], lr: 0.00010 Time 0.712 (0.804) Data 0.000 (0.062) Loss 0.9011 (0.6209) Prec@1 71.875 (82.172) Prec@5 93.750 (96.388) Epoch: [45][80/1345], lr: 0.00010 Time 0.739 (0.790) Data 0.001 (0.047) Loss 0.5512 (0.6284) Prec@1 90.625 (82.330) Prec@5 96.875 (96.277) Epoch: [45][100/1345], lr: 0.00010 Time 0.889 (0.783) Data 0.000 (0.038) Loss 0.6448 (0.6143) Prec@1 81.250 (82.673) Prec@5 96.875 (96.318) Epoch: [45][120/1345], lr: 0.00010 Time 0.726 (0.778) Data 0.000 (0.031) Loss 0.4612 (0.6084) Prec@1 87.500 (82.851) Prec@5 98.438 (96.346) Epoch: [45][140/1345], lr: 0.00010 Time 0.772 (0.772) Data 0.001 (0.027) Loss 0.5512 (0.6054) Prec@1 89.062 (83.012) Prec@5 95.312 (96.288) Epoch: [45][160/1345], lr: 0.00010 Time 0.718 (0.770) Data 0.000 (0.024) Loss 0.5043 (0.6034) Prec@1 89.062 (83.094) Prec@5 98.438 (96.322) Epoch: [45][180/1345], lr: 0.00010 Time 0.716 (0.767) Data 0.001 (0.021) Loss 0.7215 (0.6097) Prec@1 82.812 (82.951) Prec@5 93.750 (96.202) Epoch: [45][200/1345], lr: 0.00010 Time 0.714 (0.764) Data 0.000 (0.019) Loss 0.3718 (0.6048) Prec@1 85.938 (83.100) Prec@5 100.000 (96.214) Epoch: [45][220/1345], lr: 0.00010 Time 0.713 (0.762) Data 0.000 (0.017) Loss 0.5717 (0.6057) Prec@1 82.812 (83.025) Prec@5 95.312 (96.253) Epoch: [45][240/1345], lr: 0.00010 Time 0.740 (0.761) Data 0.000 (0.016) Loss 0.5203 (0.6036) Prec@1 89.062 (83.117) Prec@5 96.875 (96.259) Epoch: [45][260/1345], lr: 0.00010 Time 0.759 (0.760) Data 0.001 (0.015) Loss 0.7196 (0.6023) Prec@1 82.812 (83.190) Prec@5 95.312 (96.234) Epoch: [45][280/1345], lr: 0.00010 Time 0.710 (0.759) Data 0.000 (0.014) Loss 0.4088 (0.6007) Prec@1 92.188 (83.263) Prec@5 98.438 (96.297) Epoch: [45][300/1345], lr: 0.00010 Time 0.821 (0.758) Data 0.000 (0.013) Loss 0.6236 (0.6027) Prec@1 78.125 (83.238) Prec@5 96.875 (96.288) Epoch: [45][320/1345], lr: 0.00010 Time 0.815 (0.757) Data 0.000 (0.012) Loss 0.5176 (0.6056) Prec@1 85.938 (83.217) Prec@5 98.438 (96.203) Epoch: [45][340/1345], lr: 0.00010 Time 0.772 (0.756) Data 0.000 (0.011) Loss 0.5469 (0.6060) Prec@1 85.938 (83.115) Prec@5 100.000 (96.197) Epoch: [45][360/1345], lr: 0.00010 Time 0.716 (0.756) Data 0.000 (0.011) Loss 0.7064 (0.6028) Prec@1 84.375 (83.215) Prec@5 96.875 (96.208) Epoch: [45][380/1345], lr: 0.00010 Time 0.715 (0.755) Data 0.000 (0.010) Loss 0.6753 (0.6027) Prec@1 82.812 (83.235) Prec@5 98.438 (96.235) Epoch: [45][400/1345], lr: 0.00010 Time 0.769 (0.755) Data 0.000 (0.010) Loss 0.7658 (0.6049) Prec@1 82.812 (83.183) Prec@5 95.312 (96.240) Epoch: [45][420/1345], lr: 0.00010 Time 0.730 (0.754) Data 0.000 (0.009) Loss 0.6619 (0.6058) Prec@1 85.938 (83.198) Prec@5 93.750 (96.207) Epoch: [45][440/1345], lr: 0.00010 Time 0.713 (0.754) Data 0.000 (0.009) Loss 0.4248 (0.6069) Prec@1 85.938 (83.160) Prec@5 96.875 (96.166) Epoch: [45][460/1345], lr: 0.00010 Time 0.841 (0.754) Data 0.000 (0.009) Loss 0.5102 (0.6079) Prec@1 85.938 (83.155) Prec@5 96.875 (96.150) Epoch: [45][480/1345], lr: 0.00010 Time 0.851 (0.753) Data 0.000 (0.008) Loss 0.5217 (0.6065) Prec@1 85.938 (83.176) Prec@5 96.875 (96.170) Epoch: [45][500/1345], lr: 0.00010 Time 0.727 (0.753) Data 0.000 (0.008) Loss 0.7392 (0.6060) Prec@1 79.688 (83.165) Prec@5 95.312 (96.183) Epoch: [45][520/1345], lr: 0.00010 Time 0.710 (0.753) Data 0.000 (0.008) Loss 0.5651 (0.6043) Prec@1 84.375 (83.220) Prec@5 96.875 (96.203) Epoch: [45][540/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.007) Loss 0.4254 (0.6058) Prec@1 89.062 (83.194) Prec@5 96.875 (96.182) Epoch: [45][560/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.007) Loss 0.4183 (0.6056) Prec@1 84.375 (83.200) Prec@5 100.000 (96.190) Epoch: [45][580/1345], lr: 0.00010 Time 0.763 (0.752) Data 0.000 (0.007) Loss 0.4902 (0.6055) Prec@1 84.375 (83.232) Prec@5 98.438 (96.178) Epoch: [45][600/1345], lr: 0.00010 Time 0.711 (0.751) Data 0.000 (0.007) Loss 0.7367 (0.6067) Prec@1 76.562 (83.205) Prec@5 95.312 (96.178) Epoch: [45][620/1345], lr: 0.00010 Time 0.836 (0.751) Data 0.000 (0.007) Loss 0.5940 (0.6071) Prec@1 84.375 (83.182) Prec@5 98.438 (96.193) Epoch: [45][640/1345], lr: 0.00010 Time 0.852 (0.751) Data 0.000 (0.006) Loss 0.5308 (0.6072) Prec@1 84.375 (83.185) Prec@5 93.750 (96.188) Epoch: [45][660/1345], lr: 0.00010 Time 0.737 (0.751) Data 0.000 (0.006) Loss 0.8178 (0.6076) Prec@1 76.562 (83.172) Prec@5 92.188 (96.175) Epoch: [45][680/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.006) Loss 0.9233 (0.6086) Prec@1 76.562 (83.177) Prec@5 93.750 (96.159) Epoch: [45][700/1345], lr: 0.00010 Time 0.739 (0.750) Data 0.000 (0.006) Loss 0.6086 (0.6084) Prec@1 82.812 (83.183) Prec@5 96.875 (96.184) Epoch: [45][720/1345], lr: 0.00010 Time 0.741 (0.750) Data 0.000 (0.006) Loss 0.6469 (0.6093) Prec@1 82.812 (83.166) Prec@5 95.312 (96.177) Epoch: [45][740/1345], lr: 0.00010 Time 0.746 (0.750) Data 0.000 (0.006) Loss 1.0205 (0.6110) Prec@1 71.875 (83.146) Prec@5 89.062 (96.150) Epoch: [45][760/1345], lr: 0.00010 Time 0.734 (0.750) Data 0.000 (0.005) Loss 0.5611 (0.6128) Prec@1 85.938 (83.102) Prec@5 93.750 (96.111) Epoch: [45][780/1345], lr: 0.00010 Time 0.724 (0.750) Data 0.000 (0.005) Loss 0.4720 (0.6130) Prec@1 90.625 (83.107) Prec@5 98.438 (96.111) Epoch: [45][800/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.005) Loss 0.4349 (0.6132) Prec@1 90.625 (83.115) Prec@5 96.875 (96.104) Epoch: [45][820/1345], lr: 0.00010 Time 0.863 (0.749) Data 0.000 (0.005) Loss 0.7185 (0.6133) Prec@1 78.125 (83.125) Prec@5 95.312 (96.097) Epoch: [45][840/1345], lr: 0.00010 Time 0.861 (0.749) Data 0.000 (0.005) Loss 0.6611 (0.6133) Prec@1 81.250 (83.130) Prec@5 96.875 (96.098) Epoch: [45][860/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.000 (0.005) Loss 0.5182 (0.6145) Prec@1 87.500 (83.123) Prec@5 93.750 (96.086) Epoch: [45][880/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.005) Loss 0.4368 (0.6144) Prec@1 89.062 (83.123) Prec@5 100.000 (96.095) Epoch: [45][900/1345], lr: 0.00010 Time 0.724 (0.749) Data 0.000 (0.005) Loss 0.5209 (0.6142) Prec@1 82.812 (83.114) Prec@5 98.438 (96.081) Epoch: [45][920/1345], lr: 0.00010 Time 0.732 (0.749) Data 0.000 (0.005) Loss 0.7575 (0.6145) Prec@1 81.250 (83.094) Prec@5 93.750 (96.081) Epoch: [45][940/1345], lr: 0.00010 Time 0.786 (0.749) Data 0.001 (0.004) Loss 0.5824 (0.6152) Prec@1 85.938 (83.072) Prec@5 95.312 (96.076) Epoch: [45][960/1345], lr: 0.00010 Time 0.859 (0.749) Data 0.000 (0.004) Loss 0.5925 (0.6161) Prec@1 82.812 (83.053) Prec@5 98.438 (96.070) Epoch: [45][980/1345], lr: 0.00010 Time 0.880 (0.749) Data 0.001 (0.004) Loss 0.4113 (0.6157) Prec@1 89.062 (83.040) Prec@5 98.438 (96.075) Epoch: [45][1000/1345], lr: 0.00010 Time 0.716 (0.749) Data 0.000 (0.004) Loss 0.7718 (0.6153) Prec@1 76.562 (83.039) Prec@5 92.188 (96.070) Epoch: [45][1020/1345], lr: 0.00010 Time 0.721 (0.749) Data 0.000 (0.004) Loss 0.6395 (0.6158) Prec@1 84.375 (83.015) Prec@5 98.438 (96.073) Epoch: [45][1040/1345], lr: 0.00010 Time 0.730 (0.749) Data 0.000 (0.004) Loss 0.7245 (0.6157) Prec@1 78.125 (82.984) Prec@5 95.312 (96.082) Epoch: [45][1060/1345], lr: 0.00010 Time 0.725 (0.749) Data 0.001 (0.004) Loss 0.5870 (0.6159) Prec@1 85.938 (82.973) Prec@5 93.750 (96.074) Epoch: [45][1080/1345], lr: 0.00010 Time 0.715 (0.749) Data 0.000 (0.004) Loss 0.6772 (0.6165) Prec@1 78.125 (82.957) Prec@5 98.438 (96.068) Epoch: [45][1100/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.004) Loss 0.4576 (0.6168) Prec@1 89.062 (82.954) Prec@5 98.438 (96.072) Epoch: [45][1120/1345], lr: 0.00010 Time 0.807 (0.749) Data 0.000 (0.004) Loss 0.4864 (0.6173) Prec@1 87.500 (82.944) Prec@5 95.312 (96.068) Epoch: [45][1140/1345], lr: 0.00010 Time 0.816 (0.749) Data 0.000 (0.004) Loss 0.6402 (0.6170) Prec@1 82.812 (82.974) Prec@5 95.312 (96.066) Epoch: [45][1160/1345], lr: 0.00010 Time 0.775 (0.749) Data 0.000 (0.004) Loss 0.5862 (0.6165) Prec@1 85.938 (82.973) Prec@5 95.312 (96.080) Epoch: [45][1180/1345], lr: 0.00010 Time 0.736 (0.749) Data 0.000 (0.004) Loss 0.5525 (0.6160) Prec@1 87.500 (83.002) Prec@5 95.312 (96.083) Epoch: [45][1200/1345], lr: 0.00010 Time 0.743 (0.749) Data 0.000 (0.004) Loss 0.5094 (0.6155) Prec@1 84.375 (83.030) Prec@5 96.875 (96.081) Epoch: [45][1220/1345], lr: 0.00010 Time 0.750 (0.749) Data 0.000 (0.004) Loss 0.4080 (0.6151) Prec@1 89.062 (83.033) Prec@5 98.438 (96.088) Epoch: [45][1240/1345], lr: 0.00010 Time 0.737 (0.749) Data 0.000 (0.003) Loss 0.5517 (0.6154) Prec@1 85.938 (83.015) Prec@5 93.750 (96.083) Epoch: [45][1260/1345], lr: 0.00010 Time 0.714 (0.749) Data 0.000 (0.003) Loss 0.9831 (0.6161) Prec@1 71.875 (82.995) Prec@5 92.188 (96.070) Epoch: [45][1280/1345], lr: 0.00010 Time 0.803 (0.748) Data 0.000 (0.003) Loss 0.8100 (0.6157) Prec@1 71.875 (83.004) Prec@5 98.438 (96.079) Epoch: [45][1300/1345], lr: 0.00010 Time 0.836 (0.748) Data 0.000 (0.003) Loss 0.5407 (0.6156) Prec@1 84.375 (83.023) Prec@5 96.875 (96.088) Epoch: [45][1320/1345], lr: 0.00010 Time 0.712 (0.748) Data 0.000 (0.003) Loss 0.7105 (0.6158) Prec@1 78.125 (83.024) Prec@5 96.875 (96.088) Epoch: [45][1340/1345], lr: 0.00010 Time 0.722 (0.749) Data 0.000 (0.003) Loss 0.6788 (0.6166) Prec@1 85.938 (83.008) Prec@5 95.312 (96.076) No BN layer Freezing. Test: [0/181] Time 3.364 (3.3637) Loss 2.2792 (2.2792) Prec@1 51.562 (51.562) Prec@5 82.812 (82.812) Test: [20/181] Time 0.696 (0.5816) Loss 2.2979 (2.1008) Prec@1 51.562 (52.976) Prec@5 75.000 (80.952) Test: [40/181] Time 1.328 (0.5354) Loss 2.5214 (2.1801) Prec@1 53.125 (52.058) Prec@5 73.438 (80.069) Test: [60/181] Time 0.984 (0.5101) Loss 2.7937 (2.2293) Prec@1 53.125 (50.948) Prec@5 70.312 (79.508) Test: [80/181] Time 0.844 (0.4944) Loss 1.7161 (2.2340) Prec@1 60.938 (50.733) Prec@5 89.062 (79.167) Test: [100/181] Time 0.886 (0.4866) Loss 2.9412 (2.2213) Prec@1 50.000 (50.820) Prec@5 71.875 (79.455) Test: [120/181] Time 0.934 (0.4820) Loss 2.9352 (2.2291) Prec@1 43.750 (50.762) Prec@5 71.875 (79.507) Test: [140/181] Time 0.903 (0.4790) Loss 3.1147 (2.2245) Prec@1 32.812 (50.488) Prec@5 71.875 (79.699) Test: [160/181] Time 0.924 (0.4773) Loss 2.4921 (2.2086) Prec@1 46.875 (50.883) Prec@5 82.812 (79.852) Testing Results: Prec@1 51.120 Prec@5 79.957 Loss 2.20028 Time 0.4733 No BN layer Freezing. Epoch: [46][0/1345], lr: 0.00010 Time 4.474 (4.474) Data 3.731 (3.731) Loss 0.7758 (0.7758) Prec@1 78.125 (78.125) Prec@5 90.625 (90.625) Epoch: [46][20/1345], lr: 0.00010 Time 0.716 (0.922) Data 0.000 (0.178) Loss 0.5572 (0.6264) Prec@1 89.062 (82.515) Prec@5 95.312 (95.610) Epoch: [46][40/1345], lr: 0.00010 Time 0.716 (0.842) Data 0.000 (0.091) Loss 0.5064 (0.6202) Prec@1 92.188 (83.232) Prec@5 93.750 (95.808) Epoch: [46][60/1345], lr: 0.00010 Time 0.754 (0.810) Data 0.000 (0.062) Loss 0.7222 (0.5974) Prec@1 84.375 (84.093) Prec@5 95.312 (96.055) Epoch: [46][80/1345], lr: 0.00010 Time 0.756 (0.796) Data 0.001 (0.047) Loss 0.5765 (0.5966) Prec@1 82.812 (84.047) Prec@5 98.438 (96.084) Epoch: [46][100/1345], lr: 0.00010 Time 0.755 (0.787) Data 0.000 (0.037) Loss 0.6560 (0.5951) Prec@1 84.375 (84.050) Prec@5 96.875 (96.163) Epoch: [46][120/1345], lr: 0.00010 Time 0.715 (0.781) Data 0.000 (0.031) Loss 0.6848 (0.5902) Prec@1 82.812 (84.091) Prec@5 95.312 (96.268) Epoch: [46][140/1345], lr: 0.00010 Time 0.730 (0.776) Data 0.001 (0.027) Loss 0.6626 (0.5997) Prec@1 79.688 (83.766) Prec@5 96.875 (96.254) Epoch: [46][160/1345], lr: 0.00010 Time 0.715 (0.772) Data 0.000 (0.024) Loss 0.6548 (0.6027) Prec@1 78.125 (83.560) Prec@5 96.875 (96.244) Epoch: [46][180/1345], lr: 0.00010 Time 0.848 (0.770) Data 0.000 (0.021) Loss 0.5124 (0.6015) Prec@1 87.500 (83.607) Prec@5 98.438 (96.279) Epoch: [46][200/1345], lr: 0.00010 Time 0.809 (0.767) Data 0.000 (0.019) Loss 0.3568 (0.6026) Prec@1 93.750 (83.543) Prec@5 98.438 (96.284) Epoch: [46][220/1345], lr: 0.00010 Time 0.714 (0.766) Data 0.000 (0.017) Loss 0.5020 (0.6032) Prec@1 87.500 (83.498) Prec@5 98.438 (96.316) Epoch: [46][240/1345], lr: 0.00010 Time 0.713 (0.765) Data 0.000 (0.016) Loss 0.6649 (0.6061) Prec@1 84.375 (83.428) Prec@5 96.875 (96.304) Epoch: [46][260/1345], lr: 0.00010 Time 0.714 (0.764) Data 0.000 (0.015) Loss 0.8051 (0.6076) Prec@1 78.125 (83.447) Prec@5 92.188 (96.252) Epoch: [46][280/1345], lr: 0.00010 Time 0.767 (0.763) Data 0.000 (0.014) Loss 0.3588 (0.6056) Prec@1 92.188 (83.446) Prec@5 100.000 (96.308) Epoch: [46][300/1345], lr: 0.00010 Time 0.715 (0.762) Data 0.001 (0.013) Loss 0.6948 (0.6044) Prec@1 81.250 (83.404) Prec@5 96.875 (96.346) Epoch: [46][320/1345], lr: 0.00010 Time 0.770 (0.761) Data 0.000 (0.012) Loss 0.5840 (0.6067) Prec@1 84.375 (83.382) Prec@5 93.750 (96.291) Epoch: [46][340/1345], lr: 0.00010 Time 0.843 (0.760) Data 0.000 (0.011) Loss 1.0540 (0.6064) Prec@1 75.000 (83.353) Prec@5 90.625 (96.266) Epoch: [46][360/1345], lr: 0.00010 Time 0.829 (0.760) Data 0.000 (0.011) Loss 0.5318 (0.6063) Prec@1 87.500 (83.392) Prec@5 100.000 (96.234) Epoch: [46][380/1345], lr: 0.00010 Time 0.775 (0.759) Data 0.000 (0.010) Loss 0.6158 (0.6075) Prec@1 79.688 (83.374) Prec@5 98.438 (96.194) Epoch: [46][400/1345], lr: 0.00010 Time 0.714 (0.759) Data 0.000 (0.010) Loss 0.4550 (0.6036) Prec@1 92.188 (83.549) Prec@5 98.438 (96.220) Epoch: [46][420/1345], lr: 0.00010 Time 0.759 (0.758) Data 0.000 (0.009) Loss 0.7216 (0.6050) Prec@1 81.250 (83.529) Prec@5 92.188 (96.177) Epoch: [46][440/1345], lr: 0.00010 Time 0.713 (0.758) Data 0.000 (0.009) Loss 0.4708 (0.6029) Prec@1 81.250 (83.585) Prec@5 100.000 (96.191) Epoch: [46][460/1345], lr: 0.00010 Time 0.713 (0.757) Data 0.001 (0.009) Loss 0.7122 (0.6027) Prec@1 84.375 (83.582) Prec@5 93.750 (96.224) Epoch: [46][480/1345], lr: 0.00010 Time 0.768 (0.757) Data 0.000 (0.008) Loss 0.6843 (0.6034) Prec@1 75.000 (83.501) Prec@5 93.750 (96.232) Epoch: [46][500/1345], lr: 0.00010 Time 0.715 (0.756) Data 0.000 (0.008) Loss 0.6017 (0.6025) Prec@1 85.938 (83.496) Prec@5 93.750 (96.236) Epoch: [46][520/1345], lr: 0.00010 Time 0.713 (0.755) Data 0.000 (0.008) Loss 1.0095 (0.6029) Prec@1 75.000 (83.481) Prec@5 93.750 (96.236) Epoch: [46][540/1345], lr: 0.00010 Time 0.844 (0.755) Data 0.000 (0.007) Loss 0.6874 (0.6036) Prec@1 81.250 (83.454) Prec@5 96.875 (96.245) Epoch: [46][560/1345], lr: 0.00010 Time 0.832 (0.755) Data 0.000 (0.007) Loss 0.6764 (0.6049) Prec@1 81.250 (83.414) Prec@5 96.875 (96.215) Epoch: [46][580/1345], lr: 0.00010 Time 0.753 (0.754) Data 0.000 (0.007) Loss 0.8156 (0.6055) Prec@1 78.125 (83.383) Prec@5 90.625 (96.189) Epoch: [46][600/1345], lr: 0.00010 Time 0.737 (0.754) Data 0.000 (0.007) Loss 0.4645 (0.6051) Prec@1 85.938 (83.390) Prec@5 96.875 (96.183) Epoch: [46][620/1345], lr: 0.00010 Time 0.744 (0.754) Data 0.000 (0.006) Loss 0.4983 (0.6061) Prec@1 87.500 (83.414) Prec@5 96.875 (96.176) Epoch: [46][640/1345], lr: 0.00010 Time 0.727 (0.754) Data 0.000 (0.006) Loss 0.5444 (0.6074) Prec@1 84.375 (83.378) Prec@5 96.875 (96.136) Epoch: [46][660/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.006) Loss 0.8765 (0.6084) Prec@1 78.125 (83.347) Prec@5 92.188 (96.121) Epoch: [46][680/1345], lr: 0.00010 Time 0.713 (0.753) Data 0.001 (0.006) Loss 0.8062 (0.6102) Prec@1 76.562 (83.283) Prec@5 92.188 (96.113) Epoch: [46][700/1345], lr: 0.00010 Time 0.877 (0.754) Data 0.000 (0.006) Loss 0.6972 (0.6103) Prec@1 82.812 (83.272) Prec@5 93.750 (96.124) Epoch: [46][720/1345], lr: 0.00010 Time 0.880 (0.753) Data 0.001 (0.006) Loss 0.6901 (0.6098) Prec@1 76.562 (83.268) Prec@5 96.875 (96.125) Epoch: [46][740/1345], lr: 0.00010 Time 0.737 (0.753) Data 0.000 (0.006) Loss 0.6702 (0.6081) Prec@1 78.125 (83.316) Prec@5 93.750 (96.143) Epoch: [46][760/1345], lr: 0.00010 Time 0.738 (0.753) Data 0.000 (0.005) Loss 0.7393 (0.6090) Prec@1 78.125 (83.274) Prec@5 95.312 (96.144) Epoch: [46][780/1345], lr: 0.00010 Time 0.711 (0.753) Data 0.000 (0.005) Loss 0.8080 (0.6092) Prec@1 79.688 (83.265) Prec@5 89.062 (96.133) Epoch: [46][800/1345], lr: 0.00010 Time 0.712 (0.753) Data 0.000 (0.005) Loss 0.6108 (0.6098) Prec@1 79.688 (83.253) Prec@5 98.438 (96.118) Epoch: [46][820/1345], lr: 0.00010 Time 0.714 (0.753) Data 0.000 (0.005) Loss 0.5494 (0.6100) Prec@1 82.812 (83.279) Prec@5 100.000 (96.112) Epoch: [46][840/1345], lr: 0.00010 Time 0.728 (0.752) Data 0.001 (0.005) Loss 0.6240 (0.6098) Prec@1 82.812 (83.271) Prec@5 95.312 (96.126) Epoch: [46][860/1345], lr: 0.00010 Time 0.741 (0.752) Data 0.000 (0.005) Loss 0.6066 (0.6099) Prec@1 82.812 (83.266) Prec@5 98.438 (96.122) Epoch: [46][880/1345], lr: 0.00010 Time 0.776 (0.752) Data 0.000 (0.005) Loss 0.6357 (0.6094) Prec@1 78.125 (83.265) Prec@5 96.875 (96.121) Epoch: [46][900/1345], lr: 0.00010 Time 0.863 (0.751) Data 0.000 (0.005) Loss 0.6717 (0.6100) Prec@1 78.125 (83.258) Prec@5 96.875 (96.119) Epoch: [46][920/1345], lr: 0.00010 Time 0.738 (0.751) Data 0.000 (0.005) Loss 0.4525 (0.6095) Prec@1 85.938 (83.281) Prec@5 100.000 (96.140) Epoch: [46][940/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.7013 (0.6096) Prec@1 79.688 (83.276) Prec@5 92.188 (96.138) Epoch: [46][960/1345], lr: 0.00010 Time 0.727 (0.751) Data 0.000 (0.004) Loss 0.4854 (0.6089) Prec@1 89.062 (83.304) Prec@5 96.875 (96.153) Epoch: [46][980/1345], lr: 0.00010 Time 0.737 (0.751) Data 0.000 (0.004) Loss 0.7198 (0.6101) Prec@1 84.375 (83.262) Prec@5 95.312 (96.144) Epoch: [46][1000/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.001 (0.004) Loss 0.4582 (0.6101) Prec@1 87.500 (83.250) Prec@5 100.000 (96.151) Epoch: [46][1020/1345], lr: 0.00010 Time 0.744 (0.751) Data 0.000 (0.004) Loss 0.5956 (0.6106) Prec@1 84.375 (83.249) Prec@5 92.188 (96.131) Epoch: [46][1040/1345], lr: 0.00010 Time 0.723 (0.751) Data 0.000 (0.004) Loss 0.5789 (0.6108) Prec@1 85.938 (83.237) Prec@5 95.312 (96.137) Epoch: [46][1060/1345], lr: 0.00010 Time 0.845 (0.751) Data 0.000 (0.004) Loss 0.7771 (0.6100) Prec@1 79.688 (83.254) Prec@5 96.875 (96.149) Epoch: [46][1080/1345], lr: 0.00010 Time 0.752 (0.750) Data 0.000 (0.004) Loss 0.7520 (0.6102) Prec@1 81.250 (83.269) Prec@5 95.312 (96.148) Epoch: [46][1100/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.004) Loss 0.6520 (0.6107) Prec@1 81.250 (83.267) Prec@5 95.312 (96.140) Epoch: [46][1120/1345], lr: 0.00010 Time 0.734 (0.750) Data 0.000 (0.004) Loss 0.5310 (0.6116) Prec@1 87.500 (83.242) Prec@5 96.875 (96.142) Epoch: [46][1140/1345], lr: 0.00010 Time 0.746 (0.750) Data 0.000 (0.004) Loss 0.4987 (0.6121) Prec@1 82.812 (83.240) Prec@5 98.438 (96.137) Epoch: [46][1160/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.004) Loss 0.8302 (0.6125) Prec@1 75.000 (83.220) Prec@5 93.750 (96.137) Epoch: [46][1180/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.001 (0.004) Loss 0.6246 (0.6120) Prec@1 79.688 (83.225) Prec@5 98.438 (96.138) Epoch: [46][1200/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.004) Loss 0.6615 (0.6122) Prec@1 81.250 (83.220) Prec@5 95.312 (96.135) Epoch: [46][1220/1345], lr: 0.00010 Time 0.804 (0.750) Data 0.000 (0.004) Loss 0.4764 (0.6118) Prec@1 89.062 (83.231) Prec@5 96.875 (96.135) Epoch: [46][1240/1345], lr: 0.00010 Time 0.746 (0.750) Data 0.001 (0.003) Loss 0.8297 (0.6115) Prec@1 75.000 (83.212) Prec@5 92.188 (96.143) Epoch: [46][1260/1345], lr: 0.00010 Time 0.739 (0.749) Data 0.000 (0.003) Loss 0.6452 (0.6111) Prec@1 82.812 (83.230) Prec@5 95.312 (96.156) Epoch: [46][1280/1345], lr: 0.00010 Time 0.712 (0.749) Data 0.000 (0.003) Loss 0.6863 (0.6111) Prec@1 81.250 (83.225) Prec@5 95.312 (96.152) Epoch: [46][1300/1345], lr: 0.00010 Time 0.851 (0.749) Data 0.000 (0.003) Loss 0.4587 (0.6106) Prec@1 89.062 (83.234) Prec@5 96.875 (96.162) Epoch: [46][1320/1345], lr: 0.00010 Time 0.715 (0.749) Data 0.000 (0.003) Loss 0.9372 (0.6111) Prec@1 78.125 (83.198) Prec@5 89.062 (96.163) Epoch: [46][1340/1345], lr: 0.00010 Time 0.711 (0.749) Data 0.000 (0.003) Loss 0.5115 (0.6101) Prec@1 82.812 (83.234) Prec@5 98.438 (96.176) No BN layer Freezing. Test: [0/181] Time 3.555 (3.5550) Loss 2.3144 (2.3144) Prec@1 51.562 (51.562) Prec@5 82.812 (82.812) Test: [20/181] Time 0.750 (0.5812) Loss 2.2772 (2.1082) Prec@1 54.688 (53.274) Prec@5 75.000 (81.399) Test: [40/181] Time 0.896 (0.5227) Loss 2.5888 (2.1846) Prec@1 54.688 (52.248) Prec@5 71.875 (80.183) Test: [60/181] Time 1.450 (0.5059) Loss 2.6452 (2.2284) Prec@1 50.000 (51.409) Prec@5 71.875 (79.585) Test: [80/181] Time 0.747 (0.4914) Loss 1.7100 (2.2291) Prec@1 62.500 (51.119) Prec@5 87.500 (79.552) Test: [100/181] Time 0.969 (0.4879) Loss 2.9923 (2.2233) Prec@1 43.750 (51.037) Prec@5 70.312 (79.703) Test: [120/181] Time 1.263 (0.4855) Loss 2.9313 (2.2358) Prec@1 42.188 (50.968) Prec@5 73.438 (79.726) Test: [140/181] Time 1.096 (0.4813) Loss 3.1073 (2.2329) Prec@1 32.812 (50.720) Prec@5 76.562 (79.920) Test: [160/181] Time 0.738 (0.4780) Loss 2.5244 (2.2194) Prec@1 48.438 (51.106) Prec@5 81.250 (80.008) Testing Results: Prec@1 51.285 Prec@5 80.009 Loss 2.20981 Time 0.4740 No BN layer Freezing. Epoch: [47][0/1345], lr: 0.00010 Time 4.085 (4.085) Data 3.054 (3.054) Loss 0.4385 (0.4385) Prec@1 89.062 (89.062) Prec@5 96.875 (96.875) Epoch: [47][20/1345], lr: 0.00010 Time 0.716 (0.908) Data 0.000 (0.146) Loss 0.5639 (0.6250) Prec@1 84.375 (83.557) Prec@5 96.875 (95.610) Epoch: [47][40/1345], lr: 0.00010 Time 0.721 (0.830) Data 0.000 (0.075) Loss 0.5893 (0.6027) Prec@1 78.125 (83.689) Prec@5 98.438 (96.227) Epoch: [47][60/1345], lr: 0.00010 Time 0.714 (0.798) Data 0.000 (0.051) Loss 0.8499 (0.6028) Prec@1 78.125 (83.325) Prec@5 89.062 (96.286) Epoch: [47][80/1345], lr: 0.00010 Time 0.803 (0.785) Data 0.000 (0.038) Loss 0.8624 (0.5994) Prec@1 70.312 (83.430) Prec@5 96.875 (96.354) Epoch: [47][100/1345], lr: 0.00010 Time 0.817 (0.776) Data 0.000 (0.031) Loss 0.8794 (0.6094) Prec@1 78.125 (83.431) Prec@5 98.438 (96.272) Epoch: [47][120/1345], lr: 0.00010 Time 0.755 (0.772) Data 0.001 (0.026) Loss 0.5487 (0.6019) Prec@1 89.062 (83.471) Prec@5 98.438 (96.307) Epoch: [47][140/1345], lr: 0.00010 Time 0.738 (0.770) Data 0.000 (0.022) Loss 0.8283 (0.6026) Prec@1 71.875 (83.411) Prec@5 95.312 (96.299) Epoch: [47][160/1345], lr: 0.00010 Time 0.713 (0.766) Data 0.000 (0.019) Loss 0.7523 (0.5988) Prec@1 82.812 (83.472) Prec@5 93.750 (96.273) Epoch: [47][180/1345], lr: 0.00010 Time 0.776 (0.764) Data 0.000 (0.017) Loss 0.4059 (0.5980) Prec@1 85.938 (83.512) Prec@5 98.438 (96.297) Epoch: [47][200/1345], lr: 0.00010 Time 0.715 (0.762) Data 0.000 (0.016) Loss 0.4703 (0.6002) Prec@1 85.938 (83.473) Prec@5 96.875 (96.253) Epoch: [47][220/1345], lr: 0.00010 Time 0.712 (0.761) Data 0.001 (0.014) Loss 0.5575 (0.6049) Prec@1 85.938 (83.307) Prec@5 96.875 (96.232) Epoch: [47][240/1345], lr: 0.00010 Time 0.825 (0.760) Data 0.000 (0.013) Loss 0.5658 (0.6046) Prec@1 82.812 (83.364) Prec@5 96.875 (96.194) Epoch: [47][260/1345], lr: 0.00010 Time 0.839 (0.758) Data 0.000 (0.012) Loss 0.5383 (0.6041) Prec@1 87.500 (83.399) Prec@5 96.875 (96.216) Epoch: [47][280/1345], lr: 0.00010 Time 0.715 (0.757) Data 0.000 (0.011) Loss 0.7491 (0.6085) Prec@1 81.250 (83.307) Prec@5 96.875 (96.113) Epoch: [47][300/1345], lr: 0.00010 Time 0.753 (0.756) Data 0.000 (0.011) Loss 0.6923 (0.6043) Prec@1 76.562 (83.389) Prec@5 95.312 (96.179) Epoch: [47][320/1345], lr: 0.00010 Time 0.727 (0.756) Data 0.001 (0.010) Loss 0.6749 (0.6026) Prec@1 81.250 (83.416) Prec@5 93.750 (96.198) Epoch: [47][340/1345], lr: 0.00010 Time 0.723 (0.755) Data 0.000 (0.009) Loss 0.5549 (0.6022) Prec@1 82.812 (83.399) Prec@5 96.875 (96.243) Epoch: [47][360/1345], lr: 0.00010 Time 0.738 (0.755) Data 0.000 (0.009) Loss 0.3456 (0.5994) Prec@1 87.500 (83.483) Prec@5 98.438 (96.282) Epoch: [47][380/1345], lr: 0.00010 Time 0.722 (0.755) Data 0.000 (0.008) Loss 0.4230 (0.5988) Prec@1 87.500 (83.473) Prec@5 98.438 (96.272) Epoch: [47][400/1345], lr: 0.00010 Time 0.731 (0.754) Data 0.000 (0.008) Loss 0.6133 (0.5980) Prec@1 85.938 (83.529) Prec@5 92.188 (96.294) Epoch: [47][420/1345], lr: 0.00010 Time 0.730 (0.754) Data 0.001 (0.008) Loss 0.6218 (0.5984) Prec@1 82.812 (83.540) Prec@5 95.312 (96.296) Epoch: [47][440/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.007) Loss 0.4536 (0.5999) Prec@1 89.062 (83.464) Prec@5 96.875 (96.283) Epoch: [47][460/1345], lr: 0.00010 Time 0.731 (0.753) Data 0.000 (0.007) Loss 0.6162 (0.5997) Prec@1 85.938 (83.484) Prec@5 93.750 (96.262) Epoch: [47][480/1345], lr: 0.00010 Time 0.713 (0.753) Data 0.000 (0.007) Loss 0.5183 (0.5983) Prec@1 85.938 (83.514) Prec@5 96.875 (96.261) Epoch: [47][500/1345], lr: 0.00010 Time 0.773 (0.753) Data 0.000 (0.007) Loss 0.4692 (0.5974) Prec@1 84.375 (83.520) Prec@5 98.438 (96.279) Epoch: [47][520/1345], lr: 0.00010 Time 0.775 (0.752) Data 0.000 (0.006) Loss 0.8115 (0.5974) Prec@1 78.125 (83.520) Prec@5 93.750 (96.305) Epoch: [47][540/1345], lr: 0.00010 Time 0.724 (0.752) Data 0.000 (0.006) Loss 0.8599 (0.5979) Prec@1 73.438 (83.485) Prec@5 96.875 (96.332) Epoch: [47][560/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.006) Loss 0.5561 (0.5975) Prec@1 84.375 (83.509) Prec@5 98.438 (96.349) Epoch: [47][580/1345], lr: 0.00010 Time 0.712 (0.751) Data 0.000 (0.006) Loss 0.5316 (0.5971) Prec@1 82.812 (83.522) Prec@5 98.438 (96.375) Epoch: [47][600/1345], lr: 0.00010 Time 0.743 (0.751) Data 0.001 (0.006) Loss 0.6305 (0.5983) Prec@1 82.812 (83.431) Prec@5 93.750 (96.373) Epoch: [47][620/1345], lr: 0.00010 Time 0.768 (0.750) Data 0.000 (0.005) Loss 0.4441 (0.5975) Prec@1 85.938 (83.489) Prec@5 98.438 (96.392) Epoch: [47][640/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.005) Loss 0.9245 (0.5971) Prec@1 70.312 (83.512) Prec@5 92.188 (96.375) Epoch: [47][660/1345], lr: 0.00010 Time 0.748 (0.751) Data 0.000 (0.005) Loss 0.4525 (0.5983) Prec@1 87.500 (83.493) Prec@5 96.875 (96.357) Epoch: [47][680/1345], lr: 0.00010 Time 0.734 (0.750) Data 0.001 (0.005) Loss 0.5990 (0.5982) Prec@1 85.938 (83.496) Prec@5 96.875 (96.354) Epoch: [47][700/1345], lr: 0.00010 Time 0.875 (0.751) Data 0.000 (0.005) Loss 0.6300 (0.5987) Prec@1 81.250 (83.501) Prec@5 95.312 (96.347) Epoch: [47][720/1345], lr: 0.00010 Time 0.806 (0.750) Data 0.000 (0.005) Loss 0.7373 (0.5998) Prec@1 78.125 (83.469) Prec@5 95.312 (96.340) Epoch: [47][740/1345], lr: 0.00010 Time 0.728 (0.750) Data 0.000 (0.005) Loss 0.6176 (0.5998) Prec@1 82.812 (83.485) Prec@5 95.312 (96.335) Epoch: [47][760/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.000 (0.004) Loss 0.8203 (0.5997) Prec@1 76.562 (83.472) Prec@5 89.062 (96.329) Epoch: [47][780/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.004) Loss 0.5655 (0.6006) Prec@1 89.062 (83.453) Prec@5 95.312 (96.305) Epoch: [47][800/1345], lr: 0.00010 Time 0.723 (0.750) Data 0.000 (0.004) Loss 0.4866 (0.6011) Prec@1 84.375 (83.427) Prec@5 98.438 (96.294) Epoch: [47][820/1345], lr: 0.00010 Time 0.787 (0.750) Data 0.001 (0.004) Loss 0.5144 (0.6012) Prec@1 84.375 (83.420) Prec@5 95.312 (96.287) Epoch: [47][840/1345], lr: 0.00010 Time 0.716 (0.750) Data 0.000 (0.004) Loss 0.4830 (0.6018) Prec@1 87.500 (83.416) Prec@5 96.875 (96.279) Epoch: [47][860/1345], lr: 0.00010 Time 0.708 (0.750) Data 0.000 (0.004) Loss 0.8617 (0.6033) Prec@1 81.250 (83.388) Prec@5 90.625 (96.254) Epoch: [47][880/1345], lr: 0.00010 Time 0.731 (0.749) Data 0.000 (0.004) Loss 0.7060 (0.6034) Prec@1 78.125 (83.385) Prec@5 96.875 (96.261) Epoch: [47][900/1345], lr: 0.00010 Time 0.732 (0.749) Data 0.000 (0.004) Loss 0.6911 (0.6035) Prec@1 84.375 (83.376) Prec@5 90.625 (96.252) Epoch: [47][920/1345], lr: 0.00010 Time 0.774 (0.749) Data 0.000 (0.004) Loss 0.6803 (0.6038) Prec@1 76.562 (83.350) Prec@5 95.312 (96.256) Epoch: [47][940/1345], lr: 0.00010 Time 0.718 (0.749) Data 0.000 (0.004) Loss 0.3103 (0.6040) Prec@1 92.188 (83.342) Prec@5 100.000 (96.267) Epoch: [47][960/1345], lr: 0.00010 Time 0.751 (0.749) Data 0.000 (0.004) Loss 1.1731 (0.6053) Prec@1 65.625 (83.308) Prec@5 92.188 (96.254) Epoch: [47][980/1345], lr: 0.00010 Time 0.735 (0.749) Data 0.000 (0.004) Loss 0.7318 (0.6058) Prec@1 78.125 (83.274) Prec@5 96.875 (96.249) Epoch: [47][1000/1345], lr: 0.00010 Time 0.736 (0.749) Data 0.000 (0.004) Loss 0.9087 (0.6059) Prec@1 81.250 (83.281) Prec@5 93.750 (96.262) Epoch: [47][1020/1345], lr: 0.00010 Time 0.716 (0.749) Data 0.000 (0.003) Loss 0.3392 (0.6056) Prec@1 93.750 (83.307) Prec@5 100.000 (96.266) Epoch: [47][1040/1345], lr: 0.00010 Time 0.741 (0.749) Data 0.000 (0.003) Loss 0.6422 (0.6054) Prec@1 78.125 (83.306) Prec@5 100.000 (96.276) Epoch: [47][1060/1345], lr: 0.00010 Time 0.722 (0.748) Data 0.000 (0.003) Loss 0.5337 (0.6050) Prec@1 85.938 (83.340) Prec@5 95.312 (96.268) Epoch: [47][1080/1345], lr: 0.00010 Time 0.715 (0.748) Data 0.000 (0.003) Loss 0.6845 (0.6057) Prec@1 82.812 (83.314) Prec@5 95.312 (96.264) Epoch: [47][1100/1345], lr: 0.00010 Time 0.715 (0.748) Data 0.000 (0.003) Loss 0.6779 (0.6051) Prec@1 81.250 (83.332) Prec@5 95.312 (96.270) Epoch: [47][1120/1345], lr: 0.00010 Time 0.719 (0.748) Data 0.000 (0.003) Loss 0.6939 (0.6064) Prec@1 79.688 (83.323) Prec@5 95.312 (96.242) Epoch: [47][1140/1345], lr: 0.00010 Time 0.734 (0.748) Data 0.000 (0.003) Loss 0.7421 (0.6067) Prec@1 81.250 (83.334) Prec@5 95.312 (96.223) Epoch: [47][1160/1345], lr: 0.00010 Time 0.754 (0.748) Data 0.000 (0.003) Loss 0.5327 (0.6063) Prec@1 79.688 (83.362) Prec@5 98.438 (96.233) Epoch: [47][1180/1345], lr: 0.00010 Time 0.714 (0.748) Data 0.000 (0.003) Loss 0.5101 (0.6060) Prec@1 85.938 (83.385) Prec@5 98.438 (96.237) Epoch: [47][1200/1345], lr: 0.00010 Time 0.722 (0.748) Data 0.001 (0.003) Loss 0.6388 (0.6063) Prec@1 82.812 (83.368) Prec@5 95.312 (96.234) Epoch: [47][1220/1345], lr: 0.00010 Time 0.849 (0.748) Data 0.000 (0.003) Loss 0.5945 (0.6061) Prec@1 82.812 (83.376) Prec@5 96.875 (96.235) Epoch: [47][1240/1345], lr: 0.00010 Time 0.731 (0.748) Data 0.000 (0.003) Loss 0.6686 (0.6067) Prec@1 82.812 (83.355) Prec@5 96.875 (96.227) Epoch: [47][1260/1345], lr: 0.00010 Time 0.761 (0.748) Data 0.000 (0.003) Loss 0.5047 (0.6071) Prec@1 89.062 (83.356) Prec@5 98.438 (96.218) Epoch: [47][1280/1345], lr: 0.00010 Time 0.715 (0.748) Data 0.000 (0.003) Loss 0.6654 (0.6075) Prec@1 84.375 (83.354) Prec@5 95.312 (96.198) Epoch: [47][1300/1345], lr: 0.00010 Time 0.718 (0.748) Data 0.000 (0.003) Loss 0.5286 (0.6080) Prec@1 85.938 (83.343) Prec@5 96.875 (96.186) Epoch: [47][1320/1345], lr: 0.00010 Time 0.713 (0.748) Data 0.000 (0.003) Loss 0.8093 (0.6084) Prec@1 78.125 (83.339) Prec@5 93.750 (96.181) Epoch: [47][1340/1345], lr: 0.00010 Time 0.709 (0.748) Data 0.000 (0.003) Loss 0.3745 (0.6089) Prec@1 89.062 (83.311) Prec@5 98.438 (96.171) No BN layer Freezing. Test: [0/181] Time 3.125 (3.1251) Loss 2.3270 (2.3270) Prec@1 53.125 (53.125) Prec@5 84.375 (84.375) Test: [20/181] Time 1.122 (0.6097) Loss 2.2819 (2.0932) Prec@1 51.562 (52.604) Prec@5 81.250 (81.324) Test: [40/181] Time 1.125 (0.5380) Loss 2.6187 (2.1783) Prec@1 53.125 (52.096) Prec@5 71.875 (80.259) Test: [60/181] Time 1.000 (0.5106) Loss 2.7782 (2.2270) Prec@1 51.562 (51.101) Prec@5 70.312 (79.918) Test: [80/181] Time 0.958 (0.4976) Loss 1.7625 (2.2257) Prec@1 62.500 (50.984) Prec@5 89.062 (79.823) Test: [100/181] Time 1.167 (0.4913) Loss 3.0080 (2.2170) Prec@1 45.312 (50.897) Prec@5 68.750 (79.842) Test: [120/181] Time 0.982 (0.4845) Loss 2.9283 (2.2275) Prec@1 40.625 (50.839) Prec@5 73.438 (79.804) Test: [140/181] Time 1.117 (0.4806) Loss 3.1486 (2.2267) Prec@1 31.250 (50.643) Prec@5 76.562 (79.931) Test: [160/181] Time 1.001 (0.4777) Loss 2.5498 (2.2131) Prec@1 46.875 (50.922) Prec@5 81.250 (80.124) Testing Results: Prec@1 51.128 Prec@5 80.252 Loss 2.20378 Time 0.4710 No BN layer Freezing. Epoch: [48][0/1345], lr: 0.00010 Time 3.847 (3.847) Data 3.090 (3.090) Loss 0.7695 (0.7695) Prec@1 76.562 (76.562) Prec@5 96.875 (96.875) Epoch: [48][20/1345], lr: 0.00010 Time 0.759 (0.894) Data 0.001 (0.148) Loss 0.7374 (0.6367) Prec@1 82.812 (81.994) Prec@5 92.188 (96.354) Epoch: [48][40/1345], lr: 0.00010 Time 0.749 (0.822) Data 0.001 (0.076) Loss 0.6587 (0.6242) Prec@1 84.375 (82.393) Prec@5 95.312 (96.418) Epoch: [48][60/1345], lr: 0.00010 Time 0.832 (0.798) Data 0.000 (0.051) Loss 0.7276 (0.6074) Prec@1 81.250 (82.633) Prec@5 93.750 (96.465) Epoch: [48][80/1345], lr: 0.00010 Time 0.869 (0.784) Data 0.001 (0.039) Loss 0.6347 (0.6004) Prec@1 76.562 (82.774) Prec@5 98.438 (96.547) Epoch: [48][100/1345], lr: 0.00010 Time 0.747 (0.776) Data 0.000 (0.031) Loss 0.6598 (0.5980) Prec@1 82.812 (82.967) Prec@5 96.875 (96.581) Epoch: [48][120/1345], lr: 0.00010 Time 0.748 (0.772) Data 0.000 (0.026) Loss 0.6143 (0.6061) Prec@1 75.000 (82.748) Prec@5 96.875 (96.346) Epoch: [48][140/1345], lr: 0.00010 Time 0.718 (0.770) Data 0.000 (0.022) Loss 0.5297 (0.6053) Prec@1 82.812 (82.868) Prec@5 96.875 (96.365) Epoch: [48][160/1345], lr: 0.00010 Time 0.734 (0.767) Data 0.000 (0.020) Loss 0.5115 (0.6045) Prec@1 89.062 (83.036) Prec@5 100.000 (96.332) Epoch: [48][180/1345], lr: 0.00010 Time 0.715 (0.764) Data 0.000 (0.018) Loss 0.7634 (0.6092) Prec@1 79.688 (82.968) Prec@5 95.312 (96.279) Epoch: [48][200/1345], lr: 0.00010 Time 0.757 (0.764) Data 0.000 (0.016) Loss 0.5650 (0.6097) Prec@1 79.688 (82.914) Prec@5 96.875 (96.276) Epoch: [48][220/1345], lr: 0.00010 Time 0.827 (0.762) Data 0.000 (0.014) Loss 0.8080 (0.6094) Prec@1 78.125 (82.911) Prec@5 90.625 (96.281) Epoch: [48][240/1345], lr: 0.00010 Time 0.842 (0.761) Data 0.000 (0.013) Loss 0.6021 (0.6102) Prec@1 81.250 (82.923) Prec@5 98.438 (96.253) Epoch: [48][260/1345], lr: 0.00010 Time 0.716 (0.759) Data 0.000 (0.012) Loss 0.5360 (0.6065) Prec@1 76.562 (82.848) Prec@5 98.438 (96.336) Epoch: [48][280/1345], lr: 0.00010 Time 0.738 (0.758) Data 0.000 (0.011) Loss 0.7212 (0.6068) Prec@1 85.938 (82.946) Prec@5 93.750 (96.325) Epoch: [48][300/1345], lr: 0.00010 Time 0.737 (0.758) Data 0.001 (0.011) Loss 0.6802 (0.6061) Prec@1 82.812 (83.046) Prec@5 95.312 (96.294) Epoch: [48][320/1345], lr: 0.00010 Time 0.715 (0.756) Data 0.000 (0.010) Loss 0.6361 (0.6067) Prec@1 82.812 (83.085) Prec@5 96.875 (96.315) Epoch: [48][340/1345], lr: 0.00010 Time 0.714 (0.755) Data 0.000 (0.010) Loss 0.6192 (0.6072) Prec@1 81.250 (83.097) Prec@5 96.875 (96.307) Epoch: [48][360/1345], lr: 0.00010 Time 0.713 (0.755) Data 0.000 (0.009) Loss 0.5977 (0.6094) Prec@1 79.688 (83.016) Prec@5 96.875 (96.256) Epoch: [48][380/1345], lr: 0.00010 Time 0.837 (0.755) Data 0.000 (0.009) Loss 0.5450 (0.6086) Prec@1 85.938 (83.005) Prec@5 96.875 (96.248) Epoch: [48][400/1345], lr: 0.00010 Time 0.831 (0.754) Data 0.000 (0.008) Loss 0.6703 (0.6093) Prec@1 84.375 (83.031) Prec@5 98.438 (96.240) Epoch: [48][420/1345], lr: 0.00010 Time 0.736 (0.754) Data 0.001 (0.008) Loss 0.9962 (0.6114) Prec@1 73.438 (82.968) Prec@5 87.500 (96.192) Epoch: [48][440/1345], lr: 0.00010 Time 0.728 (0.754) Data 0.000 (0.007) Loss 0.7416 (0.6124) Prec@1 81.250 (82.968) Prec@5 95.312 (96.166) Epoch: [48][460/1345], lr: 0.00010 Time 0.717 (0.754) Data 0.000 (0.007) Loss 0.6762 (0.6125) Prec@1 82.812 (82.955) Prec@5 92.188 (96.184) Epoch: [48][480/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.007) Loss 0.3988 (0.6109) Prec@1 89.062 (82.991) Prec@5 100.000 (96.203) Epoch: [48][500/1345], lr: 0.00010 Time 0.785 (0.753) Data 0.000 (0.007) Loss 0.9950 (0.6101) Prec@1 81.250 (82.993) Prec@5 89.062 (96.217) Epoch: [48][520/1345], lr: 0.00010 Time 0.712 (0.753) Data 0.000 (0.006) Loss 0.6868 (0.6114) Prec@1 84.375 (83.004) Prec@5 93.750 (96.188) Epoch: [48][540/1345], lr: 0.00010 Time 0.878 (0.753) Data 0.000 (0.006) Loss 0.4488 (0.6106) Prec@1 85.938 (83.012) Prec@5 98.438 (96.196) Epoch: [48][560/1345], lr: 0.00010 Time 0.869 (0.752) Data 0.000 (0.006) Loss 0.5845 (0.6114) Prec@1 84.375 (83.005) Prec@5 96.875 (96.198) Epoch: [48][580/1345], lr: 0.00010 Time 0.734 (0.752) Data 0.000 (0.006) Loss 0.6653 (0.6104) Prec@1 84.375 (83.022) Prec@5 95.312 (96.221) Epoch: [48][600/1345], lr: 0.00010 Time 0.767 (0.752) Data 0.000 (0.006) Loss 0.6030 (0.6091) Prec@1 82.812 (83.080) Prec@5 95.312 (96.243) Epoch: [48][620/1345], lr: 0.00010 Time 0.716 (0.752) Data 0.000 (0.005) Loss 0.5188 (0.6076) Prec@1 84.375 (83.124) Prec@5 98.438 (96.276) Epoch: [48][640/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.005) Loss 0.8757 (0.6077) Prec@1 73.438 (83.115) Prec@5 95.312 (96.283) Epoch: [48][660/1345], lr: 0.00010 Time 0.745 (0.752) Data 0.000 (0.005) Loss 0.6382 (0.6069) Prec@1 81.250 (83.132) Prec@5 96.875 (96.296) Epoch: [48][680/1345], lr: 0.00010 Time 0.736 (0.752) Data 0.000 (0.005) Loss 0.6058 (0.6079) Prec@1 81.250 (83.076) Prec@5 98.438 (96.281) Epoch: [48][700/1345], lr: 0.00010 Time 0.726 (0.752) Data 0.000 (0.005) Loss 0.7109 (0.6076) Prec@1 78.125 (83.082) Prec@5 96.875 (96.289) Epoch: [48][720/1345], lr: 0.00010 Time 0.724 (0.752) Data 0.000 (0.005) Loss 0.5634 (0.6080) Prec@1 82.812 (83.075) Prec@5 98.438 (96.266) Epoch: [48][740/1345], lr: 0.00010 Time 0.713 (0.751) Data 0.000 (0.005) Loss 0.4644 (0.6079) Prec@1 87.500 (83.101) Prec@5 96.875 (96.261) Epoch: [48][760/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.005) Loss 0.6385 (0.6069) Prec@1 85.938 (83.129) Prec@5 96.875 (96.280) Epoch: [48][780/1345], lr: 0.00010 Time 0.827 (0.751) Data 0.000 (0.004) Loss 0.6537 (0.6069) Prec@1 79.688 (83.117) Prec@5 96.875 (96.293) Epoch: [48][800/1345], lr: 0.00010 Time 0.842 (0.751) Data 0.000 (0.004) Loss 0.5305 (0.6059) Prec@1 81.250 (83.125) Prec@5 98.438 (96.311) Epoch: [48][820/1345], lr: 0.00010 Time 0.754 (0.751) Data 0.000 (0.004) Loss 0.4580 (0.6060) Prec@1 92.188 (83.107) Prec@5 100.000 (96.315) Epoch: [48][840/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.004) Loss 0.3437 (0.6052) Prec@1 90.625 (83.145) Prec@5 98.438 (96.312) Epoch: [48][860/1345], lr: 0.00010 Time 0.845 (0.751) Data 0.000 (0.004) Loss 0.5428 (0.6046) Prec@1 85.938 (83.157) Prec@5 95.312 (96.316) Epoch: [48][880/1345], lr: 0.00010 Time 0.761 (0.751) Data 0.000 (0.004) Loss 0.4977 (0.6044) Prec@1 85.938 (83.192) Prec@5 96.875 (96.320) Epoch: [48][900/1345], lr: 0.00010 Time 0.712 (0.751) Data 0.000 (0.004) Loss 0.6380 (0.6048) Prec@1 82.812 (83.175) Prec@5 92.188 (96.294) Epoch: [48][920/1345], lr: 0.00010 Time 0.718 (0.751) Data 0.001 (0.004) Loss 0.6505 (0.6055) Prec@1 84.375 (83.150) Prec@5 95.312 (96.288) Epoch: [48][940/1345], lr: 0.00010 Time 0.733 (0.750) Data 0.001 (0.004) Loss 0.5941 (0.6052) Prec@1 85.938 (83.179) Prec@5 98.438 (96.294) Epoch: [48][960/1345], lr: 0.00010 Time 0.729 (0.750) Data 0.001 (0.004) Loss 0.7151 (0.6048) Prec@1 79.688 (83.203) Prec@5 92.188 (96.290) Epoch: [48][980/1345], lr: 0.00010 Time 0.713 (0.750) Data 0.000 (0.004) Loss 0.6840 (0.6047) Prec@1 78.125 (83.228) Prec@5 93.750 (96.292) Epoch: [48][1000/1345], lr: 0.00010 Time 0.739 (0.750) Data 0.000 (0.004) Loss 0.7609 (0.6047) Prec@1 85.938 (83.229) Prec@5 96.875 (96.301) Epoch: [48][1020/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.003) Loss 0.4124 (0.6043) Prec@1 87.500 (83.246) Prec@5 98.438 (96.300) Epoch: [48][1040/1345], lr: 0.00010 Time 0.717 (0.750) Data 0.000 (0.003) Loss 0.5822 (0.6043) Prec@1 84.375 (83.230) Prec@5 96.875 (96.308) Epoch: [48][1060/1345], lr: 0.00010 Time 0.733 (0.750) Data 0.000 (0.003) Loss 0.8139 (0.6039) Prec@1 79.688 (83.237) Prec@5 93.750 (96.312) Epoch: [48][1080/1345], lr: 0.00010 Time 0.714 (0.750) Data 0.000 (0.003) Loss 0.5813 (0.6033) Prec@1 82.812 (83.275) Prec@5 98.438 (96.320) Epoch: [48][1100/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.003) Loss 0.4683 (0.6023) Prec@1 84.375 (83.304) Prec@5 98.438 (96.323) Epoch: [48][1120/1345], lr: 0.00010 Time 0.727 (0.749) Data 0.000 (0.003) Loss 0.4211 (0.6023) Prec@1 87.500 (83.321) Prec@5 98.438 (96.317) Epoch: [48][1140/1345], lr: 0.00010 Time 0.848 (0.749) Data 0.000 (0.003) Loss 0.6387 (0.6023) Prec@1 81.250 (83.318) Prec@5 93.750 (96.315) Epoch: [48][1160/1345], lr: 0.00010 Time 0.713 (0.749) Data 0.000 (0.003) Loss 0.6969 (0.6020) Prec@1 78.125 (83.339) Prec@5 95.312 (96.321) Epoch: [48][1180/1345], lr: 0.00010 Time 0.716 (0.749) Data 0.000 (0.003) Loss 0.6787 (0.6023) Prec@1 82.812 (83.322) Prec@5 92.188 (96.319) Epoch: [48][1200/1345], lr: 0.00010 Time 0.777 (0.749) Data 0.001 (0.003) Loss 0.9541 (0.6028) Prec@1 75.000 (83.294) Prec@5 93.750 (96.312) Epoch: [48][1220/1345], lr: 0.00010 Time 0.723 (0.749) Data 0.000 (0.003) Loss 0.4296 (0.6024) Prec@1 90.625 (83.310) Prec@5 98.438 (96.312) Epoch: [48][1240/1345], lr: 0.00010 Time 0.716 (0.749) Data 0.000 (0.003) Loss 0.4244 (0.6029) Prec@1 89.062 (83.302) Prec@5 98.438 (96.302) Epoch: [48][1260/1345], lr: 0.00010 Time 0.715 (0.749) Data 0.000 (0.003) Loss 0.7735 (0.6036) Prec@1 76.562 (83.290) Prec@5 96.875 (96.286) Epoch: [48][1280/1345], lr: 0.00010 Time 0.717 (0.749) Data 0.000 (0.003) Loss 0.8258 (0.6035) Prec@1 76.562 (83.292) Prec@5 93.750 (96.285) Epoch: [48][1300/1345], lr: 0.00010 Time 0.748 (0.749) Data 0.000 (0.003) Loss 0.7461 (0.6029) Prec@1 82.812 (83.317) Prec@5 92.188 (96.289) Epoch: [48][1320/1345], lr: 0.00010 Time 0.727 (0.749) Data 0.001 (0.003) Loss 0.5230 (0.6026) Prec@1 84.375 (83.323) Prec@5 96.875 (96.280) Epoch: [48][1340/1345], lr: 0.00010 Time 0.712 (0.748) Data 0.000 (0.003) Loss 0.6995 (0.6030) Prec@1 82.812 (83.293) Prec@5 96.875 (96.280) No BN layer Freezing. Test: [0/181] Time 2.869 (2.8691) Loss 2.3322 (2.3322) Prec@1 51.562 (51.562) Prec@5 81.250 (81.250) Test: [20/181] Time 0.940 (0.5871) Loss 2.2853 (2.1124) Prec@1 51.562 (52.381) Prec@5 79.688 (81.176) Test: [40/181] Time 0.969 (0.5250) Loss 2.6039 (2.1892) Prec@1 53.125 (51.944) Prec@5 71.875 (80.450) Test: [60/181] Time 0.918 (0.4994) Loss 2.6953 (2.2367) Prec@1 51.562 (50.948) Prec@5 70.312 (79.841) Test: [80/181] Time 0.594 (0.4923) Loss 1.7011 (2.2353) Prec@1 60.938 (50.887) Prec@5 89.062 (79.745) Test: [100/181] Time 0.605 (0.4861) Loss 3.0320 (2.2241) Prec@1 45.312 (50.882) Prec@5 70.312 (79.842) Test: [120/181] Time 0.319 (0.4791) Loss 2.8208 (2.2315) Prec@1 43.750 (50.904) Prec@5 73.438 (79.726) Test: [140/181] Time 0.645 (0.4781) Loss 3.1452 (2.2313) Prec@1 35.938 (50.609) Prec@5 75.000 (79.887) Test: [160/181] Time 0.578 (0.4755) Loss 2.5203 (2.2165) Prec@1 46.875 (50.970) Prec@5 82.812 (80.056) Testing Results: Prec@1 51.172 Prec@5 80.139 Loss 2.20831 Time 0.4727 No BN layer Freezing. Epoch: [49][0/1345], lr: 0.00010 Time 3.853 (3.853) Data 3.095 (3.095) Loss 0.5710 (0.5710) Prec@1 87.500 (87.500) Prec@5 96.875 (96.875) Epoch: [49][20/1345], lr: 0.00010 Time 0.773 (0.895) Data 0.000 (0.148) Loss 0.5939 (0.5961) Prec@1 81.250 (83.631) Prec@5 96.875 (96.429) Epoch: [49][40/1345], lr: 0.00010 Time 0.767 (0.827) Data 0.000 (0.076) Loss 0.5595 (0.5927) Prec@1 82.812 (83.613) Prec@5 96.875 (96.265) Epoch: [49][60/1345], lr: 0.00010 Time 0.715 (0.803) Data 0.000 (0.051) Loss 0.5830 (0.5897) Prec@1 84.375 (83.658) Prec@5 96.875 (96.260) Epoch: [49][80/1345], lr: 0.00010 Time 0.743 (0.789) Data 0.000 (0.039) Loss 0.6321 (0.5864) Prec@1 79.688 (83.970) Prec@5 96.875 (96.335) Epoch: [49][100/1345], lr: 0.00010 Time 0.733 (0.779) Data 0.000 (0.031) Loss 0.6188 (0.5757) Prec@1 78.125 (84.344) Prec@5 98.438 (96.566) Epoch: [49][120/1345], lr: 0.00010 Time 0.762 (0.775) Data 0.000 (0.026) Loss 0.6589 (0.5810) Prec@1 84.375 (84.117) Prec@5 93.750 (96.578) Epoch: [49][140/1345], lr: 0.00010 Time 0.734 (0.771) Data 0.000 (0.022) Loss 0.4921 (0.5881) Prec@1 82.812 (84.031) Prec@5 100.000 (96.443) Epoch: [49][160/1345], lr: 0.00010 Time 0.751 (0.767) Data 0.000 (0.020) Loss 0.8370 (0.5948) Prec@1 79.688 (83.899) Prec@5 93.750 (96.332) Epoch: [49][180/1345], lr: 0.00010 Time 0.759 (0.767) Data 0.000 (0.018) Loss 0.6329 (0.5960) Prec@1 79.688 (83.883) Prec@5 96.875 (96.323) Epoch: [49][200/1345], lr: 0.00010 Time 0.820 (0.766) Data 0.000 (0.016) Loss 0.5384 (0.5949) Prec@1 84.375 (83.862) Prec@5 98.438 (96.416) Epoch: [49][220/1345], lr: 0.00010 Time 0.723 (0.765) Data 0.000 (0.014) Loss 0.5479 (0.5967) Prec@1 85.938 (83.675) Prec@5 95.312 (96.423) Epoch: [49][240/1345], lr: 0.00010 Time 0.734 (0.763) Data 0.000 (0.013) Loss 0.5307 (0.5944) Prec@1 84.375 (83.694) Prec@5 100.000 (96.467) Epoch: [49][260/1345], lr: 0.00010 Time 0.714 (0.761) Data 0.000 (0.012) Loss 0.4363 (0.5959) Prec@1 85.938 (83.645) Prec@5 100.000 (96.474) Epoch: [49][280/1345], lr: 0.00010 Time 0.724 (0.760) Data 0.000 (0.011) Loss 0.4213 (0.5965) Prec@1 90.625 (83.697) Prec@5 98.438 (96.408) Epoch: [49][300/1345], lr: 0.00010 Time 0.715 (0.759) Data 0.000 (0.011) Loss 0.7489 (0.5959) Prec@1 78.125 (83.674) Prec@5 95.312 (96.397) Epoch: [49][320/1345], lr: 0.00010 Time 0.713 (0.758) Data 0.000 (0.010) Loss 0.8065 (0.5963) Prec@1 78.125 (83.684) Prec@5 93.750 (96.398) Epoch: [49][340/1345], lr: 0.00010 Time 0.710 (0.757) Data 0.000 (0.010) Loss 0.6718 (0.5964) Prec@1 78.125 (83.665) Prec@5 96.875 (96.371) Epoch: [49][360/1345], lr: 0.00010 Time 0.720 (0.757) Data 0.000 (0.009) Loss 0.5233 (0.5997) Prec@1 84.375 (83.574) Prec@5 96.875 (96.347) Epoch: [49][380/1345], lr: 0.00010 Time 0.754 (0.756) Data 0.000 (0.009) Loss 0.6615 (0.6021) Prec@1 78.125 (83.551) Prec@5 93.750 (96.313) Epoch: [49][400/1345], lr: 0.00010 Time 0.772 (0.755) Data 0.000 (0.008) Loss 0.6759 (0.6044) Prec@1 82.812 (83.510) Prec@5 95.312 (96.275) Epoch: [49][420/1345], lr: 0.00010 Time 0.715 (0.755) Data 0.000 (0.008) Loss 0.5160 (0.6061) Prec@1 85.938 (83.440) Prec@5 96.875 (96.244) Epoch: [49][440/1345], lr: 0.00010 Time 0.714 (0.754) Data 0.000 (0.007) Loss 0.6957 (0.6054) Prec@1 84.375 (83.510) Prec@5 95.312 (96.251) Epoch: [49][460/1345], lr: 0.00010 Time 0.713 (0.753) Data 0.000 (0.007) Loss 0.6223 (0.6046) Prec@1 81.250 (83.565) Prec@5 98.438 (96.248) Epoch: [49][480/1345], lr: 0.00010 Time 0.754 (0.753) Data 0.001 (0.007) Loss 0.4875 (0.6051) Prec@1 85.938 (83.556) Prec@5 100.000 (96.245) Epoch: [49][500/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.007) Loss 0.6408 (0.6035) Prec@1 85.938 (83.586) Prec@5 95.312 (96.279) Epoch: [49][520/1345], lr: 0.00010 Time 0.758 (0.753) Data 0.001 (0.006) Loss 0.6485 (0.6051) Prec@1 76.562 (83.508) Prec@5 95.312 (96.278) Epoch: [49][540/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.6654 (0.6036) Prec@1 76.562 (83.491) Prec@5 100.000 (96.306) Epoch: [49][560/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.6423 (0.6047) Prec@1 84.375 (83.498) Prec@5 95.312 (96.254) Epoch: [49][580/1345], lr: 0.00010 Time 0.715 (0.752) Data 0.000 (0.006) Loss 0.4808 (0.6048) Prec@1 84.375 (83.488) Prec@5 100.000 (96.262) Epoch: [49][600/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.006) Loss 0.7644 (0.6047) Prec@1 76.562 (83.478) Prec@5 93.750 (96.246) Epoch: [49][620/1345], lr: 0.00010 Time 0.904 (0.753) Data 0.001 (0.005) Loss 0.3489 (0.6043) Prec@1 90.625 (83.494) Prec@5 98.438 (96.256) Epoch: [49][640/1345], lr: 0.00010 Time 0.857 (0.753) Data 0.000 (0.005) Loss 0.6945 (0.6066) Prec@1 85.938 (83.437) Prec@5 95.312 (96.222) Epoch: [49][660/1345], lr: 0.00010 Time 0.715 (0.753) Data 0.000 (0.005) Loss 0.5379 (0.6055) Prec@1 82.812 (83.467) Prec@5 96.875 (96.220) Epoch: [49][680/1345], lr: 0.00010 Time 0.714 (0.752) Data 0.000 (0.005) Loss 0.4426 (0.6050) Prec@1 89.062 (83.471) Prec@5 98.438 (96.230) Epoch: [49][700/1345], lr: 0.00010 Time 0.712 (0.752) Data 0.000 (0.005) Loss 0.3300 (0.6053) Prec@1 90.625 (83.446) Prec@5 98.438 (96.226) Epoch: [49][720/1345], lr: 0.00010 Time 0.763 (0.752) Data 0.000 (0.005) Loss 0.6786 (0.6053) Prec@1 78.125 (83.452) Prec@5 96.875 (96.227) Epoch: [49][740/1345], lr: 0.00010 Time 0.774 (0.752) Data 0.000 (0.005) Loss 0.7292 (0.6053) Prec@1 78.125 (83.454) Prec@5 96.875 (96.242) Epoch: [49][760/1345], lr: 0.00010 Time 0.713 (0.752) Data 0.000 (0.005) Loss 0.5929 (0.6044) Prec@1 85.938 (83.506) Prec@5 96.875 (96.247) Epoch: [49][780/1345], lr: 0.00010 Time 0.883 (0.752) Data 0.000 (0.004) Loss 0.7419 (0.6037) Prec@1 78.125 (83.531) Prec@5 95.312 (96.255) Epoch: [49][800/1345], lr: 0.00010 Time 0.870 (0.751) Data 0.001 (0.004) Loss 0.6611 (0.6037) Prec@1 87.500 (83.538) Prec@5 93.750 (96.257) Epoch: [49][820/1345], lr: 0.00010 Time 0.726 (0.751) Data 0.000 (0.004) Loss 0.5959 (0.6040) Prec@1 82.812 (83.541) Prec@5 95.312 (96.247) Epoch: [49][840/1345], lr: 0.00010 Time 0.732 (0.751) Data 0.000 (0.004) Loss 0.5842 (0.6046) Prec@1 85.938 (83.517) Prec@5 96.875 (96.232) Epoch: [49][860/1345], lr: 0.00010 Time 0.733 (0.751) Data 0.000 (0.004) Loss 0.6471 (0.6039) Prec@1 84.375 (83.540) Prec@5 95.312 (96.247) Epoch: [49][880/1345], lr: 0.00010 Time 0.714 (0.751) Data 0.000 (0.004) Loss 0.7712 (0.6049) Prec@1 75.000 (83.504) Prec@5 96.875 (96.242) Epoch: [49][900/1345], lr: 0.00010 Time 0.758 (0.751) Data 0.001 (0.004) Loss 0.5742 (0.6045) Prec@1 84.375 (83.504) Prec@5 95.312 (96.239) Epoch: [49][920/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.004) Loss 1.0402 (0.6047) Prec@1 68.750 (83.493) Prec@5 90.625 (96.252) Epoch: [49][940/1345], lr: 0.00010 Time 0.820 (0.751) Data 0.000 (0.004) Loss 0.6403 (0.6054) Prec@1 78.125 (83.467) Prec@5 98.438 (96.249) Epoch: [49][960/1345], lr: 0.00010 Time 0.734 (0.751) Data 0.000 (0.004) Loss 0.4071 (0.6053) Prec@1 89.062 (83.473) Prec@5 96.875 (96.239) Epoch: [49][980/1345], lr: 0.00010 Time 0.715 (0.751) Data 0.001 (0.004) Loss 0.3947 (0.6037) Prec@1 85.938 (83.502) Prec@5 96.875 (96.254) Epoch: [49][1000/1345], lr: 0.00010 Time 0.737 (0.751) Data 0.001 (0.004) Loss 0.4342 (0.6034) Prec@1 92.188 (83.506) Prec@5 98.438 (96.257) Epoch: [49][1020/1345], lr: 0.00010 Time 0.852 (0.751) Data 0.001 (0.004) Loss 0.8137 (0.6037) Prec@1 78.125 (83.492) Prec@5 92.188 (96.260) Epoch: [49][1040/1345], lr: 0.00010 Time 0.840 (0.751) Data 0.001 (0.003) Loss 0.5276 (0.6035) Prec@1 84.375 (83.488) Prec@5 95.312 (96.255) Epoch: [49][1060/1345], lr: 0.00010 Time 0.715 (0.751) Data 0.000 (0.003) Loss 0.6288 (0.6036) Prec@1 78.125 (83.490) Prec@5 95.312 (96.254) Epoch: [49][1080/1345], lr: 0.00010 Time 0.736 (0.751) Data 0.000 (0.003) Loss 0.8776 (0.6040) Prec@1 76.562 (83.467) Prec@5 92.188 (96.246) Epoch: [49][1100/1345], lr: 0.00010 Time 0.735 (0.751) Data 0.000 (0.003) Loss 0.4803 (0.6036) Prec@1 85.938 (83.484) Prec@5 98.438 (96.252) Epoch: [49][1120/1345], lr: 0.00010 Time 0.768 (0.751) Data 0.000 (0.003) Loss 0.4656 (0.6039) Prec@1 85.938 (83.468) Prec@5 95.312 (96.259) Epoch: [49][1140/1345], lr: 0.00010 Time 0.716 (0.751) Data 0.000 (0.003) Loss 0.5007 (0.6039) Prec@1 85.938 (83.489) Prec@5 98.438 (96.263) Epoch: [49][1160/1345], lr: 0.00010 Time 0.845 (0.751) Data 0.000 (0.003) Loss 0.5786 (0.6039) Prec@1 78.125 (83.479) Prec@5 98.438 (96.261) Epoch: [49][1180/1345], lr: 0.00010 Time 0.725 (0.751) Data 0.000 (0.003) Loss 0.7724 (0.6041) Prec@1 76.562 (83.479) Prec@5 92.188 (96.244) Epoch: [49][1200/1345], lr: 0.00010 Time 0.772 (0.750) Data 0.000 (0.003) Loss 0.6932 (0.6043) Prec@1 84.375 (83.492) Prec@5 96.875 (96.244) Epoch: [49][1220/1345], lr: 0.00010 Time 0.782 (0.750) Data 0.000 (0.003) Loss 0.7678 (0.6040) Prec@1 75.000 (83.488) Prec@5 95.312 (96.247) Epoch: [49][1240/1345], lr: 0.00010 Time 0.727 (0.750) Data 0.000 (0.003) Loss 0.7267 (0.6043) Prec@1 84.375 (83.470) Prec@5 96.875 (96.242) Epoch: [49][1260/1345], lr: 0.00010 Time 0.715 (0.750) Data 0.000 (0.003) Loss 0.7005 (0.6038) Prec@1 84.375 (83.468) Prec@5 93.750 (96.250) Epoch: [49][1280/1345], lr: 0.00010 Time 0.712 (0.750) Data 0.000 (0.003) Loss 0.4535 (0.6047) Prec@1 87.500 (83.433) Prec@5 100.000 (96.246) Epoch: [49][1300/1345], lr: 0.00010 Time 0.862 (0.749) Data 0.001 (0.003) Loss 0.4903 (0.6045) Prec@1 89.062 (83.455) Prec@5 96.875 (96.247) Epoch: [49][1320/1345], lr: 0.00010 Time 0.734 (0.749) Data 0.000 (0.003) Loss 0.8703 (0.6049) Prec@1 76.562 (83.450) Prec@5 92.188 (96.234) Epoch: [49][1340/1345], lr: 0.00010 Time 0.710 (0.749) Data 0.000 (0.003) Loss 0.5726 (0.6044) Prec@1 85.938 (83.449) Prec@5 98.438 (96.241) No BN layer Freezing. Test: [0/181] Time 3.578 (3.5780) Loss 2.2648 (2.2648) Prec@1 54.688 (54.688) Prec@5 84.375 (84.375) Test: [20/181] Time 1.231 (0.6075) Loss 2.2842 (2.0733) Prec@1 51.562 (52.902) Prec@5 81.250 (81.473) Test: [40/181] Time 0.905 (0.5311) Loss 2.5571 (2.1546) Prec@1 54.688 (51.944) Prec@5 71.875 (80.412) Test: [60/181] Time 0.808 (0.5075) Loss 2.6880 (2.2015) Prec@1 51.562 (51.076) Prec@5 70.312 (79.867) Test: [80/181] Time 0.961 (0.4936) Loss 1.7384 (2.2048) Prec@1 59.375 (50.887) Prec@5 85.938 (79.784) Test: [100/181] Time 1.009 (0.4899) Loss 2.9066 (2.1927) Prec@1 45.312 (50.913) Prec@5 70.312 (79.920) Test: [120/181] Time 1.248 (0.4875) Loss 2.9044 (2.2029) Prec@1 42.188 (51.085) Prec@5 75.000 (79.946) Test: [140/181] Time 1.209 (0.4833) Loss 3.1637 (2.2022) Prec@1 32.812 (50.842) Prec@5 73.438 (80.142) Test: [160/181] Time 0.733 (0.4793) Loss 2.5228 (2.1873) Prec@1 45.312 (51.252) Prec@5 82.812 (80.270) Testing Results: Prec@1 51.467 Prec@5 80.312 Loss 2.17791 Time 0.4755 validation score saved in net_runs/SELFY_resnet50_something_run1/validation/score.pt