卷积神经网络ResNet50训练CIFAR-100图像分类Pytorch实现

Downloading: "https://download.pytorch.org/models/resnet50-0676ba61.pth" to .cache\torch\hub\checkpoints\resnet50-0676ba61.pth 100%|███████████████████████████████████████████████████████████████████████████████| 97.8M/97.8M [00:12<00:00, 8.10MB/s] Epoch [1/50], Step [100/391], Loss: 4.2021 Epoch [1/50], Step [200/391], Loss: 3.5162 Epoch [1/50], Step [300/391], Loss: 3.1082 Validation Accuracy: 34.01% Epoch [2/50], Step [100/391], Loss: 2.4628 Epoch [2/50], Step [200/391], Loss: 2.2840 Epoch [2/50], Step [300/391], Loss: 2.0992 Validation Accuracy: 45.02% Epoch [3/50], Step [100/391], Loss: 2.1158 Epoch [3/50], Step [200/391], Loss: 1.9520 Epoch [3/50], Step [300/391], Loss: 1.8579 Validation Accuracy: 50.55% Epoch [4/50], Step [100/391], Loss: 1.6844 Epoch [4/50], Step [200/391], Loss: 1.9367 Epoch [4/50], Step [300/391], Loss: 1.7156 Validation Accuracy: 53.46% Epoch [5/50], Step [100/391], Loss: 1.4833 Epoch [5/50], Step [200/391], Loss: 1.4753 Epoch [5/50], Step [300/391], Loss: 1.5026 Validation Accuracy: 55.22% Epoch [6/50], Step [100/391], Loss: 1.5737 Epoch [6/50], Step [200/391], Loss: 1.3611 Epoch [6/50], Step [300/391], Loss: 1.3267 Validation Accuracy: 56.85% Epoch [7/50], Step [100/391], Loss: 1.2998 Epoch [7/50], Step [200/391], Loss: 1.1657 Epoch [7/50], Step [300/391], Loss: 1.1932 Validation Accuracy: 57.69% Epoch [8/50], Step [100/391], Loss: 1.1266 Epoch [8/50], Step [200/391], Loss: 1.3422 Epoch [8/50], Step [300/391], Loss: 1.1819 Validation Accuracy: 58.41% Epoch [9/50], Step [100/391], Loss: 0.8817 Epoch [9/50], Step [200/391], Loss: 1.1186 Epoch [9/50], Step [300/391], Loss: 1.3608 Validation Accuracy: 58.55% Epoch [10/50], Step [100/391], Loss: 1.0705 Epoch [10/50], Step [200/391], Loss: 0.9905 Epoch [10/50], Step [300/391], Loss: 0.9832 Validation Accuracy: 59.62% Epoch [11/50], Step [100/391], Loss: 0.8682 Epoch [11/50], Step [200/391], Loss: 0.8636 Epoch [11/50], Step [300/391], Loss: 0.9976 Validation Accuracy: 59.78% Epoch [12/50], Step [100/391], Loss: 0.8009 Epoch [12/50], Step [200/391], Loss: 0.8933 Epoch [12/50], Step [300/391], Loss: 1.0769 Validation Accuracy: 60.07% Epoch [13/50], Step [100/391], Loss: 0.7024 Epoch [13/50], Step [200/391], Loss: 0.9065 Epoch [13/50], Step [300/391], Loss: 0.7613 Validation Accuracy: 60.62% Epoch [14/50], Step [100/391], Loss: 0.8446 Epoch [14/50], Step [200/391], Loss: 0.8146 Epoch [14/50], Step [300/391], Loss: 0.6286 Validation Accuracy: 60.95% Epoch [15/50], Step [100/391], Loss: 0.7601 Epoch [15/50], Step [200/391], Loss: 0.7088 Epoch [15/50], Step [300/391], Loss: 0.6358 Validation Accuracy: 61.28% Epoch [16/50], Step [100/391], Loss: 0.5757 Epoch [16/50], Step [200/391], Loss: 0.6356 Epoch [16/50], Step [300/391], Loss: 0.5989 Validation Accuracy: 60.43% Epoch [17/50], Step [100/391], Loss: 0.5597 Epoch [17/50], Step [200/391], Loss: 0.5805 Epoch [17/50], Step [300/391], Loss: 0.8085 Validation Accuracy: 60.99% Epoch [18/50], Step [100/391], Loss: 0.7153 Epoch [18/50], Step [200/391], Loss: 0.4137 Epoch [18/50], Step [300/391], Loss: 0.6954 Validation Accuracy: 61.38% Epoch [19/50], Step [100/391], Loss: 0.4247 Epoch [19/50], Step [200/391], Loss: 0.3900 Epoch [19/50], Step [300/391], Loss: 0.5415 Validation Accuracy: 61.10% Epoch [20/50], Step [100/391], Loss: 0.4829 Epoch [20/50], Step [200/391], Loss: 0.5556 Epoch [20/50], Step [300/391], Loss: 0.4416 Validation Accuracy: 61.28% Epoch [21/50], Step [100/391], Loss: 0.5183 Epoch [21/50], Step [200/391], Loss: 0.4215 Epoch [21/50], Step [300/391], Loss: 0.3845 Validation Accuracy: 61.21% Epoch [22/50], Step [100/391], Loss: 0.3864 Epoch [22/50], Step [200/391], Loss: 0.3378 Epoch [22/50], Step [300/391], Loss: 0.4182 Validation Accuracy: 61.33% Epoch [23/50], Step [100/391], Loss: 0.2676 Epoch [23/50], Step [200/391], Loss: 0.4187 Epoch [23/50], Step [300/391], Loss: 0.3868 Validation Accuracy: 61.45% Epoch [24/50], Step [100/391], Loss: 0.4253 Epoch [24/50], Step [200/391], Loss: 0.3521 Epoch [24/50], Step [300/391], Loss: 0.3772 Validation Accuracy: 61.54% Epoch [25/50], Step [100/391], Loss: 0.4616 Epoch [25/50], Step [200/391], Loss: 0.3489 Epoch [25/50], Step [300/391], Loss: 0.2740 Validation Accuracy: 61.32% Epoch [26/50], Step [100/391], Loss: 0.2591 Epoch [26/50], Step [200/391], Loss: 0.3009 Epoch [26/50], Step [300/391], Loss: 0.2503 Validation Accuracy: 61.79% Epoch [27/50], Step [100/391], Loss: 0.2868 Epoch [27/50], Step [200/391], Loss: 0.3383 Epoch [27/50], Step [300/391], Loss: 0.3256 Validation Accuracy: 61.49% Epoch [28/50], Step [100/391], Loss: 0.3169 Epoch [28/50], Step [200/391], Loss: 0.2003 Epoch [28/50], Step [300/391], Loss: 0.2982 Validation Accuracy: 61.38% Epoch [29/50], Step [100/391], Loss: 0.2476 Epoch [29/50], Step [200/391], Loss: 0.3769 Epoch [29/50], Step [300/391], Loss: 0.3777 Validation Accuracy: 61.71% Epoch [30/50], Step [100/391], Loss: 0.1876 Epoch [30/50], Step [200/391], Loss: 0.2358 Epoch [30/50], Step [300/391], Loss: 0.3366 Validation Accuracy: 61.47% Epoch [31/50], Step [100/391], Loss: 0.2167 Epoch [31/50], Step [200/391], Loss: 0.3194 Epoch [31/50], Step [300/391], Loss: 0.1910 Validation Accuracy: 61.57% Epoch [32/50], Step [100/391], Loss: 0.3247 Epoch [32/50], Step [200/391], Loss: 0.3121 Epoch [32/50], Step [300/391], Loss: 0.1938 Validation Accuracy: 62.36% Epoch [33/50], Step [100/391], Loss: 0.1650 Epoch [33/50], Step [200/391], Loss: 0.2211 Epoch [33/50], Step [300/391], Loss: 0.2623 Validation Accuracy: 61.90% Epoch [34/50], Step [100/391], Loss: 0.1727 Epoch [34/50], Step [200/391], Loss: 0.1730 Epoch [34/50], Step [300/391], Loss: 0.2433 Validation Accuracy: 61.57% Epoch [35/50], Step [100/391], Loss: 0.1655 Epoch [35/50], Step [200/391], Loss: 0.2852 Epoch [35/50], Step [300/391], Loss: 0.1833 Validation Accuracy: 61.90% Epoch [36/50], Step [100/391], Loss: 0.2079 Epoch [36/50], Step [200/391], Loss: 0.1138 Epoch [36/50], Step [300/391], Loss: 0.1901 Validation Accuracy: 62.01% Epoch [37/50], Step [100/391], Loss: 0.2267 Epoch [37/50], Step [200/391], Loss: 0.0925 Epoch [37/50], Step [300/391], Loss: 0.1216 Validation Accuracy: 61.64% Epoch [38/50], Step [100/391], Loss: 0.1777 Epoch [38/50], Step [200/391], Loss: 0.1542 Epoch [38/50], Step [300/391], Loss: 0.1572 Validation Accuracy: 61.84% Epoch [39/50], Step [100/391], Loss: 0.1672 Epoch [39/50], Step [200/391], Loss: 0.0887 Epoch [39/50], Step [300/391], Loss: 0.0516 Validation Accuracy: 61.47% Epoch [40/50], Step [100/391], Loss: 0.1339 Epoch [40/50], Step [200/391], Loss: 0.1437 Epoch [40/50], Step [300/391], Loss: 0.0863 Validation Accuracy: 61.85% Epoch [41/50], Step [100/391], Loss: 0.1820 Epoch [41/50], Step [200/391], Loss: 0.1874 Epoch [41/50], Step [300/391], Loss: 0.1178 Validation Accuracy: 61.64% Epoch [42/50], Step [100/391], Loss: 0.1174 Epoch [42/50], Step [200/391], Loss: 0.1430 Epoch [42/50], Step [300/391], Loss: 0.0906 Validation Accuracy: 62.58% Epoch [43/50], Step [100/391], Loss: 0.0689 Epoch [43/50], Step [200/391], Loss: 0.1139 Epoch [43/50], Step [300/391], Loss: 0.2274 Validation Accuracy: 62.26% Epoch [44/50], Step [100/391], Loss: 0.0972 Epoch [44/50], Step [200/391], Loss: 0.1784 Epoch [44/50], Step [300/391], Loss: 0.0997 Validation Accuracy: 61.96% Epoch [45/50], Step [100/391], Loss: 0.0850 Epoch [45/50], Step [200/391], Loss: 0.0822 Epoch [45/50], Step [300/391], Loss: 0.0806 Validation Accuracy: 62.60% Epoch [46/50], Step [100/391], Loss: 0.0957 Epoch [46/50], Step [200/391], Loss: 0.0794 Epoch [46/50], Step [300/391], Loss: 0.1143 Validation Accuracy: 61.97% Epoch [47/50], Step [100/391], Loss: 0.1071 Epoch [47/50], Step [200/391]