Tuning Results:
Training Results:
Model: Resnet18, Optimizer: Adam, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Epoch [1/10] Train Loss: 1.4875 | Train Acc: 0.5105 || Val Loss: 1.0609 | Val Acc: 0.6685
Epoch [2/10] Train Loss: 0.9183 | Train Acc: 0.7060 || Val Loss: 0.8947 | Val Acc: 0.6896
Epoch [3/10] Train Loss: 0.7671 | Train Acc: 0.7589 || Val Loss: 0.7970 | Val Acc: 0.7275
Epoch [4/10] Train Loss: 0.7090 | Train Acc: 0.7733 || Val Loss: 0.7780 | Val Acc: 0.7289
Epoch [5/10] Train Loss: 0.6497 | Train Acc: 0.7886 || Val Loss: 0.7384 | Val Acc: 0.7486
Epoch [6/10] Train Loss: 0.6210 | Train Acc: 0.7949 || Val Loss: 0.7083 | Val Acc: 0.7486
Epoch [7/10] Train Loss: 0.5885 | Train Acc: 0.8073 || Val Loss: 0.6725 | Val Acc: 0.7711
Epoch [8/10] Train Loss: 0.5578 | Train Acc: 0.8184 || Val Loss: 0.6601 | Val Acc: 0.7711
Epoch [9/10] Train Loss: 0.5399 | Train Acc: 0.8286 || Val Loss: 0.6684 | Val Acc: 0.7669
Epoch [10/10] Train Loss: 0.5217 | Train Acc: 0.8307 || Val Loss: 0.6453 | Val Acc: 0.7809
Model: Resnet34, Optimizer: Adam, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Epoch [1/10] Train Loss: 1.5161 | Train Acc: 0.4838 || Val Loss: 1.0724 | Val Acc: 0.6559
Epoch [2/10] Train Loss: 0.9527 | Train Acc: 0.7014 || Val Loss: 0.8885 | Val Acc: 0.7205
Epoch [3/10] Train Loss: 0.8055 | Train Acc: 0.7408 || Val Loss: 0.7935 | Val Acc: 0.7430
Epoch [4/10] Train Loss: 0.7289 | Train Acc: 0.7565 || Val Loss: 0.7362 | Val Acc: 0.7542
Epoch [5/10] Train Loss: 0.6668 | Train Acc: 0.7874 || Val Loss: 0.7131 | Val Acc: 0.7514
Epoch [6/10] Train Loss: 0.6355 | Train Acc: 0.7916 || Val Loss: 0.6835 | Val Acc: 0.7711
Epoch [7/10] Train Loss: 0.5966 | Train Acc: 0.8070 || Val Loss: 0.6797 | Val Acc: 0.7626
Epoch [8/10] Train Loss: 0.5753 | Train Acc: 0.8112 || Val Loss: 0.6353 | Val Acc: 0.7893
Epoch [9/10] Train Loss: 0.5539 | Train Acc: 0.8154 || Val Loss: 0.6399 | Val Acc: 0.7725
Epoch [10/10] Train Loss: 0.5388 | Train Acc: 0.8235 || Val Loss: 0.6349 | Val Acc: 0.7739
Model: Resnet18, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Epoch [1/10] Train Loss: 1.6352 | Train Acc: 0.4447 || Val Loss: 1.2033 | Val Acc: 0.6250
Epoch [2/10] Train Loss: 1.0579 | Train Acc: 0.6795 || Val Loss: 1.0016 | Val Acc: 0.6615
Epoch [3/10] Train Loss: 0.8972 | Train Acc: 0.7225 || Val Loss: 0.9109 | Val Acc: 0.7079
Epoch [4/10] Train Loss: 0.8099 | Train Acc: 0.7495 || Val Loss: 0.8346 | Val Acc: 0.7022
Epoch [5/10] Train Loss: 0.7521 | Train Acc: 0.7619 || Val Loss: 0.8015 | Val Acc: 0.7402
Epoch [6/10] Train Loss: 0.7069 | Train Acc: 0.7739 || Val Loss: 0.7677 | Val Acc: 0.7458
Epoch [7/10] Train Loss: 0.6825 | Train Acc: 0.7856 || Val Loss: 0.7444 | Val Acc: 0.7416
Epoch [8/10] Train Loss: 0.6442 | Train Acc: 0.7946 || Val Loss: 0.7207 | Val Acc: 0.7472
Epoch [9/10] Train Loss: 0.6276 | Train Acc: 0.7913 || Val Loss: 0.7023 | Val Acc: 0.7598
Epoch [10/10] Train Loss: 0.6086 | Train Acc: 0.8010 || Val Loss: 0.6934 | Val Acc: 0.7626
Model: Resnet34, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Epoch [1/10] Train Loss: 1.6471 | Train Acc: 0.4465 || Val Loss: 1.1980 | Val Acc: 0.6194
Epoch [2/10] Train Loss: 1.0852 | Train Acc: 0.6575 || Val Loss: 0.9555 | Val Acc: 0.7065
Epoch [3/10] Train Loss: 0.9140 | Train Acc: 0.7186 || Val Loss: 0.8672 | Val Acc: 0.7079
Epoch [4/10] Train Loss: 0.8110 | Train Acc: 0.7474 || Val Loss: 0.8150 | Val Acc: 0.7388
Epoch [5/10] Train Loss: 0.7553 | Train Acc: 0.7631 || Val Loss: 0.7703 | Val Acc: 0.7458
Epoch [6/10] Train Loss: 0.7125 | Train Acc: 0.7808 || Val Loss: 0.7471 | Val Acc: 0.7346
Epoch [7/10] Train Loss: 0.6902 | Train Acc: 0.7715 || Val Loss: 0.7138 | Val Acc: 0.7542
Epoch [8/10] Train Loss: 0.6637 | Train Acc: 0.7868 || Val Loss: 0.7103 | Val Acc: 0.7486
Epoch [9/10] Train Loss: 0.6376 | Train Acc: 0.7946 || Val Loss: 0.6933 | Val Acc: 0.7683
Epoch [10/10] Train Loss: 0.6263 | Train Acc: 0.8001 || Val Loss: 0.6761 | Val Acc: 0.7640
Model: EfficientNet-B0, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Epoch [1/10] Train Loss: 1.2779 | Train Acc: 0.5884 || Val Loss: 0.8364 | Val Acc: 0.7416
Epoch [2/10] Train Loss: 0.8042 | Train Acc: 0.7444 || Val Loss: 0.7096 | Val Acc: 0.7654
Epoch [3/10] Train Loss: 0.6783 | Train Acc: 0.7847 || Val Loss: 0.6610 | Val Acc: 0.7781
Epoch [4/10] Train Loss: 0.6201 | Train Acc: 0.7940 || Val Loss: 0.6338 | Val Acc: 0.7893
Epoch [5/10] Train Loss: 0.5606 | Train Acc: 0.8199 || Val Loss: 0.6179 | Val Acc: 0.7837
Epoch [6/10] Train Loss: 0.5618 | Train Acc: 0.8202 || Val Loss: 0.5950 | Val Acc: 0.7879
Epoch [7/10] Train Loss: 0.5154 | Train Acc: 0.8310 || Val Loss: 0.5694 | Val Acc: 0.8020
Epoch [8/10] Train Loss: 0.4972 | Train Acc: 0.8331 || Val Loss: 0.5692 | Val Acc: 0.7992
Epoch [9/10] Train Loss: 0.4882 | Train Acc: 0.8322 || Val Loss: 0.5493 | Val Acc: 0.8118
Epoch [10/10] Train Loss: 0.4818 | Train Acc: 0.8391 || Val Loss: 0.5536 | Val Acc: 0.8048
Evaluation Results:
Model: Resnet18, Optimizer: Adam, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Accuracy: 0.7899 Precision: 0.7906 Recall: 0.7899 F1_score: 0.7897
Model: Resnet34, Optimizer: Adam, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Accuracy: 0.7507 Precision: 0.7560 Recall: 0.7507 F1_score: 0.7496
Model: Resnet18, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Accuracy: 0.7661 Precision: 0.7659 Recall: 0.7661 F1_score: 0.7631
Model: Resnet34, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.001, Number of Epochs: 10:
Accuracy: 0.7241 Precision: 0.7281 Recall: 0.7241 F1_score: 0.7230
Model: EfficientNet-B0, Optimizer: SGD, Batch Size: 32, Learning Rate: 0.01, Number of Epochs: 10:
Accuracy: 0.7899 Precision: 0.7910 Recall: 0.7899 F1_score: 0.7890