diff --git a/results/adam/resnet/confusion_matrix.png b/results/adam/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..474b6b691891b638c1dca9555de61af984dc647d --- /dev/null +++ b/results/adam/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:63b46a395203681f633ec241764ad903b7af6d0f9fbea6382e0f7d7906edd13d +size 68027 diff --git a/results/adam/resnet/log.csv b/results/adam/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..74658e8c00fca38c3a1b581f0d63e03df78c3013 --- /dev/null +++ b/results/adam/resnet/log.csv @@ -0,0 +1,25 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6890696138143539,0.475,0.3261648745519713,0.14264919305520904,0.000125 +1,0.6811029762029648,0.725,0.4623655913978494,0.32501875297975436,0.0002916666666666667 +2,0.6568811237812042,0.8,0.7992831541218638,0.6214050202455499,0.0004583333333333333 +3,0.6208048164844513,0.8,0.7240143369175627,0.5963154145107828,0.0004996859161456965 +4,0.5614998787641525,0.8,0.7168458781362007,0.5632248226620223,0.0004982915790812436 +5,0.47156069427728653,0.775,0.7060931899641577,0.4964012453781352,0.0004957883115509159 +6,0.3866201490163803,0.8,0.7491039426523297,0.5352577903290147,0.0004921872937551814 +7,0.30171431228518486,0.775,0.8028673835125448,0.479156658068799,0.000487504608713676 +8,0.23811906203627586,0.775,0.8028673835125448,0.479156658068799,0.00048176117043453436 +9,0.16715852729976177,0.775,0.8781362007168458,0.504246263803566,0.000474982630507352 +10,0.1148978192359209,0.825,0.8458781362007168,0.663197652654773,0.00046719926353695914 +11,0.11084206961095333,0.825,0.8172043010752688,0.653639707612957,0.00045844583192968674 +12,0.08324608393013477,0.775,0.8387096774193549,0.5406067411965342,0.00044876143063602076 +13,0.08638920076191425,0.85,0.8888888888888888,0.7479091995221028,0.00043818931254306284 +14,0.050563013181090355,0.825,0.8709677419354839,0.6957139245790865,0.00042677669529663686 +15,0.02975934650748968,0.775,0.8136200716845878,0.5708939951268827,0.0004145745504158204 +16,0.04655684158205986,0.8,0.8172043010752688,0.6729920575381941,0.0004016373756417668 +17,0.019254492479376495,0.85,0.8602150537634409,0.7383512544802867,0.00038802295153756415 +18,0.04804687201976776,0.775,0.7634408602150538,0.4660144836363019,0.0003737920834262134 +19,0.02171836607158184,0.8,0.6845878136200717,0.5137524139849287,0.0003590083298192957 +20,0.0381931948941201,0.8,0.7132616487455197,0.5233103590267446,0.0003437377185492303 +21,0.018138925079256296,0.8,0.7275985663082437,0.5280893315476526,0.0003280484518729466 +22,0.008973158313892782,0.8,0.7168458781362008,0.5245051021569717,0.00031201060186404833 +23,0.03882891917601228,0.775,0.7491039426523297,0.4612355111153939,0.0002956957974539226 diff --git a/results/adam/resnet/metrics.json b/results/adam/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..4ec6c68c8b495dc3a70873934db1f825e27ba37c --- /dev/null +++ b/results/adam/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.825, + "balanced_accuracy": 0.7885304659498208, + "precision_macro": 0.7523510971786834, + "recall_macro": 0.7885304659498208, + "f1_macro": 0.7666666666666666, + "precision_weighted": 0.8411442006269592, + "recall_weighted": 0.825, + "f1_weighted": 0.8308333333333333, + "cohen_kappa": 0.5348837209302326, + "quadratic_weighted_kappa": 0.5348837209302326, + "mcc": 0.5396701942924549, + "auroc": 0.9193548387096774, + "auprc": 0.8159779901898516, + "sensitivity": 0.7222222222222222, + "specificity": 0.8548387096774194, + "precision_pos": 0.5909090909090909, + "f1_pos": 0.65, + "per_class": { + "0": { + "precision": 0.9137931034482759, + "recall": 0.8548387096774194, + "f1-score": 0.8833333333333333, + "support": 62.0 + }, + "1": { + "precision": 0.5909090909090909, + "recall": 0.7222222222222222, + "f1-score": 0.65, + "support": 18.0 + }, + "accuracy": 0.825, + "macro avg": { + "precision": 0.7523510971786834, + "recall": 0.7885304659498208, + "f1-score": 0.7666666666666666, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8411442006269592, + "recall": 0.825, + "f1-score": 0.8308333333333333, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/adam/resnet/pr.png b/results/adam/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..8a74f66a11fe7c08d8e317946f99df141dc55d80 --- /dev/null +++ b/results/adam/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ec7a7c5d0a5be51e3bfa9694b36044d5501dacaeba58264214569be37057003 +size 45285 diff --git a/results/adam/resnet/roc.png b/results/adam/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..1f8da3f7165d8e021a0b840e3b540def256e3c6f --- /dev/null +++ b/results/adam/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dab3374c01a27986b12a3ab22e2781b63f9b39363576750818dc3e25f3e7322f +size 57353 diff --git a/results/adam/resnet/test_pred.npz b/results/adam/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..019b4834e08409f6f6112a3b89e02d454a61d088 --- /dev/null +++ b/results/adam/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc71c3bfe92bb42e70800536b1635bb8a2bc982678c561ff87ab1d2a12a7fd3d +size 1790 diff --git a/results/adam/resnet/train.log b/results/adam/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9276f7d1c657998050853d33fc4bd67f8c0124e1 --- /dev/null +++ b/results/adam/resnet/train.log @@ -0,0 +1,128 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=280 val=40 test=80 classes=['0', '1'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6891 val_acc=0.4750 val_auc=0.3262 score=0.1426 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6811 val_acc=0.7250 val_auc=0.4624 score=0.3250 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6569 val_acc=0.8000 val_auc=0.7993 score=0.6214 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6208 val_acc=0.8000 val_auc=0.7240 score=0.5963 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.5615 val_acc=0.8000 val_auc=0.7168 score=0.5632 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.4716 val_acc=0.7750 val_auc=0.7061 score=0.4964 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.3866 val_acc=0.8000 val_auc=0.7491 score=0.5353 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.3017 val_acc=0.7750 val_auc=0.8029 score=0.4792 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.2381 val_acc=0.7750 val_auc=0.8029 score=0.4792 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1672 val_acc=0.7750 val_auc=0.8781 score=0.5042 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1149 val_acc=0.8250 val_auc=0.8459 score=0.6632 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1108 val_acc=0.8250 val_auc=0.8172 score=0.6536 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.0832 val_acc=0.7750 val_auc=0.8387 score=0.5406 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.0864 val_acc=0.8500 val_auc=0.8889 score=0.7479 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0506 val_acc=0.8250 val_auc=0.8710 score=0.6957 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0298 val_acc=0.7750 val_auc=0.8136 score=0.5709 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0466 val_acc=0.8000 val_auc=0.8172 score=0.6730 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0193 val_acc=0.8500 val_auc=0.8602 score=0.7384 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0480 val_acc=0.7750 val_auc=0.7634 score=0.4660 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0217 val_acc=0.8000 val_auc=0.6846 score=0.5138 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0382 val_acc=0.8000 val_auc=0.7133 score=0.5233 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0181 val_acc=0.8000 val_auc=0.7276 score=0.5281 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0090 val_acc=0.8000 val_auc=0.7168 score=0.5245 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0388 val_acc=0.7750 val_auc=0.7491 score=0.4612 +[resnet] early stop at ep23 (best ep13 score=0.7479) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=13 best_val_score=0.7479 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/adam/resnet acc=0.8250 auroc=0.9193548387096774 f1_macro=0.7667 qwk=0.5348837209302326 diff --git a/results/adam/retfound/confusion_matrix.png b/results/adam/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..733b83ea877bc153647ac339aa6088dd1906b567 --- /dev/null +++ b/results/adam/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7142b7b7972e6151b7e2e8d6ac8a164aec7a1fea97226fe5dd05f1952e67a98 +size 68153 diff --git a/results/adam/retfound/confusion_matrix_test.jpg b/results/adam/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..dca75d74bd4dbe952a45d142f6b5e9f8f16695af --- /dev/null +++ b/results/adam/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32bf90581835c5cfef3b2cfa602abce96759dac4a4a2fe9e2c20659164953ba4 +size 257614 diff --git a/results/adam/retfound/log.txt b/results/adam/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..642c68ea3f6fc466c7dacbf53b86f233ea9a17da --- /dev/null +++ b/results/adam/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.7343749999999997e-05, "train_loss": 0.6853790283203125, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.984375e-05, "train_loss": 0.600438117980957, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015234375, "train_loss": 0.5224623680114746, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021484375, "train_loss": 0.5393064022064209, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.00027734375000000003, "train_loss": 0.5149924755096436, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003398437499999999, "train_loss": 0.482379674911499, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040234375, "train_loss": 0.46404457092285156, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046484375000000003, "train_loss": 0.46213990449905396, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.00052734375, "train_loss": 0.41957637667655945, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.00058984375, "train_loss": 0.47334104776382446, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247369453830207, "train_loss": 0.4200931787490845, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006229352074360887, "train_loss": 0.4227827489376068, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006192226158713984, "train_loss": 0.3990233540534973, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006136220600505078, "train_loss": 0.39853784441947937, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006061680692624268, "train_loss": 0.37906284630298615, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005969065998390672, "train_loss": 0.3678862899541855, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005858947518191751, "train_loss": 0.3812587708234787, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005732004169076044, "train_loss": 0.37992827594280243, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005589018599003793, "train_loss": 0.29254330694675446, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005430872361562023, "train_loss": 0.33580098673701286, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005258540480893521, "train_loss": 0.3463967442512512, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005073085440348776, "train_loss": 0.31276238709688187, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004875650631922804, "train_loss": 0.31376905739307404, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00046674533068632887, "train_loss": 0.3495359867811203, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004449777070911855, "train_loss": 0.32266679406166077, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.00042239639704478405, "train_loss": 0.2978495806455612, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.00039914062183260795, "train_loss": 0.2612730134278536, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003753537610421703, "train_loss": 0.32327523455023766, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035118246858017804, "train_loss": 0.25920983776450157, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003267757685024298, "train_loss": 0.29856251925230026, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0003022841362309562, "train_loss": 0.27047010883688927, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002778585708230051, "train_loss": 0.2923644706606865, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002536496640116411, "train_loss": 0.28958597406744957, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022980667175763478, "train_loss": 0.2692524269223213, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00020647659403683159, "train_loss": 0.26890668272972107, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018380326853641835, "train_loss": 0.2811972163617611, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016192648384775085, "train_loss": 0.2884734384715557, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.000140981117623204, "train_loss": 0.2802345249801874, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012109630501059311, "train_loss": 0.28150836005806923, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010239464249204608, "train_loss": 0.2860140986740589, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.499143203592362e-05, "train_loss": 0.27556246146559715, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.899397022184248e-05, "train_loss": 0.2808167040348053, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.450088672158235e-05, "train_loss": 0.24387812614440918, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.1601536214361626e-05, "train_loss": 0.23795588314533234, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.0375447485526644e-05, "train_loss": 0.2574918810278177, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.0891833105144005e-05, "train_loss": 0.2669173013418913, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3209162709490563e-05, "train_loss": 0.2666049748659134, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.374802516302661e-06, "train_loss": 0.2570477966219187, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.4247232962929436e-06, "train_loss": 0.2644502613693476, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.3832786013874152e-06, "train_loss": 0.2804808057844639, "epoch": 49, "n_parameters": 303303682} diff --git a/results/adam/retfound/metrics.json b/results/adam/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..fbf6e32a18f31177fb01f58967bf87dd62ed25a4 --- /dev/null +++ b/results/adam/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.925, + "balanced_accuracy": 0.89247311827957, + "precision_macro": 0.89247311827957, + "recall_macro": 0.89247311827957, + "f1_macro": 0.89247311827957, + "precision_weighted": 0.925, + "recall_weighted": 0.925, + "f1_weighted": 0.925, + "cohen_kappa": 0.7849462365591398, + "quadratic_weighted_kappa": 0.7849462365591398, + "mcc": 0.7849462365591398, + "auroc": 0.9516129032258065, + "auprc": 0.9214129072681705, + "sensitivity": 0.8333333333333334, + "specificity": 0.9516129032258065, + "precision_pos": 0.8333333333333334, + "f1_pos": 0.8333333333333334, + "per_class": { + "0": { + "precision": 0.9516129032258065, + "recall": 0.9516129032258065, + "f1-score": 0.9516129032258065, + "support": 62.0 + }, + "1": { + "precision": 0.8333333333333334, + "recall": 0.8333333333333334, + "f1-score": 0.8333333333333334, + "support": 18.0 + }, + "accuracy": 0.925, + "macro avg": { + "precision": 0.89247311827957, + "recall": 0.89247311827957, + "f1-score": 0.89247311827957, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.925, + "recall": 0.925, + "f1-score": 0.925, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/adam/retfound/metrics_test.csv b/results/adam/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..1da15a9eb6ee759ca1bac98d71e7cb6f2b9aaa58 --- /dev/null +++ b/results/adam/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.24544517199198404,0.925,0.89247311827957,0.9516129032258065,0.075,0.810989010989011,0.89247311827957,0.89247311827957,0.9511565422863844,0.7849462365591398 diff --git a/results/adam/retfound/metrics_val.csv b/results/adam/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..1504c8554fabdb27101c5f803242490b41a375e0 --- /dev/null +++ b/results/adam/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6949386596679688,0.775,0.43661971830985913,0.7526881720430108,0.225,0.3875,0.3875,0.5,0.702140345769078,0.0 +0.7357521057128906,0.775,0.43661971830985913,0.7535842293906809,0.225,0.3875,0.3875,0.5,0.7023472297814682,0.0 +0.8536128997802734,0.775,0.43661971830985913,0.7974910394265233,0.225,0.3875,0.3875,0.5,0.7664684375686055,0.0 +0.8364524841308594,0.775,0.43661971830985913,0.8279569892473118,0.225,0.3875,0.3875,0.5,0.8171861361036572,0.0 +0.726959228515625,0.775,0.43661971830985913,0.85752688172043,0.225,0.3875,0.3875,0.5,0.8267116528482032,0.0 +0.6651697158813477,0.775,0.43661971830985913,0.870967741935484,0.225,0.3875,0.3875,0.5,0.8399610377402678,0.0 +0.5009684562683105,0.825,0.7584124245038826,0.8611111111111112,0.175,0.6278280542986425,0.75,0.7688172043010753,0.7919264893748371,0.5172413793103448 +0.6560912132263184,0.85,0.7058823529411764,0.870967741935484,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.8455413148826051,0.43661971830985913 +0.4889563322067261,0.85,0.7849462365591398,0.881720430107527,0.15,0.6617647058823529,0.7849462365591398,0.7849462365591398,0.8786670749538394,0.5698924731182795 +0.5198326110839844,0.9,0.8268398268398269,0.8530465949820789,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8587794130254447,0.6595744680851063 +0.7482777833938599,0.8,0.5428571428571428,0.8996415770609318,0.2,0.45299145299145294,0.8974358974358974,0.5555555555555556,0.8843861452756653,0.16230366492146608 +0.45795178413391113,0.9,0.8268398268398269,0.9068100358422939,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8862751464020953,0.6595744680851063 +0.5212460160255432,0.9,0.8268398268398269,0.8996415770609318,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.881716849965241,0.6595744680851063 +0.5566548109054565,0.875,0.7948717948717949,0.8924731182795699,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8804819825795989,0.5934959349593496 +0.6318426728248596,0.9,0.8268398268398269,0.8853046594982079,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8765545253605149,0.6595744680851063 +0.6386967301368713,0.875,0.7703788748564868,0.8996415770609318,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.8778435406015122,0.5535714285714286 +0.5159331560134888,0.9,0.8268398268398269,0.9175627240143369,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.906900466227048,0.6595744680851063 +0.39144355058670044,0.95,0.921875,0.9283154121863799,0.05,0.8585858585858586,0.9696969696969697,0.8888888888888888,0.9285526168442524,0.8443579766536965 +0.5359691083431244,0.875,0.7703788748564868,0.9292114695340502,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9285066651853976,0.5535714285714286 +0.4337925612926483,0.95,0.921875,0.9283154121863799,0.05,0.8585858585858586,0.9696969696969697,0.8888888888888888,0.9284603359145964,0.8443579766536965 +0.6121203303337097,0.875,0.7703788748564868,0.9283154121863799,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9284603359145964,0.5535714285714286 +0.42638447880744934,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145964,0.7560975609756098 +0.3997582793235779,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145963,0.7560975609756098 +0.4086921811103821,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145964,0.7560975609756098 +0.41143083572387695,0.9,0.8268398268398269,0.9283154121863799,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9284603359145964,0.6595744680851063 +0.5094860941171646,0.9,0.8268398268398269,0.931899641577061,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9305555924889939,0.6595744680851063 +0.5649861097335815,0.9,0.8268398268398269,0.9292114695340502,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9233688544403764,0.6595744680851063 +0.43275587260723114,0.9,0.8268398268398269,0.9390681003584229,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9251959935525449,0.6595744680851063 +0.21898877620697021,0.9,0.8566308243727598,0.9498207885304659,0.1,0.7575757575757576,0.8566308243727598,0.8566308243727598,0.9370107576036646,0.7132616487455197 +0.28504690527915955,0.925,0.8879551820728291,0.9498207885304659,0.075,0.8045454545454545,0.90625,0.8727598566308243,0.9346306558217193,0.7761194029850746 +0.3365369886159897,0.9,0.84375,0.9498207885304659,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.940824081249271,0.688715953307393 +0.3427543491125107,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9369379825912034,0.7560975609756098 +0.35773538053035736,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9357159201324631,0.7560975609756098 +0.42521847784519196,0.9,0.8268398268398269,0.946236559139785,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9409427467992981,0.6595744680851063 +0.4029083847999573,0.925,0.8769230769230769,0.9390681003584229,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9352552443376021,0.7560975609756098 +0.3561532497406006,0.925,0.8769230769230769,0.9390681003584229,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9352552443376021,0.7560975609756098 +0.2910751849412918,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9409427467992981,0.7560975609756098 +0.3142062872648239,0.925,0.8769230769230769,0.9498207885304659,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9431059973754814,0.7560975609756098 +0.2977108359336853,0.925,0.8769230769230769,0.953405017921147,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9454120273143385,0.7560975609756098 +0.2965516448020935,0.925,0.8769230769230769,0.953405017921147,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9454120273143386,0.7560975609756098 +0.3348654955625534,0.925,0.8769230769230769,0.9498207885304659,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9424153227047307,0.7560975609756098 +0.3686094284057617,0.9,0.8268398268398269,0.953405017921147,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9454425747239183,0.6595744680851063 +0.4043458551168442,0.9,0.8268398268398269,0.953405017921147,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9454120273143387,0.6595744680851063 +0.416122242808342,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492624,0.6595744680851063 +0.4057278037071228,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492624,0.6595744680851063 +0.38761037588119507,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492622,0.6595744680851063 +0.3810284584760666,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492622,0.6595744680851063 +0.3793119490146637,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 +0.3780297338962555,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 +0.37748883664608,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 diff --git a/results/adam/retfound/pr.png b/results/adam/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..82d2b670c777139265363636bc117fa6727529db --- /dev/null +++ b/results/adam/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9ef8742fd307bb9273e0d952fdb9b2e45d58356db6b2c3f3d2bdf053b1a0e98 +size 41633 diff --git a/results/adam/retfound/roc.png b/results/adam/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..94d89a59442837c9bfb01a10c3cdf55dc732b2f9 --- /dev/null +++ b/results/adam/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9144441574909e862a8170b3e1e6474ce4e034938d1af53ad1199b41c2d6f5a0 +size 57049 diff --git a/results/adam/retfound/test_pred.npz b/results/adam/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..2ba49d6fcec500556d3182bc99e421c3c9415e69 --- /dev/null +++ b/results/adam/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a9f339d43b3b03ac4ee519789551950bc677e6e276a250d13a0f02784aa9777 +size 1470 diff --git a/results/adam/retfound/train.log b/results/adam/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..3d6a8373e00d2b6cb475cd3007a83d58034791fe --- /dev/null +++ b/results/adam/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:52:22.761018092 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:52:24.578155] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:52:24.578459] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/AMD/adamdataset', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/adam', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:52:32.164895] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:36.886727] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:38.199921] Sampler_train = +[13:52:38.245683] len of train_set: 256 +[13:52:38.780666] [Adaptation] Full fine-tuning: training all parameters. +[13:52:38.781731] number of trainable params (M): 303.30 +[13:52:38.781809] base lr: 5.00e-03 +[13:52:38.781866] actual lr: 6.25e-04 +[13:52:38.781915] accumulate grad iterations: 1 +[13:52:38.781968] effective batch size: 32 +[13:52:38.785609] criterion = CrossEntropyLoss() +[13:52:38.785695] Start training for 50 epochs +[13:52:38.787747] log_dir: ./output_logs/retfound +[13:52:42.314240] Epoch: [0] [0/8] eta: 0:00:28 lr: 0.000000 loss: 0.6928 (0.6928) time: 3.5255 data: 2.5943 max mem: 7340 +[13:52:43.413951] Epoch: [0] [7/8] eta: 0:00:00 lr: 0.000055 loss: 0.6831 (0.6854) time: 0.5781 data: 0.3244 max mem: 9671 +[13:52:43.512082] Epoch: [0] Total time: 0:00:04 (0.5905 s / it) +[13:52:43.521651] Averaged stats: lr: 0.000055 loss: 0.6831 (0.6854) +[13:52:46.014370] val: [0/2] eta: 0:00:04 loss: 0.6396 (0.6396) time: 2.4781 data: 2.4249 max mem: 9671 +[13:52:46.132148] val: [1/2] eta: 0:00:01 loss: 0.6396 (0.6949) time: 1.2977 data: 1.2125 max mem: 9671 +[13:52:46.239339] val: Total time: 0:00:02 (1.3518 s / it) +[13:52:46.251957] val loss: 0.6949386596679688 +[13:52:46.252128] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7527, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7021, Kappa: 0.0000, Score: 0.3964 +[13:52:47.965179] Best epoch = 0, Best score = 0.3964 +[13:52:48.038141] log_dir: ./output_logs/retfound +[13:52:50.410461] Epoch: [1] [0/8] eta: 0:00:18 lr: 0.000063 loss: 0.6628 (0.6628) time: 2.3713 data: 2.2170 max mem: 9671 +[13:52:51.430322] Epoch: [1] [7/8] eta: 0:00:00 lr: 0.000117 loss: 0.6032 (0.6004) time: 0.4238 data: 0.2799 max mem: 9671 +[13:52:51.535674] Epoch: [1] Total time: 0:00:03 (0.4372 s / it) +[13:52:51.545072] Averaged stats: lr: 0.000117 loss: 0.6032 (0.6004) +[13:52:54.157044] val: [0/2] eta: 0:00:05 loss: 0.4481 (0.4481) time: 2.5953 data: 2.5589 max mem: 9671 +[13:52:54.172452] val: [1/2] eta: 0:00:01 loss: 0.4481 (0.7358) time: 1.3051 data: 1.2795 max mem: 9671 +[13:52:54.277531] val: Total time: 0:00:02 (1.3583 s / it) +[13:52:54.286713] val loss: 0.7357521057128906 +[13:52:54.286912] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7536, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7023, Kappa: 0.0000, Score: 0.3967 +[13:52:55.923888] Best epoch = 1, Best score = 0.3967 +[13:52:55.990198] log_dir: ./output_logs/retfound +[13:52:58.626576] Epoch: [2] [0/8] eta: 0:00:21 lr: 0.000125 loss: 0.6224 (0.6224) time: 2.6348 data: 2.4891 max mem: 9671 +[13:52:59.620834] Epoch: [2] [7/8] eta: 0:00:00 lr: 0.000180 loss: 0.4999 (0.5225) time: 0.4535 data: 0.3113 max mem: 9671 +[13:52:59.734195] Epoch: [2] Total time: 0:00:03 (0.4680 s / it) +[13:52:59.743035] Averaged stats: lr: 0.000180 loss: 0.4999 (0.5225) +[13:53:02.391033] val: [0/2] eta: 0:00:05 loss: 0.2783 (0.2783) time: 2.6326 data: 2.5969 max mem: 9671 +[13:53:02.406238] val: [1/2] eta: 0:00:01 loss: 0.2783 (0.8536) time: 1.3236 data: 1.2985 max mem: 9671 +[13:53:02.512965] val: Total time: 0:00:02 (1.3777 s / it) +[13:53:02.522133] val loss: 0.8536128997802734 +[13:53:02.522307] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7975, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7665, Kappa: 0.0000, Score: 0.4114 +[13:53:04.204960] Best epoch = 2, Best score = 0.4114 +[13:53:04.282572] log_dir: ./output_logs/retfound +[13:53:06.786738] Epoch: [3] [0/8] eta: 0:00:20 lr: 0.000188 loss: 0.6936 (0.6936) time: 2.5028 data: 2.3583 max mem: 9671 +[13:53:07.780820] Epoch: [3] [7/8] eta: 0:00:00 lr: 0.000242 loss: 0.5377 (0.5393) time: 0.4370 data: 0.2949 max mem: 9671 +[13:53:07.899014] Epoch: [3] Total time: 0:00:03 (0.4520 s / it) +[13:53:07.908946] Averaged stats: lr: 0.000242 loss: 0.5377 (0.5393) +[13:53:10.532422] val: [0/2] eta: 0:00:05 loss: 0.2665 (0.2665) time: 2.6082 data: 2.5740 max mem: 9671 +[13:53:10.548060] val: [1/2] eta: 0:00:01 loss: 0.2665 (0.8365) time: 1.3116 data: 1.2871 max mem: 9671 +[13:53:10.653453] val: Total time: 0:00:02 (1.3650 s / it) +[13:53:10.663550] val loss: 0.8364524841308594 +[13:53:10.663729] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8280, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8172, Kappa: 0.0000, Score: 0.4215 +[13:53:12.317371] Best epoch = 3, Best score = 0.4215 +[13:53:12.384755] log_dir: ./output_logs/retfound +[13:53:14.974660] Epoch: [4] [0/8] eta: 0:00:20 lr: 0.000250 loss: 0.6812 (0.6812) time: 2.5889 data: 2.4463 max mem: 9671 +[13:53:15.968159] Epoch: [4] [7/8] eta: 0:00:00 lr: 0.000305 loss: 0.4633 (0.5150) time: 0.4477 data: 0.3059 max mem: 9671 +[13:53:16.081679] Epoch: [4] Total time: 0:00:03 (0.4621 s / it) +[13:53:16.083453] Averaged stats: lr: 0.000305 loss: 0.4633 (0.5150) +[13:53:18.543742] val: [0/2] eta: 0:00:04 loss: 0.3205 (0.3205) time: 2.4449 data: 2.4113 max mem: 9671 +[13:53:18.559316] val: [1/2] eta: 0:00:01 loss: 0.3205 (0.7270) time: 1.2300 data: 1.2057 max mem: 9671 +[13:53:18.681860] val: Total time: 0:00:02 (1.2918 s / it) +[13:53:18.691322] val loss: 0.726959228515625 +[13:53:18.691552] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8575, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8267, Kappa: 0.0000, Score: 0.4314 +[13:53:20.400076] Best epoch = 4, Best score = 0.4314 +[13:53:20.474867] log_dir: ./output_logs/retfound +[13:53:22.996871] Epoch: [5] [0/8] eta: 0:00:20 lr: 0.000313 loss: 0.4578 (0.4578) time: 2.5209 data: 2.3763 max mem: 9671 +[13:53:23.989126] Epoch: [5] [7/8] eta: 0:00:00 lr: 0.000367 loss: 0.4578 (0.4824) time: 0.4391 data: 0.2971 max mem: 9671 +[13:53:24.104994] Epoch: [5] Total time: 0:00:03 (0.4537 s / it) +[13:53:24.113508] Averaged stats: lr: 0.000367 loss: 0.4578 (0.4824) +[13:53:26.614435] val: [0/2] eta: 0:00:04 loss: 0.2633 (0.2633) time: 2.4870 data: 2.4526 max mem: 9671 +[13:53:26.630096] val: [1/2] eta: 0:00:01 loss: 0.2633 (0.6652) time: 1.2511 data: 1.2264 max mem: 9671 +[13:53:26.737026] val: Total time: 0:00:02 (1.3052 s / it) +[13:53:26.747038] val loss: 0.6651697158813477 +[13:53:26.747262] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8710, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8400, Kappa: 0.0000, Score: 0.4359 +[13:53:28.383271] Best epoch = 5, Best score = 0.4359 +[13:53:28.445868] log_dir: ./output_logs/retfound +[13:53:30.973733] Epoch: [6] [0/8] eta: 0:00:20 lr: 0.000375 loss: 0.3839 (0.3839) time: 2.5270 data: 2.3822 max mem: 9671 +[13:53:31.960696] Epoch: [6] [7/8] eta: 0:00:00 lr: 0.000430 loss: 0.4587 (0.4640) time: 0.4391 data: 0.2979 max mem: 9671 +[13:53:32.064165] Epoch: [6] Total time: 0:00:03 (0.4523 s / it) +[13:53:32.072460] Averaged stats: lr: 0.000430 loss: 0.4587 (0.4640) +[13:53:34.624255] val: [0/2] eta: 0:00:05 loss: 0.3132 (0.3132) time: 2.5418 data: 2.5052 max mem: 9671 +[13:53:34.639707] val: [1/2] eta: 0:00:01 loss: 0.3132 (0.5010) time: 1.2783 data: 1.2527 max mem: 9671 +[13:53:34.745214] val: Total time: 0:00:02 (1.3317 s / it) +[13:53:34.754126] val loss: 0.5009684562683105 +[13:53:34.754313] Accuracy: 0.8250, F1 Score: 0.7584, ROC AUC: 0.8611, Hamming Loss: 0.1750, + Jaccard Score: 0.6278, Precision: 0.7500, Recall: 0.7688, + Average Precision: 0.7919, Kappa: 0.5172, Score: 0.7123 +[13:53:36.493464] Best epoch = 6, Best score = 0.7123 +[13:53:36.564071] log_dir: ./output_logs/retfound +[13:53:39.042952] Epoch: [7] [0/8] eta: 0:00:19 lr: 0.000438 loss: 0.5242 (0.5242) time: 2.4778 data: 2.3262 max mem: 9671 +[13:53:40.031939] Epoch: [7] [7/8] eta: 0:00:00 lr: 0.000492 loss: 0.4543 (0.4621) time: 0.4332 data: 0.2909 max mem: 9671 +[13:53:40.137973] Epoch: [7] Total time: 0:00:03 (0.4467 s / it) +[13:53:40.147565] Averaged stats: lr: 0.000492 loss: 0.4543 (0.4621) +[13:53:42.640030] val: [0/2] eta: 0:00:04 loss: 0.1982 (0.1982) time: 2.4775 data: 2.4483 max mem: 9671 +[13:53:42.655557] val: [1/2] eta: 0:00:01 loss: 0.1982 (0.6561) time: 1.2462 data: 1.2242 max mem: 9671 +[13:53:42.758934] val: Total time: 0:00:02 (1.2986 s / it) +[13:53:42.769226] val loss: 0.6560912132263184 +[13:53:42.769409] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.8710, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.8455, Kappa: 0.4366, Score: 0.6712 +[13:53:42.814484] Best epoch = 6, Best score = 0.7123 +[13:53:43.081773] log_dir: ./output_logs/retfound +[13:53:45.552484] Epoch: [8] [0/8] eta: 0:00:19 lr: 0.000500 loss: 0.5544 (0.5544) time: 2.4697 data: 2.3249 max mem: 9671 +[13:53:46.541272] Epoch: [8] [7/8] eta: 0:00:00 lr: 0.000555 loss: 0.4068 (0.4196) time: 0.4322 data: 0.2907 max mem: 9671 +[13:53:46.655118] Epoch: [8] Total time: 0:00:03 (0.4466 s / it) +[13:53:46.662978] Averaged stats: lr: 0.000555 loss: 0.4068 (0.4196) +[13:53:49.242886] val: [0/2] eta: 0:00:05 loss: 0.2713 (0.2713) time: 2.5684 data: 2.5325 max mem: 9671 +[13:53:49.258470] val: [1/2] eta: 0:00:01 loss: 0.2713 (0.4890) time: 1.2917 data: 1.2663 max mem: 9671 +[13:53:49.372089] val: Total time: 0:00:02 (1.3492 s / it) +[13:53:49.380923] val loss: 0.4889563322067261 +[13:53:49.381102] Accuracy: 0.8500, F1 Score: 0.7849, ROC AUC: 0.8817, Hamming Loss: 0.1500, + Jaccard Score: 0.6618, Precision: 0.7849, Recall: 0.7849, + Average Precision: 0.8787, Kappa: 0.5699, Score: 0.7455 +[13:53:51.051394] Best epoch = 8, Best score = 0.7455 +[13:53:51.121339] log_dir: ./output_logs/retfound +[13:53:53.520589] Epoch: [9] [0/8] eta: 0:00:19 lr: 0.000562 loss: 0.5163 (0.5163) time: 2.3982 data: 2.2547 max mem: 9671 +[13:53:54.566192] Epoch: [9] [7/8] eta: 0:00:00 lr: 0.000617 loss: 0.4599 (0.4733) time: 0.4304 data: 0.2886 max mem: 9671 +[13:53:54.688760] Epoch: [9] Total time: 0:00:03 (0.4459 s / it) +[13:53:54.696927] Averaged stats: lr: 0.000617 loss: 0.4599 (0.4733) +[13:53:57.125472] val: [0/2] eta: 0:00:04 loss: 0.2701 (0.2701) time: 2.4178 data: 2.3866 max mem: 9671 +[13:53:57.140932] val: [1/2] eta: 0:00:01 loss: 0.2701 (0.5198) time: 1.2163 data: 1.1934 max mem: 9671 +[13:53:57.253153] val: Total time: 0:00:02 (1.2731 s / it) +[13:53:57.262050] val loss: 0.5198326110839844 +[13:53:57.262237] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8530, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8588, Kappa: 0.6596, Score: 0.7798 +[13:53:58.950254] Best epoch = 9, Best score = 0.7798 +[13:53:59.023367] log_dir: ./output_logs/retfound +[13:54:01.410360] Epoch: [10] [0/8] eta: 0:00:19 lr: 0.000625 loss: 0.4179 (0.4179) time: 2.3858 data: 2.2390 max mem: 9671 +[13:54:02.502642] Epoch: [10] [7/8] eta: 0:00:00 lr: 0.000624 loss: 0.4179 (0.4201) time: 0.4347 data: 0.2928 max mem: 9671 +[13:54:02.623844] Epoch: [10] Total time: 0:00:03 (0.4500 s / it) +[13:54:02.633329] Averaged stats: lr: 0.000624 loss: 0.4179 (0.4201) +[13:54:05.117772] val: [0/2] eta: 0:00:04 loss: 0.1527 (0.1527) time: 2.4700 data: 2.4356 max mem: 9671 +[13:54:05.133105] val: [1/2] eta: 0:00:01 loss: 0.1527 (0.7483) time: 1.2424 data: 1.2179 max mem: 9671 +[13:54:05.236326] val: Total time: 0:00:02 (1.2948 s / it) +[13:54:05.245930] val loss: 0.7482777833938599 +[13:54:05.246138] Accuracy: 0.8000, F1 Score: 0.5429, ROC AUC: 0.8996, Hamming Loss: 0.2000, + Jaccard Score: 0.4530, Precision: 0.8974, Recall: 0.5556, + Average Precision: 0.8844, Kappa: 0.1623, Score: 0.5349 +[13:54:05.291221] Best epoch = 9, Best score = 0.7798 +[13:54:05.569993] log_dir: ./output_logs/retfound +[13:54:08.083871] Epoch: [11] [0/8] eta: 0:00:20 lr: 0.000624 loss: 0.3514 (0.3514) time: 2.5129 data: 2.3671 max mem: 9671 +[13:54:09.077670] Epoch: [11] [7/8] eta: 0:00:00 lr: 0.000622 loss: 0.3698 (0.4228) time: 0.4382 data: 0.2960 max mem: 9671 +[13:54:09.192574] Epoch: [11] Total time: 0:00:03 (0.4528 s / it) +[13:54:09.202005] Averaged stats: lr: 0.000622 loss: 0.3698 (0.4228) +[13:54:11.770500] val: [0/2] eta: 0:00:05 loss: 0.2029 (0.2029) time: 2.5521 data: 2.5161 max mem: 9671 +[13:54:11.786192] val: [1/2] eta: 0:00:01 loss: 0.2029 (0.4580) time: 1.2836 data: 1.2581 max mem: 9671 +[13:54:11.897793] val: Total time: 0:00:02 (1.3401 s / it) +[13:54:11.906585] val loss: 0.45795178413391113 +[13:54:11.906772] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9068, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8863, Kappa: 0.6596, Score: 0.7977 +[13:54:13.615563] Best epoch = 11, Best score = 0.7977 +[13:54:13.684221] log_dir: ./output_logs/retfound +[13:54:16.320430] Epoch: [12] [0/8] eta: 0:00:21 lr: 0.000621 loss: 0.3098 (0.3098) time: 2.6353 data: 2.4874 max mem: 9671 +[13:54:17.312384] Epoch: [12] [7/8] eta: 0:00:00 lr: 0.000617 loss: 0.3706 (0.3990) time: 0.4533 data: 0.3110 max mem: 9671 +[13:54:17.456694] Epoch: [12] Total time: 0:00:03 (0.4715 s / it) +[13:54:17.465963] Averaged stats: lr: 0.000617 loss: 0.3706 (0.3990) +[13:54:20.043599] val: [0/2] eta: 0:00:05 loss: 0.1898 (0.1898) time: 2.5621 data: 2.5275 max mem: 9671 +[13:54:20.058777] val: [1/2] eta: 0:00:01 loss: 0.1898 (0.5212) time: 1.2883 data: 1.2638 max mem: 9671 +[13:54:20.170252] val: Total time: 0:00:02 (1.3449 s / it) +[13:54:20.179390] val loss: 0.5212460160255432 +[13:54:20.179612] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8996, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8817, Kappa: 0.6596, Score: 0.7954 +[13:54:20.221615] Best epoch = 11, Best score = 0.7977 +[13:54:20.484237] log_dir: ./output_logs/retfound +[13:54:23.015330] Epoch: [13] [0/8] eta: 0:00:20 lr: 0.000616 loss: 0.5697 (0.5697) time: 2.5301 data: 2.3832 max mem: 9671 +[13:54:24.015100] Epoch: [13] [7/8] eta: 0:00:00 lr: 0.000611 loss: 0.3551 (0.3985) time: 0.4411 data: 0.2981 max mem: 9671 +[13:54:24.133416] Epoch: [13] Total time: 0:00:03 (0.4561 s / it) +[13:54:24.141941] Averaged stats: lr: 0.000611 loss: 0.3551 (0.3985) +[13:54:26.708491] val: [0/2] eta: 0:00:05 loss: 0.1778 (0.1778) time: 2.5494 data: 2.5153 max mem: 9671 +[13:54:26.723630] val: [1/2] eta: 0:00:01 loss: 0.1778 (0.5567) time: 1.2820 data: 1.2577 max mem: 9671 +[13:54:26.830847] val: Total time: 0:00:02 (1.3364 s / it) +[13:54:26.839850] val loss: 0.5566548109054565 +[13:54:26.840024] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.8925, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8805, Kappa: 0.5935, Score: 0.7603 +[13:54:26.888237] Best epoch = 11, Best score = 0.7977 +[13:54:27.147032] log_dir: ./output_logs/retfound +[13:54:29.720747] Epoch: [14] [0/8] eta: 0:00:20 lr: 0.000610 loss: 0.4109 (0.4109) time: 2.5726 data: 2.4270 max mem: 9671 +[13:54:30.716226] Epoch: [14] [7/8] eta: 0:00:00 lr: 0.000602 loss: 0.3012 (0.3791) time: 0.4458 data: 0.3035 max mem: 9671 +[13:54:30.842805] Epoch: [14] Total time: 0:00:03 (0.4620 s / it) +[13:54:30.851739] Averaged stats: lr: 0.000602 loss: 0.3012 (0.3791) +[13:54:33.413005] val: [0/2] eta: 0:00:05 loss: 0.1636 (0.1636) time: 2.5527 data: 2.5190 max mem: 9671 +[13:54:33.428677] val: [1/2] eta: 0:00:01 loss: 0.1636 (0.6318) time: 1.2839 data: 1.2595 max mem: 9671 +[13:54:33.535470] val: Total time: 0:00:02 (1.3379 s / it) +[13:54:33.547601] val loss: 0.6318426728248596 +[13:54:33.547804] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8853, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8766, Kappa: 0.6596, Score: 0.7906 +[13:54:33.599757] Best epoch = 11, Best score = 0.7977 +[13:54:33.838256] log_dir: ./output_logs/retfound +[13:54:36.458393] Epoch: [15] [0/8] eta: 0:00:20 lr: 0.000601 loss: 0.4683 (0.4683) time: 2.6190 data: 2.4740 max mem: 9671 +[13:54:37.452191] Epoch: [15] [7/8] eta: 0:00:00 lr: 0.000592 loss: 0.3347 (0.3679) time: 0.4515 data: 0.3093 max mem: 9671 +[13:54:37.570649] Epoch: [15] Total time: 0:00:03 (0.4665 s / it) +[13:54:37.580224] Averaged stats: lr: 0.000592 loss: 0.3347 (0.3679) +[13:54:40.163649] val: [0/2] eta: 0:00:05 loss: 0.1511 (0.1511) time: 2.5753 data: 2.5403 max mem: 9671 +[13:54:40.179134] val: [1/2] eta: 0:00:01 loss: 0.1511 (0.6387) time: 1.2951 data: 1.2702 max mem: 9671 +[13:54:40.285602] val: Total time: 0:00:02 (1.3490 s / it) +[13:54:40.294439] val loss: 0.6386967301368713 +[13:54:40.294604] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.8996, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.8778, Kappa: 0.5536, Score: 0.7412 +[13:54:40.340349] Best epoch = 11, Best score = 0.7977 +[13:54:40.618860] log_dir: ./output_logs/retfound +[13:54:43.157979] Epoch: [16] [0/8] eta: 0:00:20 lr: 0.000591 loss: 0.3331 (0.3331) time: 2.5381 data: 2.4690 max mem: 9671 +[13:54:44.021078] Epoch: [16] [7/8] eta: 0:00:00 lr: 0.000581 loss: 0.2960 (0.3813) time: 0.4251 data: 0.3087 max mem: 9671 +[13:54:44.133910] Epoch: [16] Total time: 0:00:03 (0.4394 s / it) +[13:54:44.143191] Averaged stats: lr: 0.000581 loss: 0.2960 (0.3813) +[13:54:46.605231] val: [0/2] eta: 0:00:04 loss: 0.1563 (0.1563) time: 2.4550 data: 2.4191 max mem: 9671 +[13:54:46.620838] val: [1/2] eta: 0:00:01 loss: 0.1563 (0.5159) time: 1.2350 data: 1.2096 max mem: 9671 +[13:54:46.728162] val: Total time: 0:00:02 (1.2893 s / it) +[13:54:46.736911] val loss: 0.5159331560134888 +[13:54:46.737115] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9176, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9069, Kappa: 0.6596, Score: 0.8013 +[13:54:48.494622] Best epoch = 16, Best score = 0.8013 +[13:54:48.571901] log_dir: ./output_logs/retfound +[13:54:50.996269] Epoch: [17] [0/8] eta: 0:00:19 lr: 0.000579 loss: 0.4858 (0.4858) time: 2.4234 data: 2.2808 max mem: 9671 +[13:54:52.031801] Epoch: [17] [7/8] eta: 0:00:00 lr: 0.000567 loss: 0.3880 (0.3799) time: 0.4322 data: 0.2906 max mem: 9671 +[13:54:52.150980] Epoch: [17] Total time: 0:00:03 (0.4474 s / it) +[13:54:52.160055] Averaged stats: lr: 0.000567 loss: 0.3880 (0.3799) +[13:54:54.695276] val: [0/2] eta: 0:00:05 loss: 0.1928 (0.1928) time: 2.5192 data: 2.4843 max mem: 9671 +[13:54:54.708310] val: [1/2] eta: 0:00:01 loss: 0.1928 (0.3914) time: 1.2658 data: 1.2422 max mem: 9671 +[13:54:54.822279] val: Total time: 0:00:02 (1.3235 s / it) +[13:54:54.831482] val loss: 0.39144355058670044 +[13:54:54.831681] Accuracy: 0.9500, F1 Score: 0.9219, ROC AUC: 0.9283, Hamming Loss: 0.0500, + Jaccard Score: 0.8586, Precision: 0.9697, Recall: 0.8889, + Average Precision: 0.9286, Kappa: 0.8444, Score: 0.8982 +[13:54:56.620098] Best epoch = 17, Best score = 0.8982 +[13:54:56.694716] log_dir: ./output_logs/retfound +[13:54:59.189733] Epoch: [18] [0/8] eta: 0:00:19 lr: 0.000565 loss: 0.2349 (0.2349) time: 2.4940 data: 2.3495 max mem: 9671 +[13:55:00.181047] Epoch: [18] [7/8] eta: 0:00:00 lr: 0.000552 loss: 0.2442 (0.2925) time: 0.4355 data: 0.2938 max mem: 9671 +[13:55:00.290467] Epoch: [18] Total time: 0:00:03 (0.4494 s / it) +[13:55:00.298980] Averaged stats: lr: 0.000552 loss: 0.2442 (0.2925) +[13:55:03.000826] val: [0/2] eta: 0:00:05 loss: 0.1443 (0.1443) time: 2.6931 data: 2.6570 max mem: 9671 +[13:55:03.016392] val: [1/2] eta: 0:00:01 loss: 0.1443 (0.5360) time: 1.3540 data: 1.3286 max mem: 9671 +[13:55:03.127080] val: Total time: 0:00:02 (1.4101 s / it) +[13:55:03.136536] val loss: 0.5359691083431244 +[13:55:03.136716] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9292, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9285, Kappa: 0.5536, Score: 0.7511 +[13:55:03.173665] Best epoch = 17, Best score = 0.8982 +[13:55:03.482112] log_dir: ./output_logs/retfound +[13:55:06.095371] Epoch: [19] [0/8] eta: 0:00:20 lr: 0.000550 loss: 0.4059 (0.4059) time: 2.6123 data: 2.4664 max mem: 9671 +[13:55:07.091937] Epoch: [19] [7/8] eta: 0:00:00 lr: 0.000536 loss: 0.2734 (0.3358) time: 0.4510 data: 0.3084 max mem: 9671 +[13:55:07.199220] Epoch: [19] Total time: 0:00:03 (0.4646 s / it) +[13:55:07.208456] Averaged stats: lr: 0.000536 loss: 0.2734 (0.3358) +[13:55:09.758064] val: [0/2] eta: 0:00:05 loss: 0.1677 (0.1677) time: 2.5429 data: 2.5071 max mem: 9671 +[13:55:09.773819] val: [1/2] eta: 0:00:01 loss: 0.1677 (0.4338) time: 1.2791 data: 1.2536 max mem: 9671 +[13:55:09.884624] val: Total time: 0:00:02 (1.3351 s / it) +[13:55:09.893282] val loss: 0.4337925612926483 +[13:55:09.893461] Accuracy: 0.9500, F1 Score: 0.9219, ROC AUC: 0.9283, Hamming Loss: 0.0500, + Jaccard Score: 0.8586, Precision: 0.9697, Recall: 0.8889, + Average Precision: 0.9285, Kappa: 0.8444, Score: 0.8982 +[13:55:09.934690] Best epoch = 17, Best score = 0.8982 +[13:55:10.199747] log_dir: ./output_logs/retfound +[13:55:12.743926] Epoch: [20] [0/8] eta: 0:00:20 lr: 0.000534 loss: 0.3245 (0.3245) time: 2.5432 data: 2.3978 max mem: 9671 +[13:55:13.733418] Epoch: [20] [7/8] eta: 0:00:00 lr: 0.000518 loss: 0.3245 (0.3464) time: 0.4415 data: 0.2998 max mem: 9671 +[13:55:13.845317] Epoch: [20] Total time: 0:00:03 (0.4557 s / it) +[13:55:13.853666] Averaged stats: lr: 0.000518 loss: 0.3245 (0.3464) +[13:55:16.388226] val: [0/2] eta: 0:00:05 loss: 0.1523 (0.1523) time: 2.5229 data: 2.4889 max mem: 9671 +[13:55:16.403753] val: [1/2] eta: 0:00:01 loss: 0.1523 (0.6121) time: 1.2690 data: 1.2445 max mem: 9671 +[13:55:16.509932] val: Total time: 0:00:02 (1.3227 s / it) +[13:55:16.518766] val loss: 0.6121203303337097 +[13:55:16.518938] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9283, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9285, Kappa: 0.5536, Score: 0.7508 +[13:55:16.558346] Best epoch = 17, Best score = 0.8982 +[13:55:16.810840] log_dir: ./output_logs/retfound +[13:55:19.504352] Epoch: [21] [0/8] eta: 0:00:21 lr: 0.000516 loss: 0.2419 (0.2419) time: 2.6925 data: 2.5481 max mem: 9671 +[13:55:20.496562] Epoch: [21] [7/8] eta: 0:00:00 lr: 0.000499 loss: 0.2997 (0.3128) time: 0.4605 data: 0.3186 max mem: 9671 +[13:55:20.613083] Epoch: [21] Total time: 0:00:03 (0.4753 s / it) +[13:55:20.622872] Averaged stats: lr: 0.000499 loss: 0.2997 (0.3128) +[13:55:23.312561] val: [0/2] eta: 0:00:05 loss: 0.1637 (0.1637) time: 2.6733 data: 2.6381 max mem: 9671 +[13:55:23.327922] val: [1/2] eta: 0:00:01 loss: 0.1637 (0.4264) time: 1.3440 data: 1.3191 max mem: 9671 +[13:55:23.434524] val: Total time: 0:00:02 (1.3980 s / it) +[13:55:23.443528] val loss: 0.42638447880744934 +[13:55:23.443708] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[13:55:23.487013] Best epoch = 17, Best score = 0.8982 +[13:55:23.766583] log_dir: ./output_logs/retfound +[13:55:26.376142] Epoch: [22] [0/8] eta: 0:00:20 lr: 0.000496 loss: 0.2969 (0.2969) time: 2.6085 data: 2.4631 max mem: 9671 +[13:55:27.366310] Epoch: [22] [7/8] eta: 0:00:00 lr: 0.000479 loss: 0.3222 (0.3138) time: 0.4497 data: 0.3080 max mem: 9671 +[13:55:27.486017] Epoch: [22] Total time: 0:00:03 (0.4649 s / it) +[13:55:27.494899] Averaged stats: lr: 0.000479 loss: 0.3222 (0.3138) +[13:55:30.091932] val: [0/2] eta: 0:00:05 loss: 0.1736 (0.1736) time: 2.5837 data: 2.5485 max mem: 9671 +[13:55:30.107332] val: [1/2] eta: 0:00:01 loss: 0.1736 (0.3998) time: 1.2993 data: 1.2743 max mem: 9671 +[13:55:30.212236] val: Total time: 0:00:02 (1.3524 s / it) +[13:55:30.221133] val loss: 0.3997582793235779 +[13:55:30.221325] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[13:55:30.266705] Best epoch = 17, Best score = 0.8982 +[13:55:30.537400] log_dir: ./output_logs/retfound +[13:55:33.166547] Epoch: [23] [0/8] eta: 0:00:21 lr: 0.000476 loss: 0.2342 (0.2342) time: 2.6282 data: 2.4815 max mem: 9671 +[13:55:34.153806] Epoch: [23] [7/8] eta: 0:00:00 lr: 0.000457 loss: 0.3115 (0.3495) time: 0.4518 data: 0.3103 max mem: 9671 +[13:55:34.270581] Epoch: [23] Total time: 0:00:03 (0.4666 s / it) +[13:55:34.280309] Averaged stats: lr: 0.000457 loss: 0.3115 (0.3495) +[13:55:36.882060] val: [0/2] eta: 0:00:05 loss: 0.1643 (0.1643) time: 2.5911 data: 2.5562 max mem: 9671 +[13:55:36.897610] val: [1/2] eta: 0:00:01 loss: 0.1643 (0.4087) time: 1.3030 data: 1.2782 max mem: 9671 +[13:55:37.008262] val: Total time: 0:00:02 (1.3590 s / it) +[13:55:37.018835] val loss: 0.4086921811103821 +[13:55:37.019056] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[13:55:37.061416] Best epoch = 17, Best score = 0.8982 +[13:55:37.323509] log_dir: ./output_logs/retfound +[13:55:40.024435] Epoch: [24] [0/8] eta: 0:00:21 lr: 0.000455 loss: 0.3266 (0.3266) time: 2.6998 data: 2.5546 max mem: 9671 +[13:55:41.032889] Epoch: [24] [7/8] eta: 0:00:00 lr: 0.000435 loss: 0.3266 (0.3227) time: 0.4634 data: 0.3210 max mem: 9671 +[13:55:41.148261] Epoch: [24] Total time: 0:00:03 (0.4781 s / it) +[13:55:41.157306] Averaged stats: lr: 0.000435 loss: 0.3266 (0.3227) +[13:55:43.769570] val: [0/2] eta: 0:00:05 loss: 0.1567 (0.1567) time: 2.5970 data: 2.5624 max mem: 9671 +[13:55:43.785143] val: [1/2] eta: 0:00:01 loss: 0.1567 (0.4114) time: 1.3060 data: 1.2813 max mem: 9671 +[13:55:44.078920] val: Total time: 0:00:02 (1.4535 s / it) +[13:55:44.087839] val loss: 0.41143083572387695 +[13:55:44.088016] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9283, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9285, Kappa: 0.6596, Score: 0.8049 +[13:55:44.431151] Best epoch = 17, Best score = 0.8982 +[13:55:45.088004] log_dir: ./output_logs/retfound +[13:55:47.741439] Epoch: [25] [0/8] eta: 0:00:21 lr: 0.000432 loss: 0.2123 (0.2123) time: 2.6524 data: 2.5082 max mem: 9671 +[13:55:48.730249] Epoch: [25] [7/8] eta: 0:00:00 lr: 0.000412 loss: 0.2637 (0.2978) time: 0.4551 data: 0.3136 max mem: 9671 +[13:55:48.840251] Epoch: [25] Total time: 0:00:03 (0.4690 s / it) +[13:55:48.841157] Averaged stats: lr: 0.000412 loss: 0.2637 (0.2978) +[13:55:51.401680] val: [0/2] eta: 0:00:05 loss: 0.1612 (0.1612) time: 2.5498 data: 2.5137 max mem: 9671 +[13:55:51.417144] val: [1/2] eta: 0:00:01 loss: 0.1612 (0.5095) time: 1.2823 data: 1.2569 max mem: 9671 +[13:55:51.521274] val: Total time: 0:00:02 (1.3351 s / it) +[13:55:51.531217] val loss: 0.5094860941171646 +[13:55:51.531390] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9319, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9306, Kappa: 0.6596, Score: 0.8061 +[13:55:51.568777] Best epoch = 17, Best score = 0.8982 +[13:55:51.856946] log_dir: ./output_logs/retfound +[13:55:54.556280] Epoch: [26] [0/8] eta: 0:00:21 lr: 0.000409 loss: 0.3678 (0.3678) time: 2.6982 data: 2.5529 max mem: 9671 +[13:55:55.549899] Epoch: [26] [7/8] eta: 0:00:00 lr: 0.000389 loss: 0.2022 (0.2613) time: 0.4614 data: 0.3192 max mem: 9671 +[13:55:55.668925] Epoch: [26] Total time: 0:00:03 (0.4765 s / it) +[13:55:55.677318] Averaged stats: lr: 0.000389 loss: 0.2022 (0.2613) +[13:55:58.327322] val: [0/2] eta: 0:00:05 loss: 0.1778 (0.1778) time: 2.6385 data: 2.6048 max mem: 9671 +[13:55:58.342718] val: [1/2] eta: 0:00:01 loss: 0.1778 (0.5650) time: 1.3267 data: 1.3025 max mem: 9671 +[13:55:58.451513] val: Total time: 0:00:02 (1.3817 s / it) +[13:55:58.460350] val loss: 0.5649861097335815 +[13:55:58.460536] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9292, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9234, Kappa: 0.6596, Score: 0.8052 +[13:55:58.503458] Best epoch = 17, Best score = 0.8982 +[13:55:58.775419] log_dir: ./output_logs/retfound +[13:56:01.560598] Epoch: [27] [0/8] eta: 0:00:22 lr: 0.000386 loss: 0.3435 (0.3435) time: 2.7842 data: 2.6406 max mem: 9671 +[13:56:02.553688] Epoch: [27] [7/8] eta: 0:00:00 lr: 0.000365 loss: 0.3040 (0.3233) time: 0.4721 data: 0.3302 max mem: 9671 +[13:56:02.670899] Epoch: [27] Total time: 0:00:03 (0.4869 s / it) +[13:56:02.680284] Averaged stats: lr: 0.000365 loss: 0.3040 (0.3233) +[13:56:05.332181] val: [0/2] eta: 0:00:05 loss: 0.1657 (0.1657) time: 2.6396 data: 2.6043 max mem: 9671 +[13:56:05.347827] val: [1/2] eta: 0:00:01 loss: 0.1657 (0.4328) time: 1.3273 data: 1.3022 max mem: 9671 +[13:56:05.456625] val: Total time: 0:00:02 (1.3824 s / it) +[13:56:05.465596] val loss: 0.43275587260723114 +[13:56:05.465841] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9391, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9252, Kappa: 0.6596, Score: 0.8085 +[13:56:05.506627] Best epoch = 17, Best score = 0.8982 +[13:56:05.768898] log_dir: ./output_logs/retfound +[13:56:08.273372] Epoch: [28] [0/8] eta: 0:00:20 lr: 0.000362 loss: 0.2153 (0.2153) time: 2.5035 data: 2.3590 max mem: 9671 +[13:56:09.265644] Epoch: [28] [7/8] eta: 0:00:00 lr: 0.000341 loss: 0.2252 (0.2592) time: 0.4369 data: 0.2950 max mem: 9671 +[13:56:09.378523] Epoch: [28] Total time: 0:00:03 (0.4512 s / it) +[13:56:09.386861] Averaged stats: lr: 0.000341 loss: 0.2252 (0.2592) +[13:56:12.018122] val: [0/2] eta: 0:00:05 loss: 0.2523 (0.2523) time: 2.6136 data: 2.5779 max mem: 9671 +[13:56:12.033706] val: [1/2] eta: 0:00:01 loss: 0.1857 (0.2190) time: 1.3143 data: 1.2890 max mem: 9671 +[13:56:12.142252] val: Total time: 0:00:02 (1.3693 s / it) +[13:56:12.151085] val loss: 0.21898877620697021 +[13:56:12.151264] Accuracy: 0.9000, F1 Score: 0.8566, ROC AUC: 0.9498, Hamming Loss: 0.1000, + Jaccard Score: 0.7576, Precision: 0.8566, Recall: 0.8566, + Average Precision: 0.9370, Kappa: 0.7133, Score: 0.8399 +[13:56:12.189265] Best epoch = 17, Best score = 0.8982 +[13:56:12.455804] log_dir: ./output_logs/retfound +[13:56:14.814913] Epoch: [29] [0/8] eta: 0:00:18 lr: 0.000337 loss: 0.2674 (0.2674) time: 2.3578 data: 2.2118 max mem: 9671 +[13:56:16.083719] Epoch: [29] [7/8] eta: 0:00:00 lr: 0.000316 loss: 0.2900 (0.2986) time: 0.4532 data: 0.3119 max mem: 9671 +[13:56:16.195376] Epoch: [29] Total time: 0:00:03 (0.4674 s / it) +[13:56:16.203868] Averaged stats: lr: 0.000316 loss: 0.2900 (0.2986) +[13:56:18.768094] val: [0/2] eta: 0:00:05 loss: 0.1922 (0.1922) time: 2.5473 data: 2.5170 max mem: 9671 +[13:56:18.780819] val: [1/2] eta: 0:00:01 loss: 0.1922 (0.2850) time: 1.2797 data: 1.2585 max mem: 9671 +[13:56:18.892676] val: Total time: 0:00:02 (1.3363 s / it) +[13:56:18.902135] val loss: 0.28504690527915955 +[13:56:18.902386] Accuracy: 0.9250, F1 Score: 0.8880, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.8045, Precision: 0.9062, Recall: 0.8728, + Average Precision: 0.9346, Kappa: 0.7761, Score: 0.8713 +[13:56:18.948168] Best epoch = 17, Best score = 0.8982 +[13:56:19.202588] log_dir: ./output_logs/retfound +[13:56:21.767984] Epoch: [30] [0/8] eta: 0:00:20 lr: 0.000313 loss: 0.2006 (0.2006) time: 2.5642 data: 2.4200 max mem: 9671 +[13:56:22.758710] Epoch: [30] [7/8] eta: 0:00:00 lr: 0.000292 loss: 0.2116 (0.2705) time: 0.4443 data: 0.3026 max mem: 9671 +[13:56:22.870294] Epoch: [30] Total time: 0:00:03 (0.4584 s / it) +[13:56:22.878863] Averaged stats: lr: 0.000292 loss: 0.2116 (0.2705) +[13:56:25.501204] val: [0/2] eta: 0:00:05 loss: 0.1967 (0.1967) time: 2.6052 data: 2.5712 max mem: 9671 +[13:56:25.516928] val: [1/2] eta: 0:00:01 loss: 0.1967 (0.3365) time: 1.3102 data: 1.2857 max mem: 9671 +[13:56:25.618390] val: Total time: 0:00:02 (1.3615 s / it) +[13:56:25.627316] val loss: 0.3365369886159897 +[13:56:25.627495] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9498, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9408, Kappa: 0.6887, Score: 0.8274 +[13:56:25.666863] Best epoch = 17, Best score = 0.8982 +[13:56:25.905752] log_dir: ./output_logs/retfound +[13:56:28.512268] Epoch: [31] [0/8] eta: 0:00:20 lr: 0.000289 loss: 0.5076 (0.5076) time: 2.6055 data: 2.4603 max mem: 9671 +[13:56:29.504432] Epoch: [31] [7/8] eta: 0:00:00 lr: 0.000267 loss: 0.2211 (0.2924) time: 0.4496 data: 0.3076 max mem: 9671 +[13:56:29.615107] Epoch: [31] Total time: 0:00:03 (0.4636 s / it) +[13:56:29.624337] Averaged stats: lr: 0.000267 loss: 0.2211 (0.2924) +[13:56:32.263759] val: [0/2] eta: 0:00:05 loss: 0.1812 (0.1812) time: 2.6227 data: 2.5889 max mem: 9671 +[13:56:32.276817] val: [1/2] eta: 0:00:01 loss: 0.1812 (0.3428) time: 1.3176 data: 1.2945 max mem: 9671 +[13:56:32.382412] val: Total time: 0:00:02 (1.3710 s / it) +[13:56:32.391100] val loss: 0.3427543491125107 +[13:56:32.391293] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9369, Kappa: 0.7561, Score: 0.8598 +[13:56:32.427747] Best epoch = 17, Best score = 0.8982 +[13:56:32.701895] log_dir: ./output_logs/retfound +[13:56:35.191139] Epoch: [32] [0/8] eta: 0:00:19 lr: 0.000264 loss: 0.1763 (0.1763) time: 2.4883 data: 2.3441 max mem: 9671 +[13:56:36.182969] Epoch: [32] [7/8] eta: 0:00:00 lr: 0.000243 loss: 0.3075 (0.2896) time: 0.4349 data: 0.2931 max mem: 9671 +[13:56:36.297918] Epoch: [32] Total time: 0:00:03 (0.4495 s / it) +[13:56:36.307398] Averaged stats: lr: 0.000243 loss: 0.3075 (0.2896) +[13:56:39.028118] val: [0/2] eta: 0:00:05 loss: 0.1718 (0.1718) time: 2.7050 data: 2.6708 max mem: 9671 +[13:56:39.043524] val: [1/2] eta: 0:00:01 loss: 0.1718 (0.3577) time: 1.3599 data: 1.3355 max mem: 9671 +[13:56:39.147275] val: Total time: 0:00:02 (1.4125 s / it) +[13:56:39.156031] val loss: 0.35773538053035736 +[13:56:39.156229] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9357, Kappa: 0.7561, Score: 0.8598 +[13:56:39.198000] Best epoch = 17, Best score = 0.8982 +[13:56:39.453581] log_dir: ./output_logs/retfound +[13:56:41.822783] Epoch: [33] [0/8] eta: 0:00:18 lr: 0.000240 loss: 0.1372 (0.1372) time: 2.3680 data: 2.2236 max mem: 9671 +[13:56:42.866195] Epoch: [33] [7/8] eta: 0:00:00 lr: 0.000220 loss: 0.2494 (0.2693) time: 0.4263 data: 0.2845 max mem: 9671 +[13:56:42.973127] Epoch: [33] Total time: 0:00:03 (0.4399 s / it) +[13:56:42.980214] Averaged stats: lr: 0.000220 loss: 0.2494 (0.2693) +[13:56:45.685841] val: [0/2] eta: 0:00:05 loss: 0.1649 (0.1649) time: 2.6894 data: 2.6553 max mem: 9671 +[13:56:45.701073] val: [1/2] eta: 0:00:01 loss: 0.1649 (0.4252) time: 1.3519 data: 1.3277 max mem: 9671 +[13:56:45.808981] val: Total time: 0:00:02 (1.4068 s / it) +[13:56:45.820281] val loss: 0.42521847784519196 +[13:56:45.820468] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9462, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9409, Kappa: 0.6596, Score: 0.8109 +[13:56:45.861508] Best epoch = 17, Best score = 0.8982 +[13:56:46.149671] log_dir: ./output_logs/retfound +[13:56:48.772298] Epoch: [34] [0/8] eta: 0:00:20 lr: 0.000217 loss: 0.2293 (0.2293) time: 2.6219 data: 2.4779 max mem: 9671 +[13:56:49.762117] Epoch: [34] [7/8] eta: 0:00:00 lr: 0.000196 loss: 0.2456 (0.2689) time: 0.4514 data: 0.3099 max mem: 9671 +[13:56:49.877441] Epoch: [34] Total time: 0:00:03 (0.4660 s / it) +[13:56:49.886558] Averaged stats: lr: 0.000196 loss: 0.2456 (0.2689) +[13:56:52.464416] val: [0/2] eta: 0:00:05 loss: 0.1657 (0.1657) time: 2.5624 data: 2.5281 max mem: 9671 +[13:56:52.480069] val: [1/2] eta: 0:00:01 loss: 0.1657 (0.4029) time: 1.2887 data: 1.2641 max mem: 9671 +[13:56:52.584949] val: Total time: 0:00:02 (1.3418 s / it) +[13:56:52.593747] val loss: 0.4029083847999573 +[13:56:52.593937] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9391, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9353, Kappa: 0.7561, Score: 0.8574 +[13:56:52.637970] Best epoch = 17, Best score = 0.8982 +[13:56:52.908137] log_dir: ./output_logs/retfound +[13:56:55.634125] Epoch: [35] [0/8] eta: 0:00:21 lr: 0.000194 loss: 0.4496 (0.4496) time: 2.7249 data: 2.5847 max mem: 9671 +[13:56:56.632537] Epoch: [35] [7/8] eta: 0:00:00 lr: 0.000174 loss: 0.2822 (0.2812) time: 0.4653 data: 0.3232 max mem: 9671 +[13:56:56.743234] Epoch: [35] Total time: 0:00:03 (0.4794 s / it) +[13:56:56.752586] Averaged stats: lr: 0.000174 loss: 0.2822 (0.2812) +[13:56:59.396594] val: [0/2] eta: 0:00:05 loss: 0.1606 (0.1606) time: 2.6246 data: 2.5896 max mem: 9671 +[13:56:59.412344] val: [1/2] eta: 0:00:01 loss: 0.1606 (0.3562) time: 1.3198 data: 1.2949 max mem: 9671 +[13:56:59.523423] val: Total time: 0:00:02 (1.3761 s / it) +[13:56:59.532239] val loss: 0.3561532497406006 +[13:56:59.532441] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9391, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9353, Kappa: 0.7561, Score: 0.8574 +[13:56:59.573985] Best epoch = 17, Best score = 0.8982 +[13:56:59.835305] log_dir: ./output_logs/retfound +[13:57:02.358442] Epoch: [36] [0/8] eta: 0:00:20 lr: 0.000171 loss: 0.4470 (0.4470) time: 2.5222 data: 2.3781 max mem: 9671 +[13:57:03.348156] Epoch: [36] [7/8] eta: 0:00:00 lr: 0.000153 loss: 0.2518 (0.2885) time: 0.4389 data: 0.2973 max mem: 9671 +[13:57:03.475106] Epoch: [36] Total time: 0:00:03 (0.4550 s / it) +[13:57:03.485225] Averaged stats: lr: 0.000153 loss: 0.2518 (0.2885) +[13:57:06.118664] val: [0/2] eta: 0:00:05 loss: 0.1652 (0.1652) time: 2.6214 data: 2.5865 max mem: 9671 +[13:57:06.134092] val: [1/2] eta: 0:00:01 loss: 0.1652 (0.2911) time: 1.3181 data: 1.2933 max mem: 9671 +[13:57:06.238176] val: Total time: 0:00:02 (1.3709 s / it) +[13:57:06.246976] val loss: 0.2910751849412918 +[13:57:06.247172] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9409, Kappa: 0.7561, Score: 0.8598 +[13:57:06.292415] Best epoch = 17, Best score = 0.8982 +[13:57:06.545620] log_dir: ./output_logs/retfound +[13:57:09.305220] Epoch: [37] [0/8] eta: 0:00:22 lr: 0.000150 loss: 0.3017 (0.3017) time: 2.7585 data: 2.6143 max mem: 9671 +[13:57:10.289578] Epoch: [37] [7/8] eta: 0:00:00 lr: 0.000132 loss: 0.2246 (0.2802) time: 0.4678 data: 0.3269 max mem: 9671 +[13:57:10.402850] Epoch: [37] Total time: 0:00:03 (0.4821 s / it) +[13:57:10.411895] Averaged stats: lr: 0.000132 loss: 0.2246 (0.2802) +[13:57:12.977439] val: [0/2] eta: 0:00:05 loss: 0.1532 (0.1532) time: 2.5546 data: 2.5194 max mem: 9671 +[13:57:12.993077] val: [1/2] eta: 0:00:01 loss: 0.1532 (0.3142) time: 1.2847 data: 1.2598 max mem: 9671 +[13:57:13.101946] val: Total time: 0:00:02 (1.3400 s / it) +[13:57:13.111613] val loss: 0.3142062872648239 +[13:57:13.111822] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9431, Kappa: 0.7561, Score: 0.8609 +[13:57:13.155390] Best epoch = 17, Best score = 0.8982 +[13:57:13.419455] log_dir: ./output_logs/retfound +[13:57:15.735752] Epoch: [38] [0/8] eta: 0:00:18 lr: 0.000130 loss: 0.1331 (0.1331) time: 2.3153 data: 2.1710 max mem: 9671 +[13:57:16.860145] Epoch: [38] [7/8] eta: 0:00:00 lr: 0.000113 loss: 0.2899 (0.2815) time: 0.4299 data: 0.2878 max mem: 9671 +[13:57:16.974810] Epoch: [38] Total time: 0:00:03 (0.4444 s / it) +[13:57:16.982721] Averaged stats: lr: 0.000113 loss: 0.2899 (0.2815) +[13:57:19.584262] val: [0/2] eta: 0:00:05 loss: 0.1554 (0.1554) time: 2.5887 data: 2.5524 max mem: 9671 +[13:57:19.599922] val: [1/2] eta: 0:00:01 loss: 0.1554 (0.2977) time: 1.3018 data: 1.2763 max mem: 9671 +[13:57:19.714194] val: Total time: 0:00:02 (1.3597 s / it) +[13:57:19.723596] val loss: 0.2977108359336853 +[13:57:19.723793] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9534, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9454, Kappa: 0.7561, Score: 0.8621 +[13:57:19.762774] Best epoch = 17, Best score = 0.8982 +[13:57:20.029088] log_dir: ./output_logs/retfound +[13:57:22.694535] Epoch: [39] [0/8] eta: 0:00:21 lr: 0.000110 loss: 0.2601 (0.2601) time: 2.6645 data: 2.5190 max mem: 9671 +[13:57:23.691181] Epoch: [39] [7/8] eta: 0:00:00 lr: 0.000095 loss: 0.2684 (0.2860) time: 0.4575 data: 0.3150 max mem: 9671 +[13:57:23.801353] Epoch: [39] Total time: 0:00:03 (0.4715 s / it) +[13:57:23.809432] Averaged stats: lr: 0.000095 loss: 0.2684 (0.2860) +[13:57:26.361756] val: [0/2] eta: 0:00:05 loss: 0.1555 (0.1555) time: 2.5356 data: 2.5008 max mem: 9671 +[13:57:26.377155] val: [1/2] eta: 0:00:01 loss: 0.1555 (0.2966) time: 1.2752 data: 1.2505 max mem: 9671 +[13:57:26.480480] val: Total time: 0:00:02 (1.3276 s / it) +[13:57:26.489272] val loss: 0.2965516448020935 +[13:57:26.489492] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9534, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9454, Kappa: 0.7561, Score: 0.8621 +[13:57:26.525424] Best epoch = 17, Best score = 0.8982 +[13:57:26.832626] log_dir: ./output_logs/retfound +[13:57:29.340345] Epoch: [40] [0/8] eta: 0:00:20 lr: 0.000092 loss: 0.2017 (0.2017) time: 2.5066 data: 2.3740 max mem: 9671 +[13:57:30.337085] Epoch: [40] [7/8] eta: 0:00:00 lr: 0.000078 loss: 0.2769 (0.2756) time: 0.4378 data: 0.2969 max mem: 9671 +[13:57:30.455456] Epoch: [40] Total time: 0:00:03 (0.4528 s / it) +[13:57:30.465251] Averaged stats: lr: 0.000078 loss: 0.2769 (0.2756) +[13:57:33.157634] val: [0/2] eta: 0:00:05 loss: 0.1451 (0.1451) time: 2.6779 data: 2.6430 max mem: 9671 +[13:57:33.173109] val: [1/2] eta: 0:00:01 loss: 0.1451 (0.3349) time: 1.3464 data: 1.3216 max mem: 9671 +[13:57:33.286367] val: Total time: 0:00:02 (1.4037 s / it) +[13:57:33.295163] val loss: 0.3348654955625534 +[13:57:33.295354] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9424, Kappa: 0.7561, Score: 0.8609 +[13:57:33.343824] Best epoch = 17, Best score = 0.8982 +[13:57:33.603696] log_dir: ./output_logs/retfound +[13:57:36.339120] Epoch: [41] [0/8] eta: 0:00:21 lr: 0.000076 loss: 0.2176 (0.2176) time: 2.7342 data: 2.5890 max mem: 9671 +[13:57:37.326870] Epoch: [41] [7/8] eta: 0:00:00 lr: 0.000062 loss: 0.2356 (0.2808) time: 0.4652 data: 0.3238 max mem: 9671 +[13:57:37.452503] Epoch: [41] Total time: 0:00:03 (0.4811 s / it) +[13:57:37.461739] Averaged stats: lr: 0.000062 loss: 0.2356 (0.2808) +[13:57:40.106831] val: [0/2] eta: 0:00:05 loss: 0.1402 (0.1402) time: 2.6276 data: 2.5935 max mem: 9671 +[13:57:40.122170] val: [1/2] eta: 0:00:01 loss: 0.1402 (0.3686) time: 1.3212 data: 1.2968 max mem: 9671 +[13:57:40.236666] val: Total time: 0:00:02 (1.3791 s / it) +[13:57:40.245478] val loss: 0.3686094284057617 +[13:57:40.245655] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9534, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9454, Kappa: 0.6596, Score: 0.8133 +[13:57:40.289347] Best epoch = 17, Best score = 0.8982 +[13:57:40.540190] log_dir: ./output_logs/retfound +[13:57:43.045722] Epoch: [42] [0/8] eta: 0:00:20 lr: 0.000061 loss: 0.2027 (0.2027) time: 2.5042 data: 2.3563 max mem: 9671 +[13:57:44.034881] Epoch: [42] [7/8] eta: 0:00:00 lr: 0.000049 loss: 0.2027 (0.2439) time: 0.4366 data: 0.2946 max mem: 9671 +[13:57:44.153749] Epoch: [42] Total time: 0:00:03 (0.4517 s / it) +[13:57:44.163441] Averaged stats: lr: 0.000049 loss: 0.2027 (0.2439) +[13:57:46.812922] val: [0/2] eta: 0:00:05 loss: 0.1391 (0.1391) time: 2.6341 data: 2.5981 max mem: 9671 +[13:57:46.828705] val: [1/2] eta: 0:00:01 loss: 0.1391 (0.4043) time: 1.3246 data: 1.2991 max mem: 9671 +[13:57:46.938953] val: Total time: 0:00:02 (1.3805 s / it) +[13:57:46.947754] val loss: 0.4043458551168442 +[13:57:46.947939] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9534, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9454, Kappa: 0.6596, Score: 0.8133 +[13:57:46.998329] Best epoch = 17, Best score = 0.8982 +[13:57:47.255679] log_dir: ./output_logs/retfound +[13:57:49.865228] Epoch: [43] [0/8] eta: 0:00:20 lr: 0.000047 loss: 0.1854 (0.1854) time: 2.6086 data: 2.4643 max mem: 9671 +[13:57:50.857479] Epoch: [43] [7/8] eta: 0:00:00 lr: 0.000036 loss: 0.2205 (0.2380) time: 0.4500 data: 0.3081 max mem: 9671 +[13:57:50.967578] Epoch: [43] Total time: 0:00:03 (0.4640 s / it) +[13:57:50.975726] Averaged stats: lr: 0.000036 loss: 0.2205 (0.2380) +[13:57:53.598847] val: [0/2] eta: 0:00:05 loss: 0.1402 (0.1402) time: 2.6119 data: 2.5781 max mem: 9671 +[13:57:53.614524] val: [1/2] eta: 0:00:01 loss: 0.1402 (0.4161) time: 1.3135 data: 1.2891 max mem: 9671 +[13:57:53.722129] val: Total time: 0:00:02 (1.3680 s / it) +[13:57:53.730996] val loss: 0.416122242808342 +[13:57:53.731195] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[13:57:53.768099] Best epoch = 17, Best score = 0.8982 +[13:57:54.041269] log_dir: ./output_logs/retfound +[13:57:56.740457] Epoch: [44] [0/8] eta: 0:00:21 lr: 0.000035 loss: 0.3020 (0.3020) time: 2.6982 data: 2.5546 max mem: 9671 +[13:57:57.730640] Epoch: [44] [7/8] eta: 0:00:00 lr: 0.000026 loss: 0.2097 (0.2575) time: 0.4610 data: 0.3194 max mem: 9671 +[13:57:57.841698] Epoch: [44] Total time: 0:00:03 (0.4750 s / it) +[13:57:57.850816] Averaged stats: lr: 0.000026 loss: 0.2097 (0.2575) +[13:58:00.402354] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.5351 data: 2.5013 max mem: 9671 +[13:58:00.418008] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.4057) time: 1.2751 data: 1.2507 max mem: 9671 +[13:58:00.527311] val: Total time: 0:00:02 (1.3304 s / it) +[13:58:00.536186] val loss: 0.4057278037071228 +[13:58:00.536388] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[13:58:00.580468] Best epoch = 17, Best score = 0.8982 +[13:58:00.844867] log_dir: ./output_logs/retfound +[13:58:03.374393] Epoch: [45] [0/8] eta: 0:00:20 lr: 0.000025 loss: 0.2266 (0.2266) time: 2.5286 data: 2.3811 max mem: 9671 +[13:58:04.366539] Epoch: [45] [7/8] eta: 0:00:00 lr: 0.000017 loss: 0.2266 (0.2669) time: 0.4400 data: 0.2977 max mem: 9671 +[13:58:04.509172] Epoch: [45] Total time: 0:00:03 (0.4580 s / it) +[13:58:04.517954] Averaged stats: lr: 0.000017 loss: 0.2266 (0.2669) +[13:58:07.050917] val: [0/2] eta: 0:00:05 loss: 0.1414 (0.1414) time: 2.5190 data: 2.4836 max mem: 9671 +[13:58:07.066464] val: [1/2] eta: 0:00:01 loss: 0.1414 (0.3876) time: 1.2670 data: 1.2419 max mem: 9671 +[13:58:07.170584] val: Total time: 0:00:02 (1.3197 s / it) +[13:58:07.179685] val loss: 0.38761037588119507 +[13:58:07.179893] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[13:58:07.223819] Best epoch = 17, Best score = 0.8982 +[13:58:07.484293] log_dir: ./output_logs/retfound +[13:58:10.183208] Epoch: [46] [0/8] eta: 0:00:21 lr: 0.000016 loss: 0.2934 (0.2934) time: 2.6979 data: 2.5539 max mem: 9671 +[13:58:11.180959] Epoch: [46] [7/8] eta: 0:00:00 lr: 0.000010 loss: 0.2377 (0.2666) time: 0.4618 data: 0.3193 max mem: 9671 +[13:58:11.289298] Epoch: [46] Total time: 0:00:03 (0.4756 s / it) +[13:58:11.297206] Averaged stats: lr: 0.000010 loss: 0.2377 (0.2666) +[13:58:13.894239] val: [0/2] eta: 0:00:05 loss: 0.1412 (0.1412) time: 2.5814 data: 2.5478 max mem: 9671 +[13:58:13.909786] val: [1/2] eta: 0:00:01 loss: 0.1412 (0.3810) time: 1.2982 data: 1.2740 max mem: 9671 +[13:58:14.014601] val: Total time: 0:00:02 (1.3513 s / it) +[13:58:14.024088] val loss: 0.3810284584760666 +[13:58:14.024264] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[13:58:14.059744] Best epoch = 17, Best score = 0.8982 +[13:58:14.332951] log_dir: ./output_logs/retfound +[13:58:16.822406] Epoch: [47] [0/8] eta: 0:00:19 lr: 0.000010 loss: 0.2115 (0.2115) time: 2.4885 data: 2.3437 max mem: 9671 +[13:58:17.876939] Epoch: [47] [7/8] eta: 0:00:00 lr: 0.000005 loss: 0.2115 (0.2570) time: 0.4428 data: 0.3010 max mem: 9671 +[13:58:17.995207] Epoch: [47] Total time: 0:00:03 (0.4578 s / it) +[13:58:18.003435] Averaged stats: lr: 0.000005 loss: 0.2115 (0.2570) +[13:58:20.696537] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.6646 data: 2.6306 max mem: 9671 +[13:58:20.712303] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3793) time: 1.3398 data: 1.3154 max mem: 9671 +[13:58:20.818223] val: Total time: 0:00:02 (1.3935 s / it) +[13:58:20.833242] val loss: 0.3793119490146637 +[13:58:20.833492] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[13:58:20.873481] Best epoch = 17, Best score = 0.8982 +[13:58:21.160139] log_dir: ./output_logs/retfound +[13:58:23.571144] Epoch: [48] [0/8] eta: 0:00:19 lr: 0.000005 loss: 0.3700 (0.3700) time: 2.4099 data: 2.2657 max mem: 9671 +[13:58:24.560774] Epoch: [48] [7/8] eta: 0:00:00 lr: 0.000002 loss: 0.2148 (0.2645) time: 0.4248 data: 0.2835 max mem: 9671 +[13:58:24.677673] Epoch: [48] Total time: 0:00:03 (0.4397 s / it) +[13:58:24.687313] Averaged stats: lr: 0.000002 loss: 0.2148 (0.2645) +[13:58:27.307905] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.6085 data: 2.5741 max mem: 9671 +[13:58:27.323530] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3780) time: 1.3117 data: 1.2871 max mem: 9671 +[13:58:27.428244] val: Total time: 0:00:02 (1.3648 s / it) +[13:58:27.437292] val loss: 0.3780297338962555 +[13:58:27.437497] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[13:58:27.475820] Best epoch = 17, Best score = 0.8982 +[13:58:27.737422] log_dir: ./output_logs/retfound +[13:58:30.133935] Epoch: [49] [0/8] eta: 0:00:19 lr: 0.000002 loss: 0.1158 (0.1158) time: 2.3955 data: 2.2504 max mem: 9671 +[13:58:31.123119] Epoch: [49] [7/8] eta: 0:00:00 lr: 0.000001 loss: 0.2527 (0.2805) time: 0.4230 data: 0.2814 max mem: 9671 +[13:58:31.242759] Epoch: [49] Total time: 0:00:03 (0.4381 s / it) +[13:58:31.251532] Averaged stats: lr: 0.000001 loss: 0.2527 (0.2805) +[13:58:33.827005] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.5645 data: 2.5297 max mem: 9671 +[13:58:33.842600] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3775) time: 1.2898 data: 1.2649 max mem: 9671 +[13:58:33.957716] val: Total time: 0:00:02 (1.3480 s / it) +[13:58:33.966648] val loss: 0.37748883664608 +[13:58:33.966878] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[13:58:34.007113] Best epoch = 17, Best score = 0.8982 +[13:58:37.849680] Test with the best model, epoch = 17: +[13:58:40.407636] test: [0/3] eta: 0:00:07 loss: 0.2582 (0.2582) time: 2.5484 data: 2.5134 max mem: 9671 +[13:58:40.550538] test: [2/3] eta: 0:00:00 loss: 0.2582 (0.2454) time: 0.8969 data: 0.8379 max mem: 9671 +[13:58:40.645493] test: Total time: 0:00:02 (0.9290 s / it) +[13:58:40.654785] val loss: 0.24544517199198404 +[13:58:40.654887] Accuracy: 0.9250, F1 Score: 0.8925, ROC AUC: 0.9516, Hamming Loss: 0.0750, + Jaccard Score: 0.8110, Precision: 0.8925, Recall: 0.8925, + Average Precision: 0.9512, Kappa: 0.7849, Score: 0.8763 +[13:58:41.372124] Training time 0:06:02 +[rank0]:[W615 13:58:41.789217613 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/adam/retfound acc=0.9250 auroc=0.9516129032258065 f1_macro=0.8925 qwk=0.7849462365591398 diff --git a/results/adam/vit/confusion_matrix.png b/results/adam/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..e8a208602b1874377b6b866b683dc2f461be78b7 --- /dev/null +++ b/results/adam/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01b849ecc3b93121314108bdaf949eb826f6d28918114264525636aa6f71afa1 +size 68003 diff --git a/results/adam/vit/log.csv b/results/adam/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..108663060fd79a2a3e24e75f63a49ebfdb6b92bb --- /dev/null +++ b/results/adam/vit/log.csv @@ -0,0 +1,33 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8326617628335953,0.5,0.37275985663082434,0.20603382384588553,5.545808836528073e-08 +1,0.7702071815729141,0.525,0.7204301075268817,0.45818926807259003,1.29402206185655e-07 +2,0.6546564847230911,0.7,0.7849462365591398,0.5773837351677463,2.0334632400602932e-07 +3,0.6182350814342499,0.75,0.870967741935484,0.6940345499098299,2.7729044182640365e-07 +4,0.5662039071321487,0.75,0.8960573476702509,0.6544608924687236,3.512345596467779e-07 +5,0.5842887908220291,0.725,0.8745519713261649,0.6756845538858233,3.694672444929302e-07 +6,0.5582687333226204,0.825,0.899641577060932,0.7416449759260039,3.683426684987483e-07 +7,0.4713408350944519,0.775,0.9175627240143369,0.7299333069266117,3.663241848222764e-07 +8,0.5112659633159637,0.775,0.9175627240143369,0.7299333069266117,3.6342162731308893e-07 +9,0.465877428650856,0.9,0.9139784946236559,0.8279569892473119,3.5964913693960667e-07 +10,0.4645930454134941,0.75,0.9247311827956989,0.6824673300830383,3.550250928957199e-07 +11,0.4524441137909889,0.775,0.946236559139785,0.7109049537056887,3.495720230592029e-07 +12,0.4489521011710167,0.8,0.942652329749104,0.7321708645126134,3.4331649423815874e-07 +13,0.4123004525899887,0.825,0.946236559139785,0.7571766366189548,3.362889827402e-07 +14,0.412119522690773,0.9,0.9569892473118279,0.8422939068100358,3.285237258949416e-07 +15,0.38696958124637604,0.8,0.9641577060931898,0.7667567252402153,3.2005855525317277e-07 +16,0.3891528472304344,0.9,0.946236559139785,0.8387096774193549,3.1093471227534435e-07 +17,0.38670214265584946,0.775,0.9283154121863799,0.7049312380545537,3.0119664740731875e-07 +18,0.4111583083868027,0.925,0.935483870967742,0.866519485341882,2.908918035222662e-07 +19,0.40045975893735886,0.875,0.942652329749104,0.8084207298721364,2.800703847837553e-07 +20,0.38688723742961884,0.85,0.96415770609319,0.7884955365825584,2.687851120561149e-07 +21,0.3998527154326439,0.925,0.9605734767025089,0.8748826872534708,2.570909660536824e-07 +22,0.37270648032426834,0.875,0.9569892473118279,0.8131997023930445,2.450449194802903e-07 +23,0.36275216937065125,0.9,0.9569892473118279,0.8422939068100358,2.3270565946397963e-07 +24,0.39110884070396423,0.825,0.9498207885304659,0.7723605633561171,2.2013330163921197e-07 +25,0.39257583022117615,0.875,0.9283154121863799,0.8036417573512283,2.0738909726954043e-07 +26,0.3709259107708931,0.875,0.9283154121863799,0.8036417573512283,1.9453513483761127e-07 +27,0.35911235213279724,0.9,0.931899641577061,0.8339307048984468,1.8163403755631687e-07 +28,0.350225068628788,0.9,0.935483870967742,0.8351254480286738,1.6874865827479806e-07 +29,0.3573886975646019,0.875,0.9390681003584229,0.8072259867419094,1.5594177326568057e-07 +30,0.3484802022576332,0.925,0.9318996415770608,0.8653247422116549,1.4327577638538533e-07 +31,0.3666532337665558,0.875,0.942652329749104,0.8084207298721364,1.3081237509753143e-07 diff --git a/results/adam/vit/metrics.json b/results/adam/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..fa387bcb8fd7425d05b0b97e4264e6d762a6a6bf --- /dev/null +++ b/results/adam/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.9125, + "balanced_accuracy": 0.8844086021505376, + "precision_macro": 0.8701466781708369, + "recall_macro": 0.8844086021505376, + "f1_macro": 0.8769501208525599, + "precision_weighted": 0.9145168248490079, + "recall_weighted": 0.9125, + "f1_weighted": 0.9133267413755218, + "cohen_kappa": 0.7539543057996485, + "quadratic_weighted_kappa": 0.7539543057996485, + "mcc": 0.7544204852635336, + "auroc": 0.9318996415770608, + "auprc": 0.885447451927631, + "sensitivity": 0.8333333333333334, + "specificity": 0.9354838709677419, + "precision_pos": 0.7894736842105263, + "f1_pos": 0.8108108108108109, + "per_class": { + "0": { + "precision": 0.9508196721311475, + "recall": 0.9354838709677419, + "f1-score": 0.943089430894309, + "support": 62.0 + }, + "1": { + "precision": 0.7894736842105263, + "recall": 0.8333333333333334, + "f1-score": 0.8108108108108109, + "support": 18.0 + }, + "accuracy": 0.9125, + "macro avg": { + "precision": 0.8701466781708369, + "recall": 0.8844086021505376, + "f1-score": 0.8769501208525599, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.9145168248490079, + "recall": 0.9125, + "f1-score": 0.9133267413755218, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/adam/vit/pr.png b/results/adam/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..3c111f8607c642cf0cc3ca352ab966538d8de8d4 --- /dev/null +++ b/results/adam/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:822ccfdcbfbfa2cb1fcc682125bb453ffb6b3cfd7ca82926e12dbaf4afe3be8a +size 42694 diff --git a/results/adam/vit/roc.png b/results/adam/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..c10d1e438b63c7a1b8a483ece9aff81790fdd0c6 --- /dev/null +++ b/results/adam/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:56c1ee837c540ab36ab93e9471f702c9d8acbfe194d32fd75c878f8d0686c718 +size 57102 diff --git a/results/adam/vit/test_pred.npz b/results/adam/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..60e79dbf75d78b507400dc51b7bc8c48163de570 --- /dev/null +++ b/results/adam/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f70c97d3201c9d3359e705ebe55c3068c2b42b85f9d461b377bf407028d9864 +size 1790 diff --git a/results/adam/vit/train.log b/results/adam/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d7de37f4f0d8e4be40a1135d6d4f3f9185c8f3f6 --- /dev/null +++ b/results/adam/vit/train.log @@ -0,0 +1,169 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=280 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8327 val_acc=0.5000 val_auc=0.3728 score=0.2060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7702 val_acc=0.5250 val_auc=0.7204 score=0.4582 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6547 val_acc=0.7000 val_auc=0.7849 score=0.5774 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6182 val_acc=0.7500 val_auc=0.8710 score=0.6940 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5662 val_acc=0.7500 val_auc=0.8961 score=0.6545 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5843 val_acc=0.7250 val_auc=0.8746 score=0.6757 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5583 val_acc=0.8250 val_auc=0.8996 score=0.7416 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4713 val_acc=0.7750 val_auc=0.9176 score=0.7299 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5113 val_acc=0.7750 val_auc=0.9176 score=0.7299 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4659 val_acc=0.9000 val_auc=0.9140 score=0.8280 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4646 val_acc=0.7500 val_auc=0.9247 score=0.6825 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4524 val_acc=0.7750 val_auc=0.9462 score=0.7109 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.4490 val_acc=0.8000 val_auc=0.9427 score=0.7322 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4123 val_acc=0.8250 val_auc=0.9462 score=0.7572 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4121 val_acc=0.9000 val_auc=0.9570 score=0.8423 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.3870 val_acc=0.8000 val_auc=0.9642 score=0.7668 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3892 val_acc=0.9000 val_auc=0.9462 score=0.8387 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.3867 val_acc=0.7750 val_auc=0.9283 score=0.7049 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4112 val_acc=0.9250 val_auc=0.9355 score=0.8665 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4005 val_acc=0.8750 val_auc=0.9427 score=0.8084 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3869 val_acc=0.8500 val_auc=0.9642 score=0.7885 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.3999 val_acc=0.9250 val_auc=0.9606 score=0.8749 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3727 val_acc=0.8750 val_auc=0.9570 score=0.8132 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.3628 val_acc=0.9000 val_auc=0.9570 score=0.8423 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3911 val_acc=0.8250 val_auc=0.9498 score=0.7724 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3926 val_acc=0.8750 val_auc=0.9283 score=0.8036 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.3709 val_acc=0.8750 val_auc=0.9283 score=0.8036 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3591 val_acc=0.9000 val_auc=0.9319 score=0.8339 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3502 val_acc=0.9000 val_auc=0.9355 score=0.8351 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3574 val_acc=0.8750 val_auc=0.9391 score=0.8072 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.3485 val_acc=0.9250 val_auc=0.9319 score=0.8653 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.3667 val_acc=0.8750 val_auc=0.9427 score=0.8084 +[vit] early stop at ep31 (best ep21 score=0.8749) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=21 best_val_score=0.8749 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/adam/vit acc=0.9125 auroc=0.9318996415770608 f1_macro=0.8770 qwk=0.7539543057996485 diff --git a/results/airogs/resnet/confusion_matrix.png b/results/airogs/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..9298eed833d0759f5807aaa9fbde87b871a1e5fb --- /dev/null +++ b/results/airogs/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:065bbc07afbd5eb98ee33d083ee5e1bda244b0618bffacbac339d998dfe6d8c2 +size 65584 diff --git a/results/airogs/resnet/log.csv b/results/airogs/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..3b8816b10b3045ee1f40e37ad3140a8afec2c688 --- /dev/null +++ b/results/airogs/resnet/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6808460920284956,0.6759259259259259,0.746721536351166,0.5907443539542305,0.00016452991452991454 +1,0.4906524530588052,0.7814814814814814,0.8890123456790124,0.7427663474923492,0.0003311965811965812 +2,0.33204989918531513,0.8425925925925926,0.93039780521262,0.8192603594033093,0.0004978632478632479 +3,0.2671536950346751,0.8648148148148148,0.9504252400548697,0.8482860311849528,0.0004983526080898481 +4,0.22774717135307115,0.8796296296296297,0.9583333333333334,0.8655903352674731,0.0004933469513590679 +5,0.2098067749578219,0.8148148148148148,0.9573113854595336,0.7986563461528249,0.00048505044456893006 +6,0.19355002456368545,0.8962962962962963,0.9551851851851852,0.8813409456887719,0.00047357528374124746 +7,0.17832664304818863,0.9055555555555556,0.961673525377229,0.8927669952883797,0.00045907665074076113 +8,0.15717215769183943,0.8907407407407407,0.9637722908093278,0.878473874372551,0.0004417506147066312 +9,0.13631468368932986,0.9055555555555556,0.9600891632373114,0.8922466523141829,0.0004218314805536838 +10,0.12444841427107652,0.8981481481481481,0.9627366255144033,0.8857073354541486,0.00039958862040038377 +11,0.1062700557641876,0.8944444444444445,0.9630315500685871,0.8821186109794805,0.0003753228307730364 +12,0.10710507903534633,0.9037037037037037,0.9671604938271605,0.8927289968986077,0.0003493622648487805 +13,0.09017811846943238,0.8962962962962963,0.9655349794238683,0.8847960974760416,0.00032205799474680896 +14,0.07573812407178757,0.9055555555555556,0.9629080932784636,0.8930874928973532,0.00029377926388021573 +15,0.06202182715806442,0.9,0.965281207133059,0.8884266117907099,0.0002649084935722644 +16,0.06268224585801363,0.9018518518518519,0.9617764060356653,0.8890970726384194,0.00023583611146402853 +17,0.051908947073687345,0.9018518518518519,0.9618792866941014,0.8891448352212108,0.00020695527165031925 +18,0.0447326914813274,0.8981481481481481,0.9648628257887517,0.8864347088534984,0.00017865653794500142 +19,0.040650204612085454,0.9055555555555556,0.9656995884773661,0.8941211133673578,0.0001513226021754179 +20,0.038443614943669394,0.9018518518518519,0.9647599451303155,0.8900727103856783,0.00012532310893192616 +21,0.029121919320179865,0.9074074074074074,0.9637448559670783,0.8953155838742269,0.00010100965675893233 +22,0.024985664142056916,0.912962962962963,0.9632510288065844,0.9007084298907317,7.871104338774113e-05 +23,0.02308889929778301,0.9074074074074074,0.9619478737997257,0.8947195544754152,5.872881931128568e-05 +24,0.022915765022238094,0.8962962962962963,0.9614814814814815,0.883444931495246,4.133320983101688e-05 +25,0.018773427710701257,0.9018518518518519,0.9620644718792868,0.8892011773347791,2.675946072326743e-05 +26,0.019186586308746766,0.9074074074074074,0.9629561042524006,0.8950577485564994,1.5204656943687972e-05 +27,0.019509275467732012,0.9055555555555556,0.9618724279835391,0.8928436656223644,6.825057391317641e-06 +28,0.017345010422361203,0.9092592592592592,0.9616666666666667,0.8964788880738178,1.733981775034199e-06 +29,0.018072081849170037,0.9055555555555556,0.9617009602194787,0.8927865097010109,2.781588896993981e-10 diff --git a/results/airogs/resnet/metrics.json b/results/airogs/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..4a7578201e0b517d30e919633f2c3ea0e3990ba8 --- /dev/null +++ b/results/airogs/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.9, + "balanced_accuracy": 0.9, + "precision_macro": 0.9001024262211126, + "recall_macro": 0.9, + "f1_macro": 0.8999935995903738, + "precision_weighted": 0.9001024262211126, + "recall_weighted": 0.9, + "f1_weighted": 0.8999935995903738, + "cohen_kappa": 0.8, + "quadratic_weighted_kappa": 0.8, + "mcc": 0.8001024196649953, + "auroc": 0.961382, + "auprc": 0.9595785211616094, + "sensitivity": 0.908, + "specificity": 0.892, + "precision_pos": 0.8937007874015748, + "f1_pos": 0.9007936507936508, + "per_class": { + "0": { + "precision": 0.9065040650406504, + "recall": 0.892, + "f1-score": 0.8991935483870968, + "support": 500.0 + }, + "1": { + "precision": 0.8937007874015748, + "recall": 0.908, + "f1-score": 0.9007936507936508, + "support": 500.0 + }, + "accuracy": 0.9, + "macro avg": { + "precision": 0.9001024262211126, + "recall": 0.9, + "f1-score": 0.8999935995903738, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.9001024262211126, + "recall": 0.9, + "f1-score": 0.8999935995903738, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/airogs/resnet/pr.png b/results/airogs/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..cd8804c25d5b4d93c1f226f63d151bb1e5d81256 --- /dev/null +++ b/results/airogs/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61db524f69cd2f15ef17d8df77c0d41e9ec4475cf64a3c631a7ebd8c121c3257 +size 45459 diff --git a/results/airogs/resnet/roc.png b/results/airogs/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..7224ce6aff96c0f60ef2a8e85a27729574053524 --- /dev/null +++ b/results/airogs/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c27506f819783ef9ac202964970b07ddde47cd0257b18f0d090e2d52e9febf09 +size 60949 diff --git a/results/airogs/resnet/test_pred.npz b/results/airogs/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..ff2c15206d37e4b3a65544d64a9aa28f64b0ec5d --- /dev/null +++ b/results/airogs/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f70e1306ddc5c4e23bbf67aad991d222fe6c77648d37c5e7e7fbec79e162377d +size 16510 diff --git a/results/airogs/resnet/train.log b/results/airogs/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..04ae0a2d404aafbcbc1862037907e8e7475e96d9 --- /dev/null +++ b/results/airogs/resnet/train.log @@ -0,0 +1,157 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=5000 val=540 test=1000 classes=['0', '1'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6808 val_acc=0.6759 val_auc=0.7467 score=0.5907 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.4907 val_acc=0.7815 val_auc=0.8890 score=0.7428 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.3320 val_acc=0.8426 val_auc=0.9304 score=0.8193 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.2672 val_acc=0.8648 val_auc=0.9504 score=0.8483 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.2277 val_acc=0.8796 val_auc=0.9583 score=0.8656 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.2098 val_acc=0.8148 val_auc=0.9573 score=0.7987 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.1936 val_acc=0.8963 val_auc=0.9552 score=0.8813 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.1783 val_acc=0.9056 val_auc=0.9617 score=0.8928 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.1572 val_acc=0.8907 val_auc=0.9638 score=0.8785 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1363 val_acc=0.9056 val_auc=0.9601 score=0.8922 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1244 val_acc=0.8981 val_auc=0.9627 score=0.8857 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1063 val_acc=0.8944 val_auc=0.9630 score=0.8821 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.1071 val_acc=0.9037 val_auc=0.9672 score=0.8927 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.0902 val_acc=0.8963 val_auc=0.9655 score=0.8848 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0757 val_acc=0.9056 val_auc=0.9629 score=0.8931 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0620 val_acc=0.9000 val_auc=0.9653 score=0.8884 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0627 val_acc=0.9019 val_auc=0.9618 score=0.8891 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0519 val_acc=0.9019 val_auc=0.9619 score=0.8891 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0447 val_acc=0.8981 val_auc=0.9649 score=0.8864 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0407 val_acc=0.9056 val_auc=0.9657 score=0.8941 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0384 val_acc=0.9019 val_auc=0.9648 score=0.8901 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0291 val_acc=0.9074 val_auc=0.9637 score=0.8953 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0250 val_acc=0.9130 val_auc=0.9633 score=0.9007 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0231 val_acc=0.9074 val_auc=0.9619 score=0.8947 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0229 val_acc=0.8963 val_auc=0.9615 score=0.8834 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0188 val_acc=0.9019 val_auc=0.9621 score=0.8892 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0192 val_acc=0.9074 val_auc=0.9630 score=0.8951 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0195 val_acc=0.9056 val_auc=0.9619 score=0.8928 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0173 val_acc=0.9093 val_auc=0.9617 score=0.8965 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0181 val_acc=0.9056 val_auc=0.9617 score=0.8928 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=22 best_val_score=0.9007 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/airogs/resnet acc=0.9000 auroc=0.961382 f1_macro=0.9000 qwk=0.8 diff --git a/results/airogs/retfound/confusion_matrix.png b/results/airogs/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7a69b6a5f996d8983b1615ac77be7eff01ed2cf7 --- /dev/null +++ b/results/airogs/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b42880b2ed890eff5ed141947fe5dbe728465da48a286e573325f6b941269d0f +size 66364 diff --git a/results/airogs/retfound/confusion_matrix_test.jpg b/results/airogs/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..25d3f0b897147fab7387e85f28310c7b7f1eac96 --- /dev/null +++ b/results/airogs/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d518ea0035d9db32a50fcbb1896fa44a1e98fc6d69671c646a28a12caeb6a51 +size 260464 diff --git a/results/airogs/retfound/log.txt b/results/airogs/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..0c8d66950394ee444a80aec02df2b748b374444c --- /dev/null +++ b/results/airogs/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 3.104967948717949e-05, "train_loss": 0.6869127811529697, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.354967948717946e-05, "train_loss": 0.6098002164791791, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.0001560496794871795, "train_loss": 0.5515275139075059, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021854967948717953, "train_loss": 0.5289115004050426, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002810496794871795, "train_loss": 0.5009684100365027, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00034354967948717954, "train_loss": 0.5024133649392005, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040604967948717954, "train_loss": 0.49966024053402436, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004685496794871794, "train_loss": 0.4900955099325914, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005310496794871795, "train_loss": 0.48253656178712845, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005935496794871795, "train_loss": 0.4739962577437743, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006237308188025447, "train_loss": 0.4772948145102232, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006161042181059726, "train_loss": 0.4518326301223192, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006010141856144047, "train_loss": 0.46381485901581937, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005788322880087305, "train_loss": 0.4572136507202417, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005501047172254988, "train_loss": 0.4376392687360446, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005155388413987338, "train_loss": 0.4237176972704056, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0004759857871088555, "train_loss": 0.43037567211267275, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004324194818215422, "train_loss": 0.4216204109864357, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003859126725612448, "train_loss": 0.42659685015678406, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0003376105113192029, "train_loss": 0.40269207667845947, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.000288702357610878, "train_loss": 0.4008018556886759, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024039249249742462, "train_loss": 0.3933946737685265, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0001938704651891872, "train_loss": 0.38881010562181473, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015028180239625032, "train_loss": 0.381559801980471, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011069980165905447, "train_loss": 0.38073944586973923, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.609910320087817e-05, "train_loss": 0.38250331026621354, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.733169105089629e-05, "train_loss": 0.37621724328551537, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5105914369808592e-05, "train_loss": 0.37985354212996286, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 9.969045542627247e-06, "train_loss": 0.37229995333995575, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.2938045162649166e-06, "train_loss": 0.36475676097548926, "epoch": 29, "n_parameters": 303303682} diff --git a/results/airogs/retfound/metrics.json b/results/airogs/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..80482b59cdf02c496623c42cd326f98b872d3481 --- /dev/null +++ b/results/airogs/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.908, + "balanced_accuracy": 0.908, + "precision_macro": 0.9084182202575437, + "recall_macro": 0.908, + "f1_macro": 0.9079764419691441, + "precision_weighted": 0.9084182202575437, + "recall_weighted": 0.908, + "f1_weighted": 0.9079764419691442, + "cohen_kappa": 0.8160000000000001, + "quadratic_weighted_kappa": 0.8160000000000001, + "mcc": 0.8164181131383057, + "auroc": 0.970758, + "auprc": 0.9715458663728014, + "sensitivity": 0.892, + "specificity": 0.924, + "precision_pos": 0.9214876033057852, + "f1_pos": 0.9065040650406504, + "per_class": { + "0": { + "precision": 0.8953488372093024, + "recall": 0.924, + "f1-score": 0.9094488188976378, + "support": 500.0 + }, + "1": { + "precision": 0.9214876033057852, + "recall": 0.892, + "f1-score": 0.9065040650406504, + "support": 500.0 + }, + "accuracy": 0.908, + "macro avg": { + "precision": 0.9084182202575437, + "recall": 0.908, + "f1-score": 0.9079764419691441, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.9084182202575437, + "recall": 0.908, + "f1-score": 0.9079764419691442, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/airogs/retfound/metrics_test.csv b/results/airogs/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..e45cb3eb84acd227645a97cd8b0ba8b686ffe152 --- /dev/null +++ b/results/airogs/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.2516249555628747,0.908,0.9079764419691441,0.9707290000000001,0.092,0.8314656502892113,0.9084182202575437,0.908,0.9703819077045409,0.8160000000000001 diff --git a/results/airogs/retfound/metrics_val.csv b/results/airogs/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..952953a656d4fee56aace064ad323e3d4752a0ca --- /dev/null +++ b/results/airogs/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6267830974915448,0.7444444444444445,0.7425764463666387,0.8363786008230453,0.25555555555555554,0.5910378348136234,0.7517518083182639,0.7444444444444445,0.8347590615489223,0.48888888888888893 +0.36065684171283946,0.85,0.8499377313631513,0.923576817558299,0.15,0.739048557792073,0.8505818986246032,0.85,0.923376540042333,0.7 +0.30178015372332406,0.8796296296296297,0.8794473122932831,0.9466872427983539,0.12037037037037036,0.7848647643222898,0.8819401316606632,0.8796296296296297,0.946405607115115,0.7592592592592593 +0.26234125126810637,0.8888888888888888,0.8888873647100783,0.9587654320987654,0.1111111111111111,0.7999977777530862,0.8889102282704127,0.8888888888888888,0.9599995369798717,0.7777777777777778 +0.2907397606793572,0.8814814814814815,0.8813186813186813,0.9527469135802469,0.11851851851851852,0.787846856340007,0.8835862068965518,0.8814814814814815,0.9541414159026794,0.762962962962963 +0.27061493037378087,0.8944444444444445,0.8941798941798942,0.9617318244170097,0.10555555555555556,0.8086538461538462,0.8984287317620651,0.8944444444444444,0.9627803186031139,0.7888888888888889 +0.25615671759142594,0.8925925925925926,0.8925557461406518,0.96340877914952,0.10740740740740741,0.8059658073755818,0.8931318681318681,0.8925925925925926,0.9640757048493638,0.7851851851851852 +0.24167594839544856,0.8907407407407407,0.8905752526969799,0.9655452674897119,0.10925925925925926,0.8027625851099454,0.8931188672214632,0.8907407407407407,0.9666023085851674,0.7814814814814814 +0.25277270420509224,0.9018518518518519,0.9017031931006769,0.9657064471879286,0.09814814814814815,0.821023439101615,0.9042976027822631,0.9018518518518519,0.9664421166860533,0.8037037037037037 +0.26290999637807116,0.8907407407407407,0.8905060572214107,0.9655521262002744,0.10925925925925926,0.8026612615027249,0.8941196817710135,0.8907407407407407,0.9669754289687871,0.7814814814814814 +0.25161123801680174,0.9018518518518519,0.9018353564213465,0.9649142661179699,0.09814814814814815,0.8212229806598408,0.9021221397098187,0.9018518518518519,0.9658003944161854,0.8037037037037037 +0.23271823586786494,0.912962962962963,0.9129483349396846,0.9687551440329218,0.08703703703703704,0.8398409381168002,0.9132407242179242,0.912962962962963,0.9692950646828473,0.825925925925926 +0.22827284958432703,0.9037037037037037,0.903638934263085,0.9681035665294924,0.0962962962962963,0.8242260212180388,0.9047920334507042,0.9037037037037037,0.9682872052596343,0.8074074074074074 +0.23409848265788136,0.912962962962963,0.9129483349396846,0.9672187928669409,0.08703703703703704,0.8398409381168002,0.9132407242179242,0.912962962962963,0.9680735164326912,0.825925925925926 +0.23617559057824752,0.9,0.8999876527966415,0.9679492455418381,0.1,0.8181632653061224,0.900197628458498,0.8999999999999999,0.9691229672185688,0.8 +0.22935323767802296,0.9185185185185185,0.9185084578343006,0.9698319615912209,0.08148148148148149,0.8492991613395109,0.9187252964426877,0.9185185185185185,0.9709637428818148,0.837037037037037 +0.2348311854635968,0.9203703703703704,0.9203241933768199,0.9689231824417011,0.07962962962962963,0.8524137525020605,0.9213471559582571,0.9203703703703703,0.9700474266686755,0.8407407407407408 +0.25106555968523026,0.9074074074074074,0.9074061372583986,0.967681755829904,0.09259259259259259,0.8305065269350984,0.9074297629499561,0.9074074074074074,0.9684479216576163,0.8148148148148149 +0.23531180181924036,0.9203703703703704,0.9202913724507484,0.9707510288065844,0.07962962962962963,0.8523616018845701,0.9220434920328875,0.9203703703703703,0.9716494849123236,0.8407407407407408 +0.21684428611222437,0.9148148148148149,0.9148148148148149,0.9712791495198903,0.08518518518518518,0.8430034129692833,0.9148148148148149,0.9148148148148149,0.9723191297072907,0.8296296296296296 +0.2322498717728783,0.9111111111111111,0.9111098917680627,0.9698285322359397,0.08888888888888889,0.8367328049979754,0.9111336698858647,0.9111111111111112,0.9704452446619843,0.8222222222222222 +0.23778783990179791,0.9166666666666666,0.9165404469722729,0.9706310013717421,0.08333333333333333,0.8459553402148725,0.9192025835299962,0.9166666666666667,0.9709067553336702,0.8333333333333334 +0.2386385027100058,0.9185185185185185,0.9185084578343006,0.9701268861454047,0.08148148148148149,0.8492991613395109,0.9187252964426877,0.9185185185185185,0.970491594177331,0.837037037037037 +0.23999610104981592,0.9222222222222223,0.9221699084432609,0.9707338820301783,0.07777777777777778,0.8555865393704509,0.9233604753521127,0.9222222222222223,0.9711381442387186,0.8444444444444444 +0.23967866029809504,0.9203703703703704,0.9203088803088804,0.9703326474622771,0.07962962962962963,0.8523894201328371,0.9216718266253869,0.9203703703703704,0.9701346387665579,0.8407407407407408 +0.23302041739225388,0.9166666666666666,0.9166663808860798,0.9717661179698217,0.08333333333333333,0.846153396605732,0.9166723823372063,0.9166666666666667,0.9719504025999759,0.8333333333333334 +0.23026803880929947,0.924074074074074,0.9240529776789849,0.9718792866941015,0.07592592592592592,0.8588301528979495,0.9245457916203189,0.924074074074074,0.9721254668793979,0.8481481481481481 +0.22829091548919678,0.9277777777777778,0.927757710475132,0.9717832647462278,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.9718737444256429,0.8555555555555556 +0.22767887702759573,0.9277777777777778,0.927757710475132,0.971786694101509,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.971843318291873,0.8555555555555556 +0.2282544427058276,0.9277777777777778,0.927757710475132,0.9718518518518517,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.9719820208335395,0.8555555555555556 diff --git a/results/airogs/retfound/pr.png b/results/airogs/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a7b8722b6e008818786512019ce060492f13009a --- /dev/null +++ b/results/airogs/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0814fce5a15fb592055797d516f1aae0378941c53a5d9790107042caa8eaf745 +size 43330 diff --git a/results/airogs/retfound/roc.png b/results/airogs/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..d44878c06bb7944fb573ac6929feec5044d0a255 --- /dev/null +++ b/results/airogs/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96cf71dccc3e6d1955819e049dad85405cec7cab0b5e796170e7e06c8d26b874 +size 60366 diff --git a/results/airogs/retfound/test_pred.npz b/results/airogs/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..63f31414cdb2125dff4da4fb0d029def414291ff --- /dev/null +++ b/results/airogs/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:465d825381cbf129f91200b2a0d6f9cd0c6592b691865e383f9f66e830b2ab0c +size 12510 diff --git a/results/airogs/retfound/train.log b/results/airogs/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0eb6150f3d954e1d13cfca55082552e3dace92f7 --- /dev/null +++ b/results/airogs/retfound/train.log @@ -0,0 +1,716 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:52:23.093345041 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:52:24.582098] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:52:24.582378] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/Glaucoma/eyepacs-airogs-light', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/airogs', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:52:31.637533] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:36.886581] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:38.245461] Sampler_train = +[13:52:38.288706] len of train_set: 4992 +[13:52:38.960732] [Adaptation] Full fine-tuning: training all parameters. +[13:52:38.962191] number of trainable params (M): 303.30 +[13:52:38.962291] base lr: 5.00e-03 +[13:52:38.962369] actual lr: 6.25e-04 +[13:52:38.962438] accumulate grad iterations: 1 +[13:52:38.962508] effective batch size: 32 +[13:52:38.966035] criterion = CrossEntropyLoss() +[13:52:38.966141] Start training for 30 epochs +[13:52:38.968658] log_dir: ./output_logs/retfound +[13:52:41.395490] Epoch: [0] [ 0/156] eta: 0:06:18 lr: 0.000000 loss: 0.6929 (0.6929) time: 2.4232 data: 1.0332 max mem: 7340 +[13:52:44.363108] Epoch: [0] [ 20/156] eta: 0:00:34 lr: 0.000008 loss: 0.6930 (0.6929) time: 0.1483 data: 0.0002 max mem: 9669 +[13:52:47.214439] Epoch: [0] [ 40/156] eta: 0:00:23 lr: 0.000016 loss: 0.6920 (0.6928) time: 0.1425 data: 0.0002 max mem: 9669 +[13:52:50.063763] Epoch: [0] [ 60/156] eta: 0:00:17 lr: 0.000024 loss: 0.6920 (0.6928) time: 0.1424 data: 0.0002 max mem: 9669 +[13:52:52.898232] Epoch: [0] [ 80/156] eta: 0:00:13 lr: 0.000032 loss: 0.6909 (0.6920) time: 0.1417 data: 0.0002 max mem: 9669 +[13:52:55.387001] Epoch: [0] [100/156] eta: 0:00:09 lr: 0.000040 loss: 0.6899 (0.6917) time: 0.1244 data: 0.0002 max mem: 9669 +[13:52:58.232687] Epoch: [0] [120/156] eta: 0:00:05 lr: 0.000048 loss: 0.6860 (0.6908) time: 0.1422 data: 0.0002 max mem: 9669 +[13:53:01.079729] Epoch: [0] [140/156] eta: 0:00:02 lr: 0.000056 loss: 0.6767 (0.6888) time: 0.1423 data: 0.0002 max mem: 9669 +[13:53:03.216255] Epoch: [0] [155/156] eta: 0:00:00 lr: 0.000062 loss: 0.6643 (0.6869) time: 0.1426 data: 0.0001 max mem: 9669 +[13:53:03.363542] Epoch: [0] Total time: 0:00:24 (0.1564 s / it) +[13:53:03.372222] Averaged stats: lr: 0.000062 loss: 0.6643 (0.6869) +[13:53:04.338277] val: [ 0/17] eta: 0:00:16 loss: 0.6420 (0.6420) time: 0.9530 data: 0.9048 max mem: 9669 +[13:53:04.689602] val: [10/17] eta: 0:00:00 loss: 0.6479 (0.6470) time: 0.1185 data: 0.0827 max mem: 9669 +[13:53:04.974467] val: [16/17] eta: 0:00:00 loss: 0.6328 (0.6268) time: 0.0934 data: 0.0536 max mem: 9669 +[13:53:05.083131] val: Total time: 0:00:01 (0.0999 s / it) +[13:53:05.102235] val loss: 0.6267830974915448 +[13:53:05.102488] Accuracy: 0.7444, F1 Score: 0.7426, ROC AUC: 0.8364, Hamming Loss: 0.2556, + Jaccard Score: 0.5910, Precision: 0.7518, Recall: 0.7444, + Average Precision: 0.8348, Kappa: 0.4889, Score: 0.6893 +[13:53:06.912673] Best epoch = 0, Best score = 0.6893 +[13:53:06.990615] log_dir: ./output_logs/retfound +[13:53:08.122482] Epoch: [1] [ 0/156] eta: 0:02:56 lr: 0.000063 loss: 0.6670 (0.6670) time: 1.1309 data: 0.9324 max mem: 9669 +[13:53:10.974458] Epoch: [1] [ 20/156] eta: 0:00:25 lr: 0.000071 loss: 0.6449 (0.6479) time: 0.1426 data: 0.0002 max mem: 9669 +[13:53:13.811828] Epoch: [1] [ 40/156] eta: 0:00:19 lr: 0.000079 loss: 0.6459 (0.6417) time: 0.1418 data: 0.0001 max mem: 9669 +[13:53:16.656392] Epoch: [1] [ 60/156] eta: 0:00:15 lr: 0.000087 loss: 0.6346 (0.6419) time: 0.1422 data: 0.0002 max mem: 9669 +[13:53:19.504205] Epoch: [1] [ 80/156] eta: 0:00:11 lr: 0.000095 loss: 0.6075 (0.6338) time: 0.1423 data: 0.0002 max mem: 9669 +[13:53:22.362911] Epoch: [1] [100/156] eta: 0:00:08 lr: 0.000103 loss: 0.5754 (0.6276) time: 0.1429 data: 0.0002 max mem: 9669 +[13:53:25.215046] Epoch: [1] [120/156] eta: 0:00:05 lr: 0.000111 loss: 0.5869 (0.6222) time: 0.1426 data: 0.0002 max mem: 9669 +[13:53:28.060807] Epoch: [1] [140/156] eta: 0:00:02 lr: 0.000119 loss: 0.5824 (0.6155) time: 0.1422 data: 0.0001 max mem: 9669 +[13:53:30.192188] Epoch: [1] [155/156] eta: 0:00:00 lr: 0.000125 loss: 0.5344 (0.6098) time: 0.1421 data: 0.0001 max mem: 9669 +[13:53:30.321490] Epoch: [1] Total time: 0:00:23 (0.1496 s / it) +[13:53:30.331294] Averaged stats: lr: 0.000125 loss: 0.5344 (0.6098) +[13:53:31.278994] val: [ 0/17] eta: 0:00:15 loss: 0.2565 (0.2565) time: 0.9393 data: 0.9057 max mem: 9669 +[13:53:31.625790] val: [10/17] eta: 0:00:00 loss: 0.3011 (0.3592) time: 0.1169 data: 0.0825 max mem: 9669 +[13:53:31.820467] val: [16/17] eta: 0:00:00 loss: 0.3327 (0.3607) time: 0.0870 data: 0.0534 max mem: 9669 +[13:53:31.945759] val: Total time: 0:00:01 (0.0945 s / it) +[13:53:31.961685] val loss: 0.36065684171283946 +[13:53:31.962004] Accuracy: 0.8500, F1 Score: 0.8499, ROC AUC: 0.9236, Hamming Loss: 0.1500, + Jaccard Score: 0.7390, Precision: 0.8506, Recall: 0.8500, + Average Precision: 0.9234, Kappa: 0.7000, Score: 0.8245 +[13:53:33.782439] Best epoch = 1, Best score = 0.8245 +[13:53:33.858897] log_dir: ./output_logs/retfound +[13:53:35.052659] Epoch: [2] [ 0/156] eta: 0:03:06 lr: 0.000125 loss: 0.5732 (0.5732) time: 1.1927 data: 1.0531 max mem: 9669 +[13:53:37.907418] Epoch: [2] [ 20/156] eta: 0:00:26 lr: 0.000133 loss: 0.5497 (0.5636) time: 0.1427 data: 0.0002 max mem: 9669 +[13:53:40.759191] Epoch: [2] [ 40/156] eta: 0:00:19 lr: 0.000141 loss: 0.5511 (0.5642) time: 0.1425 data: 0.0002 max mem: 9669 +[13:53:43.599907] Epoch: [2] [ 60/156] eta: 0:00:15 lr: 0.000149 loss: 0.5691 (0.5663) time: 0.1420 data: 0.0002 max mem: 9669 +[13:53:46.445910] Epoch: [2] [ 80/156] eta: 0:00:11 lr: 0.000157 loss: 0.5262 (0.5577) time: 0.1423 data: 0.0002 max mem: 9669 +[13:53:49.305334] Epoch: [2] [100/156] eta: 0:00:08 lr: 0.000165 loss: 0.5398 (0.5551) time: 0.1429 data: 0.0002 max mem: 9669 +[13:53:52.148579] Epoch: [2] [120/156] eta: 0:00:05 lr: 0.000173 loss: 0.5339 (0.5563) time: 0.1421 data: 0.0002 max mem: 9669 +[13:53:54.994128] Epoch: [2] [140/156] eta: 0:00:02 lr: 0.000181 loss: 0.5275 (0.5525) time: 0.1422 data: 0.0001 max mem: 9669 +[13:53:57.128410] Epoch: [2] [155/156] eta: 0:00:00 lr: 0.000187 loss: 0.5123 (0.5515) time: 0.1425 data: 0.0001 max mem: 9669 +[13:53:57.245475] Epoch: [2] Total time: 0:00:23 (0.1499 s / it) +[13:53:57.254263] Averaged stats: lr: 0.000187 loss: 0.5123 (0.5515) +[13:53:58.278272] val: [ 0/17] eta: 0:00:17 loss: 0.1746 (0.1746) time: 1.0076 data: 0.9732 max mem: 9669 +[13:53:58.626204] val: [10/17] eta: 0:00:00 loss: 0.2856 (0.2840) time: 0.1232 data: 0.0886 max mem: 9669 +[13:53:58.831212] val: [16/17] eta: 0:00:00 loss: 0.3181 (0.3018) time: 0.0917 data: 0.0574 max mem: 9669 +[13:53:58.954388] val: Total time: 0:00:01 (0.0991 s / it) +[13:53:58.977600] val loss: 0.30178015372332406 +[13:53:58.977812] Accuracy: 0.8796, F1 Score: 0.8794, ROC AUC: 0.9467, Hamming Loss: 0.1204, + Jaccard Score: 0.7849, Precision: 0.8819, Recall: 0.8796, + Average Precision: 0.9464, Kappa: 0.7593, Score: 0.8618 +[13:54:00.882868] Best epoch = 2, Best score = 0.8618 +[13:54:00.954549] log_dir: ./output_logs/retfound +[13:54:02.025738] Epoch: [3] [ 0/156] eta: 0:02:46 lr: 0.000188 loss: 0.5378 (0.5378) time: 1.0702 data: 0.9241 max mem: 9669 +[13:54:04.875556] Epoch: [3] [ 20/156] eta: 0:00:25 lr: 0.000196 loss: 0.5189 (0.5252) time: 0.1424 data: 0.0002 max mem: 9669 +[13:54:07.724782] Epoch: [3] [ 40/156] eta: 0:00:19 lr: 0.000204 loss: 0.4852 (0.5192) time: 0.1424 data: 0.0002 max mem: 9669 +[13:54:10.581770] Epoch: [3] [ 60/156] eta: 0:00:15 lr: 0.000212 loss: 0.5251 (0.5225) time: 0.1428 data: 0.0002 max mem: 9669 +[13:54:13.432572] Epoch: [3] [ 80/156] eta: 0:00:11 lr: 0.000220 loss: 0.5494 (0.5297) time: 0.1425 data: 0.0002 max mem: 9669 +[13:54:16.281449] Epoch: [3] [100/156] eta: 0:00:08 lr: 0.000228 loss: 0.4800 (0.5266) time: 0.1424 data: 0.0002 max mem: 9669 +[13:54:19.131196] Epoch: [3] [120/156] eta: 0:00:05 lr: 0.000236 loss: 0.4913 (0.5219) time: 0.1424 data: 0.0002 max mem: 9669 +[13:54:21.976702] Epoch: [3] [140/156] eta: 0:00:02 lr: 0.000244 loss: 0.5545 (0.5290) time: 0.1422 data: 0.0002 max mem: 9669 +[13:54:24.116737] Epoch: [3] [155/156] eta: 0:00:00 lr: 0.000250 loss: 0.5226 (0.5289) time: 0.1424 data: 0.0001 max mem: 9669 +[13:54:24.237422] Epoch: [3] Total time: 0:00:23 (0.1492 s / it) +[13:54:24.245365] Averaged stats: lr: 0.000250 loss: 0.5226 (0.5289) +[13:54:25.272937] val: [ 0/17] eta: 0:00:17 loss: 0.1953 (0.1953) time: 1.0139 data: 0.9788 max mem: 9669 +[13:54:25.618721] val: [10/17] eta: 0:00:00 loss: 0.2957 (0.2924) time: 0.1235 data: 0.0891 max mem: 9669 +[13:54:25.822904] val: [16/17] eta: 0:00:00 loss: 0.2227 (0.2623) time: 0.0919 data: 0.0577 max mem: 9669 +[13:54:25.945494] val: Total time: 0:00:01 (0.0993 s / it) +[13:54:25.969642] val loss: 0.26234125126810637 +[13:54:25.969964] Accuracy: 0.8889, F1 Score: 0.8889, ROC AUC: 0.9588, Hamming Loss: 0.1111, + Jaccard Score: 0.8000, Precision: 0.8889, Recall: 0.8889, + Average Precision: 0.9600, Kappa: 0.7778, Score: 0.8751 +[13:54:27.800378] Best epoch = 3, Best score = 0.8751 +[13:54:27.864235] log_dir: ./output_logs/retfound +[13:54:29.013635] Epoch: [4] [ 0/156] eta: 0:02:59 lr: 0.000250 loss: 0.5731 (0.5731) time: 1.1484 data: 1.0049 max mem: 9669 +[13:54:31.859977] Epoch: [4] [ 20/156] eta: 0:00:25 lr: 0.000258 loss: 0.5041 (0.5161) time: 0.1423 data: 0.0002 max mem: 9669 +[13:54:34.707968] Epoch: [4] [ 40/156] eta: 0:00:19 lr: 0.000266 loss: 0.5131 (0.5237) time: 0.1424 data: 0.0001 max mem: 9669 +[13:54:37.550266] Epoch: [4] [ 60/156] eta: 0:00:15 lr: 0.000274 loss: 0.5255 (0.5249) time: 0.1421 data: 0.0002 max mem: 9669 +[13:54:40.392898] Epoch: [4] [ 80/156] eta: 0:00:11 lr: 0.000282 loss: 0.4897 (0.5184) time: 0.1421 data: 0.0002 max mem: 9669 +[13:54:43.238037] Epoch: [4] [100/156] eta: 0:00:08 lr: 0.000290 loss: 0.4656 (0.5108) time: 0.1422 data: 0.0002 max mem: 9669 +[13:54:46.085535] Epoch: [4] [120/156] eta: 0:00:05 lr: 0.000298 loss: 0.4011 (0.4958) time: 0.1423 data: 0.0001 max mem: 9669 +[13:54:48.938586] Epoch: [4] [140/156] eta: 0:00:02 lr: 0.000306 loss: 0.5319 (0.5011) time: 0.1426 data: 0.0002 max mem: 9669 +[13:54:51.070452] Epoch: [4] [155/156] eta: 0:00:00 lr: 0.000312 loss: 0.4962 (0.5010) time: 0.1421 data: 0.0002 max mem: 9669 +[13:54:51.199201] Epoch: [4] Total time: 0:00:23 (0.1496 s / it) +[13:54:51.208901] Averaged stats: lr: 0.000312 loss: 0.4962 (0.5010) +[13:54:52.138874] val: [ 0/17] eta: 0:00:15 loss: 0.1446 (0.1446) time: 0.9229 data: 0.8897 max mem: 9669 +[13:54:52.489329] val: [10/17] eta: 0:00:00 loss: 0.2819 (0.2813) time: 0.1157 data: 0.0816 max mem: 9669 +[13:54:52.695387] val: [16/17] eta: 0:00:00 loss: 0.2819 (0.2907) time: 0.0869 data: 0.0529 max mem: 9669 +[13:54:52.823107] val: Total time: 0:00:01 (0.0946 s / it) +[13:54:52.837677] val loss: 0.2907397606793572 +[13:54:52.837967] Accuracy: 0.8815, F1 Score: 0.8813, ROC AUC: 0.9527, Hamming Loss: 0.1185, + Jaccard Score: 0.7878, Precision: 0.8836, Recall: 0.8815, + Average Precision: 0.9541, Kappa: 0.7630, Score: 0.8657 +[13:54:52.880181] Best epoch = 3, Best score = 0.8751 +[13:54:53.150498] log_dir: ./output_logs/retfound +[13:54:54.235572] Epoch: [5] [ 0/156] eta: 0:02:49 lr: 0.000313 loss: 0.5227 (0.5227) time: 1.0842 data: 0.9375 max mem: 9669 +[13:54:57.087326] Epoch: [5] [ 20/156] eta: 0:00:25 lr: 0.000321 loss: 0.4854 (0.4888) time: 0.1425 data: 0.0002 max mem: 9669 +[13:54:59.933755] Epoch: [5] [ 40/156] eta: 0:00:19 lr: 0.000329 loss: 0.4325 (0.4803) time: 0.1423 data: 0.0002 max mem: 9669 +[13:55:02.778826] Epoch: [5] [ 60/156] eta: 0:00:15 lr: 0.000337 loss: 0.4913 (0.4887) time: 0.1422 data: 0.0001 max mem: 9669 +[13:55:05.303518] Epoch: [5] [ 80/156] eta: 0:00:11 lr: 0.000345 loss: 0.5478 (0.5029) time: 0.1262 data: 0.0001 max mem: 9669 +[13:55:08.151695] Epoch: [5] [100/156] eta: 0:00:08 lr: 0.000353 loss: 0.4967 (0.4999) time: 0.1424 data: 0.0002 max mem: 9669 +[13:55:11.001042] Epoch: [5] [120/156] eta: 0:00:05 lr: 0.000361 loss: 0.5006 (0.5023) time: 0.1424 data: 0.0001 max mem: 9669 +[13:55:13.849678] Epoch: [5] [140/156] eta: 0:00:02 lr: 0.000369 loss: 0.4839 (0.5014) time: 0.1424 data: 0.0001 max mem: 9669 +[13:55:15.992142] Epoch: [5] [155/156] eta: 0:00:00 lr: 0.000375 loss: 0.5141 (0.5024) time: 0.1425 data: 0.0001 max mem: 9669 +[13:55:16.130040] Epoch: [5] Total time: 0:00:22 (0.1473 s / it) +[13:55:16.139584] Averaged stats: lr: 0.000375 loss: 0.5141 (0.5024) +[13:55:16.960636] val: [ 0/17] eta: 0:00:13 loss: 0.3622 (0.3622) time: 0.8045 data: 0.7711 max mem: 9669 +[13:55:17.432604] val: [10/17] eta: 0:00:00 loss: 0.3603 (0.3440) time: 0.1160 data: 0.0815 max mem: 9669 +[13:55:17.635012] val: [16/17] eta: 0:00:00 loss: 0.2352 (0.2706) time: 0.0869 data: 0.0528 max mem: 9669 +[13:55:17.749395] val: Total time: 0:00:01 (0.0938 s / it) +[13:55:17.763496] val loss: 0.27061493037378087 +[13:55:17.763760] Accuracy: 0.8944, F1 Score: 0.8942, ROC AUC: 0.9617, Hamming Loss: 0.1056, + Jaccard Score: 0.8087, Precision: 0.8984, Recall: 0.8944, + Average Precision: 0.9628, Kappa: 0.7889, Score: 0.8816 +[13:55:19.751322] Best epoch = 5, Best score = 0.8816 +[13:55:19.809192] log_dir: ./output_logs/retfound +[13:55:20.924269] Epoch: [6] [ 0/156] eta: 0:02:53 lr: 0.000375 loss: 0.5117 (0.5117) time: 1.1141 data: 0.9664 max mem: 9669 +[13:55:23.772400] Epoch: [6] [ 20/156] eta: 0:00:25 lr: 0.000383 loss: 0.4926 (0.5112) time: 0.1424 data: 0.0002 max mem: 9669 +[13:55:26.619109] Epoch: [6] [ 40/156] eta: 0:00:19 lr: 0.000391 loss: 0.4482 (0.4945) time: 0.1423 data: 0.0001 max mem: 9669 +[13:55:29.440913] Epoch: [6] [ 60/156] eta: 0:00:15 lr: 0.000399 loss: 0.5284 (0.5101) time: 0.1410 data: 0.0002 max mem: 9669 +[13:55:32.298061] Epoch: [6] [ 80/156] eta: 0:00:11 lr: 0.000407 loss: 0.4940 (0.5119) time: 0.1428 data: 0.0002 max mem: 9669 +[13:55:35.141464] Epoch: [6] [100/156] eta: 0:00:08 lr: 0.000415 loss: 0.4631 (0.5055) time: 0.1421 data: 0.0002 max mem: 9669 +[13:55:37.985736] Epoch: [6] [120/156] eta: 0:00:05 lr: 0.000423 loss: 0.4742 (0.5046) time: 0.1422 data: 0.0002 max mem: 9669 +[13:55:40.835281] Epoch: [6] [140/156] eta: 0:00:02 lr: 0.000431 loss: 0.4733 (0.5021) time: 0.1424 data: 0.0002 max mem: 9669 +[13:55:42.966908] Epoch: [6] [155/156] eta: 0:00:00 lr: 0.000437 loss: 0.4729 (0.4997) time: 0.1422 data: 0.0001 max mem: 9669 +[13:55:43.104983] Epoch: [6] Total time: 0:00:23 (0.1493 s / it) +[13:55:43.114456] Averaged stats: lr: 0.000437 loss: 0.4729 (0.4997) +[13:55:43.885626] val: [ 0/17] eta: 0:00:12 loss: 0.3536 (0.3536) time: 0.7597 data: 0.7251 max mem: 9669 +[13:55:44.252320] val: [10/17] eta: 0:00:00 loss: 0.2742 (0.3143) time: 0.1023 data: 0.0679 max mem: 9669 +[13:55:44.450208] val: [16/17] eta: 0:00:00 loss: 0.2401 (0.2562) time: 0.0778 data: 0.0440 max mem: 9669 +[13:55:44.812462] val: Total time: 0:00:01 (0.0992 s / it) +[13:55:44.826756] val loss: 0.25615671759142594 +[13:55:44.827058] Accuracy: 0.8926, F1 Score: 0.8926, ROC AUC: 0.9634, Hamming Loss: 0.1074, + Jaccard Score: 0.8060, Precision: 0.8931, Recall: 0.8926, + Average Precision: 0.9641, Kappa: 0.7852, Score: 0.8804 +[13:55:44.871860] Best epoch = 5, Best score = 0.8816 +[13:55:45.141890] log_dir: ./output_logs/retfound +[13:55:46.206399] Epoch: [7] [ 0/156] eta: 0:02:45 lr: 0.000438 loss: 0.5038 (0.5038) time: 1.0634 data: 0.9182 max mem: 9669 +[13:55:49.058186] Epoch: [7] [ 20/156] eta: 0:00:25 lr: 0.000446 loss: 0.5144 (0.5279) time: 0.1425 data: 0.0002 max mem: 9669 +[13:55:51.906508] Epoch: [7] [ 40/156] eta: 0:00:19 lr: 0.000454 loss: 0.4662 (0.5053) time: 0.1424 data: 0.0001 max mem: 9669 +[13:55:54.759114] Epoch: [7] [ 60/156] eta: 0:00:15 lr: 0.000462 loss: 0.5224 (0.5106) time: 0.1426 data: 0.0002 max mem: 9669 +[13:55:57.613879] Epoch: [7] [ 80/156] eta: 0:00:11 lr: 0.000470 loss: 0.4508 (0.5010) time: 0.1427 data: 0.0001 max mem: 9669 +[13:56:00.469639] Epoch: [7] [100/156] eta: 0:00:08 lr: 0.000478 loss: 0.4589 (0.4949) time: 0.1427 data: 0.0002 max mem: 9669 +[13:56:03.318132] Epoch: [7] [120/156] eta: 0:00:05 lr: 0.000486 loss: 0.4859 (0.4906) time: 0.1424 data: 0.0002 max mem: 9669 +[13:56:06.177076] Epoch: [7] [140/156] eta: 0:00:02 lr: 0.000494 loss: 0.4952 (0.4913) time: 0.1429 data: 0.0002 max mem: 9669 +[13:56:08.315388] Epoch: [7] [155/156] eta: 0:00:00 lr: 0.000500 loss: 0.4568 (0.4901) time: 0.1428 data: 0.0001 max mem: 9669 +[13:56:08.436039] Epoch: [7] Total time: 0:00:23 (0.1493 s / it) +[13:56:08.444602] Averaged stats: lr: 0.000500 loss: 0.4568 (0.4901) +[13:56:09.373712] val: [ 0/17] eta: 0:00:15 loss: 0.3089 (0.3089) time: 0.9205 data: 0.8858 max mem: 9669 +[13:56:09.715665] val: [10/17] eta: 0:00:00 loss: 0.2991 (0.2929) time: 0.1147 data: 0.0809 max mem: 9669 +[13:56:09.920766] val: [16/17] eta: 0:00:00 loss: 0.2143 (0.2417) time: 0.0863 data: 0.0524 max mem: 9669 +[13:56:10.057973] val: Total time: 0:00:01 (0.0944 s / it) +[13:56:10.072173] val loss: 0.24167594839544856 +[13:56:10.072416] Accuracy: 0.8907, F1 Score: 0.8906, ROC AUC: 0.9655, Hamming Loss: 0.1093, + Jaccard Score: 0.8028, Precision: 0.8931, Recall: 0.8907, + Average Precision: 0.9666, Kappa: 0.7815, Score: 0.8792 +[13:56:10.114218] Best epoch = 5, Best score = 0.8816 +[13:56:10.395372] log_dir: ./output_logs/retfound +[13:56:11.550437] Epoch: [8] [ 0/156] eta: 0:03:00 lr: 0.000500 loss: 0.3634 (0.3634) time: 1.1541 data: 1.0095 max mem: 9669 +[13:56:14.399205] Epoch: [8] [ 20/156] eta: 0:00:25 lr: 0.000508 loss: 0.4905 (0.4941) time: 0.1424 data: 0.0002 max mem: 9669 +[13:56:17.244542] Epoch: [8] [ 40/156] eta: 0:00:19 lr: 0.000516 loss: 0.4808 (0.4790) time: 0.1422 data: 0.0001 max mem: 9669 +[13:56:20.089234] Epoch: [8] [ 60/156] eta: 0:00:15 lr: 0.000524 loss: 0.4981 (0.4807) time: 0.1422 data: 0.0001 max mem: 9669 +[13:56:22.955967] Epoch: [8] [ 80/156] eta: 0:00:11 lr: 0.000532 loss: 0.5331 (0.4941) time: 0.1433 data: 0.0002 max mem: 9669 +[13:56:25.809101] Epoch: [8] [100/156] eta: 0:00:08 lr: 0.000540 loss: 0.4574 (0.4899) time: 0.1426 data: 0.0002 max mem: 9669 +[13:56:28.660286] Epoch: [8] [120/156] eta: 0:00:05 lr: 0.000548 loss: 0.4651 (0.4857) time: 0.1425 data: 0.0002 max mem: 9669 +[13:56:31.515312] Epoch: [8] [140/156] eta: 0:00:02 lr: 0.000556 loss: 0.4696 (0.4866) time: 0.1427 data: 0.0002 max mem: 9669 +[13:56:33.646683] Epoch: [8] [155/156] eta: 0:00:00 lr: 0.000562 loss: 0.4585 (0.4825) time: 0.1421 data: 0.0001 max mem: 9669 +[13:56:33.775728] Epoch: [8] Total time: 0:00:23 (0.1499 s / it) +[13:56:33.784229] Averaged stats: lr: 0.000562 loss: 0.4585 (0.4825) +[13:56:34.757027] val: [ 0/17] eta: 0:00:16 loss: 0.2019 (0.2019) time: 0.9608 data: 0.9268 max mem: 9669 +[13:56:35.098065] val: [10/17] eta: 0:00:00 loss: 0.2579 (0.3103) time: 0.1183 data: 0.0844 max mem: 9669 +[13:56:35.303058] val: [16/17] eta: 0:00:00 loss: 0.2253 (0.2528) time: 0.0886 data: 0.0546 max mem: 9669 +[13:56:35.435163] val: Total time: 0:00:01 (0.0965 s / it) +[13:56:35.449333] val loss: 0.25277270420509224 +[13:56:35.449567] Accuracy: 0.9019, F1 Score: 0.9017, ROC AUC: 0.9657, Hamming Loss: 0.0981, + Jaccard Score: 0.8210, Precision: 0.9043, Recall: 0.9019, + Average Precision: 0.9664, Kappa: 0.8037, Score: 0.8904 +[13:56:37.274087] Best epoch = 8, Best score = 0.8904 +[13:56:37.338324] log_dir: ./output_logs/retfound +[13:56:38.450234] Epoch: [9] [ 0/156] eta: 0:02:53 lr: 0.000562 loss: 0.4432 (0.4432) time: 1.1108 data: 0.9656 max mem: 9669 +[13:56:41.295205] Epoch: [9] [ 20/156] eta: 0:00:25 lr: 0.000571 loss: 0.4576 (0.4668) time: 0.1422 data: 0.0002 max mem: 9669 +[13:56:44.148284] Epoch: [9] [ 40/156] eta: 0:00:19 lr: 0.000579 loss: 0.4368 (0.4555) time: 0.1426 data: 0.0002 max mem: 9669 +[13:56:46.992058] Epoch: [9] [ 60/156] eta: 0:00:15 lr: 0.000587 loss: 0.4287 (0.4578) time: 0.1421 data: 0.0001 max mem: 9669 +[13:56:49.826542] Epoch: [9] [ 80/156] eta: 0:00:11 lr: 0.000595 loss: 0.4820 (0.4673) time: 0.1417 data: 0.0002 max mem: 9669 +[13:56:52.668918] Epoch: [9] [100/156] eta: 0:00:08 lr: 0.000603 loss: 0.4720 (0.4682) time: 0.1421 data: 0.0001 max mem: 9669 +[13:56:55.517904] Epoch: [9] [120/156] eta: 0:00:05 lr: 0.000611 loss: 0.5268 (0.4759) time: 0.1424 data: 0.0001 max mem: 9669 +[13:56:58.362021] Epoch: [9] [140/156] eta: 0:00:02 lr: 0.000619 loss: 0.4779 (0.4769) time: 0.1422 data: 0.0002 max mem: 9669 +[13:57:00.495031] Epoch: [9] [155/156] eta: 0:00:00 lr: 0.000625 loss: 0.4439 (0.4740) time: 0.1422 data: 0.0002 max mem: 9669 +[13:57:00.618510] Epoch: [9] Total time: 0:00:23 (0.1492 s / it) +[13:57:00.628174] Averaged stats: lr: 0.000625 loss: 0.4439 (0.4740) +[13:57:01.591549] val: [ 0/17] eta: 0:00:16 loss: 0.3370 (0.3370) time: 0.9456 data: 0.9096 max mem: 9669 +[13:57:01.939226] val: [10/17] eta: 0:00:00 loss: 0.2942 (0.3432) time: 0.1175 data: 0.0828 max mem: 9669 +[13:57:02.144639] val: [16/17] eta: 0:00:00 loss: 0.2737 (0.2629) time: 0.0881 data: 0.0537 max mem: 9669 +[13:57:02.270073] val: Total time: 0:00:01 (0.0956 s / it) +[13:57:02.284546] val loss: 0.26290999637807116 +[13:57:02.284829] Accuracy: 0.8907, F1 Score: 0.8905, ROC AUC: 0.9656, Hamming Loss: 0.1093, + Jaccard Score: 0.8027, Precision: 0.8941, Recall: 0.8907, + Average Precision: 0.9670, Kappa: 0.7815, Score: 0.8792 +[13:57:02.326756] Best epoch = 8, Best score = 0.8904 +[13:57:02.615162] log_dir: ./output_logs/retfound +[13:57:03.661265] Epoch: [10] [ 0/156] eta: 0:02:43 lr: 0.000625 loss: 0.4573 (0.4573) time: 1.0451 data: 0.8999 max mem: 9669 +[13:57:06.508360] Epoch: [10] [ 20/156] eta: 0:00:25 lr: 0.000625 loss: 0.4759 (0.5153) time: 0.1423 data: 0.0001 max mem: 9669 +[13:57:09.364293] Epoch: [10] [ 40/156] eta: 0:00:19 lr: 0.000625 loss: 0.4536 (0.4997) time: 0.1428 data: 0.0002 max mem: 9669 +[13:57:12.212589] Epoch: [10] [ 60/156] eta: 0:00:15 lr: 0.000624 loss: 0.4371 (0.4820) time: 0.1424 data: 0.0002 max mem: 9669 +[13:57:15.060206] Epoch: [10] [ 80/156] eta: 0:00:11 lr: 0.000624 loss: 0.4576 (0.4800) time: 0.1423 data: 0.0001 max mem: 9669 +[13:57:17.603254] Epoch: [10] [100/156] eta: 0:00:08 lr: 0.000623 loss: 0.4665 (0.4773) time: 0.1271 data: 0.0002 max mem: 9669 +[13:57:20.428375] Epoch: [10] [120/156] eta: 0:00:05 lr: 0.000623 loss: 0.4646 (0.4758) time: 0.1412 data: 0.0002 max mem: 9669 +[13:57:23.265496] Epoch: [10] [140/156] eta: 0:00:02 lr: 0.000622 loss: 0.4452 (0.4737) time: 0.1418 data: 0.0002 max mem: 9669 +[13:57:25.401664] Epoch: [10] [155/156] eta: 0:00:00 lr: 0.000621 loss: 0.4679 (0.4773) time: 0.1424 data: 0.0001 max mem: 9669 +[13:57:25.527935] Epoch: [10] Total time: 0:00:22 (0.1469 s / it) +[13:57:25.536317] Averaged stats: lr: 0.000621 loss: 0.4679 (0.4773) +[13:57:26.531190] val: [ 0/17] eta: 0:00:16 loss: 0.1854 (0.1854) time: 0.9782 data: 0.9439 max mem: 9669 +[13:57:26.871659] val: [10/17] eta: 0:00:00 loss: 0.2088 (0.2325) time: 0.1198 data: 0.0859 max mem: 9669 +[13:57:27.076608] val: [16/17] eta: 0:00:00 loss: 0.2786 (0.2516) time: 0.0895 data: 0.0557 max mem: 9669 +[13:57:27.203034] val: Total time: 0:00:01 (0.0971 s / it) +[13:57:27.217059] val loss: 0.25161123801680174 +[13:57:27.217317] Accuracy: 0.9019, F1 Score: 0.9018, ROC AUC: 0.9649, Hamming Loss: 0.0981, + Jaccard Score: 0.8212, Precision: 0.9021, Recall: 0.9019, + Average Precision: 0.9658, Kappa: 0.8037, Score: 0.8902 +[13:57:27.256969] Best epoch = 8, Best score = 0.8904 +[13:57:27.596756] log_dir: ./output_logs/retfound +[13:57:28.743864] Epoch: [11] [ 0/156] eta: 0:02:58 lr: 0.000621 loss: 0.4665 (0.4665) time: 1.1458 data: 0.9998 max mem: 9669 +[13:57:31.595558] Epoch: [11] [ 20/156] eta: 0:00:25 lr: 0.000620 loss: 0.4409 (0.4736) time: 0.1425 data: 0.0002 max mem: 9669 +[13:57:34.437466] Epoch: [11] [ 40/156] eta: 0:00:19 lr: 0.000619 loss: 0.4341 (0.4579) time: 0.1421 data: 0.0002 max mem: 9669 +[13:57:37.289257] Epoch: [11] [ 60/156] eta: 0:00:15 lr: 0.000618 loss: 0.4596 (0.4576) time: 0.1425 data: 0.0002 max mem: 9669 +[13:57:40.129571] Epoch: [11] [ 80/156] eta: 0:00:11 lr: 0.000616 loss: 0.4355 (0.4541) time: 0.1420 data: 0.0002 max mem: 9669 +[13:57:42.975441] Epoch: [11] [100/156] eta: 0:00:08 lr: 0.000615 loss: 0.4711 (0.4532) time: 0.1422 data: 0.0002 max mem: 9669 +[13:57:45.824675] Epoch: [11] [120/156] eta: 0:00:05 lr: 0.000613 loss: 0.4498 (0.4599) time: 0.1424 data: 0.0002 max mem: 9669 +[13:57:48.671048] Epoch: [11] [140/156] eta: 0:00:02 lr: 0.000611 loss: 0.3991 (0.4530) time: 0.1423 data: 0.0001 max mem: 9669 +[13:57:50.806850] Epoch: [11] [155/156] eta: 0:00:00 lr: 0.000610 loss: 0.3987 (0.4518) time: 0.1423 data: 0.0001 max mem: 9669 +[13:57:50.931587] Epoch: [11] Total time: 0:00:23 (0.1496 s / it) +[13:57:50.939546] Averaged stats: lr: 0.000610 loss: 0.3987 (0.4518) +[13:57:51.845585] val: [ 0/17] eta: 0:00:15 loss: 0.2339 (0.2339) time: 0.8892 data: 0.8541 max mem: 9669 +[13:57:52.320911] val: [10/17] eta: 0:00:00 loss: 0.2979 (0.2607) time: 0.1240 data: 0.0894 max mem: 9669 +[13:57:52.520232] val: [16/17] eta: 0:00:00 loss: 0.2054 (0.2327) time: 0.0919 data: 0.0579 max mem: 9669 +[13:57:52.640729] val: Total time: 0:00:01 (0.0991 s / it) +[13:57:52.655238] val loss: 0.23271823586786494 +[13:57:52.655435] Accuracy: 0.9130, F1 Score: 0.9129, ROC AUC: 0.9688, Hamming Loss: 0.0870, + Jaccard Score: 0.8398, Precision: 0.9132, Recall: 0.9130, + Average Precision: 0.9693, Kappa: 0.8259, Score: 0.9025 +[13:57:54.498736] Best epoch = 11, Best score = 0.9025 +[13:57:54.572853] log_dir: ./output_logs/retfound +[13:57:55.711471] Epoch: [12] [ 0/156] eta: 0:02:57 lr: 0.000610 loss: 0.4175 (0.4175) time: 1.1374 data: 0.9923 max mem: 9669 +[13:57:58.558032] Epoch: [12] [ 20/156] eta: 0:00:25 lr: 0.000608 loss: 0.4299 (0.4135) time: 0.1423 data: 0.0001 max mem: 9669 +[13:58:01.407707] Epoch: [12] [ 40/156] eta: 0:00:19 lr: 0.000606 loss: 0.4877 (0.4603) time: 0.1424 data: 0.0002 max mem: 9669 +[13:58:04.254639] Epoch: [12] [ 60/156] eta: 0:00:15 lr: 0.000603 loss: 0.4499 (0.4613) time: 0.1423 data: 0.0002 max mem: 9669 +[13:58:07.102371] Epoch: [12] [ 80/156] eta: 0:00:11 lr: 0.000601 loss: 0.4070 (0.4535) time: 0.1423 data: 0.0002 max mem: 9669 +[13:58:09.955830] Epoch: [12] [100/156] eta: 0:00:08 lr: 0.000599 loss: 0.4636 (0.4610) time: 0.1426 data: 0.0002 max mem: 9669 +[13:58:12.803260] Epoch: [12] [120/156] eta: 0:00:05 lr: 0.000596 loss: 0.4434 (0.4627) time: 0.1423 data: 0.0002 max mem: 9669 +[13:58:15.647844] Epoch: [12] [140/156] eta: 0:00:02 lr: 0.000593 loss: 0.4253 (0.4616) time: 0.1422 data: 0.0001 max mem: 9669 +[13:58:17.785967] Epoch: [12] [155/156] eta: 0:00:00 lr: 0.000591 loss: 0.4490 (0.4638) time: 0.1423 data: 0.0001 max mem: 9669 +[13:58:17.924603] Epoch: [12] Total time: 0:00:23 (0.1497 s / it) +[13:58:17.933376] Averaged stats: lr: 0.000591 loss: 0.4490 (0.4638) +[13:58:18.953285] val: [ 0/17] eta: 0:00:17 loss: 0.2895 (0.2895) time: 1.0063 data: 0.9696 max mem: 9669 +[13:58:19.297278] val: [10/17] eta: 0:00:00 loss: 0.2840 (0.2778) time: 0.1227 data: 0.0883 max mem: 9669 +[13:58:19.501653] val: [16/17] eta: 0:00:00 loss: 0.1948 (0.2283) time: 0.0914 data: 0.0572 max mem: 9669 +[13:58:19.629774] val: Total time: 0:00:01 (0.0991 s / it) +[13:58:19.644571] val loss: 0.22827284958432703 +[13:58:19.644827] Accuracy: 0.9037, F1 Score: 0.9036, ROC AUC: 0.9681, Hamming Loss: 0.0963, + Jaccard Score: 0.8242, Precision: 0.9048, Recall: 0.9037, + Average Precision: 0.9683, Kappa: 0.8074, Score: 0.8930 +[13:58:19.681919] Best epoch = 11, Best score = 0.9025 +[13:58:19.965673] log_dir: ./output_logs/retfound +[13:58:21.094196] Epoch: [13] [ 0/156] eta: 0:02:55 lr: 0.000591 loss: 0.3888 (0.3888) time: 1.1273 data: 0.9827 max mem: 9669 +[13:58:23.942487] Epoch: [13] [ 20/156] eta: 0:00:25 lr: 0.000588 loss: 0.4828 (0.4831) time: 0.1424 data: 0.0001 max mem: 9669 +[13:58:26.787334] Epoch: [13] [ 40/156] eta: 0:00:19 lr: 0.000585 loss: 0.4745 (0.4803) time: 0.1422 data: 0.0001 max mem: 9669 +[13:58:29.636195] Epoch: [13] [ 60/156] eta: 0:00:15 lr: 0.000582 loss: 0.4478 (0.4713) time: 0.1424 data: 0.0002 max mem: 9669 +[13:58:32.486502] Epoch: [13] [ 80/156] eta: 0:00:11 lr: 0.000579 loss: 0.4615 (0.4669) time: 0.1425 data: 0.0002 max mem: 9669 +[13:58:35.337158] Epoch: [13] [100/156] eta: 0:00:08 lr: 0.000575 loss: 0.4323 (0.4634) time: 0.1425 data: 0.0002 max mem: 9669 +[13:58:38.183163] Epoch: [13] [120/156] eta: 0:00:05 lr: 0.000572 loss: 0.3835 (0.4549) time: 0.1423 data: 0.0002 max mem: 9669 +[13:58:41.035947] Epoch: [13] [140/156] eta: 0:00:02 lr: 0.000568 loss: 0.4419 (0.4550) time: 0.1426 data: 0.0002 max mem: 9669 +[13:58:43.170754] Epoch: [13] [155/156] eta: 0:00:00 lr: 0.000566 loss: 0.4427 (0.4572) time: 0.1426 data: 0.0001 max mem: 9669 +[13:58:43.300912] Epoch: [13] Total time: 0:00:23 (0.1496 s / it) +[13:58:43.309737] Averaged stats: lr: 0.000566 loss: 0.4427 (0.4572) +[13:58:44.302035] val: [ 0/17] eta: 0:00:16 loss: 0.2312 (0.2312) time: 0.9759 data: 0.9421 max mem: 9669 +[13:58:44.649432] val: [10/17] eta: 0:00:00 loss: 0.2646 (0.2644) time: 0.1202 data: 0.0858 max mem: 9669 +[13:58:44.854202] val: [16/17] eta: 0:00:00 loss: 0.2106 (0.2341) time: 0.0898 data: 0.0556 max mem: 9669 +[13:58:44.983550] val: Total time: 0:00:01 (0.0975 s / it) +[13:58:44.997890] val loss: 0.23409848265788136 +[13:58:44.998218] Accuracy: 0.9130, F1 Score: 0.9129, ROC AUC: 0.9672, Hamming Loss: 0.0870, + Jaccard Score: 0.8398, Precision: 0.9132, Recall: 0.9130, + Average Precision: 0.9681, Kappa: 0.8259, Score: 0.9020 +[13:58:45.038269] Best epoch = 11, Best score = 0.9025 +[13:58:45.342678] log_dir: ./output_logs/retfound +[13:58:46.534879] Epoch: [14] [ 0/156] eta: 0:03:05 lr: 0.000565 loss: 0.4599 (0.4599) time: 1.1909 data: 1.0462 max mem: 9669 +[13:58:49.373496] Epoch: [14] [ 20/156] eta: 0:00:26 lr: 0.000562 loss: 0.4364 (0.4398) time: 0.1419 data: 0.0001 max mem: 9669 +[13:58:52.219422] Epoch: [14] [ 40/156] eta: 0:00:19 lr: 0.000558 loss: 0.4232 (0.4337) time: 0.1423 data: 0.0002 max mem: 9669 +[13:58:55.071447] Epoch: [14] [ 60/156] eta: 0:00:15 lr: 0.000554 loss: 0.4154 (0.4344) time: 0.1426 data: 0.0001 max mem: 9669 +[13:58:57.916497] Epoch: [14] [ 80/156] eta: 0:00:11 lr: 0.000550 loss: 0.4435 (0.4407) time: 0.1422 data: 0.0002 max mem: 9669 +[13:59:00.766092] Epoch: [14] [100/156] eta: 0:00:08 lr: 0.000546 loss: 0.4026 (0.4389) time: 0.1424 data: 0.0003 max mem: 9669 +[13:59:03.610302] Epoch: [14] [120/156] eta: 0:00:05 lr: 0.000541 loss: 0.4247 (0.4368) time: 0.1422 data: 0.0002 max mem: 9669 +[13:59:06.463103] Epoch: [14] [140/156] eta: 0:00:02 lr: 0.000537 loss: 0.3900 (0.4370) time: 0.1426 data: 0.0001 max mem: 9669 +[13:59:08.602426] Epoch: [14] [155/156] eta: 0:00:00 lr: 0.000534 loss: 0.4216 (0.4376) time: 0.1427 data: 0.0001 max mem: 9669 +[13:59:08.742351] Epoch: [14] Total time: 0:00:23 (0.1500 s / it) +[13:59:08.750941] Averaged stats: lr: 0.000534 loss: 0.4216 (0.4376) +[13:59:09.733719] val: [ 0/17] eta: 0:00:16 loss: 0.2281 (0.2281) time: 0.9659 data: 0.9308 max mem: 9669 +[13:59:10.080707] val: [10/17] eta: 0:00:00 loss: 0.2639 (0.2634) time: 0.1193 data: 0.0847 max mem: 9669 +[13:59:10.279523] val: [16/17] eta: 0:00:00 loss: 0.2171 (0.2362) time: 0.0888 data: 0.0549 max mem: 9669 +[13:59:10.401738] val: Total time: 0:00:01 (0.0962 s / it) +[13:59:10.415748] val loss: 0.23617559057824752 +[13:59:10.415960] Accuracy: 0.9000, F1 Score: 0.9000, ROC AUC: 0.9679, Hamming Loss: 0.1000, + Jaccard Score: 0.8182, Precision: 0.9002, Recall: 0.9000, + Average Precision: 0.9691, Kappa: 0.8000, Score: 0.8893 +[13:59:10.452709] Best epoch = 11, Best score = 0.9025 +[13:59:10.752824] log_dir: ./output_logs/retfound +[13:59:11.895015] Epoch: [15] [ 0/156] eta: 0:02:58 lr: 0.000534 loss: 0.5035 (0.5035) time: 1.1411 data: 0.9952 max mem: 9669 +[13:59:14.743822] Epoch: [15] [ 20/156] eta: 0:00:25 lr: 0.000529 loss: 0.4108 (0.4235) time: 0.1424 data: 0.0002 max mem: 9669 +[13:59:17.595554] Epoch: [15] [ 40/156] eta: 0:00:19 lr: 0.000525 loss: 0.4453 (0.4318) time: 0.1425 data: 0.0002 max mem: 9669 +[13:59:20.450100] Epoch: [15] [ 60/156] eta: 0:00:15 lr: 0.000520 loss: 0.4378 (0.4427) time: 0.1427 data: 0.0002 max mem: 9669 +[13:59:22.867269] Epoch: [15] [ 80/156] eta: 0:00:11 lr: 0.000515 loss: 0.3557 (0.4318) time: 0.1208 data: 0.0001 max mem: 9669 +[13:59:25.712012] Epoch: [15] [100/156] eta: 0:00:08 lr: 0.000510 loss: 0.3944 (0.4249) time: 0.1422 data: 0.0002 max mem: 9669 +[13:59:28.556268] Epoch: [15] [120/156] eta: 0:00:05 lr: 0.000505 loss: 0.4038 (0.4225) time: 0.1422 data: 0.0002 max mem: 9669 +[13:59:31.403837] Epoch: [15] [140/156] eta: 0:00:02 lr: 0.000500 loss: 0.4266 (0.4221) time: 0.1423 data: 0.0002 max mem: 9669 +[13:59:33.539935] Epoch: [15] [155/156] eta: 0:00:00 lr: 0.000497 loss: 0.4199 (0.4237) time: 0.1425 data: 0.0001 max mem: 9669 +[13:59:33.679693] Epoch: [15] Total time: 0:00:22 (0.1470 s / it) +[13:59:33.689206] Averaged stats: lr: 0.000497 loss: 0.4199 (0.4237) +[13:59:34.703761] val: [ 0/17] eta: 0:00:17 loss: 0.2239 (0.2239) time: 1.0018 data: 0.9658 max mem: 9669 +[13:59:35.084875] val: [10/17] eta: 0:00:00 loss: 0.2239 (0.2526) time: 0.1257 data: 0.0912 max mem: 9669 +[13:59:35.290368] val: [16/17] eta: 0:00:00 loss: 0.2040 (0.2294) time: 0.0934 data: 0.0590 max mem: 9669 +[13:59:35.405531] val: Total time: 0:00:01 (0.1003 s / it) +[13:59:35.419759] val loss: 0.22935323767802296 +[13:59:35.420030] Accuracy: 0.9185, F1 Score: 0.9185, ROC AUC: 0.9698, Hamming Loss: 0.0815, + Jaccard Score: 0.8493, Precision: 0.9187, Recall: 0.9185, + Average Precision: 0.9710, Kappa: 0.8370, Score: 0.9085 +[13:59:37.239327] Best epoch = 15, Best score = 0.9085 +[13:59:37.312410] log_dir: ./output_logs/retfound +[13:59:38.538692] Epoch: [16] [ 0/156] eta: 0:03:11 lr: 0.000496 loss: 0.3148 (0.3148) time: 1.2253 data: 1.0807 max mem: 9669 +[13:59:41.387779] Epoch: [16] [ 20/156] eta: 0:00:26 lr: 0.000491 loss: 0.3946 (0.4188) time: 0.1424 data: 0.0002 max mem: 9669 +[13:59:44.235358] Epoch: [16] [ 40/156] eta: 0:00:19 lr: 0.000486 loss: 0.4368 (0.4331) time: 0.1423 data: 0.0002 max mem: 9669 +[13:59:47.077935] Epoch: [16] [ 60/156] eta: 0:00:15 lr: 0.000481 loss: 0.4131 (0.4302) time: 0.1421 data: 0.0002 max mem: 9669 +[13:59:49.925043] Epoch: [16] [ 80/156] eta: 0:00:11 lr: 0.000475 loss: 0.4221 (0.4310) time: 0.1423 data: 0.0002 max mem: 9669 +[13:59:52.788471] Epoch: [16] [100/156] eta: 0:00:08 lr: 0.000470 loss: 0.3896 (0.4337) time: 0.1431 data: 0.0002 max mem: 9669 +[13:59:55.636437] Epoch: [16] [120/156] eta: 0:00:05 lr: 0.000465 loss: 0.4177 (0.4348) time: 0.1423 data: 0.0002 max mem: 9669 +[13:59:58.488572] Epoch: [16] [140/156] eta: 0:00:02 lr: 0.000459 loss: 0.3996 (0.4352) time: 0.1426 data: 0.0001 max mem: 9669 +[14:00:00.625654] Epoch: [16] [155/156] eta: 0:00:00 lr: 0.000455 loss: 0.3996 (0.4304) time: 0.1424 data: 0.0001 max mem: 9669 +[14:00:00.753079] Epoch: [16] Total time: 0:00:23 (0.1503 s / it) +[14:00:00.758859] Averaged stats: lr: 0.000455 loss: 0.3996 (0.4304) +[14:00:01.787699] val: [ 0/17] eta: 0:00:17 loss: 0.3264 (0.3264) time: 1.0129 data: 0.9787 max mem: 9669 +[14:00:02.129569] val: [10/17] eta: 0:00:00 loss: 0.2887 (0.2838) time: 0.1231 data: 0.0891 max mem: 9669 +[14:00:02.334474] val: [16/17] eta: 0:00:00 loss: 0.1784 (0.2348) time: 0.0917 data: 0.0577 max mem: 9669 +[14:00:02.450112] val: Total time: 0:00:01 (0.0986 s / it) +[14:00:02.464116] val loss: 0.2348311854635968 +[14:00:02.464374] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9689, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9213, Recall: 0.9204, + Average Precision: 0.9700, Kappa: 0.8407, Score: 0.9100 +[14:00:04.374135] Best epoch = 16, Best score = 0.9100 +[14:00:04.447274] log_dir: ./output_logs/retfound +[14:00:05.651532] Epoch: [17] [ 0/156] eta: 0:03:07 lr: 0.000455 loss: 0.3511 (0.3511) time: 1.2031 data: 1.0565 max mem: 9669 +[14:00:08.496356] Epoch: [17] [ 20/156] eta: 0:00:26 lr: 0.000449 loss: 0.3710 (0.3823) time: 0.1422 data: 0.0002 max mem: 9669 +[14:00:11.341458] Epoch: [17] [ 40/156] eta: 0:00:19 lr: 0.000443 loss: 0.3868 (0.3839) time: 0.1422 data: 0.0002 max mem: 9669 +[14:00:14.187970] Epoch: [17] [ 60/156] eta: 0:00:15 lr: 0.000438 loss: 0.4111 (0.4096) time: 0.1423 data: 0.0001 max mem: 9669 +[14:00:17.028949] Epoch: [17] [ 80/156] eta: 0:00:11 lr: 0.000432 loss: 0.4047 (0.4131) time: 0.1420 data: 0.0002 max mem: 9669 +[14:00:19.481576] Epoch: [17] [100/156] eta: 0:00:08 lr: 0.000426 loss: 0.4097 (0.4126) time: 0.1226 data: 0.0002 max mem: 9669 +[14:00:22.328286] Epoch: [17] [120/156] eta: 0:00:05 lr: 0.000420 loss: 0.3984 (0.4150) time: 0.1423 data: 0.0002 max mem: 9669 +[14:00:25.171976] Epoch: [17] [140/156] eta: 0:00:02 lr: 0.000414 loss: 0.4486 (0.4206) time: 0.1421 data: 0.0002 max mem: 9669 +[14:00:27.304013] Epoch: [17] [155/156] eta: 0:00:00 lr: 0.000410 loss: 0.4067 (0.4216) time: 0.1423 data: 0.0001 max mem: 9669 +[14:00:27.433555] Epoch: [17] Total time: 0:00:22 (0.1473 s / it) +[14:00:27.442395] Averaged stats: lr: 0.000410 loss: 0.4067 (0.4216) +[14:00:28.459035] val: [ 0/17] eta: 0:00:16 loss: 0.1918 (0.1918) time: 0.9995 data: 0.9645 max mem: 9669 +[14:00:28.804998] val: [10/17] eta: 0:00:00 loss: 0.2945 (0.2638) time: 0.1222 data: 0.0878 max mem: 9669 +[14:00:29.009728] val: [16/17] eta: 0:00:00 loss: 0.2452 (0.2511) time: 0.0911 data: 0.0569 max mem: 9669 +[14:00:29.128985] val: Total time: 0:00:01 (0.0982 s / it) +[14:00:29.143987] val loss: 0.25106555968523026 +[14:00:29.144204] Accuracy: 0.9074, F1 Score: 0.9074, ROC AUC: 0.9677, Hamming Loss: 0.0926, + Jaccard Score: 0.8305, Precision: 0.9074, Recall: 0.9074, + Average Precision: 0.9684, Kappa: 0.8148, Score: 0.8966 +[14:00:29.193039] Best epoch = 16, Best score = 0.9100 +[14:00:29.468915] log_dir: ./output_logs/retfound +[14:00:30.640126] Epoch: [18] [ 0/156] eta: 0:03:02 lr: 0.000409 loss: 0.6675 (0.6675) time: 1.1702 data: 1.0284 max mem: 9669 +[14:00:33.486891] Epoch: [18] [ 20/156] eta: 0:00:26 lr: 0.000403 loss: 0.4165 (0.4052) time: 0.1423 data: 0.0002 max mem: 9669 +[14:00:36.333861] Epoch: [18] [ 40/156] eta: 0:00:19 lr: 0.000397 loss: 0.4269 (0.4197) time: 0.1423 data: 0.0002 max mem: 9669 +[14:00:39.186443] Epoch: [18] [ 60/156] eta: 0:00:15 lr: 0.000391 loss: 0.4218 (0.4280) time: 0.1426 data: 0.0002 max mem: 9669 +[14:00:42.035752] Epoch: [18] [ 80/156] eta: 0:00:11 lr: 0.000385 loss: 0.3917 (0.4246) time: 0.1424 data: 0.0002 max mem: 9669 +[14:00:44.878919] Epoch: [18] [100/156] eta: 0:00:08 lr: 0.000379 loss: 0.4344 (0.4275) time: 0.1421 data: 0.0002 max mem: 9669 +[14:00:47.733584] Epoch: [18] [120/156] eta: 0:00:05 lr: 0.000373 loss: 0.4358 (0.4303) time: 0.1427 data: 0.0002 max mem: 9669 +[14:00:50.581069] Epoch: [18] [140/156] eta: 0:00:02 lr: 0.000367 loss: 0.4121 (0.4294) time: 0.1423 data: 0.0001 max mem: 9669 +[14:00:52.714428] Epoch: [18] [155/156] eta: 0:00:00 lr: 0.000362 loss: 0.3924 (0.4266) time: 0.1421 data: 0.0001 max mem: 9669 +[14:00:52.853426] Epoch: [18] Total time: 0:00:23 (0.1499 s / it) +[14:00:52.861954] Averaged stats: lr: 0.000362 loss: 0.3924 (0.4266) +[14:00:53.826973] val: [ 0/17] eta: 0:00:16 loss: 0.2619 (0.2619) time: 0.9504 data: 0.9148 max mem: 9669 +[14:00:54.180927] val: [10/17] eta: 0:00:00 loss: 0.2619 (0.2848) time: 0.1185 data: 0.0839 max mem: 9669 +[14:00:54.384626] val: [16/17] eta: 0:00:00 loss: 0.1932 (0.2353) time: 0.0886 data: 0.0543 max mem: 9669 +[14:00:54.501586] val: Total time: 0:00:01 (0.0957 s / it) +[14:00:54.515711] val loss: 0.23531180181924036 +[14:00:54.516049] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9708, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9220, Recall: 0.9204, + Average Precision: 0.9716, Kappa: 0.8407, Score: 0.9106 +[14:00:56.307530] Best epoch = 18, Best score = 0.9106 +[14:00:56.388744] log_dir: ./output_logs/retfound +[14:00:57.542420] Epoch: [19] [ 0/156] eta: 0:02:59 lr: 0.000362 loss: 0.4672 (0.4672) time: 1.1526 data: 1.0071 max mem: 9669 +[14:01:00.384638] Epoch: [19] [ 20/156] eta: 0:00:25 lr: 0.000356 loss: 0.4151 (0.4028) time: 0.1421 data: 0.0002 max mem: 9669 +[14:01:03.227846] Epoch: [19] [ 40/156] eta: 0:00:19 lr: 0.000349 loss: 0.3841 (0.3963) time: 0.1421 data: 0.0001 max mem: 9669 +[14:01:06.067925] Epoch: [19] [ 60/156] eta: 0:00:15 lr: 0.000343 loss: 0.4096 (0.4003) time: 0.1420 data: 0.0004 max mem: 9669 +[14:01:08.914592] Epoch: [19] [ 80/156] eta: 0:00:11 lr: 0.000337 loss: 0.3970 (0.3987) time: 0.1423 data: 0.0001 max mem: 9669 +[14:01:11.742160] Epoch: [19] [100/156] eta: 0:00:08 lr: 0.000331 loss: 0.4069 (0.3998) time: 0.1413 data: 0.0002 max mem: 9669 +[14:01:14.586360] Epoch: [19] [120/156] eta: 0:00:05 lr: 0.000324 loss: 0.3498 (0.3965) time: 0.1422 data: 0.0001 max mem: 9669 +[14:01:17.430512] Epoch: [19] [140/156] eta: 0:00:02 lr: 0.000318 loss: 0.4158 (0.3991) time: 0.1422 data: 0.0002 max mem: 9669 +[14:01:19.564756] Epoch: [19] [155/156] eta: 0:00:00 lr: 0.000313 loss: 0.4382 (0.4027) time: 0.1422 data: 0.0001 max mem: 9669 +[14:01:19.693524] Epoch: [19] Total time: 0:00:23 (0.1494 s / it) +[14:01:19.700988] Averaged stats: lr: 0.000313 loss: 0.4382 (0.4027) +[14:01:20.725514] val: [ 0/17] eta: 0:00:17 loss: 0.1684 (0.1684) time: 1.0067 data: 0.9705 max mem: 9669 +[14:01:21.066854] val: [10/17] eta: 0:00:00 loss: 0.2131 (0.2261) time: 0.1225 data: 0.0884 max mem: 9669 +[14:01:21.273605] val: [16/17] eta: 0:00:00 loss: 0.2131 (0.2168) time: 0.0914 data: 0.0573 max mem: 9669 +[14:01:21.387809] val: Total time: 0:00:01 (0.0982 s / it) +[14:01:21.402148] val loss: 0.21684428611222437 +[14:01:21.402412] Accuracy: 0.9148, F1 Score: 0.9148, ROC AUC: 0.9713, Hamming Loss: 0.0852, + Jaccard Score: 0.8430, Precision: 0.9148, Recall: 0.9148, + Average Precision: 0.9723, Kappa: 0.8296, Score: 0.9052 +[14:01:21.449695] Best epoch = 18, Best score = 0.9106 +[14:01:21.719157] log_dir: ./output_logs/retfound +[14:01:22.953279] Epoch: [20] [ 0/156] eta: 0:03:12 lr: 0.000313 loss: 0.4803 (0.4803) time: 1.2331 data: 1.0823 max mem: 9669 +[14:01:25.798169] Epoch: [20] [ 20/156] eta: 0:00:26 lr: 0.000307 loss: 0.3682 (0.3869) time: 0.1422 data: 0.0001 max mem: 9669 +[14:01:28.626670] Epoch: [20] [ 40/156] eta: 0:00:19 lr: 0.000300 loss: 0.3869 (0.3848) time: 0.1414 data: 0.0002 max mem: 9669 +[14:01:31.466087] Epoch: [20] [ 60/156] eta: 0:00:15 lr: 0.000294 loss: 0.3871 (0.3932) time: 0.1419 data: 0.0002 max mem: 9669 +[14:01:34.297973] Epoch: [20] [ 80/156] eta: 0:00:11 lr: 0.000288 loss: 0.4182 (0.3983) time: 0.1415 data: 0.0002 max mem: 9669 +[14:01:37.135112] Epoch: [20] [100/156] eta: 0:00:08 lr: 0.000282 loss: 0.3950 (0.4008) time: 0.1418 data: 0.0001 max mem: 9669 +[14:01:39.973376] Epoch: [20] [120/156] eta: 0:00:05 lr: 0.000275 loss: 0.4196 (0.4051) time: 0.1419 data: 0.0002 max mem: 9669 +[14:01:42.822632] Epoch: [20] [140/156] eta: 0:00:02 lr: 0.000269 loss: 0.3612 (0.4010) time: 0.1424 data: 0.0002 max mem: 9669 +[14:01:44.955935] Epoch: [20] [155/156] eta: 0:00:00 lr: 0.000265 loss: 0.4231 (0.4008) time: 0.1421 data: 0.0001 max mem: 9669 +[14:01:45.080502] Epoch: [20] Total time: 0:00:23 (0.1498 s / it) +[14:01:45.089459] Averaged stats: lr: 0.000265 loss: 0.4231 (0.4008) +[14:01:46.151168] val: [ 0/17] eta: 0:00:17 loss: 0.1891 (0.1891) time: 1.0490 data: 1.0150 max mem: 9669 +[14:01:46.539146] val: [10/17] eta: 0:00:00 loss: 0.2415 (0.2510) time: 0.1306 data: 0.0964 max mem: 9669 +[14:01:46.745093] val: [16/17] eta: 0:00:00 loss: 0.2240 (0.2322) time: 0.0966 data: 0.0625 max mem: 9669 +[14:01:46.876490] val: Total time: 0:00:01 (0.1044 s / it) +[14:01:46.900138] val loss: 0.2322498717728783 +[14:01:46.900377] Accuracy: 0.9111, F1 Score: 0.9111, ROC AUC: 0.9698, Hamming Loss: 0.0889, + Jaccard Score: 0.8367, Precision: 0.9111, Recall: 0.9111, + Average Precision: 0.9704, Kappa: 0.8222, Score: 0.9011 +[14:01:46.944824] Best epoch = 18, Best score = 0.9106 +[14:01:47.219331] log_dir: ./output_logs/retfound +[14:01:48.487106] Epoch: [21] [ 0/156] eta: 0:03:17 lr: 0.000264 loss: 0.3076 (0.3076) time: 1.2667 data: 1.1222 max mem: 9669 +[14:01:51.335364] Epoch: [21] [ 20/156] eta: 0:00:26 lr: 0.000258 loss: 0.3947 (0.3811) time: 0.1424 data: 0.0003 max mem: 9669 +[14:01:54.183864] Epoch: [21] [ 40/156] eta: 0:00:19 lr: 0.000252 loss: 0.3580 (0.3706) time: 0.1424 data: 0.0002 max mem: 9669 +[14:01:57.035109] Epoch: [21] [ 60/156] eta: 0:00:15 lr: 0.000246 loss: 0.3954 (0.3747) time: 0.1425 data: 0.0003 max mem: 9669 +[14:01:59.886513] Epoch: [21] [ 80/156] eta: 0:00:11 lr: 0.000240 loss: 0.4262 (0.3865) time: 0.1425 data: 0.0002 max mem: 9669 +[14:02:02.741869] Epoch: [21] [100/156] eta: 0:00:08 lr: 0.000233 loss: 0.4166 (0.3932) time: 0.1427 data: 0.0002 max mem: 9669 +[14:02:05.591608] Epoch: [21] [120/156] eta: 0:00:05 lr: 0.000227 loss: 0.3974 (0.3968) time: 0.1424 data: 0.0002 max mem: 9669 +[14:02:08.442491] Epoch: [21] [140/156] eta: 0:00:02 lr: 0.000221 loss: 0.3612 (0.3919) time: 0.1425 data: 0.0002 max mem: 9669 +[14:02:10.580026] Epoch: [21] [155/156] eta: 0:00:00 lr: 0.000217 loss: 0.3752 (0.3934) time: 0.1424 data: 0.0002 max mem: 9669 +[14:02:10.711240] Epoch: [21] Total time: 0:00:23 (0.1506 s / it) +[14:02:10.720374] Averaged stats: lr: 0.000217 loss: 0.3752 (0.3934) +[14:02:11.949526] val: [ 0/17] eta: 0:00:20 loss: 0.3412 (0.3412) time: 1.2106 data: 1.1743 max mem: 9669 +[14:02:12.297074] val: [10/17] eta: 0:00:00 loss: 0.2534 (0.3023) time: 0.1416 data: 0.1069 max mem: 9669 +[14:02:12.501152] val: [16/17] eta: 0:00:00 loss: 0.1760 (0.2378) time: 0.1036 data: 0.0692 max mem: 9669 +[14:02:12.612917] val: Total time: 0:00:01 (0.1103 s / it) +[14:02:12.628199] val loss: 0.23778783990179791 +[14:02:12.628423] Accuracy: 0.9167, F1 Score: 0.9165, ROC AUC: 0.9706, Hamming Loss: 0.0833, + Jaccard Score: 0.8460, Precision: 0.9192, Recall: 0.9167, + Average Precision: 0.9709, Kappa: 0.8333, Score: 0.9068 +[14:02:12.670375] Best epoch = 18, Best score = 0.9106 +[14:02:12.951894] log_dir: ./output_logs/retfound +[14:02:14.307803] Epoch: [22] [ 0/156] eta: 0:03:31 lr: 0.000217 loss: 0.2743 (0.2743) time: 1.3544 data: 1.2081 max mem: 9669 +[14:02:17.162130] Epoch: [22] [ 20/156] eta: 0:00:27 lr: 0.000211 loss: 0.3164 (0.3383) time: 0.1427 data: 0.0002 max mem: 9669 +[14:02:19.669448] Epoch: [22] [ 40/156] eta: 0:00:18 lr: 0.000205 loss: 0.3849 (0.3751) time: 0.1253 data: 0.0002 max mem: 9669 +[14:02:22.514322] Epoch: [22] [ 60/156] eta: 0:00:15 lr: 0.000199 loss: 0.3804 (0.3815) time: 0.1422 data: 0.0002 max mem: 9669 +[14:02:25.368939] Epoch: [22] [ 80/156] eta: 0:00:11 lr: 0.000193 loss: 0.3572 (0.3854) time: 0.1427 data: 0.0001 max mem: 9669 +[14:02:28.221028] Epoch: [22] [100/156] eta: 0:00:08 lr: 0.000187 loss: 0.4054 (0.3910) time: 0.1426 data: 0.0002 max mem: 9669 +[14:02:31.065958] Epoch: [22] [120/156] eta: 0:00:05 lr: 0.000182 loss: 0.3655 (0.3888) time: 0.1422 data: 0.0002 max mem: 9669 +[14:02:33.908281] Epoch: [22] [140/156] eta: 0:00:02 lr: 0.000176 loss: 0.3798 (0.3888) time: 0.1421 data: 0.0002 max mem: 9669 +[14:02:36.049417] Epoch: [22] [155/156] eta: 0:00:00 lr: 0.000172 loss: 0.3645 (0.3888) time: 0.1423 data: 0.0001 max mem: 9669 +[14:02:36.200619] Epoch: [22] Total time: 0:00:23 (0.1490 s / it) +[14:02:36.203377] Averaged stats: lr: 0.000172 loss: 0.3645 (0.3888) +[14:02:37.398164] val: [ 0/17] eta: 0:00:20 loss: 0.2359 (0.2359) time: 1.1774 data: 1.1408 max mem: 9669 +[14:02:37.745100] val: [10/17] eta: 0:00:00 loss: 0.2614 (0.2715) time: 0.1385 data: 0.1039 max mem: 9669 +[14:02:37.944934] val: [16/17] eta: 0:00:00 loss: 0.2143 (0.2386) time: 0.1013 data: 0.0673 max mem: 9669 +[14:02:38.066292] val: Total time: 0:00:01 (0.1086 s / it) +[14:02:38.082495] val loss: 0.2386385027100058 +[14:02:38.082693] Accuracy: 0.9185, F1 Score: 0.9185, ROC AUC: 0.9701, Hamming Loss: 0.0815, + Jaccard Score: 0.8493, Precision: 0.9187, Recall: 0.9185, + Average Precision: 0.9705, Kappa: 0.8370, Score: 0.9086 +[14:02:38.127692] Best epoch = 18, Best score = 0.9106 +[14:02:38.441539] log_dir: ./output_logs/retfound +[14:02:39.670705] Epoch: [23] [ 0/156] eta: 0:03:11 lr: 0.000171 loss: 0.3711 (0.3711) time: 1.2281 data: 1.0819 max mem: 9669 +[14:02:42.511175] Epoch: [23] [ 20/156] eta: 0:00:26 lr: 0.000166 loss: 0.3567 (0.3692) time: 0.1420 data: 0.0002 max mem: 9669 +[14:02:45.361456] Epoch: [23] [ 40/156] eta: 0:00:19 lr: 0.000160 loss: 0.3816 (0.3839) time: 0.1425 data: 0.0002 max mem: 9669 +[14:02:48.206160] Epoch: [23] [ 60/156] eta: 0:00:15 lr: 0.000155 loss: 0.3849 (0.3865) time: 0.1422 data: 0.0002 max mem: 9669 +[14:02:51.051731] Epoch: [23] [ 80/156] eta: 0:00:11 lr: 0.000149 loss: 0.3736 (0.3917) time: 0.1422 data: 0.0002 max mem: 9669 +[14:02:53.899827] Epoch: [23] [100/156] eta: 0:00:08 lr: 0.000144 loss: 0.3504 (0.3864) time: 0.1424 data: 0.0002 max mem: 9669 +[14:02:56.748532] Epoch: [23] [120/156] eta: 0:00:05 lr: 0.000139 loss: 0.3553 (0.3820) time: 0.1424 data: 0.0002 max mem: 9669 +[14:02:59.601194] Epoch: [23] [140/156] eta: 0:00:02 lr: 0.000134 loss: 0.3917 (0.3831) time: 0.1426 data: 0.0002 max mem: 9669 +[14:03:01.736159] Epoch: [23] [155/156] eta: 0:00:00 lr: 0.000130 loss: 0.3600 (0.3816) time: 0.1425 data: 0.0002 max mem: 9669 +[14:03:01.879512] Epoch: [23] Total time: 0:00:23 (0.1502 s / it) +[14:03:01.889415] Averaged stats: lr: 0.000130 loss: 0.3600 (0.3816) +[14:03:03.047057] val: [ 0/17] eta: 0:00:19 loss: 0.3133 (0.3133) time: 1.1446 data: 1.1097 max mem: 9669 +[14:03:03.407535] val: [10/17] eta: 0:00:00 loss: 0.2837 (0.2940) time: 0.1368 data: 0.1020 max mem: 9669 +[14:03:03.611438] val: [16/17] eta: 0:00:00 loss: 0.1864 (0.2400) time: 0.1004 data: 0.0661 max mem: 9669 +[14:03:03.745928] val: Total time: 0:00:01 (0.1085 s / it) +[14:03:03.768859] val loss: 0.23999610104981592 +[14:03:03.769106] Accuracy: 0.9222, F1 Score: 0.9222, ROC AUC: 0.9707, Hamming Loss: 0.0778, + Jaccard Score: 0.8556, Precision: 0.9234, Recall: 0.9222, + Average Precision: 0.9711, Kappa: 0.8444, Score: 0.9124 +[14:03:05.544551] Best epoch = 23, Best score = 0.9124 +[14:03:05.627528] log_dir: ./output_logs/retfound +[14:03:06.917559] Epoch: [24] [ 0/156] eta: 0:03:21 lr: 0.000130 loss: 0.4154 (0.4154) time: 1.2889 data: 1.1416 max mem: 9669 +[14:03:09.772975] Epoch: [24] [ 20/156] eta: 0:00:26 lr: 0.000125 loss: 0.3271 (0.3633) time: 0.1427 data: 0.0002 max mem: 9669 +[14:03:12.624910] Epoch: [24] [ 40/156] eta: 0:00:19 lr: 0.000120 loss: 0.3947 (0.3858) time: 0.1425 data: 0.0001 max mem: 9669 +[14:03:15.478619] Epoch: [24] [ 60/156] eta: 0:00:15 lr: 0.000115 loss: 0.3738 (0.3892) time: 0.1426 data: 0.0002 max mem: 9669 +[14:03:18.328566] Epoch: [24] [ 80/156] eta: 0:00:11 lr: 0.000110 loss: 0.3735 (0.3839) time: 0.1425 data: 0.0002 max mem: 9669 +[14:03:20.830644] Epoch: [24] [100/156] eta: 0:00:08 lr: 0.000105 loss: 0.3391 (0.3819) time: 0.1250 data: 0.0002 max mem: 9669 +[14:03:23.678090] Epoch: [24] [120/156] eta: 0:00:05 lr: 0.000101 loss: 0.3435 (0.3804) time: 0.1423 data: 0.0002 max mem: 9669 +[14:03:26.523163] Epoch: [24] [140/156] eta: 0:00:02 lr: 0.000096 loss: 0.3509 (0.3794) time: 0.1422 data: 0.0002 max mem: 9669 +[14:03:28.659454] Epoch: [24] [155/156] eta: 0:00:00 lr: 0.000093 loss: 0.3651 (0.3807) time: 0.1424 data: 0.0002 max mem: 9669 +[14:03:28.805777] Epoch: [24] Total time: 0:00:23 (0.1486 s / it) +[14:03:28.815165] Averaged stats: lr: 0.000093 loss: 0.3651 (0.3807) +[14:03:30.028510] val: [ 0/17] eta: 0:00:20 loss: 0.2879 (0.2879) time: 1.1999 data: 1.1658 max mem: 9669 +[14:03:30.381705] val: [10/17] eta: 0:00:00 loss: 0.2879 (0.2952) time: 0.1411 data: 0.1062 max mem: 9669 +[14:03:30.586868] val: [16/17] eta: 0:00:00 loss: 0.1953 (0.2397) time: 0.1033 data: 0.0688 max mem: 9669 +[14:03:30.713017] val: Total time: 0:00:01 (0.1109 s / it) +[14:03:30.728328] val loss: 0.23967866029809504 +[14:03:30.728532] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9703, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9217, Recall: 0.9204, + Average Precision: 0.9701, Kappa: 0.8407, Score: 0.9105 +[14:03:30.768583] Best epoch = 23, Best score = 0.9124 +[14:03:31.061672] log_dir: ./output_logs/retfound +[14:03:32.411416] Epoch: [25] [ 0/156] eta: 0:03:30 lr: 0.000092 loss: 0.2706 (0.2706) time: 1.3486 data: 1.2017 max mem: 9669 +[14:03:35.254649] Epoch: [25] [ 20/156] eta: 0:00:27 lr: 0.000088 loss: 0.4050 (0.3961) time: 0.1421 data: 0.0002 max mem: 9669 +[14:03:38.105113] Epoch: [25] [ 40/156] eta: 0:00:19 lr: 0.000084 loss: 0.3465 (0.3772) time: 0.1425 data: 0.0002 max mem: 9669 +[14:03:40.946787] Epoch: [25] [ 60/156] eta: 0:00:15 lr: 0.000079 loss: 0.4074 (0.3937) time: 0.1420 data: 0.0002 max mem: 9669 +[14:03:43.801000] Epoch: [25] [ 80/156] eta: 0:00:11 lr: 0.000075 loss: 0.4057 (0.3941) time: 0.1427 data: 0.0003 max mem: 9669 +[14:03:46.652036] Epoch: [25] [100/156] eta: 0:00:08 lr: 0.000071 loss: 0.3420 (0.3892) time: 0.1425 data: 0.0002 max mem: 9669 +[14:03:49.501739] Epoch: [25] [120/156] eta: 0:00:05 lr: 0.000067 loss: 0.3155 (0.3838) time: 0.1424 data: 0.0002 max mem: 9669 +[14:03:52.363150] Epoch: [25] [140/156] eta: 0:00:02 lr: 0.000064 loss: 0.3565 (0.3815) time: 0.1430 data: 0.0002 max mem: 9669 +[14:03:54.500214] Epoch: [25] [155/156] eta: 0:00:00 lr: 0.000061 loss: 0.3476 (0.3825) time: 0.1426 data: 0.0001 max mem: 9669 +[14:03:54.652508] Epoch: [25] Total time: 0:00:23 (0.1512 s / it) +[14:03:54.661502] Averaged stats: lr: 0.000061 loss: 0.3476 (0.3825) +[14:03:55.853591] val: [ 0/17] eta: 0:00:19 loss: 0.1976 (0.1976) time: 1.1761 data: 1.1401 max mem: 9669 +[14:03:56.201969] val: [10/17] eta: 0:00:00 loss: 0.2580 (0.2567) time: 0.1385 data: 0.1038 max mem: 9669 +[14:03:56.405115] val: [16/17] eta: 0:00:00 loss: 0.2172 (0.2330) time: 0.1015 data: 0.0672 max mem: 9669 +[14:03:56.538552] val: Total time: 0:00:01 (0.1095 s / it) +[14:03:56.563147] val loss: 0.23302041739225388 +[14:03:56.563377] Accuracy: 0.9167, F1 Score: 0.9167, ROC AUC: 0.9718, Hamming Loss: 0.0833, + Jaccard Score: 0.8462, Precision: 0.9167, Recall: 0.9167, + Average Precision: 0.9720, Kappa: 0.8333, Score: 0.9073 +[14:03:56.609391] Best epoch = 23, Best score = 0.9124 +[14:03:56.895992] log_dir: ./output_logs/retfound +[14:03:58.193751] Epoch: [26] [ 0/156] eta: 0:03:22 lr: 0.000061 loss: 0.5383 (0.5383) time: 1.2967 data: 1.1511 max mem: 9669 +[14:04:01.039682] Epoch: [26] [ 20/156] eta: 0:00:26 lr: 0.000057 loss: 0.3742 (0.3995) time: 0.1423 data: 0.0002 max mem: 9669 +[14:04:03.887424] Epoch: [26] [ 40/156] eta: 0:00:19 lr: 0.000053 loss: 0.3442 (0.3819) time: 0.1423 data: 0.0002 max mem: 9669 +[14:04:06.732947] Epoch: [26] [ 60/156] eta: 0:00:15 lr: 0.000050 loss: 0.3598 (0.3880) time: 0.1422 data: 0.0002 max mem: 9669 +[14:04:09.587617] Epoch: [26] [ 80/156] eta: 0:00:11 lr: 0.000047 loss: 0.3191 (0.3707) time: 0.1427 data: 0.0002 max mem: 9669 +[14:04:12.435555] Epoch: [26] [100/156] eta: 0:00:08 lr: 0.000043 loss: 0.3691 (0.3724) time: 0.1424 data: 0.0002 max mem: 9669 +[14:04:15.275237] Epoch: [26] [120/156] eta: 0:00:05 lr: 0.000040 loss: 0.3316 (0.3727) time: 0.1419 data: 0.0002 max mem: 9669 +[14:04:18.117718] Epoch: [26] [140/156] eta: 0:00:02 lr: 0.000037 loss: 0.3928 (0.3753) time: 0.1421 data: 0.0002 max mem: 9669 +[14:04:20.261207] Epoch: [26] [155/156] eta: 0:00:00 lr: 0.000035 loss: 0.3924 (0.3762) time: 0.1427 data: 0.0002 max mem: 9669 +[14:04:20.391298] Epoch: [26] Total time: 0:00:23 (0.1506 s / it) +[14:04:20.399717] Averaged stats: lr: 0.000035 loss: 0.3924 (0.3762) +[14:04:21.700179] val: [ 0/17] eta: 0:00:21 loss: 0.2516 (0.2516) time: 1.2881 data: 1.2523 max mem: 9669 +[14:04:22.041464] val: [10/17] eta: 0:00:01 loss: 0.2528 (0.2728) time: 0.1481 data: 0.1140 max mem: 9669 +[14:04:22.246625] val: [16/17] eta: 0:00:00 loss: 0.2173 (0.2303) time: 0.1078 data: 0.0738 max mem: 9669 +[14:04:22.382159] val: Total time: 0:00:01 (0.1159 s / it) +[14:04:22.396778] val loss: 0.23026803880929947 +[14:04:22.396985] Accuracy: 0.9241, F1 Score: 0.9241, ROC AUC: 0.9719, Hamming Loss: 0.0759, + Jaccard Score: 0.8588, Precision: 0.9245, Recall: 0.9241, + Average Precision: 0.9721, Kappa: 0.8481, Score: 0.9147 +[14:04:24.378716] Best epoch = 26, Best score = 0.9147 +[14:04:24.454489] log_dir: ./output_logs/retfound +[14:04:25.791357] Epoch: [27] [ 0/156] eta: 0:03:28 lr: 0.000035 loss: 0.3272 (0.3272) time: 1.3359 data: 1.1909 max mem: 9669 +[14:04:28.944094] Epoch: [27] [ 20/156] eta: 0:00:29 lr: 0.000032 loss: 0.3556 (0.3730) time: 0.1576 data: 0.0001 max mem: 9669 +[14:04:31.798287] Epoch: [27] [ 40/156] eta: 0:00:20 lr: 0.000030 loss: 0.3624 (0.3686) time: 0.1426 data: 0.0002 max mem: 9669 +[14:04:34.641258] Epoch: [27] [ 60/156] eta: 0:00:16 lr: 0.000027 loss: 0.3282 (0.3590) time: 0.1421 data: 0.0002 max mem: 9669 +[14:04:37.485865] Epoch: [27] [ 80/156] eta: 0:00:12 lr: 0.000025 loss: 0.3779 (0.3642) time: 0.1422 data: 0.0002 max mem: 9669 +[14:04:40.332607] Epoch: [27] [100/156] eta: 0:00:08 lr: 0.000022 loss: 0.3637 (0.3665) time: 0.1423 data: 0.0002 max mem: 9669 +[14:04:43.180707] Epoch: [27] [120/156] eta: 0:00:05 lr: 0.000020 loss: 0.3731 (0.3708) time: 0.1424 data: 0.0002 max mem: 9669 +[14:04:46.024253] Epoch: [27] [140/156] eta: 0:00:02 lr: 0.000018 loss: 0.3775 (0.3725) time: 0.1421 data: 0.0002 max mem: 9669 +[14:04:48.156757] Epoch: [27] [155/156] eta: 0:00:00 lr: 0.000016 loss: 0.4017 (0.3799) time: 0.1421 data: 0.0002 max mem: 9669 +[14:04:48.310432] Epoch: [27] Total time: 0:00:23 (0.1529 s / it) +[14:04:48.318681] Averaged stats: lr: 0.000016 loss: 0.4017 (0.3799) +[14:04:49.430976] val: [ 0/17] eta: 0:00:18 loss: 0.2363 (0.2363) time: 1.1001 data: 1.0653 max mem: 9669 +[14:04:49.801246] val: [10/17] eta: 0:00:00 loss: 0.2602 (0.2659) time: 0.1336 data: 0.0992 max mem: 9669 +[14:04:50.004692] val: [16/17] eta: 0:00:00 loss: 0.2213 (0.2283) time: 0.0984 data: 0.0642 max mem: 9669 +[14:04:50.137586] val: Total time: 0:00:01 (0.1063 s / it) +[14:04:50.152642] val loss: 0.22829091548919678 +[14:04:50.152914] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9718, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9719, Kappa: 0.8556, Score: 0.9184 +[14:04:52.178398] Best epoch = 27, Best score = 0.9184 +[14:04:52.246907] log_dir: ./output_logs/retfound +[14:04:53.620607] Epoch: [28] [ 0/156] eta: 0:03:34 lr: 0.000016 loss: 0.5803 (0.5803) time: 1.3725 data: 1.2181 max mem: 9669 +[14:04:56.467381] Epoch: [28] [ 20/156] eta: 0:00:27 lr: 0.000014 loss: 0.3818 (0.3711) time: 0.1423 data: 0.0002 max mem: 9669 +[14:04:59.310740] Epoch: [28] [ 40/156] eta: 0:00:19 lr: 0.000013 loss: 0.3826 (0.3896) time: 0.1421 data: 0.0002 max mem: 9669 +[14:05:02.150665] Epoch: [28] [ 60/156] eta: 0:00:15 lr: 0.000011 loss: 0.3227 (0.3742) time: 0.1419 data: 0.0003 max mem: 9669 +[14:05:04.999963] Epoch: [28] [ 80/156] eta: 0:00:11 lr: 0.000009 loss: 0.3472 (0.3726) time: 0.1424 data: 0.0002 max mem: 9669 +[14:05:07.842091] Epoch: [28] [100/156] eta: 0:00:08 lr: 0.000008 loss: 0.3824 (0.3784) time: 0.1421 data: 0.0002 max mem: 9669 +[14:05:10.689542] Epoch: [28] [120/156] eta: 0:00:05 lr: 0.000007 loss: 0.3396 (0.3717) time: 0.1423 data: 0.0002 max mem: 9669 +[14:05:13.539219] Epoch: [28] [140/156] eta: 0:00:02 lr: 0.000006 loss: 0.3599 (0.3713) time: 0.1424 data: 0.0002 max mem: 9669 +[14:05:15.678783] Epoch: [28] [155/156] eta: 0:00:00 lr: 0.000005 loss: 0.3999 (0.3723) time: 0.1425 data: 0.0001 max mem: 9669 +[14:05:15.848117] Epoch: [28] Total time: 0:00:23 (0.1513 s / it) +[14:05:15.857348] Averaged stats: lr: 0.000005 loss: 0.3999 (0.3723) +[14:05:17.182925] val: [ 0/17] eta: 0:00:22 loss: 0.2405 (0.2405) time: 1.3076 data: 1.2726 max mem: 9669 +[14:05:17.531097] val: [10/17] eta: 0:00:01 loss: 0.2569 (0.2667) time: 0.1504 data: 0.1158 max mem: 9669 +[14:05:17.736291] val: [16/17] eta: 0:00:00 loss: 0.2159 (0.2277) time: 0.1094 data: 0.0750 max mem: 9669 +[14:05:17.879076] val: Total time: 0:00:02 (0.1179 s / it) +[14:05:17.894058] val loss: 0.22767887702759573 +[14:05:17.894287] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9718, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9718, Kappa: 0.8556, Score: 0.9184 +[14:05:19.736814] Best epoch = 28, Best score = 0.9184 +[14:05:19.808695] log_dir: ./output_logs/retfound +[14:05:21.151897] Epoch: [29] [ 0/156] eta: 0:03:29 lr: 0.000005 loss: 0.3711 (0.3711) time: 1.3421 data: 1.1970 max mem: 9669 +[14:05:24.001202] Epoch: [29] [ 20/156] eta: 0:00:27 lr: 0.000004 loss: 0.3350 (0.3650) time: 0.1424 data: 0.0002 max mem: 9669 +[14:05:26.837594] Epoch: [29] [ 40/156] eta: 0:00:19 lr: 0.000003 loss: 0.3628 (0.3587) time: 0.1418 data: 0.0003 max mem: 9669 +[14:05:29.689098] Epoch: [29] [ 60/156] eta: 0:00:15 lr: 0.000002 loss: 0.3660 (0.3623) time: 0.1425 data: 0.0002 max mem: 9669 +[14:05:32.538475] Epoch: [29] [ 80/156] eta: 0:00:11 lr: 0.000002 loss: 0.3133 (0.3579) time: 0.1424 data: 0.0002 max mem: 9669 +[14:05:35.392030] Epoch: [29] [100/156] eta: 0:00:08 lr: 0.000001 loss: 0.3825 (0.3635) time: 0.1426 data: 0.0002 max mem: 9669 +[14:05:38.247857] Epoch: [29] [120/156] eta: 0:00:05 lr: 0.000001 loss: 0.4103 (0.3692) time: 0.1427 data: 0.0002 max mem: 9669 +[14:05:41.094588] Epoch: [29] [140/156] eta: 0:00:02 lr: 0.000001 loss: 0.3245 (0.3648) time: 0.1423 data: 0.0002 max mem: 9669 +[14:05:43.236370] Epoch: [29] [155/156] eta: 0:00:00 lr: 0.000001 loss: 0.3686 (0.3648) time: 0.1422 data: 0.0002 max mem: 9669 +[14:05:43.390679] Epoch: [29] Total time: 0:00:23 (0.1512 s / it) +[14:05:43.398674] Averaged stats: lr: 0.000001 loss: 0.3686 (0.3648) +[14:05:44.522495] val: [ 0/17] eta: 0:00:18 loss: 0.2422 (0.2422) time: 1.1064 data: 1.0713 max mem: 9669 +[14:05:44.871971] val: [10/17] eta: 0:00:00 loss: 0.2575 (0.2679) time: 0.1323 data: 0.0978 max mem: 9669 +[14:05:45.075080] val: [16/17] eta: 0:00:00 loss: 0.2153 (0.2283) time: 0.0975 data: 0.0633 max mem: 9669 +[14:05:45.206529] val: Total time: 0:00:01 (0.1054 s / it) +[14:05:45.221314] val loss: 0.2282544427058276 +[14:05:45.221543] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9719, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9720, Kappa: 0.8556, Score: 0.9184 +[14:05:47.072535] Best epoch = 29, Best score = 0.9184 +[14:05:50.206771] Test with the best model, epoch = 29: +[14:05:51.178915] test: [ 0/32] eta: 0:00:30 loss: 0.1497 (0.1497) time: 0.9565 data: 0.9186 max mem: 9669 +[14:05:51.526147] test: [10/32] eta: 0:00:02 loss: 0.2625 (0.2276) time: 0.1184 data: 0.0837 max mem: 9669 +[14:05:51.870804] test: [20/32] eta: 0:00:00 loss: 0.2398 (0.2350) time: 0.0345 data: 0.0002 max mem: 9669 +[14:05:52.219085] test: [30/32] eta: 0:00:00 loss: 0.2097 (0.2370) time: 0.0346 data: 0.0002 max mem: 9669 +[14:05:52.333496] test: [31/32] eta: 0:00:00 loss: 0.2398 (0.2516) time: 0.0385 data: 0.0002 max mem: 9669 +[14:05:52.450048] test: Total time: 0:00:02 (0.0696 s / it) +[14:05:52.471620] val loss: 0.2516249555628747 +[14:05:52.471780] Accuracy: 0.9080, F1 Score: 0.9080, ROC AUC: 0.9707, Hamming Loss: 0.0920, + Jaccard Score: 0.8315, Precision: 0.9084, Recall: 0.9080, + Average Precision: 0.9704, Kappa: 0.8160, Score: 0.8982 +[14:05:53.253630] Training time 0:13:14 +[rank0]:[W615 14:05:53.798870046 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/airogs/retfound acc=0.9080 auroc=0.970758 f1_macro=0.9080 qwk=0.8160000000000001 diff --git a/results/airogs/vit/confusion_matrix.png b/results/airogs/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..72857d12e6da60cff8709772d3b66b1c489f06be --- /dev/null +++ b/results/airogs/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:201bd9c92310372e01d3e4a70e83653e1cca02b5db0e461c01a0be0d8cf64087 +size 65570 diff --git a/results/airogs/vit/log.csv b/results/airogs/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..24c833171ea91d2de25d3aced35efec5ec32141a --- /dev/null +++ b/results/airogs/vit/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.5149044979077119,0.7555555555555555,0.9283127572016461,0.7275739375598578,0.00016452991452991454 +1,0.4293446271465375,0.7777777777777778,0.8612139917695473,0.730651257121246,0.0003311965811965812 +2,0.47342362388586384,0.8314814814814815,0.9137174211248286,0.8026648946418073,0.0004978632478632479 +3,0.4123932318045543,0.7203703703703703,0.897798353909465,0.6806619301131271,0.0004983526080898481 +4,0.4123909076054891,0.7685185185185185,0.9021193415637859,0.7329783871837222,0.0004933469513590679 +5,0.3883578437261092,0.8018518518518518,0.8913443072702333,0.7655822842242596,0.00048505044456893006 +6,0.3691409961917462,0.825925925925926,0.9223113854595336,0.7994866233365525,0.00047357528374124746 +7,0.3690838359105281,0.8222222222222222,0.9215363511659809,0.7954690942345263,0.00045907665074076113 +8,0.36253918898411286,0.8314814814814815,0.9298148148148148,0.807657637232313,0.0004417506147066312 +9,0.36002588711487943,0.8222222222222222,0.9029149519890262,0.7897228429051469,0.0004218314805536838 +10,0.35260091225306195,0.8425925925925926,0.9319135802469136,0.8194615078938696,0.00039958862040038377 +11,0.3455605086607811,0.8333333333333334,0.928840877914952,0.8093656100986347,0.0003753228307730364 +12,0.3429072205072794,0.825925925925926,0.9123045267489712,0.7963409516049861,0.0003493622648487805 +13,0.32870546518227994,0.8555555555555555,0.9356172839506173,0.8339249744455345,0.00032205799474680896 +14,0.3155752603824322,0.8444444444444444,0.9200617283950617,0.8177271288382398,0.00029377926388021573 +15,0.302010148190535,0.837037037037037,0.9320370370370371,0.8140830140485313,0.0002649084935722644 +16,0.2751752075094443,0.8611111111111112,0.9386351165980795,0.8403607175132226,0.00023583611146402853 +17,0.27313649979157323,0.8555555555555555,0.936042524005487,0.8342040121704085,0.00020695527165031925 +18,0.2583170288648361,0.8592592592592593,0.9443072702331962,0.8406892234885023,0.00017865653794500142 +19,0.23375292361164704,0.8703703703703703,0.9430864197530865,0.8513273376183618,0.0001513226021754179 +20,0.21957338123749465,0.8685185185185185,0.9455624142661181,0.8502628140997213,0.00012532310893192616 +21,0.2069929434129825,0.8648148148148148,0.9504389574759945,0.8481976769287843,0.00010100965675893233 +22,0.1835992902230758,0.8777777777777778,0.947002743484225,0.8600313907899384,7.871104338774113e-05 +23,0.17139388950398335,0.8740740740740741,0.9502606310013718,0.8573811247399808,5.872881931128568e-05 +24,0.15429930737576422,0.8759259259259259,0.9511522633744856,0.8596261783319527,4.133320983101688e-05 +25,0.1345765313658959,0.8685185185185185,0.9509807956104253,0.8521353042468528,2.675946072326743e-05 +26,0.12600348491030625,0.8685185185185185,0.9509053497942387,0.852087252906954,1.5204656943687972e-05 +27,0.12048482914001514,0.8722222222222222,0.9517832647462279,0.8560584903419605,6.825057391317641e-06 +28,0.10587343633270417,0.8796296296296297,0.9514883401920439,0.8633983039148619,1.733981775034199e-06 +29,0.10432042496708724,0.8796296296296297,0.9512688614540467,0.8633251443355295,2.781588896993981e-10 diff --git a/results/airogs/vit/metrics.json b/results/airogs/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..bd2031bd2a1ad15da5672a2304af8e7b584502bd --- /dev/null +++ b/results/airogs/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.9, + "balanced_accuracy": 0.9, + "precision_macro": 0.9000576082955947, + "recall_macro": 0.9, + "f1_macro": 0.8999963998703953, + "precision_weighted": 0.9000576082955946, + "recall_weighted": 0.9, + "f1_weighted": 0.8999963998703953, + "cohen_kappa": 0.8, + "quadratic_weighted_kappa": 0.8, + "mcc": 0.8000576062215466, + "auroc": 0.9599679999999999, + "auprc": 0.962499989751263, + "sensitivity": 0.906, + "specificity": 0.894, + "precision_pos": 0.8952569169960475, + "f1_pos": 0.9005964214711729, + "per_class": { + "0": { + "precision": 0.9048582995951417, + "recall": 0.894, + "f1-score": 0.8993963782696177, + "support": 500.0 + }, + "1": { + "precision": 0.8952569169960475, + "recall": 0.906, + "f1-score": 0.9005964214711729, + "support": 500.0 + }, + "accuracy": 0.9, + "macro avg": { + "precision": 0.9000576082955947, + "recall": 0.9, + "f1-score": 0.8999963998703953, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.9000576082955946, + "recall": 0.9, + "f1-score": 0.8999963998703953, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/airogs/vit/pr.png b/results/airogs/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..f466ac294e129c6077b822fdc811551ba4b77be8 --- /dev/null +++ b/results/airogs/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:211643544d74e3dd397d7c27278f2a069892c270c09ec9610621d37d66abaa3f +size 44606 diff --git a/results/airogs/vit/roc.png b/results/airogs/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..0f8b1bd25c2af2f9d0e2a645f144c246d5a22480 --- /dev/null +++ b/results/airogs/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:40ef75a1c090a9ef2e9875c4a6efc0f142377cbd1db8be230c7a0d92668a3a42 +size 61437 diff --git a/results/airogs/vit/test_pred.npz b/results/airogs/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..9db7bc36090e9ee2ea11e8860921918e670ae27c --- /dev/null +++ b/results/airogs/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95914c1c34fc1fa6ada2fe1f6724299f448eae40085cffc4e8660442c11c69ad +size 16510 diff --git a/results/airogs/vit/train.log b/results/airogs/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..ec0bcf69631c400170d4b4bab1475d6c02708c37 --- /dev/null +++ b/results/airogs/vit/train.log @@ -0,0 +1,124 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=5000 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.6943 val_acc=0.7093 val_auc=0.7683 score=0.6320 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.5653 val_acc=0.7852 val_auc=0.8664 score=0.7405 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.5144 val_acc=0.8019 val_auc=0.8899 score=0.7644 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.4824 val_acc=0.8278 val_auc=0.9118 score=0.7984 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.4513 val_acc=0.8481 val_auc=0.9219 score=0.8221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.4164 val_acc=0.8463 val_auc=0.9283 score=0.8221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.3999 val_acc=0.8537 val_auc=0.9365 score=0.8320 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.3716 val_acc=0.8704 val_auc=0.9364 score=0.8489 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3723 val_acc=0.8611 val_auc=0.9361 score=0.8394 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.3535 val_acc=0.8611 val_auc=0.9355 score=0.8396 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3337 val_acc=0.8704 val_auc=0.9292 score=0.8465 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3219 val_acc=0.8556 val_auc=0.9335 score=0.8327 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3160 val_acc=0.8815 val_auc=0.9400 score=0.8614 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.3047 val_acc=0.8759 val_auc=0.9381 score=0.8551 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.2920 val_acc=0.8741 val_auc=0.9396 score=0.8539 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.2911 val_acc=0.8685 val_auc=0.9397 score=0.8484 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.2848 val_acc=0.8759 val_auc=0.9394 score=0.8557 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.2747 val_acc=0.8722 val_auc=0.9328 score=0.8497 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.2690 val_acc=0.8759 val_auc=0.9359 score=0.8544 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.2715 val_acc=0.8759 val_auc=0.9286 score=0.8520 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.2643 val_acc=0.8778 val_auc=0.9328 score=0.8554 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.2558 val_acc=0.8759 val_auc=0.9286 score=0.8520 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.2579 val_acc=0.8833 val_auc=0.9300 score=0.8599 +[vit] early stop at ep22 (best ep12 score=0.8614) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=12 best_val_score=0.8614 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/airogs/vit acc=0.8730 auroc=0.945154 f1_macro=0.8730 qwk=0.746 diff --git a/results/aptos/resnet/confusion_matrix.png b/results/aptos/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..c225e9f35f14fc0c04f57a8dc6badd6c7b5d8644 --- /dev/null +++ b/results/aptos/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4aeff92c87baf2e259a3bbdd2a97f700833dd079a6b9ef05869989d9877ea4f1 +size 100198 diff --git a/results/aptos/resnet/log.csv b/results/aptos/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..b3f382c6203a230010bbbb2d75dd700ff62e0770 --- /dev/null +++ b/results/aptos/resnet/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.6022112846374512,0.546448087431694,0.707079345096656,0.4578987430807475,0.00016296296296296295 +1,1.481985298792521,0.5437158469945356,0.8583620293281019,0.5206668333626183,0.00032962962962962964 +2,1.0903354975912305,0.5737704918032787,0.8030857451883981,0.5153118263583831,0.0004962962962962963 +3,0.8820486081971063,0.674863387978142,0.8618185474159583,0.628819203940555,0.0004983838038674314 +4,0.798465426100625,0.7486338797814208,0.8958376298611948,0.7164580866024649,0.0004934094785985891 +5,0.7188928928640154,0.7213114754098361,0.9134211122196007,0.6993799339529772,0.00048514345769673907 +6,0.5981913288434346,0.7868852459016393,0.9134103285436345,0.7240562027491079,0.0004736975249143533 +7,0.5255022015836504,0.7786885245901639,0.9108596857061437,0.7433230297988072,0.0004592264668570001 +8,0.46181285315089754,0.7622950819672131,0.9222729901925139,0.7318364302587627,0.0004419259797600385 +9,0.4122929010126326,0.7923497267759563,0.9224038254664986,0.7557271324289929,0.00042203002303274785 +10,0.35467921131187014,0.76775956284153,0.9131709330618574,0.7328492877585223,0.00039980765535867175 +11,0.2823121948374642,0.7896174863387978,0.9193939034097734,0.7542423123454284,0.0003755593961384883 +12,0.26580536630418566,0.7896174863387978,0.9154692930884206,0.741548401515562,0.0003496131614806527 +13,0.214219435552756,0.7868852459016393,0.920296111622056,0.7479175198248273,0.0003223198296985647 +14,0.20638181881772147,0.8087431693989071,0.9245912696369467,0.7664348039087657,0.000294048496283306 +15,0.16843497256437936,0.8060109289617486,0.916837076808525,0.7615347065821667,0.0002651814825203012 +16,0.14184371584819422,0.8060109289617486,0.9198374109698909,0.7597350478156475,0.0002361091652497979 +17,0.11139711323711607,0.8142076502732241,0.9186266407366391,0.7728060463495772,0.00020722469768978796 +18,0.11305203234983816,0.8005464480874317,0.9196866300613529,0.7630379030488843,0.00017891869271318173 +19,0.09528732962078518,0.8032786885245902,0.9182721864845776,0.7610588035421563,0.0001515739404787926 +20,0.08816374635530842,0.8060109289617486,0.9154554612100976,0.7638113289083576,0.00012556023185110866 +21,0.09019071285923322,0.8060109289617486,0.9157080960336517,0.7626259843203638,0.00010122935761322141 +22,0.08065369634164704,0.825136612021858,0.9163249911242524,0.7821297779496253,7.891035109997563e-05 +23,0.08097618880371253,0.8169398907103825,0.9175168093315029,0.7692978448027649,5.890503858656696e-05 +24,0.07066128568516837,0.8087431693989071,0.9168501427524222,0.763592717697461,4.148395760594798e-05 +25,0.07508751153945922,0.8114754098360656,0.9148993859567416,0.766509643802978,2.688269839279783e-05 +26,0.059848319242397946,0.8169398907103825,0.9143473537667471,0.7732221777097076,1.5298717929749516e-05 +27,0.06323818796210819,0.8278688524590164,0.9161340559730864,0.7812288586535097,6.888669680433918e-06 +28,0.06027366067800257,0.825136612021858,0.9152360310739749,0.777074034392648,1.7662851201208087e-06 +29,0.06106692339397139,0.819672131147541,0.9148171856006198,0.7728869213285655,8.357126202174214e-10 diff --git a/results/aptos/resnet/metrics.json b/results/aptos/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..53a4a6d0b372a48afb269721287e583f0def4b7b --- /dev/null +++ b/results/aptos/resnet/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 366, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.8169398907103825, + "balanced_accuracy": 0.6237892764578934, + "precision_macro": 0.6285607953355876, + "recall_macro": 0.6237892764578934, + "f1_macro": 0.6196658312447786, + "precision_weighted": 0.8194850988984747, + "recall_weighted": 0.8169398907103825, + "f1_weighted": 0.8156237588506785, + "cohen_kappa": 0.7092448333511187, + "quadratic_weighted_kappa": 0.8599771796764883, + "mcc": 0.7100650292891252, + "auroc_macro_ovr": 0.9188339984292486, + "auroc_weighted_ovr": 0.962685518802221, + "auprc_macro": 0.6550468875814671, + "auroc_per_class": { + "0": 0.9991875545391629, + "1": 0.9267857142857143, + "2": 0.9493058130432992, + "3": 0.8507500421371987, + "4": 0.8681408681408682 + }, + "per_class": { + "0": { + "precision": 0.98, + "recall": 0.9849246231155779, + "f1-score": 0.9824561403508771, + "support": 199.0 + }, + "1": { + "precision": 0.5454545454545454, + "recall": 0.6, + "f1-score": 0.5714285714285714, + "support": 30.0 + }, + "2": { + "precision": 0.7415730337078652, + "recall": 0.7586206896551724, + "f1-score": 0.75, + "support": 87.0 + }, + "3": { + "precision": 0.30434782608695654, + "recall": 0.4117647058823529, + "f1-score": 0.35, + "support": 17.0 + }, + "4": { + "precision": 0.5714285714285714, + "recall": 0.36363636363636365, + "f1-score": 0.4444444444444444, + "support": 33.0 + }, + "accuracy": 0.8169398907103825, + "macro avg": { + "precision": 0.6285607953355876, + "recall": 0.6237892764578934, + "f1-score": 0.6196658312447786, + "support": 366.0 + }, + "weighted avg": { + "precision": 0.8194850988984747, + "recall": 0.8169398907103825, + "f1-score": 0.8156237588506785, + "support": 366.0 + } + } +} \ No newline at end of file diff --git a/results/aptos/resnet/pr.png b/results/aptos/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..b595858126b13e182133eb07922930bfc5b508ea --- /dev/null +++ b/results/aptos/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46dee154fdec024f3cc9460be2ad40a78840e0b7a55356a1df88359aad198e7d +size 98404 diff --git a/results/aptos/resnet/roc.png b/results/aptos/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..0d548aa9fa2a42c04d32e99ce1142fe49dde5bc7 --- /dev/null +++ b/results/aptos/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c393998d200dfb8dd75375cf2051207326c97ca85e1e546c26621d543cfd82c6 +size 84348 diff --git a/results/aptos/resnet/test_pred.npz b/results/aptos/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e92361d9a60cef433448ef0f03e37d1ef1759507 --- /dev/null +++ b/results/aptos/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:18816aacbd1058167463c5a8c56af3c7826e0bbf16b6e8ce68f964c8b3d02271 +size 10758 diff --git a/results/aptos/resnet/train.log b/results/aptos/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..940f7e3b76d02add19316d8a75a0153ad04e4bc1 --- /dev/null +++ b/results/aptos/resnet/train.log @@ -0,0 +1,157 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=2930 val=366 test=366 classes=['0', '1', '2', '3', '4'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=1.6022 val_acc=0.5464 val_auc=0.7071 score=0.4579 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=1.4820 val_acc=0.5437 val_auc=0.8584 score=0.5207 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=1.0903 val_acc=0.5738 val_auc=0.8031 score=0.5153 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.8820 val_acc=0.6749 val_auc=0.8618 score=0.6288 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.7985 val_acc=0.7486 val_auc=0.8958 score=0.7165 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.7189 val_acc=0.7213 val_auc=0.9134 score=0.6994 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.5982 val_acc=0.7869 val_auc=0.9134 score=0.7241 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.5255 val_acc=0.7787 val_auc=0.9109 score=0.7433 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.4618 val_acc=0.7623 val_auc=0.9223 score=0.7318 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.4123 val_acc=0.7923 val_auc=0.9224 score=0.7557 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.3547 val_acc=0.7678 val_auc=0.9132 score=0.7328 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.2823 val_acc=0.7896 val_auc=0.9194 score=0.7542 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.2658 val_acc=0.7896 val_auc=0.9155 score=0.7415 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.2142 val_acc=0.7869 val_auc=0.9203 score=0.7479 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.2064 val_acc=0.8087 val_auc=0.9246 score=0.7664 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.1684 val_acc=0.8060 val_auc=0.9168 score=0.7615 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.1418 val_acc=0.8060 val_auc=0.9198 score=0.7597 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.1114 val_acc=0.8142 val_auc=0.9186 score=0.7728 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.1131 val_acc=0.8005 val_auc=0.9197 score=0.7630 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0953 val_acc=0.8033 val_auc=0.9183 score=0.7611 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0882 val_acc=0.8060 val_auc=0.9155 score=0.7638 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0902 val_acc=0.8060 val_auc=0.9157 score=0.7626 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0807 val_acc=0.8251 val_auc=0.9163 score=0.7821 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0810 val_acc=0.8169 val_auc=0.9175 score=0.7693 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0707 val_acc=0.8087 val_auc=0.9169 score=0.7636 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0751 val_acc=0.8115 val_auc=0.9149 score=0.7665 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0598 val_acc=0.8169 val_auc=0.9143 score=0.7732 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0632 val_acc=0.8279 val_auc=0.9161 score=0.7812 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0603 val_acc=0.8251 val_auc=0.9152 score=0.7771 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0611 val_acc=0.8197 val_auc=0.9148 score=0.7729 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=22 best_val_score=0.7821 -> saved test_pred.npz (366 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/aptos/resnet acc=0.8169 auroc_macro_ovr=0.9188339984292486 f1_macro=0.6197 qwk=0.8599771796764883 diff --git a/results/aptos/retfound/confusion_matrix.png b/results/aptos/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..ca82f1f2b39e3e898e9d7420c010443278c26bd0 --- /dev/null +++ b/results/aptos/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8dbd7f186d324a980c74a5fab14e99b3ee8428ed5d7088d9413c894e258144e +size 97806 diff --git a/results/aptos/retfound/confusion_matrix_test.jpg b/results/aptos/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..fb020a3d1ba15cead7073a1cb94f2194305ea36d --- /dev/null +++ b/results/aptos/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f315c3f9f6c2be9e5ee7624fd700b6b2e84921032300a4de4c6c789dc1132ce4 +size 344206 diff --git a/results/aptos/retfound/log.txt b/results/aptos/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..0e23026241dbb1cbbc96dafa60dad2bcb7f50607 --- /dev/null +++ b/results/aptos/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 3.090659340659339e-05, "train_loss": 1.4437423538375687, "epoch": 0, "n_parameters": 303306757} +{"train_lr": 9.340659340659341e-05, "train_loss": 1.1404525002280435, "epoch": 1, "n_parameters": 303306757} +{"train_lr": 0.0001559065934065934, "train_loss": 0.9711205854520693, "epoch": 2, "n_parameters": 303306757} +{"train_lr": 0.0002184065934065934, "train_loss": 0.8870967508672358, "epoch": 3, "n_parameters": 303306757} +{"train_lr": 0.0002809065934065934, "train_loss": 0.8573799447698908, "epoch": 4, "n_parameters": 303306757} +{"train_lr": 0.00034340659340659343, "train_loss": 0.8209066390991211, "epoch": 5, "n_parameters": 303306757} +{"train_lr": 0.0004059065934065934, "train_loss": 0.7632210408593272, "epoch": 6, "n_parameters": 303306757} +{"train_lr": 0.00046840659340659344, "train_loss": 0.76272807769723, "epoch": 7, "n_parameters": 303306757} +{"train_lr": 0.0005309065934065935, "train_loss": 0.74477078298946, "epoch": 8, "n_parameters": 303306757} +{"train_lr": 0.0005934065934065934, "train_loss": 0.7408228568978362, "epoch": 9, "n_parameters": 303306757} +{"train_lr": 0.0006237395619465893, "train_loss": 0.7283486353827047, "epoch": 10, "n_parameters": 303306757} +{"train_lr": 0.000616130334049658, "train_loss": 0.7347935877003513, "epoch": 11, "n_parameters": 303306757} +{"train_lr": 0.0006010570312965236, "train_loss": 0.7046759285769619, "epoch": 12, "n_parameters": 303306757} +{"train_lr": 0.0005788908084263925, "train_loss": 0.6981337610836867, "epoch": 13, "n_parameters": 303306757} +{"train_lr": 0.0005501774714118027, "train_loss": 0.6896972538350703, "epoch": 14, "n_parameters": 303306757} +{"train_lr": 0.0005156240379041557, "train_loss": 0.6764704454076159, "epoch": 15, "n_parameters": 303306757} +{"train_lr": 0.00047608132811268123, "train_loss": 0.6712526210716793, "epoch": 16, "n_parameters": 303306757} +{"train_lr": 0.00043252301478717445, "train_loss": 0.6556910430337046, "epoch": 17, "n_parameters": 303306757} +{"train_lr": 0.0003860216481634639, "train_loss": 0.6352031695973742, "epoch": 18, "n_parameters": 303306757} +{"train_lr": 0.00033772224621701313, "train_loss": 0.6475628640625503, "epoch": 19, "n_parameters": 303306757} +{"train_lr": 0.0002888141005202681, "train_loss": 0.6242544359558231, "epoch": 20, "n_parameters": 303306757} +{"train_lr": 0.00024050149193711215, "train_loss": 0.6223455117299006, "epoch": 21, "n_parameters": 303306757} +{"train_lr": 0.0001939740372312182, "train_loss": 0.6187543328646775, "epoch": 22, "n_parameters": 303306757} +{"train_lr": 0.00015037739675321466, "train_loss": 0.6264173951777783, "epoch": 23, "n_parameters": 303306757} +{"train_lr": 0.00011078506448062389, "train_loss": 0.6020110452568138, "epoch": 24, "n_parameters": 303306757} +{"train_lr": 7.617193503341465e-05, "train_loss": 0.5804416881157801, "epoch": 25, "n_parameters": 303306757} +{"train_lr": 4.7390298532967774e-05, "train_loss": 0.5938168105843303, "epoch": 26, "n_parameters": 303306757} +{"train_lr": 2.5148854390699378e-05, "train_loss": 0.5917471162565462, "epoch": 27, "n_parameters": 303306757} +{"train_lr": 9.995260776513236e-06, "train_loss": 0.5873115884733724, "epoch": 28, "n_parameters": 303306757} +{"train_lr": 2.3026494570844648e-06, "train_loss": 0.5672782646430717, "epoch": 29, "n_parameters": 303306757} diff --git a/results/aptos/retfound/metrics.json b/results/aptos/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..3fe4fd18ad30206ea4daa9444cbe6dbf7bb4c3c6 --- /dev/null +++ b/results/aptos/retfound/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 366, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.8360655737704918, + "balanced_accuracy": 0.6323164905602447, + "precision_macro": 0.695496802900507, + "recall_macro": 0.6323164905602447, + "f1_macro": 0.6495150573381631, + "precision_weighted": 0.8475874112520463, + "recall_weighted": 0.8360655737704918, + "f1_weighted": 0.8342751067728178, + "cohen_kappa": 0.7384188395611726, + "quadratic_weighted_kappa": 0.9055872387300082, + "mcc": 0.74110594625819, + "auroc_macro_ovr": 0.9477939000875149, + "auroc_weighted_ovr": 0.9722894078176452, + "auprc_macro": 0.6933866773014803, + "auroc_per_class": { + "0": 0.9977582523395421, + "1": 0.9365079365079365, + "2": 0.9485230503028055, + "3": 0.910416315523344, + "4": 0.9457639457639457 + }, + "per_class": { + "0": { + "precision": 0.9948979591836735, + "recall": 0.9798994974874372, + "f1-score": 0.9873417721518988, + "support": 199.0 + }, + "1": { + "precision": 0.5161290322580645, + "recall": 0.5333333333333333, + "f1-score": 0.5245901639344263, + "support": 30.0 + }, + "2": { + "precision": 0.6886792452830188, + "recall": 0.8390804597701149, + "f1-score": 0.7564766839378239, + "support": 87.0 + }, + "3": { + "precision": 0.3333333333333333, + "recall": 0.29411764705882354, + "f1-score": 0.3125, + "support": 17.0 + }, + "4": { + "precision": 0.9444444444444444, + "recall": 0.5151515151515151, + "f1-score": 0.6666666666666666, + "support": 33.0 + }, + "accuracy": 0.8360655737704918, + "macro avg": { + "precision": 0.695496802900507, + "recall": 0.6323164905602447, + "f1-score": 0.6495150573381631, + "support": 366.0 + }, + "weighted avg": { + "precision": 0.8475874112520463, + "recall": 0.8360655737704918, + "f1-score": 0.8342751067728178, + "support": 366.0 + } + } +} \ No newline at end of file diff --git a/results/aptos/retfound/metrics_test.csv b/results/aptos/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..a8c91c054ab7366870dd500f9ea02ef2d0621e74 --- /dev/null +++ b/results/aptos/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.4738252603759368,0.8360655737704918,0.6495150573381631,0.9477939000875149,0.06557377049180328,0.5248148148148147,0.695496802900507,0.6323164905602447,0.6933866773014803,0.7384188395611726 diff --git a/results/aptos/retfound/metrics_val.csv b/results/aptos/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..fd80f6e6ce1c6ee6ebb79d97360e86f6744740dc --- /dev/null +++ b/results/aptos/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +1.2315035263697307,0.46994535519125685,0.1278810408921933,0.7944955098877226,0.21202185792349726,0.09398907103825137,0.09398907103825137,0.2,0.3867211592562908,0.0 +0.8268127143383026,0.7131147540983607,0.31758756321408504,0.8652248581521149,0.11475409836065574,0.2742408376963351,0.2826472962066182,0.37191413237924864,0.5284428193254381,0.5372109826589595 +0.7377102474371592,0.73224043715847,0.32853228962818004,0.8997824657852792,0.10710382513661203,0.2899198167239404,0.29526655510399236,0.38613595706618964,0.5879869407179579,0.5719196066262472 +0.6795995533466339,0.7486338797814208,0.3904799006806798,0.9005039822390696,0.1005464480874317,0.32763220439691026,0.4987890137328339,0.4192128801431127,0.5791918168687276,0.6006546644844517 +0.6329679985841116,0.7568306010928961,0.41432941771710324,0.9204328974720948,0.09726775956284153,0.3460937163356089,0.5502568791361895,0.4325185279836442,0.6135659221895626,0.6170287810383748 +0.5618535578250885,0.8142076502732241,0.5667318426969825,0.92790390567637,0.07431693989071038,0.46752030062478467,0.6227950030102348,0.5568042422693587,0.6349961065438106,0.7126096997690531 +0.5954677388072014,0.7568306010928961,0.5639948300811599,0.9250012477031875,0.09726775956284153,0.457286328470092,0.6159944498221792,0.6198971958274283,0.6531166355265329,0.6513091695390609 +0.5684055685997009,0.7950819672131147,0.5348954032391149,0.9206100662528025,0.08196721311475409,0.4389907821925917,0.5909459489339891,0.5477293636595961,0.6369177305275313,0.6875142297709576 +0.5389992818236351,0.7923497267759563,0.6364107898204008,0.9247583098424185,0.08306010928961749,0.5133968009220322,0.6606400814652759,0.6290669795320959,0.6463221534762981,0.6933253952503804 +0.5605512013038,0.8114754098360656,0.6010082188566035,0.9268778818476437,0.07540983606557378,0.4930794459676261,0.6284147983103561,0.6021141649048626,0.6299065714186599,0.7171814455618519 +0.504239079852899,0.819672131147541,0.6325237341246921,0.9344415395253165,0.07213114754098361,0.5177883915289142,0.6520402878628704,0.6288571893223056,0.6678765860311349,0.7300402324541797 +0.5520134015629689,0.825136612021858,0.5798682476943346,0.9331538746422373,0.06994535519125683,0.48337567746668875,0.6254219616232264,0.5737273191924355,0.65670687582927,0.7311047846450547 +0.4989891368895769,0.8415300546448088,0.6323676305420727,0.9370933790841003,0.0633879781420765,0.5383029344634263,0.6412314220306548,0.6319690774341937,0.6752265526512111,0.7590301269098917 +0.5247880853712559,0.825136612021858,0.6032101977763482,0.9346754703392849,0.06994535519125683,0.5048369172057328,0.6104859028873434,0.6197712752363915,0.673344184909815,0.735912872894541 +0.5234006810933352,0.8060109289617486,0.654662768098131,0.9359920793644824,0.07759562841530054,0.5233124474188134,0.6967677401437932,0.6477043886346212,0.6760868714189436,0.7107395698828978 +0.5002523548901081,0.8142076502732241,0.6465158486699762,0.9412562989929161,0.07431693989071038,0.5281743078147934,0.6492952137116849,0.6473897033199358,0.6813090707268843,0.7242815678106929 +0.5038238490621249,0.8142076502732241,0.63906137893754,0.9374069728471144,0.07431693989071038,0.521465141398729,0.6709810858041723,0.6213917477870966,0.6789923401379563,0.7214487173747594 +0.49730656916896504,0.855191256830601,0.6785600810365932,0.9355819622737875,0.05792349726775956,0.5695853053865225,0.7236845196147522,0.6638482447784773,0.6751971272161368,0.7816229116945107 +0.47698313370347023,0.8469945355191257,0.6371604233391712,0.9467824090053538,0.06120218579234973,0.5440311433820313,0.645118404300097,0.6387183746486073,0.6961187449610393,0.7676927959377975 +0.5147952182839314,0.8169398907103825,0.6598394393316873,0.9418474120682031,0.073224043715847,0.5384701649760428,0.6920233100310476,0.6457313616615942,0.6915379774582466,0.727998757681301 +0.489178951519231,0.8469945355191257,0.6959509436129092,0.9413605481615592,0.06120218579234973,0.5792024642256492,0.7194379066085483,0.6795825105127429,0.689474308277099,0.771239787490513 +0.513525198524197,0.8224043715846995,0.6098809776342919,0.9454805620386011,0.07103825136612021,0.5051220671527431,0.6863694828372408,0.5947323606625933,0.6972445027012601,0.7289321361834009 +0.4880747947221001,0.8360655737704918,0.6393503808820774,0.943279215871262,0.06557377049180328,0.5369102353585112,0.6737282626847433,0.6289631298933624,0.7005359724299262,0.7517241379310345 +0.4953958783298731,0.8278688524590164,0.6795574225462315,0.9431010413623419,0.06885245901639345,0.5571730454274257,0.7031119695606233,0.6640580349882675,0.709452066013587,0.7433094358106603 +0.4656097547461589,0.8333333333333334,0.6564027898330224,0.9453025400212768,0.06666666666666667,0.54566119410947,0.6725753020349463,0.647529563459796,0.7145985729024804,0.7505363368195227 +0.4665253447989623,0.8333333333333334,0.6562145731089313,0.9461388571082263,0.06666666666666667,0.5461104037948519,0.6709646577772432,0.647529563459796,0.7131284837290878,0.7507646966889192 +0.4697543305034439,0.8387978142076503,0.6617498121915941,0.9456165061116473,0.06448087431693988,0.5538863973863974,0.6738892778885283,0.6536923541574704,0.7139850460854195,0.7589632540072331 +0.470587865759929,0.8415300546448088,0.673406579201493,0.9451668050462748,0.0633879781420765,0.5622311740349383,0.6884806749488714,0.6627832632483794,0.7145072732148162,0.7632441837121635 +0.47001992507527274,0.8360655737704918,0.660265264749677,0.9449971388523671,0.06557377049180328,0.5517855327338086,0.6723661157963483,0.652529563459796,0.7127790064006028,0.75524943159021 +0.468911811709404,0.8360655737704918,0.660265264749677,0.9452751955659519,0.06557377049180328,0.5517855327338086,0.6723661157963483,0.652529563459796,0.7131238259712397,0.75524943159021 diff --git a/results/aptos/retfound/pr.png b/results/aptos/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..804af2e92b9a03ff10bab0a80235a1c730bf724f --- /dev/null +++ b/results/aptos/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08bf39ab89dd8f4e3cf05ff9079bd939a22916a6fa04ed38280f1bd84dc66074 +size 93213 diff --git a/results/aptos/retfound/roc.png b/results/aptos/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..4aadfb0703845f9eac9cfcd296ff64a85a939988 --- /dev/null +++ b/results/aptos/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b65974fa828a5cc8ac878141cc912e494a2903bea3a3bddca9d5fccedae76c33 +size 84230 diff --git a/results/aptos/retfound/test_pred.npz b/results/aptos/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..25dc3e39517ba41a45907e75f8a5b86021edb7d4 --- /dev/null +++ b/results/aptos/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:775b3148139cba7d7ce6bf28ddfe708ea50cfb0f5e0dbfffe227b42ffd332b49 +size 7098 diff --git a/results/aptos/retfound/train.log b/results/aptos/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9f77e8c7a9827f19cf540213532230a1e1e6c3c1 --- /dev/null +++ b/results/aptos/retfound/train.log @@ -0,0 +1,624 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:59:21.971034836 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:59:22.298495] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:59:22.298777] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/DR/aptos2019', +nb_classes=5, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/aptos', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:59:25.281518] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:59:26.859608] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:59:27.251022] Sampler_train = +[13:59:27.296125] len of train_set: 2912 +[13:59:27.978446] [Adaptation] Full fine-tuning: training all parameters. +[13:59:27.979495] number of trainable params (M): 303.31 +[13:59:27.979593] base lr: 5.00e-03 +[13:59:27.979655] actual lr: 6.25e-04 +[13:59:27.979767] accumulate grad iterations: 1 +[13:59:27.979833] effective batch size: 32 +[13:59:27.982881] criterion = CrossEntropyLoss() +[13:59:27.982974] Start training for 30 epochs +[13:59:27.985175] log_dir: ./output_logs/retfound +[13:59:34.375218] Epoch: [0] [ 0/91] eta: 0:09:41 lr: 0.000000 loss: 1.6094 (1.6094) time: 6.3890 data: 5.4752 max mem: 7340 +[13:59:42.116219] Epoch: [0] [20/91] eta: 0:00:47 lr: 0.000014 loss: 1.6014 (1.5983) time: 0.3870 data: 0.2444 max mem: 9672 +[13:59:49.299302] Epoch: [0] [40/91] eta: 0:00:26 lr: 0.000027 loss: 1.5284 (1.5627) time: 0.3591 data: 0.2173 max mem: 9672 +[13:59:56.642497] Epoch: [0] [60/91] eta: 0:00:14 lr: 0.000041 loss: 1.3995 (1.5058) time: 0.3671 data: 0.2250 max mem: 9672 +[14:00:03.791827] Epoch: [0] [80/91] eta: 0:00:04 lr: 0.000055 loss: 1.3095 (1.4644) time: 0.3574 data: 0.2147 max mem: 9672 +[14:00:07.299650] Epoch: [0] [90/91] eta: 0:00:00 lr: 0.000062 loss: 1.3134 (1.4437) time: 0.3600 data: 0.2188 max mem: 9672 +[14:00:07.381785] Epoch: [0] Total time: 0:00:39 (0.4329 s / it) +[14:00:07.390479] Averaged stats: lr: 0.000062 loss: 1.3134 (1.4437) +[14:00:11.171458] val: [ 0/12] eta: 0:00:45 loss: 0.6172 (0.6172) time: 3.7655 data: 3.7207 max mem: 9672 +[14:00:17.995038] val: [10/12] eta: 0:00:01 loss: 1.2725 (1.1463) time: 0.9626 data: 0.9272 max mem: 9672 +[14:00:18.097475] val: [11/12] eta: 0:00:00 loss: 1.2725 (1.2315) time: 0.8909 data: 0.8499 max mem: 9672 +[14:00:18.171458] val: Total time: 0:00:10 (0.8972 s / it) +[14:00:18.189049] val loss: 1.2315035263697307 +[14:00:18.189290] Accuracy: 0.4699, F1 Score: 0.1279, ROC AUC: 0.7945, Hamming Loss: 0.2120, + Jaccard Score: 0.0940, Precision: 0.0940, Recall: 0.2000, + Average Precision: 0.3867, Kappa: 0.0000, Score: 0.3075 +[14:00:19.815056] Best epoch = 0, Best score = 0.3075 +[14:00:19.893911] log_dir: ./output_logs/retfound +[14:00:24.035072] Epoch: [1] [ 0/91] eta: 0:06:16 lr: 0.000063 loss: 1.3719 (1.3719) time: 4.1402 data: 3.9045 max mem: 9672 +[14:00:31.354187] Epoch: [1] [20/91] eta: 0:00:38 lr: 0.000076 loss: 1.2227 (1.2384) time: 0.3659 data: 0.2236 max mem: 9672 +[14:00:39.641512] Epoch: [1] [40/91] eta: 0:00:24 lr: 0.000090 loss: 1.1427 (1.1971) time: 0.4143 data: 0.2725 max mem: 9672 +[14:00:47.031348] Epoch: [1] [60/91] eta: 0:00:13 lr: 0.000104 loss: 1.1218 (1.1747) time: 0.3694 data: 0.2276 max mem: 9672 +[14:00:54.175419] Epoch: [1] [80/91] eta: 0:00:04 lr: 0.000117 loss: 1.0636 (1.1518) time: 0.3572 data: 0.2152 max mem: 9672 +[14:00:58.007934] Epoch: [1] [90/91] eta: 0:00:00 lr: 0.000124 loss: 1.1021 (1.1405) time: 0.3721 data: 0.2299 max mem: 9672 +[14:00:58.089051] Epoch: [1] Total time: 0:00:38 (0.4197 s / it) +[14:00:58.098012] Averaged stats: lr: 0.000124 loss: 1.1021 (1.1405) +[14:01:01.151614] val: [ 0/12] eta: 0:00:36 loss: 0.0658 (0.0658) time: 3.0413 data: 3.0061 max mem: 9672 +[14:01:07.032147] val: [10/12] eta: 0:00:01 loss: 0.7556 (0.7274) time: 0.8110 data: 0.7768 max mem: 9672 +[14:01:07.050141] val: [11/12] eta: 0:00:00 loss: 0.7556 (0.8268) time: 0.7449 data: 0.7121 max mem: 9672 +[14:01:07.133253] val: Total time: 0:00:09 (0.7520 s / it) +[14:01:07.148799] val loss: 0.8268127143383026 +[14:01:07.149002] Accuracy: 0.7131, F1 Score: 0.3176, ROC AUC: 0.8652, Hamming Loss: 0.1148, + Jaccard Score: 0.2742, Precision: 0.2826, Recall: 0.3719, + Average Precision: 0.5284, Kappa: 0.5372, Score: 0.5733 +[14:01:08.800912] Best epoch = 1, Best score = 0.5733 +[14:01:08.867536] log_dir: ./output_logs/retfound +[14:01:12.880228] Epoch: [2] [ 0/91] eta: 0:06:05 lr: 0.000125 loss: 1.0250 (1.0250) time: 4.0117 data: 3.8655 max mem: 9672 +[14:01:18.756669] Epoch: [2] [20/91] eta: 0:00:33 lr: 0.000139 loss: 1.0101 (1.0346) time: 0.2938 data: 0.1749 max mem: 9672 +[14:01:26.610668] Epoch: [2] [40/91] eta: 0:00:22 lr: 0.000152 loss: 0.9718 (1.0007) time: 0.3927 data: 0.2507 max mem: 9672 +[14:01:34.047109] Epoch: [2] [60/91] eta: 0:00:12 lr: 0.000166 loss: 0.9803 (0.9813) time: 0.3718 data: 0.2296 max mem: 9672 +[14:01:40.873048] Epoch: [2] [80/91] eta: 0:00:04 lr: 0.000180 loss: 0.9735 (0.9730) time: 0.3413 data: 0.1991 max mem: 9672 +[14:01:44.426142] Epoch: [2] [90/91] eta: 0:00:00 lr: 0.000187 loss: 0.9615 (0.9711) time: 0.3637 data: 0.2214 max mem: 9672 +[14:01:44.512691] Epoch: [2] Total time: 0:00:35 (0.3917 s / it) +[14:01:44.521850] Averaged stats: lr: 0.000187 loss: 0.9615 (0.9711) +[14:01:47.596871] val: [ 0/12] eta: 0:00:36 loss: 0.0255 (0.0255) time: 3.0595 data: 3.0237 max mem: 9672 +[14:01:53.630564] val: [10/12] eta: 0:00:01 loss: 0.7558 (0.6465) time: 0.8266 data: 0.7926 max mem: 9672 +[14:01:53.648666] val: [11/12] eta: 0:00:00 loss: 0.7558 (0.7377) time: 0.7592 data: 0.7265 max mem: 9672 +[14:01:53.728281] val: Total time: 0:00:09 (0.7660 s / it) +[14:01:53.744283] val loss: 0.7377102474371592 +[14:01:53.744492] Accuracy: 0.7322, F1 Score: 0.3285, ROC AUC: 0.8998, Hamming Loss: 0.1071, + Jaccard Score: 0.2899, Precision: 0.2953, Recall: 0.3861, + Average Precision: 0.5880, Kappa: 0.5719, Score: 0.6001 +[14:01:55.563566] Best epoch = 2, Best score = 0.6001 +[14:01:55.632552] log_dir: ./output_logs/retfound +[14:01:59.426734] Epoch: [3] [ 0/91] eta: 0:05:45 lr: 0.000188 loss: 0.7898 (0.7898) time: 3.7932 data: 3.6476 max mem: 9672 +[14:02:07.018637] Epoch: [3] [20/91] eta: 0:00:38 lr: 0.000201 loss: 0.8770 (0.9199) time: 0.3795 data: 0.2372 max mem: 9672 +[14:02:14.448540] Epoch: [3] [40/91] eta: 0:00:23 lr: 0.000215 loss: 0.8578 (0.8869) time: 0.3714 data: 0.2301 max mem: 9672 +[14:02:21.270124] Epoch: [3] [60/91] eta: 0:00:13 lr: 0.000229 loss: 0.8679 (0.8845) time: 0.3410 data: 0.1987 max mem: 9672 +[14:02:27.636833] Epoch: [3] [80/91] eta: 0:00:04 lr: 0.000242 loss: 0.8204 (0.8726) time: 0.3183 data: 0.1759 max mem: 9672 +[14:02:31.775146] Epoch: [3] [90/91] eta: 0:00:00 lr: 0.000249 loss: 0.9362 (0.8871) time: 0.3527 data: 0.2104 max mem: 9672 +[14:02:31.873759] Epoch: [3] Total time: 0:00:36 (0.3983 s / it) +[14:02:31.882597] Averaged stats: lr: 0.000249 loss: 0.9362 (0.8871) +[14:02:35.150902] val: [ 0/12] eta: 0:00:39 loss: 0.0152 (0.0152) time: 3.2519 data: 3.2171 max mem: 9672 +[14:02:41.455790] val: [10/12] eta: 0:00:01 loss: 0.5209 (0.5914) time: 0.8687 data: 0.8341 max mem: 9672 +[14:02:41.473561] val: [11/12] eta: 0:00:00 loss: 0.5209 (0.6796) time: 0.7978 data: 0.7646 max mem: 9672 +[14:02:41.552174] val: Total time: 0:00:09 (0.8045 s / it) +[14:02:41.567358] val loss: 0.6795995533466339 +[14:02:41.567559] Accuracy: 0.7486, F1 Score: 0.3905, ROC AUC: 0.9005, Hamming Loss: 0.1005, + Jaccard Score: 0.3276, Precision: 0.4988, Recall: 0.4192, + Average Precision: 0.5792, Kappa: 0.6007, Score: 0.6305 +[14:02:43.304427] Best epoch = 3, Best score = 0.6305 +[14:02:43.378931] log_dir: ./output_logs/retfound +[14:02:47.342109] Epoch: [4] [ 0/91] eta: 0:06:00 lr: 0.000250 loss: 0.7178 (0.7178) time: 3.9621 data: 3.8185 max mem: 9672 +[14:02:54.907988] Epoch: [4] [20/91] eta: 0:00:38 lr: 0.000264 loss: 0.7925 (0.8376) time: 0.3782 data: 0.2365 max mem: 9672 +[14:03:02.659105] Epoch: [4] [40/91] eta: 0:00:23 lr: 0.000277 loss: 0.8144 (0.8371) time: 0.3875 data: 0.2457 max mem: 9672 +[14:03:09.317678] Epoch: [4] [60/91] eta: 0:00:13 lr: 0.000291 loss: 0.8468 (0.8539) time: 0.3329 data: 0.1908 max mem: 9672 +[14:03:16.010823] Epoch: [4] [80/91] eta: 0:00:04 lr: 0.000305 loss: 0.9016 (0.8595) time: 0.3346 data: 0.1927 max mem: 9672 +[14:03:19.656398] Epoch: [4] [90/91] eta: 0:00:00 lr: 0.000312 loss: 0.8429 (0.8574) time: 0.3538 data: 0.2282 max mem: 9672 +[14:03:19.754804] Epoch: [4] Total time: 0:00:36 (0.3997 s / it) +[14:03:19.763715] Averaged stats: lr: 0.000312 loss: 0.8429 (0.8574) +[14:03:22.826737] val: [ 0/12] eta: 0:00:36 loss: 0.0272 (0.0272) time: 3.0458 data: 3.0099 max mem: 9672 +[14:03:29.102771] val: [10/12] eta: 0:00:01 loss: 0.5438 (0.5342) time: 0.8474 data: 0.8125 max mem: 9672 +[14:03:29.123132] val: [11/12] eta: 0:00:00 loss: 0.5438 (0.6330) time: 0.7784 data: 0.7448 max mem: 9672 +[14:03:29.208602] val: Total time: 0:00:09 (0.7857 s / it) +[14:03:29.224355] val loss: 0.6329679985841116 +[14:03:29.224575] Accuracy: 0.7568, F1 Score: 0.4143, ROC AUC: 0.9204, Hamming Loss: 0.0973, + Jaccard Score: 0.3461, Precision: 0.5503, Recall: 0.4325, + Average Precision: 0.6136, Kappa: 0.6170, Score: 0.6506 +[14:03:31.110908] Best epoch = 4, Best score = 0.6506 +[14:03:31.178101] log_dir: ./output_logs/retfound +[14:03:35.594921] Epoch: [5] [ 0/91] eta: 0:06:41 lr: 0.000313 loss: 0.6966 (0.6966) time: 4.4158 data: 4.2704 max mem: 9672 +[14:03:42.784994] Epoch: [5] [20/91] eta: 0:00:39 lr: 0.000326 loss: 0.7755 (0.7861) time: 0.3595 data: 0.2171 max mem: 9672 +[14:03:49.466328] Epoch: [5] [40/91] eta: 0:00:22 lr: 0.000340 loss: 0.8301 (0.8201) time: 0.3340 data: 0.1921 max mem: 9672 +[14:03:56.571642] Epoch: [5] [60/91] eta: 0:00:12 lr: 0.000354 loss: 0.8220 (0.8177) time: 0.3552 data: 0.2134 max mem: 9672 +[14:04:03.830285] Epoch: [5] [80/91] eta: 0:00:04 lr: 0.000367 loss: 0.7394 (0.8138) time: 0.3629 data: 0.2201 max mem: 9672 +[14:04:07.316869] Epoch: [5] [90/91] eta: 0:00:00 lr: 0.000374 loss: 0.8870 (0.8209) time: 0.3537 data: 0.2112 max mem: 9672 +[14:04:07.400704] Epoch: [5] Total time: 0:00:36 (0.3980 s / it) +[14:04:07.409587] Averaged stats: lr: 0.000374 loss: 0.8870 (0.8209) +[14:04:10.611063] val: [ 0/12] eta: 0:00:38 loss: 0.0141 (0.0141) time: 3.1853 data: 3.1539 max mem: 9672 +[14:04:16.767905] val: [10/12] eta: 0:00:01 loss: 0.4332 (0.4869) time: 0.8492 data: 0.8150 max mem: 9672 +[14:04:16.786049] val: [11/12] eta: 0:00:00 loss: 0.4332 (0.5619) time: 0.7799 data: 0.7471 max mem: 9672 +[14:04:16.870001] val: Total time: 0:00:09 (0.7871 s / it) +[14:04:16.888833] val loss: 0.5618535578250885 +[14:04:16.889097] Accuracy: 0.8142, F1 Score: 0.5667, ROC AUC: 0.9279, Hamming Loss: 0.0743, + Jaccard Score: 0.4675, Precision: 0.6228, Recall: 0.5568, + Average Precision: 0.6350, Kappa: 0.7126, Score: 0.7357 +[14:04:18.667160] Best epoch = 5, Best score = 0.7357 +[14:04:18.736949] log_dir: ./output_logs/retfound +[14:04:22.650688] Epoch: [6] [ 0/91] eta: 0:05:56 lr: 0.000375 loss: 0.7652 (0.7652) time: 3.9127 data: 3.7685 max mem: 9672 +[14:04:29.096645] Epoch: [6] [20/91] eta: 0:00:35 lr: 0.000389 loss: 0.7707 (0.7641) time: 0.3222 data: 0.1798 max mem: 9672 +[14:04:35.797521] Epoch: [6] [40/91] eta: 0:00:21 lr: 0.000402 loss: 0.7663 (0.7666) time: 0.3350 data: 0.1928 max mem: 9672 +[14:04:43.208832] Epoch: [6] [60/91] eta: 0:00:12 lr: 0.000416 loss: 0.6955 (0.7563) time: 0.3705 data: 0.2282 max mem: 9672 +[14:04:50.510642] Epoch: [6] [80/91] eta: 0:00:04 lr: 0.000430 loss: 0.7794 (0.7584) time: 0.3650 data: 0.2229 max mem: 9672 +[14:04:54.291186] Epoch: [6] [90/91] eta: 0:00:00 lr: 0.000437 loss: 0.7794 (0.7632) time: 0.3556 data: 0.2135 max mem: 9672 +[14:04:54.377051] Epoch: [6] Total time: 0:00:35 (0.3916 s / it) +[14:04:54.385989] Averaged stats: lr: 0.000437 loss: 0.7794 (0.7632) +[14:04:57.434549] val: [ 0/12] eta: 0:00:36 loss: 0.0317 (0.0317) time: 3.0322 data: 2.9980 max mem: 9672 +[14:05:03.666573] val: [10/12] eta: 0:00:01 loss: 0.7199 (0.5995) time: 0.8421 data: 0.8077 max mem: 9672 +[14:05:03.687094] val: [11/12] eta: 0:00:00 loss: 0.5516 (0.5955) time: 0.7736 data: 0.7404 max mem: 9672 +[14:05:03.766114] val: Total time: 0:00:09 (0.7804 s / it) +[14:05:03.781512] val loss: 0.5954677388072014 +[14:05:03.781772] Accuracy: 0.7568, F1 Score: 0.5640, ROC AUC: 0.9250, Hamming Loss: 0.0973, + Jaccard Score: 0.4573, Precision: 0.6160, Recall: 0.6199, + Average Precision: 0.6531, Kappa: 0.6513, Score: 0.7134 +[14:05:03.829047] Best epoch = 5, Best score = 0.7357 +[14:05:04.113277] log_dir: ./output_logs/retfound +[14:05:07.801075] Epoch: [7] [ 0/91] eta: 0:05:35 lr: 0.000438 loss: 0.7387 (0.7387) time: 3.6869 data: 3.5438 max mem: 9672 +[14:05:14.139003] Epoch: [7] [20/91] eta: 0:00:33 lr: 0.000451 loss: 0.8164 (0.7937) time: 0.3169 data: 0.1747 max mem: 9672 +[14:05:21.684159] Epoch: [7] [40/91] eta: 0:00:21 lr: 0.000465 loss: 0.7664 (0.7852) time: 0.3772 data: 0.2347 max mem: 9672 +[14:05:28.864071] Epoch: [7] [60/91] eta: 0:00:12 lr: 0.000479 loss: 0.7639 (0.7731) time: 0.3589 data: 0.2169 max mem: 9672 +[14:05:36.195499] Epoch: [7] [80/91] eta: 0:00:04 lr: 0.000492 loss: 0.7229 (0.7576) time: 0.3665 data: 0.2245 max mem: 9672 +[14:05:39.274220] Epoch: [7] [90/91] eta: 0:00:00 lr: 0.000499 loss: 0.7662 (0.7627) time: 0.3233 data: 0.1814 max mem: 9672 +[14:05:39.357758] Epoch: [7] Total time: 0:00:35 (0.3873 s / it) +[14:05:39.366920] Averaged stats: lr: 0.000499 loss: 0.7662 (0.7627) +[14:05:42.467825] val: [ 0/12] eta: 0:00:37 loss: 0.0527 (0.0527) time: 3.0853 data: 3.0511 max mem: 9672 +[14:05:48.581213] val: [10/12] eta: 0:00:01 loss: 0.5065 (0.5435) time: 0.8362 data: 0.8016 max mem: 9672 +[14:05:48.601957] val: [11/12] eta: 0:00:00 loss: 0.5065 (0.5684) time: 0.7682 data: 0.7348 max mem: 9672 +[14:05:48.685992] val: Total time: 0:00:09 (0.7753 s / it) +[14:05:48.701676] val loss: 0.5684055685997009 +[14:05:48.701929] Accuracy: 0.7951, F1 Score: 0.5349, ROC AUC: 0.9206, Hamming Loss: 0.0820, + Jaccard Score: 0.4390, Precision: 0.5909, Recall: 0.5477, + Average Precision: 0.6369, Kappa: 0.6875, Score: 0.7143 +[14:05:48.735373] Best epoch = 5, Best score = 0.7357 +[14:05:49.009558] log_dir: ./output_logs/retfound +[14:05:53.173161] Epoch: [8] [ 0/91] eta: 0:06:18 lr: 0.000500 loss: 0.7980 (0.7980) time: 4.1626 data: 4.0178 max mem: 9672 +[14:06:00.036114] Epoch: [8] [20/91] eta: 0:00:37 lr: 0.000514 loss: 0.7540 (0.7525) time: 0.3431 data: 0.2009 max mem: 9672 +[14:06:07.686406] Epoch: [8] [40/91] eta: 0:00:23 lr: 0.000527 loss: 0.7310 (0.7568) time: 0.3825 data: 0.2401 max mem: 9672 +[14:06:15.209499] Epoch: [8] [60/91] eta: 0:00:13 lr: 0.000541 loss: 0.7176 (0.7489) time: 0.3761 data: 0.2337 max mem: 9672 +[14:06:22.742953] Epoch: [8] [80/91] eta: 0:00:04 lr: 0.000555 loss: 0.7418 (0.7461) time: 0.3766 data: 0.2350 max mem: 9672 +[14:06:26.509525] Epoch: [8] [90/91] eta: 0:00:00 lr: 0.000562 loss: 0.7055 (0.7448) time: 0.3719 data: 0.2294 max mem: 9672 +[14:06:26.590396] Epoch: [8] Total time: 0:00:37 (0.4130 s / it) +[14:06:26.600215] Averaged stats: lr: 0.000562 loss: 0.7055 (0.7448) +[14:06:29.694083] val: [ 0/12] eta: 0:00:36 loss: 0.0115 (0.0115) time: 3.0771 data: 3.0417 max mem: 9672 +[14:06:35.708906] val: [10/12] eta: 0:00:01 loss: 0.7741 (0.5050) time: 0.8265 data: 0.7919 max mem: 9672 +[14:06:35.729443] val: [11/12] eta: 0:00:00 loss: 0.7741 (0.5390) time: 0.7592 data: 0.7259 max mem: 9672 +[14:06:35.814088] val: Total time: 0:00:09 (0.7665 s / it) +[14:06:35.829400] val loss: 0.5389992818236351 +[14:06:35.829603] Accuracy: 0.7923, F1 Score: 0.6364, ROC AUC: 0.9248, Hamming Loss: 0.0831, + Jaccard Score: 0.5134, Precision: 0.6606, Recall: 0.6291, + Average Precision: 0.6463, Kappa: 0.6933, Score: 0.7515 +[14:06:37.553065] Best epoch = 8, Best score = 0.7515 +[14:06:37.621704] log_dir: ./output_logs/retfound +[14:06:41.864707] Epoch: [9] [ 0/91] eta: 0:06:26 lr: 0.000562 loss: 0.9525 (0.9525) time: 4.2419 data: 4.0966 max mem: 9672 +[14:06:48.205219] Epoch: [9] [20/91] eta: 0:00:35 lr: 0.000576 loss: 0.7448 (0.7528) time: 0.3170 data: 0.1744 max mem: 9672 +[14:06:55.818043] Epoch: [9] [40/91] eta: 0:00:22 lr: 0.000590 loss: 0.7558 (0.7546) time: 0.3806 data: 0.2385 max mem: 9672 +[14:07:03.426494] Epoch: [9] [60/91] eta: 0:00:13 lr: 0.000604 loss: 0.7262 (0.7594) time: 0.3804 data: 0.2378 max mem: 9672 +[14:07:10.256666] Epoch: [9] [80/91] eta: 0:00:04 lr: 0.000617 loss: 0.7083 (0.7469) time: 0.3414 data: 0.1990 max mem: 9672 +[14:07:14.476818] Epoch: [9] [90/91] eta: 0:00:00 lr: 0.000624 loss: 0.7083 (0.7408) time: 0.3654 data: 0.2225 max mem: 9672 +[14:07:14.553890] Epoch: [9] Total time: 0:00:36 (0.4058 s / it) +[14:07:14.563325] Averaged stats: lr: 0.000624 loss: 0.7083 (0.7408) +[14:07:17.738124] val: [ 0/12] eta: 0:00:37 loss: 0.1410 (0.1410) time: 3.1561 data: 3.1215 max mem: 9672 +[14:07:23.783197] val: [10/12] eta: 0:00:01 loss: 0.5690 (0.5011) time: 0.8364 data: 0.8019 max mem: 9672 +[14:07:23.804020] val: [11/12] eta: 0:00:00 loss: 0.5690 (0.5606) time: 0.7684 data: 0.7350 max mem: 9672 +[14:07:23.888454] val: Total time: 0:00:09 (0.7756 s / it) +[14:07:23.911754] val loss: 0.5605512013038 +[14:07:23.912030] Accuracy: 0.8115, F1 Score: 0.6010, ROC AUC: 0.9269, Hamming Loss: 0.0754, + Jaccard Score: 0.4931, Precision: 0.6284, Recall: 0.6021, + Average Precision: 0.6299, Kappa: 0.7172, Score: 0.7484 +[14:07:23.953398] Best epoch = 8, Best score = 0.7515 +[14:07:24.206699] log_dir: ./output_logs/retfound +[14:07:28.431860] Epoch: [10] [ 0/91] eta: 0:06:24 lr: 0.000625 loss: 0.4936 (0.4936) time: 4.2242 data: 4.0999 max mem: 9672 +[14:07:36.066258] Epoch: [10] [20/91] eta: 0:00:40 lr: 0.000625 loss: 0.7060 (0.7350) time: 0.3817 data: 0.2398 max mem: 9672 +[14:07:42.933447] Epoch: [10] [40/91] eta: 0:00:23 lr: 0.000624 loss: 0.7020 (0.7405) time: 0.3433 data: 0.2009 max mem: 9672 +[14:07:49.208044] Epoch: [10] [60/91] eta: 0:00:12 lr: 0.000623 loss: 0.7473 (0.7341) time: 0.3137 data: 0.1718 max mem: 9672 +[14:07:55.932097] Epoch: [10] [80/91] eta: 0:00:04 lr: 0.000622 loss: 0.6388 (0.7204) time: 0.3362 data: 0.1939 max mem: 9672 +[14:07:59.533695] Epoch: [10] [90/91] eta: 0:00:00 lr: 0.000621 loss: 0.7075 (0.7283) time: 0.3538 data: 0.2118 max mem: 9672 +[14:07:59.611084] Epoch: [10] Total time: 0:00:35 (0.3891 s / it) +[14:07:59.620271] Averaged stats: lr: 0.000621 loss: 0.7075 (0.7283) +[14:08:02.753634] val: [ 0/12] eta: 0:00:37 loss: 0.0221 (0.0221) time: 3.1129 data: 3.0755 max mem: 9672 +[14:08:08.796941] val: [10/12] eta: 0:00:01 loss: 0.6431 (0.4712) time: 0.8323 data: 0.7976 max mem: 9672 +[14:08:08.819882] val: [11/12] eta: 0:00:00 loss: 0.6431 (0.5042) time: 0.7648 data: 0.7312 max mem: 9672 +[14:08:08.895870] val: Total time: 0:00:09 (0.7713 s / it) +[14:08:08.910761] val loss: 0.504239079852899 +[14:08:08.910962] Accuracy: 0.8197, F1 Score: 0.6325, ROC AUC: 0.9344, Hamming Loss: 0.0721, + Jaccard Score: 0.5178, Precision: 0.6520, Recall: 0.6289, + Average Precision: 0.6679, Kappa: 0.7300, Score: 0.7657 +[14:08:10.649436] Best epoch = 10, Best score = 0.7657 +[14:08:10.722771] log_dir: ./output_logs/retfound +[14:08:15.742388] Epoch: [11] [ 0/91] eta: 0:07:36 lr: 0.000621 loss: 0.8411 (0.8411) time: 5.0186 data: 4.8775 max mem: 9672 +[14:08:22.459869] Epoch: [11] [20/91] eta: 0:00:39 lr: 0.000619 loss: 0.7493 (0.7794) time: 0.3358 data: 0.1937 max mem: 9672 +[14:08:29.603746] Epoch: [11] [40/91] eta: 0:00:23 lr: 0.000617 loss: 0.7291 (0.7594) time: 0.3571 data: 0.2152 max mem: 9672 +[14:08:36.836358] Epoch: [11] [60/91] eta: 0:00:13 lr: 0.000614 loss: 0.6661 (0.7322) time: 0.3616 data: 0.2205 max mem: 9672 +[14:08:43.494971] Epoch: [11] [80/91] eta: 0:00:04 lr: 0.000612 loss: 0.8034 (0.7373) time: 0.3329 data: 0.1911 max mem: 9672 +[14:08:46.321445] Epoch: [11] [90/91] eta: 0:00:00 lr: 0.000610 loss: 0.7284 (0.7348) time: 0.2816 data: 0.1392 max mem: 9672 +[14:08:46.404001] Epoch: [11] Total time: 0:00:35 (0.3921 s / it) +[14:08:46.412171] Averaged stats: lr: 0.000610 loss: 0.7284 (0.7348) +[14:08:49.523655] val: [ 0/12] eta: 0:00:37 loss: 0.0066 (0.0066) time: 3.0945 data: 3.0579 max mem: 9672 +[14:08:55.662705] val: [10/12] eta: 0:00:01 loss: 0.3304 (0.4487) time: 0.8394 data: 0.8047 max mem: 9672 +[14:08:55.680718] val: [11/12] eta: 0:00:00 loss: 0.3304 (0.5520) time: 0.7709 data: 0.7377 max mem: 9672 +[14:08:55.761350] val: Total time: 0:00:09 (0.7777 s / it) +[14:08:55.775901] val loss: 0.5520134015629689 +[14:08:55.776137] Accuracy: 0.8251, F1 Score: 0.5799, ROC AUC: 0.9332, Hamming Loss: 0.0699, + Jaccard Score: 0.4834, Precision: 0.6254, Recall: 0.5737, + Average Precision: 0.6567, Kappa: 0.7311, Score: 0.7480 +[14:08:55.819452] Best epoch = 10, Best score = 0.7657 +[14:08:56.132420] log_dir: ./output_logs/retfound +[14:09:00.762940] Epoch: [12] [ 0/91] eta: 0:07:01 lr: 0.000610 loss: 0.7904 (0.7904) time: 4.6294 data: 4.4827 max mem: 9672 +[14:09:07.997830] Epoch: [12] [20/91] eta: 0:00:40 lr: 0.000606 loss: 0.7158 (0.7017) time: 0.3617 data: 0.2431 max mem: 9672 +[14:09:14.353321] Epoch: [12] [40/91] eta: 0:00:22 lr: 0.000602 loss: 0.6729 (0.7007) time: 0.3177 data: 0.1763 max mem: 9672 +[14:09:21.143484] Epoch: [12] [60/91] eta: 0:00:12 lr: 0.000598 loss: 0.6434 (0.7094) time: 0.3395 data: 0.1973 max mem: 9672 +[14:09:28.038131] Epoch: [12] [80/91] eta: 0:00:04 lr: 0.000594 loss: 0.6479 (0.6966) time: 0.3447 data: 0.2032 max mem: 9672 +[14:09:31.643562] Epoch: [12] [90/91] eta: 0:00:00 lr: 0.000591 loss: 0.6858 (0.7047) time: 0.3690 data: 0.2271 max mem: 9672 +[14:09:31.720386] Epoch: [12] Total time: 0:00:35 (0.3911 s / it) +[14:09:31.729679] Averaged stats: lr: 0.000591 loss: 0.6858 (0.7047) +[14:09:34.739239] val: [ 0/12] eta: 0:00:35 loss: 0.0022 (0.0022) time: 2.9936 data: 2.9594 max mem: 9672 +[14:09:40.743321] val: [10/12] eta: 0:00:01 loss: 0.5109 (0.4613) time: 0.8179 data: 0.7834 max mem: 9672 +[14:09:40.763777] val: [11/12] eta: 0:00:00 loss: 0.5109 (0.4990) time: 0.7514 data: 0.7182 max mem: 9672 +[14:09:40.837488] val: Total time: 0:00:09 (0.7577 s / it) +[14:09:40.852721] val loss: 0.4989891368895769 +[14:09:40.852908] Accuracy: 0.8415, F1 Score: 0.6324, ROC AUC: 0.9371, Hamming Loss: 0.0634, + Jaccard Score: 0.5383, Precision: 0.6412, Recall: 0.6320, + Average Precision: 0.6752, Kappa: 0.7590, Score: 0.7762 +[14:09:42.487650] Best epoch = 12, Best score = 0.7762 +[14:09:42.560226] log_dir: ./output_logs/retfound +[14:09:46.522056] Epoch: [13] [ 0/91] eta: 0:06:00 lr: 0.000591 loss: 0.7664 (0.7664) time: 3.9606 data: 3.8136 max mem: 9672 +[14:09:53.125025] Epoch: [13] [20/91] eta: 0:00:35 lr: 0.000586 loss: 0.6660 (0.6816) time: 0.3301 data: 0.2096 max mem: 9672 +[14:09:59.834397] Epoch: [13] [40/91] eta: 0:00:21 lr: 0.000581 loss: 0.6605 (0.6725) time: 0.3354 data: 0.1935 max mem: 9672 +[14:10:06.336346] Epoch: [13] [60/91] eta: 0:00:12 lr: 0.000575 loss: 0.6882 (0.6810) time: 0.3250 data: 0.1828 max mem: 9672 +[14:10:14.707388] Epoch: [13] [80/91] eta: 0:00:04 lr: 0.000569 loss: 0.6988 (0.6893) time: 0.4185 data: 0.2764 max mem: 9672 +[14:10:18.237460] Epoch: [13] [90/91] eta: 0:00:00 lr: 0.000566 loss: 0.6661 (0.6981) time: 0.3703 data: 0.2280 max mem: 9672 +[14:10:18.312104] Epoch: [13] Total time: 0:00:35 (0.3929 s / it) +[14:10:18.319823] Averaged stats: lr: 0.000566 loss: 0.6661 (0.6981) +[14:10:21.316869] val: [ 0/12] eta: 0:00:35 loss: 0.0071 (0.0071) time: 2.9809 data: 2.9458 max mem: 9672 +[14:10:27.182164] val: [10/12] eta: 0:00:01 loss: 0.5306 (0.5171) time: 0.8041 data: 0.7705 max mem: 9672 +[14:10:27.202985] val: [11/12] eta: 0:00:00 loss: 0.5306 (0.5248) time: 0.7388 data: 0.7063 max mem: 9672 +[14:10:27.275999] val: Total time: 0:00:08 (0.7450 s / it) +[14:10:27.292357] val loss: 0.5247880853712559 +[14:10:27.292552] Accuracy: 0.8251, F1 Score: 0.6032, ROC AUC: 0.9347, Hamming Loss: 0.0699, + Jaccard Score: 0.5048, Precision: 0.6105, Recall: 0.6198, + Average Precision: 0.6733, Kappa: 0.7359, Score: 0.7579 +[14:10:27.336927] Best epoch = 12, Best score = 0.7762 +[14:10:27.575879] log_dir: ./output_logs/retfound +[14:10:31.349762] Epoch: [14] [ 0/91] eta: 0:05:43 lr: 0.000565 loss: 0.8356 (0.8356) time: 3.7730 data: 3.6292 max mem: 9672 +[14:10:39.157290] Epoch: [14] [20/91] eta: 0:00:39 lr: 0.000559 loss: 0.7097 (0.6962) time: 0.3903 data: 0.2696 max mem: 9672 +[14:10:46.239583] Epoch: [14] [40/91] eta: 0:00:23 lr: 0.000552 loss: 0.6538 (0.6749) time: 0.3541 data: 0.2128 max mem: 9672 +[14:10:53.824648] Epoch: [14] [60/91] eta: 0:00:13 lr: 0.000545 loss: 0.6295 (0.6719) time: 0.3792 data: 0.2372 max mem: 9672 +[14:10:59.888271] Epoch: [14] [80/91] eta: 0:00:04 lr: 0.000538 loss: 0.7297 (0.6909) time: 0.3031 data: 0.1619 max mem: 9672 +[14:11:03.164317] Epoch: [14] [90/91] eta: 0:00:00 lr: 0.000534 loss: 0.7159 (0.6897) time: 0.3008 data: 0.1595 max mem: 9672 +[14:11:03.245435] Epoch: [14] Total time: 0:00:35 (0.3920 s / it) +[14:11:03.255127] Averaged stats: lr: 0.000534 loss: 0.7159 (0.6897) +[14:11:06.255099] val: [ 0/12] eta: 0:00:35 loss: 0.0015 (0.0015) time: 2.9844 data: 2.9522 max mem: 9672 +[14:11:12.313862] val: [10/12] eta: 0:00:01 loss: 0.6190 (0.4964) time: 0.8220 data: 0.7877 max mem: 9672 +[14:11:12.331606] val: [11/12] eta: 0:00:00 loss: 0.6190 (0.5234) time: 0.7550 data: 0.7220 max mem: 9672 +[14:11:12.407298] val: Total time: 0:00:09 (0.7614 s / it) +[14:11:12.422002] val loss: 0.5234006810933352 +[14:11:12.422221] Accuracy: 0.8060, F1 Score: 0.6547, ROC AUC: 0.9360, Hamming Loss: 0.0776, + Jaccard Score: 0.5233, Precision: 0.6968, Recall: 0.6477, + Average Precision: 0.6761, Kappa: 0.7107, Score: 0.7671 +[14:11:12.476498] Best epoch = 12, Best score = 0.7762 +[14:11:12.746407] log_dir: ./output_logs/retfound +[14:11:16.471791] Epoch: [15] [ 0/91] eta: 0:05:38 lr: 0.000534 loss: 0.4600 (0.4600) time: 3.7244 data: 3.5781 max mem: 9672 +[14:11:23.970227] Epoch: [15] [20/91] eta: 0:00:37 lr: 0.000526 loss: 0.6203 (0.6508) time: 0.3749 data: 0.2331 max mem: 9672 +[14:11:30.304925] Epoch: [15] [40/91] eta: 0:00:21 lr: 0.000518 loss: 0.7660 (0.7024) time: 0.3167 data: 0.1764 max mem: 9672 +[14:11:37.450497] Epoch: [15] [60/91] eta: 0:00:12 lr: 0.000510 loss: 0.6541 (0.6905) time: 0.3572 data: 0.2152 max mem: 9672 +[14:11:44.184283] Epoch: [15] [80/91] eta: 0:00:04 lr: 0.000501 loss: 0.6523 (0.6860) time: 0.3367 data: 0.1945 max mem: 9672 +[14:11:47.662529] Epoch: [15] [90/91] eta: 0:00:00 lr: 0.000497 loss: 0.5760 (0.6765) time: 0.3513 data: 0.2092 max mem: 9672 +[14:11:47.742366] Epoch: [15] Total time: 0:00:34 (0.3846 s / it) +[14:11:47.751447] Averaged stats: lr: 0.000497 loss: 0.5760 (0.6765) +[14:11:50.880873] val: [ 0/12] eta: 0:00:37 loss: 0.0013 (0.0013) time: 3.1127 data: 3.0762 max mem: 9672 +[14:11:57.148429] val: [10/12] eta: 0:00:01 loss: 0.6298 (0.4533) time: 0.8527 data: 0.8184 max mem: 9672 +[14:11:57.169041] val: [11/12] eta: 0:00:00 loss: 0.6298 (0.5003) time: 0.7833 data: 0.7502 max mem: 9672 +[14:11:57.242564] val: Total time: 0:00:09 (0.7896 s / it) +[14:11:57.258275] val loss: 0.5002523548901081 +[14:11:57.258468] Accuracy: 0.8142, F1 Score: 0.6465, ROC AUC: 0.9413, Hamming Loss: 0.0743, + Jaccard Score: 0.5282, Precision: 0.6493, Recall: 0.6474, + Average Precision: 0.6813, Kappa: 0.7243, Score: 0.7707 +[14:11:57.301402] Best epoch = 12, Best score = 0.7762 +[14:11:57.586906] log_dir: ./output_logs/retfound +[14:12:01.369985] Epoch: [16] [ 0/91] eta: 0:05:44 lr: 0.000496 loss: 0.7322 (0.7322) time: 3.7821 data: 3.6381 max mem: 9672 +[14:12:08.553691] Epoch: [16] [20/91] eta: 0:00:37 lr: 0.000488 loss: 0.6899 (0.6611) time: 0.3591 data: 0.2171 max mem: 9672 +[14:12:15.683670] Epoch: [16] [40/91] eta: 0:00:22 lr: 0.000479 loss: 0.6794 (0.6616) time: 0.3565 data: 0.2146 max mem: 9672 +[14:12:22.341991] Epoch: [16] [60/91] eta: 0:00:12 lr: 0.000469 loss: 0.6606 (0.6695) time: 0.3329 data: 0.1910 max mem: 9672 +[14:12:30.063176] Epoch: [16] [80/91] eta: 0:00:04 lr: 0.000460 loss: 0.5693 (0.6584) time: 0.3860 data: 0.2437 max mem: 9672 +[14:12:33.764105] Epoch: [16] [90/91] eta: 0:00:00 lr: 0.000455 loss: 0.6183 (0.6713) time: 0.3451 data: 0.2023 max mem: 9672 +[14:12:33.844247] Epoch: [16] Total time: 0:00:36 (0.3984 s / it) +[14:12:33.852866] Averaged stats: lr: 0.000455 loss: 0.6183 (0.6713) +[14:12:36.882705] val: [ 0/12] eta: 0:00:36 loss: 0.0131 (0.0131) time: 3.0120 data: 2.9762 max mem: 9672 +[14:12:42.955544] val: [10/12] eta: 0:00:01 loss: 0.5364 (0.4610) time: 0.8258 data: 0.7911 max mem: 9672 +[14:12:42.975751] val: [11/12] eta: 0:00:00 loss: 0.5364 (0.5038) time: 0.7586 data: 0.7252 max mem: 9672 +[14:12:43.053699] val: Total time: 0:00:09 (0.7653 s / it) +[14:12:43.068411] val loss: 0.5038238490621249 +[14:12:43.068601] Accuracy: 0.8142, F1 Score: 0.6391, ROC AUC: 0.9374, Hamming Loss: 0.0743, + Jaccard Score: 0.5215, Precision: 0.6710, Recall: 0.6214, + Average Precision: 0.6790, Kappa: 0.7214, Score: 0.7660 +[14:12:43.114019] Best epoch = 12, Best score = 0.7762 +[14:12:43.392527] log_dir: ./output_logs/retfound +[14:12:47.360051] Epoch: [17] [ 0/91] eta: 0:06:00 lr: 0.000455 loss: 0.5115 (0.5115) time: 3.9663 data: 3.8211 max mem: 9672 +[14:12:54.218973] Epoch: [17] [20/91] eta: 0:00:36 lr: 0.000445 loss: 0.6402 (0.6604) time: 0.3429 data: 0.2006 max mem: 9672 +[14:13:01.004896] Epoch: [17] [40/91] eta: 0:00:21 lr: 0.000435 loss: 0.6356 (0.6545) time: 0.3393 data: 0.1975 max mem: 9672 +[14:13:08.499322] Epoch: [17] [60/91] eta: 0:00:12 lr: 0.000425 loss: 0.6368 (0.6582) time: 0.3747 data: 0.2325 max mem: 9672 +[14:13:16.471375] Epoch: [17] [80/91] eta: 0:00:04 lr: 0.000415 loss: 0.6391 (0.6553) time: 0.3986 data: 0.2562 max mem: 9672 +[14:13:20.284781] Epoch: [17] [90/91] eta: 0:00:00 lr: 0.000410 loss: 0.6610 (0.6557) time: 0.3842 data: 0.2418 max mem: 9672 +[14:13:20.364924] Epoch: [17] Total time: 0:00:36 (0.4063 s / it) +[14:13:20.374446] Averaged stats: lr: 0.000410 loss: 0.6610 (0.6557) +[14:13:23.491912] val: [ 0/12] eta: 0:00:37 loss: 0.0044 (0.0044) time: 3.0988 data: 3.0629 max mem: 9672 +[14:13:29.512023] val: [10/12] eta: 0:00:01 loss: 0.5145 (0.4436) time: 0.8289 data: 0.7941 max mem: 9672 +[14:13:29.530153] val: [11/12] eta: 0:00:00 loss: 0.5145 (0.4973) time: 0.7613 data: 0.7279 max mem: 9672 +[14:13:29.605667] val: Total time: 0:00:09 (0.7678 s / it) +[14:13:29.621165] val loss: 0.49730656916896504 +[14:13:29.621393] Accuracy: 0.8552, F1 Score: 0.6786, ROC AUC: 0.9356, Hamming Loss: 0.0579, + Jaccard Score: 0.5696, Precision: 0.7237, Recall: 0.6638, + Average Precision: 0.6752, Kappa: 0.7816, Score: 0.7986 +[14:13:31.267421] Best epoch = 17, Best score = 0.7986 +[14:13:31.349068] log_dir: ./output_logs/retfound +[14:13:35.464633] Epoch: [18] [ 0/91] eta: 0:06:14 lr: 0.000409 loss: 0.7354 (0.7354) time: 4.1145 data: 3.9733 max mem: 9672 +[14:13:42.314643] Epoch: [18] [20/91] eta: 0:00:37 lr: 0.000399 loss: 0.5803 (0.6091) time: 0.3425 data: 0.2005 max mem: 9672 +[14:13:48.673132] Epoch: [18] [40/91] eta: 0:00:21 lr: 0.000389 loss: 0.6370 (0.6392) time: 0.3179 data: 0.1770 max mem: 9672 +[14:13:55.698321] Epoch: [18] [60/91] eta: 0:00:12 lr: 0.000378 loss: 0.6567 (0.6405) time: 0.3512 data: 0.2091 max mem: 9672 +[14:14:03.335391] Epoch: [18] [80/91] eta: 0:00:04 lr: 0.000368 loss: 0.5816 (0.6346) time: 0.3818 data: 0.2439 max mem: 9672 +[14:14:06.624289] Epoch: [18] [90/91] eta: 0:00:00 lr: 0.000362 loss: 0.5720 (0.6352) time: 0.3490 data: 0.2075 max mem: 9672 +[14:14:06.708952] Epoch: [18] Total time: 0:00:35 (0.3886 s / it) +[14:14:06.717857] Averaged stats: lr: 0.000362 loss: 0.5720 (0.6352) +[14:14:09.911136] val: [ 0/12] eta: 0:00:38 loss: 0.0010 (0.0010) time: 3.1773 data: 3.1409 max mem: 9672 +[14:14:15.944691] val: [10/12] eta: 0:00:01 loss: 0.4178 (0.4272) time: 0.8373 data: 0.8024 max mem: 9672 +[14:14:15.965349] val: [11/12] eta: 0:00:00 loss: 0.4178 (0.4770) time: 0.7692 data: 0.7355 max mem: 9672 +[14:14:16.041031] val: Total time: 0:00:09 (0.7756 s / it) +[14:14:16.056942] val loss: 0.47698313370347023 +[14:14:16.057157] Accuracy: 0.8470, F1 Score: 0.6372, ROC AUC: 0.9468, Hamming Loss: 0.0612, + Jaccard Score: 0.5440, Precision: 0.6451, Recall: 0.6387, + Average Precision: 0.6961, Kappa: 0.7677, Score: 0.7839 +[14:14:16.099152] Best epoch = 17, Best score = 0.7986 +[14:14:16.371973] log_dir: ./output_logs/retfound +[14:14:20.349567] Epoch: [19] [ 0/91] eta: 0:06:01 lr: 0.000362 loss: 0.7090 (0.7090) time: 3.9767 data: 3.8306 max mem: 9672 +[14:14:27.112151] Epoch: [19] [20/91] eta: 0:00:36 lr: 0.000351 loss: 0.5642 (0.6137) time: 0.3381 data: 0.1958 max mem: 9672 +[14:14:34.470501] Epoch: [19] [40/91] eta: 0:00:22 lr: 0.000340 loss: 0.5771 (0.6118) time: 0.3679 data: 0.2252 max mem: 9672 +[14:14:41.305389] Epoch: [19] [60/91] eta: 0:00:12 lr: 0.000330 loss: 0.6752 (0.6335) time: 0.3417 data: 0.1996 max mem: 9672 +[14:14:48.453755] Epoch: [19] [80/91] eta: 0:00:04 lr: 0.000319 loss: 0.6186 (0.6441) time: 0.3574 data: 0.2152 max mem: 9672 +[14:14:52.163869] Epoch: [19] [90/91] eta: 0:00:00 lr: 0.000314 loss: 0.6186 (0.6476) time: 0.3662 data: 0.2240 max mem: 9672 +[14:14:52.241054] Epoch: [19] Total time: 0:00:35 (0.3942 s / it) +[14:14:52.250169] Averaged stats: lr: 0.000314 loss: 0.6186 (0.6476) +[14:14:55.268791] val: [ 0/12] eta: 0:00:36 loss: 0.0034 (0.0034) time: 3.0025 data: 2.9659 max mem: 9672 +[14:15:01.302594] val: [10/12] eta: 0:00:01 loss: 0.5661 (0.4323) time: 0.8214 data: 0.7866 max mem: 9672 +[14:15:01.323200] val: [11/12] eta: 0:00:00 loss: 0.5661 (0.5148) time: 0.7546 data: 0.7211 max mem: 9672 +[14:15:01.405662] val: Total time: 0:00:09 (0.7617 s / it) +[14:15:01.420217] val loss: 0.5147952182839314 +[14:15:01.420406] Accuracy: 0.8169, F1 Score: 0.6598, ROC AUC: 0.9418, Hamming Loss: 0.0732, + Jaccard Score: 0.5385, Precision: 0.6920, Recall: 0.6457, + Average Precision: 0.6915, Kappa: 0.7280, Score: 0.7766 +[14:15:01.464755] Best epoch = 17, Best score = 0.7986 +[14:15:01.724374] log_dir: ./output_logs/retfound +[14:15:05.554063] Epoch: [20] [ 0/91] eta: 0:05:48 lr: 0.000313 loss: 0.8235 (0.8235) time: 3.8287 data: 3.6835 max mem: 9672 +[14:15:12.386786] Epoch: [20] [20/91] eta: 0:00:36 lr: 0.000302 loss: 0.6022 (0.6409) time: 0.3416 data: 0.1994 max mem: 9672 +[14:15:19.094826] Epoch: [20] [40/91] eta: 0:00:21 lr: 0.000291 loss: 0.5904 (0.6285) time: 0.3354 data: 0.1934 max mem: 9672 +[14:15:26.413293] Epoch: [20] [60/91] eta: 0:00:12 lr: 0.000281 loss: 0.5684 (0.6159) time: 0.3659 data: 0.2236 max mem: 9672 +[14:15:33.733953] Epoch: [20] [80/91] eta: 0:00:04 lr: 0.000270 loss: 0.5991 (0.6195) time: 0.3660 data: 0.2233 max mem: 9672 +[14:15:36.981662] Epoch: [20] [90/91] eta: 0:00:00 lr: 0.000265 loss: 0.6295 (0.6243) time: 0.3327 data: 0.1898 max mem: 9672 +[14:15:37.060147] Epoch: [20] Total time: 0:00:35 (0.3883 s / it) +[14:15:37.068167] Averaged stats: lr: 0.000265 loss: 0.6295 (0.6243) +[14:15:40.076104] val: [ 0/12] eta: 0:00:35 loss: 0.0045 (0.0045) time: 2.9899 data: 2.9555 max mem: 9672 +[14:15:46.312643] val: [10/12] eta: 0:00:01 loss: 0.5132 (0.4262) time: 0.8387 data: 0.8040 max mem: 9672 +[14:15:46.330649] val: [11/12] eta: 0:00:00 loss: 0.5132 (0.4892) time: 0.7703 data: 0.7370 max mem: 9672 +[14:15:46.401519] val: Total time: 0:00:09 (0.7763 s / it) +[14:15:46.416533] val loss: 0.489178951519231 +[14:15:46.416813] Accuracy: 0.8470, F1 Score: 0.6960, ROC AUC: 0.9414, Hamming Loss: 0.0612, + Jaccard Score: 0.5792, Precision: 0.7194, Recall: 0.6796, + Average Precision: 0.6895, Kappa: 0.7712, Score: 0.8029 +[14:15:48.067649] Best epoch = 20, Best score = 0.8029 +[14:15:48.146897] log_dir: ./output_logs/retfound +[14:15:52.503592] Epoch: [21] [ 0/91] eta: 0:06:36 lr: 0.000264 loss: 0.5288 (0.5288) time: 4.3559 data: 4.2124 max mem: 9672 +[14:15:59.695050] Epoch: [21] [20/91] eta: 0:00:39 lr: 0.000254 loss: 0.5579 (0.6066) time: 0.3595 data: 0.2175 max mem: 9672 +[14:16:06.805828] Epoch: [21] [40/91] eta: 0:00:23 lr: 0.000243 loss: 0.5969 (0.6263) time: 0.3555 data: 0.2132 max mem: 9672 +[14:16:14.512305] Epoch: [21] [60/91] eta: 0:00:13 lr: 0.000233 loss: 0.6827 (0.6423) time: 0.3853 data: 0.2437 max mem: 9672 +[14:16:21.485640] Epoch: [21] [80/91] eta: 0:00:04 lr: 0.000222 loss: 0.5784 (0.6264) time: 0.3486 data: 0.2060 max mem: 9672 +[14:16:25.252516] Epoch: [21] [90/91] eta: 0:00:00 lr: 0.000217 loss: 0.5645 (0.6223) time: 0.3571 data: 0.2146 max mem: 9672 +[14:16:25.329861] Epoch: [21] Total time: 0:00:37 (0.4086 s / it) +[14:16:25.338487] Averaged stats: lr: 0.000217 loss: 0.5645 (0.6223) +[14:16:28.367647] val: [ 0/12] eta: 0:00:36 loss: 0.0017 (0.0017) time: 3.0129 data: 2.9783 max mem: 9672 +[14:16:34.401750] val: [10/12] eta: 0:00:01 loss: 0.2886 (0.4467) time: 0.8224 data: 0.7878 max mem: 9672 +[14:16:34.419682] val: [11/12] eta: 0:00:00 loss: 0.2886 (0.5135) time: 0.7553 data: 0.7222 max mem: 9672 +[14:16:34.494346] val: Total time: 0:00:09 (0.7617 s / it) +[14:16:34.509079] val loss: 0.513525198524197 +[14:16:34.509242] Accuracy: 0.8224, F1 Score: 0.6099, ROC AUC: 0.9455, Hamming Loss: 0.0710, + Jaccard Score: 0.5051, Precision: 0.6864, Recall: 0.5947, + Average Precision: 0.6972, Kappa: 0.7289, Score: 0.7614 +[14:16:34.553259] Best epoch = 20, Best score = 0.8029 +[14:16:34.833940] log_dir: ./output_logs/retfound +[14:16:39.071052] Epoch: [22] [ 0/91] eta: 0:06:25 lr: 0.000217 loss: 0.7775 (0.7775) time: 4.2361 data: 4.0912 max mem: 9672 +[14:16:45.454441] Epoch: [22] [20/91] eta: 0:00:35 lr: 0.000206 loss: 0.5637 (0.6171) time: 0.3191 data: 0.1768 max mem: 9672 +[14:16:53.453321] Epoch: [22] [40/91] eta: 0:00:23 lr: 0.000196 loss: 0.6217 (0.6269) time: 0.3999 data: 0.2580 max mem: 9672 +[14:17:00.891179] Epoch: [22] [60/91] eta: 0:00:13 lr: 0.000186 loss: 0.5745 (0.6343) time: 0.3719 data: 0.2300 max mem: 9672 +[14:17:07.543566] Epoch: [22] [80/91] eta: 0:00:04 lr: 0.000177 loss: 0.6036 (0.6290) time: 0.3326 data: 0.1905 max mem: 9672 +[14:17:10.600015] Epoch: [22] [90/91] eta: 0:00:00 lr: 0.000172 loss: 0.5636 (0.6188) time: 0.3239 data: 0.1822 max mem: 9672 +[14:17:10.675509] Epoch: [22] Total time: 0:00:35 (0.3939 s / it) +[14:17:10.684530] Averaged stats: lr: 0.000172 loss: 0.5636 (0.6188) +[14:17:13.748856] val: [ 0/12] eta: 0:00:36 loss: 0.0044 (0.0044) time: 3.0449 data: 3.0088 max mem: 9672 +[14:17:19.760074] val: [10/12] eta: 0:00:01 loss: 0.3889 (0.4115) time: 0.8232 data: 0.7886 max mem: 9672 +[14:17:19.775467] val: [11/12] eta: 0:00:00 loss: 0.3889 (0.4881) time: 0.7558 data: 0.7229 max mem: 9672 +[14:17:19.877976] val: Total time: 0:00:09 (0.7645 s / it) +[14:17:19.901512] val loss: 0.4880747947221001 +[14:17:19.901722] Accuracy: 0.8361, F1 Score: 0.6394, ROC AUC: 0.9433, Hamming Loss: 0.0656, + Jaccard Score: 0.5369, Precision: 0.6737, Recall: 0.6290, + Average Precision: 0.7005, Kappa: 0.7517, Score: 0.7781 +[14:17:19.948461] Best epoch = 20, Best score = 0.8029 +[14:17:20.200530] log_dir: ./output_logs/retfound +[14:17:24.170641] Epoch: [23] [ 0/91] eta: 0:06:01 lr: 0.000171 loss: 0.6048 (0.6048) time: 3.9691 data: 3.8295 max mem: 9672 +[14:17:30.583983] Epoch: [23] [20/91] eta: 0:00:35 lr: 0.000162 loss: 0.5653 (0.5903) time: 0.3206 data: 0.1786 max mem: 9672 +[14:17:37.292263] Epoch: [23] [40/91] eta: 0:00:21 lr: 0.000153 loss: 0.7458 (0.6646) time: 0.3354 data: 0.1938 max mem: 9672 +[14:17:44.395296] Epoch: [23] [60/91] eta: 0:00:12 lr: 0.000143 loss: 0.5606 (0.6389) time: 0.3551 data: 0.2128 max mem: 9672 +[14:17:51.312102] Epoch: [23] [80/91] eta: 0:00:04 lr: 0.000134 loss: 0.5609 (0.6290) time: 0.3458 data: 0.2075 max mem: 9672 +[14:17:54.983281] Epoch: [23] [90/91] eta: 0:00:00 lr: 0.000130 loss: 0.5713 (0.6264) time: 0.3587 data: 0.2171 max mem: 9672 +[14:17:55.060411] Epoch: [23] Total time: 0:00:34 (0.3831 s / it) +[14:17:55.070009] Averaged stats: lr: 0.000130 loss: 0.5713 (0.6264) +[14:17:58.088822] val: [ 0/12] eta: 0:00:36 loss: 0.0051 (0.0051) time: 3.0019 data: 2.9667 max mem: 9672 +[14:18:04.123717] val: [10/12] eta: 0:00:01 loss: 0.5309 (0.4222) time: 0.8215 data: 0.7871 max mem: 9672 +[14:18:04.141884] val: [11/12] eta: 0:00:00 loss: 0.5309 (0.4954) time: 0.7545 data: 0.7215 max mem: 9672 +[14:18:04.223103] val: Total time: 0:00:09 (0.7614 s / it) +[14:18:04.238941] val loss: 0.4953958783298731 +[14:18:04.239131] Accuracy: 0.8279, F1 Score: 0.6796, ROC AUC: 0.9431, Hamming Loss: 0.0689, + Jaccard Score: 0.5572, Precision: 0.7031, Recall: 0.6641, + Average Precision: 0.7095, Kappa: 0.7433, Score: 0.7887 +[14:18:04.278600] Best epoch = 20, Best score = 0.8029 +[14:18:04.533936] log_dir: ./output_logs/retfound +[14:18:08.092404] Epoch: [24] [ 0/91] eta: 0:05:23 lr: 0.000130 loss: 0.7083 (0.7083) time: 3.5575 data: 3.4115 max mem: 9672 +[14:18:15.062515] Epoch: [24] [20/91] eta: 0:00:35 lr: 0.000121 loss: 0.5672 (0.5995) time: 0.3485 data: 0.2065 max mem: 9672 +[14:18:21.222396] Epoch: [24] [40/91] eta: 0:00:20 lr: 0.000113 loss: 0.5992 (0.5990) time: 0.3079 data: 0.1658 max mem: 9672 +[14:18:28.469634] Epoch: [24] [60/91] eta: 0:00:12 lr: 0.000104 loss: 0.5812 (0.5861) time: 0.3623 data: 0.2378 max mem: 9672 +[14:18:35.566079] Epoch: [24] [80/91] eta: 0:00:04 lr: 0.000097 loss: 0.6075 (0.5928) time: 0.3548 data: 0.2125 max mem: 9672 +[14:18:38.875334] Epoch: [24] [90/91] eta: 0:00:00 lr: 0.000093 loss: 0.6982 (0.6020) time: 0.3425 data: 0.2002 max mem: 9672 +[14:18:38.953081] Epoch: [24] Total time: 0:00:34 (0.3782 s / it) +[14:18:38.961882] Averaged stats: lr: 0.000093 loss: 0.6982 (0.6020) +[14:18:41.882489] val: [ 0/12] eta: 0:00:34 loss: 0.0049 (0.0049) time: 2.9048 data: 2.8708 max mem: 9672 +[14:18:47.890966] val: [10/12] eta: 0:00:01 loss: 0.5051 (0.4277) time: 0.8102 data: 0.7759 max mem: 9672 +[14:18:47.911600] val: [11/12] eta: 0:00:00 loss: 0.5051 (0.4656) time: 0.7444 data: 0.7113 max mem: 9672 +[14:18:47.989654] val: Total time: 0:00:09 (0.7510 s / it) +[14:18:48.006839] val loss: 0.4656097547461589 +[14:18:48.007038] Accuracy: 0.8333, F1 Score: 0.6564, ROC AUC: 0.9453, Hamming Loss: 0.0667, + Jaccard Score: 0.5457, Precision: 0.6726, Recall: 0.6475, + Average Precision: 0.7146, Kappa: 0.7505, Score: 0.7841 +[14:18:48.052530] Best epoch = 20, Best score = 0.8029 +[14:18:48.309510] log_dir: ./output_logs/retfound +[14:18:52.436968] Epoch: [25] [ 0/91] eta: 0:06:15 lr: 0.000092 loss: 0.7129 (0.7129) time: 4.1266 data: 3.9807 max mem: 9672 +[14:18:58.895641] Epoch: [25] [20/91] eta: 0:00:35 lr: 0.000085 loss: 0.5180 (0.5764) time: 0.3229 data: 0.1809 max mem: 9672 +[14:19:06.694136] Epoch: [25] [40/91] eta: 0:00:22 lr: 0.000078 loss: 0.5699 (0.5936) time: 0.3899 data: 0.2477 max mem: 9672 +[14:19:13.404145] Epoch: [25] [60/91] eta: 0:00:12 lr: 0.000071 loss: 0.5863 (0.5908) time: 0.3355 data: 0.1934 max mem: 9672 +[14:19:21.067096] Epoch: [25] [80/91] eta: 0:00:04 lr: 0.000064 loss: 0.5094 (0.5873) time: 0.3831 data: 0.2410 max mem: 9672 +[14:19:24.920383] Epoch: [25] [90/91] eta: 0:00:00 lr: 0.000061 loss: 0.5148 (0.5804) time: 0.3559 data: 0.2139 max mem: 9672 +[14:19:25.005082] Epoch: [25] Total time: 0:00:36 (0.4032 s / it) +[14:19:25.013360] Averaged stats: lr: 0.000061 loss: 0.5148 (0.5804) +[14:19:28.157577] val: [ 0/12] eta: 0:00:37 loss: 0.0044 (0.0044) time: 3.1262 data: 3.0899 max mem: 9672 +[14:19:34.201611] val: [10/12] eta: 0:00:01 loss: 0.4970 (0.4166) time: 0.8336 data: 0.7993 max mem: 9672 +[14:19:34.222376] val: [11/12] eta: 0:00:00 loss: 0.4970 (0.4665) time: 0.7658 data: 0.7327 max mem: 9672 +[14:19:34.296398] val: Total time: 0:00:09 (0.7721 s / it) +[14:19:34.311031] val loss: 0.4665253447989623 +[14:19:34.311211] Accuracy: 0.8333, F1 Score: 0.6562, ROC AUC: 0.9461, Hamming Loss: 0.0667, + Jaccard Score: 0.5461, Precision: 0.6710, Recall: 0.6475, + Average Precision: 0.7131, Kappa: 0.7508, Score: 0.7844 +[14:19:34.352968] Best epoch = 20, Best score = 0.8029 +[14:19:34.615941] log_dir: ./output_logs/retfound +[14:19:38.075073] Epoch: [26] [ 0/91] eta: 0:05:14 lr: 0.000061 loss: 0.4910 (0.4910) time: 3.4581 data: 3.3184 max mem: 9672 +[14:19:45.236782] Epoch: [26] [20/91] eta: 0:00:35 lr: 0.000054 loss: 0.5952 (0.6149) time: 0.3580 data: 0.2321 max mem: 9672 +[14:19:53.194620] Epoch: [26] [40/91] eta: 0:00:23 lr: 0.000049 loss: 0.5282 (0.5882) time: 0.3978 data: 0.2566 max mem: 9672 +[14:19:59.834325] Epoch: [26] [60/91] eta: 0:00:12 lr: 0.000043 loss: 0.6217 (0.5954) time: 0.3319 data: 0.1896 max mem: 9672 +[14:20:06.541321] Epoch: [26] [80/91] eta: 0:00:04 lr: 0.000038 loss: 0.5417 (0.6003) time: 0.3353 data: 0.1934 max mem: 9672 +[14:20:09.471194] Epoch: [26] [90/91] eta: 0:00:00 lr: 0.000035 loss: 0.5417 (0.5938) time: 0.2900 data: 0.1475 max mem: 9672 +[14:20:09.552606] Epoch: [26] Total time: 0:00:34 (0.3839 s / it) +[14:20:09.561300] Averaged stats: lr: 0.000035 loss: 0.5417 (0.5938) +[14:20:12.670793] val: [ 0/12] eta: 0:00:37 loss: 0.0040 (0.0040) time: 3.0935 data: 3.0573 max mem: 9672 +[14:20:18.677898] val: [10/12] eta: 0:00:01 loss: 0.5400 (0.4176) time: 0.8273 data: 0.7942 max mem: 9672 +[14:20:18.687566] val: [11/12] eta: 0:00:00 loss: 0.5400 (0.4698) time: 0.7591 data: 0.7281 max mem: 9672 +[14:20:18.763589] val: Total time: 0:00:09 (0.7656 s / it) +[14:20:18.786236] val loss: 0.4697543305034439 +[14:20:18.786436] Accuracy: 0.8388, F1 Score: 0.6617, ROC AUC: 0.9456, Hamming Loss: 0.0645, + Jaccard Score: 0.5539, Precision: 0.6739, Recall: 0.6537, + Average Precision: 0.7140, Kappa: 0.7590, Score: 0.7888 +[14:20:18.827092] Best epoch = 20, Best score = 0.8029 +[14:20:19.079213] log_dir: ./output_logs/retfound +[14:20:23.603594] Epoch: [27] [ 0/91] eta: 0:06:51 lr: 0.000035 loss: 0.8099 (0.8099) time: 4.5234 data: 4.3801 max mem: 9672 +[14:20:29.957620] Epoch: [27] [20/91] eta: 0:00:36 lr: 0.000030 loss: 0.5915 (0.6246) time: 0.3177 data: 0.1755 max mem: 9672 +[14:20:36.966395] Epoch: [27] [40/91] eta: 0:00:22 lr: 0.000026 loss: 0.5851 (0.6130) time: 0.3504 data: 0.2086 max mem: 9672 +[14:20:44.265775] Epoch: [27] [60/91] eta: 0:00:12 lr: 0.000022 loss: 0.6211 (0.6170) time: 0.3649 data: 0.2227 max mem: 9672 +[14:20:51.667785] Epoch: [27] [80/91] eta: 0:00:04 lr: 0.000018 loss: 0.5192 (0.5952) time: 0.3701 data: 0.2277 max mem: 9672 +[14:20:55.349553] Epoch: [27] [90/91] eta: 0:00:00 lr: 0.000016 loss: 0.5192 (0.5917) time: 0.3651 data: 0.2226 max mem: 9672 +[14:20:55.433182] Epoch: [27] Total time: 0:00:36 (0.3995 s / it) +[14:20:55.441928] Averaged stats: lr: 0.000016 loss: 0.5192 (0.5917) +[14:20:58.438173] val: [ 0/12] eta: 0:00:35 loss: 0.0047 (0.0047) time: 2.9791 data: 2.9436 max mem: 9672 +[14:21:04.470482] val: [10/12] eta: 0:00:01 loss: 0.5242 (0.4170) time: 0.8192 data: 0.7845 max mem: 9672 +[14:21:04.488553] val: [11/12] eta: 0:00:00 loss: 0.5242 (0.4706) time: 0.7524 data: 0.7191 max mem: 9672 +[14:21:04.561582] val: Total time: 0:00:09 (0.7586 s / it) +[14:21:04.576385] val loss: 0.470587865759929 +[14:21:04.576654] Accuracy: 0.8415, F1 Score: 0.6734, ROC AUC: 0.9452, Hamming Loss: 0.0634, + Jaccard Score: 0.5622, Precision: 0.6885, Recall: 0.6628, + Average Precision: 0.7145, Kappa: 0.7632, Score: 0.7939 +[14:21:04.622827] Best epoch = 20, Best score = 0.8029 +[14:21:04.888658] log_dir: ./output_logs/retfound +[14:21:09.030525] Epoch: [28] [ 0/91] eta: 0:06:16 lr: 0.000016 loss: 0.4917 (0.4917) time: 4.1408 data: 3.9947 max mem: 9672 +[14:21:16.870804] Epoch: [28] [20/91] eta: 0:00:40 lr: 0.000013 loss: 0.4976 (0.5350) time: 0.3920 data: 0.2500 max mem: 9672 +[14:21:23.818628] Epoch: [28] [40/91] eta: 0:00:23 lr: 0.000010 loss: 0.6066 (0.5824) time: 0.3473 data: 0.2049 max mem: 9672 +[14:21:30.267555] Epoch: [28] [60/91] eta: 0:00:12 lr: 0.000008 loss: 0.5936 (0.5825) time: 0.3224 data: 0.1806 max mem: 9672 +[14:21:37.638843] Epoch: [28] [80/91] eta: 0:00:04 lr: 0.000006 loss: 0.6226 (0.5881) time: 0.3685 data: 0.2445 max mem: 9672 +[14:21:40.690586] Epoch: [28] [90/91] eta: 0:00:00 lr: 0.000005 loss: 0.6543 (0.5873) time: 0.3078 data: 0.1834 max mem: 9672 +[14:21:40.772100] Epoch: [28] Total time: 0:00:35 (0.3943 s / it) +[14:21:40.781199] Averaged stats: lr: 0.000005 loss: 0.6543 (0.5873) +[14:21:43.864888] val: [ 0/12] eta: 0:00:36 loss: 0.0056 (0.0056) time: 3.0675 data: 3.0312 max mem: 9672 +[14:21:49.882433] val: [10/12] eta: 0:00:01 loss: 0.5236 (0.4185) time: 0.8259 data: 0.7919 max mem: 9672 +[14:21:49.903288] val: [11/12] eta: 0:00:00 loss: 0.5236 (0.4700) time: 0.7587 data: 0.7260 max mem: 9672 +[14:21:49.977340] val: Total time: 0:00:09 (0.7650 s / it) +[14:21:49.991955] val loss: 0.47001992507527274 +[14:21:49.992142] Accuracy: 0.8361, F1 Score: 0.6603, ROC AUC: 0.9450, Hamming Loss: 0.0656, + Jaccard Score: 0.5518, Precision: 0.6724, Recall: 0.6525, + Average Precision: 0.7128, Kappa: 0.7552, Score: 0.7868 +[14:21:50.034343] Best epoch = 20, Best score = 0.8029 +[14:21:50.318733] log_dir: ./output_logs/retfound +[14:21:53.882600] Epoch: [29] [ 0/91] eta: 0:05:24 lr: 0.000005 loss: 0.4039 (0.4039) time: 3.5629 data: 3.4185 max mem: 9672 +[14:22:02.374725] Epoch: [29] [20/91] eta: 0:00:40 lr: 0.000003 loss: 0.5815 (0.5432) time: 0.4246 data: 0.2821 max mem: 9672 +[14:22:10.424813] Epoch: [29] [40/91] eta: 0:00:25 lr: 0.000002 loss: 0.6099 (0.5764) time: 0.4024 data: 0.2604 max mem: 9672 +[14:22:16.550895] Epoch: [29] [60/91] eta: 0:00:13 lr: 0.000001 loss: 0.6322 (0.5857) time: 0.3063 data: 0.1827 max mem: 9672 +[14:22:23.107720] Epoch: [29] [80/91] eta: 0:00:04 lr: 0.000001 loss: 0.5222 (0.5732) time: 0.3278 data: 0.1858 max mem: 9672 +[14:22:26.515748] Epoch: [29] [90/91] eta: 0:00:00 lr: 0.000001 loss: 0.5222 (0.5673) time: 0.3499 data: 0.2079 max mem: 9672 +[14:22:26.590835] Epoch: [29] Total time: 0:00:36 (0.3986 s / it) +[14:22:26.592716] Averaged stats: lr: 0.000001 loss: 0.5222 (0.5673) +[14:22:29.442123] val: [ 0/12] eta: 0:00:33 loss: 0.0057 (0.0057) time: 2.8321 data: 2.7971 max mem: 9672 +[14:22:35.427356] val: [10/12] eta: 0:00:01 loss: 0.5069 (0.4176) time: 0.8015 data: 0.7670 max mem: 9672 +[14:22:35.448071] val: [11/12] eta: 0:00:00 loss: 0.5069 (0.4689) time: 0.7364 data: 0.7031 max mem: 9672 +[14:22:35.528270] val: Total time: 0:00:08 (0.7432 s / it) +[14:22:35.542913] val loss: 0.468911811709404 +[14:22:35.543112] Accuracy: 0.8361, F1 Score: 0.6603, ROC AUC: 0.9453, Hamming Loss: 0.0656, + Jaccard Score: 0.5518, Precision: 0.6724, Recall: 0.6525, + Average Precision: 0.7131, Kappa: 0.7552, Score: 0.7869 +[14:22:35.585236] Best epoch = 20, Best score = 0.8029 +[14:22:39.189935] Test with the best model, epoch = 20: +[14:22:42.338297] test: [ 0/12] eta: 0:00:37 loss: 0.0019 (0.0019) time: 3.1334 data: 3.0990 max mem: 9672 +[14:22:47.870333] test: [10/12] eta: 0:00:01 loss: 0.1041 (0.3849) time: 0.7877 data: 0.7554 max mem: 9672 +[14:22:47.880361] test: [11/12] eta: 0:00:00 loss: 0.1041 (0.4738) time: 0.7229 data: 0.6925 max mem: 9672 +[14:22:47.956536] test: Total time: 0:00:08 (0.7294 s / it) +[14:22:47.972502] val loss: 0.4738252603759368 +[14:22:47.972657] Accuracy: 0.8361, F1 Score: 0.6495, ROC AUC: 0.9478, Hamming Loss: 0.0656, + Jaccard Score: 0.5248, Precision: 0.6955, Recall: 0.6323, + Average Precision: 0.6934, Kappa: 0.7384, Score: 0.7786 +[14:22:48.703197] Training time 0:23:20 +[rank0]:[W615 14:22:49.089502189 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/aptos/retfound acc=0.8361 auroc_macro_ovr=0.9477939000875149 f1_macro=0.6495 qwk=0.9055872387300082 diff --git a/results/aptos/vit/confusion_matrix.png b/results/aptos/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..a26d29138c35dbbe2b562e34abe1c1d742a7451e --- /dev/null +++ b/results/aptos/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dcc8115b99efe6e583683f19cd3b3dd15156ebdbc61c45775209e7380d5c95c1 +size 95292 diff --git a/results/aptos/vit/log.csv b/results/aptos/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..b1beff72fc37af906284ba6e8d481618c7f6f268 --- /dev/null +++ b/results/aptos/vit/log.csv @@ -0,0 +1,18 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.9414930529064602,0.366120218579235,0.691584930218475,0.3932175381512096,7.230091520214377e-08 +1,1.5824870374467639,0.5355191256830601,0.8443007456208633,0.5403723764378828,1.4624503302251806e-07 +2,1.4336420350604586,0.5546448087431693,0.8784619774530464,0.5551728464859785,2.2018915084289233e-07 +3,1.3304660346772936,0.7486338797814208,0.8698132272686244,0.6982646912065512,2.9413326866326666e-07 +4,1.2724972354041206,0.6311475409836066,0.8620025072490781,0.6186938169505419,3.6807738648364093e-07 +5,1.228438933690389,0.7185792349726776,0.8730791497210626,0.688594392164727,3.68326895715127e-07 +6,1.1796417395273844,0.773224043715847,0.9033635574849272,0.7417797248104895,3.640405452192437e-07 +7,1.1395060976346334,0.7404371584699454,0.9007815984200768,0.7170117123215412,3.569284151332278e-07 +8,1.0889184501436022,0.7377049180327869,0.9070575400697996,0.7150473619816492,3.4710266799711624e-07 +9,1.0612807499037848,0.773224043715847,0.8917355261723398,0.7387249897157826,3.3471826171286236e-07 +10,1.0361893720097013,0.76775956284153,0.8833464442622343,0.7315647984199867,3.199705057656546e-07 +11,1.019848510954115,0.7431693989071039,0.8799943840863638,0.7063742593570885,3.030919810766626e-07 +12,0.9996129274368286,0.7540983606557377,0.8828037034925351,0.7156845473990883,2.8434887206297395e-07 +13,0.9594410591655307,0.7622950819672131,0.8905838601784287,0.7229269663828678,2.6403676875032503e-07 +14,0.9264983508321974,0.7459016393442623,0.8955212078072705,0.7141067506689697,2.4247600514181857e-07 +15,0.9140548547108968,0.7540983606557377,0.8776981790388578,0.7175668592141894,2.200066073593367e-07 +16,0.8953906761275398,0.7704918032786885,0.8819356874951536,0.7300404688909238,1.9698293122847687e-07 diff --git a/results/aptos/vit/metrics.json b/results/aptos/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..c372935ddf94caf962b8c5a695eeb379ecfb8dc1 --- /dev/null +++ b/results/aptos/vit/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 366, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.7923497267759563, + "balanced_accuracy": 0.6169725386413288, + "precision_macro": 0.6192561508157517, + "recall_macro": 0.6169725386413288, + "f1_macro": 0.604525638141757, + "precision_weighted": 0.8190554374982756, + "recall_weighted": 0.7923497267759563, + "f1_weighted": 0.8002778101173526, + "cohen_kappa": 0.6750770955985422, + "quadratic_weighted_kappa": 0.87478506591312, + "mcc": 0.677229279266987, + "auroc_macro_ovr": 0.9164958759762334, + "auroc_weighted_ovr": 0.952627313127194, + "auprc_macro": 0.6056136099262176, + "auroc_per_class": { + "0": 0.9952757800980953, + "1": 0.8774801587301587, + "2": 0.9144934701108228, + "3": 0.9078880836002022, + "4": 0.8873418873418875 + }, + "per_class": { + "0": { + "precision": 0.9847715736040609, + "recall": 0.9748743718592965, + "f1-score": 0.9797979797979798, + "support": 199.0 + }, + "1": { + "precision": 0.42857142857142855, + "recall": 0.6, + "f1-score": 0.5, + "support": 30.0 + }, + "2": { + "precision": 0.7272727272727273, + "recall": 0.6436781609195402, + "f1-score": 0.6829268292682927, + "support": 87.0 + }, + "3": { + "precision": 0.2413793103448276, + "recall": 0.4117647058823529, + "f1-score": 0.30434782608695654, + "support": 17.0 + }, + "4": { + "precision": 0.7142857142857143, + "recall": 0.45454545454545453, + "f1-score": 0.5555555555555556, + "support": 33.0 + }, + "accuracy": 0.7923497267759563, + "macro avg": { + "precision": 0.6192561508157517, + "recall": 0.6169725386413288, + "f1-score": 0.604525638141757, + "support": 366.0 + }, + "weighted avg": { + "precision": 0.8190554374982756, + "recall": 0.7923497267759563, + "f1-score": 0.8002778101173526, + "support": 366.0 + } + } +} \ No newline at end of file diff --git a/results/aptos/vit/pr.png b/results/aptos/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..cbe2fc755d7aa6ef1560d9cb4de6593d211a7f66 --- /dev/null +++ b/results/aptos/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32ed707c09741412693a35c632da84e63817ea78d9c659129ed6f63459f1f105 +size 110427 diff --git a/results/aptos/vit/roc.png b/results/aptos/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..7e981e0cba326a3e08ecbb5dfa7c19aad6e8d203 --- /dev/null +++ b/results/aptos/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd708f105d66c8095d2a2e265b0f800dcce735ef09761fa9fe90ffc7de76bd2f +size 85356 diff --git a/results/aptos/vit/test_pred.npz b/results/aptos/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..c73164af5f080f2c58f1ec4a26f39795b694e922 --- /dev/null +++ b/results/aptos/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6395f7f9d79608634e423a04d239093b153c13fe5e58f61855c39ebbaa84fa36 +size 10758 diff --git a/results/aptos/vit/train.log b/results/aptos/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f28b3c340af292ee74b98ef461365aff9e36e193 --- /dev/null +++ b/results/aptos/vit/train.log @@ -0,0 +1,94 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=2930 val=366 test=366 classes=['0', '1', '2', '3', '4'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=1.9415 val_acc=0.3661 val_auc=0.6916 score=0.3932 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=1.5825 val_acc=0.5355 val_auc=0.8443 score=0.5404 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=1.4336 val_acc=0.5546 val_auc=0.8785 score=0.5552 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.3305 val_acc=0.7486 val_auc=0.8698 score=0.6983 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=1.2725 val_acc=0.6311 val_auc=0.8620 score=0.6187 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=1.2284 val_acc=0.7186 val_auc=0.8731 score=0.6886 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=1.1796 val_acc=0.7732 val_auc=0.9034 score=0.7418 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=1.1395 val_acc=0.7404 val_auc=0.9008 score=0.7170 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=1.0889 val_acc=0.7377 val_auc=0.9071 score=0.7150 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=1.0613 val_acc=0.7732 val_auc=0.8917 score=0.7387 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=1.0362 val_acc=0.7678 val_auc=0.8833 score=0.7316 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=1.0198 val_acc=0.7432 val_auc=0.8800 score=0.7064 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.9996 val_acc=0.7541 val_auc=0.8828 score=0.7157 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.9594 val_acc=0.7623 val_auc=0.8906 score=0.7229 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.9265 val_acc=0.7459 val_auc=0.8955 score=0.7141 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.9141 val_acc=0.7541 val_auc=0.8777 score=0.7176 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.8954 val_acc=0.7705 val_auc=0.8819 score=0.7300 +[vit] early stop at ep16 (best ep6 score=0.7418) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=6 best_val_score=0.7418 -> saved test_pred.npz (366 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/aptos/vit acc=0.7923 auroc_macro_ovr=0.9164958759762334 f1_macro=0.6045 qwk=0.87478506591312 diff --git a/results/deepdrid/resnet/confusion_matrix.png b/results/deepdrid/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..81060813f50fc5ba9d9e0e8b301f5fe4eb49fd05 --- /dev/null +++ b/results/deepdrid/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9dfdb40d8b78453204f19a87f0c0e0f1aecf3c3306492e382937c03555087dd +size 103722 diff --git a/results/deepdrid/resnet/log.csv b/results/deepdrid/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..f6fbff66980d33cd2e393eb32daf73ea2cdafa43 --- /dev/null +++ b/results/deepdrid/resnet/log.csv @@ -0,0 +1,26 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.6088274916013081,0.3175,0.577595389983238,0.29653961804088197,0.0001574074074074074 +1,1.587100128332774,0.5375,0.750095647608512,0.5012446326740919,0.00032407407407407406 +2,1.5047050846947565,0.5325,0.7991009510230981,0.514190578660458,0.0004907407407407408 +3,1.2848913073539734,0.5225,0.810219842781881,0.5166209165414865,0.0004995020076872309 +4,1.0490628017319574,0.53,0.8373545101455075,0.5481707700227358,0.0004978914025677908 +5,0.883466018570794,0.5775,0.856319971632213,0.5924268519826278,0.0004951736542679293 +6,0.7467797365453508,0.525,0.8319800920396998,0.5621002576747337,0.0004913609009112972 +7,0.6041025502814187,0.575,0.8375832468253416,0.585507909761232,0.00048647017117994497 +8,0.4941401481628418,0.57,0.8615875337982306,0.5921050886414511,0.00048052330826009557 +9,0.41659408973322976,0.6025,0.8516482747941172,0.6127488894719882,0.000473546872285172 +10,0.32821205755074817,0.61,0.8636702251689459,0.6260974785922264,0.0004655720217117939 +11,0.29310083306497997,0.6325,0.8623173208899967,0.6292558107795595,0.00045663437415854646 +12,0.2857719072037273,0.6,0.8395493764168647,0.6054825887714085,0.00044677384732904935 +13,0.22614803496334288,0.605,0.8480078835931082,0.6033334664380684,0.00043603448072980245 +14,0.19276066083047125,0.645,0.8554379196473427,0.6418155458895053,0.00042446423897905887 +15,0.15560390137963825,0.5775,0.8193927319393535,0.5866991230385118,0.00041211479758519645 +16,0.11282555013895035,0.5925,0.8439370434178096,0.5897997723405428,0.0003990413121513526 +17,0.1249556862231758,0.6175,0.837508530211997,0.6078336535932065,0.0003853021720371118 +18,0.09002056303951475,0.64,0.8387059266855286,0.6262253936085712,0.00037095873957745116 +19,0.07843600120395422,0.645,0.8464061989903628,0.6383237532068925,0.00035607507602365685 +20,0.05829336690819926,0.63,0.8435061468991398,0.6278831850337193,0.00034071765543022435 +21,0.08001760921130578,0.6425,0.8302532910983373,0.6265969236856778,0.0003249550677655946 +22,0.07459283211371964,0.6225,0.8380082353645519,0.613871590160307,0.0003088577125727041 +23,0.047830414958298206,0.635,0.8480499660454267,0.6278871395507433,0.00029249748454753677 +24,0.05430900098548995,0.6325,0.8461306407679642,0.6298586631390614,0.00027594745243995884 diff --git a/results/deepdrid/resnet/metrics.json b/results/deepdrid/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..da1bb81913f81ebe10cf69e5260ae2aa405362c9 --- /dev/null +++ b/results/deepdrid/resnet/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 400, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.7025, + "balanced_accuracy": 0.5311111111111111, + "precision_macro": 0.5570393205296086, + "recall_macro": 0.5311111111111111, + "f1_macro": 0.5341998001249733, + "precision_weighted": 0.7027937218099811, + "recall_weighted": 0.7025, + "f1_weighted": 0.6991151542461292, + "cohen_kappa": 0.5604639137179582, + "quadratic_weighted_kappa": 0.8270492156215463, + "mcc": 0.5616415990289286, + "auroc_macro_ovr": 0.8759508947086289, + "auroc_weighted_ovr": 0.909941973543145, + "auprc_macro": 0.5809977088302861, + "auroc_per_class": { + "0": 0.9382499999999999, + "1": 0.8013583638583639, + "2": 0.880589430894309, + "3": 0.9400829945799458, + "4": 0.8194736842105264 + }, + "per_class": { + "0": { + "precision": 0.8854166666666666, + "recall": 0.85, + "f1-score": 0.8673469387755102, + "support": 200.0 + }, + "1": { + "precision": 0.1875, + "recall": 0.16666666666666666, + "f1-score": 0.17647058823529413, + "support": 36.0 + }, + "2": { + "precision": 0.5287356321839081, + "recall": 0.6388888888888888, + "f1-score": 0.5786163522012578, + "support": 72.0 + }, + "3": { + "precision": 0.6835443037974683, + "recall": 0.75, + "f1-score": 0.7152317880794702, + "support": 72.0 + }, + "4": { + "precision": 0.5, + "recall": 0.25, + "f1-score": 0.3333333333333333, + "support": 20.0 + }, + "accuracy": 0.7025, + "macro avg": { + "precision": 0.5570393205296086, + "recall": 0.5311111111111111, + "f1-score": 0.5341998001249733, + "support": 400.0 + }, + "weighted avg": { + "precision": 0.7027937218099811, + "recall": 0.7025, + "f1-score": 0.6991151542461292, + "support": 400.0 + } + } +} \ No newline at end of file diff --git a/results/deepdrid/resnet/pr.png b/results/deepdrid/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..982b31da58dc74e051ce7376216d24af738eff1b --- /dev/null +++ b/results/deepdrid/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a69365b413b7fe2a7fb15a61747ae676a810267a4dce9b8693642c64a557e88 +size 105725 diff --git a/results/deepdrid/resnet/roc.png b/results/deepdrid/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..09f00a1b6da7f094f4151e0c7591ad2a62eb61ab --- /dev/null +++ b/results/deepdrid/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35a1a054780335c6f03936a25db3c8c50dd5652bd8cb5fc8ad379f1f75075d29 +size 87405 diff --git a/results/deepdrid/resnet/test_pred.npz b/results/deepdrid/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..69b51d498a53b02d56afa189bab553e499962b3c --- /dev/null +++ b/results/deepdrid/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:83c4a9536c2a0245434b403a91ca75b7215512d3f8475ebb1b3ff0a9ab1c4e1b +size 11710 diff --git a/results/deepdrid/resnet/train.log b/results/deepdrid/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1b180253a6352391822648dee23630fca009ea09 --- /dev/null +++ b/results/deepdrid/resnet/train.log @@ -0,0 +1,133 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=1200 val=400 test=400 classes=['0', '1', '2', '3', '4'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=1.6088 val_acc=0.3175 val_auc=0.5776 score=0.2965 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=1.5871 val_acc=0.5375 val_auc=0.7501 score=0.5012 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=1.5047 val_acc=0.5325 val_auc=0.7991 score=0.5142 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=1.2849 val_acc=0.5225 val_auc=0.8102 score=0.5166 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=1.0491 val_acc=0.5300 val_auc=0.8374 score=0.5482 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.8835 val_acc=0.5775 val_auc=0.8563 score=0.5924 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.7468 val_acc=0.5250 val_auc=0.8320 score=0.5621 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6041 val_acc=0.5750 val_auc=0.8376 score=0.5855 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.4941 val_acc=0.5700 val_auc=0.8616 score=0.5921 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.4166 val_acc=0.6025 val_auc=0.8516 score=0.6127 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.3282 val_acc=0.6100 val_auc=0.8637 score=0.6261 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.2931 val_acc=0.6325 val_auc=0.8623 score=0.6293 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.2858 val_acc=0.6000 val_auc=0.8395 score=0.6055 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.2261 val_acc=0.6050 val_auc=0.8480 score=0.6033 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.1928 val_acc=0.6450 val_auc=0.8554 score=0.6418 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.1556 val_acc=0.5775 val_auc=0.8194 score=0.5867 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.1128 val_acc=0.5925 val_auc=0.8439 score=0.5898 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.1250 val_acc=0.6175 val_auc=0.8375 score=0.6078 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0900 val_acc=0.6400 val_auc=0.8387 score=0.6262 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0784 val_acc=0.6450 val_auc=0.8464 score=0.6383 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0583 val_acc=0.6300 val_auc=0.8435 score=0.6279 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0800 val_acc=0.6425 val_auc=0.8303 score=0.6266 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0746 val_acc=0.6225 val_auc=0.8380 score=0.6139 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0478 val_acc=0.6350 val_auc=0.8480 score=0.6279 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0543 val_acc=0.6325 val_auc=0.8461 score=0.6299 +[resnet] early stop at ep24 (best ep14 score=0.6418) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=14 best_val_score=0.6418 -> saved test_pred.npz (400 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/deepdrid/resnet acc=0.7025 auroc_macro_ovr=0.8759508947086289 f1_macro=0.5342 qwk=0.8270492156215463 diff --git a/results/deepdrid/retfound/confusion_matrix.png b/results/deepdrid/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..50fc07fa424e55816547c0fe7c5390da58e24242 --- /dev/null +++ b/results/deepdrid/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e1a68187c0d7d8d989daae1881c0c5a6bf2af2e666624429d95da865f6018c49 +size 102444 diff --git a/results/deepdrid/retfound/confusion_matrix_test.jpg b/results/deepdrid/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b835b8a474dfdede9fda5cb3d41e1d3d503d2e3e --- /dev/null +++ b/results/deepdrid/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c42dd77b3cb5e80dfbd2fa3501a506b9c56d92ba860dd40f1313b9ffd9029d7 +size 342361 diff --git a/results/deepdrid/retfound/log.txt b/results/deepdrid/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..79f18058e42c04dabef72ec7056917c024e11e57 --- /dev/null +++ b/results/deepdrid/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 3.040540540540541e-05, "train_loss": 1.5490099932696368, "epoch": 0, "n_parameters": 303306757} +{"train_lr": 9.290540540540539e-05, "train_loss": 1.393124142208615, "epoch": 1, "n_parameters": 303306757} +{"train_lr": 0.00015540540540540543, "train_loss": 1.2970332583865605, "epoch": 2, "n_parameters": 303306757} +{"train_lr": 0.0002179054054054054, "train_loss": 1.2249928551751215, "epoch": 3, "n_parameters": 303306757} +{"train_lr": 0.0002804054054054054, "train_loss": 1.1643589638374947, "epoch": 4, "n_parameters": 303306757} +{"train_lr": 0.0003429054054054054, "train_loss": 1.134743020341203, "epoch": 5, "n_parameters": 303306757} +{"train_lr": 0.00040540540540540544, "train_loss": 1.1119098276705355, "epoch": 6, "n_parameters": 303306757} +{"train_lr": 0.00046790540540540533, "train_loss": 1.115264841028162, "epoch": 7, "n_parameters": 303306757} +{"train_lr": 0.0005304054054054054, "train_loss": 1.0972523238207843, "epoch": 8, "n_parameters": 303306757} +{"train_lr": 0.0005929054054054054, "train_loss": 1.057661501136986, "epoch": 9, "n_parameters": 303306757} +{"train_lr": 0.0006246922169294559, "train_loss": 1.041992509687269, "epoch": 10, "n_parameters": 303306757} +{"train_lr": 0.0006227965251389128, "train_loss": 1.021438946595063, "epoch": 11, "n_parameters": 303306757} +{"train_lr": 0.0006189908347530905, "train_loss": 1.024420132508149, "epoch": 12, "n_parameters": 303306757} +{"train_lr": 0.000613298609118737, "train_loss": 1.016700245238639, "epoch": 13, "n_parameters": 303306757} +{"train_lr": 0.0006057549426997458, "train_loss": 0.9689739781457025, "epoch": 14, "n_parameters": 303306757} +{"train_lr": 0.0005964063447081134, "train_loss": 0.957852840423584, "epoch": 15, "n_parameters": 303306757} +{"train_lr": 0.0005853104523591827, "train_loss": 0.9439821646020219, "epoch": 16, "n_parameters": 303306757} +{"train_lr": 0.0005725356755190434, "train_loss": 0.9653627292529957, "epoch": 17, "n_parameters": 303306757} +{"train_lr": 0.000558160774934959, "train_loss": 0.9284957357355066, "epoch": 18, "n_parameters": 303306757} +{"train_lr": 0.0005422743766491704, "train_loss": 0.9342711245691454, "epoch": 19, "n_parameters": 303306757} +{"train_lr": 0.000524974425589873, "train_loss": 0.9503510384946257, "epoch": 20, "n_parameters": 303306757} +{"train_lr": 0.0005063675817081646, "train_loss": 0.9013161465928361, "epoch": 21, "n_parameters": 303306757} +{"train_lr": 0.0004865685623839792, "train_loss": 0.9048521695910273, "epoch": 22, "n_parameters": 303306757} +{"train_lr": 0.00046569943515529296, "train_loss": 0.8896518559069246, "epoch": 23, "n_parameters": 303306757} +{"train_lr": 0.0004438888651311591, "train_loss": 0.9006354873244827, "epoch": 24, "n_parameters": 303306757} +{"train_lr": 0.00042127132172852736, "train_loss": 0.8943660742527729, "epoch": 25, "n_parameters": 303306757} +{"train_lr": 0.00039798624962357145, "train_loss": 0.8588034494503124, "epoch": 26, "n_parameters": 303306757} +{"train_lr": 0.0003741772090288904, "train_loss": 0.8647741095439808, "epoch": 27, "n_parameters": 303306757} +{"train_lr": 0.0003499909905970598, "train_loss": 0.8946552389376873, "epoch": 28, "n_parameters": 303306757} +{"train_lr": 0.0003255767104074456, "train_loss": 0.8423824809693001, "epoch": 29, "n_parameters": 303306757} +{"train_lr": 0.0003010848906159891, "train_loss": 0.8773198401605761, "epoch": 30, "n_parameters": 303306757} +{"train_lr": 0.00027666653143606097, "train_loss": 0.800272930312801, "epoch": 31, "n_parameters": 303306757} +{"train_lr": 0.00025247218017193375, "train_loss": 0.82258244624009, "epoch": 32, "n_parameters": 303306757} +{"train_lr": 0.00022865100304458917, "train_loss": 0.7866110656712506, "epoch": 33, "n_parameters": 303306757} +{"train_lr": 0.00020534986553236243, "train_loss": 0.8328869938850403, "epoch": 34, "n_parameters": 303306757} +{"train_lr": 0.0001827124268964309, "train_loss": 0.7969237533775536, "epoch": 35, "n_parameters": 303306757} +{"train_lr": 0.00016087825447369744, "train_loss": 0.8122462259756552, "epoch": 36, "n_parameters": 303306757} +{"train_lr": 0.00013998196319774713, "train_loss": 0.8002816342018746, "epoch": 37, "n_parameters": 303306757} +{"train_lr": 0.00012015238565301837, "train_loss": 0.7887475925522882, "epoch": 38, "n_parameters": 303306757} +{"train_lr": 0.00010151177777907797, "train_loss": 0.7995240720542701, "epoch": 39, "n_parameters": 303306757} +{"train_lr": 8.41750651220999e-05, "train_loss": 0.7871757036930805, "epoch": 40, "n_parameters": 303306757} +{"train_lr": 6.82491342806564e-05, "train_loss": 0.7756367100251688, "epoch": 41, "n_parameters": 303306757} +{"train_lr": 5.38321739142944e-05, "train_loss": 0.7588708432945045, "epoch": 42, "n_parameters": 303306757} +{"train_lr": 4.101306937779829e-05, "train_loss": 0.8100593653885094, "epoch": 43, "n_parameters": 303306757} +{"train_lr": 2.987085471342045e-05, "train_loss": 0.7595094200727102, "epoch": 44, "n_parameters": 303306757} +{"train_lr": 2.0474225379729095e-05, "train_loss": 0.7296602725982666, "epoch": 45, "n_parameters": 303306757} +{"train_lr": 1.2881114721262776e-05, "train_loss": 0.764088637120015, "epoch": 46, "n_parameters": 303306757} +{"train_lr": 7.138336790196452e-06, "train_loss": 0.7429992443806416, "epoch": 47, "n_parameters": 303306757} +{"train_lr": 3.2812977221425987e-06, "train_loss": 0.7327528821455466, "epoch": 48, "n_parameters": 303306757} +{"train_lr": 1.3337774455514033e-06, "train_loss": 0.7304508476643949, "epoch": 49, "n_parameters": 303306757} diff --git a/results/deepdrid/retfound/metrics.json b/results/deepdrid/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..bf8d9946098c1f0ba132bb8cb7814240d2d2dfbc --- /dev/null +++ b/results/deepdrid/retfound/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 400, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.7525, + "balanced_accuracy": 0.6375555555555555, + "precision_macro": 0.692025556711968, + "recall_macro": 0.6375555555555555, + "f1_macro": 0.6534401129901979, + "precision_weighted": 0.7627709624869904, + "recall_weighted": 0.7525, + "f1_weighted": 0.754459272570478, + "cohen_kappa": 0.6338757396449703, + "quadratic_weighted_kappa": 0.8441825745387079, + "mcc": 0.6344217048033686, + "auroc_macro_ovr": 0.927122747483938, + "auroc_weighted_ovr": 0.9462435985308016, + "auprc_macro": 0.7323304575833615, + "auroc_per_class": { + "0": 0.9588375, + "1": 0.8526404151404151, + "2": 0.9327151084010841, + "3": 0.9817496612466124, + "4": 0.9096710526315789 + }, + "per_class": { + "0": { + "precision": 0.8520408163265306, + "recall": 0.835, + "f1-score": 0.8434343434343434, + "support": 200.0 + }, + "1": { + "precision": 0.2926829268292683, + "recall": 0.3333333333333333, + "f1-score": 0.3116883116883117, + "support": 36.0 + }, + "2": { + "precision": 0.65, + "recall": 0.7222222222222222, + "f1-score": 0.6842105263157895, + "support": 72.0 + }, + "3": { + "precision": 0.8472222222222222, + "recall": 0.8472222222222222, + "f1-score": 0.8472222222222222, + "support": 72.0 + }, + "4": { + "precision": 0.8181818181818182, + "recall": 0.45, + "f1-score": 0.5806451612903226, + "support": 20.0 + }, + "accuracy": 0.7525, + "macro avg": { + "precision": 0.692025556711968, + "recall": 0.6375555555555555, + "f1-score": 0.6534401129901979, + "support": 400.0 + }, + "weighted avg": { + "precision": 0.7627709624869904, + "recall": 0.7525, + "f1-score": 0.754459272570478, + "support": 400.0 + } + } +} \ No newline at end of file diff --git a/results/deepdrid/retfound/metrics_test.csv b/results/deepdrid/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..10799139ced841928908703a538bb00161c7ac96 --- /dev/null +++ b/results/deepdrid/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6675626131204458,0.7525,0.6534401129901979,0.927122747483938,0.099,0.5155807389327671,0.692025556711968,0.6375555555555555,0.7323304575833615,0.6338757396449703 diff --git a/results/deepdrid/retfound/metrics_val.csv b/results/deepdrid/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..3fa0dd7d53d7174cf7771564a1866db149c73153 --- /dev/null +++ b/results/deepdrid/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +1.4799616887019231,0.435,0.12125435540069687,0.7256067590300261,0.226,0.087,0.087,0.2,0.4069847649442016,0.0 +1.3491656963641827,0.435,0.12125435540069687,0.7558459342398764,0.226,0.087,0.087,0.2,0.4268120021361913,0.0 +1.163876460148738,0.5375,0.28681479546889005,0.8022851800658914,0.185,0.2089697075177837,0.3525766541151157,0.33752976452949995,0.47654416972969127,0.3248544787694104 +1.1296216524564302,0.5525,0.2953072625698324,0.8264598487821544,0.179,0.2158295281582953,0.3569693094629156,0.3441690919246259,0.5247145979277056,0.3394833948339483 +1.0159740447998047,0.5725,0.3512782868877552,0.8540913439022308,0.171,0.2604516129032258,0.32629049111807734,0.41298027456859804,0.5510043675733071,0.3994099466142176 +1.0000089131868803,0.67,0.48797552460813093,0.8681850928469015,0.132,0.3663164088132145,0.48883567390920335,0.5024620042919716,0.5978422658961067,0.5204533895226331 +0.9199759409977839,0.6425,0.4612483952596155,0.8781344310314992,0.143,0.3427589617496191,0.49852317352808495,0.4875180057030309,0.6086146849269094,0.4924488455873218 +0.9154136364276593,0.625,0.3913359261617002,0.8782575284375314,0.15,0.2991025641025641,0.3556098057166107,0.4468530440662022,0.6138883362500148,0.4625774784135287 +0.9254515721247747,0.6825,0.4642345787976856,0.8879005477926205,0.127,0.35626561713853117,0.6043196225858765,0.48075491665931736,0.6284599546533347,0.5191761632529437 +0.8754060451801007,0.67,0.5714501519665374,0.8860578012773471,0.132,0.41843293692608763,0.5853846153846154,0.5687318105653055,0.6333253292160272,0.5409494001043296 +0.8791058613703802,0.6425,0.4708043341695262,0.8833950266375693,0.143,0.34942101464845454,0.45547391022800865,0.494954581532763,0.6331891573610875,0.4846011064857365 +0.8248477532313421,0.695,0.5978542874616174,0.8992852355324124,0.122,0.44779969624313304,0.6565941337085616,0.5969612252697182,0.6656091477882264,0.5756226520105747 +0.8334504274221567,0.6875,0.5911660925355388,0.9012566520558284,0.125,0.4415369643968769,0.6486071763707046,0.59135079519064,0.6562075740392783,0.5707859767194313 +0.9208796116021963,0.655,0.5783051999840895,0.888791059450797,0.138,0.42098374960280605,0.6116692421053135,0.5873192815357028,0.6507661719430179,0.5232419547079856 +0.8168353667626014,0.71,0.5802790304400023,0.9042157794230457,0.116,0.44128379274877083,0.6680566717740779,0.5791277890466532,0.6591282118737105,0.5905761934174534 +0.8617628904489371,0.6925,0.49672138188028636,0.9013852584045761,0.123,0.3797973099546664,0.5672417637335178,0.5061272305023958,0.6573555421683588,0.5383402769958339 +0.8912057693188007,0.66,0.5684569054134272,0.8932558975744047,0.136,0.41794466404356234,0.6581716184362535,0.5652579592556662,0.6462278365879187,0.5330792735262815 +0.8286451009603647,0.705,0.6054924417955043,0.9001686552395045,0.118,0.45608576206402285,0.6417473327870045,0.6009268894964283,0.6617509820454843,0.5886280046715126 +0.8380820384392371,0.7,0.6071041007487262,0.898972824887794,0.12,0.45361661541733256,0.623556440870596,0.6113995943204868,0.6609469366371583,0.5797142056598487 +0.8050662554227389,0.695,0.6100463908719191,0.9030173106442788,0.122,0.4532943612227499,0.6502129722092658,0.6067222271217332,0.6600648337755116,0.5730458975660117 +0.8401335202730619,0.7,0.5886761159736217,0.8994519304180436,0.12,0.44028917530556877,0.6503176174979054,0.57968015992004,0.6580140871367474,0.5743095833555047 +0.8426446181077224,0.6925,0.5924573003804696,0.9047333907012198,0.123,0.44200341773434343,0.6346832106427325,0.5923744010347767,0.6814630522473372,0.5714659001829109 +0.8606914648642907,0.6975,0.5793643723658224,0.8978633574390216,0.121,0.4313380075573341,0.6259970073897577,0.5637181409295352,0.655448505414559,0.5646150801504056 +0.8863991590646597,0.66,0.5526611008329666,0.8928754481312475,0.136,0.40019791774239566,0.5656631129044923,0.5571678866449129,0.6516276341273042,0.5157384987893462 +0.864152715756343,0.7125,0.601320606388309,0.9033098011802909,0.115,0.4556543093956618,0.6448400200609893,0.582365581914925,0.6760355594669628,0.5865168539325842 +0.868835618862739,0.6625,0.576862175734149,0.893720470700945,0.135,0.42035192882542105,0.5729610955493307,0.5925022782726284,0.643717317310771,0.5300506500966007 +0.920516325877263,0.66,0.536023940042228,0.896234873210855,0.136,0.3871777347614139,0.5753626216993212,0.5392562542258282,0.6332119626278463,0.5083508061600752 +0.8833090846355145,0.7025,0.5876561100810977,0.903089126549945,0.119,0.4364712858991666,0.6738381514287836,0.5706746626686656,0.6688527982870313,0.5719347470278242 +0.8810019951600295,0.69,0.6036470371206858,0.8976576007360808,0.124,0.44721404639721785,0.6294262568294132,0.602454067084105,0.6600174958669787,0.5645914533515924 +0.8745717452122614,0.6975,0.6185543313376122,0.9007072425792384,0.121,0.4614531475636319,0.6452148355439207,0.6161398712408502,0.6672245934008212,0.5766196049615984 +0.8607793404505804,0.69,0.6071987536389647,0.9016309643682077,0.124,0.4501576054764529,0.6042885319808396,0.6116356527618543,0.6579981702573281,0.5681323465389638 +0.9119756038372333,0.7125,0.5900047168118675,0.903659616523869,0.115,0.4430655230756523,0.6403962366903908,0.5707666754857865,0.6597864340172013,0.5811175056458076 +0.9346982836723328,0.69,0.5894541173491994,0.9002171689406838,0.124,0.4359777453886172,0.6030864459535936,0.5802910309551106,0.6569684434035665,0.5571903010391743 +0.9309894305009109,0.7,0.5983199642933981,0.9002132824412568,0.12,0.4493082009927605,0.6093678895282746,0.5991518946409149,0.6469786986381135,0.5807347623290184 +0.8780909501589261,0.69,0.5781459376525827,0.9019444138128373,0.124,0.43294409097523906,0.5850854341736694,0.5735749772172737,0.6474852539915625,0.5638486836320149 +0.8805368542671204,0.69,0.5809417780839927,0.9038568845254724,0.124,0.433266762219796,0.5858823529411765,0.5802275332921775,0.6556551631836405,0.5654611718530979 +0.8798409883792584,0.715,0.6304607182349884,0.9036771340631568,0.114,0.47772322868603484,0.6410456792215526,0.6336290678190316,0.6592764463823997,0.6023440770196735 +0.8637872292445257,0.72,0.6261061300200279,0.9066034085259709,0.112,0.47406534238469183,0.6402271583324215,0.6185322044859923,0.6669427411214446,0.6047849253678677 +0.8765294414300185,0.705,0.6078582347931342,0.9051736436754245,0.118,0.45541993157488136,0.6222508779322938,0.6016553487961901,0.6649374945432966,0.5835318615772849 +0.9051186717473544,0.7,0.5853910944967109,0.901968042037,0.12,0.43493936494807545,0.6190153506589119,0.5679774818473117,0.6562639273437267,0.5625785991579637 +0.8759457239737878,0.72,0.6270981609549058,0.9047121697101517,0.112,0.4745756700798333,0.6369025859777881,0.6196545844724697,0.660380533145238,0.6042682495936684 +0.8970959117779365,0.705,0.5963438409070287,0.9026630897518839,0.118,0.4448824203651872,0.62326736820148,0.5800038216186024,0.6565435361381879,0.5742531389810939 +0.8994409396098211,0.7175,0.6154021172434018,0.9032110611147793,0.113,0.46379846874396924,0.6312865497076023,0.603791927565629,0.6561723880476217,0.5968748885162856 +0.9050251979094285,0.7025,0.6091228400259403,0.9022832192280841,0.119,0.4559540925754564,0.61720643341333,0.6041843783990357,0.6542717268172756,0.5800839832033593 +0.9210980809651889,0.7075,0.6070694845550714,0.9011843039129938,0.117,0.4547585146201392,0.6218618613653294,0.5958635388188259,0.6490710204937897,0.5819861019310812 +0.9264009961715112,0.7025,0.6020924241509438,0.901302554425173,0.119,0.44893774480537746,0.6190651464495207,0.5899811858776495,0.6498593047128468,0.5742549461557727 +0.9228192797073951,0.705,0.6037013050717077,0.901648371088873,0.118,0.45121246877375903,0.6220362757039543,0.5910056736337713,0.6502566330728419,0.5780138039552265 +0.9199902690373934,0.7075,0.6060933377347647,0.9016578474134291,0.117,0.45414081901850184,0.6211007386338546,0.5947141135314695,0.6500159730702054,0.5822547531911095 +0.9209988438166105,0.705,0.6032980396773141,0.9017303418741193,0.118,0.4509150125668889,0.6188204444258402,0.5917729370608813,0.649936385586606,0.5785940038926486 +0.9206776114610525,0.7075,0.6060933377347647,0.9017511744923944,0.117,0.45414081901850184,0.6211007386338546,0.5947141135314695,0.6498583754565053,0.5822547531911095 diff --git a/results/deepdrid/retfound/pr.png b/results/deepdrid/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..26159bb01869b9ba6fcb8edb61927016c0e0e039 --- /dev/null +++ b/results/deepdrid/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc1cdc14e31b0232f3c2fb48952655fbe42ffc7a7d52212e24480013640d1050 +size 85614 diff --git a/results/deepdrid/retfound/roc.png b/results/deepdrid/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..770ccd9df7cb7a1179f542a4a58b3dc236747d91 --- /dev/null +++ b/results/deepdrid/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac42e3a8157fde634eabf87eca1e99d78104eeed59026e148514be5f0f10da45 +size 86948 diff --git a/results/deepdrid/retfound/test_pred.npz b/results/deepdrid/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..efeea5bca00c4e620343d62cdfcc236240876814 --- /dev/null +++ b/results/deepdrid/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcaff01e3088fe568b4c44d4d7ab03c2f1e896c1498913c6b9210778c42f7d29 +size 7710 diff --git a/results/deepdrid/retfound/train.log b/results/deepdrid/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..ffc7830b8d221aa6ec7bfdbf0db29137fbf14502 --- /dev/null +++ b/results/deepdrid/retfound/train.log @@ -0,0 +1,834 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 14:01:51.338596173 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:01:52.577094] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:01:52.577344] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/DR/deepdrid', +nb_classes=5, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/deepdrid', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:01:55.505158] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:01:57.088275] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:01:57.464460] Sampler_train = +[14:01:57.505823] len of train_set: 1184 +[14:01:58.122056] [Adaptation] Full fine-tuning: training all parameters. +[14:01:58.123654] number of trainable params (M): 303.31 +[14:01:58.123745] base lr: 5.00e-03 +[14:01:58.123832] actual lr: 6.25e-04 +[14:01:58.123894] accumulate grad iterations: 1 +[14:01:58.123957] effective batch size: 32 +[14:01:58.127044] criterion = CrossEntropyLoss() +[14:01:58.127132] Start training for 50 epochs +[14:01:58.129339] log_dir: ./output_logs/retfound +[14:02:00.870553] Epoch: [0] [ 0/37] eta: 0:01:41 lr: 0.000000 loss: 1.6095 (1.6095) time: 2.7403 data: 1.9373 max mem: 7340 +[14:02:03.754696] Epoch: [0] [20/37] eta: 0:00:04 lr: 0.000034 loss: 1.5944 (1.5933) time: 0.1441 data: 0.0001 max mem: 9672 +[14:02:06.048603] Epoch: [0] [36/37] eta: 0:00:00 lr: 0.000061 loss: 1.5412 (1.5490) time: 0.1436 data: 0.0002 max mem: 9672 +[14:02:06.116997] Epoch: [0] Total time: 0:00:07 (0.2159 s / it) +[14:02:06.126587] Averaged stats: lr: 0.000061 loss: 1.5412 (1.5490) +[14:02:08.277631] val: [ 0/13] eta: 0:00:27 loss: 1.1863 (1.1863) time: 2.1338 data: 2.0852 max mem: 9672 +[14:02:10.074513] val: [10/13] eta: 0:00:01 loss: 1.5634 (1.4209) time: 0.3572 data: 0.3214 max mem: 9672 +[14:02:10.870572] val: [12/13] eta: 0:00:00 loss: 1.5649 (1.4800) time: 0.3635 data: 0.2720 max mem: 9672 +[14:02:10.946956] val: Total time: 0:00:04 (0.3696 s / it) +[14:02:10.966815] val loss: 1.4799616887019231 +[14:02:10.967013] Accuracy: 0.4350, F1 Score: 0.1213, ROC AUC: 0.7256, Hamming Loss: 0.2260, + Jaccard Score: 0.0870, Precision: 0.0870, Recall: 0.2000, + Average Precision: 0.4070, Kappa: 0.0000, Score: 0.2823 +[14:02:12.867206] Best epoch = 0, Best score = 0.2823 +[14:02:12.943008] log_dir: ./output_logs/retfound +[14:02:14.822839] Epoch: [1] [ 0/37] eta: 0:01:09 lr: 0.000063 loss: 1.3747 (1.3747) time: 1.8785 data: 1.7130 max mem: 9672 +[14:02:17.686036] Epoch: [1] [20/37] eta: 0:00:03 lr: 0.000096 loss: 1.3894 (1.4175) time: 0.1431 data: 0.0002 max mem: 9672 +[14:02:19.974891] Epoch: [1] [36/37] eta: 0:00:00 lr: 0.000123 loss: 1.3510 (1.3931) time: 0.1431 data: 0.0001 max mem: 9672 +[14:02:20.057697] Epoch: [1] Total time: 0:00:07 (0.1923 s / it) +[14:02:20.062410] Averaged stats: lr: 0.000123 loss: 1.3510 (1.3931) +[14:02:22.148286] val: [ 0/13] eta: 0:00:26 loss: 0.7874 (0.7874) time: 2.0672 data: 2.0322 max mem: 9672 +[14:02:23.743824] val: [10/13] eta: 0:00:00 loss: 1.4057 (1.2126) time: 0.3329 data: 0.2987 max mem: 9672 +[14:02:23.806149] val: [12/13] eta: 0:00:00 loss: 1.4416 (1.3492) time: 0.2864 data: 0.2533 max mem: 9672 +[14:02:23.882584] val: Total time: 0:00:03 (0.2925 s / it) +[14:02:23.898986] val loss: 1.3491656963641827 +[14:02:23.899265] Accuracy: 0.4350, F1 Score: 0.1213, ROC AUC: 0.7558, Hamming Loss: 0.2260, + Jaccard Score: 0.0870, Precision: 0.0870, Recall: 0.2000, + Average Precision: 0.4268, Kappa: 0.0000, Score: 0.2924 +[14:02:25.847725] Best epoch = 1, Best score = 0.2924 +[14:02:25.913827] log_dir: ./output_logs/retfound +[14:02:27.811179] Epoch: [2] [ 0/37] eta: 0:01:10 lr: 0.000125 loss: 1.3836 (1.3836) time: 1.8963 data: 1.7474 max mem: 9672 +[14:02:30.670623] Epoch: [2] [20/37] eta: 0:00:03 lr: 0.000159 loss: 1.3326 (1.3308) time: 0.1429 data: 0.0002 max mem: 9672 +[14:02:32.962771] Epoch: [2] [36/37] eta: 0:00:00 lr: 0.000186 loss: 1.2639 (1.2970) time: 0.1429 data: 0.0001 max mem: 9672 +[14:02:33.043097] Epoch: [2] Total time: 0:00:07 (0.1927 s / it) +[14:02:33.051443] Averaged stats: lr: 0.000186 loss: 1.2639 (1.2970) +[14:02:35.172485] val: [ 0/13] eta: 0:00:27 loss: 0.6546 (0.6546) time: 2.1044 data: 2.0682 max mem: 9672 +[14:02:36.856379] val: [10/13] eta: 0:00:01 loss: 1.1687 (1.0533) time: 0.3443 data: 0.3096 max mem: 9672 +[14:02:36.912712] val: [12/13] eta: 0:00:00 loss: 1.1833 (1.1639) time: 0.2956 data: 0.2620 max mem: 9672 +[14:02:36.998923] val: Total time: 0:00:03 (0.3024 s / it) +[14:02:37.015840] val loss: 1.163876460148738 +[14:02:37.016100] Accuracy: 0.5375, F1 Score: 0.2868, ROC AUC: 0.8023, Hamming Loss: 0.1850, + Jaccard Score: 0.2090, Precision: 0.3526, Recall: 0.3375, + Average Precision: 0.4765, Kappa: 0.3249, Score: 0.4713 +[14:02:38.902080] Best epoch = 2, Best score = 0.4713 +[14:02:38.965885] log_dir: ./output_logs/retfound +[14:02:40.870147] Epoch: [3] [ 0/37] eta: 0:01:10 lr: 0.000188 loss: 1.3325 (1.3325) time: 1.9031 data: 1.7529 max mem: 9672 +[14:02:43.738669] Epoch: [3] [20/37] eta: 0:00:03 lr: 0.000221 loss: 1.1988 (1.2459) time: 0.1434 data: 0.0002 max mem: 9672 +[14:02:45.620304] Epoch: [3] [36/37] eta: 0:00:00 lr: 0.000248 loss: 1.1465 (1.2250) time: 0.1230 data: 0.0001 max mem: 9672 +[14:02:45.704134] Epoch: [3] Total time: 0:00:06 (0.1821 s / it) +[14:02:45.713156] Averaged stats: lr: 0.000248 loss: 1.1465 (1.2250) +[14:02:47.662683] val: [ 0/13] eta: 0:00:25 loss: 0.5287 (0.5287) time: 1.9334 data: 1.8988 max mem: 9672 +[14:02:49.308830] val: [10/13] eta: 0:00:00 loss: 0.9931 (0.9958) time: 0.3253 data: 0.2907 max mem: 9672 +[14:02:49.400985] val: [12/13] eta: 0:00:00 loss: 1.0564 (1.1296) time: 0.2823 data: 0.2496 max mem: 9672 +[14:02:49.480288] val: Total time: 0:00:03 (0.2886 s / it) +[14:02:49.509857] val loss: 1.1296216524564302 +[14:02:49.510099] Accuracy: 0.5525, F1 Score: 0.2953, ROC AUC: 0.8265, Hamming Loss: 0.1790, + Jaccard Score: 0.2158, Precision: 0.3570, Recall: 0.3442, + Average Precision: 0.5247, Kappa: 0.3395, Score: 0.4871 +[14:02:51.343092] Best epoch = 3, Best score = 0.4871 +[14:02:51.407079] log_dir: ./output_logs/retfound +[14:02:53.351912] Epoch: [4] [ 0/37] eta: 0:01:11 lr: 0.000250 loss: 1.2185 (1.2185) time: 1.9435 data: 1.7931 max mem: 9672 +[14:02:56.238765] Epoch: [4] [20/37] eta: 0:00:03 lr: 0.000284 loss: 1.1092 (1.1587) time: 0.1443 data: 0.0003 max mem: 9672 +[14:02:58.533035] Epoch: [4] [36/37] eta: 0:00:00 lr: 0.000311 loss: 1.1992 (1.1644) time: 0.1435 data: 0.0002 max mem: 9672 +[14:02:58.621545] Epoch: [4] Total time: 0:00:07 (0.1950 s / it) +[14:02:58.630043] Averaged stats: lr: 0.000311 loss: 1.1992 (1.1644) +[14:03:00.794463] val: [ 0/13] eta: 0:00:27 loss: 0.7497 (0.7497) time: 2.1464 data: 2.1115 max mem: 9672 +[14:03:02.456958] val: [10/13] eta: 0:00:01 loss: 0.8920 (0.9953) time: 0.3462 data: 0.3115 max mem: 9672 +[14:03:02.528115] val: [12/13] eta: 0:00:00 loss: 0.8920 (1.0160) time: 0.2984 data: 0.2647 max mem: 9672 +[14:03:02.611399] val: Total time: 0:00:03 (0.3049 s / it) +[14:03:02.632292] val loss: 1.0159740447998047 +[14:03:02.632498] Accuracy: 0.5725, F1 Score: 0.3513, ROC AUC: 0.8541, Hamming Loss: 0.1710, + Jaccard Score: 0.2605, Precision: 0.3263, Recall: 0.4130, + Average Precision: 0.5510, Kappa: 0.3994, Score: 0.5349 +[14:03:04.340287] Best epoch = 4, Best score = 0.5349 +[14:03:04.409877] log_dir: ./output_logs/retfound +[14:03:06.450164] Epoch: [5] [ 0/37] eta: 0:01:15 lr: 0.000313 loss: 1.1002 (1.1002) time: 2.0391 data: 1.8987 max mem: 9672 +[14:03:09.337547] Epoch: [5] [20/37] eta: 0:00:03 lr: 0.000346 loss: 1.2011 (1.1842) time: 0.1443 data: 0.0002 max mem: 9672 +[14:03:11.623959] Epoch: [5] [36/37] eta: 0:00:00 lr: 0.000373 loss: 1.0937 (1.1347) time: 0.1434 data: 0.0002 max mem: 9672 +[14:03:11.716480] Epoch: [5] Total time: 0:00:07 (0.1975 s / it) +[14:03:11.726050] Averaged stats: lr: 0.000373 loss: 1.0937 (1.1347) +[14:03:13.733107] val: [ 0/13] eta: 0:00:25 loss: 0.4737 (0.4737) time: 1.9883 data: 1.9532 max mem: 9672 +[14:03:15.355402] val: [10/13] eta: 0:00:00 loss: 1.0871 (0.8889) time: 0.3282 data: 0.2934 max mem: 9672 +[14:03:15.501296] val: [12/13] eta: 0:00:00 loss: 1.0871 (1.0000) time: 0.2888 data: 0.2549 max mem: 9672 +[14:03:15.578140] val: Total time: 0:00:03 (0.2949 s / it) +[14:03:15.595078] val loss: 1.0000089131868803 +[14:03:15.595333] Accuracy: 0.6700, F1 Score: 0.4880, ROC AUC: 0.8682, Hamming Loss: 0.1320, + Jaccard Score: 0.3663, Precision: 0.4888, Recall: 0.5025, + Average Precision: 0.5978, Kappa: 0.5205, Score: 0.6255 +[14:03:17.449049] Best epoch = 5, Best score = 0.6255 +[14:03:17.515119] log_dir: ./output_logs/retfound +[14:03:19.362436] Epoch: [6] [ 0/37] eta: 0:01:08 lr: 0.000375 loss: 0.9421 (0.9421) time: 1.8462 data: 1.6998 max mem: 9672 +[14:03:22.231955] Epoch: [6] [20/37] eta: 0:00:03 lr: 0.000409 loss: 1.0819 (1.1029) time: 0.1434 data: 0.0002 max mem: 9672 +[14:03:24.517387] Epoch: [6] [36/37] eta: 0:00:00 lr: 0.000436 loss: 1.1510 (1.1119) time: 0.1431 data: 0.0001 max mem: 9672 +[14:03:24.598311] Epoch: [6] Total time: 0:00:07 (0.1914 s / it) +[14:03:24.607956] Averaged stats: lr: 0.000436 loss: 1.1510 (1.1119) +[14:03:26.514769] val: [ 0/13] eta: 0:00:24 loss: 0.6130 (0.6130) time: 1.8895 data: 1.8544 max mem: 9672 +[14:03:28.074055] val: [10/13] eta: 0:00:00 loss: 0.7494 (0.8575) time: 0.3135 data: 0.2787 max mem: 9672 +[14:03:28.235697] val: [12/13] eta: 0:00:00 loss: 0.7494 (0.9200) time: 0.2776 data: 0.2439 max mem: 9672 +[14:03:28.312439] val: Total time: 0:00:03 (0.2837 s / it) +[14:03:28.328623] val loss: 0.9199759409977839 +[14:03:28.328838] Accuracy: 0.6425, F1 Score: 0.4612, ROC AUC: 0.8781, Hamming Loss: 0.1430, + Jaccard Score: 0.3428, Precision: 0.4985, Recall: 0.4875, + Average Precision: 0.6086, Kappa: 0.4924, Score: 0.6106 +[14:03:28.372609] Best epoch = 5, Best score = 0.6255 +[14:03:28.623673] log_dir: ./output_logs/retfound +[14:03:30.459307] Epoch: [7] [ 0/37] eta: 0:01:07 lr: 0.000438 loss: 0.7618 (0.7618) time: 1.8344 data: 1.6857 max mem: 9672 +[14:03:33.325402] Epoch: [7] [20/37] eta: 0:00:03 lr: 0.000471 loss: 1.1075 (1.0945) time: 0.1433 data: 0.0002 max mem: 9672 +[14:03:35.614139] Epoch: [7] [36/37] eta: 0:00:00 lr: 0.000498 loss: 1.1211 (1.1153) time: 0.1430 data: 0.0001 max mem: 9672 +[14:03:35.695126] Epoch: [7] Total time: 0:00:07 (0.1911 s / it) +[14:03:35.704791] Averaged stats: lr: 0.000498 loss: 1.1211 (1.1153) +[14:03:37.653837] val: [ 0/13] eta: 0:00:25 loss: 0.6058 (0.6058) time: 1.9324 data: 1.8981 max mem: 9672 +[14:03:39.242949] val: [10/13] eta: 0:00:00 loss: 0.8344 (0.8658) time: 0.3201 data: 0.2855 max mem: 9672 +[14:03:39.374831] val: [12/13] eta: 0:00:00 loss: 0.8344 (0.9154) time: 0.2809 data: 0.2473 max mem: 9672 +[14:03:39.470498] val: Total time: 0:00:03 (0.2884 s / it) +[14:03:39.487094] val loss: 0.9154136364276593 +[14:03:39.487302] Accuracy: 0.6250, F1 Score: 0.3913, ROC AUC: 0.8783, Hamming Loss: 0.1500, + Jaccard Score: 0.2991, Precision: 0.3556, Recall: 0.4469, + Average Precision: 0.6139, Kappa: 0.4626, Score: 0.5774 +[14:03:39.531037] Best epoch = 5, Best score = 0.6255 +[14:03:39.803355] log_dir: ./output_logs/retfound +[14:03:41.755043] Epoch: [8] [ 0/37] eta: 0:01:12 lr: 0.000500 loss: 1.1173 (1.1173) time: 1.9506 data: 1.8046 max mem: 9672 +[14:03:44.623320] Epoch: [8] [20/37] eta: 0:00:03 lr: 0.000534 loss: 1.0183 (1.0436) time: 0.1434 data: 0.0002 max mem: 9672 +[14:03:46.917968] Epoch: [8] [36/37] eta: 0:00:00 lr: 0.000561 loss: 1.1403 (1.0973) time: 0.1431 data: 0.0001 max mem: 9672 +[14:03:47.013109] Epoch: [8] Total time: 0:00:07 (0.1949 s / it) +[14:03:47.021341] Averaged stats: lr: 0.000561 loss: 1.1403 (1.0973) +[14:03:49.113020] val: [ 0/13] eta: 0:00:26 loss: 0.3809 (0.3809) time: 2.0749 data: 2.0397 max mem: 9672 +[14:03:50.890730] val: [10/13] eta: 0:00:01 loss: 0.8314 (0.8032) time: 0.3501 data: 0.3154 max mem: 9672 +[14:03:50.948848] val: [12/13] eta: 0:00:00 loss: 0.8314 (0.9255) time: 0.3007 data: 0.2669 max mem: 9672 +[14:03:51.030105] val: Total time: 0:00:03 (0.3071 s / it) +[14:03:51.058480] val loss: 0.9254515721247747 +[14:03:51.058909] Accuracy: 0.6825, F1 Score: 0.4642, ROC AUC: 0.8879, Hamming Loss: 0.1270, + Jaccard Score: 0.3563, Precision: 0.6043, Recall: 0.4808, + Average Precision: 0.6285, Kappa: 0.5192, Score: 0.6238 +[14:03:51.095111] Best epoch = 5, Best score = 0.6255 +[14:03:51.343432] log_dir: ./output_logs/retfound +[14:03:53.191559] Epoch: [9] [ 0/37] eta: 0:01:08 lr: 0.000562 loss: 1.2318 (1.2318) time: 1.8472 data: 1.7034 max mem: 9672 +[14:03:56.052928] Epoch: [9] [20/37] eta: 0:00:03 lr: 0.000596 loss: 1.0546 (1.0556) time: 0.1430 data: 0.0002 max mem: 9672 +[14:03:58.344753] Epoch: [9] [36/37] eta: 0:00:00 lr: 0.000623 loss: 1.0974 (1.0577) time: 0.1431 data: 0.0001 max mem: 9672 +[14:03:58.440853] Epoch: [9] Total time: 0:00:07 (0.1918 s / it) +[14:03:58.449625] Averaged stats: lr: 0.000623 loss: 1.0974 (1.0577) +[14:04:00.641224] val: [ 0/13] eta: 0:00:28 loss: 0.6497 (0.6497) time: 2.1787 data: 2.1438 max mem: 9672 +[14:04:02.397410] val: [10/13] eta: 0:00:01 loss: 0.8190 (0.8222) time: 0.3577 data: 0.3230 max mem: 9672 +[14:04:02.453388] val: [12/13] eta: 0:00:00 loss: 0.8190 (0.8754) time: 0.3069 data: 0.2734 max mem: 9672 +[14:04:02.537515] val: Total time: 0:00:04 (0.3135 s / it) +[14:04:02.553363] val loss: 0.8754060451801007 +[14:04:02.553565] Accuracy: 0.6700, F1 Score: 0.5715, ROC AUC: 0.8861, Hamming Loss: 0.1320, + Jaccard Score: 0.4184, Precision: 0.5854, Recall: 0.5687, + Average Precision: 0.6333, Kappa: 0.5409, Score: 0.6662 +[14:04:04.316084] Best epoch = 9, Best score = 0.6662 +[14:04:04.380692] log_dir: ./output_logs/retfound +[14:04:06.176743] Epoch: [10] [ 0/37] eta: 0:01:06 lr: 0.000625 loss: 1.0183 (1.0183) time: 1.7950 data: 1.6511 max mem: 9672 +[14:04:09.036515] Epoch: [10] [20/37] eta: 0:00:03 lr: 0.000625 loss: 1.0833 (1.0810) time: 0.1429 data: 0.0001 max mem: 9672 +[14:04:11.324597] Epoch: [10] [36/37] eta: 0:00:00 lr: 0.000624 loss: 0.9523 (1.0420) time: 0.1428 data: 0.0001 max mem: 9672 +[14:04:11.410808] Epoch: [10] Total time: 0:00:07 (0.1900 s / it) +[14:04:11.419426] Averaged stats: lr: 0.000624 loss: 0.9523 (1.0420) +[14:04:13.308446] val: [ 0/13] eta: 0:00:24 loss: 0.5154 (0.5154) time: 1.8726 data: 1.8387 max mem: 9672 +[14:04:14.943530] val: [10/13] eta: 0:00:00 loss: 0.7395 (0.8136) time: 0.3188 data: 0.2843 max mem: 9672 +[14:04:14.999856] val: [12/13] eta: 0:00:00 loss: 0.7395 (0.8791) time: 0.2741 data: 0.2406 max mem: 9672 +[14:04:15.073971] val: Total time: 0:00:03 (0.2799 s / it) +[14:04:15.091592] val loss: 0.8791058613703802 +[14:04:15.091828] Accuracy: 0.6425, F1 Score: 0.4708, ROC AUC: 0.8834, Hamming Loss: 0.1430, + Jaccard Score: 0.3494, Precision: 0.4555, Recall: 0.4950, + Average Precision: 0.6332, Kappa: 0.4846, Score: 0.6129 +[14:04:15.132989] Best epoch = 9, Best score = 0.6662 +[14:04:15.396738] log_dir: ./output_logs/retfound +[14:04:17.204385] Epoch: [11] [ 0/37] eta: 0:01:06 lr: 0.000624 loss: 0.8419 (0.8419) time: 1.8067 data: 1.6617 max mem: 9672 +[14:04:20.043465] Epoch: [11] [20/37] eta: 0:00:03 lr: 0.000623 loss: 0.9605 (1.0181) time: 0.1419 data: 0.0002 max mem: 9672 +[14:04:22.335887] Epoch: [11] [36/37] eta: 0:00:00 lr: 0.000621 loss: 1.0317 (1.0214) time: 0.1429 data: 0.0001 max mem: 9672 +[14:04:22.421269] Epoch: [11] Total time: 0:00:07 (0.1898 s / it) +[14:04:22.430914] Averaged stats: lr: 0.000621 loss: 1.0317 (1.0214) +[14:04:24.278291] val: [ 0/13] eta: 0:00:23 loss: 0.5038 (0.5038) time: 1.8349 data: 1.8012 max mem: 9672 +[14:04:26.076031] val: [10/13] eta: 0:00:00 loss: 0.7720 (0.7207) time: 0.3302 data: 0.2952 max mem: 9672 +[14:04:26.132167] val: [12/13] eta: 0:00:00 loss: 0.7720 (0.8248) time: 0.2837 data: 0.2498 max mem: 9672 +[14:04:26.207381] val: Total time: 0:00:03 (0.2896 s / it) +[14:04:26.223183] val loss: 0.8248477532313421 +[14:04:26.223493] Accuracy: 0.6950, F1 Score: 0.5979, ROC AUC: 0.8993, Hamming Loss: 0.1220, + Jaccard Score: 0.4478, Precision: 0.6566, Recall: 0.5970, + Average Precision: 0.6656, Kappa: 0.5756, Score: 0.6909 +[14:04:28.101963] Best epoch = 11, Best score = 0.6909 +[14:04:28.174609] log_dir: ./output_logs/retfound +[14:04:29.998093] Epoch: [12] [ 0/37] eta: 0:01:07 lr: 0.000621 loss: 1.2073 (1.2073) time: 1.8222 data: 1.6736 max mem: 9672 +[14:04:32.885280] Epoch: [12] [20/37] eta: 0:00:03 lr: 0.000619 loss: 0.9371 (0.9965) time: 0.1443 data: 0.0007 max mem: 9672 +[14:04:35.173361] Epoch: [12] [36/37] eta: 0:00:00 lr: 0.000617 loss: 1.0075 (1.0244) time: 0.1432 data: 0.0002 max mem: 9672 +[14:04:35.251651] Epoch: [12] Total time: 0:00:07 (0.1913 s / it) +[14:04:35.259833] Averaged stats: lr: 0.000617 loss: 1.0075 (1.0244) +[14:04:37.191075] val: [ 0/13] eta: 0:00:24 loss: 0.7321 (0.7321) time: 1.9136 data: 1.8793 max mem: 9672 +[14:04:38.789194] val: [10/13] eta: 0:00:00 loss: 0.6338 (0.7260) time: 0.3192 data: 0.2845 max mem: 9672 +[14:04:38.844997] val: [12/13] eta: 0:00:00 loss: 0.6673 (0.8335) time: 0.2743 data: 0.2408 max mem: 9672 +[14:04:38.927107] val: Total time: 0:00:03 (0.2808 s / it) +[14:04:38.946572] val loss: 0.8334504274221567 +[14:04:38.946770] Accuracy: 0.6875, F1 Score: 0.5912, ROC AUC: 0.9013, Hamming Loss: 0.1250, + Jaccard Score: 0.4415, Precision: 0.6486, Recall: 0.5914, + Average Precision: 0.6562, Kappa: 0.5708, Score: 0.6877 +[14:04:38.990169] Best epoch = 11, Best score = 0.6909 +[14:04:39.229912] log_dir: ./output_logs/retfound +[14:04:41.084445] Epoch: [13] [ 0/37] eta: 0:01:08 lr: 0.000616 loss: 0.8524 (0.8524) time: 1.8535 data: 1.7100 max mem: 9672 +[14:04:43.937680] Epoch: [13] [20/37] eta: 0:00:03 lr: 0.000613 loss: 1.0332 (1.0156) time: 0.1426 data: 0.0002 max mem: 9672 +[14:04:46.219456] Epoch: [13] [36/37] eta: 0:00:00 lr: 0.000610 loss: 1.0210 (1.0167) time: 0.1426 data: 0.0001 max mem: 9672 +[14:04:46.299370] Epoch: [13] Total time: 0:00:07 (0.1911 s / it) +[14:04:46.308077] Averaged stats: lr: 0.000610 loss: 1.0210 (1.0167) +[14:04:48.171018] val: [ 0/13] eta: 0:00:23 loss: 0.4166 (0.4166) time: 1.8460 data: 1.8113 max mem: 9672 +[14:04:49.753498] val: [10/13] eta: 0:00:00 loss: 0.8734 (0.8161) time: 0.3116 data: 0.2772 max mem: 9672 +[14:04:49.809659] val: [12/13] eta: 0:00:00 loss: 0.8734 (0.9209) time: 0.2680 data: 0.2346 max mem: 9672 +[14:04:49.898040] val: Total time: 0:00:03 (0.2749 s / it) +[14:04:49.916687] val loss: 0.9208796116021963 +[14:04:49.916895] Accuracy: 0.6550, F1 Score: 0.5783, ROC AUC: 0.8888, Hamming Loss: 0.1380, + Jaccard Score: 0.4210, Precision: 0.6117, Recall: 0.5873, + Average Precision: 0.6508, Kappa: 0.5232, Score: 0.6634 +[14:04:49.963043] Best epoch = 11, Best score = 0.6909 +[14:04:50.196264] log_dir: ./output_logs/retfound +[14:04:52.243783] Epoch: [14] [ 0/37] eta: 0:01:15 lr: 0.000610 loss: 1.0611 (1.0611) time: 2.0464 data: 1.9034 max mem: 9672 +[14:04:55.094285] Epoch: [14] [20/37] eta: 0:00:03 lr: 0.000605 loss: 1.0312 (0.9865) time: 0.1425 data: 0.0002 max mem: 9672 +[14:04:57.373340] Epoch: [14] [36/37] eta: 0:00:00 lr: 0.000602 loss: 0.9145 (0.9690) time: 0.1425 data: 0.0001 max mem: 9672 +[14:04:57.452726] Epoch: [14] Total time: 0:00:07 (0.1961 s / it) +[14:04:57.461862] Averaged stats: lr: 0.000602 loss: 0.9145 (0.9690) +[14:04:59.319351] val: [ 0/13] eta: 0:00:23 loss: 0.4882 (0.4882) time: 1.8461 data: 1.8097 max mem: 9672 +[14:05:00.863819] val: [10/13] eta: 0:00:00 loss: 0.6639 (0.6727) time: 0.3082 data: 0.2735 max mem: 9672 +[14:05:00.919295] val: [12/13] eta: 0:00:00 loss: 0.6662 (0.8168) time: 0.2650 data: 0.2314 max mem: 9672 +[14:05:01.000621] val: Total time: 0:00:03 (0.2714 s / it) +[14:05:01.016513] val loss: 0.8168353667626014 +[14:05:01.016717] Accuracy: 0.7100, F1 Score: 0.5803, ROC AUC: 0.9042, Hamming Loss: 0.1160, + Jaccard Score: 0.4413, Precision: 0.6681, Recall: 0.5791, + Average Precision: 0.6591, Kappa: 0.5906, Score: 0.6917 +[14:05:02.972589] Best epoch = 14, Best score = 0.6917 +[14:05:03.036078] log_dir: ./output_logs/retfound +[14:05:04.859083] Epoch: [15] [ 0/37] eta: 0:01:07 lr: 0.000601 loss: 0.8427 (0.8427) time: 1.8216 data: 1.6708 max mem: 9672 +[14:05:07.726501] Epoch: [15] [20/37] eta: 0:00:03 lr: 0.000596 loss: 0.9677 (0.9511) time: 0.1433 data: 0.0002 max mem: 9672 +[14:05:10.018807] Epoch: [15] [36/37] eta: 0:00:00 lr: 0.000591 loss: 0.8993 (0.9579) time: 0.1432 data: 0.0001 max mem: 9672 +[14:05:10.097829] Epoch: [15] Total time: 0:00:07 (0.1909 s / it) +[14:05:10.107272] Averaged stats: lr: 0.000591 loss: 0.8993 (0.9579) +[14:05:11.986731] val: [ 0/13] eta: 0:00:24 loss: 0.3283 (0.3283) time: 1.8629 data: 1.8263 max mem: 9672 +[14:05:13.670986] val: [10/13] eta: 0:00:00 loss: 0.6941 (0.7243) time: 0.3224 data: 0.2875 max mem: 9672 +[14:05:13.727372] val: [12/13] eta: 0:00:00 loss: 0.6941 (0.8618) time: 0.2771 data: 0.2433 max mem: 9672 +[14:05:13.810132] val: Total time: 0:00:03 (0.2836 s / it) +[14:05:13.828344] val loss: 0.8617628904489371 +[14:05:13.828587] Accuracy: 0.6925, F1 Score: 0.4967, ROC AUC: 0.9014, Hamming Loss: 0.1230, + Jaccard Score: 0.3798, Precision: 0.5672, Recall: 0.5061, + Average Precision: 0.6574, Kappa: 0.5383, Score: 0.6455 +[14:05:13.871593] Best epoch = 14, Best score = 0.6917 +[14:05:14.112589] log_dir: ./output_logs/retfound +[14:05:16.051774] Epoch: [16] [ 0/37] eta: 0:01:11 lr: 0.000591 loss: 0.8572 (0.8572) time: 1.9374 data: 1.7794 max mem: 9672 +[14:05:18.918955] Epoch: [16] [20/37] eta: 0:00:03 lr: 0.000585 loss: 0.8895 (0.8941) time: 0.1433 data: 0.0002 max mem: 9672 +[14:05:21.206592] Epoch: [16] [36/37] eta: 0:00:00 lr: 0.000579 loss: 0.9455 (0.9440) time: 0.1430 data: 0.0001 max mem: 9672 +[14:05:21.284163] Epoch: [16] Total time: 0:00:07 (0.1938 s / it) +[14:05:21.293571] Averaged stats: lr: 0.000579 loss: 0.9455 (0.9440) +[14:05:23.233758] val: [ 0/13] eta: 0:00:24 loss: 0.6158 (0.6158) time: 1.9224 data: 1.8891 max mem: 9672 +[14:05:24.863970] val: [10/13] eta: 0:00:00 loss: 0.6769 (0.7426) time: 0.3229 data: 0.2889 max mem: 9672 +[14:05:24.919960] val: [12/13] eta: 0:00:00 loss: 0.7799 (0.8912) time: 0.2775 data: 0.2445 max mem: 9672 +[14:05:24.998617] val: Total time: 0:00:03 (0.2837 s / it) +[14:05:25.014754] val loss: 0.8912057693188007 +[14:05:25.014958] Accuracy: 0.6600, F1 Score: 0.5685, ROC AUC: 0.8933, Hamming Loss: 0.1360, + Jaccard Score: 0.4179, Precision: 0.6582, Recall: 0.5653, + Average Precision: 0.6462, Kappa: 0.5331, Score: 0.6649 +[14:05:25.056468] Best epoch = 14, Best score = 0.6917 +[14:05:25.289051] log_dir: ./output_logs/retfound +[14:05:27.232902] Epoch: [17] [ 0/37] eta: 0:01:11 lr: 0.000579 loss: 1.1849 (1.1849) time: 1.9424 data: 1.7944 max mem: 9672 +[14:05:30.106187] Epoch: [17] [20/37] eta: 0:00:03 lr: 0.000572 loss: 0.9461 (0.9927) time: 0.1436 data: 0.0002 max mem: 9672 +[14:05:32.396258] Epoch: [17] [36/37] eta: 0:00:00 lr: 0.000566 loss: 0.9018 (0.9654) time: 0.1430 data: 0.0001 max mem: 9672 +[14:05:32.474509] Epoch: [17] Total time: 0:00:07 (0.1942 s / it) +[14:05:32.482557] Averaged stats: lr: 0.000566 loss: 0.9018 (0.9654) +[14:05:34.370833] val: [ 0/13] eta: 0:00:24 loss: 0.6977 (0.6977) time: 1.8709 data: 1.8355 max mem: 9672 +[14:05:35.889433] val: [10/13] eta: 0:00:00 loss: 0.6477 (0.7423) time: 0.3081 data: 0.2734 max mem: 9672 +[14:05:36.063304] val: [12/13] eta: 0:00:00 loss: 0.6477 (0.8286) time: 0.2740 data: 0.2404 max mem: 9672 +[14:05:36.148535] val: Total time: 0:00:03 (0.2807 s / it) +[14:05:36.173593] val loss: 0.8286451009603647 +[14:05:36.173868] Accuracy: 0.7050, F1 Score: 0.6055, ROC AUC: 0.9002, Hamming Loss: 0.1180, + Jaccard Score: 0.4561, Precision: 0.6417, Recall: 0.6009, + Average Precision: 0.6618, Kappa: 0.5886, Score: 0.6981 +[14:05:38.173653] Best epoch = 17, Best score = 0.6981 +[14:05:38.251255] log_dir: ./output_logs/retfound +[14:05:40.106358] Epoch: [18] [ 0/37] eta: 0:01:08 lr: 0.000565 loss: 0.9553 (0.9553) time: 1.8538 data: 1.7068 max mem: 9672 +[14:05:42.976704] Epoch: [18] [20/37] eta: 0:00:03 lr: 0.000557 loss: 0.9586 (0.9413) time: 0.1435 data: 0.0002 max mem: 9672 +[14:05:45.265237] Epoch: [18] [36/37] eta: 0:00:00 lr: 0.000551 loss: 0.8949 (0.9285) time: 0.1431 data: 0.0001 max mem: 9672 +[14:05:45.350056] Epoch: [18] Total time: 0:00:07 (0.1919 s / it) +[14:05:45.358137] Averaged stats: lr: 0.000551 loss: 0.8949 (0.9285) +[14:05:47.148529] val: [ 0/13] eta: 0:00:23 loss: 0.6214 (0.6214) time: 1.7773 data: 1.7494 max mem: 9672 +[14:05:48.869326] val: [10/13] eta: 0:00:00 loss: 0.7783 (0.7711) time: 0.3179 data: 0.2837 max mem: 9672 +[14:05:48.925753] val: [12/13] eta: 0:00:00 loss: 0.7783 (0.8381) time: 0.2733 data: 0.2401 max mem: 9672 +[14:05:49.004854] val: Total time: 0:00:03 (0.2796 s / it) +[14:05:49.020918] val loss: 0.8380820384392371 +[14:05:49.021182] Accuracy: 0.7000, F1 Score: 0.6071, ROC AUC: 0.8990, Hamming Loss: 0.1200, + Jaccard Score: 0.4536, Precision: 0.6236, Recall: 0.6114, + Average Precision: 0.6609, Kappa: 0.5797, Score: 0.6953 +[14:05:49.061459] Best epoch = 17, Best score = 0.6981 +[14:05:49.304401] log_dir: ./output_logs/retfound +[14:05:51.201099] Epoch: [19] [ 0/37] eta: 0:01:10 lr: 0.000550 loss: 0.9944 (0.9944) time: 1.8954 data: 1.7471 max mem: 9672 +[14:05:54.061718] Epoch: [19] [20/37] eta: 0:00:03 lr: 0.000541 loss: 0.8783 (0.9033) time: 0.1430 data: 0.0002 max mem: 9672 +[14:05:56.355255] Epoch: [19] [36/37] eta: 0:00:00 lr: 0.000534 loss: 0.9156 (0.9343) time: 0.1433 data: 0.0001 max mem: 9672 +[14:05:56.441691] Epoch: [19] Total time: 0:00:07 (0.1929 s / it) +[14:05:56.450248] Averaged stats: lr: 0.000534 loss: 0.9156 (0.9343) +[14:05:58.447562] val: [ 0/13] eta: 0:00:25 loss: 0.5418 (0.5418) time: 1.9848 data: 1.9484 max mem: 9672 +[14:06:00.121243] val: [10/13] eta: 0:00:00 loss: 0.7241 (0.7438) time: 0.3325 data: 0.2980 max mem: 9672 +[14:06:00.171810] val: [12/13] eta: 0:00:00 loss: 0.7241 (0.8051) time: 0.2852 data: 0.2522 max mem: 9672 +[14:06:00.246508] val: Total time: 0:00:03 (0.2911 s / it) +[14:06:00.263066] val loss: 0.8050662554227389 +[14:06:00.263270] Accuracy: 0.6950, F1 Score: 0.6100, ROC AUC: 0.9030, Hamming Loss: 0.1220, + Jaccard Score: 0.4533, Precision: 0.6502, Recall: 0.6067, + Average Precision: 0.6601, Kappa: 0.5730, Score: 0.6954 +[14:06:00.301388] Best epoch = 17, Best score = 0.6981 +[14:06:00.543320] log_dir: ./output_logs/retfound +[14:06:02.451182] Epoch: [20] [ 0/37] eta: 0:01:10 lr: 0.000534 loss: 0.9398 (0.9398) time: 1.9066 data: 1.7595 max mem: 9672 +[14:06:05.321754] Epoch: [20] [20/37] eta: 0:00:03 lr: 0.000524 loss: 0.9581 (0.9747) time: 0.1435 data: 0.0002 max mem: 9672 +[14:06:07.614383] Epoch: [20] [36/37] eta: 0:00:00 lr: 0.000516 loss: 0.9048 (0.9504) time: 0.1434 data: 0.0001 max mem: 9672 +[14:06:07.696627] Epoch: [20] Total time: 0:00:07 (0.1933 s / it) +[14:06:07.705038] Averaged stats: lr: 0.000516 loss: 0.9048 (0.9504) +[14:06:09.747890] val: [ 0/13] eta: 0:00:26 loss: 0.5241 (0.5241) time: 2.0302 data: 1.9948 max mem: 9672 +[14:06:11.284259] val: [10/13] eta: 0:00:00 loss: 0.9172 (0.7345) time: 0.3242 data: 0.2902 max mem: 9672 +[14:06:11.382738] val: [12/13] eta: 0:00:00 loss: 0.9172 (0.8401) time: 0.2818 data: 0.2487 max mem: 9672 +[14:06:11.458094] val: Total time: 0:00:03 (0.2878 s / it) +[14:06:11.474417] val loss: 0.8401335202730619 +[14:06:11.474625] Accuracy: 0.7000, F1 Score: 0.5887, ROC AUC: 0.8995, Hamming Loss: 0.1200, + Jaccard Score: 0.4403, Precision: 0.6503, Recall: 0.5797, + Average Precision: 0.6580, Kappa: 0.5743, Score: 0.6875 +[14:06:11.520039] Best epoch = 17, Best score = 0.6981 +[14:06:11.772064] log_dir: ./output_logs/retfound +[14:06:13.765917] Epoch: [21] [ 0/37] eta: 0:01:13 lr: 0.000516 loss: 0.7251 (0.7251) time: 1.9928 data: 1.8474 max mem: 9672 +[14:06:16.632438] Epoch: [21] [20/37] eta: 0:00:03 lr: 0.000505 loss: 0.9665 (0.9407) time: 0.1433 data: 0.0002 max mem: 9672 +[14:06:18.925452] Epoch: [21] [36/37] eta: 0:00:00 lr: 0.000497 loss: 0.8323 (0.9013) time: 0.1431 data: 0.0001 max mem: 9672 +[14:06:19.010788] Epoch: [21] Total time: 0:00:07 (0.1956 s / it) +[14:06:19.019543] Averaged stats: lr: 0.000497 loss: 0.8323 (0.9013) +[14:06:20.996421] val: [ 0/13] eta: 0:00:25 loss: 0.5092 (0.5092) time: 1.9683 data: 1.9319 max mem: 9672 +[14:06:22.449614] val: [10/13] eta: 0:00:00 loss: 0.8189 (0.7422) time: 0.3110 data: 0.2760 max mem: 9672 +[14:06:22.726220] val: [12/13] eta: 0:00:00 loss: 0.8189 (0.8426) time: 0.2844 data: 0.2505 max mem: 9672 +[14:06:22.804322] val: Total time: 0:00:03 (0.2905 s / it) +[14:06:22.820069] val loss: 0.8426446181077224 +[14:06:22.820270] Accuracy: 0.6925, F1 Score: 0.5925, ROC AUC: 0.9047, Hamming Loss: 0.1230, + Jaccard Score: 0.4420, Precision: 0.6347, Recall: 0.5924, + Average Precision: 0.6815, Kappa: 0.5715, Score: 0.6896 +[14:06:22.864342] Best epoch = 17, Best score = 0.6981 +[14:06:23.110992] log_dir: ./output_logs/retfound +[14:06:24.962098] Epoch: [22] [ 0/37] eta: 0:01:08 lr: 0.000496 loss: 0.6942 (0.6942) time: 1.8501 data: 1.7062 max mem: 9672 +[14:06:27.820483] Epoch: [22] [20/37] eta: 0:00:03 lr: 0.000486 loss: 0.8678 (0.8940) time: 0.1429 data: 0.0002 max mem: 9672 +[14:06:30.117675] Epoch: [22] [36/37] eta: 0:00:00 lr: 0.000477 loss: 0.8802 (0.9049) time: 0.1432 data: 0.0001 max mem: 9672 +[14:06:30.198116] Epoch: [22] Total time: 0:00:07 (0.1915 s / it) +[14:06:30.206865] Averaged stats: lr: 0.000477 loss: 0.8802 (0.9049) +[14:06:32.330043] val: [ 0/13] eta: 0:00:27 loss: 0.5761 (0.5761) time: 2.1154 data: 2.0800 max mem: 9672 +[14:06:33.840854] val: [10/13] eta: 0:00:00 loss: 0.6778 (0.7629) time: 0.3296 data: 0.2948 max mem: 9672 +[14:06:33.901675] val: [12/13] eta: 0:00:00 loss: 0.6778 (0.8607) time: 0.2835 data: 0.2498 max mem: 9672 +[14:06:33.982445] val: Total time: 0:00:03 (0.2899 s / it) +[14:06:33.998348] val loss: 0.8606914648642907 +[14:06:33.998550] Accuracy: 0.6975, F1 Score: 0.5794, ROC AUC: 0.8979, Hamming Loss: 0.1210, + Jaccard Score: 0.4313, Precision: 0.6260, Recall: 0.5637, + Average Precision: 0.6554, Kappa: 0.5646, Score: 0.6806 +[14:06:34.044080] Best epoch = 17, Best score = 0.6981 +[14:06:34.286382] log_dir: ./output_logs/retfound +[14:06:36.118104] Epoch: [23] [ 0/37] eta: 0:01:07 lr: 0.000476 loss: 0.8260 (0.8260) time: 1.8307 data: 1.6836 max mem: 9672 +[14:06:38.981128] Epoch: [23] [20/37] eta: 0:00:03 lr: 0.000465 loss: 0.8332 (0.8214) time: 0.1431 data: 0.0002 max mem: 9672 +[14:06:41.275071] Epoch: [23] [36/37] eta: 0:00:00 lr: 0.000455 loss: 0.9324 (0.8897) time: 0.1435 data: 0.0001 max mem: 9672 +[14:06:41.363907] Epoch: [23] Total time: 0:00:07 (0.1913 s / it) +[14:06:41.371891] Averaged stats: lr: 0.000455 loss: 0.9324 (0.8897) +[14:06:43.344888] val: [ 0/13] eta: 0:00:25 loss: 0.5752 (0.5752) time: 1.9553 data: 1.9195 max mem: 9672 +[14:06:44.939933] val: [10/13] eta: 0:00:00 loss: 0.9839 (0.8322) time: 0.3227 data: 0.2879 max mem: 9672 +[14:06:45.027769] val: [12/13] eta: 0:00:00 loss: 0.9839 (0.8864) time: 0.2797 data: 0.2459 max mem: 9672 +[14:06:45.105188] val: Total time: 0:00:03 (0.2859 s / it) +[14:06:45.120915] val loss: 0.8863991590646597 +[14:06:45.121184] Accuracy: 0.6600, F1 Score: 0.5527, ROC AUC: 0.8929, Hamming Loss: 0.1360, + Jaccard Score: 0.4002, Precision: 0.5657, Recall: 0.5572, + Average Precision: 0.6516, Kappa: 0.5157, Score: 0.6538 +[14:06:45.166348] Best epoch = 17, Best score = 0.6981 +[14:06:45.396090] log_dir: ./output_logs/retfound +[14:06:47.293125] Epoch: [24] [ 0/37] eta: 0:01:10 lr: 0.000455 loss: 0.9915 (0.9915) time: 1.8955 data: 1.7472 max mem: 9672 +[14:06:50.163321] Epoch: [24] [20/37] eta: 0:00:03 lr: 0.000443 loss: 0.9178 (0.9001) time: 0.1434 data: 0.0002 max mem: 9672 +[14:06:52.454527] Epoch: [24] [36/37] eta: 0:00:00 lr: 0.000433 loss: 0.9092 (0.9006) time: 0.1433 data: 0.0001 max mem: 9672 +[14:06:52.539284] Epoch: [24] Total time: 0:00:07 (0.1931 s / it) +[14:06:52.548735] Averaged stats: lr: 0.000433 loss: 0.9092 (0.9006) +[14:06:54.597837] val: [ 0/13] eta: 0:00:26 loss: 0.4372 (0.4372) time: 2.0324 data: 1.9981 max mem: 9672 +[14:06:56.187069] val: [10/13] eta: 0:00:00 loss: 0.7616 (0.7447) time: 0.3292 data: 0.2945 max mem: 9672 +[14:06:56.501462] val: [12/13] eta: 0:00:00 loss: 0.7616 (0.8642) time: 0.3027 data: 0.2690 max mem: 9672 +[14:06:56.581410] val: Total time: 0:00:04 (0.3090 s / it) +[14:06:56.597826] val loss: 0.864152715756343 +[14:06:56.598117] Accuracy: 0.7125, F1 Score: 0.6013, ROC AUC: 0.9033, Hamming Loss: 0.1150, + Jaccard Score: 0.4557, Precision: 0.6448, Recall: 0.5824, + Average Precision: 0.6760, Kappa: 0.5865, Score: 0.6970 +[14:06:56.650394] Best epoch = 17, Best score = 0.6981 +[14:06:56.898238] log_dir: ./output_logs/retfound +[14:06:58.696788] Epoch: [25] [ 0/37] eta: 0:01:06 lr: 0.000432 loss: 0.8581 (0.8581) time: 1.7975 data: 1.7097 max mem: 9672 +[14:07:01.480494] Epoch: [25] [20/37] eta: 0:00:03 lr: 0.000420 loss: 0.8584 (0.8732) time: 0.1391 data: 0.0185 max mem: 9672 +[14:07:03.763499] Epoch: [25] [36/37] eta: 0:00:00 lr: 0.000410 loss: 0.9064 (0.8944) time: 0.1431 data: 0.0001 max mem: 9672 +[14:07:03.841720] Epoch: [25] Total time: 0:00:06 (0.1877 s / it) +[14:07:03.851281] Averaged stats: lr: 0.000410 loss: 0.9064 (0.8944) +[14:07:05.994219] val: [ 0/13] eta: 0:00:27 loss: 0.5810 (0.5810) time: 2.1294 data: 2.0952 max mem: 9672 +[14:07:07.547218] val: [10/13] eta: 0:00:01 loss: 0.9300 (0.8279) time: 0.3347 data: 0.3000 max mem: 9672 +[14:07:07.645855] val: [12/13] eta: 0:00:00 loss: 0.9300 (0.8688) time: 0.2907 data: 0.2571 max mem: 9672 +[14:07:07.726749] val: Total time: 0:00:03 (0.2971 s / it) +[14:07:07.742358] val loss: 0.868835618862739 +[14:07:07.742667] Accuracy: 0.6625, F1 Score: 0.5769, ROC AUC: 0.8937, Hamming Loss: 0.1350, + Jaccard Score: 0.4204, Precision: 0.5730, Recall: 0.5925, + Average Precision: 0.6437, Kappa: 0.5301, Score: 0.6669 +[14:07:07.788908] Best epoch = 17, Best score = 0.6981 +[14:07:08.009261] log_dir: ./output_logs/retfound +[14:07:09.980821] Epoch: [26] [ 0/37] eta: 0:01:12 lr: 0.000409 loss: 0.7798 (0.7798) time: 1.9698 data: 1.8251 max mem: 9672 +[14:07:12.840027] Epoch: [26] [20/37] eta: 0:00:03 lr: 0.000397 loss: 0.8075 (0.8634) time: 0.1429 data: 0.0001 max mem: 9672 +[14:07:15.134006] Epoch: [26] [36/37] eta: 0:00:00 lr: 0.000386 loss: 0.8487 (0.8588) time: 0.1434 data: 0.0001 max mem: 9672 +[14:07:15.216060] Epoch: [26] Total time: 0:00:07 (0.1948 s / it) +[14:07:15.225909] Averaged stats: lr: 0.000386 loss: 0.8487 (0.8588) +[14:07:17.292479] val: [ 0/13] eta: 0:00:26 loss: 0.4989 (0.4989) time: 2.0499 data: 2.0132 max mem: 9672 +[14:07:18.807468] val: [10/13] eta: 0:00:00 loss: 1.0650 (0.8567) time: 0.3240 data: 0.2897 max mem: 9672 +[14:07:19.011265] val: [12/13] eta: 0:00:00 loss: 1.0650 (0.9205) time: 0.2898 data: 0.2565 max mem: 9672 +[14:07:19.093351] val: Total time: 0:00:03 (0.2963 s / it) +[14:07:19.119887] val loss: 0.920516325877263 +[14:07:19.120111] Accuracy: 0.6600, F1 Score: 0.5360, ROC AUC: 0.8962, Hamming Loss: 0.1360, + Jaccard Score: 0.3872, Precision: 0.5754, Recall: 0.5393, + Average Precision: 0.6332, Kappa: 0.5084, Score: 0.6469 +[14:07:19.172060] Best epoch = 17, Best score = 0.6981 +[14:07:19.414502] log_dir: ./output_logs/retfound +[14:07:21.212821] Epoch: [27] [ 0/37] eta: 0:01:06 lr: 0.000386 loss: 1.1308 (1.1308) time: 1.7972 data: 1.6488 max mem: 9672 +[14:07:24.079341] Epoch: [27] [20/37] eta: 0:00:03 lr: 0.000373 loss: 0.8274 (0.8282) time: 0.1433 data: 0.0002 max mem: 9672 +[14:07:26.367746] Epoch: [27] [36/37] eta: 0:00:00 lr: 0.000362 loss: 0.8816 (0.8648) time: 0.1429 data: 0.0001 max mem: 9672 +[14:07:26.452246] Epoch: [27] Total time: 0:00:07 (0.1902 s / it) +[14:07:26.460393] Averaged stats: lr: 0.000362 loss: 0.8816 (0.8648) +[14:07:28.601567] val: [ 0/13] eta: 0:00:27 loss: 0.7022 (0.7022) time: 2.1247 data: 2.0894 max mem: 9672 +[14:07:30.143345] val: [10/13] eta: 0:00:00 loss: 0.6240 (0.7898) time: 0.3332 data: 0.2986 max mem: 9672 +[14:07:30.353300] val: [12/13] eta: 0:00:00 loss: 0.7022 (0.8833) time: 0.2981 data: 0.2644 max mem: 9672 +[14:07:30.431108] val: Total time: 0:00:03 (0.3042 s / it) +[14:07:30.446939] val loss: 0.8833090846355145 +[14:07:30.447152] Accuracy: 0.7025, F1 Score: 0.5877, ROC AUC: 0.9031, Hamming Loss: 0.1190, + Jaccard Score: 0.4365, Precision: 0.6738, Recall: 0.5707, + Average Precision: 0.6689, Kappa: 0.5719, Score: 0.6876 +[14:07:30.498260] Best epoch = 17, Best score = 0.6981 +[14:07:30.750853] log_dir: ./output_logs/retfound +[14:07:32.502692] Epoch: [28] [ 0/37] eta: 0:01:04 lr: 0.000362 loss: 0.8730 (0.8730) time: 1.7508 data: 1.6059 max mem: 9672 +[14:07:35.359212] Epoch: [28] [20/37] eta: 0:00:03 lr: 0.000349 loss: 0.8459 (0.8680) time: 0.1428 data: 0.0002 max mem: 9672 +[14:07:37.655315] Epoch: [28] [36/37] eta: 0:00:00 lr: 0.000338 loss: 0.9051 (0.8947) time: 0.1433 data: 0.0001 max mem: 9672 +[14:07:37.732897] Epoch: [28] Total time: 0:00:06 (0.1887 s / it) +[14:07:37.733850] Averaged stats: lr: 0.000338 loss: 0.9051 (0.8947) +[14:07:39.639343] val: [ 0/13] eta: 0:00:24 loss: 0.7128 (0.7128) time: 1.8888 data: 1.8546 max mem: 9672 +[14:07:41.318043] val: [10/13] eta: 0:00:00 loss: 0.7128 (0.8253) time: 0.3243 data: 0.2896 max mem: 9672 +[14:07:41.374206] val: [12/13] eta: 0:00:00 loss: 0.7625 (0.8810) time: 0.2786 data: 0.2451 max mem: 9672 +[14:07:41.447416] val: Total time: 0:00:03 (0.2844 s / it) +[14:07:41.468134] val loss: 0.8810019951600295 +[14:07:41.468352] Accuracy: 0.6900, F1 Score: 0.6036, ROC AUC: 0.8977, Hamming Loss: 0.1240, + Jaccard Score: 0.4472, Precision: 0.6294, Recall: 0.6025, + Average Precision: 0.6600, Kappa: 0.5646, Score: 0.6886 +[14:07:41.516001] Best epoch = 17, Best score = 0.6981 +[14:07:41.746828] log_dir: ./output_logs/retfound +[14:07:43.588523] Epoch: [29] [ 0/37] eta: 0:01:08 lr: 0.000337 loss: 1.0053 (1.0053) time: 1.8405 data: 1.6951 max mem: 9672 +[14:07:46.450027] Epoch: [29] [20/37] eta: 0:00:03 lr: 0.000324 loss: 0.8402 (0.8374) time: 0.1430 data: 0.0001 max mem: 9672 +[14:07:48.742175] Epoch: [29] [36/37] eta: 0:00:00 lr: 0.000314 loss: 0.8376 (0.8424) time: 0.1428 data: 0.0001 max mem: 9672 +[14:07:48.822528] Epoch: [29] Total time: 0:00:07 (0.1912 s / it) +[14:07:48.830600] Averaged stats: lr: 0.000314 loss: 0.8376 (0.8424) +[14:07:50.705848] val: [ 0/13] eta: 0:00:24 loss: 0.7090 (0.7090) time: 1.8578 data: 1.8224 max mem: 9672 +[14:07:52.327805] val: [10/13] eta: 0:00:00 loss: 0.6683 (0.7898) time: 0.3163 data: 0.2819 max mem: 9672 +[14:07:52.383287] val: [12/13] eta: 0:00:00 loss: 0.7090 (0.8746) time: 0.2718 data: 0.2386 max mem: 9672 +[14:07:52.455475] val: Total time: 0:00:03 (0.2776 s / it) +[14:07:52.471406] val loss: 0.8745717452122614 +[14:07:52.471685] Accuracy: 0.6975, F1 Score: 0.6186, ROC AUC: 0.9007, Hamming Loss: 0.1210, + Jaccard Score: 0.4615, Precision: 0.6452, Recall: 0.6161, + Average Precision: 0.6672, Kappa: 0.5766, Score: 0.6986 +[14:07:54.377421] Best epoch = 29, Best score = 0.6986 +[14:07:54.456926] log_dir: ./output_logs/retfound +[14:07:56.292270] Epoch: [30] [ 0/37] eta: 0:01:07 lr: 0.000313 loss: 0.7726 (0.7726) time: 1.8339 data: 1.6886 max mem: 9672 +[14:07:59.156750] Epoch: [30] [20/37] eta: 0:00:03 lr: 0.000300 loss: 0.8179 (0.8489) time: 0.1432 data: 0.0002 max mem: 9672 +[14:08:01.452835] Epoch: [30] [36/37] eta: 0:00:00 lr: 0.000289 loss: 0.8498 (0.8773) time: 0.1434 data: 0.0001 max mem: 9672 +[14:08:01.531466] Epoch: [30] Total time: 0:00:07 (0.1912 s / it) +[14:08:01.541017] Averaged stats: lr: 0.000289 loss: 0.8498 (0.8773) +[14:08:03.709836] val: [ 0/13] eta: 0:00:27 loss: 0.5845 (0.5845) time: 2.1523 data: 2.1176 max mem: 9672 +[14:08:05.478390] val: [10/13] eta: 0:00:01 loss: 0.7456 (0.7818) time: 0.3564 data: 0.3215 max mem: 9672 +[14:08:05.534972] val: [12/13] eta: 0:00:00 loss: 0.7456 (0.8608) time: 0.3058 data: 0.2721 max mem: 9672 +[14:08:05.613750] val: Total time: 0:00:04 (0.3121 s / it) +[14:08:05.642513] val loss: 0.8607793404505804 +[14:08:05.642783] Accuracy: 0.6900, F1 Score: 0.6072, ROC AUC: 0.9016, Hamming Loss: 0.1240, + Jaccard Score: 0.4502, Precision: 0.6043, Recall: 0.6116, + Average Precision: 0.6580, Kappa: 0.5681, Score: 0.6923 +[14:08:05.688361] Best epoch = 29, Best score = 0.6986 +[14:08:05.929383] log_dir: ./output_logs/retfound +[14:08:07.855188] Epoch: [31] [ 0/37] eta: 0:01:11 lr: 0.000289 loss: 0.8747 (0.8747) time: 1.9246 data: 1.7752 max mem: 9672 +[14:08:10.720820] Epoch: [31] [20/37] eta: 0:00:03 lr: 0.000275 loss: 0.7811 (0.8106) time: 0.1432 data: 0.0002 max mem: 9672 +[14:08:13.007440] Epoch: [31] [36/37] eta: 0:00:00 lr: 0.000265 loss: 0.7550 (0.8003) time: 0.1427 data: 0.0001 max mem: 9672 +[14:08:13.088949] Epoch: [31] Total time: 0:00:07 (0.1935 s / it) +[14:08:13.098071] Averaged stats: lr: 0.000265 loss: 0.7550 (0.8003) +[14:08:15.267599] val: [ 0/13] eta: 0:00:27 loss: 0.5363 (0.5363) time: 2.1513 data: 2.1154 max mem: 9672 +[14:08:16.880784] val: [10/13] eta: 0:00:01 loss: 0.7788 (0.8130) time: 0.3422 data: 0.3072 max mem: 9672 +[14:08:16.936898] val: [12/13] eta: 0:00:00 loss: 0.7788 (0.9120) time: 0.2938 data: 0.2599 max mem: 9672 +[14:08:17.016981] val: Total time: 0:00:03 (0.3001 s / it) +[14:08:17.034558] val loss: 0.9119756038372333 +[14:08:17.034786] Accuracy: 0.7125, F1 Score: 0.5900, ROC AUC: 0.9037, Hamming Loss: 0.1150, + Jaccard Score: 0.4431, Precision: 0.6404, Recall: 0.5708, + Average Precision: 0.6598, Kappa: 0.5811, Score: 0.6916 +[14:08:17.079568] Best epoch = 29, Best score = 0.6986 +[14:08:17.309923] log_dir: ./output_logs/retfound +[14:08:19.183639] Epoch: [32] [ 0/37] eta: 0:01:09 lr: 0.000264 loss: 0.9246 (0.9246) time: 1.8726 data: 1.7252 max mem: 9672 +[14:08:22.045564] Epoch: [32] [20/37] eta: 0:00:03 lr: 0.000251 loss: 0.8607 (0.8576) time: 0.1430 data: 0.0002 max mem: 9672 +[14:08:24.338830] Epoch: [32] [36/37] eta: 0:00:00 lr: 0.000241 loss: 0.8066 (0.8226) time: 0.1432 data: 0.0001 max mem: 9672 +[14:08:24.413996] Epoch: [32] Total time: 0:00:07 (0.1920 s / it) +[14:08:24.423161] Averaged stats: lr: 0.000241 loss: 0.8066 (0.8226) +[14:08:26.305756] val: [ 0/13] eta: 0:00:24 loss: 0.4929 (0.4929) time: 1.8636 data: 1.8290 max mem: 9672 +[14:08:27.938568] val: [10/13] eta: 0:00:00 loss: 0.8407 (0.8517) time: 0.3178 data: 0.2831 max mem: 9672 +[14:08:28.068058] val: [12/13] eta: 0:00:00 loss: 0.8407 (0.9347) time: 0.2788 data: 0.2452 max mem: 9672 +[14:08:28.147166] val: Total time: 0:00:03 (0.2850 s / it) +[14:08:28.163748] val loss: 0.9346982836723328 +[14:08:28.164010] Accuracy: 0.6900, F1 Score: 0.5895, ROC AUC: 0.9002, Hamming Loss: 0.1240, + Jaccard Score: 0.4360, Precision: 0.6031, Recall: 0.5803, + Average Precision: 0.6570, Kappa: 0.5572, Score: 0.6823 +[14:08:28.207404] Best epoch = 29, Best score = 0.6986 +[14:08:28.429944] log_dir: ./output_logs/retfound +[14:08:30.367718] Epoch: [33] [ 0/37] eta: 0:01:11 lr: 0.000240 loss: 0.6998 (0.6998) time: 1.9367 data: 1.7902 max mem: 9672 +[14:08:33.224160] Epoch: [33] [20/37] eta: 0:00:03 lr: 0.000227 loss: 0.7801 (0.7587) time: 0.1428 data: 0.0002 max mem: 9672 +[14:08:35.509043] Epoch: [33] [36/37] eta: 0:00:00 lr: 0.000217 loss: 0.7871 (0.7866) time: 0.1430 data: 0.0001 max mem: 9672 +[14:08:35.588308] Epoch: [33] Total time: 0:00:07 (0.1935 s / it) +[14:08:35.597382] Averaged stats: lr: 0.000217 loss: 0.7871 (0.7866) +[14:08:37.622355] val: [ 0/13] eta: 0:00:26 loss: 0.5458 (0.5458) time: 2.0169 data: 2.0001 max mem: 9672 +[14:08:39.331261] val: [10/13] eta: 0:00:01 loss: 0.9234 (0.8165) time: 0.3386 data: 0.3141 max mem: 9672 +[14:08:39.387722] val: [12/13] eta: 0:00:00 loss: 0.9234 (0.9310) time: 0.2908 data: 0.2658 max mem: 9672 +[14:08:39.464157] val: Total time: 0:00:03 (0.2969 s / it) +[14:08:39.480129] val loss: 0.9309894305009109 +[14:08:39.480326] Accuracy: 0.7000, F1 Score: 0.5983, ROC AUC: 0.9002, Hamming Loss: 0.1200, + Jaccard Score: 0.4493, Precision: 0.6094, Recall: 0.5992, + Average Precision: 0.6470, Kappa: 0.5807, Score: 0.6931 +[14:08:39.524287] Best epoch = 29, Best score = 0.6986 +[14:08:39.776784] log_dir: ./output_logs/retfound +[14:08:41.701903] Epoch: [34] [ 0/37] eta: 0:01:11 lr: 0.000217 loss: 0.7521 (0.7521) time: 1.9239 data: 1.7728 max mem: 9672 +[14:08:44.566730] Epoch: [34] [20/37] eta: 0:00:03 lr: 0.000204 loss: 0.7812 (0.8051) time: 0.1432 data: 0.0002 max mem: 9672 +[14:08:46.856029] Epoch: [34] [36/37] eta: 0:00:00 lr: 0.000194 loss: 0.8021 (0.8329) time: 0.1430 data: 0.0001 max mem: 9672 +[14:08:46.934520] Epoch: [34] Total time: 0:00:07 (0.1934 s / it) +[14:08:46.942600] Averaged stats: lr: 0.000194 loss: 0.8021 (0.8329) +[14:08:48.901488] val: [ 0/13] eta: 0:00:25 loss: 0.6481 (0.6481) time: 1.9470 data: 1.9123 max mem: 9672 +[14:08:50.545435] val: [10/13] eta: 0:00:00 loss: 0.8007 (0.7811) time: 0.3264 data: 0.2917 max mem: 9672 +[14:08:50.629258] val: [12/13] eta: 0:00:00 loss: 0.8007 (0.8781) time: 0.2826 data: 0.2490 max mem: 9672 +[14:08:50.704518] val: Total time: 0:00:03 (0.2885 s / it) +[14:08:50.720175] val loss: 0.8780909501589261 +[14:08:50.720365] Accuracy: 0.6900, F1 Score: 0.5781, ROC AUC: 0.9019, Hamming Loss: 0.1240, + Jaccard Score: 0.4329, Precision: 0.5851, Recall: 0.5736, + Average Precision: 0.6475, Kappa: 0.5638, Score: 0.6813 +[14:08:50.766594] Best epoch = 29, Best score = 0.6986 +[14:08:51.013936] log_dir: ./output_logs/retfound +[14:08:52.791885] Epoch: [35] [ 0/37] eta: 0:01:05 lr: 0.000194 loss: 0.8176 (0.8176) time: 1.7768 data: 1.6300 max mem: 9672 +[14:08:55.652657] Epoch: [35] [20/37] eta: 0:00:03 lr: 0.000181 loss: 0.7531 (0.7788) time: 0.1430 data: 0.0001 max mem: 9672 +[14:08:57.937284] Epoch: [35] [36/37] eta: 0:00:00 lr: 0.000172 loss: 0.8333 (0.7969) time: 0.1428 data: 0.0001 max mem: 9672 +[14:08:58.019494] Epoch: [35] Total time: 0:00:07 (0.1893 s / it) +[14:08:58.028045] Averaged stats: lr: 0.000172 loss: 0.8333 (0.7969) +[14:09:00.032748] val: [ 0/13] eta: 0:00:25 loss: 0.7637 (0.7637) time: 1.9879 data: 1.9527 max mem: 9672 +[14:09:01.677724] val: [10/13] eta: 0:00:00 loss: 0.7637 (0.7914) time: 0.3302 data: 0.2954 max mem: 9672 +[14:09:01.733571] val: [12/13] eta: 0:00:00 loss: 0.7637 (0.8805) time: 0.2836 data: 0.2500 max mem: 9672 +[14:09:01.814328] val: Total time: 0:00:03 (0.2900 s / it) +[14:09:01.830088] val loss: 0.8805368542671204 +[14:09:01.830277] Accuracy: 0.6900, F1 Score: 0.5809, ROC AUC: 0.9039, Hamming Loss: 0.1240, + Jaccard Score: 0.4333, Precision: 0.5859, Recall: 0.5802, + Average Precision: 0.6557, Kappa: 0.5655, Score: 0.6834 +[14:09:01.865808] Best epoch = 29, Best score = 0.6986 +[14:09:02.130843] log_dir: ./output_logs/retfound +[14:09:04.111655] Epoch: [36] [ 0/37] eta: 0:01:13 lr: 0.000171 loss: 0.7475 (0.7475) time: 1.9798 data: 1.8363 max mem: 9672 +[14:09:06.974432] Epoch: [36] [20/37] eta: 0:00:03 lr: 0.000160 loss: 0.7808 (0.7936) time: 0.1431 data: 0.0002 max mem: 9672 +[14:09:09.262003] Epoch: [36] [36/37] eta: 0:00:00 lr: 0.000151 loss: 0.7835 (0.8122) time: 0.1428 data: 0.0001 max mem: 9672 +[14:09:09.337056] Epoch: [36] Total time: 0:00:07 (0.1948 s / it) +[14:09:09.345547] Averaged stats: lr: 0.000151 loss: 0.7835 (0.8122) +[14:09:11.267571] val: [ 0/13] eta: 0:00:24 loss: 0.5683 (0.5683) time: 1.9051 data: 1.8736 max mem: 9672 +[14:09:12.832233] val: [10/13] eta: 0:00:00 loss: 0.8524 (0.8058) time: 0.3154 data: 0.2808 max mem: 9672 +[14:09:12.902268] val: [12/13] eta: 0:00:00 loss: 0.8661 (0.8798) time: 0.2722 data: 0.2390 max mem: 9672 +[14:09:12.980049] val: Total time: 0:00:03 (0.2783 s / it) +[14:09:13.003611] val loss: 0.8798409883792584 +[14:09:13.003818] Accuracy: 0.7150, F1 Score: 0.6305, ROC AUC: 0.9037, Hamming Loss: 0.1140, + Jaccard Score: 0.4777, Precision: 0.6410, Recall: 0.6336, + Average Precision: 0.6593, Kappa: 0.6023, Score: 0.7122 +[14:09:14.781449] Best epoch = 36, Best score = 0.7122 +[14:09:14.846776] log_dir: ./output_logs/retfound +[14:09:16.694339] Epoch: [37] [ 0/37] eta: 0:01:08 lr: 0.000150 loss: 0.6358 (0.6358) time: 1.8464 data: 1.6989 max mem: 9672 +[14:09:19.556660] Epoch: [37] [20/37] eta: 0:00:03 lr: 0.000139 loss: 0.7702 (0.7981) time: 0.1431 data: 0.0002 max mem: 9672 +[14:09:21.844331] Epoch: [37] [36/37] eta: 0:00:00 lr: 0.000130 loss: 0.7207 (0.8003) time: 0.1427 data: 0.0001 max mem: 9672 +[14:09:21.915386] Epoch: [37] Total time: 0:00:07 (0.1910 s / it) +[14:09:21.924551] Averaged stats: lr: 0.000130 loss: 0.7207 (0.8003) +[14:09:24.025537] val: [ 0/13] eta: 0:00:27 loss: 0.5637 (0.5637) time: 2.0845 data: 2.0493 max mem: 9672 +[14:09:25.640899] val: [10/13] eta: 0:00:01 loss: 0.8017 (0.7758) time: 0.3363 data: 0.3016 max mem: 9672 +[14:09:25.696689] val: [12/13] eta: 0:00:00 loss: 0.8017 (0.8638) time: 0.2888 data: 0.2552 max mem: 9672 +[14:09:25.772173] val: Total time: 0:00:03 (0.2947 s / it) +[14:09:25.789645] val loss: 0.8637872292445257 +[14:09:25.789984] Accuracy: 0.7200, F1 Score: 0.6261, ROC AUC: 0.9066, Hamming Loss: 0.1120, + Jaccard Score: 0.4741, Precision: 0.6402, Recall: 0.6185, + Average Precision: 0.6669, Kappa: 0.6048, Score: 0.7125 +[14:09:27.591628] Best epoch = 37, Best score = 0.7125 +[14:09:27.659610] log_dir: ./output_logs/retfound +[14:09:29.459903] Epoch: [38] [ 0/37] eta: 0:01:06 lr: 0.000130 loss: 0.6356 (0.6356) time: 1.7993 data: 1.6532 max mem: 9672 +[14:09:32.322606] Epoch: [38] [20/37] eta: 0:00:03 lr: 0.000119 loss: 0.7571 (0.7513) time: 0.1431 data: 0.0002 max mem: 9672 +[14:09:34.606751] Epoch: [38] [36/37] eta: 0:00:00 lr: 0.000111 loss: 0.7688 (0.7887) time: 0.1427 data: 0.0001 max mem: 9672 +[14:09:34.681159] Epoch: [38] Total time: 0:00:07 (0.1898 s / it) +[14:09:34.687329] Averaged stats: lr: 0.000111 loss: 0.7688 (0.7887) +[14:09:36.560620] val: [ 0/13] eta: 0:00:24 loss: 0.6333 (0.6333) time: 1.8575 data: 1.8209 max mem: 9672 +[14:09:38.241740] val: [10/13] eta: 0:00:00 loss: 0.7610 (0.7924) time: 0.3216 data: 0.2870 max mem: 9672 +[14:09:38.298467] val: [12/13] eta: 0:00:00 loss: 0.8057 (0.8765) time: 0.2765 data: 0.2429 max mem: 9672 +[14:09:38.371130] val: Total time: 0:00:03 (0.2822 s / it) +[14:09:38.387765] val loss: 0.8765294414300185 +[14:09:38.387957] Accuracy: 0.7050, F1 Score: 0.6079, ROC AUC: 0.9052, Hamming Loss: 0.1180, + Jaccard Score: 0.4554, Precision: 0.6223, Recall: 0.6017, + Average Precision: 0.6649, Kappa: 0.5835, Score: 0.6989 +[14:09:38.439209] Best epoch = 37, Best score = 0.7125 +[14:09:38.676491] log_dir: ./output_logs/retfound +[14:09:40.465134] Epoch: [39] [ 0/37] eta: 0:01:06 lr: 0.000110 loss: 0.7601 (0.7601) time: 1.7875 data: 1.6454 max mem: 9672 +[14:09:43.329657] Epoch: [39] [20/37] eta: 0:00:03 lr: 0.000100 loss: 0.8214 (0.8245) time: 0.1432 data: 0.0002 max mem: 9672 +[14:09:45.614787] Epoch: [39] [36/37] eta: 0:00:00 lr: 0.000093 loss: 0.7520 (0.7995) time: 0.1430 data: 0.0001 max mem: 9672 +[14:09:45.690165] Epoch: [39] Total time: 0:00:07 (0.1896 s / it) +[14:09:45.699281] Averaged stats: lr: 0.000093 loss: 0.7520 (0.7995) +[14:09:47.616719] val: [ 0/13] eta: 0:00:24 loss: 0.4143 (0.4143) time: 1.9006 data: 1.8660 max mem: 9672 +[14:09:49.181555] val: [10/13] eta: 0:00:00 loss: 0.9031 (0.8156) time: 0.3150 data: 0.2806 max mem: 9672 +[14:09:49.380512] val: [12/13] eta: 0:00:00 loss: 0.9031 (0.9051) time: 0.2818 data: 0.2483 max mem: 9672 +[14:09:49.453349] val: Total time: 0:00:03 (0.2875 s / it) +[14:09:49.469137] val loss: 0.9051186717473544 +[14:09:49.469379] Accuracy: 0.7000, F1 Score: 0.5854, ROC AUC: 0.9020, Hamming Loss: 0.1200, + Jaccard Score: 0.4349, Precision: 0.6190, Recall: 0.5680, + Average Precision: 0.6563, Kappa: 0.5626, Score: 0.6833 +[14:09:49.527553] Best epoch = 37, Best score = 0.7125 +[14:09:49.743683] log_dir: ./output_logs/retfound +[14:09:51.563873] Epoch: [40] [ 0/37] eta: 0:01:07 lr: 0.000092 loss: 0.7548 (0.7548) time: 1.8190 data: 1.6708 max mem: 9672 +[14:09:54.429278] Epoch: [40] [20/37] eta: 0:00:03 lr: 0.000083 loss: 0.7747 (0.7756) time: 0.1432 data: 0.0002 max mem: 9672 +[14:09:56.712649] Epoch: [40] [36/37] eta: 0:00:00 lr: 0.000076 loss: 0.7346 (0.7872) time: 0.1427 data: 0.0001 max mem: 9672 +[14:09:56.789106] Epoch: [40] Total time: 0:00:07 (0.1904 s / it) +[14:09:56.797132] Averaged stats: lr: 0.000076 loss: 0.7346 (0.7872) +[14:09:58.760410] val: [ 0/13] eta: 0:00:25 loss: 0.5215 (0.5215) time: 1.9474 data: 1.9109 max mem: 9672 +[14:10:00.361010] val: [10/13] eta: 0:00:00 loss: 0.8089 (0.7765) time: 0.3225 data: 0.2878 max mem: 9672 +[14:10:00.417437] val: [12/13] eta: 0:00:00 loss: 0.8089 (0.8759) time: 0.2772 data: 0.2436 max mem: 9672 +[14:10:00.492475] val: Total time: 0:00:03 (0.2831 s / it) +[14:10:00.508206] val loss: 0.8759457239737878 +[14:10:00.508418] Accuracy: 0.7200, F1 Score: 0.6271, ROC AUC: 0.9047, Hamming Loss: 0.1120, + Jaccard Score: 0.4746, Precision: 0.6369, Recall: 0.6197, + Average Precision: 0.6604, Kappa: 0.6043, Score: 0.7120 +[14:10:00.553905] Best epoch = 37, Best score = 0.7125 +[14:10:00.780430] log_dir: ./output_logs/retfound +[14:10:02.621064] Epoch: [41] [ 0/37] eta: 0:01:08 lr: 0.000076 loss: 0.7115 (0.7115) time: 1.8389 data: 1.6938 max mem: 9672 +[14:10:05.482914] Epoch: [41] [20/37] eta: 0:00:03 lr: 0.000067 loss: 0.7623 (0.7721) time: 0.1430 data: 0.0002 max mem: 9672 +[14:10:07.773204] Epoch: [41] [36/37] eta: 0:00:00 lr: 0.000061 loss: 0.7684 (0.7756) time: 0.1429 data: 0.0001 max mem: 9672 +[14:10:07.851635] Epoch: [41] Total time: 0:00:07 (0.1911 s / it) +[14:10:07.861479] Averaged stats: lr: 0.000061 loss: 0.7684 (0.7756) +[14:10:09.778460] val: [ 0/13] eta: 0:00:24 loss: 0.4754 (0.4754) time: 1.9064 data: 1.8713 max mem: 9672 +[14:10:11.518768] val: [10/13] eta: 0:00:00 loss: 0.8328 (0.8052) time: 0.3314 data: 0.2970 max mem: 9672 +[14:10:11.575281] val: [12/13] eta: 0:00:00 loss: 0.8328 (0.8971) time: 0.2848 data: 0.2514 max mem: 9672 +[14:10:11.648341] val: Total time: 0:00:03 (0.2905 s / it) +[14:10:11.664074] val loss: 0.8970959117779365 +[14:10:11.664308] Accuracy: 0.7050, F1 Score: 0.5963, ROC AUC: 0.9027, Hamming Loss: 0.1180, + Jaccard Score: 0.4449, Precision: 0.6233, Recall: 0.5800, + Average Precision: 0.6565, Kappa: 0.5743, Score: 0.6911 +[14:10:11.709407] Best epoch = 37, Best score = 0.7125 +[14:10:11.944913] log_dir: ./output_logs/retfound +[14:10:13.757805] Epoch: [42] [ 0/37] eta: 0:01:07 lr: 0.000061 loss: 0.5968 (0.5968) time: 1.8118 data: 1.6681 max mem: 9672 +[14:10:16.622766] Epoch: [42] [20/37] eta: 0:00:03 lr: 0.000053 loss: 0.7290 (0.7485) time: 0.1432 data: 0.0001 max mem: 9672 +[14:10:18.908659] Epoch: [42] [36/37] eta: 0:00:00 lr: 0.000047 loss: 0.7469 (0.7589) time: 0.1429 data: 0.0001 max mem: 9672 +[14:10:18.982813] Epoch: [42] Total time: 0:00:07 (0.1902 s / it) +[14:10:18.990864] Averaged stats: lr: 0.000047 loss: 0.7469 (0.7589) +[14:10:20.904723] val: [ 0/13] eta: 0:00:24 loss: 0.5244 (0.5244) time: 1.8974 data: 1.8607 max mem: 9672 +[14:10:22.556381] val: [10/13] eta: 0:00:00 loss: 0.8266 (0.8057) time: 0.3226 data: 0.2876 max mem: 9672 +[14:10:22.613165] val: [12/13] eta: 0:00:00 loss: 0.8266 (0.8994) time: 0.2773 data: 0.2434 max mem: 9672 +[14:10:22.688846] val: Total time: 0:00:03 (0.2833 s / it) +[14:10:22.704645] val loss: 0.8994409396098211 +[14:10:22.704897] Accuracy: 0.7175, F1 Score: 0.6154, ROC AUC: 0.9032, Hamming Loss: 0.1130, + Jaccard Score: 0.4638, Precision: 0.6313, Recall: 0.6038, + Average Precision: 0.6562, Kappa: 0.5969, Score: 0.7052 +[14:10:22.750592] Best epoch = 37, Best score = 0.7125 +[14:10:22.990631] log_dir: ./output_logs/retfound +[14:10:24.840417] Epoch: [43] [ 0/37] eta: 0:01:08 lr: 0.000047 loss: 0.7050 (0.7050) time: 1.8488 data: 1.7039 max mem: 9672 +[14:10:27.698115] Epoch: [43] [20/37] eta: 0:00:03 lr: 0.000040 loss: 0.7917 (0.7803) time: 0.1428 data: 0.0002 max mem: 9672 +[14:10:29.987180] Epoch: [43] [36/37] eta: 0:00:00 lr: 0.000035 loss: 0.7852 (0.8101) time: 0.1432 data: 0.0001 max mem: 9672 +[14:10:30.066544] Epoch: [43] Total time: 0:00:07 (0.1912 s / it) +[14:10:30.075427] Averaged stats: lr: 0.000035 loss: 0.7852 (0.8101) +[14:10:31.999019] val: [ 0/13] eta: 0:00:24 loss: 0.5167 (0.5167) time: 1.9076 data: 1.8727 max mem: 9672 +[14:10:33.749385] val: [10/13] eta: 0:00:00 loss: 0.8503 (0.8043) time: 0.3325 data: 0.2982 max mem: 9672 +[14:10:33.806042] val: [12/13] eta: 0:00:00 loss: 0.8503 (0.9050) time: 0.2856 data: 0.2524 max mem: 9672 +[14:10:33.879155] val: Total time: 0:00:03 (0.2914 s / it) +[14:10:33.895060] val loss: 0.9050251979094285 +[14:10:33.895324] Accuracy: 0.7025, F1 Score: 0.6091, ROC AUC: 0.9023, Hamming Loss: 0.1190, + Jaccard Score: 0.4560, Precision: 0.6172, Recall: 0.6042, + Average Precision: 0.6543, Kappa: 0.5801, Score: 0.6972 +[14:10:33.939935] Best epoch = 37, Best score = 0.7125 +[14:10:34.161517] log_dir: ./output_logs/retfound +[14:10:36.127550] Epoch: [44] [ 0/37] eta: 0:01:12 lr: 0.000035 loss: 0.8958 (0.8958) time: 1.9650 data: 1.8175 max mem: 9672 +[14:10:38.990002] Epoch: [44] [20/37] eta: 0:00:03 lr: 0.000029 loss: 0.7582 (0.7449) time: 0.1431 data: 0.0002 max mem: 9672 +[14:10:41.280197] Epoch: [44] [36/37] eta: 0:00:00 lr: 0.000025 loss: 0.7745 (0.7595) time: 0.1432 data: 0.0001 max mem: 9672 +[14:10:41.359462] Epoch: [44] Total time: 0:00:07 (0.1945 s / it) +[14:10:41.367622] Averaged stats: lr: 0.000025 loss: 0.7745 (0.7595) +[14:10:43.305470] val: [ 0/13] eta: 0:00:24 loss: 0.4718 (0.4718) time: 1.9216 data: 1.8868 max mem: 9672 +[14:10:44.896639] val: [10/13] eta: 0:00:00 loss: 0.8992 (0.8176) time: 0.3193 data: 0.2850 max mem: 9672 +[14:10:44.953172] val: [12/13] eta: 0:00:00 loss: 0.8992 (0.9211) time: 0.2745 data: 0.2412 max mem: 9672 +[14:10:45.039490] val: Total time: 0:00:03 (0.2812 s / it) +[14:10:45.055996] val loss: 0.9210980809651889 +[14:10:45.056248] Accuracy: 0.7075, F1 Score: 0.6071, ROC AUC: 0.9012, Hamming Loss: 0.1170, + Jaccard Score: 0.4548, Precision: 0.6219, Recall: 0.5959, + Average Precision: 0.6491, Kappa: 0.5820, Score: 0.6967 +[14:10:45.099521] Best epoch = 37, Best score = 0.7125 +[14:10:45.345789] log_dir: ./output_logs/retfound +[14:10:47.181431] Epoch: [45] [ 0/37] eta: 0:01:07 lr: 0.000025 loss: 0.6301 (0.6301) time: 1.8344 data: 1.6904 max mem: 9672 +[14:10:50.040783] Epoch: [45] [20/37] eta: 0:00:03 lr: 0.000020 loss: 0.6942 (0.7260) time: 0.1429 data: 0.0002 max mem: 9672 +[14:10:52.326647] Epoch: [45] [36/37] eta: 0:00:00 lr: 0.000016 loss: 0.6942 (0.7297) time: 0.1427 data: 0.0001 max mem: 9672 +[14:10:52.403657] Epoch: [45] Total time: 0:00:07 (0.1907 s / it) +[14:10:52.413238] Averaged stats: lr: 0.000016 loss: 0.6942 (0.7297) +[14:10:54.467331] val: [ 0/13] eta: 0:00:26 loss: 0.4792 (0.4792) time: 2.0374 data: 2.0026 max mem: 9672 +[14:10:56.059676] val: [10/13] eta: 0:00:00 loss: 0.8967 (0.8258) time: 0.3299 data: 0.2953 max mem: 9672 +[14:10:56.116021] val: [12/13] eta: 0:00:00 loss: 0.8967 (0.9264) time: 0.2834 data: 0.2499 max mem: 9672 +[14:10:56.193307] val: Total time: 0:00:03 (0.2895 s / it) +[14:10:56.209250] val loss: 0.9264009961715112 +[14:10:56.209437] Accuracy: 0.7025, F1 Score: 0.6021, ROC AUC: 0.9013, Hamming Loss: 0.1190, + Jaccard Score: 0.4489, Precision: 0.6191, Recall: 0.5900, + Average Precision: 0.6499, Kappa: 0.5743, Score: 0.6925 +[14:10:56.255405] Best epoch = 37, Best score = 0.7125 +[14:10:56.606509] log_dir: ./output_logs/retfound +[14:10:58.573074] Epoch: [46] [ 0/37] eta: 0:01:12 lr: 0.000016 loss: 0.9281 (0.9281) time: 1.9654 data: 1.8154 max mem: 9672 +[14:11:01.421823] Epoch: [46] [20/37] eta: 0:00:03 lr: 0.000012 loss: 0.7444 (0.7603) time: 0.1424 data: 0.0001 max mem: 9672 +[14:11:03.710682] Epoch: [46] [36/37] eta: 0:00:00 lr: 0.000010 loss: 0.7550 (0.7641) time: 0.1430 data: 0.0001 max mem: 9672 +[14:11:03.788264] Epoch: [46] Total time: 0:00:07 (0.1941 s / it) +[14:11:03.797873] Averaged stats: lr: 0.000010 loss: 0.7550 (0.7641) +[14:11:05.671032] val: [ 0/13] eta: 0:00:24 loss: 0.4975 (0.4975) time: 1.8603 data: 1.8242 max mem: 9672 +[14:11:07.225006] val: [10/13] eta: 0:00:00 loss: 0.8758 (0.8182) time: 0.3103 data: 0.2756 max mem: 9672 +[14:11:07.280268] val: [12/13] eta: 0:00:00 loss: 0.8758 (0.9228) time: 0.2668 data: 0.2333 max mem: 9672 +[14:11:07.355958] val: Total time: 0:00:03 (0.2728 s / it) +[14:11:07.374584] val loss: 0.9228192797073951 +[14:11:07.374784] Accuracy: 0.7050, F1 Score: 0.6037, ROC AUC: 0.9016, Hamming Loss: 0.1180, + Jaccard Score: 0.4512, Precision: 0.6220, Recall: 0.5910, + Average Precision: 0.6503, Kappa: 0.5780, Score: 0.6945 +[14:11:07.414198] Best epoch = 37, Best score = 0.7125 +[14:11:07.671739] log_dir: ./output_logs/retfound +[14:11:09.594663] Epoch: [47] [ 0/37] eta: 0:01:11 lr: 0.000010 loss: 0.8618 (0.8618) time: 1.9215 data: 1.7729 max mem: 9672 +[14:11:12.454925] Epoch: [47] [20/37] eta: 0:00:03 lr: 0.000007 loss: 0.6834 (0.7274) time: 0.1430 data: 0.0002 max mem: 9672 +[14:11:14.743387] Epoch: [47] [36/37] eta: 0:00:00 lr: 0.000005 loss: 0.7418 (0.7430) time: 0.1432 data: 0.0001 max mem: 9672 +[14:11:14.822714] Epoch: [47] Total time: 0:00:07 (0.1933 s / it) +[14:11:14.832495] Averaged stats: lr: 0.000005 loss: 0.7418 (0.7430) +[14:11:16.731779] val: [ 0/13] eta: 0:00:24 loss: 0.4983 (0.4983) time: 1.8831 data: 1.8474 max mem: 9672 +[14:11:18.339284] val: [10/13] eta: 0:00:00 loss: 0.8743 (0.8179) time: 0.3172 data: 0.2827 max mem: 9672 +[14:11:18.387713] val: [12/13] eta: 0:00:00 loss: 0.8743 (0.9200) time: 0.2721 data: 0.2392 max mem: 9672 +[14:11:18.463206] val: Total time: 0:00:03 (0.2781 s / it) +[14:11:18.479008] val loss: 0.9199902690373934 +[14:11:18.479198] Accuracy: 0.7075, F1 Score: 0.6061, ROC AUC: 0.9017, Hamming Loss: 0.1170, + Jaccard Score: 0.4541, Precision: 0.6211, Recall: 0.5947, + Average Precision: 0.6500, Kappa: 0.5823, Score: 0.6967 +[14:11:18.521330] Best epoch = 37, Best score = 0.7125 +[14:11:18.754602] log_dir: ./output_logs/retfound +[14:11:20.629928] Epoch: [48] [ 0/37] eta: 0:01:09 lr: 0.000005 loss: 0.6941 (0.6941) time: 1.8743 data: 1.7284 max mem: 9672 +[14:11:23.484027] Epoch: [48] [20/37] eta: 0:00:03 lr: 0.000003 loss: 0.7478 (0.7505) time: 0.1427 data: 0.0002 max mem: 9672 +[14:11:25.769983] Epoch: [48] [36/37] eta: 0:00:00 lr: 0.000002 loss: 0.7160 (0.7328) time: 0.1430 data: 0.0001 max mem: 9672 +[14:11:25.847456] Epoch: [48] Total time: 0:00:07 (0.1917 s / it) +[14:11:25.857299] Averaged stats: lr: 0.000002 loss: 0.7160 (0.7328) +[14:11:27.727639] val: [ 0/13] eta: 0:00:24 loss: 0.5002 (0.5002) time: 1.8542 data: 1.8195 max mem: 9672 +[14:11:29.338357] val: [10/13] eta: 0:00:00 loss: 0.8746 (0.8190) time: 0.3149 data: 0.2803 max mem: 9672 +[14:11:29.501488] val: [12/13] eta: 0:00:00 loss: 0.8746 (0.9210) time: 0.2790 data: 0.2453 max mem: 9672 +[14:11:29.579596] val: Total time: 0:00:03 (0.2851 s / it) +[14:11:29.595628] val loss: 0.9209988438166105 +[14:11:29.595963] Accuracy: 0.7050, F1 Score: 0.6033, ROC AUC: 0.9017, Hamming Loss: 0.1180, + Jaccard Score: 0.4509, Precision: 0.6188, Recall: 0.5918, + Average Precision: 0.6499, Kappa: 0.5786, Score: 0.6945 +[14:11:29.643104] Best epoch = 37, Best score = 0.7125 +[14:11:29.868307] log_dir: ./output_logs/retfound +[14:11:31.837904] Epoch: [49] [ 0/37] eta: 0:01:12 lr: 0.000002 loss: 0.5892 (0.5892) time: 1.9685 data: 1.8240 max mem: 9672 +[14:11:34.700435] Epoch: [49] [20/37] eta: 0:00:03 lr: 0.000001 loss: 0.7042 (0.7401) time: 0.1431 data: 0.0002 max mem: 9672 +[14:11:36.992496] Epoch: [49] [36/37] eta: 0:00:00 lr: 0.000001 loss: 0.6819 (0.7305) time: 0.1431 data: 0.0001 max mem: 9672 +[14:11:37.064983] Epoch: [49] Total time: 0:00:07 (0.1945 s / it) +[14:11:37.072831] Averaged stats: lr: 0.000001 loss: 0.6819 (0.7305) +[14:11:38.968479] val: [ 0/13] eta: 0:00:24 loss: 0.5024 (0.5024) time: 1.8790 data: 1.8447 max mem: 9672 +[14:11:40.579758] val: [10/13] eta: 0:00:00 loss: 0.8731 (0.8190) time: 0.3172 data: 0.2831 max mem: 9672 +[14:11:40.684816] val: [12/13] eta: 0:00:00 loss: 0.8731 (0.9207) time: 0.2765 data: 0.2433 max mem: 9672 +[14:11:40.762535] val: Total time: 0:00:03 (0.2826 s / it) +[14:11:40.778216] val loss: 0.9206776114610525 +[14:11:40.778531] Accuracy: 0.7075, F1 Score: 0.6061, ROC AUC: 0.9018, Hamming Loss: 0.1170, + Jaccard Score: 0.4541, Precision: 0.6211, Recall: 0.5947, + Average Precision: 0.6499, Kappa: 0.5823, Score: 0.6967 +[14:11:40.823976] Best epoch = 37, Best score = 0.7125 +[14:11:44.277310] Test with the best model, epoch = 37: +[14:11:46.175325] test: [ 0/13] eta: 0:00:24 loss: 0.1307 (0.1307) time: 1.8829 data: 1.8527 max mem: 9672 +[14:11:47.791348] test: [10/13] eta: 0:00:00 loss: 0.3910 (0.4938) time: 0.3180 data: 0.2839 max mem: 9672 +[14:11:47.846417] test: [12/13] eta: 0:00:00 loss: 0.4296 (0.6676) time: 0.2733 data: 0.2402 max mem: 9672 +[14:11:47.916717] test: Total time: 0:00:03 (0.2788 s / it) +[14:11:47.932945] val loss: 0.6675626131204458 +[14:11:47.933060] Accuracy: 0.7525, F1 Score: 0.6534, ROC AUC: 0.9271, Hamming Loss: 0.0990, + Jaccard Score: 0.5156, Precision: 0.6920, Recall: 0.6376, + Average Precision: 0.7323, Kappa: 0.6339, Score: 0.7381 +[14:11:48.922659] Training time 0:09:50 +[rank0]:[W615 14:11:49.245705618 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/deepdrid/retfound acc=0.7525 auroc_macro_ovr=0.927122747483938 f1_macro=0.6534 qwk=0.8441825745387079 diff --git a/results/deepdrid/vit/confusion_matrix.png b/results/deepdrid/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..fcc1b2e1a2599efb26f1ce401f67a08f417b353d --- /dev/null +++ b/results/deepdrid/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfa620edaecbc6f2cfe66fabf10c85264f1abb2ebcd14ff5f24762b9a1e78c12 +size 101890 diff --git a/results/deepdrid/vit/log.csv b/results/deepdrid/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..376ed8aedbb08fd89147e67e70e71707e92c3864 --- /dev/null +++ b/results/deepdrid/vit/log.csv @@ -0,0 +1,27 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,2.0803818504015603,0.185,0.5739966902766089,0.2518016292006565,6.983611127479795e-08 +1,1.7019634975327387,0.3975,0.7280216852227811,0.4271544773726819,1.4378022909517224e-07 +2,1.533027668793996,0.51,0.7892368606803464,0.505197821243187,2.1772434691554654e-07 +3,1.4116172525617812,0.5,0.8324596058835816,0.5305108027902496,2.9166846473592086e-07 +4,1.3413185675938923,0.56,0.8362530846944933,0.5617025522167124,3.6561258255629513e-07 +5,1.2947539819611444,0.4175,0.8579166400130305,0.48081541211267637,3.693189057311058e-07 +6,1.23852742711703,0.465,0.8658643034607177,0.5328476633680715,3.68019948958147e-07 +7,1.1577609843677945,0.5925,0.8700112311814749,0.6086996944017987,3.658286567600604e-07 +8,1.1600263913472493,0.59,0.8763281164593844,0.6161851079158226,3.627557048921669e-07 +9,1.0727454490131803,0.6025,0.862096879072524,0.6225841033193683,3.5881606446707524e-07 +10,1.0135341717137232,0.58,0.870676110540449,0.603686736276276,3.540289290169249e-07 +11,0.9804271823830075,0.6275,0.8646321241304233,0.6391532979572025,3.48417620984428e-07 +12,0.9483394556575351,0.5725,0.8495869827402055,0.5894356030596243,3.4200947809827206e-07 +13,0.9129275050428178,0.57,0.8446447766867673,0.5978608407360307,3.3483572018645417e-07 +14,0.9095177484883202,0.5725,0.862086902979124,0.5801633034846807,3.269312970764171e-07 +15,0.8800272345542908,0.6375,0.8524909361517308,0.6447925732209469,3.183347183230031e-07 +16,0.8455840680334303,0.5825,0.8464793915087696,0.6056952370788574,3.090878655937738e-07 +17,0.8239458998044332,0.5975,0.8459816999119916,0.6140812925986582,2.992357886257356e-07 +18,0.809093670712577,0.595,0.8297329154059019,0.6073917778458907,2.8882648574754933e-07 +19,0.7940123379230499,0.6175,0.8342350329139997,0.6051379117171364,2.779106700364982e-07 +20,0.7567668689621819,0.5975,0.8311546262926243,0.6155365501910288,2.6654152224947245e-07 +21,0.767350971698761,0.62,0.8315257039379057,0.6170051727725754,2.547744317316677e-07 +22,0.7695494757758247,0.6025,0.8312001454201878,0.6175996900540998,2.4266672656526296e-07 +23,0.7245845860905118,0.62,0.8354076206945594,0.6170609956045302,2.3027739427276662e-07 +24,0.7427315579520332,0.5725,0.8365173688336792,0.5834167115779776,2.1766679443573644e-07 +25,0.720173544353909,0.5925,0.8317641659594754,0.6005629615547913,2.0489636462896535e-07 diff --git a/results/deepdrid/vit/metrics.json b/results/deepdrid/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..320a03ade7489c235cdf631cb46882769920c9d3 --- /dev/null +++ b/results/deepdrid/vit/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 400, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.7075, + "balanced_accuracy": 0.5952222222222223, + "precision_macro": 0.6279808462873511, + "recall_macro": 0.5952222222222223, + "f1_macro": 0.594512587124109, + "precision_weighted": 0.7504432754117417, + "recall_weighted": 0.7075, + "f1_weighted": 0.719688454340353, + "cohen_kappa": 0.583555792845702, + "quadratic_weighted_kappa": 0.8241790177274049, + "mcc": 0.5888460708608552, + "auroc_macro_ovr": 0.8944440780108496, + "auroc_weighted_ovr": 0.9254467789431365, + "auprc_macro": 0.6584721242239061, + "auroc_per_class": { + "0": 0.95545, + "1": 0.7995268620268621, + "2": 0.9073932926829269, + "3": 0.9380081300813008, + "4": 0.8718421052631579 + }, + "per_class": { + "0": { + "precision": 0.9216867469879518, + "recall": 0.765, + "f1-score": 0.8360655737704918, + "support": 200.0 + }, + "1": { + "precision": 0.20754716981132076, + "recall": 0.3055555555555556, + "f1-score": 0.24719101123595505, + "support": 36.0 + }, + "2": { + "precision": 0.5930232558139535, + "recall": 0.7083333333333334, + "f1-score": 0.6455696202531646, + "support": 72.0 + }, + "3": { + "precision": 0.7176470588235294, + "recall": 0.8472222222222222, + "f1-score": 0.7770700636942676, + "support": 72.0 + }, + "4": { + "precision": 0.7, + "recall": 0.35, + "f1-score": 0.4666666666666667, + "support": 20.0 + }, + "accuracy": 0.7075, + "macro avg": { + "precision": 0.6279808462873511, + "recall": 0.5952222222222223, + "f1-score": 0.594512587124109, + "support": 400.0 + }, + "weighted avg": { + "precision": 0.7504432754117417, + "recall": 0.7075, + "f1-score": 0.719688454340353, + "support": 400.0 + } + } +} \ No newline at end of file diff --git a/results/deepdrid/vit/pr.png b/results/deepdrid/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..8864a2b09891a930cf6fced73ff018921f687f6d --- /dev/null +++ b/results/deepdrid/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f13752c6f22e399e7c772b1d79202eeddc2f924d7e3105435fe51e6efe4e911e +size 93128 diff --git a/results/deepdrid/vit/roc.png b/results/deepdrid/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..cc9cfcc38b404541aab24572c8f63f1551df1182 --- /dev/null +++ b/results/deepdrid/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9c2b00a15a45428b16ca2b392b5e8eed5334b232ba82576b6fc668df33cbb07 +size 86035 diff --git a/results/deepdrid/vit/test_pred.npz b/results/deepdrid/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..3288635ca8cc1696a17b176fb5799968686feffb --- /dev/null +++ b/results/deepdrid/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d762d8ff7aca6e6f20f9dde4b1deae8e18456f08b28e9cb8da09ca8dae622e8e +size 11710 diff --git a/results/deepdrid/vit/train.log b/results/deepdrid/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0981a66de720db23fa6badbd5b48a51005416829 --- /dev/null +++ b/results/deepdrid/vit/train.log @@ -0,0 +1,139 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=1200 val=400 test=400 classes=['0', '1', '2', '3', '4'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=2.0804 val_acc=0.1850 val_auc=0.5740 score=0.2518 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=1.7020 val_acc=0.3975 val_auc=0.7280 score=0.4272 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=1.5330 val_acc=0.5100 val_auc=0.7892 score=0.5052 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.4116 val_acc=0.5000 val_auc=0.8325 score=0.5305 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=1.3413 val_acc=0.5600 val_auc=0.8363 score=0.5617 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=1.2948 val_acc=0.4175 val_auc=0.8579 score=0.4808 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=1.2385 val_acc=0.4650 val_auc=0.8659 score=0.5328 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=1.1578 val_acc=0.5925 val_auc=0.8700 score=0.6087 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=1.1600 val_acc=0.5900 val_auc=0.8763 score=0.6162 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=1.0727 val_acc=0.6025 val_auc=0.8621 score=0.6226 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=1.0135 val_acc=0.5800 val_auc=0.8707 score=0.6037 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.9804 val_acc=0.6275 val_auc=0.8646 score=0.6392 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.9483 val_acc=0.5725 val_auc=0.8496 score=0.5894 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.9129 val_acc=0.5700 val_auc=0.8446 score=0.5979 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.9095 val_acc=0.5725 val_auc=0.8621 score=0.5802 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.8800 val_acc=0.6375 val_auc=0.8525 score=0.6448 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.8456 val_acc=0.5825 val_auc=0.8465 score=0.6057 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.8239 val_acc=0.5975 val_auc=0.8460 score=0.6141 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.8091 val_acc=0.5950 val_auc=0.8297 score=0.6074 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.7940 val_acc=0.6175 val_auc=0.8342 score=0.6051 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.7568 val_acc=0.5975 val_auc=0.8312 score=0.6155 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.7674 val_acc=0.6200 val_auc=0.8315 score=0.6170 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.7695 val_acc=0.6025 val_auc=0.8312 score=0.6176 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.7246 val_acc=0.6200 val_auc=0.8354 score=0.6171 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.7427 val_acc=0.5725 val_auc=0.8365 score=0.5834 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.7202 val_acc=0.5925 val_auc=0.8318 score=0.6006 +[vit] early stop at ep25 (best ep15 score=0.6448) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=15 best_val_score=0.6448 -> saved test_pred.npz (400 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/deepdrid/vit acc=0.7075 auroc_macro_ovr=0.8944440780108496 f1_macro=0.5945 qwk=0.8241790177274049 diff --git a/results/downsample/adam/005/resnet/confusion_matrix.png b/results/downsample/adam/005/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..f0e32282eb15459e45afcf81bf8aa4247811b198 --- /dev/null +++ b/results/downsample/adam/005/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fb816a5fe9296042905080047fdfb441f87f5db1d5484d82cfccec308fa90b3 +size 69550 diff --git a/results/downsample/adam/005/resnet/log.csv b/results/downsample/adam/005/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..9f5a1aa27103a074ec4eeafbf5f86601ab767cfc --- /dev/null +++ b/results/downsample/adam/005/resnet/log.csv @@ -0,0 +1,37 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.674975574016571,0.45,0.4229390681003584,0.19908278652528746,0.0 +1,0.6994781494140625,0.575,0.3942652329749104,0.21171291884124957,0.00016666666666666666 +2,0.668548583984375,0.75,0.4050179211469534,0.26215631029007147,0.0003333333333333333 +3,0.664993405342102,0.775,0.39784946236559143,0.2781563935584835,0.0005 +4,0.648101806640625,0.775,0.4623655913978495,0.29966176990256954,0.0004997919498361457 +5,0.6358826756477356,0.775,0.5232974910394266,0.31997240311642855,0.0004991681456235483 +6,0.6389831900596619,0.775,0.5949820788530467,0.34386726572096854,0.000498129625622757 +7,0.7107492685317993,0.775,0.5967741935483871,0.34446463728608206,0.0004966781183478222 +8,0.6175918579101562,0.775,0.5842293906810035,0.34028303633028756,0.0004948160396893552 +9,0.6038055419921875,0.775,0.6200716845878136,0.3522304676325576,0.0004925464888935161 +10,0.5691101551055908,0.775,0.5770609318996416,0.3378935500698336,0.0004898732434036243 +11,0.6901549696922302,0.775,0.5232974910394266,0.31997240311642855,0.00048680075257297753 +12,0.6168212294578552,0.775,0.5197132616487454,0.3187776599862015,0.0004833341302593417 +13,0.4717712700366974,0.775,0.5340501792114696,0.3235566325071096,0.0004794791463134399 +14,0.5169678330421448,0.775,0.5232974910394266,0.31997240311642855,0.00047524221697560476 +15,0.4846116006374359,0.775,0.5304659498207885,0.32236188937688254,0.0004706303941965803 +16,0.48383787274360657,0.775,0.5519713261648745,0.32953034815824456,0.00046565135390024513 +17,0.5024078488349915,0.775,0.5806451612903225,0.3390882932000605,0.0004603133832077953 +18,0.4555511474609375,0.775,0.6308243727598566,0.3558146970232386,0.00045462536664464835 +19,0.4466705620288849,0.775,0.6594982078853047,0.36537264206505454,0.0004485967713530281 +20,0.398458868265152,0.775,0.6702508960573477,0.3689568714557356,0.00044223763133484053 +21,0.4153594970703125,0.775,0.6917562724014338,0.37612533023709765,0.0004355585307510675 +22,0.35573965311050415,0.775,0.6881720430107526,0.3749305871068706,0.00042857058630547593 +23,0.341287225484848,0.775,0.6738351254480287,0.37015161458596263,0.00042128542874196107 +24,0.3069210350513458,0.775,0.6845878136200717,0.4397301347713079,0.0004137151834863213 +25,0.27249911427497864,0.775,0.6953405017921147,0.44331436416198894,0.0004058724504646834 +26,0.2697181701660156,0.775,0.6881720430107526,0.4409248779015349,0.00039777028313216917 +27,0.2695114314556122,0.775,0.6881720430107526,0.4409248779015349,0.00038942216674670737 +28,0.2731521725654602,0.775,0.6666666666666666,0.4337564191201729,0.000380841995924153 +29,0.25873494148254395,0.775,0.6630824372759856,0.4325616759899458,0.0003720440515120703 +30,0.23027192056179047,0.775,0.6702508960573477,0.4349511622503999,0.00036304297682067144 +31,0.32815074920654297,0.75,0.6630824372759856,0.34817781566641554,0.00035385375325047166 +32,0.13995744287967682,0.75,0.6523297491039426,0.34459358627573455,0.0003444916753572266 +33,0.19190749526023865,0.75,0.6344086021505376,0.33861987062459953,0.00033497232539565416 +34,0.13100530207157135,0.775,0.6308243727598566,0.3558146970232386,0.00032531154738430856 +35,0.13373374938964844,0.775,0.6630824372759856,0.3665673851952816,0.0003155254207347755 diff --git a/results/downsample/adam/005/resnet/metrics.json b/results/downsample/adam/005/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..8bc07b58658fa7d2fd9b5c943c70194f577fc112 --- /dev/null +++ b/results/downsample/adam/005/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.825, + "balanced_accuracy": 0.6505376344086021, + "precision_macro": 0.7916666666666667, + "recall_macro": 0.6505376344086021, + "f1_macro": 0.6785304247990815, + "precision_weighted": 0.8145833333333334, + "recall_weighted": 0.825, + "f1_weighted": 0.7978760045924225, + "cohen_kappa": 0.375, + "quadratic_weighted_kappa": 0.375, + "mcc": 0.41907903806247476, + "auroc": 0.7840501792114696, + "auprc": 0.5617672383607307, + "sensitivity": 0.3333333333333333, + "specificity": 0.967741935483871, + "precision_pos": 0.75, + "f1_pos": 0.46153846153846156, + "per_class": { + "0": { + "precision": 0.8333333333333334, + "recall": 0.967741935483871, + "f1-score": 0.8955223880597015, + "support": 62.0 + }, + "1": { + "precision": 0.75, + "recall": 0.3333333333333333, + "f1-score": 0.46153846153846156, + "support": 18.0 + }, + "accuracy": 0.825, + "macro avg": { + "precision": 0.7916666666666667, + "recall": 0.6505376344086021, + "f1-score": 0.6785304247990815, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8145833333333334, + "recall": 0.825, + "f1-score": 0.7978760045924225, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/005/resnet/pr.png b/results/downsample/adam/005/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..40156051fb66e15be47c61187808a69137bb0a90 --- /dev/null +++ b/results/downsample/adam/005/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f3ec7ebdb9e0892e1b9733610bf8c25fca3e934903d7b2161d181f7f9fe465a +size 53696 diff --git a/results/downsample/adam/005/resnet/roc.png b/results/downsample/adam/005/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..d641dd86a75863ea2ceebcf5ef59815ca0b6d8d0 --- /dev/null +++ b/results/downsample/adam/005/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ad61b1840c8b5f9990199573f1b4a740fd66e60c644b6d9f071adccd588e524 +size 57528 diff --git a/results/downsample/adam/005/resnet/test_pred.npz b/results/downsample/adam/005/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..3dd20ca78634cbe7e9d04c7dcf62fe5ea5ff679b --- /dev/null +++ b/results/downsample/adam/005/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7bcb1064e3c82b3c8d700e3952438c755c855bac7d5fb5611f3a43e3d57bedc +size 1790 diff --git a/results/downsample/adam/005/resnet/train.log b/results/downsample/adam/005/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b16e2c962e4193787f6c1eb80a1583a4b099892c --- /dev/null +++ b/results/downsample/adam/005/resnet/train.log @@ -0,0 +1,189 @@ +[resnet] train=14 val=40 test=80 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6750 val_acc=0.4500 val_auc=0.4229 score=0.1991 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6995 val_acc=0.5750 val_auc=0.3943 score=0.2117 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6685 val_acc=0.7500 val_auc=0.4050 score=0.2622 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6650 val_acc=0.7750 val_auc=0.3978 score=0.2782 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6481 val_acc=0.7750 val_auc=0.4624 score=0.2997 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6359 val_acc=0.7750 val_auc=0.5233 score=0.3200 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6390 val_acc=0.7750 val_auc=0.5950 score=0.3439 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.7107 val_acc=0.7750 val_auc=0.5968 score=0.3445 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6176 val_acc=0.7750 val_auc=0.5842 score=0.3403 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6038 val_acc=0.7750 val_auc=0.6201 score=0.3522 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.5691 val_acc=0.7750 val_auc=0.5771 score=0.3379 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.6902 val_acc=0.7750 val_auc=0.5233 score=0.3200 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.6168 val_acc=0.7750 val_auc=0.5197 score=0.3188 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.4718 val_acc=0.7750 val_auc=0.5341 score=0.3236 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.5170 val_acc=0.7750 val_auc=0.5233 score=0.3200 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.4846 val_acc=0.7750 val_auc=0.5305 score=0.3224 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.4838 val_acc=0.7750 val_auc=0.5520 score=0.3295 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.5024 val_acc=0.7750 val_auc=0.5806 score=0.3391 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.4556 val_acc=0.7750 val_auc=0.6308 score=0.3558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.4467 val_acc=0.7750 val_auc=0.6595 score=0.3654 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.3985 val_acc=0.7750 val_auc=0.6703 score=0.3690 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.4154 val_acc=0.7750 val_auc=0.6918 score=0.3761 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.3557 val_acc=0.7750 val_auc=0.6882 score=0.3749 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.3413 val_acc=0.7750 val_auc=0.6738 score=0.3702 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.3069 val_acc=0.7750 val_auc=0.6846 score=0.4397 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.2725 val_acc=0.7750 val_auc=0.6953 score=0.4433 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.2697 val_acc=0.7750 val_auc=0.6882 score=0.4409 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.2695 val_acc=0.7750 val_auc=0.6882 score=0.4409 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.2732 val_acc=0.7750 val_auc=0.6667 score=0.4338 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.2587 val_acc=0.7750 val_auc=0.6631 score=0.4326 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep30 loss=0.2303 val_acc=0.7750 val_auc=0.6703 score=0.4350 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep31 loss=0.3282 val_acc=0.7500 val_auc=0.6631 score=0.3482 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep32 loss=0.1400 val_acc=0.7500 val_auc=0.6523 score=0.3446 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep33 loss=0.1919 val_acc=0.7500 val_auc=0.6344 score=0.3386 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep34 loss=0.1310 val_acc=0.7750 val_auc=0.6308 score=0.3558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep35 loss=0.1337 val_acc=0.7750 val_auc=0.6631 score=0.3666 +[resnet] early stop at ep35 (best ep25 score=0.4433) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=25 best_val_score=0.4433 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/005/resnet acc=0.8250 auroc=0.7840501792114696 f1_macro=0.6785 qwk=0.375 diff --git a/results/downsample/adam/005/retfound/confusion_matrix.png b/results/downsample/adam/005/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..c7f2220cab082dcb6dab58860cbca5896ad1fa7e --- /dev/null +++ b/results/downsample/adam/005/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fc24241757f80b79119cda6b1fa30d9181f1c22a24509445663e09c83f9beff +size 68400 diff --git a/results/downsample/adam/005/retfound/confusion_matrix_test.jpg b/results/downsample/adam/005/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a44063cdd273cf01b218e16dedbcc5eb44589de0 --- /dev/null +++ b/results/downsample/adam/005/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbf6a7f569061967235b05e90273681d0d6dd737bd7ced46cbededfec57d2cd +size 241221 diff --git a/results/downsample/adam/005/retfound/log.txt b/results/downsample/adam/005/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..bbedc5b453ced1c8813fd6f695a52776d526866c --- /dev/null +++ b/results/downsample/adam/005/retfound/log.txt @@ -0,0 +1,80 @@ +{"train_lr": 0.0, "train_loss": 0.692626953125, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 1.5625e-05, "train_loss": 0.69256591796875, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 3.125e-05, "train_loss": 0.6929931640625, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 4.6875e-05, "train_loss": 0.69293212890625, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 6.25e-05, "train_loss": 0.6856689453125, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 7.8125e-05, "train_loss": 0.66888427734375, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 9.375e-05, "train_loss": 0.6649169921875, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00010937500000000002, "train_loss": 0.66925048828125, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000125, "train_loss": 0.60601806640625, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000140625, "train_loss": 0.52679443359375, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.00015625, "train_loss": 0.564697265625, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.00015617183679026957, "train_loss": 0.569183349609375, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.00015593750457138045, "train_loss": 0.563995361328125, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.00015554747525723593, "train_loss": 0.471588134765625, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.00015500253431496863, "train_loss": 0.585601806640625, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.00015430377918311406, "train_loss": 0.412109375, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00015345261706151562, "train_loss": 0.4241943359375, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0001524507620774113, "train_loss": 0.5689849853515625, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0001513002318334096, "train_loss": 0.5692138671875, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00015000334334430636, "train_loss": 0.5765533447265625, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0001485627083709253, "train_loss": 0.35736083984375, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00014698122816037928, "train_loss": 0.3632354736328125, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00014526208760334486, "train_loss": 0.6175384521484375, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0001434087488201161, "train_loss": 0.3665008544921875, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0001414249441883553, "train_loss": 0.5311508178710938, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.00013931466882658082, "train_loss": 0.7560195922851562, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0001370821725485303, "train_loss": 0.5588150024414062, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.00013473195130460128, "train_loss": 0.5511016845703125, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00013226873812760537, "train_loss": 0.8189773559570312, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0001296974936010702, "train_loss": 0.5426406860351562, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0001270233958692842, "train_loss": 0.836395263671875, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.00012425183020920322, "train_loss": 0.5393142700195312, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00012138837818521947, "train_loss": 0.320587158203125, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00011843880640863348, "train_loss": 0.5794830322265625, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00011540905492446652, "train_loss": 0.3460845947265625, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00011230522524900044, "train_loss": 0.388214111328125, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00010913356808213583, "train_loss": 0.486053466796875, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00010590047071931423, "train_loss": 0.5696868896484375, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.0001026124441883553, "train_loss": 0.400543212890625, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 9.927611013711315e-05, "train_loss": 0.5305023193359375, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 9.58981874983589e-05, "train_loss": 0.6786041259765625, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 9.248547895874417e-05, "train_loss": 0.2126312255859375, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 8.904485725909551e-05, "train_loss": 0.4027557373046875, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 8.558325135362903e-05, "train_loss": 0.517547607421875, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 8.210763245595873e-05, "train_loss": 0.207427978515625, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 7.8625e-05, "train_loss": 0.349609375, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 7.514236754404128e-05, "train_loss": 0.3785858154296875, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.166674864637099e-05, "train_loss": 0.4681854248046875, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 6.82051427409045e-05, "train_loss": 0.5128173828125, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 6.476452104125584e-05, "train_loss": 0.30804443359375, "epoch": 49, "n_parameters": 303303682} +{"train_lr": 6.13518125016411e-05, "train_loss": 0.6630706787109375, "epoch": 50, "n_parameters": 303303682} +{"train_lr": 5.7973889862886864e-05, "train_loss": 0.484100341796875, "epoch": 51, "n_parameters": 303303682} +{"train_lr": 5.4637555811644714e-05, "train_loss": 0.19134521484375, "epoch": 52, "n_parameters": 303303682} +{"train_lr": 5.134952928068578e-05, "train_loss": 0.7011566162109375, "epoch": 53, "n_parameters": 303303682} +{"train_lr": 4.81164319178642e-05, "train_loss": 0.6509246826171875, "epoch": 54, "n_parameters": 303303682} +{"train_lr": 4.4944774750999556e-05, "train_loss": 0.46341705322265625, "epoch": 55, "n_parameters": 303303682} +{"train_lr": 4.184094507553348e-05, "train_loss": 0.31853485107421875, "epoch": 56, "n_parameters": 303303682} +{"train_lr": 3.881119359136654e-05, "train_loss": 0.61370849609375, "epoch": 57, "n_parameters": 303303682} +{"train_lr": 3.586162181478055e-05, "train_loss": 0.3902435302734375, "epoch": 58, "n_parameters": 303303682} +{"train_lr": 3.299816979079678e-05, "train_loss": 0.47113037109375, "epoch": 59, "n_parameters": 303303682} +{"train_lr": 3.0226604130715816e-05, "train_loss": 0.359344482421875, "epoch": 60, "n_parameters": 303303682} +{"train_lr": 2.755250639892981e-05, "train_loss": 0.254364013671875, "epoch": 61, "n_parameters": 303303682} +{"train_lr": 2.4981261872394632e-05, "train_loss": 0.5357818603515625, "epoch": 62, "n_parameters": 303303682} +{"train_lr": 2.251804869539874e-05, "train_loss": 0.41448974609375, "epoch": 63, "n_parameters": 303303682} +{"train_lr": 2.016782745146971e-05, "train_loss": 0.4392852783203125, "epoch": 64, "n_parameters": 303303682} +{"train_lr": 1.7935331173419187e-05, "train_loss": 0.4860076904296875, "epoch": 65, "n_parameters": 303303682} +{"train_lr": 1.5825055811644713e-05, "train_loss": 0.488128662109375, "epoch": 66, "n_parameters": 303303682} +{"train_lr": 1.3841251179883884e-05, "train_loss": 0.50140380859375, "epoch": 67, "n_parameters": 303303682} +{"train_lr": 1.1987912396655145e-05, "train_loss": 0.5718994140625, "epoch": 68, "n_parameters": 303303682} +{"train_lr": 1.0268771839620714e-05, "train_loss": 0.338226318359375, "epoch": 69, "n_parameters": 303303682} +{"train_lr": 8.687291629074723e-06, "train_loss": 0.83282470703125, "epoch": 70, "n_parameters": 303303682} +{"train_lr": 7.246656655693649e-06, "train_loss": 0.5216903686523438, "epoch": 71, "n_parameters": 303303682} +{"train_lr": 5.949768166590399e-06, "train_loss": 0.752685546875, "epoch": 72, "n_parameters": 303303682} +{"train_lr": 4.799237922588707e-06, "train_loss": 0.501495361328125, "epoch": 73, "n_parameters": 303303682} +{"train_lr": 3.797382938484392e-06, "train_loss": 0.452606201171875, "epoch": 74, "n_parameters": 303303682} +{"train_lr": 2.9462208168859414e-06, "train_loss": 0.4839019775390625, "epoch": 75, "n_parameters": 303303682} +{"train_lr": 2.247465685031372e-06, "train_loss": 0.6468963623046875, "epoch": 76, "n_parameters": 303303682} +{"train_lr": 1.702524742764069e-06, "train_loss": 0.349700927734375, "epoch": 77, "n_parameters": 303303682} +{"train_lr": 1.3124954286195674e-06, "train_loss": 0.3069000244140625, "epoch": 78, "n_parameters": 303303682} +{"train_lr": 1.078163209730455e-06, "train_loss": 0.4912872314453125, "epoch": 79, "n_parameters": 303303682} diff --git a/results/downsample/adam/005/retfound/metrics.json b/results/downsample/adam/005/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..5f001c4a7271537c2f0fa70dda956e111c90d4c4 --- /dev/null +++ b/results/downsample/adam/005/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.775, + "balanced_accuracy": 0.5, + "precision_macro": 0.3875, + "recall_macro": 0.5, + "f1_macro": 0.43661971830985913, + "precision_weighted": 0.6006250000000001, + "recall_weighted": 0.775, + "f1_weighted": 0.6767605633802816, + "cohen_kappa": 0.0, + "quadratic_weighted_kappa": 0.0, + "mcc": 0.0, + "auroc": 0.8006272401433692, + "auprc": 0.5656874873347463, + "sensitivity": 0.0, + "specificity": 1.0, + "precision_pos": null, + "f1_pos": 0.0, + "per_class": { + "0": { + "precision": 0.775, + "recall": 1.0, + "f1-score": 0.8732394366197183, + "support": 62.0 + }, + "1": { + "precision": 0.0, + "recall": 0.0, + "f1-score": 0.0, + "support": 18.0 + }, + "accuracy": 0.775, + "macro avg": { + "precision": 0.3875, + "recall": 0.5, + "f1-score": 0.43661971830985913, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.6006250000000001, + "recall": 0.775, + "f1-score": 0.6767605633802816, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/005/retfound/metrics_test.csv b/results/downsample/adam/005/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..329f2c569639d184cf4bdb1f10373facde179e74 --- /dev/null +++ b/results/downsample/adam/005/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5066802978515625,0.775,0.43661971830985913,0.8004032258064516,0.225,0.3875,0.3875,0.5,0.751091757746893,0.0 diff --git a/results/downsample/adam/005/retfound/metrics_val.csv b/results/downsample/adam/005/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..0e52e77b32bc8812af0b7517a9c9fcaa06e7e4cd --- /dev/null +++ b/results/downsample/adam/005/retfound/metrics_val.csv @@ -0,0 +1,81 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.69271240234375,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.69271240234375,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.69271240234375,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.684765625,0.775,0.43661971830985913,0.739247311827957,0.225,0.3875,0.3875,0.5,0.6350664419212807,0.0 +0.67490234375,0.775,0.43661971830985913,0.7616487455197133,0.225,0.3875,0.3875,0.5,0.6420733444549983,0.0 +0.6625,0.775,0.43661971830985913,0.7329749103942652,0.225,0.3875,0.3875,0.5,0.6504178605514476,0.0 +0.6482177734375,0.775,0.43661971830985913,0.728494623655914,0.225,0.3875,0.3875,0.5,0.6426430256471809,0.0 +0.63408203125,0.775,0.43661971830985913,0.7267025089605734,0.225,0.3875,0.3875,0.5,0.6404688874185195,0.0 +0.61885986328125,0.775,0.43661971830985913,0.7123655913978495,0.225,0.3875,0.3875,0.5,0.6270396527951705,0.0 +0.6021728515625,0.775,0.43661971830985913,0.7043010752688172,0.225,0.3875,0.3875,0.5,0.6341072926336042,0.0 +0.58443603515625,0.775,0.43661971830985913,0.7114695340501792,0.225,0.3875,0.3875,0.5,0.6385168975988678,0.0 +0.569580078125,0.775,0.43661971830985913,0.7168458781362008,0.225,0.3875,0.3875,0.5,0.6688902448625683,0.0 +0.55726318359375,0.775,0.43661971830985913,0.7222222222222222,0.225,0.3875,0.3875,0.5,0.6801615303332806,0.0 +0.546624755859375,0.775,0.43661971830985913,0.7204301075268817,0.225,0.3875,0.3875,0.5,0.6744803809093077,0.0 +0.53834228515625,0.775,0.43661971830985913,0.7267025089605735,0.225,0.3875,0.3875,0.5,0.6738094132806908,0.0 +0.531390380859375,0.775,0.43661971830985913,0.7338709677419355,0.225,0.3875,0.3875,0.5,0.683371638884291,0.0 +0.526361083984375,0.775,0.43661971830985913,0.7329749103942653,0.225,0.3875,0.3875,0.5,0.6800867158639099,0.0 +0.5225982666015625,0.775,0.43661971830985913,0.7383512544802868,0.225,0.3875,0.3875,0.5,0.6862512592421874,0.0 +0.52034912109375,0.775,0.43661971830985913,0.7419354838709677,0.225,0.3875,0.3875,0.5,0.6869584327250328,0.0 +0.519140625,0.775,0.43661971830985913,0.75,0.225,0.3875,0.3875,0.5,0.6918723523242736,0.0 +0.51806640625,0.775,0.43661971830985913,0.7589605734767024,0.225,0.3875,0.3875,0.5,0.7002820735889214,0.0 +0.5171112060546875,0.775,0.43661971830985913,0.7724014336917564,0.225,0.3875,0.3875,0.5,0.7053183790582955,0.0 +0.51668701171875,0.775,0.43661971830985913,0.7786738351254481,0.225,0.3875,0.3875,0.5,0.7155871402776786,0.0 +0.51632080078125,0.775,0.43661971830985913,0.7903225806451614,0.225,0.3875,0.3875,0.5,0.7209692201210935,0.0 +0.515655517578125,0.775,0.43661971830985913,0.7912186379928317,0.225,0.3875,0.3875,0.5,0.7214663105308788,0.0 +0.514117431640625,0.775,0.43661971830985913,0.7956989247311828,0.225,0.3875,0.3875,0.5,0.7238135090044948,0.0 +0.5122894287109375,0.775,0.43661971830985913,0.8055555555555556,0.225,0.3875,0.3875,0.5,0.729866742677312,0.0 +0.51025390625,0.775,0.43661971830985913,0.8091397849462365,0.225,0.3875,0.3875,0.5,0.7310165582618798,0.0 +0.5083740234375,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.735294747217488,0.0 +0.506781005859375,0.775,0.43661971830985913,0.8127240143369175,0.225,0.3875,0.3875,0.5,0.7312358334276996,0.0 +0.50521240234375,0.775,0.43661971830985913,0.8073476702508959,0.225,0.3875,0.3875,0.5,0.7270442033599376,0.0 +0.5040130615234375,0.775,0.43661971830985913,0.810931899641577,0.225,0.3875,0.3875,0.5,0.731657749908968,0.0 +0.503277587890625,0.775,0.43661971830985913,0.8118279569892473,0.225,0.3875,0.3875,0.5,0.731230247147,0.0 +0.5026458740234375,0.775,0.43661971830985913,0.8118279569892473,0.225,0.3875,0.3875,0.5,0.7311431808460118,0.0 +0.502099609375,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7345654262511795,0.0 +0.5014556884765625,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7351168461574381,0.0 +0.5006378173828125,0.775,0.43661971830985913,0.8198924731182795,0.225,0.3875,0.3875,0.5,0.7355049779497445,0.0 +0.49990234375,0.775,0.43661971830985913,0.8225806451612903,0.225,0.3875,0.3875,0.5,0.7374739098232062,0.0 +0.4991119384765625,0.775,0.43661971830985913,0.8189964157706093,0.225,0.3875,0.3875,0.5,0.7089874874326093,0.0 +0.49827880859375,0.775,0.43661971830985913,0.818100358422939,0.225,0.3875,0.3875,0.5,0.706844535058291,0.0 +0.497674560546875,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7028762810900371,0.0 +0.4966705322265625,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7033817232285438,0.0 +0.4956756591796875,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7002953034754573,0.0 +0.494671630859375,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7029693331992182,0.0 +0.493377685546875,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7028762810900371,0.0 +0.4920562744140625,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7028762810900371,0.0 +0.490740966796875,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.6997898613369506,0.0 +0.489349365234375,0.775,0.43661971830985913,0.818100358422939,0.225,0.3875,0.3875,0.5,0.7019266134737028,0.0 +0.488134765625,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.6997898613369506,0.0 +0.4869110107421875,0.775,0.43661971830985913,0.818100358422939,0.225,0.3875,0.3875,0.5,0.7019266134737028,0.0 +0.485931396484375,0.775,0.43661971830985913,0.8198924731182796,0.225,0.3875,0.3875,0.5,0.7055464006956405,0.0 +0.484954833984375,0.775,0.43661971830985913,0.8181003584229392,0.225,0.3875,0.3875,0.5,0.7046546102877211,0.0 +0.4839019775390625,0.775,0.43661971830985913,0.8189964157706093,0.225,0.3875,0.3875,0.5,0.7025139480328115,0.0 +0.483056640625,0.775,0.43661971830985913,0.8189964157706093,0.225,0.3875,0.3875,0.5,0.7019897161913754,0.0 +0.4823028564453125,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7031947984448447,0.0 +0.481634521484375,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7018966640821942,0.0 +0.4809295654296875,0.775,0.43661971830985913,0.8154121863799284,0.225,0.3875,0.3875,0.5,0.6992265444765907,0.0 +0.4802825927734375,0.775,0.43661971830985913,0.8154121863799284,0.225,0.3875,0.3875,0.5,0.6994443929807296,0.0 +0.479693603515625,0.775,0.43661971830985913,0.8127240143369175,0.225,0.3875,0.3875,0.5,0.6958246057587918,0.0 +0.4792266845703125,0.775,0.43661971830985913,0.8118279569892473,0.225,0.3875,0.3875,0.5,0.6957277346941788,0.0 +0.4787872314453125,0.775,0.43661971830985913,0.8109318996415771,0.225,0.3875,0.3875,0.5,0.6952403188841824,0.0 +0.47832489013671875,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.6933884670323305,0.0 +0.47794647216796876,0.775,0.43661971830985913,0.8118279569892473,0.225,0.3875,0.3875,0.5,0.6933884670323305,0.0 +0.4776092529296875,0.775,0.43661971830985913,0.8109318996415771,0.225,0.3875,0.3875,0.5,0.6937468899713986,0.0 +0.477386474609375,0.775,0.43661971830985913,0.8118279569892473,0.225,0.3875,0.3875,0.5,0.6970921707360342,0.0 +0.47721099853515625,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7070363515630673,0.0 +0.4770660400390625,0.775,0.43661971830985913,0.8136200716845878,0.225,0.3875,0.3875,0.5,0.7062874258990314,0.0 +0.4768798828125,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.707197268744358,0.0 +0.47679443359375,0.775,0.43661971830985913,0.8154121863799283,0.225,0.3875,0.3875,0.5,0.7069723474668052,0.0 +0.47665557861328123,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7093756192249158,0.0 +0.47664031982421873,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7093756192249158,0.0 +0.47658538818359375,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.708824199318657,0.0 +0.47655792236328126,0.775,0.43661971830985913,0.8154121863799283,0.225,0.3875,0.3875,0.5,0.7069723474668052,0.0 +0.47658843994140626,0.775,0.43661971830985913,0.8154121863799283,0.225,0.3875,0.3875,0.5,0.7069723474668052,0.0 +0.47654571533203127,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7093756192249157,0.0 +0.47653350830078123,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7075237673730639,0.0 +0.47650909423828125,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7075237673730639,0.0 +0.47650909423828125,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7075237673730639,0.0 +0.47648162841796876,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7075237673730639,0.0 +0.47645416259765627,0.775,0.43661971830985913,0.8154121863799283,0.225,0.3875,0.3875,0.5,0.7069723474668052,0.0 diff --git a/results/downsample/adam/005/retfound/pr.png b/results/downsample/adam/005/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..571c4b64b8a5eb0d7080e441884fecaa5227dd53 --- /dev/null +++ b/results/downsample/adam/005/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a613a4b4b705dfb81f5a66197480a8f68e8ba7ed6c991268246b4c7cb874309 +size 49659 diff --git a/results/downsample/adam/005/retfound/roc.png b/results/downsample/adam/005/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..57e4bf8e4a3dc453f5bd967dacdaab954144dd7b --- /dev/null +++ b/results/downsample/adam/005/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39850e71114d41256b2e4a4892d552127743cf44b2b49f21ba8b315f95e919cf +size 59379 diff --git a/results/downsample/adam/005/retfound/test_pred.npz b/results/downsample/adam/005/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..370e0e9b1910bea122761abd0c55921585a99326 --- /dev/null +++ b/results/downsample/adam/005/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cce175c9df7eb436104891c3531fcb48cbee51a7ad767120f2205ad601325051 +size 1470 diff --git a/results/downsample/adam/005/retfound/train.log b/results/downsample/adam/005/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..fb7d88a12d2cc5633a2bbfce0d645a6582aee217 --- /dev/null +++ b/results/downsample/adam/005/retfound/train.log @@ -0,0 +1,1043 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:45:53.839272999 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:45:54.288181] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:45:54.288407] Namespace(batch_size=8, +epochs=80, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/adam_5', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/005', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:45:57.335366] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:45:58.867692] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:45:59.090372] Sampler_train = +[14:45:59.133296] len of train_set: 8 +[14:45:59.448547] [Adaptation] Full fine-tuning: training all parameters. +[14:45:59.449680] number of trainable params (M): 303.30 +[14:45:59.449772] base lr: 5.00e-03 +[14:45:59.449850] actual lr: 1.56e-04 +[14:45:59.449920] accumulate grad iterations: 1 +[14:45:59.449991] effective batch size: 8 +[14:45:59.453365] criterion = CrossEntropyLoss() +[14:45:59.453462] Start training for 80 epochs +[14:45:59.456044] log_dir: ./output_logs/retfound +[14:46:00.775578] Epoch: [0] [0/1] eta: 0:00:01 lr: 0.000000 loss: 0.6926 (0.6926) time: 1.3184 data: 0.8214 max mem: 3223 +[14:46:00.842237] Epoch: [0] Total time: 0:00:01 (1.3860 s / it) +[14:46:00.843630] Averaged stats: lr: 0.000000 loss: 0.6926 (0.6926) +[14:46:01.693005] val: [0/5] eta: 0:00:04 loss: 0.6924 (0.6924) time: 0.8377 data: 0.8144 max mem: 3223 +[14:46:01.784502] val: [4/5] eta: 0:00:00 loss: 0.6924 (0.6927) time: 0.1857 data: 0.1743 max mem: 3223 +[14:46:01.858960] val: Total time: 0:00:01 (0.2008 s / it) +[14:46:01.871342] val loss: 0.69271240234375 +[14:46:01.871558] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:46:05.836299] Best epoch = 0, Best score = 0.3023 +[14:46:06.120420] log_dir: ./output_logs/retfound +[14:46:08.009656] Epoch: [1] [0/1] eta: 0:00:01 lr: 0.000016 loss: 0.6926 (0.6926) time: 1.8880 data: 0.9558 max mem: 3226 +[14:46:08.091091] Epoch: [1] Total time: 0:00:01 (1.9705 s / it) +[14:46:08.092023] Averaged stats: lr: 0.000016 loss: 0.6926 (0.6926) +[14:46:09.110633] val: [0/5] eta: 0:00:05 loss: 0.6924 (0.6924) time: 1.0106 data: 0.9923 max mem: 3226 +[14:46:09.220754] val: [4/5] eta: 0:00:00 loss: 0.6924 (0.6927) time: 0.2240 data: 0.2114 max mem: 3226 +[14:46:09.338417] val: Total time: 0:00:01 (0.2478 s / it) +[14:46:09.349833] val loss: 0.69271240234375 +[14:46:09.350700] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:46:09.446599] Best epoch = 0, Best score = 0.3023 +[14:46:10.262987] log_dir: ./output_logs/retfound +[14:46:11.060533] Epoch: [2] [0/1] eta: 0:00:00 lr: 0.000031 loss: 0.6930 (0.6930) time: 0.7966 data: 0.7592 max mem: 3230 +[14:46:11.135486] Epoch: [2] Total time: 0:00:00 (0.8723 s / it) +[14:46:11.136311] Averaged stats: lr: 0.000031 loss: 0.6930 (0.6930) +[14:46:11.993243] val: [0/5] eta: 0:00:04 loss: 0.6924 (0.6924) time: 0.8461 data: 0.8334 max mem: 3230 +[14:46:12.070247] val: [4/5] eta: 0:00:00 loss: 0.6924 (0.6927) time: 0.1845 data: 0.1745 max mem: 3230 +[14:46:12.150704] val: Total time: 0:00:01 (0.2008 s / it) +[14:46:12.159377] val loss: 0.69271240234375 +[14:46:12.159621] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:46:12.199182] Best epoch = 0, Best score = 0.3023 +[14:46:12.448880] log_dir: ./output_logs/retfound +[14:46:13.356090] Epoch: [3] [0/1] eta: 0:00:00 lr: 0.000047 loss: 0.6929 (0.6929) time: 0.9063 data: 0.7915 max mem: 4749 +[14:46:13.438518] Epoch: [3] Total time: 0:00:00 (0.9895 s / it) +[14:46:13.439333] Averaged stats: lr: 0.000047 loss: 0.6929 (0.6929) +[14:46:14.360636] val: [0/5] eta: 0:00:04 loss: 0.6785 (0.6785) time: 0.9130 data: 0.8998 max mem: 4749 +[14:46:14.394696] val: [4/5] eta: 0:00:00 loss: 0.6792 (0.6848) time: 0.1893 data: 0.1800 max mem: 4749 +[14:46:14.468826] val: Total time: 0:00:01 (0.2044 s / it) +[14:46:14.478610] val loss: 0.684765625 +[14:46:14.478819] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7392, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6351, Kappa: 0.0000, Score: 0.3920 +[14:46:16.227192] Best epoch = 3, Best score = 0.3920 +[14:46:16.288433] log_dir: ./output_logs/retfound +[14:46:17.258121] Epoch: [4] [0/1] eta: 0:00:00 lr: 0.000063 loss: 0.6857 (0.6857) time: 0.9685 data: 0.8764 max mem: 5538 +[14:46:17.333850] Epoch: [4] Total time: 0:00:01 (1.0453 s / it) +[14:46:17.334891] Averaged stats: lr: 0.000063 loss: 0.6857 (0.6857) +[14:46:18.297313] val: [0/5] eta: 0:00:04 loss: 0.6606 (0.6606) time: 0.9385 data: 0.9193 max mem: 5538 +[14:46:18.372238] val: [4/5] eta: 0:00:00 loss: 0.6620 (0.6749) time: 0.2025 data: 0.1884 max mem: 5538 +[14:46:18.450241] val: Total time: 0:00:01 (0.2184 s / it) +[14:46:18.459202] val loss: 0.67490234375 +[14:46:18.459453] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7616, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6421, Kappa: 0.0000, Score: 0.3994 +[14:46:20.473818] Best epoch = 4, Best score = 0.3994 +[14:46:20.541589] log_dir: ./output_logs/retfound +[14:46:21.557717] Epoch: [5] [0/1] eta: 0:00:01 lr: 0.000078 loss: 0.6689 (0.6689) time: 1.0151 data: 0.9472 max mem: 5538 +[14:46:21.640929] Epoch: [5] Total time: 0:00:01 (1.0992 s / it) +[14:46:21.641738] Averaged stats: lr: 0.000078 loss: 0.6689 (0.6689) +[14:46:22.563838] val: [0/5] eta: 0:00:04 loss: 0.6375 (0.6375) time: 0.9076 data: 0.8939 max mem: 5538 +[14:46:22.642650] val: [4/5] eta: 0:00:00 loss: 0.6394 (0.6625) time: 0.1972 data: 0.1877 max mem: 5538 +[14:46:22.723552] val: Total time: 0:00:01 (0.2136 s / it) +[14:46:22.732232] val loss: 0.6625 +[14:46:22.732482] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7330, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6504, Kappa: 0.0000, Score: 0.3899 +[14:46:22.762284] Best epoch = 4, Best score = 0.3994 +[14:46:23.026172] log_dir: ./output_logs/retfound +[14:46:23.893810] Epoch: [6] [0/1] eta: 0:00:00 lr: 0.000094 loss: 0.6649 (0.6649) time: 0.8668 data: 0.8145 max mem: 5538 +[14:46:23.963578] Epoch: [6] Total time: 0:00:00 (0.9372 s / it) +[14:46:23.964369] Averaged stats: lr: 0.000094 loss: 0.6649 (0.6649) +[14:46:24.830782] val: [0/5] eta: 0:00:04 loss: 0.6100 (0.6100) time: 0.8589 data: 0.8457 max mem: 5538 +[14:46:24.952784] val: [4/5] eta: 0:00:00 loss: 0.6125 (0.6482) time: 0.1961 data: 0.1865 max mem: 5538 +[14:46:25.028843] val: Total time: 0:00:01 (0.2115 s / it) +[14:46:25.037523] val loss: 0.6482177734375 +[14:46:25.037811] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7285, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6426, Kappa: 0.0000, Score: 0.3884 +[14:46:25.074857] Best epoch = 4, Best score = 0.3994 +[14:46:25.341962] log_dir: ./output_logs/retfound +[14:46:26.133787] Epoch: [7] [0/1] eta: 0:00:00 lr: 0.000109 loss: 0.6693 (0.6693) time: 0.7907 data: 0.7284 max mem: 5538 +[14:46:26.202278] Epoch: [7] Total time: 0:00:00 (0.8602 s / it) +[14:46:26.203217] Averaged stats: lr: 0.000109 loss: 0.6693 (0.6693) +[14:46:27.061581] val: [0/5] eta: 0:00:04 loss: 0.5823 (0.5823) time: 0.8458 data: 0.8322 max mem: 5538 +[14:46:27.159822] val: [4/5] eta: 0:00:00 loss: 0.5854 (0.6341) time: 0.1886 data: 0.1785 max mem: 5538 +[14:46:27.243725] val: Total time: 0:00:01 (0.2057 s / it) +[14:46:27.253879] val loss: 0.63408203125 +[14:46:27.253983] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7267, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6405, Kappa: 0.0000, Score: 0.3878 +[14:46:27.295294] Best epoch = 4, Best score = 0.3994 +[14:46:27.562354] log_dir: ./output_logs/retfound +[14:46:28.401774] Epoch: [8] [0/1] eta: 0:00:00 lr: 0.000125 loss: 0.6060 (0.6060) time: 0.8385 data: 0.7862 max mem: 5538 +[14:46:28.474886] Epoch: [8] Total time: 0:00:00 (0.9124 s / it) +[14:46:28.475674] Averaged stats: lr: 0.000125 loss: 0.6060 (0.6060) +[14:46:29.372676] val: [0/5] eta: 0:00:04 loss: 0.5515 (0.5515) time: 0.8898 data: 0.8737 max mem: 5538 +[14:46:29.493645] val: [4/5] eta: 0:00:00 loss: 0.5553 (0.6189) time: 0.2020 data: 0.1914 max mem: 5538 +[14:46:29.575435] val: Total time: 0:00:01 (0.2187 s / it) +[14:46:29.584424] val loss: 0.61885986328125 +[14:46:29.584615] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7124, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6270, Kappa: 0.0000, Score: 0.3830 +[14:46:29.619439] Best epoch = 4, Best score = 0.3994 +[14:46:29.887314] log_dir: ./output_logs/retfound +[14:46:30.811617] Epoch: [9] [0/1] eta: 0:00:00 lr: 0.000141 loss: 0.5268 (0.5268) time: 0.9234 data: 0.8725 max mem: 5538 +[14:46:30.886295] Epoch: [9] Total time: 0:00:00 (0.9988 s / it) +[14:46:30.888342] Averaged stats: lr: 0.000141 loss: 0.5268 (0.5268) +[14:46:31.770449] val: [0/5] eta: 0:00:04 loss: 0.5159 (0.5159) time: 0.8642 data: 0.8496 max mem: 5538 +[14:46:31.840054] val: [4/5] eta: 0:00:00 loss: 0.5206 (0.6022) time: 0.1866 data: 0.1761 max mem: 5538 +[14:46:31.919259] val: Total time: 0:00:01 (0.2027 s / it) +[14:46:31.930444] val loss: 0.6021728515625 +[14:46:31.930675] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7043, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6341, Kappa: 0.0000, Score: 0.3803 +[14:46:31.967177] Best epoch = 4, Best score = 0.3994 +[14:46:32.241881] log_dir: ./output_logs/retfound +[14:46:33.160678] Epoch: [10] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5647 (0.5647) time: 0.9180 data: 0.8667 max mem: 5538 +[14:46:33.227582] Epoch: [10] Total time: 0:00:00 (0.9856 s / it) +[14:46:33.228331] Averaged stats: lr: 0.000156 loss: 0.5647 (0.5647) +[14:46:34.095620] val: [0/5] eta: 0:00:04 loss: 0.4760 (0.4760) time: 0.8556 data: 0.8426 max mem: 5538 +[14:46:34.190788] val: [4/5] eta: 0:00:00 loss: 0.4816 (0.5844) time: 0.1900 data: 0.1806 max mem: 5538 +[14:46:34.264360] val: Total time: 0:00:01 (0.2050 s / it) +[14:46:34.273082] val loss: 0.58443603515625 +[14:46:34.273267] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7115, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6385, Kappa: 0.0000, Score: 0.3827 +[14:46:34.308906] Best epoch = 4, Best score = 0.3994 +[14:46:34.564182] log_dir: ./output_logs/retfound +[14:46:35.485551] Epoch: [11] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5692 (0.5692) time: 0.9203 data: 0.8548 max mem: 5538 +[14:46:35.549546] Epoch: [11] Total time: 0:00:00 (0.9852 s / it) +[14:46:35.555374] Averaged stats: lr: 0.000156 loss: 0.5692 (0.5692) +[14:46:36.499107] val: [0/5] eta: 0:00:04 loss: 0.4396 (0.4396) time: 0.9306 data: 0.9111 max mem: 5538 +[14:46:36.587822] val: [4/5] eta: 0:00:00 loss: 0.4461 (0.5696) time: 0.2037 data: 0.1919 max mem: 5538 +[14:46:36.670857] val: Total time: 0:00:01 (0.2206 s / it) +[14:46:36.679670] val loss: 0.569580078125 +[14:46:36.679947] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7168, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6689, Kappa: 0.0000, Score: 0.3845 +[14:46:36.714737] Best epoch = 4, Best score = 0.3994 +[14:46:36.953955] log_dir: ./output_logs/retfound +[14:46:37.868922] Epoch: [12] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5640 (0.5640) time: 0.9140 data: 0.8475 max mem: 5538 +[14:46:37.951720] Epoch: [12] Total time: 0:00:00 (0.9976 s / it) +[14:46:37.952493] Averaged stats: lr: 0.000156 loss: 0.5640 (0.5640) +[14:46:38.794554] val: [0/5] eta: 0:00:04 loss: 0.4068 (0.4068) time: 0.8312 data: 0.8183 max mem: 5538 +[14:46:38.899564] val: [4/5] eta: 0:00:00 loss: 0.4143 (0.5573) time: 0.1871 data: 0.1777 max mem: 5538 +[14:46:38.972401] val: Total time: 0:00:01 (0.2020 s / it) +[14:46:38.981294] val loss: 0.55726318359375 +[14:46:38.981539] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7222, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6802, Kappa: 0.0000, Score: 0.3863 +[14:46:39.040954] Best epoch = 4, Best score = 0.3994 +[14:46:39.283271] log_dir: ./output_logs/retfound +[14:46:40.152864] Epoch: [13] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.4716 (0.4716) time: 0.8687 data: 0.8085 max mem: 5538 +[14:46:40.226000] Epoch: [13] Total time: 0:00:00 (0.9426 s / it) +[14:46:40.226790] Averaged stats: lr: 0.000156 loss: 0.4716 (0.4716) +[14:46:41.116401] val: [0/5] eta: 0:00:04 loss: 0.3757 (0.3757) time: 0.8744 data: 0.8613 max mem: 5538 +[14:46:41.197544] val: [4/5] eta: 0:00:00 loss: 0.3838 (0.5466) time: 0.1910 data: 0.1805 max mem: 5538 +[14:46:41.269405] val: Total time: 0:00:01 (0.2056 s / it) +[14:46:41.278195] val loss: 0.546624755859375 +[14:46:41.278394] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7204, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6745, Kappa: 0.0000, Score: 0.3857 +[14:46:41.318310] Best epoch = 4, Best score = 0.3994 +[14:46:41.579602] log_dir: ./output_logs/retfound +[14:46:42.388552] Epoch: [14] [0/1] eta: 0:00:00 lr: 0.000155 loss: 0.5856 (0.5856) time: 0.8079 data: 0.7485 max mem: 5538 +[14:46:42.466410] Epoch: [14] Total time: 0:00:00 (0.8866 s / it) +[14:46:42.467174] Averaged stats: lr: 0.000155 loss: 0.5856 (0.5856) +[14:46:43.349763] val: [0/5] eta: 0:00:04 loss: 0.3481 (0.3481) time: 0.8713 data: 0.8576 max mem: 5538 +[14:46:43.460722] val: [4/5] eta: 0:00:00 loss: 0.3571 (0.5383) time: 0.1963 data: 0.1868 max mem: 5538 +[14:46:43.540078] val: Total time: 0:00:01 (0.2125 s / it) +[14:46:43.554343] val loss: 0.53834228515625 +[14:46:43.554643] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7267, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6738, Kappa: 0.0000, Score: 0.3878 +[14:46:43.587582] Best epoch = 4, Best score = 0.3994 +[14:46:43.831640] log_dir: ./output_logs/retfound +[14:46:44.749712] Epoch: [15] [0/1] eta: 0:00:00 lr: 0.000154 loss: 0.4121 (0.4121) time: 0.9171 data: 0.8666 max mem: 5538 +[14:46:44.836128] Epoch: [15] Total time: 0:00:01 (1.0043 s / it) +[14:46:44.836885] Averaged stats: lr: 0.000154 loss: 0.4121 (0.4121) +[14:46:45.686518] val: [0/5] eta: 0:00:04 loss: 0.3218 (0.3218) time: 0.8335 data: 0.8198 max mem: 5538 +[14:46:45.771491] val: [4/5] eta: 0:00:00 loss: 0.3313 (0.5314) time: 0.1836 data: 0.1742 max mem: 5538 +[14:46:45.857854] val: Total time: 0:00:01 (0.2011 s / it) +[14:46:45.866650] val loss: 0.531390380859375 +[14:46:45.866947] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7339, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6834, Kappa: 0.0000, Score: 0.3902 +[14:46:45.898016] Best epoch = 4, Best score = 0.3994 +[14:46:46.150212] log_dir: ./output_logs/retfound +[14:46:46.995919] Epoch: [16] [0/1] eta: 0:00:00 lr: 0.000153 loss: 0.4242 (0.4242) time: 0.8447 data: 0.7938 max mem: 5538 +[14:46:47.063630] Epoch: [16] Total time: 0:00:00 (0.9133 s / it) +[14:46:47.064630] Averaged stats: lr: 0.000153 loss: 0.4242 (0.4242) +[14:46:48.003105] val: [0/5] eta: 0:00:04 loss: 0.2973 (0.2973) time: 0.9258 data: 0.9126 max mem: 5538 +[14:46:48.038162] val: [4/5] eta: 0:00:00 loss: 0.3072 (0.5264) time: 0.1921 data: 0.1826 max mem: 5538 +[14:46:48.108459] val: Total time: 0:00:01 (0.2064 s / it) +[14:46:48.117215] val loss: 0.526361083984375 +[14:46:48.117483] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7330, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6801, Kappa: 0.0000, Score: 0.3899 +[14:46:48.162899] Best epoch = 4, Best score = 0.3994 +[14:46:48.411739] log_dir: ./output_logs/retfound +[14:46:49.226276] Epoch: [17] [0/1] eta: 0:00:00 lr: 0.000152 loss: 0.5690 (0.5690) time: 0.8133 data: 0.7483 max mem: 5538 +[14:46:49.296609] Epoch: [17] Total time: 0:00:00 (0.8847 s / it) +[14:46:49.297613] Averaged stats: lr: 0.000152 loss: 0.5690 (0.5690) +[14:46:50.239509] val: [0/5] eta: 0:00:04 loss: 0.2754 (0.2754) time: 0.9168 data: 0.9038 max mem: 5538 +[14:46:50.338070] val: [4/5] eta: 0:00:00 loss: 0.2856 (0.5226) time: 0.2030 data: 0.1935 max mem: 5538 +[14:46:50.436443] val: Total time: 0:00:01 (0.2229 s / it) +[14:46:50.445151] val loss: 0.5225982666015625 +[14:46:50.445430] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7384, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6863, Kappa: 0.0000, Score: 0.3917 +[14:46:50.495037] Best epoch = 4, Best score = 0.3994 +[14:46:51.170766] log_dir: ./output_logs/retfound +[14:46:52.189877] Epoch: [18] [0/1] eta: 0:00:01 lr: 0.000151 loss: 0.5692 (0.5692) time: 1.0181 data: 0.9671 max mem: 5538 +[14:46:52.264360] Epoch: [18] Total time: 0:00:01 (1.0934 s / it) +[14:46:52.265159] Averaged stats: lr: 0.000151 loss: 0.5692 (0.5692) +[14:46:53.152736] val: [0/5] eta: 0:00:04 loss: 0.2565 (0.2565) time: 0.8796 data: 0.8592 max mem: 5538 +[14:46:53.314497] val: [4/5] eta: 0:00:00 loss: 0.2670 (0.5203) time: 0.2081 data: 0.1946 max mem: 5538 +[14:46:53.394022] val: Total time: 0:00:01 (0.2243 s / it) +[14:46:53.402806] val loss: 0.52034912109375 +[14:46:53.403084] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7419, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6870, Kappa: 0.0000, Score: 0.3929 +[14:46:53.436997] Best epoch = 4, Best score = 0.3994 +[14:46:53.707188] log_dir: ./output_logs/retfound +[14:46:54.582169] Epoch: [19] [0/1] eta: 0:00:00 lr: 0.000150 loss: 0.5766 (0.5766) time: 0.8739 data: 0.8169 max mem: 5538 +[14:46:54.653056] Epoch: [19] Total time: 0:00:00 (0.9457 s / it) +[14:46:54.654092] Averaged stats: lr: 0.000150 loss: 0.5766 (0.5766) +[14:46:55.536015] val: [0/5] eta: 0:00:04 loss: 0.2422 (0.2422) time: 0.8697 data: 0.8513 max mem: 5538 +[14:46:55.620067] val: [4/5] eta: 0:00:00 loss: 0.2528 (0.5191) time: 0.1906 data: 0.1787 max mem: 5538 +[14:46:55.695771] val: Total time: 0:00:01 (0.2061 s / it) +[14:46:55.704828] val loss: 0.519140625 +[14:46:55.705089] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7500, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6919, Kappa: 0.0000, Score: 0.3955 +[14:46:55.739653] Best epoch = 4, Best score = 0.3994 +[14:46:56.019110] log_dir: ./output_logs/retfound +[14:46:56.843039] Epoch: [20] [0/1] eta: 0:00:00 lr: 0.000149 loss: 0.3574 (0.3574) time: 0.8229 data: 0.7648 max mem: 5538 +[14:46:56.919948] Epoch: [20] Total time: 0:00:00 (0.9006 s / it) +[14:46:56.920701] Averaged stats: lr: 0.000149 loss: 0.3574 (0.3574) +[14:46:57.777570] val: [0/5] eta: 0:00:04 loss: 0.2292 (0.2292) time: 0.8430 data: 0.8296 max mem: 5538 +[14:46:57.875746] val: [4/5] eta: 0:00:00 loss: 0.2400 (0.5181) time: 0.1881 data: 0.1785 max mem: 5538 +[14:46:57.952444] val: Total time: 0:00:01 (0.2037 s / it) +[14:46:57.961165] val loss: 0.51806640625 +[14:46:57.961351] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7590, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7003, Kappa: 0.0000, Score: 0.3985 +[14:46:58.000627] Best epoch = 4, Best score = 0.3994 +[14:46:58.262938] log_dir: ./output_logs/retfound +[14:46:59.102155] Epoch: [21] [0/1] eta: 0:00:00 lr: 0.000147 loss: 0.3632 (0.3632) time: 0.8384 data: 0.7831 max mem: 5538 +[14:46:59.172637] Epoch: [21] Total time: 0:00:00 (0.9095 s / it) +[14:46:59.173570] Averaged stats: lr: 0.000147 loss: 0.3632 (0.3632) +[14:47:00.131958] val: [0/5] eta: 0:00:04 loss: 0.2168 (0.2168) time: 0.9471 data: 0.9326 max mem: 5538 +[14:47:00.338956] val: [4/5] eta: 0:00:00 loss: 0.2280 (0.5171) time: 0.2300 data: 0.2189 max mem: 5538 +[14:47:00.432687] val: Total time: 0:00:01 (0.2497 s / it) +[14:47:00.442434] val loss: 0.5171112060546875 +[14:47:00.442634] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7724, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7053, Kappa: 0.0000, Score: 0.4030 +[14:47:02.412650] Best epoch = 21, Best score = 0.4030 +[14:47:02.481436] log_dir: ./output_logs/retfound +[14:47:03.415383] Epoch: [22] [0/1] eta: 0:00:00 lr: 0.000145 loss: 0.6175 (0.6175) time: 0.9330 data: 0.8820 max mem: 5538 +[14:47:03.486597] Epoch: [22] Total time: 0:00:01 (1.0050 s / it) +[14:47:03.487341] Averaged stats: lr: 0.000145 loss: 0.6175 (0.6175) +[14:47:04.453737] val: [0/5] eta: 0:00:04 loss: 0.2069 (0.2069) time: 0.9432 data: 0.9296 max mem: 5538 +[14:47:04.548720] val: [4/5] eta: 0:00:00 loss: 0.2185 (0.5167) time: 0.2075 data: 0.1981 max mem: 5538 +[14:47:04.625991] val: Total time: 0:00:01 (0.2232 s / it) +[14:47:04.636641] val loss: 0.51668701171875 +[14:47:04.636900] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7787, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7156, Kappa: 0.0000, Score: 0.4051 +[14:47:06.558544] Best epoch = 22, Best score = 0.4051 +[14:47:06.623643] log_dir: ./output_logs/retfound +[14:47:07.555143] Epoch: [23] [0/1] eta: 0:00:00 lr: 0.000143 loss: 0.3665 (0.3665) time: 0.9303 data: 0.8624 max mem: 5538 +[14:47:07.626050] Epoch: [23] Total time: 0:00:01 (1.0022 s / it) +[14:47:07.627083] Averaged stats: lr: 0.000143 loss: 0.3665 (0.3665) +[14:47:08.517648] val: [0/5] eta: 0:00:04 loss: 0.1984 (0.1984) time: 0.8768 data: 0.8592 max mem: 5538 +[14:47:08.612026] val: [4/5] eta: 0:00:00 loss: 0.2103 (0.5163) time: 0.1941 data: 0.1825 max mem: 5538 +[14:47:08.688203] val: Total time: 0:00:01 (0.2096 s / it) +[14:47:08.697172] val loss: 0.51632080078125 +[14:47:08.697364] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7903, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7210, Kappa: 0.0000, Score: 0.4090 +[14:47:10.607896] Best epoch = 23, Best score = 0.4090 +[14:47:10.665032] log_dir: ./output_logs/retfound +[14:47:11.549034] Epoch: [24] [0/1] eta: 0:00:00 lr: 0.000141 loss: 0.5312 (0.5312) time: 0.8830 data: 0.8204 max mem: 5538 +[14:47:11.620281] Epoch: [24] Total time: 0:00:00 (0.9551 s / it) +[14:47:11.621157] Averaged stats: lr: 0.000141 loss: 0.5312 (0.5312) +[14:47:12.608823] val: [0/5] eta: 0:00:04 loss: 0.1920 (0.1920) time: 0.9755 data: 0.9575 max mem: 5538 +[14:47:12.700425] val: [4/5] eta: 0:00:00 loss: 0.2042 (0.5157) time: 0.2133 data: 0.2023 max mem: 5538 +[14:47:12.780950] val: Total time: 0:00:01 (0.2297 s / it) +[14:47:12.791987] val loss: 0.515655517578125 +[14:47:12.792201] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7912, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7215, Kappa: 0.0000, Score: 0.4093 +[14:47:14.782970] Best epoch = 24, Best score = 0.4093 +[14:47:14.848541] log_dir: ./output_logs/retfound +[14:47:15.736935] Epoch: [25] [0/1] eta: 0:00:00 lr: 0.000139 loss: 0.7560 (0.7560) time: 0.8873 data: 0.8133 max mem: 5538 +[14:47:15.809319] Epoch: [25] Total time: 0:00:00 (0.9606 s / it) +[14:47:15.810226] Averaged stats: lr: 0.000139 loss: 0.7560 (0.7560) +[14:47:16.710695] val: [0/5] eta: 0:00:04 loss: 0.1896 (0.1896) time: 0.8872 data: 0.8739 max mem: 5538 +[14:47:16.795178] val: [4/5] eta: 0:00:00 loss: 0.2022 (0.5141) time: 0.1942 data: 0.1850 max mem: 5538 +[14:47:16.971705] val: Total time: 0:00:01 (0.2298 s / it) +[14:47:16.981292] val loss: 0.514117431640625 +[14:47:16.981469] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7957, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7238, Kappa: 0.0000, Score: 0.4108 +[14:47:19.050288] Best epoch = 25, Best score = 0.4108 +[14:47:19.121494] log_dir: ./output_logs/retfound +[14:47:20.078089] Epoch: [26] [0/1] eta: 0:00:00 lr: 0.000137 loss: 0.5588 (0.5588) time: 0.9556 data: 0.8966 max mem: 5538 +[14:47:20.149146] Epoch: [26] Total time: 0:00:01 (1.0275 s / it) +[14:47:20.150115] Averaged stats: lr: 0.000137 loss: 0.5588 (0.5588) +[14:47:21.043663] val: [0/5] eta: 0:00:04 loss: 0.1892 (0.1892) time: 0.8817 data: 0.8683 max mem: 5538 +[14:47:21.123496] val: [4/5] eta: 0:00:00 loss: 0.2020 (0.5123) time: 0.1922 data: 0.1823 max mem: 5538 +[14:47:21.204631] val: Total time: 0:00:01 (0.2087 s / it) +[14:47:21.214330] val loss: 0.5122894287109375 +[14:47:21.214544] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8056, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7299, Kappa: 0.0000, Score: 0.4141 +[14:47:23.139025] Best epoch = 26, Best score = 0.4141 +[14:47:23.204168] log_dir: ./output_logs/retfound +[14:47:24.213374] Epoch: [27] [0/1] eta: 0:00:01 lr: 0.000135 loss: 0.5511 (0.5511) time: 1.0082 data: 0.9554 max mem: 5538 +[14:47:24.284697] Epoch: [27] Total time: 0:00:01 (1.0804 s / it) +[14:47:24.285596] Averaged stats: lr: 0.000135 loss: 0.5511 (0.5511) +[14:47:25.274581] val: [0/5] eta: 0:00:04 loss: 0.1913 (0.1913) time: 0.9770 data: 0.9634 max mem: 5538 +[14:47:25.353910] val: [4/5] eta: 0:00:00 loss: 0.2044 (0.5103) time: 0.2111 data: 0.2017 max mem: 5538 +[14:47:25.428004] val: Total time: 0:00:01 (0.2262 s / it) +[14:47:25.437706] val loss: 0.51025390625 +[14:47:25.437917] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8091, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7310, Kappa: 0.0000, Score: 0.4153 +[14:47:27.257752] Best epoch = 27, Best score = 0.4153 +[14:47:27.331013] log_dir: ./output_logs/retfound +[14:47:28.220711] Epoch: [28] [0/1] eta: 0:00:00 lr: 0.000132 loss: 0.8190 (0.8190) time: 0.8887 data: 0.8256 max mem: 5538 +[14:47:28.295130] Epoch: [28] Total time: 0:00:00 (0.9639 s / it) +[14:47:28.296056] Averaged stats: lr: 0.000132 loss: 0.8190 (0.8190) +[14:47:29.319394] val: [0/5] eta: 0:00:05 loss: 0.1960 (0.1960) time: 1.0081 data: 0.9949 max mem: 5538 +[14:47:29.352982] val: [4/5] eta: 0:00:00 loss: 0.2093 (0.5084) time: 0.2082 data: 0.1991 max mem: 5538 +[14:47:29.433667] val: Total time: 0:00:01 (0.2246 s / it) +[14:47:29.443285] val loss: 0.5083740234375 +[14:47:29.443487] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7353, Kappa: 0.0000, Score: 0.4176 +[14:47:31.413359] Best epoch = 28, Best score = 0.4176 +[14:47:31.498667] log_dir: ./output_logs/retfound +[14:47:32.482015] Epoch: [29] [0/1] eta: 0:00:00 lr: 0.000130 loss: 0.5426 (0.5426) time: 0.9824 data: 0.9305 max mem: 5538 +[14:47:32.551908] Epoch: [29] Total time: 0:00:01 (1.0530 s / it) +[14:47:32.552760] Averaged stats: lr: 0.000130 loss: 0.5426 (0.5426) +[14:47:33.448513] val: [0/5] eta: 0:00:04 loss: 0.2021 (0.2021) time: 0.8734 data: 0.8605 max mem: 5538 +[14:47:33.536386] val: [4/5] eta: 0:00:00 loss: 0.2156 (0.5068) time: 0.1921 data: 0.1825 max mem: 5538 +[14:47:33.614108] val: Total time: 0:00:01 (0.2079 s / it) +[14:47:33.624172] val loss: 0.506781005859375 +[14:47:33.624372] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8127, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7312, Kappa: 0.0000, Score: 0.4164 +[14:47:33.688963] Best epoch = 28, Best score = 0.4176 +[14:47:33.958939] log_dir: ./output_logs/retfound +[14:47:34.835554] Epoch: [30] [0/1] eta: 0:00:00 lr: 0.000127 loss: 0.8364 (0.8364) time: 0.8757 data: 0.8253 max mem: 5538 +[14:47:34.910993] Epoch: [30] Total time: 0:00:00 (0.9519 s / it) +[14:47:34.911763] Averaged stats: lr: 0.000127 loss: 0.8364 (0.8364) +[14:47:35.871904] val: [0/5] eta: 0:00:04 loss: 0.2111 (0.2111) time: 0.9438 data: 0.9261 max mem: 5538 +[14:47:35.977703] val: [4/5] eta: 0:00:00 loss: 0.2249 (0.5052) time: 0.2098 data: 0.1960 max mem: 5538 +[14:47:36.053711] val: Total time: 0:00:01 (0.2253 s / it) +[14:47:36.063450] val loss: 0.50521240234375 +[14:47:36.063639] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8073, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7270, Kappa: 0.0000, Score: 0.4147 +[14:47:36.123971] Best epoch = 28, Best score = 0.4176 +[14:47:36.430804] log_dir: ./output_logs/retfound +[14:47:37.536578] Epoch: [31] [0/1] eta: 0:00:01 lr: 0.000124 loss: 0.5393 (0.5393) time: 1.1049 data: 1.0540 max mem: 5538 +[14:47:37.601565] Epoch: [31] Total time: 0:00:01 (1.1706 s / it) +[14:47:37.602353] Averaged stats: lr: 0.000124 loss: 0.5393 (0.5393) +[14:47:38.657989] val: [0/5] eta: 0:00:05 loss: 0.2200 (0.2200) time: 1.0142 data: 0.9970 max mem: 5538 +[14:47:38.756388] val: [4/5] eta: 0:00:00 loss: 0.2340 (0.5040) time: 0.2224 data: 0.2113 max mem: 5538 +[14:47:38.831364] val: Total time: 0:00:01 (0.2377 s / it) +[14:47:38.842360] val loss: 0.5040130615234375 +[14:47:38.842572] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8109, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7317, Kappa: 0.0000, Score: 0.4159 +[14:47:38.909811] Best epoch = 28, Best score = 0.4176 +[14:47:39.219481] log_dir: ./output_logs/retfound +[14:47:40.205871] Epoch: [32] [0/1] eta: 0:00:00 lr: 0.000121 loss: 0.3206 (0.3206) time: 0.9855 data: 0.9337 max mem: 5538 +[14:47:40.280893] Epoch: [32] Total time: 0:00:01 (1.0613 s / it) +[14:47:40.281640] Averaged stats: lr: 0.000121 loss: 0.3206 (0.3206) +[14:47:41.213129] val: [0/5] eta: 0:00:04 loss: 0.2267 (0.2267) time: 0.9138 data: 0.8997 max mem: 5538 +[14:47:41.328010] val: [4/5] eta: 0:00:00 loss: 0.2408 (0.5033) time: 0.2056 data: 0.1956 max mem: 5538 +[14:47:41.401281] val: Total time: 0:00:01 (0.2205 s / it) +[14:47:41.410056] val loss: 0.503277587890625 +[14:47:41.410306] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8118, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7312, Kappa: 0.0000, Score: 0.4161 +[14:47:41.516117] Best epoch = 28, Best score = 0.4176 +[14:47:41.824677] log_dir: ./output_logs/retfound +[14:47:42.789528] Epoch: [33] [0/1] eta: 0:00:00 lr: 0.000118 loss: 0.5795 (0.5795) time: 0.9639 data: 0.9119 max mem: 5538 +[14:47:42.864652] Epoch: [33] Total time: 0:00:01 (1.0398 s / it) +[14:47:42.865545] Averaged stats: lr: 0.000118 loss: 0.5795 (0.5795) +[14:47:43.787552] val: [0/5] eta: 0:00:04 loss: 0.2333 (0.2333) time: 0.8965 data: 0.8837 max mem: 5538 +[14:47:43.944092] val: [4/5] eta: 0:00:00 loss: 0.2476 (0.5026) time: 0.2105 data: 0.2015 max mem: 5538 +[14:47:44.018083] val: Total time: 0:00:01 (0.2255 s / it) +[14:47:44.028245] val loss: 0.5026458740234375 +[14:47:44.028448] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8118, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7311, Kappa: 0.0000, Score: 0.4161 +[14:47:44.089860] Best epoch = 28, Best score = 0.4176 +[14:47:44.416329] log_dir: ./output_logs/retfound +[14:47:45.454412] Epoch: [34] [0/1] eta: 0:00:01 lr: 0.000115 loss: 0.3461 (0.3461) time: 1.0370 data: 0.9876 max mem: 5538 +[14:47:45.530012] Epoch: [34] Total time: 0:00:01 (1.1134 s / it) +[14:47:45.530816] Averaged stats: lr: 0.000115 loss: 0.3461 (0.3461) +[14:47:46.562027] val: [0/5] eta: 0:00:05 loss: 0.2388 (0.2388) time: 1.0021 data: 0.9871 max mem: 5538 +[14:47:46.686709] val: [4/5] eta: 0:00:00 loss: 0.2532 (0.5021) time: 0.2252 data: 0.2152 max mem: 5538 +[14:47:46.767182] val: Total time: 0:00:01 (0.2416 s / it) +[14:47:46.778690] val loss: 0.502099609375 +[14:47:46.778964] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7346, Kappa: 0.0000, Score: 0.4179 +[14:47:48.735831] Best epoch = 34, Best score = 0.4179 +[14:47:48.883633] log_dir: ./output_logs/retfound +[14:47:49.968317] Epoch: [35] [0/1] eta: 0:00:01 lr: 0.000112 loss: 0.3882 (0.3882) time: 1.0838 data: 1.0294 max mem: 5538 +[14:47:50.057677] Epoch: [35] Total time: 0:00:01 (1.1739 s / it) +[14:47:50.058488] Averaged stats: lr: 0.000112 loss: 0.3882 (0.3882) +[14:47:51.122177] val: [0/5] eta: 0:00:05 loss: 0.2426 (0.2426) time: 1.0307 data: 1.0171 max mem: 5538 +[14:47:51.267461] val: [4/5] eta: 0:00:00 loss: 0.2571 (0.5015) time: 0.2351 data: 0.2257 max mem: 5538 +[14:47:51.344858] val: Total time: 0:00:01 (0.2508 s / it) +[14:47:51.354702] val loss: 0.5014556884765625 +[14:47:51.354902] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7351, Kappa: 0.0000, Score: 0.4176 +[14:47:51.399226] Best epoch = 34, Best score = 0.4179 +[14:47:51.675255] log_dir: ./output_logs/retfound +[14:47:52.684249] Epoch: [36] [0/1] eta: 0:00:01 lr: 0.000109 loss: 0.4861 (0.4861) time: 1.0080 data: 0.9586 max mem: 5538 +[14:47:52.756377] Epoch: [36] Total time: 0:00:01 (1.0809 s / it) +[14:47:52.757177] Averaged stats: lr: 0.000109 loss: 0.4861 (0.4861) +[14:47:53.796566] val: [0/5] eta: 0:00:04 loss: 0.2469 (0.2469) time: 0.9983 data: 0.9799 max mem: 5538 +[14:47:53.836496] val: [4/5] eta: 0:00:00 loss: 0.2616 (0.5006) time: 0.2075 data: 0.1972 max mem: 5538 +[14:47:53.913020] val: Total time: 0:00:01 (0.2231 s / it) +[14:47:53.921827] val loss: 0.5006378173828125 +[14:47:53.922121] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8199, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7355, Kappa: 0.0000, Score: 0.4188 +[14:47:55.900936] Best epoch = 36, Best score = 0.4188 +[14:47:56.007339] log_dir: ./output_logs/retfound +[14:47:57.036652] Epoch: [37] [0/1] eta: 0:00:01 lr: 0.000106 loss: 0.5697 (0.5697) time: 1.0284 data: 0.9701 max mem: 5538 +[14:47:57.110026] Epoch: [37] Total time: 0:00:01 (1.1025 s / it) +[14:47:57.110775] Averaged stats: lr: 0.000106 loss: 0.5697 (0.5697) +[14:47:58.163681] val: [0/5] eta: 0:00:05 loss: 0.2514 (0.2514) time: 1.0222 data: 1.0026 max mem: 5538 +[14:47:58.274016] val: [4/5] eta: 0:00:00 loss: 0.2664 (0.4999) time: 0.2263 data: 0.2125 max mem: 5538 +[14:47:58.354665] val: Total time: 0:00:01 (0.2428 s / it) +[14:47:58.365961] val loss: 0.49990234375 +[14:47:58.366158] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8226, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7375, Kappa: 0.0000, Score: 0.4197 +[14:48:00.311677] Best epoch = 37, Best score = 0.4197 +[14:48:00.426205] log_dir: ./output_logs/retfound +[14:48:01.414053] Epoch: [38] [0/1] eta: 0:00:00 lr: 0.000103 loss: 0.4005 (0.4005) time: 0.9869 data: 0.9351 max mem: 5538 +[14:48:01.483978] Epoch: [38] Total time: 0:00:01 (1.0576 s / it) +[14:48:01.484838] Averaged stats: lr: 0.000103 loss: 0.4005 (0.4005) +[14:48:02.627716] val: [0/5] eta: 0:00:05 loss: 0.2545 (0.2545) time: 1.0864 data: 1.0730 max mem: 5538 +[14:48:02.715841] val: [4/5] eta: 0:00:00 loss: 0.2698 (0.4991) time: 0.2348 data: 0.2251 max mem: 5538 +[14:48:02.793581] val: Total time: 0:00:01 (0.2506 s / it) +[14:48:02.803191] val loss: 0.4991119384765625 +[14:48:02.803378] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8190, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7090, Kappa: 0.0000, Score: 0.4185 +[14:48:02.867056] Best epoch = 37, Best score = 0.4197 +[14:48:03.187556] log_dir: ./output_logs/retfound +[14:48:04.221500] Epoch: [39] [0/1] eta: 0:00:01 lr: 0.000099 loss: 0.5305 (0.5305) time: 1.0330 data: 0.9829 max mem: 5538 +[14:48:04.293010] Epoch: [39] Total time: 0:00:01 (1.1053 s / it) +[14:48:04.293759] Averaged stats: lr: 0.000099 loss: 0.5305 (0.5305) +[14:48:05.293493] val: [0/5] eta: 0:00:04 loss: 0.2578 (0.2578) time: 0.9736 data: 0.9594 max mem: 5538 +[14:48:05.445804] val: [4/5] eta: 0:00:00 loss: 0.2735 (0.4983) time: 0.2251 data: 0.2145 max mem: 5538 +[14:48:05.524265] val: Total time: 0:00:01 (0.2410 s / it) +[14:48:05.533716] val loss: 0.49827880859375 +[14:48:05.534179] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8181, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7068, Kappa: 0.0000, Score: 0.4182 +[14:48:05.621224] Best epoch = 37, Best score = 0.4197 +[14:48:05.934480] log_dir: ./output_logs/retfound +[14:48:06.938268] Epoch: [40] [0/1] eta: 0:00:01 lr: 0.000096 loss: 0.6786 (0.6786) time: 1.0029 data: 0.9496 max mem: 5538 +[14:48:07.012239] Epoch: [40] Total time: 0:00:01 (1.0776 s / it) +[14:48:07.013017] Averaged stats: lr: 0.000096 loss: 0.6786 (0.6786) +[14:48:07.933571] val: [0/5] eta: 0:00:04 loss: 0.2617 (0.2617) time: 0.9036 data: 0.8905 max mem: 5538 +[14:48:08.025567] val: [4/5] eta: 0:00:00 loss: 0.2777 (0.4977) time: 0.1990 data: 0.1899 max mem: 5538 +[14:48:08.101912] val: Total time: 0:00:01 (0.2145 s / it) +[14:48:08.113362] val loss: 0.497674560546875 +[14:48:08.113590] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7029, Kappa: 0.0000, Score: 0.4176 +[14:48:08.187330] Best epoch = 37, Best score = 0.4197 +[14:48:08.553765] log_dir: ./output_logs/retfound +[14:48:09.500583] Epoch: [41] [0/1] eta: 0:00:00 lr: 0.000092 loss: 0.2126 (0.2126) time: 0.9458 data: 0.8906 max mem: 5538 +[14:48:09.574019] Epoch: [41] Total time: 0:00:01 (1.0201 s / it) +[14:48:09.574783] Averaged stats: lr: 0.000092 loss: 0.2126 (0.2126) +[14:48:10.654775] val: [0/5] eta: 0:00:05 loss: 0.2628 (0.2628) time: 1.0488 data: 1.0338 max mem: 5538 +[14:48:10.710225] val: [4/5] eta: 0:00:00 loss: 0.2791 (0.4967) time: 0.2207 data: 0.2112 max mem: 5538 +[14:48:10.789828] val: Total time: 0:00:01 (0.2369 s / it) +[14:48:10.800889] val loss: 0.4966705322265625 +[14:48:10.801082] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7034, Kappa: 0.0000, Score: 0.4179 +[14:48:10.873510] Best epoch = 37, Best score = 0.4197 +[14:48:11.212165] log_dir: ./output_logs/retfound +[14:48:12.283019] Epoch: [42] [0/1] eta: 0:00:01 lr: 0.000089 loss: 0.4028 (0.4028) time: 1.0698 data: 1.0199 max mem: 5538 +[14:48:12.359495] Epoch: [42] Total time: 0:00:01 (1.1471 s / it) +[14:48:12.360227] Averaged stats: lr: 0.000089 loss: 0.4028 (0.4028) +[14:48:13.403360] val: [0/5] eta: 0:00:05 loss: 0.2629 (0.2629) time: 1.0176 data: 1.0044 max mem: 5538 +[14:48:13.580723] val: [4/5] eta: 0:00:00 loss: 0.2795 (0.4957) time: 0.2388 data: 0.2287 max mem: 5538 +[14:48:13.655631] val: Total time: 0:00:01 (0.2541 s / it) +[14:48:13.664672] val loss: 0.4956756591796875 +[14:48:13.664862] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7003, Kappa: 0.0000, Score: 0.4176 +[14:48:13.715137] Best epoch = 37, Best score = 0.4197 +[14:48:14.003034] log_dir: ./output_logs/retfound +[14:48:14.936508] Epoch: [43] [0/1] eta: 0:00:00 lr: 0.000086 loss: 0.5175 (0.5175) time: 0.9325 data: 0.8823 max mem: 5538 +[14:48:15.011916] Epoch: [43] Total time: 0:00:01 (1.0087 s / it) +[14:48:15.012667] Averaged stats: lr: 0.000086 loss: 0.5175 (0.5175) +[14:48:16.001483] val: [0/5] eta: 0:00:04 loss: 0.2632 (0.2632) time: 0.9766 data: 0.9633 max mem: 5538 +[14:48:16.072969] val: [4/5] eta: 0:00:00 loss: 0.2802 (0.4947) time: 0.2095 data: 0.2003 max mem: 5538 +[14:48:16.150294] val: Total time: 0:00:01 (0.2252 s / it) +[14:48:16.159065] val loss: 0.494671630859375 +[14:48:16.159264] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7030, Kappa: 0.0000, Score: 0.4176 +[14:48:16.245233] Best epoch = 37, Best score = 0.4197 +[14:48:16.566308] log_dir: ./output_logs/retfound +[14:48:17.576862] Epoch: [44] [0/1] eta: 0:00:01 lr: 0.000082 loss: 0.2074 (0.2074) time: 1.0095 data: 0.9567 max mem: 5538 +[14:48:17.653711] Epoch: [44] Total time: 0:00:01 (1.0872 s / it) +[14:48:17.654707] Averaged stats: lr: 0.000082 loss: 0.2074 (0.2074) +[14:48:18.699002] val: [0/5] eta: 0:00:05 loss: 0.2614 (0.2614) time: 1.0001 data: 0.9868 max mem: 5538 +[14:48:18.763864] val: [4/5] eta: 0:00:00 loss: 0.2787 (0.4934) time: 0.2129 data: 0.2037 max mem: 5538 +[14:48:18.847179] val: Total time: 0:00:01 (0.2298 s / it) +[14:48:18.856229] val loss: 0.493377685546875 +[14:48:18.856528] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7029, Kappa: 0.0000, Score: 0.4176 +[14:48:18.937817] Best epoch = 37, Best score = 0.4197 +[14:48:19.274747] log_dir: ./output_logs/retfound +[14:48:20.244982] Epoch: [45] [0/1] eta: 0:00:00 lr: 0.000079 loss: 0.3496 (0.3496) time: 0.9693 data: 0.9142 max mem: 5538 +[14:48:20.316626] Epoch: [45] Total time: 0:00:01 (1.0417 s / it) +[14:48:20.317417] Averaged stats: lr: 0.000079 loss: 0.3496 (0.3496) +[14:48:21.358457] val: [0/5] eta: 0:00:04 loss: 0.2590 (0.2590) time: 0.9892 data: 0.9759 max mem: 5538 +[14:48:21.455739] val: [4/5] eta: 0:00:00 loss: 0.2765 (0.4921) time: 0.2172 data: 0.2079 max mem: 5538 +[14:48:21.532305] val: Total time: 0:00:01 (0.2327 s / it) +[14:48:21.540968] val loss: 0.4920562744140625 +[14:48:21.541215] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7029, Kappa: 0.0000, Score: 0.4176 +[14:48:21.632466] Best epoch = 37, Best score = 0.4197 +[14:48:21.900296] log_dir: ./output_logs/retfound +[14:48:22.859958] Epoch: [46] [0/1] eta: 0:00:00 lr: 0.000075 loss: 0.3786 (0.3786) time: 0.9586 data: 0.9040 max mem: 5538 +[14:48:22.936533] Epoch: [46] Total time: 0:00:01 (1.0360 s / it) +[14:48:22.937304] Averaged stats: lr: 0.000075 loss: 0.3786 (0.3786) +[14:48:23.928887] val: [0/5] eta: 0:00:04 loss: 0.2559 (0.2559) time: 0.9756 data: 0.9627 max mem: 5538 +[14:48:24.029557] val: [4/5] eta: 0:00:00 loss: 0.2738 (0.4907) time: 0.2151 data: 0.2061 max mem: 5538 +[14:48:24.105034] val: Total time: 0:00:01 (0.2305 s / it) +[14:48:24.113748] val loss: 0.490740966796875 +[14:48:24.113961] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6998, Kappa: 0.0000, Score: 0.4176 +[14:48:24.175707] Best epoch = 37, Best score = 0.4197 +[14:48:24.435086] log_dir: ./output_logs/retfound +[14:48:25.393868] Epoch: [47] [0/1] eta: 0:00:00 lr: 0.000072 loss: 0.4682 (0.4682) time: 0.9578 data: 0.9055 max mem: 5538 +[14:48:25.465309] Epoch: [47] Total time: 0:00:01 (1.0301 s / it) +[14:48:25.466116] Averaged stats: lr: 0.000072 loss: 0.4682 (0.4682) +[14:48:26.637575] val: [0/5] eta: 0:00:05 loss: 0.2529 (0.2529) time: 1.1290 data: 1.1156 max mem: 5538 +[14:48:26.682119] val: [4/5] eta: 0:00:00 loss: 0.2711 (0.4893) time: 0.2346 data: 0.2251 max mem: 5538 +[14:48:26.759033] val: Total time: 0:00:01 (0.2502 s / it) +[14:48:26.768089] val loss: 0.489349365234375 +[14:48:26.768337] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8181, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7019, Kappa: 0.0000, Score: 0.4182 +[14:48:26.820551] Best epoch = 37, Best score = 0.4197 +[14:48:27.085425] log_dir: ./output_logs/retfound +[14:48:28.070489] Epoch: [48] [0/1] eta: 0:00:00 lr: 0.000068 loss: 0.5128 (0.5128) time: 0.9844 data: 0.9186 max mem: 5538 +[14:48:28.212928] Epoch: [48] Total time: 0:00:01 (1.1273 s / it) +[14:48:28.214228] Averaged stats: lr: 0.000068 loss: 0.5128 (0.5128) +[14:48:29.280440] val: [0/5] eta: 0:00:05 loss: 0.2504 (0.2504) time: 1.0174 data: 0.9994 max mem: 5538 +[14:48:29.381916] val: [4/5] eta: 0:00:00 loss: 0.2690 (0.4881) time: 0.2237 data: 0.2130 max mem: 5538 +[14:48:29.455961] val: Total time: 0:00:01 (0.2387 s / it) +[14:48:29.470178] val loss: 0.488134765625 +[14:48:29.470497] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6998, Kappa: 0.0000, Score: 0.4176 +[14:48:29.542555] Best epoch = 37, Best score = 0.4197 +[14:48:29.853539] log_dir: ./output_logs/retfound +[14:48:30.795231] Epoch: [49] [0/1] eta: 0:00:00 lr: 0.000065 loss: 0.3080 (0.3080) time: 0.9407 data: 0.8910 max mem: 5538 +[14:48:30.861352] Epoch: [49] Total time: 0:00:01 (1.0076 s / it) +[14:48:30.862135] Averaged stats: lr: 0.000065 loss: 0.3080 (0.3080) +[14:48:31.934283] val: [0/5] eta: 0:00:05 loss: 0.2478 (0.2478) time: 1.0322 data: 1.0191 max mem: 5538 +[14:48:32.001974] val: [4/5] eta: 0:00:00 loss: 0.2667 (0.4869) time: 0.2199 data: 0.2106 max mem: 5538 +[14:48:32.084950] val: Total time: 0:00:01 (0.2367 s / it) +[14:48:32.093705] val loss: 0.4869110107421875 +[14:48:32.093894] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8181, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7019, Kappa: 0.0000, Score: 0.4182 +[14:48:32.184377] Best epoch = 37, Best score = 0.4197 +[14:48:32.505148] log_dir: ./output_logs/retfound +[14:48:33.527779] Epoch: [50] [0/1] eta: 0:00:01 lr: 0.000061 loss: 0.6631 (0.6631) time: 1.0215 data: 0.9528 max mem: 5538 +[14:48:33.599549] Epoch: [50] Total time: 0:00:01 (1.0942 s / it) +[14:48:33.600309] Averaged stats: lr: 0.000061 loss: 0.6631 (0.6631) +[14:48:34.663633] val: [0/5] eta: 0:00:05 loss: 0.2464 (0.2464) time: 1.0379 data: 1.0249 max mem: 5538 +[14:48:34.799315] val: [4/5] eta: 0:00:00 loss: 0.2656 (0.4859) time: 0.2346 data: 0.2235 max mem: 5538 +[14:48:34.869566] val: Total time: 0:00:01 (0.2489 s / it) +[14:48:34.878255] val loss: 0.485931396484375 +[14:48:34.878496] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8199, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7055, Kappa: 0.0000, Score: 0.4188 +[14:48:34.939657] Best epoch = 37, Best score = 0.4197 +[14:48:35.287556] log_dir: ./output_logs/retfound +[14:48:36.219270] Epoch: [51] [0/1] eta: 0:00:00 lr: 0.000058 loss: 0.4841 (0.4841) time: 0.9305 data: 0.8762 max mem: 5538 +[14:48:36.290998] Epoch: [51] Total time: 0:00:01 (1.0033 s / it) +[14:48:36.291758] Averaged stats: lr: 0.000058 loss: 0.4841 (0.4841) +[14:48:37.324007] val: [0/5] eta: 0:00:05 loss: 0.2454 (0.2454) time: 1.0055 data: 0.9923 max mem: 5538 +[14:48:37.431374] val: [4/5] eta: 0:00:00 loss: 0.2649 (0.4850) time: 0.2225 data: 0.2132 max mem: 5538 +[14:48:37.507917] val: Total time: 0:00:01 (0.2380 s / it) +[14:48:37.516629] val loss: 0.484954833984375 +[14:48:37.516833] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8181, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7047, Kappa: 0.0000, Score: 0.4182 +[14:48:37.619776] Best epoch = 37, Best score = 0.4197 +[14:48:37.872345] log_dir: ./output_logs/retfound +[14:48:38.726217] Epoch: [52] [0/1] eta: 0:00:00 lr: 0.000055 loss: 0.1913 (0.1913) time: 0.8530 data: 0.8031 max mem: 5538 +[14:48:38.792593] Epoch: [52] Total time: 0:00:00 (0.9201 s / it) +[14:48:38.793498] Averaged stats: lr: 0.000055 loss: 0.1913 (0.1913) +[14:48:39.724037] val: [0/5] eta: 0:00:04 loss: 0.2432 (0.2432) time: 0.9148 data: 0.9014 max mem: 5538 +[14:48:39.778646] val: [4/5] eta: 0:00:00 loss: 0.2630 (0.4839) time: 0.1938 data: 0.1844 max mem: 5538 +[14:48:39.849968] val: Total time: 0:00:01 (0.2083 s / it) +[14:48:39.858785] val loss: 0.4839019775390625 +[14:48:39.858987] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8190, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7025, Kappa: 0.0000, Score: 0.4185 +[14:48:39.913033] Best epoch = 37, Best score = 0.4197 +[14:48:40.173243] log_dir: ./output_logs/retfound +[14:48:41.243131] Epoch: [53] [0/1] eta: 0:00:01 lr: 0.000051 loss: 0.7012 (0.7012) time: 1.0690 data: 1.0192 max mem: 5538 +[14:48:41.309899] Epoch: [53] Total time: 0:00:01 (1.1365 s / it) +[14:48:41.310622] Averaged stats: lr: 0.000051 loss: 0.7012 (0.7012) +[14:48:42.389992] val: [0/5] eta: 0:00:05 loss: 0.2425 (0.2425) time: 1.0387 data: 1.0232 max mem: 5538 +[14:48:42.480067] val: [4/5] eta: 0:00:00 loss: 0.2626 (0.4831) time: 0.2256 data: 0.2161 max mem: 5538 +[14:48:42.559463] val: Total time: 0:00:01 (0.2418 s / it) +[14:48:42.568316] val loss: 0.483056640625 +[14:48:42.568495] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8190, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7020, Kappa: 0.0000, Score: 0.4185 +[14:48:42.668302] Best epoch = 37, Best score = 0.4197 +[14:48:42.981939] log_dir: ./output_logs/retfound +[14:48:43.959433] Epoch: [54] [0/1] eta: 0:00:00 lr: 0.000048 loss: 0.6509 (0.6509) time: 0.9767 data: 0.9247 max mem: 5538 +[14:48:44.042031] Epoch: [54] Total time: 0:00:01 (1.0599 s / it) +[14:48:44.042810] Averaged stats: lr: 0.000048 loss: 0.6509 (0.6509) +[14:48:45.153803] val: [0/5] eta: 0:00:05 loss: 0.2426 (0.2426) time: 1.0808 data: 1.0677 max mem: 5538 +[14:48:45.209441] val: [4/5] eta: 0:00:00 loss: 0.2633 (0.4823) time: 0.2272 data: 0.2180 max mem: 5538 +[14:48:45.281512] val: Total time: 0:00:01 (0.2418 s / it) +[14:48:45.290170] val loss: 0.4823028564453125 +[14:48:45.290347] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7032, Kappa: 0.0000, Score: 0.4176 +[14:48:45.403111] Best epoch = 37, Best score = 0.4197 +[14:48:45.676577] log_dir: ./output_logs/retfound +[14:48:46.688317] Epoch: [55] [0/1] eta: 0:00:01 lr: 0.000045 loss: 0.4634 (0.4634) time: 1.0109 data: 0.9584 max mem: 5538 +[14:48:46.764041] Epoch: [55] Total time: 0:00:01 (1.0873 s / it) +[14:48:46.765537] Averaged stats: lr: 0.000045 loss: 0.4634 (0.4634) +[14:48:47.791625] val: [0/5] eta: 0:00:04 loss: 0.2428 (0.2428) time: 0.9924 data: 0.9792 max mem: 5538 +[14:48:47.886426] val: [4/5] eta: 0:00:00 loss: 0.2639 (0.4816) time: 0.2173 data: 0.2082 max mem: 5538 +[14:48:47.958245] val: Total time: 0:00:01 (0.2319 s / it) +[14:48:47.966952] val loss: 0.481634521484375 +[14:48:47.967189] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7019, Kappa: 0.0000, Score: 0.4179 +[14:48:48.032922] Best epoch = 37, Best score = 0.4197 +[14:48:48.304523] log_dir: ./output_logs/retfound +[14:48:49.202710] Epoch: [56] [0/1] eta: 0:00:00 lr: 0.000042 loss: 0.3185 (0.3185) time: 0.8973 data: 0.8487 max mem: 5538 +[14:48:49.281394] Epoch: [56] Total time: 0:00:00 (0.9767 s / it) +[14:48:49.282130] Averaged stats: lr: 0.000042 loss: 0.3185 (0.3185) +[14:48:50.193207] val: [0/5] eta: 0:00:04 loss: 0.2426 (0.2426) time: 0.8807 data: 0.8675 max mem: 5538 +[14:48:50.349719] val: [4/5] eta: 0:00:00 loss: 0.2640 (0.4809) time: 0.2073 data: 0.1980 max mem: 5538 +[14:48:50.423719] val: Total time: 0:00:01 (0.2224 s / it) +[14:48:50.432492] val loss: 0.4809295654296875 +[14:48:50.432723] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6992, Kappa: 0.0000, Score: 0.4173 +[14:48:50.488700] Best epoch = 37, Best score = 0.4197 +[14:48:50.788857] log_dir: ./output_logs/retfound +[14:48:51.738234] Epoch: [57] [0/1] eta: 0:00:00 lr: 0.000039 loss: 0.6137 (0.6137) time: 0.9484 data: 0.8986 max mem: 5538 +[14:48:51.807099] Epoch: [57] Total time: 0:00:01 (1.0181 s / it) +[14:48:51.807888] Averaged stats: lr: 0.000039 loss: 0.6137 (0.6137) +[14:48:52.748022] val: [0/5] eta: 0:00:04 loss: 0.2430 (0.2430) time: 0.9181 data: 0.8969 max mem: 5538 +[14:48:52.823713] val: [4/5] eta: 0:00:00 loss: 0.2647 (0.4803) time: 0.1987 data: 0.1878 max mem: 5538 +[14:48:52.895886] val: Total time: 0:00:01 (0.2134 s / it) +[14:48:52.904543] val loss: 0.4802825927734375 +[14:48:52.904783] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6994, Kappa: 0.0000, Score: 0.4173 +[14:48:52.952841] Best epoch = 37, Best score = 0.4197 +[14:48:53.196529] log_dir: ./output_logs/retfound +[14:48:54.229944] Epoch: [58] [0/1] eta: 0:00:01 lr: 0.000036 loss: 0.3902 (0.3902) time: 1.0325 data: 0.9817 max mem: 5538 +[14:48:54.310932] Epoch: [58] Total time: 0:00:01 (1.1142 s / it) +[14:48:54.311666] Averaged stats: lr: 0.000036 loss: 0.3902 (0.3902) +[14:48:55.449064] val: [0/5] eta: 0:00:05 loss: 0.2428 (0.2428) time: 1.1110 data: 1.0980 max mem: 5538 +[14:48:55.483646] val: [4/5] eta: 0:00:00 loss: 0.2648 (0.4797) time: 0.2290 data: 0.2197 max mem: 5538 +[14:48:55.559472] val: Total time: 0:00:01 (0.2444 s / it) +[14:48:55.568288] val loss: 0.479693603515625 +[14:48:55.568535] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8127, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6958, Kappa: 0.0000, Score: 0.4164 +[14:48:55.681598] Best epoch = 37, Best score = 0.4197 +[14:48:55.958944] log_dir: ./output_logs/retfound +[14:48:56.863988] Epoch: [59] [0/1] eta: 0:00:00 lr: 0.000033 loss: 0.4711 (0.4711) time: 0.9041 data: 0.8526 max mem: 5538 +[14:48:56.934717] Epoch: [59] Total time: 0:00:00 (0.9756 s / it) +[14:48:56.935591] Averaged stats: lr: 0.000033 loss: 0.4711 (0.4711) +[14:48:57.855145] val: [0/5] eta: 0:00:04 loss: 0.2425 (0.2425) time: 0.8884 data: 0.8746 max mem: 5538 +[14:48:57.957813] val: [4/5] eta: 0:00:00 loss: 0.2648 (0.4792) time: 0.1981 data: 0.1887 max mem: 5538 +[14:48:58.034046] val: Total time: 0:00:01 (0.2136 s / it) +[14:48:58.042782] val loss: 0.4792266845703125 +[14:48:58.043021] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8118, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6957, Kappa: 0.0000, Score: 0.4161 +[14:48:58.117365] Best epoch = 37, Best score = 0.4197 +[14:48:58.407859] log_dir: ./output_logs/retfound +[14:48:59.311951] Epoch: [60] [0/1] eta: 0:00:00 lr: 0.000030 loss: 0.3593 (0.3593) time: 0.9032 data: 0.8494 max mem: 5538 +[14:48:59.383226] Epoch: [60] Total time: 0:00:00 (0.9752 s / it) +[14:48:59.383983] Averaged stats: lr: 0.000030 loss: 0.3593 (0.3593) +[14:49:00.348822] val: [0/5] eta: 0:00:04 loss: 0.2416 (0.2416) time: 0.9323 data: 0.9187 max mem: 5538 +[14:49:00.514743] val: [4/5] eta: 0:00:00 loss: 0.2641 (0.4788) time: 0.2195 data: 0.2096 max mem: 5538 +[14:49:00.591564] val: Total time: 0:00:01 (0.2351 s / it) +[14:49:00.600314] val loss: 0.4787872314453125 +[14:49:00.600499] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8109, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6952, Kappa: 0.0000, Score: 0.4159 +[14:49:00.671622] Best epoch = 37, Best score = 0.4197 +[14:49:00.972345] log_dir: ./output_logs/retfound +[14:49:01.947118] Epoch: [61] [0/1] eta: 0:00:00 lr: 0.000028 loss: 0.2544 (0.2544) time: 0.9739 data: 0.9228 max mem: 5538 +[14:49:02.023323] Epoch: [61] Total time: 0:00:01 (1.0508 s / it) +[14:49:02.024083] Averaged stats: lr: 0.000028 loss: 0.2544 (0.2544) +[14:49:03.075971] val: [0/5] eta: 0:00:05 loss: 0.2401 (0.2401) time: 1.0242 data: 1.0092 max mem: 5538 +[14:49:03.206322] val: [4/5] eta: 0:00:00 loss: 0.2627 (0.4783) time: 0.2308 data: 0.2212 max mem: 5538 +[14:49:03.284480] val: Total time: 0:00:01 (0.2467 s / it) +[14:49:03.293426] val loss: 0.47832489013671875 +[14:49:03.293663] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6934, Kappa: 0.0000, Score: 0.4156 +[14:49:03.349329] Best epoch = 37, Best score = 0.4197 +[14:49:03.670563] log_dir: ./output_logs/retfound +[14:49:04.638345] Epoch: [62] [0/1] eta: 0:00:00 lr: 0.000025 loss: 0.5358 (0.5358) time: 0.9666 data: 0.9064 max mem: 5538 +[14:49:04.724014] Epoch: [62] Total time: 0:00:01 (1.0533 s / it) +[14:49:04.724879] Averaged stats: lr: 0.000025 loss: 0.5358 (0.5358) +[14:49:05.678525] val: [0/5] eta: 0:00:04 loss: 0.2389 (0.2389) time: 0.9328 data: 0.9154 max mem: 5538 +[14:49:05.780810] val: [4/5] eta: 0:00:00 loss: 0.2615 (0.4779) time: 0.2069 data: 0.1958 max mem: 5538 +[14:49:05.859410] val: Total time: 0:00:01 (0.2229 s / it) +[14:49:05.868929] val loss: 0.47794647216796876 +[14:49:05.869149] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8118, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6934, Kappa: 0.0000, Score: 0.4161 +[14:49:05.907605] Best epoch = 37, Best score = 0.4197 +[14:49:06.163499] log_dir: ./output_logs/retfound +[14:49:07.090095] Epoch: [63] [0/1] eta: 0:00:00 lr: 0.000023 loss: 0.4145 (0.4145) time: 0.9257 data: 0.8735 max mem: 5538 +[14:49:07.161865] Epoch: [63] Total time: 0:00:00 (0.9982 s / it) +[14:49:07.162639] Averaged stats: lr: 0.000023 loss: 0.4145 (0.4145) +[14:49:08.055931] val: [0/5] eta: 0:00:04 loss: 0.2378 (0.2378) time: 0.8659 data: 0.8529 max mem: 5538 +[14:49:08.178248] val: [4/5] eta: 0:00:00 loss: 0.2606 (0.4776) time: 0.1975 data: 0.1881 max mem: 5538 +[14:49:08.254883] val: Total time: 0:00:01 (0.2131 s / it) +[14:49:08.263728] val loss: 0.4776092529296875 +[14:49:08.263935] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8109, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6937, Kappa: 0.0000, Score: 0.4159 +[14:49:08.316391] Best epoch = 37, Best score = 0.4197 +[14:49:08.675145] log_dir: ./output_logs/retfound +[14:49:09.657247] Epoch: [64] [0/1] eta: 0:00:00 lr: 0.000020 loss: 0.4393 (0.4393) time: 0.9811 data: 0.9267 max mem: 5538 +[14:49:09.730128] Epoch: [64] Total time: 0:00:01 (1.0548 s / it) +[14:49:09.730906] Averaged stats: lr: 0.000020 loss: 0.4393 (0.4393) +[14:49:10.807185] val: [0/5] eta: 0:00:05 loss: 0.2367 (0.2367) time: 1.0100 data: 0.9973 max mem: 5538 +[14:49:10.964638] val: [4/5] eta: 0:00:00 loss: 0.2595 (0.4774) time: 0.2334 data: 0.2241 max mem: 5538 +[14:49:11.048730] val: Total time: 0:00:01 (0.2504 s / it) +[14:49:11.058906] val loss: 0.477386474609375 +[14:49:11.059127] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8118, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6971, Kappa: 0.0000, Score: 0.4161 +[14:49:11.137582] Best epoch = 37, Best score = 0.4197 +[14:49:11.459967] log_dir: ./output_logs/retfound +[14:49:12.570611] Epoch: [65] [0/1] eta: 0:00:01 lr: 0.000018 loss: 0.4860 (0.4860) time: 1.1097 data: 1.0603 max mem: 5538 +[14:49:12.637817] Epoch: [65] Total time: 0:00:01 (1.1776 s / it) +[14:49:12.638573] Averaged stats: lr: 0.000018 loss: 0.4860 (0.4860) +[14:49:13.755580] val: [0/5] eta: 0:00:05 loss: 0.2358 (0.2358) time: 1.0492 data: 1.0363 max mem: 5538 +[14:49:13.837070] val: [4/5] eta: 0:00:00 loss: 0.2587 (0.4772) time: 0.2260 data: 0.2168 max mem: 5538 +[14:49:13.912115] val: Total time: 0:00:01 (0.2413 s / it) +[14:49:13.920798] val loss: 0.47721099853515625 +[14:49:13.921037] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7070, Kappa: 0.0000, Score: 0.4176 +[14:49:13.981561] Best epoch = 37, Best score = 0.4197 +[14:49:14.237363] log_dir: ./output_logs/retfound +[14:49:15.179608] Epoch: [66] [0/1] eta: 0:00:00 lr: 0.000016 loss: 0.4881 (0.4881) time: 0.9414 data: 0.8910 max mem: 5538 +[14:49:15.246213] Epoch: [66] Total time: 0:00:01 (1.0087 s / it) +[14:49:15.246937] Averaged stats: lr: 0.000016 loss: 0.4881 (0.4881) +[14:49:16.249594] val: [0/5] eta: 0:00:04 loss: 0.2352 (0.2352) time: 0.9749 data: 0.9615 max mem: 5538 +[14:49:16.373377] val: [4/5] eta: 0:00:00 loss: 0.2581 (0.4771) time: 0.2196 data: 0.2104 max mem: 5538 +[14:49:16.449967] val: Total time: 0:00:01 (0.2352 s / it) +[14:49:16.458663] val loss: 0.4770660400390625 +[14:49:16.458873] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8136, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7063, Kappa: 0.0000, Score: 0.4167 +[14:49:16.499691] Best epoch = 37, Best score = 0.4197 +[14:49:16.759040] log_dir: ./output_logs/retfound +[14:49:17.852821] Epoch: [67] [0/1] eta: 0:00:01 lr: 0.000014 loss: 0.5014 (0.5014) time: 1.0927 data: 1.0315 max mem: 5538 +[14:49:17.924563] Epoch: [67] Total time: 0:00:01 (1.1654 s / it) +[14:49:17.925361] Averaged stats: lr: 0.000014 loss: 0.5014 (0.5014) +[14:49:18.986280] val: [0/5] eta: 0:00:05 loss: 0.2348 (0.2348) time: 1.0383 data: 1.0252 max mem: 5538 +[14:49:19.031982] val: [4/5] eta: 0:00:00 loss: 0.2578 (0.4769) time: 0.2167 data: 0.2076 max mem: 5538 +[14:49:19.107058] val: Total time: 0:00:01 (0.2319 s / it) +[14:49:19.116906] val loss: 0.4768798828125 +[14:49:19.117100] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7072, Kappa: 0.0000, Score: 0.4176 +[14:49:19.155966] Best epoch = 37, Best score = 0.4197 +[14:49:19.452192] log_dir: ./output_logs/retfound +[14:49:20.368526] Epoch: [68] [0/1] eta: 0:00:00 lr: 0.000012 loss: 0.5719 (0.5719) time: 0.9155 data: 0.8663 max mem: 5538 +[14:49:20.439403] Epoch: [68] Total time: 0:00:00 (0.9871 s / it) +[14:49:20.440131] Averaged stats: lr: 0.000012 loss: 0.5719 (0.5719) +[14:49:21.367039] val: [0/5] eta: 0:00:04 loss: 0.2346 (0.2346) time: 0.8694 data: 0.8562 max mem: 5538 +[14:49:21.453007] val: [4/5] eta: 0:00:00 loss: 0.2576 (0.4768) time: 0.1910 data: 0.1815 max mem: 5538 +[14:49:21.536354] val: Total time: 0:00:01 (0.2079 s / it) +[14:49:21.546868] val loss: 0.47679443359375 +[14:49:21.547115] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7070, Kappa: 0.0000, Score: 0.4173 +[14:49:21.615711] Best epoch = 37, Best score = 0.4197 +[14:49:21.865267] log_dir: ./output_logs/retfound +[14:49:22.821246] Epoch: [69] [0/1] eta: 0:00:00 lr: 0.000010 loss: 0.3382 (0.3382) time: 0.9551 data: 0.9041 max mem: 5538 +[14:49:22.891594] Epoch: [69] Total time: 0:00:01 (1.0261 s / it) +[14:49:22.892367] Averaged stats: lr: 0.000010 loss: 0.3382 (0.3382) +[14:49:23.897822] val: [0/5] eta: 0:00:04 loss: 0.2342 (0.2342) time: 0.9847 data: 0.9714 max mem: 5538 +[14:49:23.981006] val: [4/5] eta: 0:00:00 loss: 0.2573 (0.4767) time: 0.2135 data: 0.2043 max mem: 5538 +[14:49:24.058725] val: Total time: 0:00:01 (0.2292 s / it) +[14:49:24.067540] val loss: 0.47665557861328123 +[14:49:24.067780] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7094, Kappa: 0.0000, Score: 0.4179 +[14:49:24.143229] Best epoch = 37, Best score = 0.4197 +[14:49:24.415733] log_dir: ./output_logs/retfound +[14:49:25.355687] Epoch: [70] [0/1] eta: 0:00:00 lr: 0.000009 loss: 0.8328 (0.8328) time: 0.9390 data: 0.8900 max mem: 5538 +[14:49:25.430192] Epoch: [70] Total time: 0:00:01 (1.0143 s / it) +[14:49:25.430965] Averaged stats: lr: 0.000009 loss: 0.8328 (0.8328) +[14:49:26.445036] val: [0/5] eta: 0:00:04 loss: 0.2343 (0.2343) time: 0.9654 data: 0.9517 max mem: 5538 +[14:49:26.506485] val: [4/5] eta: 0:00:00 loss: 0.2575 (0.4766) time: 0.2053 data: 0.1961 max mem: 5538 +[14:49:26.581620] val: Total time: 0:00:01 (0.2205 s / it) +[14:49:26.590343] val loss: 0.47664031982421873 +[14:49:26.590569] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7094, Kappa: 0.0000, Score: 0.4179 +[14:49:26.654116] Best epoch = 37, Best score = 0.4197 +[14:49:26.942460] log_dir: ./output_logs/retfound +[14:49:27.942628] Epoch: [71] [0/1] eta: 0:00:00 lr: 0.000007 loss: 0.5217 (0.5217) time: 0.9992 data: 0.9437 max mem: 5538 +[14:49:28.016709] Epoch: [71] Total time: 0:00:01 (1.0741 s / it) +[14:49:28.017533] Averaged stats: lr: 0.000007 loss: 0.5217 (0.5217) +[14:49:29.208603] val: [0/5] eta: 0:00:05 loss: 0.2343 (0.2343) time: 1.1678 data: 1.1543 max mem: 5538 +[14:49:29.320629] val: [4/5] eta: 0:00:00 loss: 0.2575 (0.4766) time: 0.2558 data: 0.2462 max mem: 5538 +[14:49:29.395210] val: Total time: 0:00:01 (0.2710 s / it) +[14:49:29.403875] val loss: 0.47658538818359375 +[14:49:29.404112] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7088, Kappa: 0.0000, Score: 0.4176 +[14:49:29.490625] Best epoch = 37, Best score = 0.4197 +[14:49:29.787725] log_dir: ./output_logs/retfound +[14:49:30.735759] Epoch: [72] [0/1] eta: 0:00:00 lr: 0.000006 loss: 0.7527 (0.7527) time: 0.9470 data: 0.8931 max mem: 5538 +[14:49:30.805652] Epoch: [72] Total time: 0:00:01 (1.0178 s / it) +[14:49:30.806381] Averaged stats: lr: 0.000006 loss: 0.7527 (0.7527) +[14:49:31.734599] val: [0/5] eta: 0:00:04 loss: 0.2344 (0.2344) time: 0.8950 data: 0.8813 max mem: 5538 +[14:49:31.820941] val: [4/5] eta: 0:00:00 loss: 0.2577 (0.4766) time: 0.1962 data: 0.1868 max mem: 5538 +[14:49:31.901346] val: Total time: 0:00:01 (0.2125 s / it) +[14:49:31.910014] val loss: 0.47655792236328126 +[14:49:31.910212] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7070, Kappa: 0.0000, Score: 0.4173 +[14:49:31.976532] Best epoch = 37, Best score = 0.4197 +[14:49:32.298268] log_dir: ./output_logs/retfound +[14:49:33.216584] Epoch: [73] [0/1] eta: 0:00:00 lr: 0.000005 loss: 0.5015 (0.5015) time: 0.9174 data: 0.8660 max mem: 5538 +[14:49:33.285174] Epoch: [73] Total time: 0:00:00 (0.9867 s / it) +[14:49:33.285924] Averaged stats: lr: 0.000005 loss: 0.5015 (0.5015) +[14:49:34.302628] val: [0/5] eta: 0:00:04 loss: 0.2346 (0.2346) time: 0.9848 data: 0.9713 max mem: 5538 +[14:49:34.392572] val: [4/5] eta: 0:00:00 loss: 0.2579 (0.4766) time: 0.2148 data: 0.2053 max mem: 5538 +[14:49:34.466393] val: Total time: 0:00:01 (0.2298 s / it) +[14:49:34.475278] val loss: 0.47658843994140626 +[14:49:34.475508] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7070, Kappa: 0.0000, Score: 0.4173 +[14:49:34.580189] Best epoch = 37, Best score = 0.4197 +[14:49:34.905526] log_dir: ./output_logs/retfound +[14:49:36.069907] Epoch: [74] [0/1] eta: 0:00:01 lr: 0.000004 loss: 0.4526 (0.4526) time: 1.1634 data: 1.1130 max mem: 5538 +[14:49:36.134411] Epoch: [74] Total time: 0:00:01 (1.2287 s / it) +[14:49:36.135121] Averaged stats: lr: 0.000004 loss: 0.4526 (0.4526) +[14:49:37.196527] val: [0/5] eta: 0:00:05 loss: 0.2347 (0.2347) time: 1.0287 data: 1.0138 max mem: 5538 +[14:49:37.257342] val: [4/5] eta: 0:00:00 loss: 0.2580 (0.4765) time: 0.2178 data: 0.2073 max mem: 5538 +[14:49:37.335318] val: Total time: 0:00:01 (0.2337 s / it) +[14:49:37.344006] val loss: 0.47654571533203127 +[14:49:37.344259] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7094, Kappa: 0.0000, Score: 0.4179 +[14:49:37.461561] Best epoch = 37, Best score = 0.4197 +[14:49:37.769578] log_dir: ./output_logs/retfound +[14:49:38.808673] Epoch: [75] [0/1] eta: 0:00:01 lr: 0.000003 loss: 0.4839 (0.4839) time: 1.0382 data: 0.9889 max mem: 5538 +[14:49:38.883463] Epoch: [75] Total time: 0:00:01 (1.1137 s / it) +[14:49:38.884229] Averaged stats: lr: 0.000003 loss: 0.4839 (0.4839) +[14:49:39.889978] val: [0/5] eta: 0:00:04 loss: 0.2348 (0.2348) time: 0.9831 data: 0.9701 max mem: 5538 +[14:49:39.960946] val: [4/5] eta: 0:00:00 loss: 0.2580 (0.4765) time: 0.2107 data: 0.2016 max mem: 5538 +[14:49:40.036962] val: Total time: 0:00:01 (0.2261 s / it) +[14:49:40.047627] val loss: 0.47653350830078123 +[14:49:40.048141] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7075, Kappa: 0.0000, Score: 0.4176 +[14:49:40.118413] Best epoch = 37, Best score = 0.4197 +[14:49:40.407595] log_dir: ./output_logs/retfound +[14:49:41.353368] Epoch: [76] [0/1] eta: 0:00:00 lr: 0.000002 loss: 0.6469 (0.6469) time: 0.9448 data: 0.8949 max mem: 5538 +[14:49:41.422168] Epoch: [76] Total time: 0:00:01 (1.0144 s / it) +[14:49:41.422984] Averaged stats: lr: 0.000002 loss: 0.6469 (0.6469) +[14:49:42.289607] val: [0/5] eta: 0:00:04 loss: 0.2348 (0.2348) time: 0.8480 data: 0.8334 max mem: 5538 +[14:49:42.411383] val: [4/5] eta: 0:00:00 loss: 0.2582 (0.4765) time: 0.1938 data: 0.1832 max mem: 5538 +[14:49:42.487395] val: Total time: 0:00:01 (0.2093 s / it) +[14:49:42.497264] val loss: 0.47650909423828125 +[14:49:42.497591] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7075, Kappa: 0.0000, Score: 0.4176 +[14:49:42.576541] Best epoch = 37, Best score = 0.4197 +[14:49:42.888306] log_dir: ./output_logs/retfound +[14:49:43.839155] Epoch: [77] [0/1] eta: 0:00:00 lr: 0.000002 loss: 0.3497 (0.3497) time: 0.9500 data: 0.8991 max mem: 5538 +[14:49:43.916950] Epoch: [77] Total time: 0:00:01 (1.0285 s / it) +[14:49:43.917763] Averaged stats: lr: 0.000002 loss: 0.3497 (0.3497) +[14:49:44.973070] val: [0/5] eta: 0:00:05 loss: 0.2348 (0.2348) time: 1.0222 data: 1.0083 max mem: 5538 +[14:49:45.066873] val: [4/5] eta: 0:00:00 loss: 0.2582 (0.4765) time: 0.2231 data: 0.2138 max mem: 5538 +[14:49:45.137880] val: Total time: 0:00:01 (0.2375 s / it) +[14:49:45.146938] val loss: 0.47650909423828125 +[14:49:45.147116] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7075, Kappa: 0.0000, Score: 0.4176 +[14:49:45.249627] Best epoch = 37, Best score = 0.4197 +[14:49:45.571564] log_dir: ./output_logs/retfound +[14:49:46.541822] Epoch: [78] [0/1] eta: 0:00:00 lr: 0.000001 loss: 0.3069 (0.3069) time: 0.9693 data: 0.9176 max mem: 5538 +[14:49:46.607954] Epoch: [78] Total time: 0:00:01 (1.0362 s / it) +[14:49:46.609078] Averaged stats: lr: 0.000001 loss: 0.3069 (0.3069) +[14:49:47.742443] val: [0/5] eta: 0:00:05 loss: 0.2348 (0.2348) time: 1.0592 data: 1.0463 max mem: 5538 +[14:49:47.858774] val: [4/5] eta: 0:00:00 loss: 0.2582 (0.4765) time: 0.2350 data: 0.2258 max mem: 5538 +[14:49:47.937225] val: Total time: 0:00:01 (0.2509 s / it) +[14:49:47.947948] val loss: 0.47648162841796876 +[14:49:47.948140] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7075, Kappa: 0.0000, Score: 0.4176 +[14:49:48.037953] Best epoch = 37, Best score = 0.4197 +[14:49:48.304982] log_dir: ./output_logs/retfound +[14:49:49.229358] Epoch: [79] [0/1] eta: 0:00:00 lr: 0.000001 loss: 0.4913 (0.4913) time: 0.9234 data: 0.8725 max mem: 5538 +[14:49:49.297489] Epoch: [79] Total time: 0:00:00 (0.9923 s / it) +[14:49:49.298401] Averaged stats: lr: 0.000001 loss: 0.4913 (0.4913) +[14:49:50.391489] val: [0/5] eta: 0:00:05 loss: 0.2348 (0.2348) time: 1.0712 data: 1.0576 max mem: 5538 +[14:49:50.425663] val: [4/5] eta: 0:00:00 loss: 0.2582 (0.4765) time: 0.2210 data: 0.2117 max mem: 5538 +[14:49:50.497226] val: Total time: 0:00:01 (0.2355 s / it) +[14:49:50.506080] val loss: 0.47645416259765627 +[14:49:50.506321] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8154, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7070, Kappa: 0.0000, Score: 0.4173 +[14:49:50.602896] Best epoch = 37, Best score = 0.4197 +[14:49:53.626516] Test with the best model, epoch = 37: +[14:49:54.716623] test: [ 0/10] eta: 0:00:10 loss: 0.2948 (0.2948) time: 1.0549 data: 1.0310 max mem: 5538 +[14:49:54.793208] test: [ 9/10] eta: 0:00:00 loss: 0.2768 (0.5067) time: 0.1131 data: 0.1032 max mem: 5538 +[14:49:54.863510] test: Total time: 0:00:01 (0.1202 s / it) +[14:49:54.872704] val loss: 0.5066802978515625 +[14:49:54.872832] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8004, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7511, Kappa: 0.0000, Score: 0.4123 +[14:49:55.609641] Training time 0:03:56 +[rank0]:[W701 14:49:56.078598890 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/005/retfound acc=0.7750 auroc=0.8006272401433692 f1_macro=0.4366 qwk=0.0 diff --git a/results/downsample/adam/005/vit/confusion_matrix.png b/results/downsample/adam/005/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..3cc7bd7f42e60ce371cbf2eb4bae1139dacbec4b --- /dev/null +++ b/results/downsample/adam/005/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2aca4c58da4fdc5fb484b1d175f2e04b558b09a729c439cf1413b6113da2f1d6 +size 66723 diff --git a/results/downsample/adam/005/vit/log.csv b/results/downsample/adam/005/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..0ad4b650575d6b6290de312a9a205b448c6b31c7 --- /dev/null +++ b/results/downsample/adam/005/vit/log.csv @@ -0,0 +1,20 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7596129775047302,0.3,0.3261648745519713,0.16738866653636775,0.0 +1,1.1559396982192993,0.3,0.3261648745519713,0.16738866653636775,7.39441178203743e-08 +2,1.3670729398727417,0.725,0.46953405017921146,0.3274082392402084,1.478882356407486e-07 +3,0.7330941557884216,0.675,0.6917562724014337,0.5277457918559767,2.2183235346112292e-07 +4,0.5740594863891602,0.775,0.7634408602150538,0.4660144836363019,2.957764712814972e-07 +5,0.4841795265674591,0.75,0.7455197132616487,0.4871851501913273,3.697205891018715e-07 +6,0.45928436517715454,0.725,0.7347670250896057,0.4630586769060412,3.6955843521557525e-07 +7,0.6498278975486755,0.7,0.7240143369175627,0.4778972520908004,3.6907225802970327e-07 +8,0.3159206807613373,0.75,0.7204301075268817,0.5742149823170078,3.682629104642438e-07 +9,0.5079060792922974,0.7,0.7132616487455197,0.4743130227001194,3.671318123898447e-07 +10,0.446988582611084,0.7,0.7132616487455197,0.43688650288923286,3.6568094813687817e-07 +11,0.45400136709213257,0.725,0.7168458781362006,0.4098455152258715,3.6391286301425095e-07 +12,0.30665355920791626,0.75,0.7132616487455197,0.4280328113879707,3.618306588440675e-07 +13,0.564050018787384,0.725,0.7204301075268817,0.41104025835609853,3.594379885199801e-07 +14,0.3932533264160156,0.725,0.7025089605734767,0.4050665427049635,3.567390495987718e-07 +15,0.4745815396308899,0.725,0.7025089605734767,0.4050665427049635,3.537385769364163e-07 +16,0.4452817440032959,0.725,0.6917562724014337,0.40148231331428247,3.504418343815308e-07 +17,0.501997709274292,0.75,0.6810035842293907,0.5040561558326244,3.4685460554079696e-07 +18,0.31084030866622925,0.7,0.6702508960573477,0.49096956778877265,3.4298318363255025e-07 diff --git a/results/downsample/adam/005/vit/metrics.json b/results/downsample/adam/005/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..3b46a5dcd2623688b67c4f9ca5f02e23798aaaf9 --- /dev/null +++ b/results/downsample/adam/005/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7125, + "balanced_accuracy": 0.6568100358422939, + "precision_macro": 0.6272727272727272, + "recall_macro": 0.6568100358422939, + "f1_macro": 0.6342675412442854, + "precision_weighted": 0.7522727272727272, + "recall_weighted": 0.7125, + "f1_weighted": 0.7273007354402703, + "cohen_kappa": 0.2755905511811023, + "quadratic_weighted_kappa": 0.2755905511811023, + "mcc": 0.28254302982294827, + "auroc": 0.7661290322580645, + "auprc": 0.43603463827400313, + "sensitivity": 0.5555555555555556, + "specificity": 0.7580645161290323, + "precision_pos": 0.4, + "f1_pos": 0.46511627906976744, + "per_class": { + "0": { + "precision": 0.8545454545454545, + "recall": 0.7580645161290323, + "f1-score": 0.8034188034188035, + "support": 62.0 + }, + "1": { + "precision": 0.4, + "recall": 0.5555555555555556, + "f1-score": 0.46511627906976744, + "support": 18.0 + }, + "accuracy": 0.7125, + "macro avg": { + "precision": 0.6272727272727272, + "recall": 0.6568100358422939, + "f1-score": 0.6342675412442854, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.7522727272727272, + "recall": 0.7125, + "f1-score": 0.7273007354402703, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/005/vit/pr.png b/results/downsample/adam/005/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..cad7015fef4138fdac5610a0db54eb3d76cf5c4e --- /dev/null +++ b/results/downsample/adam/005/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f96ba2b7154b6975a532be5a7069386537d284cb37af97a3fe85e509d1682c2 +size 55140 diff --git a/results/downsample/adam/005/vit/roc.png b/results/downsample/adam/005/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..306c6b5f8846eb5a8346ff36cdd708904fa1cc8f --- /dev/null +++ b/results/downsample/adam/005/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:397988ecbddcd73cf7fe34fa3e600cdaf9a45b975348e1a01db22ff308e65b9d +size 57624 diff --git a/results/downsample/adam/005/vit/test_pred.npz b/results/downsample/adam/005/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..07318386e03a74c26802e0e17563b2bdb9ed0d2f --- /dev/null +++ b/results/downsample/adam/005/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:879509f27bdee98854152b2e8e4089831e7796b71fc961bce99017df299b4173 +size 1790 diff --git a/results/downsample/adam/005/vit/train.log b/results/downsample/adam/005/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..01de3180d4b1a3266f1d4d727c32fa0e63de418e --- /dev/null +++ b/results/downsample/adam/005/vit/train.log @@ -0,0 +1,104 @@ +[vit] train=14 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7596 val_acc=0.3000 val_auc=0.3262 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=1.1559 val_acc=0.3000 val_auc=0.3262 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=1.3671 val_acc=0.7250 val_auc=0.4695 score=0.3274 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.7331 val_acc=0.6750 val_auc=0.6918 score=0.5277 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5741 val_acc=0.7750 val_auc=0.7634 score=0.4660 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.4842 val_acc=0.7500 val_auc=0.7455 score=0.4872 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.4593 val_acc=0.7250 val_auc=0.7348 score=0.4631 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.6498 val_acc=0.7000 val_auc=0.7240 score=0.4779 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3159 val_acc=0.7500 val_auc=0.7204 score=0.5742 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.5079 val_acc=0.7000 val_auc=0.7133 score=0.4743 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4470 val_acc=0.7000 val_auc=0.7133 score=0.4369 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4540 val_acc=0.7250 val_auc=0.7168 score=0.4098 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3067 val_acc=0.7500 val_auc=0.7133 score=0.4280 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.5641 val_acc=0.7250 val_auc=0.7204 score=0.4110 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.3933 val_acc=0.7250 val_auc=0.7025 score=0.4051 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4746 val_acc=0.7250 val_auc=0.7025 score=0.4051 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4453 val_acc=0.7250 val_auc=0.6918 score=0.4015 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.5020 val_acc=0.7500 val_auc=0.6810 score=0.5041 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.3108 val_acc=0.7000 val_auc=0.6703 score=0.4910 +[vit] early stop at ep18 (best ep8 score=0.5742) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=8 best_val_score=0.5742 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/005/vit acc=0.7125 auroc=0.7661290322580645 f1_macro=0.6343 qwk=0.2755905511811023 diff --git a/results/downsample/adam/010/resnet/confusion_matrix.png b/results/downsample/adam/010/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7338a9d7f628ed1c162c3e8a1ae4a57e090ed5c3 --- /dev/null +++ b/results/downsample/adam/010/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84ce0881512abb3061fe1d19062f50cddbd53eb9053be45fd0b142d019a2dafd +size 68274 diff --git a/results/downsample/adam/010/resnet/log.csv b/results/downsample/adam/010/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..8ffa9b59c7c9923e116f82b574366eb6d9e95887 --- /dev/null +++ b/results/downsample/adam/010/resnet/log.csv @@ -0,0 +1,26 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6868912577629089,0.45,0.42652329749103945,0.1654663579628517,0.0 +1,0.6759793162345886,0.4,0.34408602150537637,0.15027073340366318,0.00016666666666666666 +2,0.6899767518043518,0.4,0.37992831541218636,0.2154610935900513,0.0003333333333333333 +3,0.6854769587516785,0.425,0.4982078853046595,0.308985479219454,0.0005 +4,0.6854347586631775,0.425,0.5412186379928315,0.30400563413603915,0.0004997919498361457 +5,0.6477171182632446,0.55,0.5949820788530465,0.3916755709607522,0.0004991681456235483 +6,0.6523366570472717,0.725,0.6021505376344085,0.41885318108764213,0.000498129625622757 +7,0.6320070624351501,0.75,0.6129032258064516,0.44297965437292824,0.0004966781183478222 +8,0.6070477366447449,0.775,0.6379928315412187,0.47370112590382213,0.0004948160396893552 +9,0.6136426329612732,0.725,0.6774193548387096,0.33620009967336145,0.0004925464888935161 +10,0.5858384370803833,0.75,0.7347670250896057,0.4352012701693327,0.0004898732434036243 +11,0.6138298511505127,0.775,0.7670250896057348,0.5167118785919941,0.00048680075257297753 +12,0.5324985980987549,0.8,0.8243727598566308,0.5990671165688325,0.0004833341302593417 +13,0.5447873473167419,0.8,0.7992831541218639,0.5907039146572434,0.0004794791463134399 +14,0.4627349078655243,0.8,0.7741935483870968,0.6130418183339609,0.00047524221697560476 +15,0.5038831830024719,0.7,0.7706093189964158,0.5244223754351287,0.0004706303941965803 +16,0.4925979673862457,0.675,0.7168458781362007,0.5552109800436033,0.00046565135390024513 +17,0.46355220675468445,0.55,0.6451612903225806,0.4084019747839301,0.0004603133832077953 +18,0.42283251881599426,0.525,0.6308243727598566,0.38898788535312784,0.00045462536664464835 +19,0.4284207820892334,0.575,0.6236559139784946,0.39135142655258637,0.0004485967713530281 +20,0.37804368138313293,0.6,0.6344086021505376,0.41022729802533453,0.00044223763133484053 +21,0.3351643681526184,0.575,0.6308243727598566,0.36472409216490703,0.0004355585307510675 +22,0.2825261354446411,0.625,0.6379928315412187,0.3973976229307012,0.00042857058630547593 +23,0.32372090220451355,0.725,0.6379928315412187,0.468755436695622,0.00042128542874196107 +24,0.33558735251426697,0.75,0.6451612903225806,0.49210872453035437,0.0004137151834863213 diff --git a/results/downsample/adam/010/resnet/metrics.json b/results/downsample/adam/010/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..f0302a23e5a475648e90e183fbe34d99386014cb --- /dev/null +++ b/results/downsample/adam/010/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.775, + "balanced_accuracy": 0.6971326164874552, + "precision_macro": 0.6833333333333333, + "recall_macro": 0.6971326164874552, + "f1_macro": 0.6893874029335634, + "precision_weighted": 0.7841666666666667, + "recall_weighted": 0.775, + "f1_weighted": 0.7790767903364969, + "cohen_kappa": 0.3793103448275862, + "quadratic_weighted_kappa": 0.3793103448275862, + "mcc": 0.38021562140115595, + "auroc": 0.7706093189964158, + "auprc": 0.5517673728476293, + "sensitivity": 0.5555555555555556, + "specificity": 0.8387096774193549, + "precision_pos": 0.5, + "f1_pos": 0.5263157894736842, + "per_class": { + "0": { + "precision": 0.8666666666666667, + "recall": 0.8387096774193549, + "f1-score": 0.8524590163934426, + "support": 62.0 + }, + "1": { + "precision": 0.5, + "recall": 0.5555555555555556, + "f1-score": 0.5263157894736842, + "support": 18.0 + }, + "accuracy": 0.775, + "macro avg": { + "precision": 0.6833333333333333, + "recall": 0.6971326164874552, + "f1-score": 0.6893874029335634, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.7841666666666667, + "recall": 0.775, + "f1-score": 0.7790767903364969, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/010/resnet/pr.png b/results/downsample/adam/010/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..1d5d251376322f070e5046b8534960d23865ac7f --- /dev/null +++ b/results/downsample/adam/010/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc81a8507a373b824262896c7fbd021be4ad942fa20d150094de2b08c53680c6 +size 53386 diff --git a/results/downsample/adam/010/resnet/roc.png b/results/downsample/adam/010/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..0a1b5708b3f81959a7f2c911351da72ec3865eec --- /dev/null +++ b/results/downsample/adam/010/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:11c1e206ec030b743df8c4766ecab50bb35b331496a22db2d5f9d59d072d16f3 +size 57101 diff --git a/results/downsample/adam/010/resnet/test_pred.npz b/results/downsample/adam/010/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..d14858a9ad6caaea22dcf03a63a29fe32202833f --- /dev/null +++ b/results/downsample/adam/010/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:24992981677f82ca0925f6567acd5e7f3cfd31b7581444d4b617b5846af2da6e +size 1790 diff --git a/results/downsample/adam/010/resnet/train.log b/results/downsample/adam/010/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d4f08b782306604a803814c621fa61c63f78cc1b --- /dev/null +++ b/results/downsample/adam/010/resnet/train.log @@ -0,0 +1,134 @@ +[resnet] train=28 val=40 test=80 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6869 val_acc=0.4500 val_auc=0.4265 score=0.1655 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6760 val_acc=0.4000 val_auc=0.3441 score=0.1503 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6900 val_acc=0.4000 val_auc=0.3799 score=0.2155 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6855 val_acc=0.4250 val_auc=0.4982 score=0.3090 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6854 val_acc=0.4250 val_auc=0.5412 score=0.3040 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6477 val_acc=0.5500 val_auc=0.5950 score=0.3917 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6523 val_acc=0.7250 val_auc=0.6022 score=0.4189 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6320 val_acc=0.7500 val_auc=0.6129 score=0.4430 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6070 val_acc=0.7750 val_auc=0.6380 score=0.4737 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6136 val_acc=0.7250 val_auc=0.6774 score=0.3362 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.5858 val_acc=0.7500 val_auc=0.7348 score=0.4352 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.6138 val_acc=0.7750 val_auc=0.7670 score=0.5167 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.5325 val_acc=0.8000 val_auc=0.8244 score=0.5991 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.5448 val_acc=0.8000 val_auc=0.7993 score=0.5907 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.4627 val_acc=0.8000 val_auc=0.7742 score=0.6130 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.5039 val_acc=0.7000 val_auc=0.7706 score=0.5244 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.4926 val_acc=0.6750 val_auc=0.7168 score=0.5552 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.4636 val_acc=0.5500 val_auc=0.6452 score=0.4084 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.4228 val_acc=0.5250 val_auc=0.6308 score=0.3890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.4284 val_acc=0.5750 val_auc=0.6237 score=0.3914 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.3780 val_acc=0.6000 val_auc=0.6344 score=0.4102 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.3352 val_acc=0.5750 val_auc=0.6308 score=0.3647 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.2825 val_acc=0.6250 val_auc=0.6380 score=0.3974 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.3237 val_acc=0.7250 val_auc=0.6380 score=0.4688 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.3356 val_acc=0.7500 val_auc=0.6452 score=0.4921 +[resnet] early stop at ep24 (best ep14 score=0.6130) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=14 best_val_score=0.6130 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/010/resnet acc=0.7750 auroc=0.7706093189964158 f1_macro=0.6894 qwk=0.3793103448275862 diff --git a/results/downsample/adam/010/retfound/confusion_matrix.png b/results/downsample/adam/010/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..c7f2220cab082dcb6dab58860cbca5896ad1fa7e --- /dev/null +++ b/results/downsample/adam/010/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fc24241757f80b79119cda6b1fa30d9181f1c22a24509445663e09c83f9beff +size 68400 diff --git a/results/downsample/adam/010/retfound/confusion_matrix_test.jpg b/results/downsample/adam/010/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a44063cdd273cf01b218e16dedbcc5eb44589de0 --- /dev/null +++ b/results/downsample/adam/010/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbf6a7f569061967235b05e90273681d0d6dd737bd7ced46cbededfec57d2cd +size 241221 diff --git a/results/downsample/adam/010/retfound/log.txt b/results/downsample/adam/010/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..e906eca50c5843e4a99f51ae97228aa426935029 --- /dev/null +++ b/results/downsample/adam/010/retfound/log.txt @@ -0,0 +1,80 @@ +{"train_lr": 0.0, "train_loss": 0.69287109375, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 2.9296875e-05, "train_loss": 0.692675769329071, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 5.859375e-05, "train_loss": 0.6929361820220947, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 8.7890625e-05, "train_loss": 0.6928385496139526, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0001171875, "train_loss": 0.6651041507720947, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.000146484375, "train_loss": 0.650585949420929, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00017578125, "train_loss": 0.6193034052848816, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00020507812500000002, "train_loss": 0.5952962040901184, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000234375, "train_loss": 0.5989745855331421, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.00026367187499999996, "train_loss": 0.5298990607261658, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.00029296875, "train_loss": 0.5160481929779053, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.00029282175344836725, "train_loss": 0.40546876192092896, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.00029238105982496126, "train_loss": 0.45376789569854736, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.00029164755662809086, "train_loss": 0.38005778193473816, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.00029062272103557807, "train_loss": 0.5393717288970947, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0002893086169299184, "train_loss": 0.5471516847610474, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0002877078907418962, "train_loss": 0.41567790508270264, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.00028582376612102533, "train_loss": 0.40058594942092896, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.00028366003744354786, "train_loss": 0.690234363079071, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00028122106217106567, "train_loss": 0.6759399175643921, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002785117520751922, "train_loss": 0.09405110776424408, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0002755375633458985, "train_loss": 0.5543578863143921, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00027230448560347243, "train_loss": 0.4824666380882263, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0002688190298362208, "train_loss": 0.5095377564430237, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0002650882152882052, "train_loss": 0.5150146484375, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0002611195553234188, "train_loss": 0.6086507439613342, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0002569210422948709, "train_loss": 0.43494874238967896, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.00025250113144905187, "train_loss": 0.5195515751838684, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.0002478687238981918, "train_loss": 0.6431314945220947, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0002430331486946052, "train_loss": 0.4064778685569763, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00023800414404322103, "train_loss": 0.614697277545929, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.000232791837690134, "train_loss": 0.45710450410842896, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00022740672652667176, "train_loss": 0.45250651240348816, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022185965545005283, "train_loss": 0.6167643070220947, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00021616179552320664, "train_loss": 0.565380871295929, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00021032462147773973, "train_loss": 0.5257405638694763, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00020435988860535324, "train_loss": 0.4080444276332855, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00019827960908424974, "train_loss": 0.348876953125, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.0001920960277882052, "train_loss": 0.4750610291957855, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00018582159762702254, "train_loss": 0.3688150942325592, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 0.0001794689544680288, "train_loss": 0.5460652709007263, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 0.0001730508916891197, "train_loss": 0.5000284910202026, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 0.00016658033441459928, "train_loss": 0.36626383662223816, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 0.00016007031348569967, "train_loss": 0.668078601360321, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 0.00015353393921820096, "train_loss": 0.40086668729782104, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 0.00014698437499999998, "train_loss": 0.3045613467693329, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 0.00014043481078179906, "train_loss": 0.623583972454071, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 0.00013389843651430032, "train_loss": 0.542651355266571, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 0.0001273884155854007, "train_loss": 0.5618123412132263, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 0.00012091785831088028, "train_loss": 0.5629028081893921, "epoch": 49, "n_parameters": 303303682} +{"train_lr": 0.00011449979553197115, "train_loss": 0.4204345643520355, "epoch": 50, "n_parameters": 303303682} +{"train_lr": 0.00010814715237297744, "train_loss": 0.4891194701194763, "epoch": 51, "n_parameters": 303303682} +{"train_lr": 0.00010187272221179478, "train_loss": 0.6047933101654053, "epoch": 52, "n_parameters": 303303682} +{"train_lr": 9.568914091575024e-05, "train_loss": 0.36343181133270264, "epoch": 53, "n_parameters": 303303682} +{"train_lr": 8.960886139464677e-05, "train_loss": 0.5020670294761658, "epoch": 54, "n_parameters": 303303682} +{"train_lr": 8.364412852226022e-05, "train_loss": 0.4825683534145355, "epoch": 55, "n_parameters": 303303682} +{"train_lr": 7.780695447679333e-05, "train_loss": 0.3970988094806671, "epoch": 56, "n_parameters": 303303682} +{"train_lr": 7.210909454994717e-05, "train_loss": 0.35562336444854736, "epoch": 57, "n_parameters": 303303682} +{"train_lr": 6.656202347332823e-05, "train_loss": 0.3549031615257263, "epoch": 58, "n_parameters": 303303682} +{"train_lr": 6.1176912309866e-05, "train_loss": 0.3529052734375, "epoch": 59, "n_parameters": 303303682} +{"train_lr": 5.5964605956778944e-05, "train_loss": 0.3511148989200592, "epoch": 60, "n_parameters": 303303682} +{"train_lr": 5.0935601305394765e-05, "train_loss": 0.5986002683639526, "epoch": 61, "n_parameters": 303303682} +{"train_lr": 4.610002610180817e-05, "train_loss": 0.6416422724723816, "epoch": 62, "n_parameters": 303303682} +{"train_lr": 4.146761855094815e-05, "train_loss": 0.4926920533180237, "epoch": 63, "n_parameters": 303303682} +{"train_lr": 3.704770770512911e-05, "train_loss": 0.5485554933547974, "epoch": 64, "n_parameters": 303303682} +{"train_lr": 3.284919467658121e-05, "train_loss": 0.4323079288005829, "epoch": 65, "n_parameters": 303303682} +{"train_lr": 2.8880534711794795e-05, "train_loss": 0.32736003398895264, "epoch": 66, "n_parameters": 303303682} +{"train_lr": 2.5149720163779206e-05, "train_loss": 0.45201414823532104, "epoch": 67, "n_parameters": 303303682} +{"train_lr": 2.1664264396527578e-05, "train_loss": 0.6007487177848816, "epoch": 68, "n_parameters": 303303682} +{"train_lr": 1.843118665410151e-05, "train_loss": 0.42778727412223816, "epoch": 69, "n_parameters": 303303682} +{"train_lr": 1.5456997924807793e-05, "train_loss": 0.3225504457950592, "epoch": 70, "n_parameters": 303303682} +{"train_lr": 1.2747687828934333e-05, "train_loss": 0.3636433780193329, "epoch": 71, "n_parameters": 303303682} +{"train_lr": 1.0308712556452114e-05, "train_loss": 0.33894652128219604, "epoch": 72, "n_parameters": 303303682} +{"train_lr": 8.144983878974696e-06, "train_loss": 0.4097086489200592, "epoch": 73, "n_parameters": 303303682} +{"train_lr": 6.260859258103799e-06, "train_loss": 0.3693400025367737, "epoch": 74, "n_parameters": 303303682} +{"train_lr": 4.660133070081592e-06, "train_loss": 0.44324544072151184, "epoch": 75, "n_parameters": 303303682} +{"train_lr": 3.3460289644219218e-06, "train_loss": 0.593127429485321, "epoch": 76, "n_parameters": 303303682} +{"train_lr": 2.3211933719091574e-06, "train_loss": 0.5338460206985474, "epoch": 77, "n_parameters": 303303682} +{"train_lr": 1.5876901750387716e-06, "train_loss": 0.4762410521507263, "epoch": 78, "n_parameters": 303303682} +{"train_lr": 1.146996551632778e-06, "train_loss": 0.4970459043979645, "epoch": 79, "n_parameters": 303303682} diff --git a/results/downsample/adam/010/retfound/metrics.json b/results/downsample/adam/010/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..ace3663c47f52e9f310bb4fcc559fb3eb0c043a6 --- /dev/null +++ b/results/downsample/adam/010/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.775, + "balanced_accuracy": 0.5, + "precision_macro": 0.3875, + "recall_macro": 0.5, + "f1_macro": 0.43661971830985913, + "precision_weighted": 0.6006250000000001, + "recall_weighted": 0.775, + "f1_weighted": 0.6767605633802816, + "cohen_kappa": 0.0, + "quadratic_weighted_kappa": 0.0, + "mcc": 0.0, + "auroc": 0.8409498207885304, + "auprc": 0.6266484107916334, + "sensitivity": 0.0, + "specificity": 1.0, + "precision_pos": null, + "f1_pos": 0.0, + "per_class": { + "0": { + "precision": 0.775, + "recall": 1.0, + "f1-score": 0.8732394366197183, + "support": 62.0 + }, + "1": { + "precision": 0.0, + "recall": 0.0, + "f1-score": 0.0, + "support": 18.0 + }, + "accuracy": 0.775, + "macro avg": { + "precision": 0.3875, + "recall": 0.5, + "f1-score": 0.43661971830985913, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.6006250000000001, + "recall": 0.775, + "f1-score": 0.6767605633802816, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/010/retfound/metrics_test.csv b/results/downsample/adam/010/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..014b1e98a73d243054d1038761719a6e709e138d --- /dev/null +++ b/results/downsample/adam/010/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.616695836186409,0.775,0.43661971830985913,0.8407258064516129,0.225,0.3875,0.3875,0.5,0.788479937844087,0.0 diff --git a/results/downsample/adam/010/retfound/metrics_val.csv b/results/downsample/adam/010/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..f52f149db9e1dbc1666b1b7adf41b62dd81a44f8 --- /dev/null +++ b/results/downsample/adam/010/retfound/metrics_val.csv @@ -0,0 +1,81 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.692822257677714,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.692822257677714,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.692822257677714,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.6809787352879842,0.775,0.43661971830985913,0.7840501792114695,0.225,0.3875,0.3875,0.5,0.7000345720734933,0.0 +0.6669704914093018,0.775,0.43661971830985913,0.7508960573476703,0.225,0.3875,0.3875,0.5,0.6667353966456893,0.0 +0.6515082518259684,0.775,0.43661971830985913,0.7231182795698925,0.225,0.3875,0.3875,0.5,0.6437288412469168,0.0 +0.634960949420929,0.775,0.43661971830985913,0.7195340501792115,0.225,0.3875,0.3875,0.5,0.6571275633733351,0.0 +0.6189588705698649,0.775,0.43661971830985913,0.7410394265232975,0.225,0.3875,0.3875,0.5,0.6576355972034109,0.0 +0.6060655315717062,0.775,0.43661971830985913,0.764336917562724,0.225,0.3875,0.3875,0.5,0.7041922850496812,0.0 +0.5974229673544565,0.775,0.43661971830985913,0.7894265232974911,0.225,0.3875,0.3875,0.5,0.7217567856990241,0.0 +0.5944878359635671,0.775,0.43661971830985913,0.796594982078853,0.225,0.3875,0.3875,0.5,0.7329921020830448,0.0 +0.5988213618596395,0.775,0.43661971830985913,0.8136200716845878,0.225,0.3875,0.3875,0.5,0.7469903157534084,0.0 +0.6117580930391947,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7595986324788727,0.0 +0.6297037601470947,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.7703539582038005,0.0 +0.6493720014890035,0.775,0.43661971830985913,0.8163082437275986,0.225,0.3875,0.3875,0.5,0.7661111630947203,0.0 +0.6637071544925371,0.775,0.43661971830985913,0.8243727598566308,0.225,0.3875,0.3875,0.5,0.7709036974430559,0.0 +0.6739963019887606,0.775,0.43661971830985913,0.8207885304659498,0.225,0.3875,0.3875,0.5,0.7685004256849455,0.0 +0.679985207815965,0.775,0.43661971830985913,0.8198924731182795,0.225,0.3875,0.3875,0.5,0.7680979186410721,0.0 +0.674045130610466,0.775,0.43661971830985913,0.8189964157706093,0.225,0.3875,0.3875,0.5,0.7643539851140462,0.0 +0.6655619144439697,0.775,0.43661971830985913,0.8216845878136201,0.225,0.3875,0.3875,0.5,0.7710658461612351,0.0 +0.6626627519726753,0.775,0.43661971830985913,0.8243727598566308,0.225,0.3875,0.3875,0.5,0.7798208845095413,0.0 +0.6554667179783186,0.775,0.43661971830985913,0.8324372759856631,0.225,0.3875,0.3875,0.5,0.7853734933086588,0.0 +0.6452894409497579,0.775,0.43661971830985913,0.8440860215053764,0.225,0.3875,0.3875,0.5,0.793023792577793,0.0 +0.6360317667325338,0.775,0.43661971830985913,0.8449820788530467,0.225,0.3875,0.3875,0.5,0.8010548380838107,0.0 +0.6275186985731125,0.775,0.43661971830985913,0.8422939068100359,0.225,0.3875,0.3875,0.5,0.7997045528825963,0.0 +0.619118923942248,0.775,0.43661971830985913,0.8431899641577061,0.225,0.3875,0.3875,0.5,0.7998528658229004,0.0 +0.6093153357505798,0.775,0.43661971830985913,0.8458781362007168,0.225,0.3875,0.3875,0.5,0.8138696071872795,0.0 +0.599237730105718,0.775,0.43661971830985913,0.8449820788530467,0.225,0.3875,0.3875,0.5,0.8132504184457747,0.0 +0.5868910153706869,0.775,0.43661971830985913,0.8431899641577061,0.225,0.3875,0.3875,0.5,0.7977081554658099,0.0 +0.5778889854749044,0.775,0.43661971830985913,0.8422939068100359,0.225,0.3875,0.3875,0.5,0.7996666808114319,0.0 +0.5709526985883713,0.775,0.43661971830985913,0.8387096774193549,0.225,0.3875,0.3875,0.5,0.7975761722181229,0.0 +0.5683173437913259,0.775,0.43661971830985913,0.8387096774193549,0.225,0.3875,0.3875,0.5,0.7975761722181229,0.0 +0.5689453134934107,0.775,0.43661971830985913,0.8351254480286738,0.225,0.3875,0.3875,0.5,0.7984018447202706,0.0 +0.5681925465663274,0.775,0.43661971830985913,0.8306451612903225,0.225,0.3875,0.3875,0.5,0.7964432050401808,0.0 +0.5678521196047465,0.775,0.43661971830985913,0.8279569892473119,0.225,0.3875,0.3875,0.5,0.795273614396906,0.0 +0.5664157519737879,0.775,0.43661971830985913,0.8279569892473119,0.225,0.3875,0.3875,0.5,0.7952736143969061,0.0 +0.5664116591215134,0.775,0.43661971830985913,0.8306451612903226,0.225,0.3875,0.3875,0.5,0.8092235981049543,0.0 +0.5691636900107065,0.775,0.43661971830985913,0.8342293906810037,0.225,0.3875,0.3875,0.5,0.8095975754233435,0.0 +0.5720546990633011,0.775,0.43661971830985913,0.8378136200716846,0.225,0.3875,0.3875,0.5,0.8136885406933967,0.0 +0.5755900094906489,0.775,0.43661971830985913,0.8351254480286738,0.225,0.3875,0.3875,0.5,0.8126062896111456,0.0 +0.5773444126049677,0.775,0.43661971830985913,0.8306451612903226,0.225,0.3875,0.3875,0.5,0.8099695454209208,0.0 +0.578347439567248,0.775,0.43661971830985913,0.825268817204301,0.225,0.3875,0.3875,0.5,0.8072553953752584,0.0 +0.5806369533141454,0.775,0.43661971830985913,0.8243727598566308,0.225,0.3875,0.3875,0.5,0.8072553953752584,0.0 +0.5792928238709768,0.775,0.43661971830985913,0.8216845878136201,0.225,0.3875,0.3875,0.5,0.8054807822045181,0.0 +0.5787177334229151,0.775,0.43661971830985913,0.8172043010752688,0.225,0.3875,0.3875,0.5,0.8036789277553986,0.0 +0.5784457723299662,0.775,0.43661971830985913,0.8136200716845878,0.225,0.3875,0.3875,0.5,0.8019988435293632,0.0 +0.575597474972407,0.775,0.43661971830985913,0.8136200716845878,0.225,0.3875,0.3875,0.5,0.8019988435293632,0.0 +0.572087953488032,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5663845290740331,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5604729404052099,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003970261805171,0.0 +0.5559929559628168,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003927816983438,0.0 +0.5510979443788528,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5462958514690399,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003927816983438,0.0 +0.5421813776095709,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5387451301018397,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5357923855384191,0.775,0.43661971830985913,0.8100358422939068,0.225,0.3875,0.3875,0.5,0.8003121365370536,0.0 +0.5332065224647522,0.775,0.43661971830985913,0.8064516129032258,0.225,0.3875,0.3875,0.5,0.7986949663714963,0.0 +0.5309115995963415,0.775,0.43661971830985913,0.803763440860215,0.225,0.3875,0.3875,0.5,0.7933177666157387,0.0 +0.5292012492815653,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7948274284514874,0.0 +0.5284674863020579,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7948274284514875,0.0 +0.528290460507075,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7948274284514874,0.0 +0.5276516228914261,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7934225005914404,0.0 +0.5269436438878378,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7934225005914404,0.0 +0.5263915856679281,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157387,0.0 +0.5254543672005335,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5248047063748041,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5243035157521566,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157387,0.0 +0.5237467288970947,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5233907103538513,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5231079260508219,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5228807131449381,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5228237559398016,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5227776517470678,0.775,0.43661971830985913,0.8028673835125448,0.225,0.3875,0.3875,0.5,0.7933177666157386,0.0 +0.5226725240548452,0.775,0.43661971830985913,0.8019713261648745,0.225,0.3875,0.3875,0.5,0.7929247061657522,0.0 +0.5226142058769861,0.775,0.43661971830985913,0.8019713261648745,0.225,0.3875,0.3875,0.5,0.7929247061657522,0.0 +0.5226047188043594,0.775,0.43661971830985913,0.8010752688172043,0.225,0.3875,0.3875,0.5,0.7876337008747469,0.0 +0.5224704146385193,0.775,0.43661971830985913,0.8010752688172043,0.225,0.3875,0.3875,0.5,0.7876337008747469,0.0 +0.5223741233348846,0.775,0.43661971830985913,0.8010752688172043,0.225,0.3875,0.3875,0.5,0.7891433627104957,0.0 +0.5223171611626943,0.775,0.43661971830985913,0.8010752688172043,0.225,0.3875,0.3875,0.5,0.7891433627104957,0.0 +0.5222975065310796,0.775,0.43661971830985913,0.8010752688172043,0.225,0.3875,0.3875,0.5,0.7876337008747469,0.0 diff --git a/results/downsample/adam/010/retfound/pr.png b/results/downsample/adam/010/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..ef0016b6319aceccc28a1264a8a07bfde81a717e --- /dev/null +++ b/results/downsample/adam/010/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b94fb0050f2456c85150283725a25058f2ce4508a02f703f8e048609a55ebf9 +size 51184 diff --git a/results/downsample/adam/010/retfound/roc.png b/results/downsample/adam/010/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..cadd76bbc261e700039d75ba1cc29bdc8cec48d2 --- /dev/null +++ b/results/downsample/adam/010/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:14db407a9b4320f119ce98456d4a37547a00ab6554363711ea7b3ce3a1e9cfcd +size 59314 diff --git a/results/downsample/adam/010/retfound/test_pred.npz b/results/downsample/adam/010/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..373dd45953cda5d27bb5e8e6da3a9f97340baca7 --- /dev/null +++ b/results/downsample/adam/010/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd674b4c9187f50f2606541687a4f8fa3618f23517ddb238057cbd3723497a22 +size 1470 diff --git a/results/downsample/adam/010/retfound/train.log b/results/downsample/adam/010/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..84e1d0e7e22ab18564f0063049fe7e6920962ec5 --- /dev/null +++ b/results/downsample/adam/010/retfound/train.log @@ -0,0 +1,1043 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:44:54.977267096 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:44:54.421717] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:44:54.421941] Namespace(batch_size=15, +epochs=80, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/adam_10', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/010', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:44:57.467524] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:44:59.106134] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:44:59.349721] Sampler_train = +[14:44:59.394904] len of train_set: 15 +[14:44:59.792417] [Adaptation] Full fine-tuning: training all parameters. +[14:44:59.793497] number of trainable params (M): 303.30 +[14:44:59.793578] base lr: 5.00e-03 +[14:44:59.793631] actual lr: 2.93e-04 +[14:44:59.793677] accumulate grad iterations: 1 +[14:44:59.793720] effective batch size: 15 +[14:44:59.796696] criterion = CrossEntropyLoss() +[14:44:59.796788] Start training for 80 epochs +[14:44:59.799430] log_dir: ./output_logs/retfound +[14:45:01.538207] Epoch: [0] [0/1] eta: 0:00:01 lr: 0.000000 loss: 0.6929 (0.6929) time: 1.7376 data: 1.1325 max mem: 4514 +[14:45:01.624654] Epoch: [0] Total time: 0:00:01 (1.8251 s / it) +[14:45:01.626126] Averaged stats: lr: 0.000000 loss: 0.6929 (0.6929) +[14:45:03.074585] val: [0/3] eta: 0:00:04 loss: 0.6924 (0.6924) time: 1.4412 data: 1.4152 max mem: 4514 +[14:45:03.140808] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.5023 data: 0.4719 max mem: 4514 +[14:45:03.214658] val: Total time: 0:00:01 (0.5273 s / it) +[14:45:03.228364] val loss: 0.692822257677714 +[14:45:03.228569] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:45:04.952792] Best epoch = 0, Best score = 0.3023 +[14:45:05.026965] log_dir: ./output_logs/retfound +[14:45:06.303029] Epoch: [1] [0/1] eta: 0:00:01 lr: 0.000029 loss: 0.6927 (0.6927) time: 1.2748 data: 1.1546 max mem: 4514 +[14:45:06.370485] Epoch: [1] Total time: 0:00:01 (1.3434 s / it) +[14:45:06.374608] Averaged stats: lr: 0.000029 loss: 0.6927 (0.6927) +[14:45:07.596374] val: [0/3] eta: 0:00:03 loss: 0.6924 (0.6924) time: 1.2135 data: 1.1891 max mem: 4514 +[14:45:07.622716] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.4130 data: 0.3965 max mem: 4514 +[14:45:07.701337] val: Total time: 0:00:01 (0.4401 s / it) +[14:45:07.716393] val loss: 0.692822257677714 +[14:45:07.716626] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:45:07.752630] Best epoch = 0, Best score = 0.3023 +[14:45:08.007747] log_dir: ./output_logs/retfound +[14:45:09.227083] Epoch: [2] [0/1] eta: 0:00:01 lr: 0.000059 loss: 0.6929 (0.6929) time: 1.2184 data: 1.1747 max mem: 4514 +[14:45:09.302030] Epoch: [2] Total time: 0:00:01 (1.2941 s / it) +[14:45:09.302883] Averaged stats: lr: 0.000059 loss: 0.6929 (0.6929) +[14:45:10.543921] val: [0/3] eta: 0:00:03 loss: 0.6924 (0.6924) time: 1.2303 data: 1.2141 max mem: 4514 +[14:45:10.562549] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.4161 data: 0.4048 max mem: 4514 +[14:45:10.636932] val: Total time: 0:00:01 (0.4413 s / it) +[14:45:10.645821] val loss: 0.692822257677714 +[14:45:10.645996] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:45:10.684133] Best epoch = 0, Best score = 0.3023 +[14:45:10.935655] log_dir: ./output_logs/retfound +[14:45:12.156114] Epoch: [3] [0/1] eta: 0:00:01 lr: 0.000088 loss: 0.6928 (0.6928) time: 1.2195 data: 1.1027 max mem: 4752 +[14:45:12.238860] Epoch: [3] Total time: 0:00:01 (1.3030 s / it) +[14:45:12.239880] Averaged stats: lr: 0.000088 loss: 0.6928 (0.6928) +[14:45:13.513564] val: [0/3] eta: 0:00:03 loss: 0.6638 (0.6638) time: 1.2670 data: 1.2529 max mem: 4752 +[14:45:13.532850] val: [2/3] eta: 0:00:00 loss: 0.6638 (0.6810) time: 0.4286 data: 0.4178 max mem: 4752 +[14:45:13.610497] val: Total time: 0:00:01 (0.4548 s / it) +[14:45:13.621056] val loss: 0.6809787352879842 +[14:45:13.621248] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7841, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7000, Kappa: 0.0000, Score: 0.4069 +[14:45:15.298512] Best epoch = 3, Best score = 0.4069 +[14:45:15.361968] log_dir: ./output_logs/retfound +[14:45:16.670973] Epoch: [4] [0/1] eta: 0:00:01 lr: 0.000117 loss: 0.6651 (0.6651) time: 1.3081 data: 1.2340 max mem: 6806 +[14:45:16.739726] Epoch: [4] Total time: 0:00:01 (1.3776 s / it) +[14:45:16.740596] Averaged stats: lr: 0.000117 loss: 0.6651 (0.6651) +[14:45:17.964778] val: [0/3] eta: 0:00:03 loss: 0.6274 (0.6274) time: 1.2100 data: 1.1959 max mem: 6806 +[14:45:17.983658] val: [2/3] eta: 0:00:00 loss: 0.6275 (0.6670) time: 0.4094 data: 0.3988 max mem: 6806 +[14:45:18.059207] val: Total time: 0:00:01 (0.4350 s / it) +[14:45:18.068120] val loss: 0.6669704914093018 +[14:45:18.068304] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7509, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6667, Kappa: 0.0000, Score: 0.3958 +[14:45:18.103189] Best epoch = 3, Best score = 0.4069 +[14:45:18.369016] log_dir: ./output_logs/retfound +[14:45:19.658132] Epoch: [5] [0/1] eta: 0:00:01 lr: 0.000146 loss: 0.6506 (0.6506) time: 1.2882 data: 1.2303 max mem: 6806 +[14:45:19.730436] Epoch: [5] Total time: 0:00:01 (1.3613 s / it) +[14:45:19.731260] Averaged stats: lr: 0.000146 loss: 0.6506 (0.6506) +[14:45:21.081965] val: [0/3] eta: 0:00:03 loss: 0.5832 (0.5832) time: 1.3316 data: 1.3126 max mem: 6806 +[14:45:21.109332] val: [2/3] eta: 0:00:00 loss: 0.5835 (0.6515) time: 0.4527 data: 0.4376 max mem: 6806 +[14:45:21.183483] val: Total time: 0:00:01 (0.4781 s / it) +[14:45:21.197800] val loss: 0.6515082518259684 +[14:45:21.198028] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7231, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6437, Kappa: 0.0000, Score: 0.3866 +[14:45:21.233897] Best epoch = 3, Best score = 0.4069 +[14:45:21.516902] log_dir: ./output_logs/retfound +[14:45:22.984916] Epoch: [6] [0/1] eta: 0:00:01 lr: 0.000176 loss: 0.6193 (0.6193) time: 1.4668 data: 1.3140 max mem: 6806 +[14:45:23.081484] Epoch: [6] Total time: 0:00:01 (1.5644 s / it) +[14:45:23.098062] Averaged stats: lr: 0.000176 loss: 0.6193 (0.6193) +[14:45:24.361654] val: [0/3] eta: 0:00:03 loss: 0.5317 (0.5317) time: 1.2559 data: 1.2421 max mem: 6806 +[14:45:24.380348] val: [2/3] eta: 0:00:00 loss: 0.5321 (0.6350) time: 0.4247 data: 0.4141 max mem: 6806 +[14:45:24.450839] val: Total time: 0:00:01 (0.4486 s / it) +[14:45:24.459644] val loss: 0.634960949420929 +[14:45:24.459844] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7195, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6571, Kappa: 0.0000, Score: 0.3854 +[14:45:24.506695] Best epoch = 3, Best score = 0.4069 +[14:45:24.743500] log_dir: ./output_logs/retfound +[14:45:25.860277] Epoch: [7] [0/1] eta: 0:00:01 lr: 0.000205 loss: 0.5953 (0.5953) time: 1.1159 data: 1.0592 max mem: 6806 +[14:45:25.930442] Epoch: [7] Total time: 0:00:01 (1.1868 s / it) +[14:45:25.931264] Averaged stats: lr: 0.000205 loss: 0.5953 (0.5953) +[14:45:27.166436] val: [0/3] eta: 0:00:03 loss: 0.4745 (0.4745) time: 1.2245 data: 1.2091 max mem: 6806 +[14:45:27.187529] val: [2/3] eta: 0:00:00 loss: 0.4747 (0.6190) time: 0.4150 data: 0.4037 max mem: 6806 +[14:45:27.266589] val: Total time: 0:00:01 (0.4418 s / it) +[14:45:27.275617] val loss: 0.6189588705698649 +[14:45:27.275841] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7410, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6576, Kappa: 0.0000, Score: 0.3926 +[14:45:27.306275] Best epoch = 3, Best score = 0.4069 +[14:45:27.581287] log_dir: ./output_logs/retfound +[14:45:28.913476] Epoch: [8] [0/1] eta: 0:00:01 lr: 0.000234 loss: 0.5990 (0.5990) time: 1.3312 data: 1.2276 max mem: 6806 +[14:45:28.986245] Epoch: [8] Total time: 0:00:01 (1.4048 s / it) +[14:45:28.992270] Averaged stats: lr: 0.000234 loss: 0.5990 (0.5990) +[14:45:30.242067] val: [0/3] eta: 0:00:03 loss: 0.4165 (0.4165) time: 1.2247 data: 1.2097 max mem: 6806 +[14:45:30.260762] val: [2/3] eta: 0:00:00 loss: 0.4165 (0.6061) time: 0.4143 data: 0.4033 max mem: 6806 +[14:45:30.331917] val: Total time: 0:00:01 (0.4384 s / it) +[14:45:30.343467] val loss: 0.6060655315717062 +[14:45:30.343711] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7643, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7042, Kappa: 0.0000, Score: 0.4003 +[14:45:30.397231] Best epoch = 3, Best score = 0.4069 +[14:45:30.639075] log_dir: ./output_logs/retfound +[14:45:31.786346] Epoch: [9] [0/1] eta: 0:00:01 lr: 0.000264 loss: 0.5299 (0.5299) time: 1.1464 data: 1.0927 max mem: 6806 +[14:45:31.857073] Epoch: [9] Total time: 0:00:01 (1.2178 s / it) +[14:45:31.857939] Averaged stats: lr: 0.000264 loss: 0.5299 (0.5299) +[14:45:33.212954] val: [0/3] eta: 0:00:04 loss: 0.3593 (0.3593) time: 1.3402 data: 1.3236 max mem: 6806 +[14:45:33.237412] val: [2/3] eta: 0:00:00 loss: 0.3593 (0.5974) time: 0.4547 data: 0.4413 max mem: 6806 +[14:45:33.311374] val: Total time: 0:00:01 (0.4798 s / it) +[14:45:33.322107] val loss: 0.5974229673544565 +[14:45:33.322364] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7894, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7218, Kappa: 0.0000, Score: 0.4087 +[14:45:35.159481] Best epoch = 9, Best score = 0.4087 +[14:45:35.211164] log_dir: ./output_logs/retfound +[14:45:36.446642] Epoch: [10] [0/1] eta: 0:00:01 lr: 0.000293 loss: 0.5160 (0.5160) time: 1.2346 data: 1.1793 max mem: 6806 +[14:45:36.513629] Epoch: [10] Total time: 0:00:01 (1.3023 s / it) +[14:45:36.514492] Averaged stats: lr: 0.000293 loss: 0.5160 (0.5160) +[14:45:37.757137] val: [0/3] eta: 0:00:03 loss: 0.3019 (0.3019) time: 1.2312 data: 1.2161 max mem: 6806 +[14:45:37.776721] val: [2/3] eta: 0:00:00 loss: 0.3019 (0.5945) time: 0.4167 data: 0.4055 max mem: 6806 +[14:45:37.845353] val: Total time: 0:00:01 (0.4400 s / it) +[14:45:37.854805] val loss: 0.5944878359635671 +[14:45:37.855014] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7966, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7330, Kappa: 0.0000, Score: 0.4111 +[14:45:39.489548] Best epoch = 10, Best score = 0.4111 +[14:45:39.560621] log_dir: ./output_logs/retfound +[14:45:40.659851] Epoch: [11] [0/1] eta: 0:00:01 lr: 0.000293 loss: 0.4055 (0.4055) time: 1.0983 data: 1.0442 max mem: 6806 +[14:45:40.732973] Epoch: [11] Total time: 0:00:01 (1.1722 s / it) +[14:45:40.734001] Averaged stats: lr: 0.000293 loss: 0.4055 (0.4055) +[14:45:42.051365] val: [0/3] eta: 0:00:03 loss: 0.2505 (0.2505) time: 1.3064 data: 1.2924 max mem: 6806 +[14:45:42.070190] val: [2/3] eta: 0:00:00 loss: 0.2505 (0.5988) time: 0.4416 data: 0.4309 max mem: 6806 +[14:45:42.140090] val: Total time: 0:00:01 (0.4653 s / it) +[14:45:42.148906] val loss: 0.5988213618596395 +[14:45:42.149084] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8136, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7470, Kappa: 0.0000, Score: 0.4167 +[14:45:43.898234] Best epoch = 11, Best score = 0.4167 +[14:45:43.979246] log_dir: ./output_logs/retfound +[14:45:45.105612] Epoch: [12] [0/1] eta: 0:00:01 lr: 0.000292 loss: 0.4538 (0.4538) time: 1.1255 data: 1.0718 max mem: 6806 +[14:45:45.172248] Epoch: [12] Total time: 0:00:01 (1.1929 s / it) +[14:45:45.173056] Averaged stats: lr: 0.000292 loss: 0.4538 (0.4538) +[14:45:46.456145] val: [0/3] eta: 0:00:03 loss: 0.2068 (0.2068) time: 1.2721 data: 1.2583 max mem: 6806 +[14:45:46.474675] val: [2/3] eta: 0:00:00 loss: 0.2068 (0.6118) time: 0.4300 data: 0.4195 max mem: 6806 +[14:45:46.542350] val: Total time: 0:00:01 (0.4530 s / it) +[14:45:46.551207] val loss: 0.6117580930391947 +[14:45:46.551377] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7596, Kappa: 0.0000, Score: 0.4179 +[14:45:48.204776] Best epoch = 12, Best score = 0.4179 +[14:45:48.264057] log_dir: ./output_logs/retfound +[14:45:49.499741] Epoch: [13] [0/1] eta: 0:00:01 lr: 0.000292 loss: 0.3801 (0.3801) time: 1.2347 data: 1.1307 max mem: 6806 +[14:45:49.571358] Epoch: [13] Total time: 0:00:01 (1.3071 s / it) +[14:45:49.581003] Averaged stats: lr: 0.000292 loss: 0.3801 (0.3801) +[14:45:50.950228] val: [0/3] eta: 0:00:04 loss: 0.1712 (0.1712) time: 1.3607 data: 1.3392 max mem: 6806 +[14:45:50.979398] val: [2/3] eta: 0:00:00 loss: 0.1712 (0.6297) time: 0.4631 data: 0.4465 max mem: 6806 +[14:45:51.049512] val: Total time: 0:00:01 (0.4872 s / it) +[14:45:51.063647] val loss: 0.6297037601470947 +[14:45:51.063859] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7704, Kappa: 0.0000, Score: 0.4179 +[14:45:51.099586] Best epoch = 12, Best score = 0.4179 +[14:45:51.350932] log_dir: ./output_logs/retfound +[14:45:52.534976] Epoch: [14] [0/1] eta: 0:00:01 lr: 0.000291 loss: 0.5394 (0.5394) time: 1.1831 data: 1.1288 max mem: 6806 +[14:45:52.606386] Epoch: [14] Total time: 0:00:01 (1.2553 s / it) +[14:45:52.607402] Averaged stats: lr: 0.000291 loss: 0.5394 (0.5394) +[14:45:53.893131] val: [0/3] eta: 0:00:03 loss: 0.1452 (0.1452) time: 1.2784 data: 1.2631 max mem: 6806 +[14:45:53.913331] val: [2/3] eta: 0:00:00 loss: 0.1452 (0.6494) time: 0.4326 data: 0.4212 max mem: 6806 +[14:45:53.990145] val: Total time: 0:00:01 (0.4587 s / it) +[14:45:54.009451] val loss: 0.6493720014890035 +[14:45:54.009681] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8163, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7661, Kappa: 0.0000, Score: 0.4176 +[14:45:54.054097] Best epoch = 12, Best score = 0.4179 +[14:45:54.311081] log_dir: ./output_logs/retfound +[14:45:55.637553] Epoch: [15] [0/1] eta: 0:00:01 lr: 0.000289 loss: 0.5472 (0.5472) time: 1.3249 data: 1.1730 max mem: 6806 +[14:45:55.708346] Epoch: [15] Total time: 0:00:01 (1.3971 s / it) +[14:45:55.725170] Averaged stats: lr: 0.000289 loss: 0.5472 (0.5472) +[14:45:56.989502] val: [0/3] eta: 0:00:03 loss: 0.1300 (0.1300) time: 1.2534 data: 1.2394 max mem: 6806 +[14:45:57.008951] val: [2/3] eta: 0:00:00 loss: 0.1300 (0.6637) time: 0.4241 data: 0.4132 max mem: 6806 +[14:45:57.089127] val: Total time: 0:00:01 (0.4512 s / it) +[14:45:57.098065] val loss: 0.6637071544925371 +[14:45:57.098289] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8244, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7709, Kappa: 0.0000, Score: 0.4203 +[14:45:58.746459] Best epoch = 15, Best score = 0.4203 +[14:45:58.808942] log_dir: ./output_logs/retfound +[14:46:00.044133] Epoch: [16] [0/1] eta: 0:00:01 lr: 0.000288 loss: 0.4157 (0.4157) time: 1.2343 data: 1.1809 max mem: 6806 +[14:46:00.113108] Epoch: [16] Total time: 0:00:01 (1.3040 s / it) +[14:46:00.114057] Averaged stats: lr: 0.000288 loss: 0.4157 (0.4157) +[14:46:01.395548] val: [0/3] eta: 0:00:03 loss: 0.1207 (0.1207) time: 1.2613 data: 1.2389 max mem: 6806 +[14:46:01.432367] val: [2/3] eta: 0:00:00 loss: 0.1207 (0.6740) time: 0.4325 data: 0.4131 max mem: 6806 +[14:46:01.502750] val: Total time: 0:00:01 (0.4564 s / it) +[14:46:01.511729] val loss: 0.6739963019887606 +[14:46:01.511962] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8208, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7685, Kappa: 0.0000, Score: 0.4191 +[14:46:01.556692] Best epoch = 15, Best score = 0.4203 +[14:46:01.811933] log_dir: ./output_logs/retfound +[14:46:03.072139] Epoch: [17] [0/1] eta: 0:00:01 lr: 0.000286 loss: 0.4006 (0.4006) time: 1.2593 data: 1.1978 max mem: 6806 +[14:46:03.452993] Epoch: [17] Total time: 0:00:01 (1.6409 s / it) +[14:46:03.453766] Averaged stats: lr: 0.000286 loss: 0.4006 (0.4006) +[14:46:04.758197] val: [0/3] eta: 0:00:03 loss: 0.1146 (0.1146) time: 1.2977 data: 1.2840 max mem: 6806 +[14:46:04.776927] val: [2/3] eta: 0:00:00 loss: 0.1146 (0.6800) time: 0.4386 data: 0.4281 max mem: 6806 +[14:46:05.183570] val: Total time: 0:00:01 (0.5746 s / it) +[14:46:05.192360] val loss: 0.679985207815965 +[14:46:05.192551] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8199, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7681, Kappa: 0.0000, Score: 0.4188 +[14:46:05.558080] Best epoch = 15, Best score = 0.4203 +[14:46:06.326333] log_dir: ./output_logs/retfound +[14:46:07.530792] Epoch: [18] [0/1] eta: 0:00:01 lr: 0.000284 loss: 0.6902 (0.6902) time: 1.2036 data: 1.1486 max mem: 6806 +[14:46:07.604000] Epoch: [18] Total time: 0:00:01 (1.2775 s / it) +[14:46:07.604867] Averaged stats: lr: 0.000284 loss: 0.6902 (0.6902) +[14:46:08.991678] val: [0/3] eta: 0:00:04 loss: 0.1168 (0.1168) time: 1.3761 data: 1.3425 max mem: 6806 +[14:46:09.049605] val: [2/3] eta: 0:00:00 loss: 0.1168 (0.6740) time: 0.4778 data: 0.4476 max mem: 6806 +[14:46:09.193981] val: Total time: 0:00:01 (0.5264 s / it) +[14:46:09.205097] val loss: 0.674045130610466 +[14:46:09.205291] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8190, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7644, Kappa: 0.0000, Score: 0.4185 +[14:46:09.325347] Best epoch = 15, Best score = 0.4203 +[14:46:10.212862] log_dir: ./output_logs/retfound +[14:46:11.516863] Epoch: [19] [0/1] eta: 0:00:01 lr: 0.000281 loss: 0.6759 (0.6759) time: 1.3031 data: 1.2489 max mem: 6806 +[14:46:11.586477] Epoch: [19] Total time: 0:00:01 (1.3734 s / it) +[14:46:11.587430] Averaged stats: lr: 0.000281 loss: 0.6759 (0.6759) +[14:46:12.842562] val: [0/3] eta: 0:00:03 loss: 0.1216 (0.1216) time: 1.2436 data: 1.2295 max mem: 6806 +[14:46:12.861420] val: [2/3] eta: 0:00:00 loss: 0.1216 (0.6656) time: 0.4206 data: 0.4099 max mem: 6806 +[14:46:12.935125] val: Total time: 0:00:01 (0.4456 s / it) +[14:46:12.943844] val loss: 0.6655619144439697 +[14:46:12.944070] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8217, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7711, Kappa: 0.0000, Score: 0.4194 +[14:46:12.991252] Best epoch = 15, Best score = 0.4203 +[14:46:13.250750] log_dir: ./output_logs/retfound +[14:46:14.505618] Epoch: [20] [0/1] eta: 0:00:01 lr: 0.000279 loss: 0.0941 (0.0941) time: 1.2539 data: 1.1011 max mem: 6806 +[14:46:14.576193] Epoch: [20] Total time: 0:00:01 (1.3253 s / it) +[14:46:14.592045] Averaged stats: lr: 0.000279 loss: 0.0941 (0.0941) +[14:46:15.915419] val: [0/3] eta: 0:00:03 loss: 0.1224 (0.1224) time: 1.3120 data: 1.2826 max mem: 6806 +[14:46:15.961560] val: [2/3] eta: 0:00:00 loss: 0.1224 (0.6627) time: 0.4525 data: 0.4276 max mem: 6806 +[14:46:16.036459] val: Total time: 0:00:01 (0.4779 s / it) +[14:46:16.045413] val loss: 0.6626627519726753 +[14:46:16.045615] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8244, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7798, Kappa: 0.0000, Score: 0.4203 +[14:46:16.087168] Best epoch = 15, Best score = 0.4203 +[14:46:16.348828] log_dir: ./output_logs/retfound +[14:46:17.492207] Epoch: [21] [0/1] eta: 0:00:01 lr: 0.000276 loss: 0.5544 (0.5544) time: 1.1424 data: 1.0385 max mem: 6806 +[14:46:17.567078] Epoch: [21] Total time: 0:00:01 (1.2181 s / it) +[14:46:17.575694] Averaged stats: lr: 0.000276 loss: 0.5544 (0.5544) +[14:46:18.883265] val: [0/3] eta: 0:00:03 loss: 0.1271 (0.1271) time: 1.2966 data: 1.2754 max mem: 6806 +[14:46:18.920991] val: [2/3] eta: 0:00:00 loss: 0.1271 (0.6555) time: 0.4446 data: 0.4252 max mem: 6806 +[14:46:19.020977] val: Total time: 0:00:01 (0.4784 s / it) +[14:46:19.029869] val loss: 0.6554667179783186 +[14:46:19.030068] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8324, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7854, Kappa: 0.0000, Score: 0.4230 +[14:46:20.889086] Best epoch = 21, Best score = 0.4230 +[14:46:20.945764] log_dir: ./output_logs/retfound +[14:46:22.156113] Epoch: [22] [0/1] eta: 0:00:01 lr: 0.000272 loss: 0.4825 (0.4825) time: 1.2094 data: 1.1518 max mem: 6806 +[14:46:22.234661] Epoch: [22] Total time: 0:00:01 (1.2887 s / it) +[14:46:22.243374] Averaged stats: lr: 0.000272 loss: 0.4825 (0.4825) +[14:46:23.470871] val: [0/3] eta: 0:00:03 loss: 0.1340 (0.1340) time: 1.2167 data: 1.2025 max mem: 6806 +[14:46:23.509768] val: [2/3] eta: 0:00:00 loss: 0.1340 (0.6453) time: 0.4183 data: 0.4009 max mem: 6806 +[14:46:23.580437] val: Total time: 0:00:01 (0.4423 s / it) +[14:46:23.589579] val loss: 0.6452894409497579 +[14:46:23.589824] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8441, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7930, Kappa: 0.0000, Score: 0.4269 +[14:46:25.249162] Best epoch = 22, Best score = 0.4269 +[14:46:25.311391] log_dir: ./output_logs/retfound +[14:46:26.643754] Epoch: [23] [0/1] eta: 0:00:01 lr: 0.000269 loss: 0.5095 (0.5095) time: 1.3313 data: 1.1777 max mem: 6806 +[14:46:26.721935] Epoch: [23] Total time: 0:00:01 (1.4104 s / it) +[14:46:26.739076] Averaged stats: lr: 0.000269 loss: 0.5095 (0.5095) +[14:46:28.169440] val: [0/3] eta: 0:00:04 loss: 0.1412 (0.1412) time: 1.4182 data: 1.3968 max mem: 6806 +[14:46:28.207057] val: [2/3] eta: 0:00:00 loss: 0.1412 (0.6360) time: 0.4851 data: 0.4657 max mem: 6806 +[14:46:28.286161] val: Total time: 0:00:01 (0.5119 s / it) +[14:46:28.295571] val loss: 0.6360317667325338 +[14:46:28.295855] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8450, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8011, Kappa: 0.0000, Score: 0.4272 +[14:46:30.096680] Best epoch = 23, Best score = 0.4272 +[14:46:30.157496] log_dir: ./output_logs/retfound +[14:46:31.323297] Epoch: [24] [0/1] eta: 0:00:01 lr: 0.000265 loss: 0.5150 (0.5150) time: 1.1650 data: 1.1101 max mem: 6806 +[14:46:31.389005] Epoch: [24] Total time: 0:00:01 (1.2313 s / it) +[14:46:31.389964] Averaged stats: lr: 0.000265 loss: 0.5150 (0.5150) +[14:46:32.619116] val: [0/3] eta: 0:00:03 loss: 0.1486 (0.1486) time: 1.2107 data: 1.1842 max mem: 6806 +[14:46:32.649442] val: [2/3] eta: 0:00:00 loss: 0.1486 (0.6275) time: 0.4135 data: 0.3948 max mem: 6806 +[14:46:32.720815] val: Total time: 0:00:01 (0.4377 s / it) +[14:46:32.730347] val loss: 0.6275186985731125 +[14:46:32.730528] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8423, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7997, Kappa: 0.0000, Score: 0.4263 +[14:46:32.773127] Best epoch = 23, Best score = 0.4272 +[14:46:33.039717] log_dir: ./output_logs/retfound +[14:46:34.367748] Epoch: [25] [0/1] eta: 0:00:01 lr: 0.000261 loss: 0.6087 (0.6087) time: 1.3270 data: 1.2237 max mem: 6806 +[14:46:34.441387] Epoch: [25] Total time: 0:00:01 (1.4014 s / it) +[14:46:34.450523] Averaged stats: lr: 0.000261 loss: 0.6087 (0.6087) +[14:46:35.812894] val: [0/3] eta: 0:00:04 loss: 0.1569 (0.1569) time: 1.3525 data: 1.3179 max mem: 6806 +[14:46:35.869460] val: [2/3] eta: 0:00:00 loss: 0.1569 (0.6191) time: 0.4695 data: 0.4394 max mem: 6806 +[14:46:35.939165] val: Total time: 0:00:01 (0.4931 s / it) +[14:46:35.948069] val loss: 0.619118923942248 +[14:46:35.948248] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8432, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7999, Kappa: 0.0000, Score: 0.4266 +[14:46:35.985606] Best epoch = 23, Best score = 0.4272 +[14:46:36.238205] log_dir: ./output_logs/retfound +[14:46:37.518952] Epoch: [26] [0/1] eta: 0:00:01 lr: 0.000257 loss: 0.4349 (0.4349) time: 1.2797 data: 1.1792 max mem: 6806 +[14:46:37.592297] Epoch: [26] Total time: 0:00:01 (1.3539 s / it) +[14:46:37.609840] Averaged stats: lr: 0.000257 loss: 0.4349 (0.4349) +[14:46:38.931677] val: [0/3] eta: 0:00:03 loss: 0.1667 (0.1667) time: 1.3076 data: 1.2851 max mem: 6806 +[14:46:38.969533] val: [2/3] eta: 0:00:00 loss: 0.1667 (0.6093) time: 0.4483 data: 0.4284 max mem: 6806 +[14:46:39.040225] val: Total time: 0:00:01 (0.4723 s / it) +[14:46:39.049070] val loss: 0.6093153357505798 +[14:46:39.049249] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8459, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8139, Kappa: 0.0000, Score: 0.4275 +[14:46:40.832489] Best epoch = 26, Best score = 0.4275 +[14:46:40.900747] log_dir: ./output_logs/retfound +[14:46:42.292414] Epoch: [27] [0/1] eta: 0:00:01 lr: 0.000253 loss: 0.5196 (0.5196) time: 1.3903 data: 1.2845 max mem: 6806 +[14:46:42.370735] Epoch: [27] Total time: 0:00:01 (1.4698 s / it) +[14:46:42.380853] Averaged stats: lr: 0.000253 loss: 0.5196 (0.5196) +[14:46:43.763173] val: [0/3] eta: 0:00:04 loss: 0.1781 (0.1781) time: 1.3662 data: 1.3282 max mem: 6806 +[14:46:43.831731] val: [2/3] eta: 0:00:00 loss: 0.1781 (0.5992) time: 0.4781 data: 0.4428 max mem: 6806 +[14:46:43.904764] val: Total time: 0:00:01 (0.5028 s / it) +[14:46:43.913567] val loss: 0.599237730105718 +[14:46:43.913738] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8450, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8133, Kappa: 0.0000, Score: 0.4272 +[14:46:43.960683] Best epoch = 26, Best score = 0.4275 +[14:46:44.200935] log_dir: ./output_logs/retfound +[14:46:45.398724] Epoch: [28] [0/1] eta: 0:00:01 lr: 0.000248 loss: 0.6431 (0.6431) time: 1.1967 data: 1.0975 max mem: 6806 +[14:46:45.467693] Epoch: [28] Total time: 0:00:01 (1.2666 s / it) +[14:46:45.476444] Averaged stats: lr: 0.000248 loss: 0.6431 (0.6431) +[14:46:46.776247] val: [0/3] eta: 0:00:03 loss: 0.1923 (0.1923) time: 1.2811 data: 1.2604 max mem: 6806 +[14:46:46.832910] val: [2/3] eta: 0:00:00 loss: 0.1923 (0.5869) time: 0.4457 data: 0.4202 max mem: 6806 +[14:46:46.903949] val: Total time: 0:00:01 (0.4698 s / it) +[14:46:46.912909] val loss: 0.5868910153706869 +[14:46:46.913125] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8432, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7977, Kappa: 0.0000, Score: 0.4266 +[14:46:46.964847] Best epoch = 26, Best score = 0.4275 +[14:46:47.219888] log_dir: ./output_logs/retfound +[14:46:48.388383] Epoch: [29] [0/1] eta: 0:00:01 lr: 0.000243 loss: 0.4065 (0.4065) time: 1.1676 data: 1.1056 max mem: 6806 +[14:46:48.462114] Epoch: [29] Total time: 0:00:01 (1.2421 s / it) +[14:46:48.462988] Averaged stats: lr: 0.000243 loss: 0.4065 (0.4065) +[14:46:49.758768] val: [0/3] eta: 0:00:03 loss: 0.2013 (0.2013) time: 1.2848 data: 1.2541 max mem: 6806 +[14:46:49.815335] val: [2/3] eta: 0:00:00 loss: 0.2013 (0.5779) time: 0.4469 data: 0.4181 max mem: 6806 +[14:46:49.885859] val: Total time: 0:00:01 (0.4709 s / it) +[14:46:49.894552] val loss: 0.5778889854749044 +[14:46:49.894720] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8423, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7997, Kappa: 0.0000, Score: 0.4263 +[14:46:49.936851] Best epoch = 26, Best score = 0.4275 +[14:46:50.190425] log_dir: ./output_logs/retfound +[14:46:51.629711] Epoch: [30] [0/1] eta: 0:00:01 lr: 0.000238 loss: 0.6147 (0.6147) time: 1.4383 data: 1.3341 max mem: 6806 +[14:46:51.701287] Epoch: [30] Total time: 0:00:01 (1.5107 s / it) +[14:46:51.710158] Averaged stats: lr: 0.000238 loss: 0.6147 (0.6147) +[14:46:52.966239] val: [0/3] eta: 0:00:03 loss: 0.2084 (0.2084) time: 1.2455 data: 1.2241 max mem: 6806 +[14:46:53.003984] val: [2/3] eta: 0:00:00 loss: 0.2084 (0.5710) time: 0.4275 data: 0.4081 max mem: 6806 +[14:46:53.079240] val: Total time: 0:00:01 (0.4530 s / it) +[14:46:53.088132] val loss: 0.5709526985883713 +[14:46:53.088346] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8387, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7976, Kappa: 0.0000, Score: 0.4251 +[14:46:53.125413] Best epoch = 26, Best score = 0.4275 +[14:46:53.356950] log_dir: ./output_logs/retfound +[14:46:54.636511] Epoch: [31] [0/1] eta: 0:00:01 lr: 0.000233 loss: 0.4571 (0.4571) time: 1.2782 data: 1.1695 max mem: 6806 +[14:46:54.706754] Epoch: [31] Total time: 0:00:01 (1.3496 s / it) +[14:46:54.716328] Averaged stats: lr: 0.000233 loss: 0.4571 (0.4571) +[14:46:56.102581] val: [0/3] eta: 0:00:04 loss: 0.2076 (0.2076) time: 1.3757 data: 1.3420 max mem: 6806 +[14:46:56.160446] val: [2/3] eta: 0:00:00 loss: 0.2076 (0.5683) time: 0.4776 data: 0.4474 max mem: 6806 +[14:46:56.234976] val: Total time: 0:00:01 (0.5029 s / it) +[14:46:56.243822] val loss: 0.5683173437913259 +[14:46:56.244012] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8387, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7976, Kappa: 0.0000, Score: 0.4251 +[14:46:56.298112] Best epoch = 26, Best score = 0.4275 +[14:46:56.523781] log_dir: ./output_logs/retfound +[14:46:57.716064] Epoch: [32] [0/1] eta: 0:00:01 lr: 0.000227 loss: 0.4525 (0.4525) time: 1.1914 data: 1.1338 max mem: 6806 +[14:46:57.786565] Epoch: [32] Total time: 0:00:01 (1.2626 s / it) +[14:46:57.787372] Averaged stats: lr: 0.000227 loss: 0.4525 (0.4525) +[14:46:59.061553] val: [0/3] eta: 0:00:03 loss: 0.2028 (0.2028) time: 1.2635 data: 1.2423 max mem: 6806 +[14:46:59.094235] val: [2/3] eta: 0:00:00 loss: 0.2028 (0.5689) time: 0.4319 data: 0.4142 max mem: 6806 +[14:46:59.165947] val: Total time: 0:00:01 (0.4562 s / it) +[14:46:59.177120] val loss: 0.5689453134934107 +[14:46:59.177363] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8351, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7984, Kappa: 0.0000, Score: 0.4239 +[14:46:59.220193] Best epoch = 26, Best score = 0.4275 +[14:46:59.465410] log_dir: ./output_logs/retfound +[14:47:00.777661] Epoch: [33] [0/1] eta: 0:00:01 lr: 0.000222 loss: 0.6168 (0.6168) time: 1.3112 data: 1.2129 max mem: 6806 +[14:47:00.850017] Epoch: [33] Total time: 0:00:01 (1.3844 s / it) +[14:47:00.859691] Averaged stats: lr: 0.000222 loss: 0.6168 (0.6168) +[14:47:02.212523] val: [0/3] eta: 0:00:04 loss: 0.2006 (0.2006) time: 1.3421 data: 1.3104 max mem: 6806 +[14:47:02.270845] val: [2/3] eta: 0:00:00 loss: 0.2006 (0.5682) time: 0.4666 data: 0.4369 max mem: 6806 +[14:47:02.343541] val: Total time: 0:00:01 (0.4913 s / it) +[14:47:02.353110] val loss: 0.5681925465663274 +[14:47:02.353323] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8306, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7964, Kappa: 0.0000, Score: 0.4224 +[14:47:02.406689] Best epoch = 26, Best score = 0.4275 +[14:47:02.652666] log_dir: ./output_logs/retfound +[14:47:04.013301] Epoch: [34] [0/1] eta: 0:00:01 lr: 0.000216 loss: 0.5654 (0.5654) time: 1.3596 data: 1.2092 max mem: 6806 +[14:47:04.084828] Epoch: [34] Total time: 0:00:01 (1.4320 s / it) +[14:47:04.103358] Averaged stats: lr: 0.000216 loss: 0.5654 (0.5654) +[14:47:05.393860] val: [0/3] eta: 0:00:03 loss: 0.1989 (0.1989) time: 1.2794 data: 1.2654 max mem: 6806 +[14:47:05.412982] val: [2/3] eta: 0:00:00 loss: 0.1989 (0.5679) time: 0.4326 data: 0.4219 max mem: 6806 +[14:47:05.486419] val: Total time: 0:00:01 (0.4575 s / it) +[14:47:05.495259] val loss: 0.5678521196047465 +[14:47:05.495442] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8280, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7953, Kappa: 0.0000, Score: 0.4215 +[14:47:05.529547] Best epoch = 26, Best score = 0.4275 +[14:47:05.812487] log_dir: ./output_logs/retfound +[14:47:07.027395] Epoch: [35] [0/1] eta: 0:00:01 lr: 0.000210 loss: 0.5257 (0.5257) time: 1.2140 data: 1.1589 max mem: 6806 +[14:47:07.103067] Epoch: [35] Total time: 0:00:01 (1.2904 s / it) +[14:47:07.104107] Averaged stats: lr: 0.000210 loss: 0.5257 (0.5257) +[14:47:08.502097] val: [0/3] eta: 0:00:04 loss: 0.1978 (0.1978) time: 1.3833 data: 1.3504 max mem: 6806 +[14:47:08.550792] val: [2/3] eta: 0:00:00 loss: 0.1978 (0.5664) time: 0.4771 data: 0.4503 max mem: 6806 +[14:47:08.623859] val: Total time: 0:00:01 (0.5019 s / it) +[14:47:08.632653] val loss: 0.5664157519737879 +[14:47:08.632891] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8280, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7953, Kappa: 0.0000, Score: 0.4215 +[14:47:08.678003] Best epoch = 26, Best score = 0.4275 +[14:47:08.921971] log_dir: ./output_logs/retfound +[14:47:10.308253] Epoch: [36] [0/1] eta: 0:00:01 lr: 0.000204 loss: 0.4080 (0.4080) time: 1.3852 data: 1.2330 max mem: 6806 +[14:47:10.379067] Epoch: [36] Total time: 0:00:01 (1.4569 s / it) +[14:47:10.391923] Averaged stats: lr: 0.000204 loss: 0.4080 (0.4080) +[14:47:11.693422] val: [0/3] eta: 0:00:03 loss: 0.1945 (0.1945) time: 1.2909 data: 1.2680 max mem: 6806 +[14:47:11.731187] val: [2/3] eta: 0:00:00 loss: 0.1945 (0.5664) time: 0.4427 data: 0.4228 max mem: 6806 +[14:47:11.799894] val: Total time: 0:00:01 (0.4661 s / it) +[14:47:11.810847] val loss: 0.5664116591215134 +[14:47:11.811089] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8306, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8092, Kappa: 0.0000, Score: 0.4224 +[14:47:11.857679] Best epoch = 26, Best score = 0.4275 +[14:47:12.087921] log_dir: ./output_logs/retfound +[14:47:13.433334] Epoch: [37] [0/1] eta: 0:00:01 lr: 0.000198 loss: 0.3489 (0.3489) time: 1.3443 data: 1.2382 max mem: 6806 +[14:47:13.514353] Epoch: [37] Total time: 0:00:01 (1.4263 s / it) +[14:47:13.522302] Averaged stats: lr: 0.000198 loss: 0.3489 (0.3489) +[14:47:14.847840] val: [0/3] eta: 0:00:03 loss: 0.1872 (0.1872) time: 1.3104 data: 1.2942 max mem: 6806 +[14:47:14.866428] val: [2/3] eta: 0:00:00 loss: 0.1872 (0.5692) time: 0.4428 data: 0.4315 max mem: 6806 +[14:47:14.940950] val: Total time: 0:00:01 (0.4681 s / it) +[14:47:14.949759] val loss: 0.5691636900107065 +[14:47:14.949944] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8342, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8096, Kappa: 0.0000, Score: 0.4236 +[14:47:14.988494] Best epoch = 26, Best score = 0.4275 +[14:47:15.252937] log_dir: ./output_logs/retfound +[14:47:16.499347] Epoch: [38] [0/1] eta: 0:00:01 lr: 0.000192 loss: 0.4751 (0.4751) time: 1.2453 data: 1.1880 max mem: 6806 +[14:47:16.647209] Epoch: [38] Total time: 0:00:01 (1.3941 s / it) +[14:47:16.657079] Averaged stats: lr: 0.000192 loss: 0.4751 (0.4751) +[14:47:18.019905] val: [0/3] eta: 0:00:04 loss: 0.1801 (0.1801) time: 1.3402 data: 1.3063 max mem: 6806 +[14:47:18.078127] val: [2/3] eta: 0:00:00 loss: 0.1801 (0.5721) time: 0.4659 data: 0.4355 max mem: 6806 +[14:47:18.187812] val: Total time: 0:00:01 (0.5029 s / it) +[14:47:18.196736] val loss: 0.5720546990633011 +[14:47:18.196962] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8378, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8137, Kappa: 0.0000, Score: 0.4248 +[14:47:18.265763] Best epoch = 26, Best score = 0.4275 +[14:47:18.536279] log_dir: ./output_logs/retfound +[14:47:19.801029] Epoch: [39] [0/1] eta: 0:00:01 lr: 0.000186 loss: 0.3688 (0.3688) time: 1.2632 data: 1.1583 max mem: 6806 +[14:47:19.878217] Epoch: [39] Total time: 0:00:01 (1.3418 s / it) +[14:47:19.886820] Averaged stats: lr: 0.000186 loss: 0.3688 (0.3688) +[14:47:21.230186] val: [0/3] eta: 0:00:03 loss: 0.1731 (0.1731) time: 1.3208 data: 1.2943 max mem: 6806 +[14:47:21.286746] val: [2/3] eta: 0:00:00 loss: 0.1731 (0.5756) time: 0.4589 data: 0.4315 max mem: 6806 +[14:47:21.356845] val: Total time: 0:00:01 (0.4827 s / it) +[14:47:21.366744] val loss: 0.5755900094906489 +[14:47:21.366931] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8351, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8126, Kappa: 0.0000, Score: 0.4239 +[14:47:21.411535] Best epoch = 26, Best score = 0.4275 +[14:47:21.643081] log_dir: ./output_logs/retfound +[14:47:23.086535] Epoch: [40] [0/1] eta: 0:00:01 lr: 0.000179 loss: 0.5461 (0.5461) time: 1.4424 data: 1.3831 max mem: 6806 +[14:47:23.164774] Epoch: [40] Total time: 0:00:01 (1.5215 s / it) +[14:47:23.165656] Averaged stats: lr: 0.000179 loss: 0.5461 (0.5461) +[14:47:24.546158] val: [0/3] eta: 0:00:04 loss: 0.1690 (0.1690) time: 1.3697 data: 1.3474 max mem: 6806 +[14:47:24.586561] val: [2/3] eta: 0:00:00 loss: 0.1690 (0.5773) time: 0.4698 data: 0.4492 max mem: 6806 +[14:47:24.662268] val: Total time: 0:00:01 (0.4956 s / it) +[14:47:24.672850] val loss: 0.5773444126049677 +[14:47:24.673048] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8306, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8100, Kappa: 0.0000, Score: 0.4224 +[14:47:24.714700] Best epoch = 26, Best score = 0.4275 +[14:47:24.947841] log_dir: ./output_logs/retfound +[14:47:26.304669] Epoch: [41] [0/1] eta: 0:00:01 lr: 0.000173 loss: 0.5000 (0.5000) time: 1.3559 data: 1.2523 max mem: 6806 +[14:47:26.385989] Epoch: [41] Total time: 0:00:01 (1.4380 s / it) +[14:47:26.395780] Averaged stats: lr: 0.000173 loss: 0.5000 (0.5000) +[14:47:27.684824] val: [0/3] eta: 0:00:03 loss: 0.1660 (0.1660) time: 1.2777 data: 1.2479 max mem: 6806 +[14:47:27.742617] val: [2/3] eta: 0:00:00 loss: 0.1660 (0.5783) time: 0.4450 data: 0.4161 max mem: 6806 +[14:47:27.812933] val: Total time: 0:00:01 (0.4689 s / it) +[14:47:27.821684] val loss: 0.578347439567248 +[14:47:27.821878] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8253, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8073, Kappa: 0.0000, Score: 0.4206 +[14:47:27.865782] Best epoch = 26, Best score = 0.4275 +[14:47:28.137855] log_dir: ./output_logs/retfound +[14:47:29.432550] Epoch: [42] [0/1] eta: 0:00:01 lr: 0.000167 loss: 0.3663 (0.3663) time: 1.2937 data: 1.1976 max mem: 6806 +[14:47:29.509558] Epoch: [42] Total time: 0:00:01 (1.3716 s / it) +[14:47:29.519370] Averaged stats: lr: 0.000167 loss: 0.3663 (0.3663) +[14:47:30.946225] val: [0/3] eta: 0:00:04 loss: 0.1617 (0.1617) time: 1.4156 data: 1.3922 max mem: 6806 +[14:47:31.002347] val: [2/3] eta: 0:00:00 loss: 0.1617 (0.5806) time: 0.4904 data: 0.4642 max mem: 6806 +[14:47:31.075598] val: Total time: 0:00:01 (0.5152 s / it) +[14:47:31.084469] val loss: 0.5806369533141454 +[14:47:31.084664] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8244, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8073, Kappa: 0.0000, Score: 0.4203 +[14:47:31.130282] Best epoch = 26, Best score = 0.4275 +[14:47:31.405403] log_dir: ./output_logs/retfound +[14:47:32.806109] Epoch: [43] [0/1] eta: 0:00:01 lr: 0.000160 loss: 0.6681 (0.6681) time: 1.3997 data: 1.3413 max mem: 6806 +[14:47:32.877355] Epoch: [43] Total time: 0:00:01 (1.4718 s / it) +[14:47:32.878230] Averaged stats: lr: 0.000160 loss: 0.6681 (0.6681) +[14:47:34.204521] val: [0/3] eta: 0:00:03 loss: 0.1614 (0.1614) time: 1.3152 data: 1.3016 max mem: 6806 +[14:47:34.223497] val: [2/3] eta: 0:00:00 loss: 0.1614 (0.5793) time: 0.4446 data: 0.4340 max mem: 6806 +[14:47:34.295120] val: Total time: 0:00:01 (0.4688 s / it) +[14:47:34.304876] val loss: 0.5792928238709768 +[14:47:34.305077] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8217, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8055, Kappa: 0.0000, Score: 0.4194 +[14:47:34.340160] Best epoch = 26, Best score = 0.4275 +[14:47:34.853652] log_dir: ./output_logs/retfound +[14:47:36.227586] Epoch: [44] [0/1] eta: 0:00:01 lr: 0.000154 loss: 0.4009 (0.4009) time: 1.3729 data: 1.2442 max mem: 6806 +[14:47:36.300575] Epoch: [44] Total time: 0:00:01 (1.4467 s / it) +[14:47:36.308661] Averaged stats: lr: 0.000154 loss: 0.4009 (0.4009) +[14:47:38.073331] val: [0/3] eta: 0:00:05 loss: 0.1605 (0.1605) time: 1.7393 data: 1.7122 max mem: 6806 +[14:47:38.111117] val: [2/3] eta: 0:00:00 loss: 0.1605 (0.5787) time: 0.5921 data: 0.5709 max mem: 6806 +[14:47:38.187455] val: Total time: 0:00:01 (0.6181 s / it) +[14:47:38.201867] val loss: 0.5787177334229151 +[14:47:38.202177] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8172, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8037, Kappa: 0.0000, Score: 0.4179 +[14:47:38.264645] Best epoch = 26, Best score = 0.4275 +[14:47:38.542480] log_dir: ./output_logs/retfound +[14:47:39.997645] Epoch: [45] [0/1] eta: 0:00:01 lr: 0.000147 loss: 0.3046 (0.3046) time: 1.4538 data: 1.3014 max mem: 6806 +[14:47:40.066192] Epoch: [45] Total time: 0:00:01 (1.5235 s / it) +[14:47:40.082217] Averaged stats: lr: 0.000147 loss: 0.3046 (0.3046) +[14:47:41.490103] val: [0/3] eta: 0:00:04 loss: 0.1588 (0.1588) time: 1.3816 data: 1.3659 max mem: 6806 +[14:47:41.508750] val: [2/3] eta: 0:00:00 loss: 0.1588 (0.5784) time: 0.4666 data: 0.4554 max mem: 6806 +[14:47:41.582004] val: Total time: 0:00:01 (0.4914 s / it) +[14:47:41.591342] val loss: 0.5784457723299662 +[14:47:41.591554] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8136, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8020, Kappa: 0.0000, Score: 0.4167 +[14:47:41.651735] Best epoch = 26, Best score = 0.4275 +[14:47:41.983966] log_dir: ./output_logs/retfound +[14:47:43.464628] Epoch: [46] [0/1] eta: 0:00:01 lr: 0.000140 loss: 0.6236 (0.6236) time: 1.4797 data: 1.4116 max mem: 6806 +[14:47:43.536331] Epoch: [46] Total time: 0:00:01 (1.5522 s / it) +[14:47:43.544303] Averaged stats: lr: 0.000140 loss: 0.6236 (0.6236) +[14:47:45.272717] val: [0/3] eta: 0:00:05 loss: 0.1602 (0.1602) time: 1.6750 data: 1.6518 max mem: 6806 +[14:47:45.310629] val: [2/3] eta: 0:00:00 loss: 0.1602 (0.5756) time: 0.5708 data: 0.5507 max mem: 6806 +[14:47:45.383640] val: Total time: 0:00:01 (0.5956 s / it) +[14:47:45.392727] val loss: 0.575597474972407 +[14:47:45.392919] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8136, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8020, Kappa: 0.0000, Score: 0.4167 +[14:47:45.472039] Best epoch = 26, Best score = 0.4275 +[14:47:45.809828] log_dir: ./output_logs/retfound +[14:47:47.526451] Epoch: [47] [0/1] eta: 0:00:01 lr: 0.000134 loss: 0.5427 (0.5427) time: 1.7155 data: 1.5197 max mem: 6806 +[14:47:47.602426] Epoch: [47] Total time: 0:00:01 (1.7924 s / it) +[14:47:47.628656] Averaged stats: lr: 0.000134 loss: 0.5427 (0.5427) +[14:47:49.258170] val: [0/3] eta: 0:00:04 loss: 0.1622 (0.1622) time: 1.5960 data: 1.5729 max mem: 6806 +[14:47:49.290396] val: [2/3] eta: 0:00:00 loss: 0.1622 (0.5721) time: 0.5426 data: 0.5244 max mem: 6806 +[14:47:49.364178] val: Total time: 0:00:01 (0.5676 s / it) +[14:47:49.374021] val loss: 0.572087953488032 +[14:47:49.374202] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:47:49.436597] Best epoch = 26, Best score = 0.4275 +[14:47:49.775077] log_dir: ./output_logs/retfound +[14:47:51.189216] Epoch: [48] [0/1] eta: 0:00:01 lr: 0.000127 loss: 0.5618 (0.5618) time: 1.4133 data: 1.3551 max mem: 6806 +[14:47:51.256980] Epoch: [48] Total time: 0:00:01 (1.4817 s / it) +[14:47:51.257885] Averaged stats: lr: 0.000127 loss: 0.5618 (0.5618) +[14:47:52.905723] val: [0/3] eta: 0:00:04 loss: 0.1663 (0.1663) time: 1.6290 data: 1.6075 max mem: 6806 +[14:47:52.943432] val: [2/3] eta: 0:00:00 loss: 0.1663 (0.5664) time: 0.5554 data: 0.5360 max mem: 6806 +[14:47:53.017366] val: Total time: 0:00:01 (0.5804 s / it) +[14:47:53.029760] val loss: 0.5663845290740331 +[14:47:53.029984] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:47:53.099021] Best epoch = 26, Best score = 0.4275 +[14:47:53.430671] log_dir: ./output_logs/retfound +[14:47:54.944321] Epoch: [49] [0/1] eta: 0:00:01 lr: 0.000121 loss: 0.5629 (0.5629) time: 1.5126 data: 1.3093 max mem: 6806 +[14:47:55.122285] Epoch: [49] Total time: 0:00:01 (1.6914 s / it) +[14:47:55.142094] Averaged stats: lr: 0.000121 loss: 0.5629 (0.5629) +[14:47:56.676112] val: [0/3] eta: 0:00:04 loss: 0.1713 (0.1713) time: 1.4970 data: 1.4755 max mem: 6806 +[14:47:56.714010] val: [2/3] eta: 0:00:00 loss: 0.1713 (0.5605) time: 0.5114 data: 0.4919 max mem: 6806 +[14:47:56.785570] val: Total time: 0:00:01 (0.5357 s / it) +[14:47:56.794522] val loss: 0.5604729404052099 +[14:47:56.794733] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8004, Kappa: 0.0000, Score: 0.4156 +[14:47:56.874677] Best epoch = 26, Best score = 0.4275 +[14:47:57.245702] log_dir: ./output_logs/retfound +[14:47:58.803539] Epoch: [50] [0/1] eta: 0:00:01 lr: 0.000114 loss: 0.4204 (0.4204) time: 1.5567 data: 1.4520 max mem: 6806 +[14:47:58.892076] Epoch: [50] Total time: 0:00:01 (1.6461 s / it) +[14:47:58.893374] Averaged stats: lr: 0.000114 loss: 0.4204 (0.4204) +[14:48:00.457621] val: [0/3] eta: 0:00:04 loss: 0.1751 (0.1751) time: 1.5503 data: 1.5358 max mem: 6806 +[14:48:00.476410] val: [2/3] eta: 0:00:00 loss: 0.1751 (0.5560) time: 0.5228 data: 0.5120 max mem: 6806 +[14:48:00.548198] val: Total time: 0:00:01 (0.5471 s / it) +[14:48:00.558578] val loss: 0.5559929559628168 +[14:48:00.558857] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8004, Kappa: 0.0000, Score: 0.4156 +[14:48:00.614057] Best epoch = 26, Best score = 0.4275 +[14:48:00.973116] log_dir: ./output_logs/retfound +[14:48:02.436731] Epoch: [51] [0/1] eta: 0:00:01 lr: 0.000108 loss: 0.4891 (0.4891) time: 1.4625 data: 1.2871 max mem: 6806 +[14:48:02.506599] Epoch: [51] Total time: 0:00:01 (1.5333 s / it) +[14:48:02.521196] Averaged stats: lr: 0.000108 loss: 0.4891 (0.4891) +[14:48:04.056371] val: [0/3] eta: 0:00:04 loss: 0.1797 (0.1797) time: 1.4775 data: 1.4550 max mem: 6806 +[14:48:04.094264] val: [2/3] eta: 0:00:00 loss: 0.1797 (0.5511) time: 0.5049 data: 0.4851 max mem: 6806 +[14:48:04.164922] val: Total time: 0:00:01 (0.5289 s / it) +[14:48:04.173715] val loss: 0.5510979443788528 +[14:48:04.173921] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:48:04.244017] Best epoch = 26, Best score = 0.4275 +[14:48:04.554420] log_dir: ./output_logs/retfound +[14:48:06.156156] Epoch: [52] [0/1] eta: 0:00:01 lr: 0.000102 loss: 0.6048 (0.6048) time: 1.6005 data: 1.4949 max mem: 6806 +[14:48:06.227639] Epoch: [52] Total time: 0:00:01 (1.6730 s / it) +[14:48:06.236405] Averaged stats: lr: 0.000102 loss: 0.6048 (0.6048) +[14:48:07.795370] val: [0/3] eta: 0:00:04 loss: 0.1853 (0.1853) time: 1.4933 data: 1.4752 max mem: 6806 +[14:48:07.820507] val: [2/3] eta: 0:00:00 loss: 0.1853 (0.5463) time: 0.5059 data: 0.4918 max mem: 6806 +[14:48:07.896467] val: Total time: 0:00:01 (0.5318 s / it) +[14:48:07.905471] val loss: 0.5462958514690399 +[14:48:07.905740] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8004, Kappa: 0.0000, Score: 0.4156 +[14:48:07.978503] Best epoch = 26, Best score = 0.4275 +[14:48:08.339178] log_dir: ./output_logs/retfound +[14:48:09.819492] Epoch: [53] [0/1] eta: 0:00:01 lr: 0.000096 loss: 0.3634 (0.3634) time: 1.4794 data: 1.4118 max mem: 6806 +[14:48:09.893250] Epoch: [53] Total time: 0:00:01 (1.5539 s / it) +[14:48:09.894040] Averaged stats: lr: 0.000096 loss: 0.3634 (0.3634) +[14:48:11.294257] val: [0/3] eta: 0:00:04 loss: 0.1900 (0.1900) time: 1.3551 data: 1.3354 max mem: 6806 +[14:48:11.333644] val: [2/3] eta: 0:00:00 loss: 0.1900 (0.5422) time: 0.4646 data: 0.4453 max mem: 6806 +[14:48:11.409214] val: Total time: 0:00:01 (0.4903 s / it) +[14:48:11.418840] val loss: 0.5421813776095709 +[14:48:11.419022] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:48:11.493755] Best epoch = 26, Best score = 0.4275 +[14:48:11.809243] log_dir: ./output_logs/retfound +[14:48:13.359984] Epoch: [54] [0/1] eta: 0:00:01 lr: 0.000090 loss: 0.5021 (0.5021) time: 1.5498 data: 1.4468 max mem: 6806 +[14:48:13.435269] Epoch: [54] Total time: 0:00:01 (1.6258 s / it) +[14:48:13.444605] Averaged stats: lr: 0.000090 loss: 0.5021 (0.5021) +[14:48:14.833585] val: [0/3] eta: 0:00:04 loss: 0.1944 (0.1944) time: 1.3485 data: 1.3252 max mem: 6806 +[14:48:14.871523] val: [2/3] eta: 0:00:00 loss: 0.1944 (0.5387) time: 0.4620 data: 0.4418 max mem: 6806 +[14:48:14.955366] val: Total time: 0:00:01 (0.4903 s / it) +[14:48:14.964205] val loss: 0.5387451301018397 +[14:48:14.964406] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:48:15.005686] Best epoch = 26, Best score = 0.4275 +[14:48:15.279874] log_dir: ./output_logs/retfound +[14:48:16.746004] Epoch: [55] [0/1] eta: 0:00:01 lr: 0.000084 loss: 0.4826 (0.4826) time: 1.4652 data: 1.4110 max mem: 6806 +[14:48:16.819834] Epoch: [55] Total time: 0:00:01 (1.5398 s / it) +[14:48:16.820761] Averaged stats: lr: 0.000084 loss: 0.4826 (0.4826) +[14:48:18.522665] val: [0/3] eta: 0:00:05 loss: 0.1989 (0.1989) time: 1.6739 data: 1.6509 max mem: 6806 +[14:48:18.556703] val: [2/3] eta: 0:00:00 loss: 0.1989 (0.5358) time: 0.5691 data: 0.5504 max mem: 6806 +[14:48:18.637861] val: Total time: 0:00:01 (0.5966 s / it) +[14:48:18.646810] val loss: 0.5357923855384191 +[14:48:18.647087] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8100, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8003, Kappa: 0.0000, Score: 0.4156 +[14:48:18.727732] Best epoch = 26, Best score = 0.4275 +[14:48:19.084906] log_dir: ./output_logs/retfound +[14:48:20.648691] Epoch: [56] [0/1] eta: 0:00:01 lr: 0.000078 loss: 0.3971 (0.3971) time: 1.5625 data: 1.4568 max mem: 6806 +[14:48:20.719448] Epoch: [56] Total time: 0:00:01 (1.6344 s / it) +[14:48:20.728817] Averaged stats: lr: 0.000078 loss: 0.3971 (0.3971) +[14:48:22.133365] val: [0/3] eta: 0:00:04 loss: 0.2025 (0.2025) time: 1.3893 data: 1.3657 max mem: 6806 +[14:48:22.171131] val: [2/3] eta: 0:00:00 loss: 0.2025 (0.5332) time: 0.4755 data: 0.4553 max mem: 6806 +[14:48:22.247530] val: Total time: 0:00:01 (0.5014 s / it) +[14:48:22.256572] val loss: 0.5332065224647522 +[14:48:22.256789] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8065, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7987, Kappa: 0.0000, Score: 0.4144 +[14:48:22.313123] Best epoch = 26, Best score = 0.4275 +[14:48:22.657670] log_dir: ./output_logs/retfound +[14:48:24.066077] Epoch: [57] [0/1] eta: 0:00:01 lr: 0.000072 loss: 0.3556 (0.3556) time: 1.4076 data: 1.3529 max mem: 6806 +[14:48:24.143651] Epoch: [57] Total time: 0:00:01 (1.4858 s / it) +[14:48:24.144586] Averaged stats: lr: 0.000072 loss: 0.3556 (0.3556) +[14:48:25.607673] val: [0/3] eta: 0:00:04 loss: 0.2053 (0.2053) time: 1.4441 data: 1.4259 max mem: 6806 +[14:48:25.629663] val: [2/3] eta: 0:00:00 loss: 0.2053 (0.5309) time: 0.4885 data: 0.4754 max mem: 6806 +[14:48:25.702905] val: Total time: 0:00:01 (0.5134 s / it) +[14:48:25.711716] val loss: 0.5309115995963415 +[14:48:25.711897] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8038, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4135 +[14:48:25.781424] Best epoch = 26, Best score = 0.4275 +[14:48:26.112537] log_dir: ./output_logs/retfound +[14:48:27.575638] Epoch: [58] [0/1] eta: 0:00:01 lr: 0.000067 loss: 0.3549 (0.3549) time: 1.4622 data: 1.3592 max mem: 6806 +[14:48:27.645667] Epoch: [58] Total time: 0:00:01 (1.5330 s / it) +[14:48:27.653898] Averaged stats: lr: 0.000067 loss: 0.3549 (0.3549) +[14:48:29.271073] val: [0/3] eta: 0:00:04 loss: 0.2072 (0.2072) time: 1.5978 data: 1.5765 max mem: 6806 +[14:48:29.308698] val: [2/3] eta: 0:00:00 loss: 0.2072 (0.5292) time: 0.5449 data: 0.5256 max mem: 6806 +[14:48:29.378905] val: Total time: 0:00:01 (0.5688 s / it) +[14:48:29.388764] val loss: 0.5292012492815653 +[14:48:29.388978] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7948, Kappa: 0.0000, Score: 0.4132 +[14:48:29.505708] Best epoch = 26, Best score = 0.4275 +[14:48:29.783169] log_dir: ./output_logs/retfound +[14:48:31.250351] Epoch: [59] [0/1] eta: 0:00:01 lr: 0.000061 loss: 0.3529 (0.3529) time: 1.4661 data: 1.3692 max mem: 6806 +[14:48:31.321768] Epoch: [59] Total time: 0:00:01 (1.5384 s / it) +[14:48:31.322537] Averaged stats: lr: 0.000061 loss: 0.3529 (0.3529) +[14:48:32.654123] val: [0/3] eta: 0:00:03 loss: 0.2074 (0.2074) time: 1.3086 data: 1.2939 max mem: 6806 +[14:48:32.706652] val: [2/3] eta: 0:00:00 loss: 0.2074 (0.5285) time: 0.4535 data: 0.4425 max mem: 6806 +[14:48:32.779563] val: Total time: 0:00:01 (0.4782 s / it) +[14:48:32.789478] val loss: 0.5284674863020579 +[14:48:32.789700] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7948, Kappa: 0.0000, Score: 0.4132 +[14:48:32.858141] Best epoch = 26, Best score = 0.4275 +[14:48:33.186329] log_dir: ./output_logs/retfound +[14:48:34.702134] Epoch: [60] [0/1] eta: 0:00:01 lr: 0.000056 loss: 0.3511 (0.3511) time: 1.5148 data: 1.4136 max mem: 6806 +[14:48:34.771953] Epoch: [60] Total time: 0:00:01 (1.5855 s / it) +[14:48:34.780003] Averaged stats: lr: 0.000056 loss: 0.3511 (0.3511) +[14:48:36.421290] val: [0/3] eta: 0:00:04 loss: 0.2066 (0.2066) time: 1.5861 data: 1.5650 max mem: 6806 +[14:48:36.459184] val: [2/3] eta: 0:00:00 loss: 0.2066 (0.5283) time: 0.5412 data: 0.5218 max mem: 6806 +[14:48:36.531545] val: Total time: 0:00:01 (0.5657 s / it) +[14:48:36.540470] val loss: 0.528290460507075 +[14:48:36.540669] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7948, Kappa: 0.0000, Score: 0.4132 +[14:48:36.592261] Best epoch = 26, Best score = 0.4275 +[14:48:36.850007] log_dir: ./output_logs/retfound +[14:48:38.420742] Epoch: [61] [0/1] eta: 0:00:01 lr: 0.000051 loss: 0.5986 (0.5986) time: 1.5699 data: 1.4680 max mem: 6806 +[14:48:38.491369] Epoch: [61] Total time: 0:00:01 (1.6412 s / it) +[14:48:38.500726] Averaged stats: lr: 0.000051 loss: 0.5986 (0.5986) +[14:48:39.957720] val: [0/3] eta: 0:00:04 loss: 0.2069 (0.2069) time: 1.4464 data: 1.4321 max mem: 6806 +[14:48:39.976828] val: [2/3] eta: 0:00:00 loss: 0.2069 (0.5277) time: 0.4883 data: 0.4775 max mem: 6806 +[14:48:40.051726] val: Total time: 0:00:01 (0.5137 s / it) +[14:48:40.060747] val loss: 0.5276516228914261 +[14:48:40.060931] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7934, Kappa: 0.0000, Score: 0.4132 +[14:48:40.101809] Best epoch = 26, Best score = 0.4275 +[14:48:40.402171] log_dir: ./output_logs/retfound +[14:48:41.810655] Epoch: [62] [0/1] eta: 0:00:01 lr: 0.000046 loss: 0.6416 (0.6416) time: 1.4076 data: 1.3520 max mem: 6806 +[14:48:41.878821] Epoch: [62] Total time: 0:00:01 (1.4765 s / it) +[14:48:41.887877] Averaged stats: lr: 0.000046 loss: 0.6416 (0.6416) +[14:48:43.338949] val: [0/3] eta: 0:00:04 loss: 0.2071 (0.2071) time: 1.4324 data: 1.4093 max mem: 6806 +[14:48:43.377009] val: [2/3] eta: 0:00:00 loss: 0.2071 (0.5269) time: 0.4900 data: 0.4698 max mem: 6806 +[14:48:43.450446] val: Total time: 0:00:01 (0.5149 s / it) +[14:48:43.459401] val loss: 0.5269436438878378 +[14:48:43.459604] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7934, Kappa: 0.0000, Score: 0.4132 +[14:48:43.507039] Best epoch = 26, Best score = 0.4275 +[14:48:43.838216] log_dir: ./output_logs/retfound +[14:48:45.448532] Epoch: [63] [0/1] eta: 0:00:01 lr: 0.000041 loss: 0.4927 (0.4927) time: 1.6092 data: 1.5160 max mem: 6806 +[14:48:45.521217] Epoch: [63] Total time: 0:00:01 (1.6828 s / it) +[14:48:45.530406] Averaged stats: lr: 0.000041 loss: 0.4927 (0.4927) +[14:48:47.158424] val: [0/3] eta: 0:00:04 loss: 0.2072 (0.2072) time: 1.6025 data: 1.5707 max mem: 6806 +[14:48:47.206093] val: [2/3] eta: 0:00:00 loss: 0.2072 (0.5264) time: 0.5498 data: 0.5237 max mem: 6806 +[14:48:47.274519] val: Total time: 0:00:01 (0.5731 s / it) +[14:48:47.283292] val loss: 0.5263915856679281 +[14:48:47.283475] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:48:47.364621] Best epoch = 26, Best score = 0.4275 +[14:48:47.641406] log_dir: ./output_logs/retfound +[14:48:49.046709] Epoch: [64] [0/1] eta: 0:00:01 lr: 0.000037 loss: 0.5486 (0.5486) time: 1.4045 data: 1.3501 max mem: 6806 +[14:48:49.115400] Epoch: [64] Total time: 0:00:01 (1.4738 s / it) +[14:48:49.116171] Averaged stats: lr: 0.000037 loss: 0.5486 (0.5486) +[14:48:50.562350] val: [0/3] eta: 0:00:04 loss: 0.2080 (0.2080) time: 1.4053 data: 1.3827 max mem: 6806 +[14:48:50.600314] val: [2/3] eta: 0:00:00 loss: 0.2080 (0.5255) time: 0.4809 data: 0.4610 max mem: 6806 +[14:48:50.673266] val: Total time: 0:00:01 (0.5056 s / it) +[14:48:50.682406] val loss: 0.5254543672005335 +[14:48:50.682579] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:48:50.731979] Best epoch = 26, Best score = 0.4275 +[14:48:51.029607] log_dir: ./output_logs/retfound +[14:48:52.424876] Epoch: [65] [0/1] eta: 0:00:01 lr: 0.000033 loss: 0.4323 (0.4323) time: 1.3942 data: 1.2895 max mem: 6806 +[14:48:52.494804] Epoch: [65] Total time: 0:00:01 (1.4650 s / it) +[14:48:52.503209] Averaged stats: lr: 0.000033 loss: 0.4323 (0.4323) +[14:48:54.120863] val: [0/3] eta: 0:00:04 loss: 0.2084 (0.2084) time: 1.6061 data: 1.5830 max mem: 6806 +[14:48:54.158437] val: [2/3] eta: 0:00:00 loss: 0.2084 (0.5248) time: 0.5477 data: 0.5278 max mem: 6806 +[14:48:54.237804] val: Total time: 0:00:01 (0.5746 s / it) +[14:48:54.247898] val loss: 0.5248047063748041 +[14:48:54.248065] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:48:54.348541] Best epoch = 26, Best score = 0.4275 +[14:48:54.627133] log_dir: ./output_logs/retfound +[14:48:56.125040] Epoch: [66] [0/1] eta: 0:00:01 lr: 0.000029 loss: 0.3274 (0.3274) time: 1.4970 data: 1.4435 max mem: 6806 +[14:48:56.199025] Epoch: [66] Total time: 0:00:01 (1.5717 s / it) +[14:48:56.199789] Averaged stats: lr: 0.000029 loss: 0.3274 (0.3274) +[14:48:57.751265] val: [0/3] eta: 0:00:04 loss: 0.2085 (0.2085) time: 1.5264 data: 1.5023 max mem: 6806 +[14:48:57.793352] val: [2/3] eta: 0:00:00 loss: 0.2085 (0.5243) time: 0.5227 data: 0.5009 max mem: 6806 +[14:48:57.870402] val: Total time: 0:00:01 (0.5487 s / it) +[14:48:57.886150] val loss: 0.5243035157521566 +[14:48:57.886396] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:48:57.969314] Best epoch = 26, Best score = 0.4275 +[14:48:58.315535] log_dir: ./output_logs/retfound +[14:48:59.752234] Epoch: [67] [0/1] eta: 0:00:01 lr: 0.000025 loss: 0.4520 (0.4520) time: 1.4357 data: 1.3436 max mem: 6806 +[14:48:59.825336] Epoch: [67] Total time: 0:00:01 (1.5096 s / it) +[14:48:59.834302] Averaged stats: lr: 0.000025 loss: 0.4520 (0.4520) +[14:49:01.143991] val: [0/3] eta: 0:00:03 loss: 0.2086 (0.2086) time: 1.2941 data: 1.2720 max mem: 6806 +[14:49:01.184080] val: [2/3] eta: 0:00:00 loss: 0.2086 (0.5237) time: 0.4445 data: 0.4241 max mem: 6806 +[14:49:01.260658] val: Total time: 0:00:01 (0.4705 s / it) +[14:49:01.271443] val loss: 0.5237467288970947 +[14:49:01.271652] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:01.329211] Best epoch = 26, Best score = 0.4275 +[14:49:01.684722] log_dir: ./output_logs/retfound +[14:49:03.141285] Epoch: [68] [0/1] eta: 0:00:01 lr: 0.000022 loss: 0.6007 (0.6007) time: 1.4557 data: 1.4015 max mem: 6806 +[14:49:03.211844] Epoch: [68] Total time: 0:00:01 (1.5270 s / it) +[14:49:03.212635] Averaged stats: lr: 0.000022 loss: 0.6007 (0.6007) +[14:49:04.700361] val: [0/3] eta: 0:00:04 loss: 0.2090 (0.2090) time: 1.4596 data: 1.4456 max mem: 6806 +[14:49:04.719208] val: [2/3] eta: 0:00:00 loss: 0.2090 (0.5234) time: 0.4926 data: 0.4820 max mem: 6806 +[14:49:04.791717] val: Total time: 0:00:01 (0.5172 s / it) +[14:49:04.800506] val loss: 0.5233907103538513 +[14:49:04.800689] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:04.851953] Best epoch = 26, Best score = 0.4275 +[14:49:05.209201] log_dir: ./output_logs/retfound +[14:49:06.581392] Epoch: [69] [0/1] eta: 0:00:01 lr: 0.000018 loss: 0.4278 (0.4278) time: 1.3712 data: 1.2688 max mem: 6806 +[14:49:06.651242] Epoch: [69] Total time: 0:00:01 (1.4419 s / it) +[14:49:06.659648] Averaged stats: lr: 0.000018 loss: 0.4278 (0.4278) +[14:49:08.104844] val: [0/3] eta: 0:00:04 loss: 0.2091 (0.2091) time: 1.4345 data: 1.4124 max mem: 6806 +[14:49:08.142696] val: [2/3] eta: 0:00:00 loss: 0.2091 (0.5231) time: 0.4905 data: 0.4709 max mem: 6806 +[14:49:08.217779] val: Total time: 0:00:01 (0.5160 s / it) +[14:49:08.226563] val loss: 0.5231079260508219 +[14:49:08.226744] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:08.293102] Best epoch = 26, Best score = 0.4275 +[14:49:08.648513] log_dir: ./output_logs/retfound +[14:49:10.116712] Epoch: [70] [0/1] eta: 0:00:01 lr: 0.000015 loss: 0.3226 (0.3226) time: 1.4673 data: 1.3647 max mem: 6806 +[14:49:10.185366] Epoch: [70] Total time: 0:00:01 (1.5367 s / it) +[14:49:10.186289] Averaged stats: lr: 0.000015 loss: 0.3226 (0.3226) +[14:49:11.540180] val: [0/3] eta: 0:00:03 loss: 0.2090 (0.2090) time: 1.3245 data: 1.3099 max mem: 6806 +[14:49:11.559045] val: [2/3] eta: 0:00:00 loss: 0.2090 (0.5229) time: 0.4476 data: 0.4367 max mem: 6806 +[14:49:11.629464] val: Total time: 0:00:01 (0.4715 s / it) +[14:49:11.638160] val loss: 0.5228807131449381 +[14:49:11.638343] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:11.698333] Best epoch = 26, Best score = 0.4275 +[14:49:11.999258] log_dir: ./output_logs/retfound +[14:49:13.474002] Epoch: [71] [0/1] eta: 0:00:01 lr: 0.000013 loss: 0.3636 (0.3636) time: 1.4737 data: 1.3754 max mem: 6806 +[14:49:13.545323] Epoch: [71] Total time: 0:00:01 (1.5459 s / it) +[14:49:13.553856] Averaged stats: lr: 0.000013 loss: 0.3636 (0.3636) +[14:49:14.891941] val: [0/3] eta: 0:00:03 loss: 0.2088 (0.2088) time: 1.3104 data: 1.2882 max mem: 6806 +[14:49:14.929572] val: [2/3] eta: 0:00:00 loss: 0.2088 (0.5228) time: 0.4492 data: 0.4295 max mem: 6806 +[14:49:14.998128] val: Total time: 0:00:01 (0.4724 s / it) +[14:49:15.006919] val loss: 0.5228237559398016 +[14:49:15.007116] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:15.082775] Best epoch = 26, Best score = 0.4275 +[14:49:15.380158] log_dir: ./output_logs/retfound +[14:49:16.800112] Epoch: [72] [0/1] eta: 0:00:01 lr: 0.000010 loss: 0.3389 (0.3389) time: 1.4191 data: 1.3206 max mem: 6806 +[14:49:16.867536] Epoch: [72] Total time: 0:00:01 (1.4872 s / it) +[14:49:16.875985] Averaged stats: lr: 0.000010 loss: 0.3389 (0.3389) +[14:49:18.271859] val: [0/3] eta: 0:00:04 loss: 0.2085 (0.2085) time: 1.3852 data: 1.3717 max mem: 6806 +[14:49:18.290237] val: [2/3] eta: 0:00:00 loss: 0.2085 (0.5228) time: 0.4677 data: 0.4573 max mem: 6806 +[14:49:18.359845] val: Total time: 0:00:01 (0.4913 s / it) +[14:49:18.368547] val loss: 0.5227776517470678 +[14:49:18.368722] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8029, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7933, Kappa: 0.0000, Score: 0.4132 +[14:49:18.439467] Best epoch = 26, Best score = 0.4275 +[14:49:18.728230] log_dir: ./output_logs/retfound +[14:49:20.021803] Epoch: [73] [0/1] eta: 0:00:01 lr: 0.000008 loss: 0.4097 (0.4097) time: 1.2927 data: 1.2390 max mem: 6806 +[14:49:20.091414] Epoch: [73] Total time: 0:00:01 (1.3630 s / it) +[14:49:20.092209] Averaged stats: lr: 0.000008 loss: 0.4097 (0.4097) +[14:49:21.455067] val: [0/3] eta: 0:00:04 loss: 0.2083 (0.2083) time: 1.3418 data: 1.3182 max mem: 6806 +[14:49:21.492885] val: [2/3] eta: 0:00:00 loss: 0.2083 (0.5227) time: 0.4597 data: 0.4395 max mem: 6806 +[14:49:21.567317] val: Total time: 0:00:01 (0.4849 s / it) +[14:49:21.576199] val loss: 0.5226725240548452 +[14:49:21.576378] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8020, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7929, Kappa: 0.0000, Score: 0.4129 +[14:49:21.624249] Best epoch = 26, Best score = 0.4275 +[14:49:21.901890] log_dir: ./output_logs/retfound +[14:49:23.253879] Epoch: [74] [0/1] eta: 0:00:01 lr: 0.000006 loss: 0.3693 (0.3693) time: 1.3510 data: 1.2500 max mem: 6806 +[14:49:23.333891] Epoch: [74] Total time: 0:00:01 (1.4318 s / it) +[14:49:23.342479] Averaged stats: lr: 0.000006 loss: 0.3693 (0.3693) +[14:49:24.624938] val: [0/3] eta: 0:00:03 loss: 0.2083 (0.2083) time: 1.2552 data: 1.2323 max mem: 6806 +[14:49:24.659810] val: [2/3] eta: 0:00:00 loss: 0.2083 (0.5226) time: 0.4298 data: 0.4109 max mem: 6806 +[14:49:24.733332] val: Total time: 0:00:01 (0.4548 s / it) +[14:49:24.742151] val loss: 0.5226142058769861 +[14:49:24.742324] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8020, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7929, Kappa: 0.0000, Score: 0.4129 +[14:49:24.789881] Best epoch = 26, Best score = 0.4275 +[14:49:25.090177] log_dir: ./output_logs/retfound +[14:49:26.544845] Epoch: [75] [0/1] eta: 0:00:01 lr: 0.000005 loss: 0.4432 (0.4432) time: 1.4538 data: 1.3976 max mem: 6806 +[14:49:26.616224] Epoch: [75] Total time: 0:00:01 (1.5259 s / it) +[14:49:26.617032] Averaged stats: lr: 0.000005 loss: 0.4432 (0.4432) +[14:49:28.021218] val: [0/3] eta: 0:00:04 loss: 0.2082 (0.2082) time: 1.3835 data: 1.3697 max mem: 6806 +[14:49:28.039579] val: [2/3] eta: 0:00:00 loss: 0.2082 (0.5226) time: 0.4671 data: 0.4566 max mem: 6806 +[14:49:28.116180] val: Total time: 0:00:01 (0.4930 s / it) +[14:49:28.125005] val loss: 0.5226047188043594 +[14:49:28.125180] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8011, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7876, Kappa: 0.0000, Score: 0.4126 +[14:49:28.161978] Best epoch = 26, Best score = 0.4275 +[14:49:28.457518] log_dir: ./output_logs/retfound +[14:49:29.927950] Epoch: [76] [0/1] eta: 0:00:01 lr: 0.000003 loss: 0.5931 (0.5931) time: 1.4695 data: 1.3664 max mem: 6806 +[14:49:30.000961] Epoch: [76] Total time: 0:00:01 (1.5433 s / it) +[14:49:30.010733] Averaged stats: lr: 0.000003 loss: 0.5931 (0.5931) +[14:49:31.517768] val: [0/3] eta: 0:00:04 loss: 0.2083 (0.2083) time: 1.4852 data: 1.4642 max mem: 6806 +[14:49:31.554934] val: [2/3] eta: 0:00:00 loss: 0.2083 (0.5225) time: 0.5073 data: 0.4882 max mem: 6806 +[14:49:31.623854] val: Total time: 0:00:01 (0.5306 s / it) +[14:49:31.632721] val loss: 0.5224704146385193 +[14:49:31.632899] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8011, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7876, Kappa: 0.0000, Score: 0.4126 +[14:49:31.730817] Best epoch = 26, Best score = 0.4275 +[14:49:32.058037] log_dir: ./output_logs/retfound +[14:49:33.553327] Epoch: [77] [0/1] eta: 0:00:01 lr: 0.000002 loss: 0.5338 (0.5338) time: 1.4944 data: 1.4358 max mem: 6806 +[14:49:33.621158] Epoch: [77] Total time: 0:00:01 (1.5629 s / it) +[14:49:33.621909] Averaged stats: lr: 0.000002 loss: 0.5338 (0.5338) +[14:49:34.992457] val: [0/3] eta: 0:00:04 loss: 0.2083 (0.2083) time: 1.3502 data: 1.3364 max mem: 6806 +[14:49:35.012098] val: [2/3] eta: 0:00:00 loss: 0.2083 (0.5224) time: 0.4564 data: 0.4456 max mem: 6806 +[14:49:35.083441] val: Total time: 0:00:01 (0.4806 s / it) +[14:49:35.092512] val loss: 0.5223741233348846 +[14:49:35.092693] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8011, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7891, Kappa: 0.0000, Score: 0.4126 +[14:49:35.158072] Best epoch = 26, Best score = 0.4275 +[14:49:35.513987] log_dir: ./output_logs/retfound +[14:49:36.994020] Epoch: [78] [0/1] eta: 0:00:01 lr: 0.000002 loss: 0.4762 (0.4762) time: 1.4790 data: 1.3751 max mem: 6806 +[14:49:37.067249] Epoch: [78] Total time: 0:00:01 (1.5530 s / it) +[14:49:37.076428] Averaged stats: lr: 0.000002 loss: 0.4762 (0.4762) +[14:49:38.563518] val: [0/3] eta: 0:00:04 loss: 0.2083 (0.2083) time: 1.4600 data: 1.4393 max mem: 6806 +[14:49:38.601043] val: [2/3] eta: 0:00:00 loss: 0.2083 (0.5223) time: 0.4989 data: 0.4798 max mem: 6806 +[14:49:38.671714] val: Total time: 0:00:01 (0.5230 s / it) +[14:49:38.682411] val loss: 0.5223171611626943 +[14:49:38.682601] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8011, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7891, Kappa: 0.0000, Score: 0.4126 +[14:49:38.770459] Best epoch = 26, Best score = 0.4275 +[14:49:39.089581] log_dir: ./output_logs/retfound +[14:49:40.537605] Epoch: [79] [0/1] eta: 0:00:01 lr: 0.000001 loss: 0.4970 (0.4970) time: 1.4468 data: 1.3439 max mem: 6806 +[14:49:40.610480] Epoch: [79] Total time: 0:00:01 (1.5207 s / it) +[14:49:40.619168] Averaged stats: lr: 0.000001 loss: 0.4970 (0.4970) +[14:49:41.934864] val: [0/3] eta: 0:00:03 loss: 0.2084 (0.2084) time: 1.2974 data: 1.2836 max mem: 6806 +[14:49:41.954101] val: [2/3] eta: 0:00:00 loss: 0.2084 (0.5223) time: 0.4387 data: 0.4280 max mem: 6806 +[14:49:42.032803] val: Total time: 0:00:01 (0.4653 s / it) +[14:49:42.042127] val loss: 0.5222975065310796 +[14:49:42.042320] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8011, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7876, Kappa: 0.0000, Score: 0.4126 +[14:49:42.085274] Best epoch = 26, Best score = 0.4275 +[14:49:45.552387] Test with the best model, epoch = 26: +[14:49:47.275217] test: [0/6] eta: 0:00:10 loss: 0.1798 (0.1798) time: 1.7000 data: 1.6835 max mem: 6806 +[14:49:47.482691] test: [5/6] eta: 0:00:00 loss: 0.1798 (0.6167) time: 0.3178 data: 0.2807 max mem: 6806 +[14:49:47.552964] test: Total time: 0:00:01 (0.3297 s / it) +[14:49:47.562969] val loss: 0.616695836186409 +[14:49:47.563097] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8407, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7885, Kappa: 0.0000, Score: 0.4258 +[14:49:48.320642] Training time 0:04:48 +[rank0]:[W701 14:49:48.775845659 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/010/retfound acc=0.7750 auroc=0.8409498207885304 f1_macro=0.4366 qwk=0.0 diff --git a/results/downsample/adam/010/vit/confusion_matrix.png b/results/downsample/adam/010/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7b2e78188bd8b221136d92abcaec4515708131a5 --- /dev/null +++ b/results/downsample/adam/010/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd370a1ef3780a46d039db13ecedd18073ee54df94ad777b1b0bb633c679d2c5 +size 66049 diff --git a/results/downsample/adam/010/vit/log.csv b/results/downsample/adam/010/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..bec5ce1af0dcacc7d9efd1ba471e9e58c67ce929 --- /dev/null +++ b/results/downsample/adam/010/vit/log.csv @@ -0,0 +1,38 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7243645191192627,0.3,0.3261648745519713,0.16738866653636775,0.0 +1,0.8274303674697876,0.35,0.3978494623655914,0.2628592213617868,7.39441178203743e-08 +2,0.6562439203262329,0.35,0.5770609318996416,0.3333804310110334,1.478882356407486e-07 +3,0.7882642149925232,0.35,0.5770609318996416,0.3333804310110334,2.2183235346112292e-07 +4,0.7689127326011658,0.725,0.6702508960573477,0.3338106134129075,2.957764712814972e-07 +5,0.5189658999443054,0.725,0.6487455197132617,0.3871453957515585,3.697205891018715e-07 +6,0.802010178565979,0.75,0.6738351254480286,0.5586833216240569,3.6955843521557525e-07 +7,0.8098161220550537,0.65,0.6917562724014337,0.5099843647498272,3.6907225802970327e-07 +8,0.7024650573730469,0.625,0.7096774193548387,0.498913137988001,3.682629104642438e-07 +9,0.6406999826431274,0.625,0.7096774193548387,0.498913137988001,3.671318123898447e-07 +10,0.555525541305542,0.675,0.7491039426523298,0.5468616819396087,3.6568094813687817e-07 +11,0.381649911403656,0.75,0.7777777777777777,0.5933308724006398,3.6391286301425095e-07 +12,0.43236806988716125,0.8,0.7741935483870968,0.5436209922406036,3.618306588440675e-07 +13,0.46471673250198364,0.825,0.8100358422939068,0.5826866931667454,3.594379885199801e-07 +14,0.45462656021118164,0.825,0.8028673835125448,0.5802972069062914,3.567390495987718e-07 +15,0.45308369398117065,0.825,0.8172043010752688,0.5850761794271994,3.537385769364163e-07 +16,0.34881651401519775,0.825,0.8172043010752688,0.653639707612957,3.504418343815308e-07 +17,0.3694971799850464,0.775,0.8243727598566308,0.6310235025392602,3.4685460554079696e-07 +18,0.4109410047531128,0.725,0.8279569892473118,0.6112175638879962,3.4298318363255025e-07 +19,0.5166340470314026,0.725,0.8315412186379928,0.6124123070182232,3.3883436044633555e-07 +20,0.38884031772613525,0.725,0.8315412186379928,0.6124123070182232,3.344154144278013e-07 +21,0.46052056550979614,0.75,0.8387096774193549,0.6353450023850916,3.297340979098317e-07 +22,0.3884739875793457,0.8,0.8387096774193549,0.6801605163195562,3.2479862351232145e-07 +23,0.46891120076179504,0.85,0.8458781362007168,0.6918331928627056,3.1961764973444924e-07 +24,0.390653133392334,0.8,0.8530465949820788,0.5699053411055975,3.1420026576472814e-07 +25,0.40256136655807495,0.825,0.8566308243727598,0.6367204163906138,3.0855597553548053e-07 +26,0.3479497730731964,0.85,0.8530465949820788,0.7172485083143894,3.026946810497112e-07 +27,0.4970046579837799,0.825,0.8494623655913978,0.6885454657977244,2.9662666500962944e-07 +28,0.4633110761642456,0.8,0.8422939068100357,0.681355259449783,2.9036257277729527e-07 +29,0.37565603852272034,0.8,0.8387096774193548,0.6801605163195562,2.839133936990369e-07 +30,0.4257713556289673,0.775,0.8351254480286737,0.6558818234597349,2.7729044182640365e-07 +31,0.33132117986679077,0.75,0.8351254480286738,0.6341502592548646,2.705053360674741e-07 +32,0.30196624994277954,0.75,0.8315412186379928,0.6329555161246376,2.6356997980334384e-07 +33,0.41960060596466064,0.825,0.8387096774193548,0.7047878270778606,2.5649654000555013e-07 +34,0.28068774938583374,0.825,0.8351254480286737,0.6837664932768165,2.4929742589106926e-07 +35,0.27809739112854004,0.85,0.8422939068100357,0.7136642789237083,2.419852671523334e-07 +36,0.2821797728538513,0.85,0.8422939068100357,0.7136642789237083,2.3457289180045857e-07 diff --git a/results/downsample/adam/010/vit/metrics.json b/results/downsample/adam/010/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..85bc7a2169d536a70d37b58361cb7e88eb626919 --- /dev/null +++ b/results/downsample/adam/010/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8, + "balanced_accuracy": 0.6935483870967742, + "precision_macro": 0.7109375, + "recall_macro": 0.6935483870967742, + "f1_macro": 0.7012138188608776, + "precision_weighted": 0.792578125, + "recall_weighted": 0.8, + "f1_weighted": 0.7957049486461252, + "cohen_kappa": 0.4029850746268657, + "quadratic_weighted_kappa": 0.4029850746268657, + "mcc": 0.4041119295602435, + "auroc": 0.8221326164874552, + "auprc": 0.6603445975879552, + "sensitivity": 0.5, + "specificity": 0.8870967741935484, + "precision_pos": 0.5625, + "f1_pos": 0.5294117647058824, + "per_class": { + "0": { + "precision": 0.859375, + "recall": 0.8870967741935484, + "f1-score": 0.873015873015873, + "support": 62.0 + }, + "1": { + "precision": 0.5625, + "recall": 0.5, + "f1-score": 0.5294117647058824, + "support": 18.0 + }, + "accuracy": 0.8, + "macro avg": { + "precision": 0.7109375, + "recall": 0.6935483870967742, + "f1-score": 0.7012138188608776, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.792578125, + "recall": 0.8, + "f1-score": 0.7957049486461252, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/010/vit/pr.png b/results/downsample/adam/010/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..0710f3b5d1779f0a269bbf99192293e3e11f7dc1 --- /dev/null +++ b/results/downsample/adam/010/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3de32c7955f9ecaab9c1e49f6f121fc8e78c5437423ad37f96a9aedc0bd4cd0e +size 48420 diff --git a/results/downsample/adam/010/vit/roc.png b/results/downsample/adam/010/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..ff5568bb15c4133e1bd34ae1392d868e684e5c48 --- /dev/null +++ b/results/downsample/adam/010/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d64842793cc871cc265e10dc1e5ad372f13b2019a2d81a1953182cab58a61f78 +size 58349 diff --git a/results/downsample/adam/010/vit/test_pred.npz b/results/downsample/adam/010/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..ea80e8f6ffe09d6e3e60b6ea195c01b2879faf74 --- /dev/null +++ b/results/downsample/adam/010/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:81325336e3761f1cc0350c4c7ca49da2ec77b106dc744987aa2adf3bd2c9fe7e +size 1790 diff --git a/results/downsample/adam/010/vit/train.log b/results/downsample/adam/010/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..83d277d7e5f29d5697d840304c0519a33aa52e02 --- /dev/null +++ b/results/downsample/adam/010/vit/train.log @@ -0,0 +1,194 @@ +[vit] train=28 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7244 val_acc=0.3000 val_auc=0.3262 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.8274 val_acc=0.3500 val_auc=0.3978 score=0.2629 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6562 val_acc=0.3500 val_auc=0.5771 score=0.3334 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.7883 val_acc=0.3500 val_auc=0.5771 score=0.3334 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.7689 val_acc=0.7250 val_auc=0.6703 score=0.3338 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5190 val_acc=0.7250 val_auc=0.6487 score=0.3871 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.8020 val_acc=0.7500 val_auc=0.6738 score=0.5587 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.8098 val_acc=0.6500 val_auc=0.6918 score=0.5100 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.7025 val_acc=0.6250 val_auc=0.7097 score=0.4989 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.6407 val_acc=0.6250 val_auc=0.7097 score=0.4989 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.5555 val_acc=0.6750 val_auc=0.7491 score=0.5469 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3816 val_acc=0.7500 val_auc=0.7778 score=0.5933 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.4324 val_acc=0.8000 val_auc=0.7742 score=0.5436 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4647 val_acc=0.8250 val_auc=0.8100 score=0.5827 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4546 val_acc=0.8250 val_auc=0.8029 score=0.5803 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4531 val_acc=0.8250 val_auc=0.8172 score=0.5851 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3488 val_acc=0.8250 val_auc=0.8172 score=0.6536 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.3695 val_acc=0.7750 val_auc=0.8244 score=0.6310 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4109 val_acc=0.7250 val_auc=0.8280 score=0.6112 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.5166 val_acc=0.7250 val_auc=0.8315 score=0.6124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3888 val_acc=0.7250 val_auc=0.8315 score=0.6124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4605 val_acc=0.7500 val_auc=0.8387 score=0.6353 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3885 val_acc=0.8000 val_auc=0.8387 score=0.6802 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4689 val_acc=0.8500 val_auc=0.8459 score=0.6918 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3907 val_acc=0.8000 val_auc=0.8530 score=0.5699 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.4026 val_acc=0.8250 val_auc=0.8566 score=0.6367 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.3479 val_acc=0.8500 val_auc=0.8530 score=0.7172 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.4970 val_acc=0.8250 val_auc=0.8495 score=0.6885 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.4633 val_acc=0.8000 val_auc=0.8423 score=0.6814 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3757 val_acc=0.8000 val_auc=0.8387 score=0.6802 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.4258 val_acc=0.7750 val_auc=0.8351 score=0.6559 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.3313 val_acc=0.7500 val_auc=0.8351 score=0.6342 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep32 loss=0.3020 val_acc=0.7500 val_auc=0.8315 score=0.6330 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep33 loss=0.4196 val_acc=0.8250 val_auc=0.8387 score=0.7048 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep34 loss=0.2807 val_acc=0.8250 val_auc=0.8351 score=0.6838 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep35 loss=0.2781 val_acc=0.8500 val_auc=0.8423 score=0.7137 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep36 loss=0.2822 val_acc=0.8500 val_auc=0.8423 score=0.7137 +[vit] early stop at ep36 (best ep26 score=0.7172) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=26 best_val_score=0.7172 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/010/vit acc=0.8000 auroc=0.8221326164874552 f1_macro=0.7012 qwk=0.4029850746268657 diff --git a/results/downsample/adam/025/resnet/confusion_matrix.png b/results/downsample/adam/025/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..cd0ba064b29a16718b1c399d08b85d5bb97774de --- /dev/null +++ b/results/downsample/adam/025/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1abe73d3b561c1a666501db6004cef76710a551fda1431e012239a3373c648f5 +size 67295 diff --git a/results/downsample/adam/025/resnet/log.csv b/results/downsample/adam/025/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..d42b86d17d2d78b180ed98b4bf17b9c3acd51a45 --- /dev/null +++ b/results/downsample/adam/025/resnet/log.csv @@ -0,0 +1,43 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6914718747138977,0.475,0.4086021505376344,0.205972592909889,0.0 +1,0.691820502281189,0.5,0.3727598566308244,0.16913683870088816,0.00016666666666666666 +2,0.6871117949485779,0.675,0.3548387096774194,0.19903649952999972,0.0003333333333333333 +3,0.685950756072998,0.75,0.3046594982078853,0.22870350264371542,0.0005 +4,0.6725234985351562,0.75,0.3655913978494624,0.24901413585757445,0.0004994417196557883 +5,0.660677969455719,0.775,0.48028673835125446,0.3056354855537045,0.000497769372038695 +6,0.664720356464386,0.775,0.6129032258064515,0.3498409813721035,0.0004949904262591467 +7,0.6441075205802917,0.775,0.6810035842293907,0.37254110084641656,0.0004911172937635942 +8,0.6211735606193542,0.775,0.7383512544802868,0.39165699093004863,0.0004861672729019797 +9,0.6065897345542908,0.775,0.7921146953405018,0.4095781378834536,0.0004801624716691072 +10,0.6041999459266663,0.775,0.7956989247311828,0.41077288101368065,0.0004731297089649703 +11,0.5719426870346069,0.775,0.8387096774193548,0.4251097985764047,0.00046510039481503486 +12,0.6001823544502258,0.775,0.8172043010752689,0.41794133979504267,0.0004561103900854401 +13,0.5649173855781555,0.775,0.7849462365591399,0.40718865162299966,0.00044619984631966527 +14,0.5429705381393433,0.775,0.6917562724014337,0.37612533023709754,0.00043541302641198946 +15,0.5118892192840576,0.75,0.6487455197132617,0.34339884314550756,0.00042379810691866064 +16,0.4617708921432495,0.725,0.6379928315412187,0.3835611663608775,0.0004114069628897006 +17,0.48448413610458374,0.725,0.6379928315412187,0.3835611663608775,0.0003982949361823388 +18,0.46801692247390747,0.75,0.6308243727598566,0.48732975200944634,0.0003845205882908432 +19,0.4680742621421814,0.725,0.6344086021505375,0.4675606935653949,0.00037014543879667093 +20,0.44728195667266846,0.725,0.6236559139784947,0.49513253522885553,0.0003552336906070838 +21,0.4437512457370758,0.75,0.6415770609318996,0.5221027479091994,0.0003398519432093782 +22,0.389069527387619,0.75,0.6523297491039426,0.4944982107908083,0.0003240688952214085 +23,0.3154485821723938,0.75,0.6523297491039426,0.4944982107908083,0.0003079550375668821 +24,0.3127584159374237,0.775,0.6451612903225806,0.5147410680062136,0.00029158233864578256 +25,0.34374675154685974,0.775,0.6272401433691757,0.5087673523550785,0.0002750239229060246 +26,0.32273441553115845,0.775,0.6379928315412187,0.5123515817457596,0.0002583537442519187 +27,0.31403857469558716,0.775,0.6415770609318996,0.5135463248759865,0.00024164625574808144 +28,0.27025941014289856,0.775,0.6523297491039426,0.5171305542666675,0.0002249760770939754 +29,0.27681291103363037,0.75,0.6451612903225806,0.49210872453035437,0.0002084176613542175 +30,0.2363043576478958,0.75,0.6487455197132617,0.4933034676605814,0.00019204496243311792 +31,0.28110814094543457,0.8,0.6666666666666666,0.5464984188388443,0.00017593110477859153 +32,0.23425576090812683,0.775,0.6881720430107526,0.5290779855689375,0.00016014805679062183 +33,0.23924396932125092,0.775,0.6917562724014337,0.5302727286991645,0.0001447663093929163 +34,0.22234606742858887,0.75,0.6845878136200716,0.5052508989628514,0.00012985456120332905 +35,0.20810933411121368,0.75,0.6953405017921147,0.5088351283535325,0.00011547941170915685 +36,0.22648902237415314,0.775,0.6953405017921147,0.5314674718293916,0.0001017050638176612 +37,0.16878066956996918,0.775,0.7132616487455197,0.5374411874805266,8.85930371102994e-05 +38,0.17641764879226685,0.775,0.7132616487455197,0.5374411874805266,7.620189308133943e-05 +39,0.2414732724428177,0.75,0.7132616487455197,0.5148088440046674,6.458697358801061e-05 +40,0.1929554045200348,0.75,0.7204301075268817,0.5171983302651214,5.3800153680334754e-05 +41,0.18377751111984253,0.75,0.7168458781362007,0.5160035871348945,4.388960991455998e-05 diff --git a/results/downsample/adam/025/resnet/metrics.json b/results/downsample/adam/025/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..ac5e848e13225da43028ed9298c6647bd19e6f7b --- /dev/null +++ b/results/downsample/adam/025/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.825, + "balanced_accuracy": 0.7491039426523298, + "precision_macro": 0.7491039426523298, + "recall_macro": 0.7491039426523298, + "f1_macro": 0.7491039426523298, + "precision_weighted": 0.825, + "recall_weighted": 0.825, + "f1_weighted": 0.825, + "cohen_kappa": 0.4982078853046594, + "quadratic_weighted_kappa": 0.4982078853046594, + "mcc": 0.4982078853046595, + "auroc": 0.8494623655913978, + "auprc": 0.6600700201825375, + "sensitivity": 0.6111111111111112, + "specificity": 0.8870967741935484, + "precision_pos": 0.6111111111111112, + "f1_pos": 0.6111111111111112, + "per_class": { + "0": { + "precision": 0.8870967741935484, + "recall": 0.8870967741935484, + "f1-score": 0.8870967741935484, + "support": 62.0 + }, + "1": { + "precision": 0.6111111111111112, + "recall": 0.6111111111111112, + "f1-score": 0.6111111111111112, + "support": 18.0 + }, + "accuracy": 0.825, + "macro avg": { + "precision": 0.7491039426523298, + "recall": 0.7491039426523298, + "f1-score": 0.7491039426523298, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.825, + "recall": 0.825, + "f1-score": 0.825, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/025/resnet/pr.png b/results/downsample/adam/025/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..9ae6b448a687d5d930713f3ab6e7af3ca0024ed4 --- /dev/null +++ b/results/downsample/adam/025/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:213ad78611e37874b510ca19ef47db7bae6b724f65fc54207923360948e4a024 +size 52854 diff --git a/results/downsample/adam/025/resnet/roc.png b/results/downsample/adam/025/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..88696401292f86f973c53eaa61eb9844bd7ce6c4 --- /dev/null +++ b/results/downsample/adam/025/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:094c40e24bd262287a814af41462cbe3bcd58a1a85c000b4a26946e8c14d34be +size 57730 diff --git a/results/downsample/adam/025/resnet/test_pred.npz b/results/downsample/adam/025/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..2102c6bedd3b0b74072eb111040967d70f471685 --- /dev/null +++ b/results/downsample/adam/025/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:247542a86f412504a47a349a41e41a940aad712fbe4ad56191652f7e8cc37a2d +size 1790 diff --git a/results/downsample/adam/025/resnet/train.log b/results/downsample/adam/025/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..148602a1b8c881ad66f5c45c04039aad5999798d --- /dev/null +++ b/results/downsample/adam/025/resnet/train.log @@ -0,0 +1,219 @@ +[resnet] train=70 val=40 test=80 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6915 val_acc=0.4750 val_auc=0.4086 score=0.2060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6918 val_acc=0.5000 val_auc=0.3728 score=0.1691 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6871 val_acc=0.6750 val_auc=0.3548 score=0.1990 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6860 val_acc=0.7500 val_auc=0.3047 score=0.2287 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6725 val_acc=0.7500 val_auc=0.3656 score=0.2490 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6607 val_acc=0.7750 val_auc=0.4803 score=0.3056 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6647 val_acc=0.7750 val_auc=0.6129 score=0.3498 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6441 val_acc=0.7750 val_auc=0.6810 score=0.3725 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6212 val_acc=0.7750 val_auc=0.7384 score=0.3917 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6066 val_acc=0.7750 val_auc=0.7921 score=0.4096 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.6042 val_acc=0.7750 val_auc=0.7957 score=0.4108 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.5719 val_acc=0.7750 val_auc=0.8387 score=0.4251 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.6002 val_acc=0.7750 val_auc=0.8172 score=0.4179 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.5649 val_acc=0.7750 val_auc=0.7849 score=0.4072 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.5430 val_acc=0.7750 val_auc=0.6918 score=0.3761 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.5119 val_acc=0.7500 val_auc=0.6487 score=0.3434 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.4618 val_acc=0.7250 val_auc=0.6380 score=0.3836 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.4845 val_acc=0.7250 val_auc=0.6380 score=0.3836 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.4680 val_acc=0.7500 val_auc=0.6308 score=0.4873 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.4681 val_acc=0.7250 val_auc=0.6344 score=0.4676 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.4473 val_acc=0.7250 val_auc=0.6237 score=0.4951 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.4438 val_acc=0.7500 val_auc=0.6416 score=0.5221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.3891 val_acc=0.7500 val_auc=0.6523 score=0.4945 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.3154 val_acc=0.7500 val_auc=0.6523 score=0.4945 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.3128 val_acc=0.7750 val_auc=0.6452 score=0.5147 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.3437 val_acc=0.7750 val_auc=0.6272 score=0.5088 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.3227 val_acc=0.7750 val_auc=0.6380 score=0.5124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.3140 val_acc=0.7750 val_auc=0.6416 score=0.5135 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.2703 val_acc=0.7750 val_auc=0.6523 score=0.5171 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.2768 val_acc=0.7500 val_auc=0.6452 score=0.4921 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep30 loss=0.2363 val_acc=0.7500 val_auc=0.6487 score=0.4933 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep31 loss=0.2811 val_acc=0.8000 val_auc=0.6667 score=0.5465 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep32 loss=0.2343 val_acc=0.7750 val_auc=0.6882 score=0.5291 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep33 loss=0.2392 val_acc=0.7750 val_auc=0.6918 score=0.5303 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep34 loss=0.2223 val_acc=0.7500 val_auc=0.6846 score=0.5053 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep35 loss=0.2081 val_acc=0.7500 val_auc=0.6953 score=0.5088 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep36 loss=0.2265 val_acc=0.7750 val_auc=0.6953 score=0.5315 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep37 loss=0.1688 val_acc=0.7750 val_auc=0.7133 score=0.5374 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep38 loss=0.1764 val_acc=0.7750 val_auc=0.7133 score=0.5374 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep39 loss=0.2415 val_acc=0.7500 val_auc=0.7133 score=0.5148 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep40 loss=0.1930 val_acc=0.7500 val_auc=0.7204 score=0.5172 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep41 loss=0.1838 val_acc=0.7500 val_auc=0.7168 score=0.5160 +[resnet] early stop at ep41 (best ep31 score=0.5465) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=31 best_val_score=0.5465 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/025/resnet acc=0.8250 auroc=0.8494623655913978 f1_macro=0.7491 qwk=0.4982078853046594 diff --git a/results/downsample/adam/025/retfound/confusion_matrix.png b/results/downsample/adam/025/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..8cd1c0f10ba1369abffa786304692fe2ead74a61 --- /dev/null +++ b/results/downsample/adam/025/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7a177252fc9991b4f0282e49c832e75a801376395fa91ea7adb8ef656b91f44 +size 67542 diff --git a/results/downsample/adam/025/retfound/confusion_matrix_test.jpg b/results/downsample/adam/025/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..c96b09811118a796332dd44b50534baecbbe8da5 --- /dev/null +++ b/results/downsample/adam/025/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35d86c7d23d8eba281f704bd415ce819bc71b33be27f11393ede0755880b34ab +size 261022 diff --git a/results/downsample/adam/025/retfound/log.txt b/results/downsample/adam/025/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..9f691a1671113804edf94f8e316f5fad45e94b81 --- /dev/null +++ b/results/downsample/adam/025/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 1.5625e-05, "train_loss": 0.69281005859375, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 7.8125e-05, "train_loss": 0.6928024291992188, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014062500000000002, "train_loss": 0.6597137451171875, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00020312500000000002, "train_loss": 0.6111793518066406, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.000265625, "train_loss": 0.56707763671875, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00032812499999999997, "train_loss": 0.5063743591308594, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.000390625, "train_loss": 0.5323562622070312, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.000453125, "train_loss": 0.5140285491943359, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000515625, "train_loss": 0.5910167694091797, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000578125, "train_loss": 0.571446418762207, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006248797296535528, "train_loss": 0.5038890838623047, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006234377833473366, "train_loss": 0.5428504943847656, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006200818848757455, "train_loss": 0.49982261657714844, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006148327244688065, "train_loss": 0.49104785919189453, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006077226649459515, "train_loss": 0.4852256774902344, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005987955421884734, "train_loss": 0.4922332763671875, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005881063948767417, "train_loss": 0.45473146438598633, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005757211251584354, "train_loss": 0.464465856552124, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.000561716092339888, "train_loss": 0.4561605453491211, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005461776421055717, "train_loss": 0.48391997814178467, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005292015741682404, "train_loss": 0.45394015312194824, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005108925516318428, "train_loss": 0.3850417137145996, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004913634557086838, "train_loss": 0.39360034465789795, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00047073468976921304, "train_loss": 0.42000794410705566, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004491334370152081, "train_loss": 0.36667317152023315, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.000426692876353043, "train_loss": 0.4083402156829834, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004035513613014434, "train_loss": 0.40936511754989624, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003798515669960373, "train_loss": 0.3629714846611023, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035573961054969873, "train_loss": 0.4142100214958191, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003313641501919586, "train_loss": 0.36670446395874023, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030687546874158533, "train_loss": 0.4820868968963623, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002824245470630336, "train_loss": 0.4061429500579834, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00025816213321920806, "train_loss": 0.37096667289733887, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00023423781305952714, "train_loss": 0.33367133140563965, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00021079908797341933, "train_loss": 0.39954519271850586, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018799046549520726, "train_loss": 0.3364083170890808, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.0001659525683671057, "train_loss": 0.3989395499229431, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014482126755325402, "train_loss": 0.346619188785553, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012472684455004014, "train_loss": 0.3216207027435303, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010579318815735249, "train_loss": 0.3124226927757263, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.813703066293163e-05, "train_loss": 0.3022347688674927, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 7.186722814899894e-05, "train_loss": 0.44560563564300537, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.708408935831119e-05, "train_loss": 0.36727413535118103, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.387875725740529e-05, "train_loss": 0.3780505061149597, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.233264710990193e-05, "train_loss": 0.350458025932312, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.251694452433174e-05, "train_loss": 0.2747492790222168, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.449216657118709e-05, "train_loss": 0.3130863904953003, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 8.30778867505851e-06, "train_loss": 0.2996370196342468, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 4.001939582190023e-06, "train_loss": 0.33741360902786255, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.601166284079265e-06, "train_loss": 0.3337419927120209, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/adam/025/retfound/metrics.json b/results/downsample/adam/025/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..be39cab6e68b5526ff6fddb9f9685e4008720cbb --- /dev/null +++ b/results/downsample/adam/025/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.875, + "balanced_accuracy": 0.8602150537634409, + "precision_macro": 0.8150470219435737, + "recall_macro": 0.8602150537634409, + "f1_macro": 0.8333333333333333, + "precision_weighted": 0.8883228840125392, + "recall_weighted": 0.875, + "f1_weighted": 0.8791666666666667, + "cohen_kappa": 0.6677740863787376, + "quadratic_weighted_kappa": 0.6677740863787376, + "mcc": 0.6737497456694623, + "auroc": 0.921146953405018, + "auprc": 0.8693619531300691, + "sensitivity": 0.8333333333333334, + "specificity": 0.8870967741935484, + "precision_pos": 0.6818181818181818, + "f1_pos": 0.75, + "per_class": { + "0": { + "precision": 0.9482758620689655, + "recall": 0.8870967741935484, + "f1-score": 0.9166666666666666, + "support": 62.0 + }, + "1": { + "precision": 0.6818181818181818, + "recall": 0.8333333333333334, + "f1-score": 0.75, + "support": 18.0 + }, + "accuracy": 0.875, + "macro avg": { + "precision": 0.8150470219435737, + "recall": 0.8602150537634409, + "f1-score": 0.8333333333333333, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8883228840125392, + "recall": 0.875, + "f1-score": 0.8791666666666667, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/025/retfound/metrics_test.csv b/results/downsample/adam/025/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..43bc69e9b149cf034b7d0172ac1947c9971430f4 --- /dev/null +++ b/results/downsample/adam/025/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.30319035053253174,0.875,0.8333333333333333,0.921146953405018,0.125,0.7230769230769231,0.8150470219435737,0.8602150537634409,0.9200821070647697,0.6677740863787376 diff --git a/results/downsample/adam/025/retfound/metrics_val.csv b/results/downsample/adam/025/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..aaada85de03a278c1191c12604a67a4debd482d9 --- /dev/null +++ b/results/downsample/adam/025/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6931381225585938,0.775,0.43661971830985913,0.4704301075268817,0.225,0.3875,0.3875,0.5,0.4919117647058824,0.0 +0.6940689086914062,0.775,0.43661971830985913,0.6397849462365591,0.225,0.3875,0.3875,0.5,0.6120131643877164,0.0 +0.7008514404296875,0.775,0.43661971830985913,0.7123655913978495,0.225,0.3875,0.3875,0.5,0.679546217921357,0.0 +0.723785400390625,0.775,0.43661971830985913,0.7231182795698925,0.225,0.3875,0.3875,0.5,0.6834902498325579,0.0 +0.7757492065429688,0.775,0.43661971830985913,0.7302867383512546,0.225,0.3875,0.3875,0.5,0.6945541121766359,0.0 +0.8606357574462891,0.775,0.43661971830985913,0.764336917562724,0.225,0.3875,0.3875,0.5,0.7235162087521442,0.0 +0.9275722503662109,0.775,0.43661971830985913,0.7813620071684588,0.225,0.3875,0.3875,0.5,0.7394731333230635,0.0 +0.9076881408691406,0.775,0.43661971830985913,0.8001792114695341,0.225,0.3875,0.3875,0.5,0.7640976850397252,0.0 +0.8484306335449219,0.775,0.43661971830985913,0.8378136200716846,0.225,0.3875,0.3875,0.5,0.77912545924571,0.0 +0.7631969451904297,0.775,0.43661971830985913,0.860215053763441,0.225,0.3875,0.3875,0.5,0.8021001580714583,0.0 +0.6950168609619141,0.775,0.43661971830985913,0.860215053763441,0.225,0.3875,0.3875,0.5,0.7923548909195697,0.0 +0.6615047454833984,0.775,0.43661971830985913,0.8584229390681004,0.225,0.3875,0.3875,0.5,0.7836582910939376,0.0 +0.6754188537597656,0.775,0.43661971830985913,0.8566308243727598,0.225,0.3875,0.3875,0.5,0.7836582910939377,0.0 +0.724090576171875,0.775,0.43661971830985913,0.8530465949820789,0.225,0.3875,0.3875,0.5,0.7803156443199692,0.0 +0.7763900756835938,0.775,0.43661971830985913,0.8530465949820789,0.225,0.3875,0.3875,0.5,0.7803156443199692,0.0 +0.7758550643920898,0.775,0.43661971830985913,0.8530465949820789,0.225,0.3875,0.3875,0.5,0.7803156443199692,0.0 +0.7098367214202881,0.775,0.43661971830985913,0.8530465949820789,0.225,0.3875,0.3875,0.5,0.7766390171713557,0.0 +0.5545053482055664,0.775,0.5866819747416763,0.8449820788530467,0.225,0.47248803827751196,0.6527777777777778,0.578853046594982,0.767258665521233,0.1964285714285714 +0.44260621070861816,0.725,0.6925227113906359,0.8422939068100359,0.275,0.5386513157894737,0.6994949494949495,0.7831541218637992,0.762171757386584,0.4179894179894179 +0.41592884063720703,0.725,0.6925227113906359,0.8387096774193548,0.275,0.5386513157894737,0.6994949494949495,0.7831541218637992,0.7562650442218064,0.4179894179894179 +0.41899538040161133,0.75,0.7023809523809523,0.8422939068100359,0.25,0.554367201426025,0.6933333333333334,0.7598566308243728,0.7584561341951543,0.42028985507246375 +0.43852484226226807,0.75,0.6865203761755485,0.8494623655913979,0.25,0.5404411764705883,0.6752136752136753,0.7204301075268817,0.7674022370262524,0.3808049535603715 +0.4600367546081543,0.75,0.6865203761755485,0.8422939068100359,0.25,0.5404411764705883,0.6752136752136753,0.7204301075268817,0.7621427009470148,0.3808049535603715 +0.5097696781158447,0.75,0.6666666666666667,0.8494623655913979,0.25,0.5238095238095238,0.658307210031348,0.6810035842293907,0.7703217467588879,0.33554817275747506 +0.591159462928772,0.825,0.738562091503268,0.85663082437276,0.175,0.6083333333333334,0.75,0.7293906810035842,0.7886217530470097,0.4776119402985075 +0.6798206567764282,0.8,0.6078431372549019,0.8530465949820789,0.2,0.49473684210526314,0.7387387387387387,0.5949820788530467,0.7843360646215363,0.24882629107981213 +0.6662570238113403,0.8,0.6078431372549019,0.8530465949820788,0.2,0.49473684210526314,0.7387387387387387,0.5949820788530467,0.7935103685459151,0.24882629107981213 +0.631696343421936,0.825,0.7128205128205128,0.8458781362007168,0.175,0.5845959595959596,0.7598039215686274,0.6899641577060931,0.7791169351874279,0.43089430894308944 +0.5867966413497925,0.825,0.738562091503268,0.8494623655913978,0.175,0.6083333333333334,0.75,0.7293906810035842,0.7854061479500287,0.4776119402985075 +0.5493767261505127,0.8,0.7132616487455197,0.8422939068100359,0.2,0.5780219780219781,0.7132616487455197,0.7132616487455197,0.7851823944346485,0.42652329749103934 +0.5392974615097046,0.825,0.738562091503268,0.8422939068100359,0.175,0.6083333333333334,0.75,0.7293906810035842,0.7814786907309449,0.4776119402985075 +0.5296193361282349,0.825,0.738562091503268,0.8494623655913979,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8022931165114149,0.4776119402985075 +0.539230465888977,0.825,0.738562091503268,0.8566308243727598,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8077606136420267,0.4776119402985075 +0.5495936870574951,0.825,0.738562091503268,0.8646953405017921,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8186483034410068,0.4776119402985075 +0.5406581163406372,0.825,0.738562091503268,0.8611111111111112,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8151497381838567,0.4776119402985075 +0.5183157324790955,0.825,0.738562091503268,0.8620071684587813,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8209722382953754,0.4776119402985075 +0.5090571045875549,0.825,0.738562091503268,0.8637992831541218,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8229411701688372,0.4776119402985075 +0.5036530494689941,0.8,0.7132616487455197,0.8637992831541218,0.2,0.5780219780219781,0.7132616487455197,0.7132616487455197,0.8229411701688372,0.42652329749103934 +0.5054279565811157,0.8,0.7132616487455197,0.8637992831541219,0.2,0.5780219780219781,0.7132616487455197,0.7132616487455197,0.8229411701688372,0.42652329749103934 +0.5128768980503082,0.825,0.738562091503268,0.8655913978494624,0.175,0.6083333333333334,0.75,0.7293906810035842,0.822941170168837,0.4776119402985075 +0.5152709186077118,0.825,0.738562091503268,0.8664874551971327,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8242569596425213,0.4776119402985075 +0.5225353538990021,0.825,0.738562091503268,0.8673835125448028,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8249900974724333,0.4776119402985075 +0.5234984755516052,0.825,0.738562091503268,0.8781362007168458,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8468921290416154,0.4776119402985075 +0.5274004638195038,0.825,0.738562091503268,0.8781362007168458,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8469458924824755,0.4776119402985075 +0.5317047834396362,0.825,0.738562091503268,0.881720430107527,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8583786155281018,0.4776119402985075 +0.5344215035438538,0.825,0.738562091503268,0.881720430107527,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8583786155281018,0.4776119402985075 +0.5356009006500244,0.825,0.738562091503268,0.8790322580645162,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8542196568816302,0.4776119402985075 +0.5362502932548523,0.825,0.738562091503268,0.8781362007168458,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8542196568816302,0.4776119402985075 +0.5363571643829346,0.825,0.738562091503268,0.8781362007168458,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8542196568816303,0.4776119402985075 +0.536094069480896,0.825,0.738562091503268,0.8781362007168458,0.175,0.6083333333333334,0.75,0.7293906810035842,0.8542196568816303,0.4776119402985075 diff --git a/results/downsample/adam/025/retfound/pr.png b/results/downsample/adam/025/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..80ba27ae55ec8b26ffc7ba897da45ac375549000 --- /dev/null +++ b/results/downsample/adam/025/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:17fd47545b8704e8281f358ade6cd439f217d6010b6ec5c8236b17e3c72ac6a9 +size 43390 diff --git a/results/downsample/adam/025/retfound/roc.png b/results/downsample/adam/025/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..9ffc65725d328dfc035f21343f24fcb3de805e73 --- /dev/null +++ b/results/downsample/adam/025/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2fd467a99f8aaa292a6be6fdeaeba19006a45909a8499fbf15d1d940ca57dee +size 57179 diff --git a/results/downsample/adam/025/retfound/test_pred.npz b/results/downsample/adam/025/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..4be96c031b733bea89c2a80edd625896c48c8ce3 --- /dev/null +++ b/results/downsample/adam/025/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6879125c0c8434bb916a4bda9ff47d420f48898f62c7abd12178388d6275c07 +size 1470 diff --git a/results/downsample/adam/025/retfound/train.log b/results/downsample/adam/025/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4ce8407ab01d83f2c4be24981df97b908c390fec --- /dev/null +++ b/results/downsample/adam/025/retfound/train.log @@ -0,0 +1,732 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:42:25.258736650 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:42:26.868392] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:42:26.868877] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/adam_25', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/025', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:42:42.802746] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:44.751536] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:48.561153] Sampler_train = +[14:42:48.604601] len of train_set: 64 +[14:42:48.843563] [Adaptation] Full fine-tuning: training all parameters. +[14:42:48.844706] number of trainable params (M): 303.30 +[14:42:48.844815] base lr: 5.00e-03 +[14:42:48.844886] actual lr: 6.25e-04 +[14:42:48.844951] accumulate grad iterations: 1 +[14:42:48.845017] effective batch size: 32 +[14:42:48.847950] criterion = CrossEntropyLoss() +[14:42:48.848074] Start training for 50 epochs +[14:42:48.850024] log_dir: ./output_logs/retfound +[14:42:51.520852] Epoch: [0] [0/2] eta: 0:00:05 lr: 0.000000 loss: 0.6927 (0.6927) time: 2.6699 data: 2.0285 max mem: 7340 +[14:42:51.612441] Epoch: [0] [1/2] eta: 0:00:01 lr: 0.000031 loss: 0.6927 (0.6928) time: 1.3801 data: 1.0143 max mem: 7340 +[14:42:51.679568] Epoch: [0] Total time: 0:00:02 (1.4147 s / it) +[14:42:51.680550] Averaged stats: lr: 0.000031 loss: 0.6927 (0.6928) +[14:42:53.800535] val: [0/2] eta: 0:00:04 loss: 0.6924 (0.6924) time: 2.1057 data: 2.0804 max mem: 7340 +[14:42:53.880207] val: [1/2] eta: 0:00:01 loss: 0.6924 (0.6931) time: 1.0924 data: 1.0403 max mem: 7340 +[14:42:53.950024] val: Total time: 0:00:02 (1.1279 s / it) +[14:42:53.962258] val loss: 0.6931381225585938 +[14:42:53.962530] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.4704, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.4919, Kappa: 0.0000, Score: 0.3023 +[14:42:56.913009] Best epoch = 0, Best score = 0.3023 +[14:42:57.189541] log_dir: ./output_logs/retfound +[14:42:59.375331] Epoch: [1] [0/2] eta: 0:00:04 lr: 0.000063 loss: 0.6928 (0.6928) time: 2.1847 data: 2.0049 max mem: 7340 +[14:42:59.640603] Epoch: [1] [1/2] eta: 0:00:01 lr: 0.000094 loss: 0.6928 (0.6928) time: 1.2246 data: 1.0834 max mem: 7340 +[14:42:59.710436] Epoch: [1] Total time: 0:00:02 (1.2604 s / it) +[14:42:59.711443] Averaged stats: lr: 0.000094 loss: 0.6928 (0.6928) +[14:43:01.970246] val: [0/2] eta: 0:00:04 loss: 0.6662 (0.6662) time: 2.2361 data: 2.2174 max mem: 7340 +[14:43:01.979901] val: [1/2] eta: 0:00:01 loss: 0.6662 (0.6941) time: 1.1226 data: 1.1087 max mem: 7340 +[14:43:02.158034] val: Total time: 0:00:02 (1.2122 s / it) +[14:43:02.167248] val loss: 0.6940689086914062 +[14:43:02.167500] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.6398, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6120, Kappa: 0.0000, Score: 0.3588 +[14:43:04.371258] Best epoch = 1, Best score = 0.3588 +[14:43:04.652562] log_dir: ./output_logs/retfound +[14:43:06.856763] Epoch: [2] [0/2] eta: 0:00:04 lr: 0.000125 loss: 0.6771 (0.6771) time: 2.2031 data: 2.1223 max mem: 9665 +[14:43:06.922562] Epoch: [2] [1/2] eta: 0:00:01 lr: 0.000156 loss: 0.6423 (0.6597) time: 1.1341 data: 1.0612 max mem: 9665 +[14:43:06.995927] Epoch: [2] Total time: 0:00:02 (1.1716 s / it) +[14:43:06.996698] Averaged stats: lr: 0.000156 loss: 0.6423 (0.6597) +[14:43:09.271932] val: [0/2] eta: 0:00:04 loss: 0.5929 (0.5929) time: 2.2603 data: 2.2434 max mem: 9665 +[14:43:09.281378] val: [1/2] eta: 0:00:01 loss: 0.5929 (0.7009) time: 1.1346 data: 1.1217 max mem: 9665 +[14:43:09.397495] val: Total time: 0:00:02 (1.1932 s / it) +[14:43:09.411685] val loss: 0.7008514404296875 +[14:43:09.411962] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7124, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6795, Kappa: 0.0000, Score: 0.3830 +[14:43:12.000042] Best epoch = 2, Best score = 0.3830 +[14:43:12.218417] log_dir: ./output_logs/retfound +[14:43:14.320877] Epoch: [3] [0/2] eta: 0:00:04 lr: 0.000188 loss: 0.5962 (0.5962) time: 2.1012 data: 2.0302 max mem: 9665 +[14:43:14.412149] Epoch: [3] [1/2] eta: 0:00:01 lr: 0.000219 loss: 0.5962 (0.6112) time: 1.0957 data: 1.0273 max mem: 9665 +[14:43:14.483522] Epoch: [3] Total time: 0:00:02 (1.1325 s / it) +[14:43:14.484295] Averaged stats: lr: 0.000219 loss: 0.5962 (0.6112) +[14:43:16.799642] val: [0/2] eta: 0:00:04 loss: 0.4905 (0.4905) time: 2.2969 data: 2.2802 max mem: 9665 +[14:43:16.808859] val: [1/2] eta: 0:00:01 loss: 0.4905 (0.7238) time: 1.1528 data: 1.1401 max mem: 9665 +[14:43:16.876703] val: Total time: 0:00:02 (1.1873 s / it) +[14:43:16.886068] val loss: 0.723785400390625 +[14:43:16.886302] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7231, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6835, Kappa: 0.0000, Score: 0.3866 +[14:43:19.133289] Best epoch = 3, Best score = 0.3866 +[14:43:19.408931] log_dir: ./output_logs/retfound +[14:43:21.692390] Epoch: [4] [0/2] eta: 0:00:04 lr: 0.000250 loss: 0.5609 (0.5609) time: 2.2824 data: 2.2127 max mem: 9665 +[14:43:21.774529] Epoch: [4] [1/2] eta: 0:00:01 lr: 0.000281 loss: 0.5609 (0.5671) time: 1.1819 data: 1.1144 max mem: 9665 +[14:43:21.847950] Epoch: [4] Total time: 0:00:02 (1.2194 s / it) +[14:43:21.848777] Averaged stats: lr: 0.000281 loss: 0.5609 (0.5671) +[14:43:24.146182] val: [0/2] eta: 0:00:04 loss: 0.3799 (0.3799) time: 2.2904 data: 2.2736 max mem: 9665 +[14:43:24.155657] val: [1/2] eta: 0:00:01 loss: 0.3799 (0.7757) time: 1.1497 data: 1.1369 max mem: 9665 +[14:43:24.265917] val: Total time: 0:00:02 (1.2054 s / it) +[14:43:24.274963] val loss: 0.7757492065429688 +[14:43:24.275230] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7303, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6946, Kappa: 0.0000, Score: 0.3890 +[14:43:26.237242] Best epoch = 4, Best score = 0.3890 +[14:43:26.299215] log_dir: ./output_logs/retfound +[14:43:28.322290] Epoch: [5] [0/2] eta: 0:00:04 lr: 0.000313 loss: 0.4181 (0.4181) time: 2.0220 data: 1.9430 max mem: 9665 +[14:43:28.581181] Epoch: [5] [1/2] eta: 0:00:01 lr: 0.000344 loss: 0.4181 (0.5064) time: 1.1400 data: 1.0665 max mem: 9665 +[14:43:28.728876] Epoch: [5] Total time: 0:00:02 (1.2148 s / it) +[14:43:28.729864] Averaged stats: lr: 0.000344 loss: 0.4181 (0.5064) +[14:43:30.983291] val: [0/2] eta: 0:00:04 loss: 0.2851 (0.2851) time: 2.2383 data: 2.2213 max mem: 9665 +[14:43:30.992958] val: [1/2] eta: 0:00:01 loss: 0.2851 (0.8606) time: 1.1237 data: 1.1107 max mem: 9665 +[14:43:31.097690] val: Total time: 0:00:02 (1.1767 s / it) +[14:43:31.106681] val loss: 0.8606357574462891 +[14:43:31.106957] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7643, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7235, Kappa: 0.0000, Score: 0.4003 +[14:43:33.057406] Best epoch = 5, Best score = 0.4003 +[14:43:33.131791] log_dir: ./output_logs/retfound +[14:43:35.237812] Epoch: [6] [0/2] eta: 0:00:04 lr: 0.000375 loss: 0.3831 (0.3831) time: 2.1050 data: 2.0362 max mem: 9665 +[14:43:35.303484] Epoch: [6] [1/2] eta: 0:00:01 lr: 0.000406 loss: 0.3831 (0.5324) time: 1.0849 data: 1.0181 max mem: 9665 +[14:43:35.377088] Epoch: [6] Total time: 0:00:02 (1.1225 s / it) +[14:43:35.377863] Averaged stats: lr: 0.000406 loss: 0.3831 (0.5324) +[14:43:37.706003] val: [0/2] eta: 0:00:04 loss: 0.2370 (0.2370) time: 2.3133 data: 2.2968 max mem: 9665 +[14:43:37.715513] val: [1/2] eta: 0:00:01 loss: 0.2370 (0.9276) time: 1.1611 data: 1.1485 max mem: 9665 +[14:43:37.785780] val: Total time: 0:00:02 (1.1969 s / it) +[14:43:37.794705] val loss: 0.9275722503662109 +[14:43:37.794963] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7814, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7395, Kappa: 0.0000, Score: 0.4060 +[14:43:39.894889] Best epoch = 6, Best score = 0.4060 +[14:43:39.961852] log_dir: ./output_logs/retfound +[14:43:42.190282] Epoch: [7] [0/2] eta: 0:00:04 lr: 0.000438 loss: 0.5845 (0.5845) time: 2.2274 data: 2.1583 max mem: 9665 +[14:43:42.255827] Epoch: [7] [1/2] eta: 0:00:01 lr: 0.000469 loss: 0.4436 (0.5140) time: 1.1461 data: 1.0792 max mem: 9665 +[14:43:42.331921] Epoch: [7] Total time: 0:00:02 (1.1849 s / it) +[14:43:42.332698] Averaged stats: lr: 0.000469 loss: 0.4436 (0.5140) +[14:43:44.501749] val: [0/2] eta: 0:00:04 loss: 0.2381 (0.2381) time: 2.1507 data: 2.1175 max mem: 9665 +[14:43:44.517340] val: [1/2] eta: 0:00:01 loss: 0.2381 (0.9077) time: 1.0828 data: 1.0588 max mem: 9665 +[14:43:44.589346] val: Total time: 0:00:02 (1.1195 s / it) +[14:43:44.598281] val loss: 0.9076881408691406 +[14:43:44.598537] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8002, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7641, Kappa: 0.0000, Score: 0.4123 +[14:43:46.519236] Best epoch = 7, Best score = 0.4123 +[14:43:46.582921] log_dir: ./output_logs/retfound +[14:43:48.635501] Epoch: [8] [0/2] eta: 0:00:04 lr: 0.000500 loss: 0.6891 (0.6891) time: 2.0514 data: 1.9829 max mem: 9665 +[14:43:48.779560] Epoch: [8] [1/2] eta: 0:00:01 lr: 0.000531 loss: 0.4929 (0.5910) time: 1.0973 data: 1.0262 max mem: 9665 +[14:43:48.852948] Epoch: [8] Total time: 0:00:02 (1.1349 s / it) +[14:43:48.853735] Averaged stats: lr: 0.000531 loss: 0.4929 (0.5910) +[14:43:51.146510] val: [0/2] eta: 0:00:04 loss: 0.2592 (0.2592) time: 2.2816 data: 2.2643 max mem: 9665 +[14:43:51.156634] val: [1/2] eta: 0:00:01 loss: 0.2592 (0.8484) time: 1.1455 data: 1.1322 max mem: 9665 +[14:43:51.228363] val: Total time: 0:00:02 (1.1822 s / it) +[14:43:51.237200] val loss: 0.8484306335449219 +[14:43:51.237463] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8378, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7791, Kappa: 0.0000, Score: 0.4248 +[14:43:53.104576] Best epoch = 8, Best score = 0.4248 +[14:43:53.176363] log_dir: ./output_logs/retfound +[14:43:55.334999] Epoch: [9] [0/2] eta: 0:00:04 lr: 0.000562 loss: 0.6526 (0.6526) time: 2.1574 data: 2.0804 max mem: 9665 +[14:43:55.467910] Epoch: [9] [1/2] eta: 0:00:01 lr: 0.000594 loss: 0.4903 (0.5714) time: 1.1448 data: 1.0737 max mem: 9665 +[14:43:55.535389] Epoch: [9] Total time: 0:00:02 (1.1794 s / it) +[14:43:55.536272] Averaged stats: lr: 0.000594 loss: 0.4903 (0.5714) +[14:43:57.889592] val: [0/2] eta: 0:00:04 loss: 0.3053 (0.3053) time: 2.3426 data: 2.3256 max mem: 9665 +[14:43:57.899167] val: [1/2] eta: 0:00:01 loss: 0.3053 (0.7632) time: 1.1758 data: 1.1629 max mem: 9665 +[14:43:57.970522] val: Total time: 0:00:02 (1.2121 s / it) +[14:43:57.979446] val loss: 0.7631969451904297 +[14:43:57.979682] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8602, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8021, Kappa: 0.0000, Score: 0.4323 +[14:43:59.986105] Best epoch = 9, Best score = 0.4323 +[14:44:00.129142] log_dir: ./output_logs/retfound +[14:44:02.309912] Epoch: [10] [0/2] eta: 0:00:04 lr: 0.000625 loss: 0.5125 (0.5125) time: 2.1798 data: 2.1100 max mem: 9665 +[14:44:02.525709] Epoch: [10] [1/2] eta: 0:00:01 lr: 0.000625 loss: 0.4953 (0.5039) time: 1.1974 data: 1.1296 max mem: 9665 +[14:44:02.596335] Epoch: [10] Total time: 0:00:02 (1.2335 s / it) +[14:44:02.597177] Averaged stats: lr: 0.000625 loss: 0.4953 (0.5039) +[14:44:04.874249] val: [0/2] eta: 0:00:04 loss: 0.3534 (0.3534) time: 2.2629 data: 2.2459 max mem: 9665 +[14:44:04.884009] val: [1/2] eta: 0:00:01 loss: 0.3534 (0.6950) time: 1.1360 data: 1.1230 max mem: 9665 +[14:44:04.952903] val: Total time: 0:00:02 (1.1711 s / it) +[14:44:04.961756] val loss: 0.6950168609619141 +[14:44:04.962024] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8602, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7924, Kappa: 0.0000, Score: 0.4323 +[14:44:04.996809] Best epoch = 9, Best score = 0.4323 +[14:44:05.259922] log_dir: ./output_logs/retfound +[14:44:07.391040] Epoch: [11] [0/2] eta: 0:00:04 lr: 0.000624 loss: 0.5990 (0.5990) time: 2.1302 data: 2.0614 max mem: 9665 +[14:44:07.457135] Epoch: [11] [1/2] eta: 0:00:01 lr: 0.000623 loss: 0.4867 (0.5429) time: 1.0977 data: 1.0307 max mem: 9665 +[14:44:07.537974] Epoch: [11] Total time: 0:00:02 (1.1389 s / it) +[14:44:07.538748] Averaged stats: lr: 0.000623 loss: 0.4867 (0.5429) +[14:44:09.861681] val: [0/2] eta: 0:00:04 loss: 0.3770 (0.3770) time: 2.3119 data: 2.2936 max mem: 9665 +[14:44:09.871648] val: [1/2] eta: 0:00:01 loss: 0.3770 (0.6615) time: 1.1606 data: 1.1469 max mem: 9665 +[14:44:09.942403] val: Total time: 0:00:02 (1.1966 s / it) +[14:44:09.951463] val loss: 0.6615047454833984 +[14:44:09.951675] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8584, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7837, Kappa: 0.0000, Score: 0.4317 +[14:44:09.986423] Best epoch = 9, Best score = 0.4323 +[14:44:10.285336] log_dir: ./output_logs/retfound +[14:44:12.483570] Epoch: [12] [0/2] eta: 0:00:04 lr: 0.000621 loss: 0.5254 (0.5254) time: 2.1973 data: 2.1289 max mem: 9665 +[14:44:12.556911] Epoch: [12] [1/2] eta: 0:00:01 lr: 0.000619 loss: 0.4742 (0.4998) time: 1.1350 data: 1.0645 max mem: 9665 +[14:44:12.630246] Epoch: [12] Total time: 0:00:02 (1.1724 s / it) +[14:44:12.631066] Averaged stats: lr: 0.000619 loss: 0.4742 (0.4998) +[14:44:14.827690] val: [0/2] eta: 0:00:04 loss: 0.3317 (0.3317) time: 2.1851 data: 2.1683 max mem: 9665 +[14:44:14.837250] val: [1/2] eta: 0:00:01 loss: 0.3317 (0.6754) time: 1.0971 data: 1.0842 max mem: 9665 +[14:44:14.912635] val: Total time: 0:00:02 (1.1353 s / it) +[14:44:14.921501] val loss: 0.6754188537597656 +[14:44:14.921679] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8566, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7837, Kappa: 0.0000, Score: 0.4311 +[14:44:14.958050] Best epoch = 9, Best score = 0.4323 +[14:44:15.775262] log_dir: ./output_logs/retfound +[14:44:17.949444] Epoch: [13] [0/2] eta: 0:00:04 lr: 0.000616 loss: 0.4973 (0.4973) time: 2.1732 data: 2.1025 max mem: 9665 +[14:44:18.158985] Epoch: [13] [1/2] eta: 0:00:01 lr: 0.000613 loss: 0.4848 (0.4910) time: 1.1910 data: 1.1232 max mem: 9665 +[14:44:18.227543] Epoch: [13] Total time: 0:00:02 (1.2261 s / it) +[14:44:18.228412] Averaged stats: lr: 0.000613 loss: 0.4848 (0.4910) +[14:44:20.614298] val: [0/2] eta: 0:00:04 loss: 0.2659 (0.2659) time: 2.3706 data: 2.3535 max mem: 9665 +[14:44:20.624611] val: [1/2] eta: 0:00:01 loss: 0.2659 (0.7241) time: 1.1902 data: 1.1768 max mem: 9665 +[14:44:20.701187] val: Total time: 0:00:02 (1.2294 s / it) +[14:44:20.710125] val loss: 0.724090576171875 +[14:44:20.710315] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8530, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7803, Kappa: 0.0000, Score: 0.4299 +[14:44:20.744046] Best epoch = 9, Best score = 0.4323 +[14:44:21.026213] log_dir: ./output_logs/retfound +[14:44:23.326462] Epoch: [14] [0/2] eta: 0:00:04 lr: 0.000610 loss: 0.4897 (0.4897) time: 2.2990 data: 2.2233 max mem: 9665 +[14:44:23.394086] Epoch: [14] [1/2] eta: 0:00:01 lr: 0.000606 loss: 0.4808 (0.4852) time: 1.1829 data: 1.1117 max mem: 9665 +[14:44:23.465610] Epoch: [14] Total time: 0:00:02 (1.2196 s / it) +[14:44:23.466613] Averaged stats: lr: 0.000606 loss: 0.4808 (0.4852) +[14:44:25.753249] val: [0/2] eta: 0:00:04 loss: 0.2161 (0.2161) time: 2.2707 data: 2.2540 max mem: 9665 +[14:44:25.762830] val: [1/2] eta: 0:00:01 loss: 0.2161 (0.7764) time: 1.1398 data: 1.1270 max mem: 9665 +[14:44:25.835651] val: Total time: 0:00:02 (1.1768 s / it) +[14:44:25.844525] val loss: 0.7763900756835938 +[14:44:25.844763] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8530, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7803, Kappa: 0.0000, Score: 0.4299 +[14:44:25.879509] Best epoch = 9, Best score = 0.4323 +[14:44:26.137485] log_dir: ./output_logs/retfound +[14:44:28.472557] Epoch: [15] [0/2] eta: 0:00:04 lr: 0.000601 loss: 0.4809 (0.4809) time: 2.3340 data: 2.2646 max mem: 9665 +[14:44:28.538079] Epoch: [15] [1/2] eta: 0:00:01 lr: 0.000596 loss: 0.4809 (0.4922) time: 1.1994 data: 1.1323 max mem: 9665 +[14:44:28.612012] Epoch: [15] Total time: 0:00:02 (1.2372 s / it) +[14:44:28.612833] Averaged stats: lr: 0.000596 loss: 0.4809 (0.4922) +[14:44:30.964539] val: [0/2] eta: 0:00:04 loss: 0.1947 (0.1947) time: 2.3443 data: 2.3275 max mem: 9665 +[14:44:30.974351] val: [1/2] eta: 0:00:01 loss: 0.1947 (0.7759) time: 1.1768 data: 1.1638 max mem: 9665 +[14:44:31.045614] val: Total time: 0:00:02 (1.2130 s / it) +[14:44:31.054565] val loss: 0.7758550643920898 +[14:44:31.054830] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8530, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7803, Kappa: 0.0000, Score: 0.4299 +[14:44:31.100432] Best epoch = 9, Best score = 0.4323 +[14:44:31.392623] log_dir: ./output_logs/retfound +[14:44:33.609241] Epoch: [16] [0/2] eta: 0:00:04 lr: 0.000591 loss: 0.4155 (0.4155) time: 2.2155 data: 2.1467 max mem: 9665 +[14:44:33.760689] Epoch: [16] [1/2] eta: 0:00:01 lr: 0.000585 loss: 0.4155 (0.4547) time: 1.1831 data: 1.1162 max mem: 9665 +[14:44:33.834603] Epoch: [16] Total time: 0:00:02 (1.2209 s / it) +[14:44:33.835413] Averaged stats: lr: 0.000585 loss: 0.4155 (0.4547) +[14:44:35.998202] val: [0/2] eta: 0:00:04 loss: 0.1981 (0.1981) time: 2.1522 data: 2.1353 max mem: 9665 +[14:44:36.007787] val: [1/2] eta: 0:00:01 loss: 0.1981 (0.7098) time: 1.0806 data: 1.0677 max mem: 9665 +[14:44:36.076117] val: Total time: 0:00:02 (1.1154 s / it) +[14:44:36.084833] val loss: 0.7098367214202881 +[14:44:36.085097] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8530, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7766, Kappa: 0.0000, Score: 0.4299 +[14:44:36.131091] Best epoch = 9, Best score = 0.4323 +[14:44:36.403272] log_dir: ./output_logs/retfound +[14:44:38.672114] Epoch: [17] [0/2] eta: 0:00:04 lr: 0.000579 loss: 0.5985 (0.5985) time: 2.2678 data: 2.1980 max mem: 9665 +[14:44:38.737953] Epoch: [17] [1/2] eta: 0:00:01 lr: 0.000572 loss: 0.3304 (0.4645) time: 1.1664 data: 1.0990 max mem: 9665 +[14:44:38.806484] Epoch: [17] Total time: 0:00:02 (1.2015 s / it) +[14:44:38.807243] Averaged stats: lr: 0.000572 loss: 0.3304 (0.4645) +[14:44:41.221754] val: [0/2] eta: 0:00:04 loss: 0.2700 (0.2700) time: 2.4035 data: 2.3865 max mem: 9665 +[14:44:41.231332] val: [1/2] eta: 0:00:01 loss: 0.2700 (0.5545) time: 1.2063 data: 1.1933 max mem: 9665 +[14:44:41.303972] val: Total time: 0:00:02 (1.2432 s / it) +[14:44:41.312869] val loss: 0.5545053482055664 +[14:44:41.313101] Accuracy: 0.7750, F1 Score: 0.5867, ROC AUC: 0.8450, Hamming Loss: 0.2250, + Jaccard Score: 0.4725, Precision: 0.6528, Recall: 0.5789, + Average Precision: 0.7673, Kappa: 0.1964, Score: 0.5427 +[14:44:43.402170] Best epoch = 17, Best score = 0.5427 +[14:44:43.632019] log_dir: ./output_logs/retfound +[14:44:45.762660] Epoch: [18] [0/2] eta: 0:00:04 lr: 0.000565 loss: 0.4645 (0.4645) time: 2.1295 data: 2.0601 max mem: 9665 +[14:44:45.828628] Epoch: [18] [1/2] eta: 0:00:01 lr: 0.000558 loss: 0.4478 (0.4562) time: 1.0974 data: 1.0301 max mem: 9665 +[14:44:45.989343] Epoch: [18] Total time: 0:00:02 (1.1785 s / it) +[14:44:45.990176] Averaged stats: lr: 0.000558 loss: 0.4478 (0.4562) +[14:44:48.195439] val: [0/2] eta: 0:00:04 loss: 0.4114 (0.4114) time: 2.1835 data: 2.1666 max mem: 9665 +[14:44:48.205115] val: [1/2] eta: 0:00:01 loss: 0.4114 (0.4426) time: 1.0963 data: 1.0834 max mem: 9665 +[14:44:48.284217] val: Total time: 0:00:02 (1.1364 s / it) +[14:44:48.293393] val loss: 0.44260621070861816 +[14:44:48.293595] Accuracy: 0.7250, F1 Score: 0.6925, ROC AUC: 0.8423, Hamming Loss: 0.2750, + Jaccard Score: 0.5387, Precision: 0.6995, Recall: 0.7832, + Average Precision: 0.7622, Kappa: 0.4180, Score: 0.6509 +[14:44:50.167659] Best epoch = 18, Best score = 0.6509 +[14:44:50.239804] log_dir: ./output_logs/retfound +[14:44:52.598703] Epoch: [19] [0/2] eta: 0:00:04 lr: 0.000550 loss: 0.5536 (0.5536) time: 2.3579 data: 2.2816 max mem: 9665 +[14:44:52.709660] Epoch: [19] [1/2] eta: 0:00:01 lr: 0.000542 loss: 0.4142 (0.4839) time: 1.2341 data: 1.1638 max mem: 9665 +[14:44:52.793514] Epoch: [19] Total time: 0:00:02 (1.2768 s / it) +[14:44:52.794337] Averaged stats: lr: 0.000542 loss: 0.4142 (0.4839) +[14:44:55.223220] val: [0/2] eta: 0:00:04 loss: 0.4730 (0.4730) time: 2.4197 data: 2.4026 max mem: 9665 +[14:44:55.233778] val: [1/2] eta: 0:00:01 loss: 0.3588 (0.4159) time: 1.2148 data: 1.2013 max mem: 9665 +[14:44:55.312378] val: Total time: 0:00:02 (1.2548 s / it) +[14:44:55.321405] val loss: 0.41592884063720703 +[14:44:55.321676] Accuracy: 0.7250, F1 Score: 0.6925, ROC AUC: 0.8387, Hamming Loss: 0.2750, + Jaccard Score: 0.5387, Precision: 0.6995, Recall: 0.7832, + Average Precision: 0.7563, Kappa: 0.4180, Score: 0.6497 +[14:44:55.370935] Best epoch = 18, Best score = 0.6509 +[14:44:55.638639] log_dir: ./output_logs/retfound +[14:44:57.841777] Epoch: [20] [0/2] eta: 0:00:04 lr: 0.000534 loss: 0.4273 (0.4273) time: 2.2022 data: 2.1332 max mem: 9665 +[14:44:57.919162] Epoch: [20] [1/2] eta: 0:00:01 lr: 0.000525 loss: 0.4273 (0.4539) time: 1.1394 data: 1.0725 max mem: 9665 +[14:44:57.990541] Epoch: [20] Total time: 0:00:02 (1.1759 s / it) +[14:44:57.991359] Averaged stats: lr: 0.000525 loss: 0.4273 (0.4539) +[14:45:00.257839] val: [0/2] eta: 0:00:04 loss: 0.4074 (0.4074) time: 2.2517 data: 2.2339 max mem: 9665 +[14:45:00.269565] val: [1/2] eta: 0:00:01 loss: 0.4074 (0.4190) time: 1.1314 data: 1.1170 max mem: 9665 +[14:45:00.351249] val: Total time: 0:00:02 (1.1729 s / it) +[14:45:00.360269] val loss: 0.41899538040161133 +[14:45:00.360558] Accuracy: 0.7500, F1 Score: 0.7024, ROC AUC: 0.8423, Hamming Loss: 0.2500, + Jaccard Score: 0.5544, Precision: 0.6933, Recall: 0.7599, + Average Precision: 0.7585, Kappa: 0.4203, Score: 0.6550 +[14:45:02.245703] Best epoch = 20, Best score = 0.6550 +[14:45:02.321864] log_dir: ./output_logs/retfound +[14:45:04.520308] Epoch: [21] [0/2] eta: 0:00:04 lr: 0.000516 loss: 0.3957 (0.3957) time: 2.1974 data: 2.1281 max mem: 9665 +[14:45:04.585863] Epoch: [21] [1/2] eta: 0:00:01 lr: 0.000506 loss: 0.3744 (0.3850) time: 1.1311 data: 1.0641 max mem: 9665 +[14:45:04.661348] Epoch: [21] Total time: 0:00:02 (1.1696 s / it) +[14:45:04.662266] Averaged stats: lr: 0.000506 loss: 0.3744 (0.3850) +[14:45:06.862839] val: [0/2] eta: 0:00:04 loss: 0.3647 (0.3647) time: 2.1853 data: 2.1616 max mem: 9665 +[14:45:06.876307] val: [1/2] eta: 0:00:01 loss: 0.3647 (0.4385) time: 1.0991 data: 1.0809 max mem: 9665 +[14:45:06.965819] val: Total time: 0:00:02 (1.1445 s / it) +[14:45:06.974980] val loss: 0.43852484226226807 +[14:45:06.975168] Accuracy: 0.7500, F1 Score: 0.6865, ROC AUC: 0.8495, Hamming Loss: 0.2500, + Jaccard Score: 0.5404, Precision: 0.6752, Recall: 0.7204, + Average Precision: 0.7674, Kappa: 0.3808, Score: 0.6389 +[14:45:07.007869] Best epoch = 20, Best score = 0.6550 +[14:45:07.263426] log_dir: ./output_logs/retfound +[14:45:09.415698] Epoch: [22] [0/2] eta: 0:00:04 lr: 0.000496 loss: 0.4084 (0.4084) time: 2.1512 data: 2.0790 max mem: 9665 +[14:45:09.481414] Epoch: [22] [1/2] eta: 0:00:01 lr: 0.000486 loss: 0.3788 (0.3936) time: 1.1080 data: 1.0395 max mem: 9665 +[14:45:09.564220] Epoch: [22] Total time: 0:00:02 (1.1503 s / it) +[14:45:09.565001] Averaged stats: lr: 0.000486 loss: 0.3788 (0.3936) +[14:45:11.799712] val: [0/2] eta: 0:00:04 loss: 0.3518 (0.3518) time: 2.2230 data: 2.2041 max mem: 9665 +[14:45:11.812885] val: [1/2] eta: 0:00:01 loss: 0.3518 (0.4600) time: 1.1177 data: 1.1021 max mem: 9665 +[14:45:11.882015] val: Total time: 0:00:02 (1.1531 s / it) +[14:45:11.894113] val loss: 0.4600367546081543 +[14:45:11.894425] Accuracy: 0.7500, F1 Score: 0.6865, ROC AUC: 0.8423, Hamming Loss: 0.2500, + Jaccard Score: 0.5404, Precision: 0.6752, Recall: 0.7204, + Average Precision: 0.7621, Kappa: 0.3808, Score: 0.6365 +[14:45:11.941431] Best epoch = 20, Best score = 0.6550 +[14:45:12.207817] log_dir: ./output_logs/retfound +[14:45:14.369666] Epoch: [23] [0/2] eta: 0:00:04 lr: 0.000476 loss: 0.4155 (0.4155) time: 2.1609 data: 2.0911 max mem: 9665 +[14:45:14.435554] Epoch: [23] [1/2] eta: 0:00:01 lr: 0.000465 loss: 0.4155 (0.4200) time: 1.1131 data: 1.0456 max mem: 9665 +[14:45:14.505974] Epoch: [23] Total time: 0:00:02 (1.1490 s / it) +[14:45:14.507136] Averaged stats: lr: 0.000465 loss: 0.4155 (0.4200) +[14:45:16.729340] val: [0/2] eta: 0:00:04 loss: 0.3005 (0.3005) time: 2.2095 data: 2.1929 max mem: 9665 +[14:45:16.739113] val: [1/2] eta: 0:00:01 loss: 0.3005 (0.5098) time: 1.1094 data: 1.0965 max mem: 9665 +[14:45:16.807623] val: Total time: 0:00:02 (1.1442 s / it) +[14:45:16.816397] val loss: 0.5097696781158447 +[14:45:16.816624] Accuracy: 0.7500, F1 Score: 0.6667, ROC AUC: 0.8495, Hamming Loss: 0.2500, + Jaccard Score: 0.5238, Precision: 0.6583, Recall: 0.6810, + Average Precision: 0.7703, Kappa: 0.3355, Score: 0.6172 +[14:45:16.851810] Best epoch = 20, Best score = 0.6550 +[14:45:17.102441] log_dir: ./output_logs/retfound +[14:45:19.152223] Epoch: [24] [0/2] eta: 0:00:04 lr: 0.000455 loss: 0.3556 (0.3556) time: 2.0489 data: 1.9804 max mem: 9665 +[14:45:19.217413] Epoch: [24] [1/2] eta: 0:00:01 lr: 0.000444 loss: 0.3556 (0.3667) time: 1.0567 data: 0.9902 max mem: 9665 +[14:45:19.285052] Epoch: [24] Total time: 0:00:02 (1.0912 s / it) +[14:45:19.285830] Averaged stats: lr: 0.000444 loss: 0.3556 (0.3667) +[14:45:21.414515] val: [0/2] eta: 0:00:04 loss: 0.2516 (0.2516) time: 2.1137 data: 2.0950 max mem: 9665 +[14:45:21.429459] val: [1/2] eta: 0:00:01 loss: 0.2516 (0.5912) time: 1.0639 data: 1.0476 max mem: 9665 +[14:45:21.502182] val: Total time: 0:00:02 (1.1011 s / it) +[14:45:21.516701] val loss: 0.591159462928772 +[14:45:21.517003] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8566, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.7886, Kappa: 0.4776, Score: 0.6909 +[14:45:23.385204] Best epoch = 24, Best score = 0.6909 +[14:45:23.460276] log_dir: ./output_logs/retfound +[14:45:25.602702] Epoch: [25] [0/2] eta: 0:00:04 lr: 0.000432 loss: 0.3439 (0.3439) time: 2.1414 data: 2.0723 max mem: 9665 +[14:45:25.700581] Epoch: [25] [1/2] eta: 0:00:01 lr: 0.000421 loss: 0.3439 (0.4083) time: 1.1193 data: 1.0523 max mem: 9665 +[14:45:25.773514] Epoch: [25] Total time: 0:00:02 (1.1565 s / it) +[14:45:25.774297] Averaged stats: lr: 0.000421 loss: 0.3439 (0.4083) +[14:45:27.892853] val: [0/2] eta: 0:00:04 loss: 0.2133 (0.2133) time: 2.0997 data: 2.0828 max mem: 9665 +[14:45:27.902865] val: [1/2] eta: 0:00:01 loss: 0.2133 (0.6798) time: 1.0546 data: 1.0415 max mem: 9665 +[14:45:27.971348] val: Total time: 0:00:02 (1.0894 s / it) +[14:45:27.980334] val loss: 0.6798206567764282 +[14:45:27.980565] Accuracy: 0.8000, F1 Score: 0.6078, ROC AUC: 0.8530, Hamming Loss: 0.2000, + Jaccard Score: 0.4947, Precision: 0.7387, Recall: 0.5950, + Average Precision: 0.7843, Kappa: 0.2488, Score: 0.5699 +[14:45:28.038976] Best epoch = 24, Best score = 0.6909 +[14:45:28.296422] log_dir: ./output_logs/retfound +[14:45:30.478653] Epoch: [26] [0/2] eta: 0:00:04 lr: 0.000409 loss: 0.3464 (0.3464) time: 2.1810 data: 2.1037 max mem: 9665 +[14:45:30.560928] Epoch: [26] [1/2] eta: 0:00:01 lr: 0.000398 loss: 0.3464 (0.4094) time: 1.1313 data: 1.0519 max mem: 9665 +[14:45:30.626336] Epoch: [26] Total time: 0:00:02 (1.1649 s / it) +[14:45:30.627075] Averaged stats: lr: 0.000398 loss: 0.3464 (0.4094) +[14:45:32.868304] val: [0/2] eta: 0:00:04 loss: 0.2153 (0.2153) time: 2.2303 data: 2.2131 max mem: 9665 +[14:45:32.878186] val: [1/2] eta: 0:00:01 loss: 0.2153 (0.6663) time: 1.1198 data: 1.1066 max mem: 9665 +[14:45:32.953342] val: Total time: 0:00:02 (1.1580 s / it) +[14:45:32.962886] val loss: 0.6662570238113403 +[14:45:32.963177] Accuracy: 0.8000, F1 Score: 0.6078, ROC AUC: 0.8530, Hamming Loss: 0.2000, + Jaccard Score: 0.4947, Precision: 0.7387, Recall: 0.5950, + Average Precision: 0.7935, Kappa: 0.2488, Score: 0.5699 +[14:45:33.000097] Best epoch = 24, Best score = 0.6909 +[14:45:33.259911] log_dir: ./output_logs/retfound +[14:45:35.319357] Epoch: [27] [0/2] eta: 0:00:04 lr: 0.000386 loss: 0.4424 (0.4424) time: 2.0584 data: 1.9897 max mem: 9665 +[14:45:35.484653] Epoch: [27] [1/2] eta: 0:00:01 lr: 0.000374 loss: 0.2835 (0.3630) time: 1.1115 data: 1.0355 max mem: 9665 +[14:45:35.555307] Epoch: [27] Total time: 0:00:02 (1.1476 s / it) +[14:45:35.556121] Averaged stats: lr: 0.000374 loss: 0.2835 (0.3630) +[14:45:37.756472] val: [0/2] eta: 0:00:04 loss: 0.2301 (0.2301) time: 2.1858 data: 2.1688 max mem: 9665 +[14:45:37.766150] val: [1/2] eta: 0:00:01 loss: 0.2301 (0.6317) time: 1.0974 data: 1.0845 max mem: 9665 +[14:45:37.839867] val: Total time: 0:00:02 (1.1349 s / it) +[14:45:37.849819] val loss: 0.631696343421936 +[14:45:37.850065] Accuracy: 0.8250, F1 Score: 0.7128, ROC AUC: 0.8459, Hamming Loss: 0.1750, + Jaccard Score: 0.5846, Precision: 0.7598, Recall: 0.6900, + Average Precision: 0.7791, Kappa: 0.4309, Score: 0.6632 +[14:45:37.888385] Best epoch = 24, Best score = 0.6909 +[14:45:38.172261] log_dir: ./output_logs/retfound +[14:45:40.249488] Epoch: [28] [0/2] eta: 0:00:04 lr: 0.000362 loss: 0.5225 (0.5225) time: 2.0763 data: 2.0077 max mem: 9665 +[14:45:40.363053] Epoch: [28] [1/2] eta: 0:00:01 lr: 0.000350 loss: 0.3059 (0.4142) time: 1.0946 data: 1.0277 max mem: 9665 +[14:45:40.433762] Epoch: [28] Total time: 0:00:02 (1.1307 s / it) +[14:45:40.434546] Averaged stats: lr: 0.000350 loss: 0.3059 (0.4142) +[14:45:42.619988] val: [0/2] eta: 0:00:04 loss: 0.2534 (0.2534) time: 2.1706 data: 2.1536 max mem: 9665 +[14:45:42.629721] val: [1/2] eta: 0:00:01 loss: 0.2534 (0.5868) time: 1.0899 data: 1.0769 max mem: 9665 +[14:45:42.701315] val: Total time: 0:00:02 (1.1263 s / it) +[14:45:42.710189] val loss: 0.5867966413497925 +[14:45:42.710386] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8495, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.7854, Kappa: 0.4776, Score: 0.6885 +[14:45:42.759232] Best epoch = 24, Best score = 0.6909 +[14:45:43.036190] log_dir: ./output_logs/retfound +[14:45:45.311893] Epoch: [29] [0/2] eta: 0:00:04 lr: 0.000337 loss: 0.2719 (0.2719) time: 2.2747 data: 2.2055 max mem: 9665 +[14:45:45.383318] Epoch: [29] [1/2] eta: 0:00:01 lr: 0.000325 loss: 0.2719 (0.3667) time: 1.1727 data: 1.1028 max mem: 9665 +[14:45:45.455093] Epoch: [29] Total time: 0:00:02 (1.2094 s / it) +[14:45:45.455885] Averaged stats: lr: 0.000325 loss: 0.2719 (0.3667) +[14:45:47.565031] val: [0/2] eta: 0:00:04 loss: 0.2825 (0.2825) time: 2.0985 data: 2.0816 max mem: 9665 +[14:45:47.574631] val: [1/2] eta: 0:00:01 loss: 0.2825 (0.5494) time: 1.0538 data: 1.0408 max mem: 9665 +[14:45:47.659199] val: Total time: 0:00:02 (1.0967 s / it) +[14:45:47.668064] val loss: 0.5493767261505127 +[14:45:47.668297] Accuracy: 0.8000, F1 Score: 0.7133, ROC AUC: 0.8423, Hamming Loss: 0.2000, + Jaccard Score: 0.5780, Precision: 0.7133, Recall: 0.7133, + Average Precision: 0.7852, Kappa: 0.4265, Score: 0.6607 +[14:45:47.708804] Best epoch = 24, Best score = 0.6909 +[14:45:47.993969] log_dir: ./output_logs/retfound +[14:45:50.304640] Epoch: [30] [0/2] eta: 0:00:04 lr: 0.000313 loss: 0.3539 (0.3539) time: 2.3097 data: 2.2405 max mem: 9665 +[14:45:50.370149] Epoch: [30] [1/2] eta: 0:00:01 lr: 0.000301 loss: 0.3539 (0.4821) time: 1.1872 data: 1.1203 max mem: 9665 +[14:45:50.436918] Epoch: [30] Total time: 0:00:02 (1.2214 s / it) +[14:45:50.437712] Averaged stats: lr: 0.000301 loss: 0.3539 (0.4821) +[14:45:52.628902] val: [0/2] eta: 0:00:04 loss: 0.2814 (0.2814) time: 2.1692 data: 2.1518 max mem: 9665 +[14:45:52.638977] val: [1/2] eta: 0:00:01 loss: 0.2814 (0.5393) time: 1.0893 data: 1.0760 max mem: 9665 +[14:45:52.712405] val: Total time: 0:00:02 (1.1267 s / it) +[14:45:52.721742] val loss: 0.5392974615097046 +[14:45:52.721990] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8423, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.7815, Kappa: 0.4776, Score: 0.6862 +[14:45:52.756442] Best epoch = 24, Best score = 0.6909 +[14:45:53.032771] log_dir: ./output_logs/retfound +[14:45:55.320678] Epoch: [31] [0/2] eta: 0:00:04 lr: 0.000289 loss: 0.5192 (0.5192) time: 2.2870 data: 2.2181 max mem: 9665 +[14:45:55.386012] Epoch: [31] [1/2] eta: 0:00:01 lr: 0.000276 loss: 0.2931 (0.4061) time: 1.1758 data: 1.1091 max mem: 9665 +[14:45:55.460611] Epoch: [31] Total time: 0:00:02 (1.2138 s / it) +[14:45:55.461322] Averaged stats: lr: 0.000276 loss: 0.2931 (0.4061) +[14:45:57.760997] val: [0/2] eta: 0:00:04 loss: 0.2767 (0.2767) time: 2.2851 data: 2.2682 max mem: 9665 +[14:45:57.770630] val: [1/2] eta: 0:00:01 loss: 0.2767 (0.5296) time: 1.1471 data: 1.1342 max mem: 9665 +[14:45:57.841744] val: Total time: 0:00:02 (1.1833 s / it) +[14:45:57.850563] val loss: 0.5296193361282349 +[14:45:57.850741] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8495, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8023, Kappa: 0.4776, Score: 0.6885 +[14:45:57.883835] Best epoch = 24, Best score = 0.6909 +[14:45:58.161948] log_dir: ./output_logs/retfound +[14:46:00.435602] Epoch: [32] [0/2] eta: 0:00:04 lr: 0.000264 loss: 0.4308 (0.4308) time: 2.2728 data: 2.2031 max mem: 9665 +[14:46:00.501215] Epoch: [32] [1/2] eta: 0:00:01 lr: 0.000252 loss: 0.3111 (0.3710) time: 1.1688 data: 1.1016 max mem: 9665 +[14:46:00.574660] Epoch: [32] Total time: 0:00:02 (1.2063 s / it) +[14:46:00.575436] Averaged stats: lr: 0.000252 loss: 0.3111 (0.3710) +[14:46:02.761250] val: [0/2] eta: 0:00:04 loss: 0.2620 (0.2620) time: 2.1666 data: 2.1442 max mem: 9665 +[14:46:02.781726] val: [1/2] eta: 0:00:01 loss: 0.2620 (0.5392) time: 1.0880 data: 1.0722 max mem: 9665 +[14:46:03.015260] val: Total time: 0:00:02 (1.2106 s / it) +[14:46:03.034110] val loss: 0.539230465888977 +[14:46:03.034308] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8566, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8078, Kappa: 0.4776, Score: 0.6909 +[14:46:03.290343] Best epoch = 24, Best score = 0.6909 +[14:46:06.153075] log_dir: ./output_logs/retfound +[14:46:08.441144] Epoch: [33] [0/2] eta: 0:00:04 lr: 0.000240 loss: 0.2173 (0.2173) time: 2.2871 data: 2.2184 max mem: 9665 +[14:46:08.528762] Epoch: [33] [1/2] eta: 0:00:01 lr: 0.000228 loss: 0.2173 (0.3337) time: 1.1870 data: 1.1203 max mem: 9665 +[14:46:08.631369] Epoch: [33] Total time: 0:00:02 (1.2390 s / it) +[14:46:08.632287] Averaged stats: lr: 0.000228 loss: 0.2173 (0.3337) +[14:46:10.847317] val: [0/2] eta: 0:00:04 loss: 0.2521 (0.2521) time: 2.1925 data: 2.1756 max mem: 9665 +[14:46:10.856475] val: [1/2] eta: 0:00:01 loss: 0.2521 (0.5496) time: 1.1006 data: 1.0878 max mem: 9665 +[14:46:10.926475] val: Total time: 0:00:02 (1.1361 s / it) +[14:46:10.935275] val loss: 0.5495936870574951 +[14:46:10.935443] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8647, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8186, Kappa: 0.4776, Score: 0.6936 +[14:46:12.835865] Best epoch = 33, Best score = 0.6936 +[14:46:12.898617] log_dir: ./output_logs/retfound +[14:46:15.167313] Epoch: [34] [0/2] eta: 0:00:04 lr: 0.000217 loss: 0.2996 (0.2996) time: 2.2676 data: 2.1926 max mem: 9665 +[14:46:15.245268] Epoch: [34] [1/2] eta: 0:00:01 lr: 0.000205 loss: 0.2996 (0.3995) time: 1.1724 data: 1.0963 max mem: 9665 +[14:46:15.312957] Epoch: [34] Total time: 0:00:02 (1.2070 s / it) +[14:46:15.313769] Averaged stats: lr: 0.000205 loss: 0.2996 (0.3995) +[14:46:17.481507] val: [0/2] eta: 0:00:04 loss: 0.2546 (0.2546) time: 2.1569 data: 2.1402 max mem: 9665 +[14:46:17.491084] val: [1/2] eta: 0:00:01 loss: 0.2546 (0.5407) time: 1.0829 data: 1.0701 max mem: 9665 +[14:46:17.570271] val: Total time: 0:00:02 (1.1232 s / it) +[14:46:17.580770] val loss: 0.5406581163406372 +[14:46:17.581036] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8611, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8151, Kappa: 0.4776, Score: 0.6924 +[14:46:17.618064] Best epoch = 33, Best score = 0.6936 +[14:46:17.886261] log_dir: ./output_logs/retfound +[14:46:20.149255] Epoch: [35] [0/2] eta: 0:00:04 lr: 0.000194 loss: 0.3216 (0.3216) time: 2.2619 data: 2.1904 max mem: 9665 +[14:46:20.307066] Epoch: [35] [1/2] eta: 0:00:01 lr: 0.000182 loss: 0.3216 (0.3364) time: 1.2094 data: 1.1412 max mem: 9665 +[14:46:20.384978] Epoch: [35] Total time: 0:00:02 (1.2493 s / it) +[14:46:20.385930] Averaged stats: lr: 0.000182 loss: 0.3216 (0.3364) +[14:46:22.861951] val: [0/2] eta: 0:00:04 loss: 0.2686 (0.2686) time: 2.4634 data: 2.4457 max mem: 9665 +[14:46:22.873761] val: [1/2] eta: 0:00:01 loss: 0.2686 (0.5183) time: 1.2373 data: 1.2229 max mem: 9665 +[14:46:22.956845] val: Total time: 0:00:02 (1.2795 s / it) +[14:46:22.966935] val loss: 0.5183157324790955 +[14:46:22.967187] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8620, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8210, Kappa: 0.4776, Score: 0.6927 +[14:46:23.009409] Best epoch = 33, Best score = 0.6936 +[14:46:23.270128] log_dir: ./output_logs/retfound +[14:46:25.510821] Epoch: [36] [0/2] eta: 0:00:04 lr: 0.000171 loss: 0.3620 (0.3620) time: 2.2397 data: 2.1700 max mem: 9665 +[14:46:25.579113] Epoch: [36] [1/2] eta: 0:00:01 lr: 0.000161 loss: 0.3620 (0.3989) time: 1.1536 data: 1.0850 max mem: 9665 +[14:46:25.656887] Epoch: [36] Total time: 0:00:02 (1.1933 s / it) +[14:46:25.657644] Averaged stats: lr: 0.000161 loss: 0.3620 (0.3989) +[14:46:28.058842] val: [0/2] eta: 0:00:04 loss: 0.2775 (0.2775) time: 2.3902 data: 2.3728 max mem: 9665 +[14:46:28.069570] val: [1/2] eta: 0:00:01 loss: 0.2775 (0.5091) time: 1.2001 data: 1.1865 max mem: 9665 +[14:46:28.151195] val: Total time: 0:00:02 (1.2416 s / it) +[14:46:28.165411] val loss: 0.5090571045875549 +[14:46:28.165708] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8638, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8229, Kappa: 0.4776, Score: 0.6933 +[14:46:28.203321] Best epoch = 33, Best score = 0.6936 +[14:46:28.488746] log_dir: ./output_logs/retfound +[14:46:30.840236] Epoch: [37] [0/2] eta: 0:00:04 lr: 0.000150 loss: 0.3071 (0.3071) time: 2.3504 data: 2.2820 max mem: 9665 +[14:46:30.909891] Epoch: [37] [1/2] eta: 0:00:01 lr: 0.000140 loss: 0.3071 (0.3466) time: 1.2096 data: 1.1429 max mem: 9665 +[14:46:30.982537] Epoch: [37] Total time: 0:00:02 (1.2468 s / it) +[14:46:30.983334] Averaged stats: lr: 0.000140 loss: 0.3071 (0.3466) +[14:46:33.247832] val: [0/2] eta: 0:00:04 loss: 0.2842 (0.2842) time: 2.2536 data: 2.2367 max mem: 9665 +[14:46:33.257472] val: [1/2] eta: 0:00:01 loss: 0.2842 (0.5037) time: 1.1313 data: 1.1184 max mem: 9665 +[14:46:33.335887] val: Total time: 0:00:02 (1.1712 s / it) +[14:46:33.344924] val loss: 0.5036530494689941 +[14:46:33.345178] Accuracy: 0.8000, F1 Score: 0.7133, ROC AUC: 0.8638, Hamming Loss: 0.2000, + Jaccard Score: 0.5780, Precision: 0.7133, Recall: 0.7133, + Average Precision: 0.8229, Kappa: 0.4265, Score: 0.6679 +[14:46:33.396453] Best epoch = 33, Best score = 0.6936 +[14:46:33.667096] log_dir: ./output_logs/retfound +[14:46:36.066255] Epoch: [38] [0/2] eta: 0:00:04 lr: 0.000130 loss: 0.4011 (0.4011) time: 2.3981 data: 2.3291 max mem: 9665 +[14:46:36.134772] Epoch: [38] [1/2] eta: 0:00:01 lr: 0.000120 loss: 0.2422 (0.3216) time: 1.2328 data: 1.1646 max mem: 9665 +[14:46:36.208658] Epoch: [38] Total time: 0:00:02 (1.2707 s / it) +[14:46:36.209682] Averaged stats: lr: 0.000120 loss: 0.2422 (0.3216) +[14:46:38.375892] val: [0/2] eta: 0:00:04 loss: 0.2848 (0.2848) time: 2.1537 data: 2.1370 max mem: 9665 +[14:46:38.385679] val: [1/2] eta: 0:00:01 loss: 0.2848 (0.5054) time: 1.0815 data: 1.0686 max mem: 9665 +[14:46:38.458826] val: Total time: 0:00:02 (1.1187 s / it) +[14:46:38.467771] val loss: 0.5054279565811157 +[14:46:38.467918] Accuracy: 0.8000, F1 Score: 0.7133, ROC AUC: 0.8638, Hamming Loss: 0.2000, + Jaccard Score: 0.5780, Precision: 0.7133, Recall: 0.7133, + Average Precision: 0.8229, Kappa: 0.4265, Score: 0.6679 +[14:46:38.505847] Best epoch = 33, Best score = 0.6936 +[14:46:38.799196] log_dir: ./output_logs/retfound +[14:46:41.058815] Epoch: [39] [0/2] eta: 0:00:04 lr: 0.000110 loss: 0.2327 (0.2327) time: 2.2587 data: 2.1895 max mem: 9665 +[14:46:41.124077] Epoch: [39] [1/2] eta: 0:00:01 lr: 0.000101 loss: 0.2327 (0.3124) time: 1.1616 data: 1.0948 max mem: 9665 +[14:46:41.196267] Epoch: [39] Total time: 0:00:02 (1.1985 s / it) +[14:46:41.196964] Averaged stats: lr: 0.000101 loss: 0.2327 (0.3124) +[14:46:43.361214] val: [0/2] eta: 0:00:04 loss: 0.2794 (0.2794) time: 2.1530 data: 2.1360 max mem: 9665 +[14:46:43.370782] val: [1/2] eta: 0:00:01 loss: 0.2794 (0.5129) time: 1.0810 data: 1.0681 max mem: 9665 +[14:46:43.438669] val: Total time: 0:00:02 (1.1155 s / it) +[14:46:43.447850] val loss: 0.5128768980503082 +[14:46:43.448063] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8656, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8229, Kappa: 0.4776, Score: 0.6939 +[14:46:45.210807] Best epoch = 39, Best score = 0.6939 +[14:46:45.275804] log_dir: ./output_logs/retfound +[14:46:47.401962] Epoch: [40] [0/2] eta: 0:00:04 lr: 0.000092 loss: 0.4136 (0.4136) time: 2.1248 data: 2.0551 max mem: 9665 +[14:46:47.508817] Epoch: [40] [1/2] eta: 0:00:01 lr: 0.000084 loss: 0.1908 (0.3022) time: 1.1154 data: 1.0446 max mem: 9665 +[14:46:47.578825] Epoch: [40] Total time: 0:00:02 (1.1514 s / it) +[14:46:47.580123] Averaged stats: lr: 0.000084 loss: 0.1908 (0.3022) +[14:46:49.791329] val: [0/2] eta: 0:00:04 loss: 0.2778 (0.2778) time: 2.1983 data: 2.1811 max mem: 9665 +[14:46:49.800958] val: [1/2] eta: 0:00:01 loss: 0.2778 (0.5153) time: 1.1036 data: 1.0906 max mem: 9665 +[14:46:49.876742] val: Total time: 0:00:02 (1.1422 s / it) +[14:46:49.885662] val loss: 0.5152709186077118 +[14:46:49.885867] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8665, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8243, Kappa: 0.4776, Score: 0.6942 +[14:46:51.659434] Best epoch = 40, Best score = 0.6942 +[14:46:51.739325] log_dir: ./output_logs/retfound +[14:46:53.948296] Epoch: [41] [0/2] eta: 0:00:04 lr: 0.000076 loss: 0.4626 (0.4626) time: 2.2081 data: 2.1388 max mem: 9665 +[14:46:54.013651] Epoch: [41] [1/2] eta: 0:00:01 lr: 0.000068 loss: 0.4286 (0.4456) time: 1.1364 data: 1.0694 max mem: 9665 +[14:46:54.097097] Epoch: [41] Total time: 0:00:02 (1.1788 s / it) +[14:46:54.099282] Averaged stats: lr: 0.000068 loss: 0.4286 (0.4456) +[14:46:56.335476] val: [0/2] eta: 0:00:04 loss: 0.2715 (0.2715) time: 2.2113 data: 2.1943 max mem: 9665 +[14:46:56.344989] val: [1/2] eta: 0:00:01 loss: 0.2715 (0.5225) time: 1.1101 data: 1.0972 max mem: 9665 +[14:46:56.426234] val: Total time: 0:00:02 (1.1514 s / it) +[14:46:56.435339] val loss: 0.5225353538990021 +[14:46:56.435521] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8674, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8250, Kappa: 0.4776, Score: 0.6945 +[14:46:58.217426] Best epoch = 41, Best score = 0.6945 +[14:46:58.285698] log_dir: ./output_logs/retfound +[14:47:00.613134] Epoch: [42] [0/2] eta: 0:00:04 lr: 0.000061 loss: 0.3992 (0.3992) time: 2.3265 data: 2.2573 max mem: 9665 +[14:47:00.678227] Epoch: [42] [1/2] eta: 0:00:01 lr: 0.000054 loss: 0.3354 (0.3673) time: 1.1955 data: 1.1287 max mem: 9665 +[14:47:00.747968] Epoch: [42] Total time: 0:00:02 (1.2311 s / it) +[14:47:00.748767] Averaged stats: lr: 0.000054 loss: 0.3354 (0.3673) +[14:47:03.015397] val: [0/2] eta: 0:00:04 loss: 0.2694 (0.2694) time: 2.2518 data: 2.2290 max mem: 9665 +[14:47:03.025284] val: [1/2] eta: 0:00:01 loss: 0.2694 (0.5235) time: 1.1306 data: 1.1146 max mem: 9665 +[14:47:03.094616] val: Total time: 0:00:02 (1.1659 s / it) +[14:47:03.103445] val loss: 0.5234984755516052 +[14:47:03.103626] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8781, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8469, Kappa: 0.4776, Score: 0.6981 +[14:47:04.934703] Best epoch = 42, Best score = 0.6981 +[14:47:05.002141] log_dir: ./output_logs/retfound +[14:47:07.348688] Epoch: [43] [0/2] eta: 0:00:04 lr: 0.000047 loss: 0.3535 (0.3535) time: 2.3455 data: 2.2166 max mem: 9665 +[14:47:07.484912] Epoch: [43] [1/2] eta: 0:00:01 lr: 0.000041 loss: 0.3535 (0.3781) time: 1.2404 data: 1.1084 max mem: 9665 +[14:47:07.566162] Epoch: [43] Total time: 0:00:02 (1.2819 s / it) +[14:47:07.574337] Averaged stats: lr: 0.000041 loss: 0.3535 (0.3781) +[14:47:10.087503] val: [0/2] eta: 0:00:05 loss: 0.2659 (0.2659) time: 2.5009 data: 2.4840 max mem: 9665 +[14:47:10.097076] val: [1/2] eta: 0:00:01 loss: 0.2659 (0.5274) time: 1.2550 data: 1.2421 max mem: 9665 +[14:47:10.167403] val: Total time: 0:00:02 (1.2908 s / it) +[14:47:10.176242] val loss: 0.5274004638195038 +[14:47:10.176467] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8781, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8469, Kappa: 0.4776, Score: 0.6981 +[14:47:10.211813] Best epoch = 42, Best score = 0.6981 +[14:47:10.473780] log_dir: ./output_logs/retfound +[14:47:12.834571] Epoch: [44] [0/2] eta: 0:00:04 lr: 0.000035 loss: 0.3168 (0.3168) time: 2.3597 data: 2.2876 max mem: 9665 +[14:47:12.900359] Epoch: [44] [1/2] eta: 0:00:01 lr: 0.000030 loss: 0.3168 (0.3505) time: 1.2123 data: 1.1439 max mem: 9665 +[14:47:12.981157] Epoch: [44] Total time: 0:00:02 (1.2536 s / it) +[14:47:12.989831] Averaged stats: lr: 0.000030 loss: 0.3168 (0.3505) +[14:47:15.398502] val: [0/2] eta: 0:00:04 loss: 0.2625 (0.2625) time: 2.4013 data: 2.3844 max mem: 9665 +[14:47:15.408083] val: [1/2] eta: 0:00:01 loss: 0.2625 (0.5317) time: 1.2051 data: 1.1922 max mem: 9665 +[14:47:15.480167] val: Total time: 0:00:02 (1.2418 s / it) +[14:47:15.494595] val loss: 0.5317047834396362 +[14:47:15.494892] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8817, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8584, Kappa: 0.4776, Score: 0.6993 +[14:47:17.661695] Best epoch = 44, Best score = 0.6993 +[14:47:17.840692] log_dir: ./output_logs/retfound +[14:47:20.063767] Epoch: [45] [0/2] eta: 0:00:04 lr: 0.000025 loss: 0.2903 (0.2903) time: 2.2220 data: 2.0817 max mem: 9665 +[14:47:20.198234] Epoch: [45] [1/2] eta: 0:00:01 lr: 0.000020 loss: 0.2592 (0.2747) time: 1.1778 data: 1.0409 max mem: 9665 +[14:47:20.293790] Epoch: [45] Total time: 0:00:02 (1.2264 s / it) +[14:47:20.298293] Averaged stats: lr: 0.000020 loss: 0.2592 (0.2747) +[14:47:22.562807] val: [0/2] eta: 0:00:04 loss: 0.2610 (0.2610) time: 2.2533 data: 2.2197 max mem: 9665 +[14:47:22.578510] val: [1/2] eta: 0:00:01 loss: 0.2610 (0.5344) time: 1.1342 data: 1.1099 max mem: 9665 +[14:47:22.655418] val: Total time: 0:00:02 (1.1733 s / it) +[14:47:22.666905] val loss: 0.5344215035438538 +[14:47:22.667120] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8817, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8584, Kappa: 0.4776, Score: 0.6993 +[14:47:22.708480] Best epoch = 44, Best score = 0.6993 +[14:47:22.983273] log_dir: ./output_logs/retfound +[14:47:25.407349] Epoch: [46] [0/2] eta: 0:00:04 lr: 0.000016 loss: 0.2517 (0.2517) time: 2.4228 data: 2.3520 max mem: 9665 +[14:47:25.474438] Epoch: [46] [1/2] eta: 0:00:01 lr: 0.000013 loss: 0.2517 (0.3131) time: 1.2445 data: 1.1761 max mem: 9665 +[14:47:25.546724] Epoch: [46] Total time: 0:00:02 (1.2816 s / it) +[14:47:25.547483] Averaged stats: lr: 0.000013 loss: 0.2517 (0.3131) +[14:47:27.790504] val: [0/2] eta: 0:00:04 loss: 0.2605 (0.2605) time: 2.2319 data: 2.2148 max mem: 9665 +[14:47:27.800068] val: [1/2] eta: 0:00:01 loss: 0.2605 (0.5356) time: 1.1205 data: 1.1075 max mem: 9665 +[14:47:27.871219] val: Total time: 0:00:02 (1.1567 s / it) +[14:47:27.880178] val loss: 0.5356009006500244 +[14:47:27.880389] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8790, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8542, Kappa: 0.4776, Score: 0.6984 +[14:47:27.917860] Best epoch = 44, Best score = 0.6993 +[14:47:28.187046] log_dir: ./output_logs/retfound +[14:47:30.641863] Epoch: [47] [0/2] eta: 0:00:04 lr: 0.000010 loss: 0.3373 (0.3373) time: 2.4537 data: 2.3122 max mem: 9665 +[14:47:30.773493] Epoch: [47] [1/2] eta: 0:00:01 lr: 0.000007 loss: 0.2620 (0.2996) time: 1.2923 data: 1.1561 max mem: 9665 +[14:47:30.846506] Epoch: [47] Total time: 0:00:02 (1.3297 s / it) +[14:47:30.847279] Averaged stats: lr: 0.000007 loss: 0.2620 (0.2996) +[14:47:33.314527] val: [0/2] eta: 0:00:04 loss: 0.2602 (0.2602) time: 2.4566 data: 2.4364 max mem: 9665 +[14:47:33.324077] val: [1/2] eta: 0:00:01 loss: 0.2602 (0.5363) time: 1.2328 data: 1.2183 max mem: 9665 +[14:47:33.399223] val: Total time: 0:00:02 (1.2710 s / it) +[14:47:33.408175] val loss: 0.5362502932548523 +[14:47:33.408425] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8781, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8542, Kappa: 0.4776, Score: 0.6981 +[14:47:33.446570] Best epoch = 44, Best score = 0.6993 +[14:47:33.725219] log_dir: ./output_logs/retfound +[14:47:36.098231] Epoch: [48] [0/2] eta: 0:00:04 lr: 0.000005 loss: 0.3559 (0.3559) time: 2.3720 data: 2.2338 max mem: 9665 +[14:47:36.236155] Epoch: [48] [1/2] eta: 0:00:01 lr: 0.000003 loss: 0.3189 (0.3374) time: 1.2546 data: 1.1169 max mem: 9665 +[14:47:36.309494] Epoch: [48] Total time: 0:00:02 (1.2920 s / it) +[14:47:36.318613] Averaged stats: lr: 0.000003 loss: 0.3189 (0.3374) +[14:47:39.107676] val: [0/2] eta: 0:00:05 loss: 0.2603 (0.2603) time: 2.7667 data: 2.7500 max mem: 9665 +[14:47:39.117108] val: [1/2] eta: 0:00:01 loss: 0.2603 (0.5364) time: 1.3878 data: 1.3750 max mem: 9665 +[14:47:39.185122] val: Total time: 0:00:02 (1.4224 s / it) +[14:47:39.193953] val loss: 0.5363571643829346 +[14:47:39.194198] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8781, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8542, Kappa: 0.4776, Score: 0.6981 +[14:47:39.254771] Best epoch = 44, Best score = 0.6993 +[14:47:39.558069] log_dir: ./output_logs/retfound +[14:47:42.189092] Epoch: [49] [0/2] eta: 0:00:05 lr: 0.000002 loss: 0.3696 (0.3696) time: 2.6301 data: 2.5612 max mem: 9665 +[14:47:42.403887] Epoch: [49] [1/2] eta: 0:00:01 lr: 0.000001 loss: 0.2978 (0.3337) time: 1.4220 data: 1.3549 max mem: 9665 +[14:47:42.484412] Epoch: [49] Total time: 0:00:02 (1.4631 s / it) +[14:47:42.485169] Averaged stats: lr: 0.000001 loss: 0.2978 (0.3337) +[14:47:45.199417] val: [0/2] eta: 0:00:05 loss: 0.2603 (0.2603) time: 2.6473 data: 2.6275 max mem: 9665 +[14:47:45.213700] val: [1/2] eta: 0:00:01 loss: 0.2603 (0.5361) time: 1.3304 data: 1.3138 max mem: 9665 +[14:47:45.288936] val: Total time: 0:00:02 (1.3688 s / it) +[14:47:45.297789] val loss: 0.536094069480896 +[14:47:45.298002] Accuracy: 0.8250, F1 Score: 0.7386, ROC AUC: 0.8781, Hamming Loss: 0.1750, + Jaccard Score: 0.6083, Precision: 0.7500, Recall: 0.7294, + Average Precision: 0.8542, Kappa: 0.4776, Score: 0.6981 +[14:47:45.378283] Best epoch = 44, Best score = 0.6993 +[14:47:48.155789] Test with the best model, epoch = 44: +[14:47:51.009012] test: [0/3] eta: 0:00:08 loss: 0.2367 (0.2367) time: 2.8354 data: 2.8023 max mem: 9665 +[14:47:51.206186] test: [2/3] eta: 0:00:01 loss: 0.2869 (0.3032) time: 1.0107 data: 0.9342 max mem: 9665 +[14:47:51.275572] test: Total time: 0:00:03 (1.0342 s / it) +[14:47:51.287474] val loss: 0.30319035053253174 +[14:47:51.287599] Accuracy: 0.8750, F1 Score: 0.8333, ROC AUC: 0.9211, Hamming Loss: 0.1250, + Jaccard Score: 0.7231, Precision: 0.8150, Recall: 0.8602, + Average Precision: 0.9201, Kappa: 0.6678, Score: 0.8074 +[14:47:52.022218] Training time 0:05:03 +[rank0]:[W701 14:47:52.498299935 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) diff --git a/results/downsample/adam/025/vit/confusion_matrix.png b/results/downsample/adam/025/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..01ef2b028acfc9e529189b9793a8fba8f8fcb699 --- /dev/null +++ b/results/downsample/adam/025/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:920bf98e31a117f5b9c876a5e8e88f108a46202523ea04ba682ade4f312404a8 +size 68242 diff --git a/results/downsample/adam/025/vit/log.csv b/results/downsample/adam/025/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..a9e00f0d8b392495c3591c341e222d4ab62e1a69 --- /dev/null +++ b/results/downsample/adam/025/vit/log.csv @@ -0,0 +1,39 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.0639675855636597,0.3,0.3261648745519713,0.16738866653636775,0.0 +1,0.7906926274299622,0.275,0.37275985663082434,0.21831853779572438,7.39441178203743e-08 +2,0.9115864038467407,0.35,0.4802867383512545,0.30112236649490437,1.478882356407486e-07 +3,0.6448042392730713,0.675,0.6272401433691757,0.42734192634138984,2.2183235346112292e-07 +4,0.6287508010864258,0.775,0.6738351254480287,0.4856485572060922,2.957764712814972e-07 +5,0.5496615767478943,0.775,0.6738351254480287,0.4856485572060922,3.697205891018715e-07 +6,0.5109244585037231,0.775,0.7096774193548387,0.3820990458882327,3.6927027871539135e-07 +7,0.558362603187561,0.7,0.7060931899641577,0.38844535824631915,3.679215414228887e-07 +8,0.7030777931213379,0.7,0.6989247311827957,0.5487099000422984,3.6568094813687817e-07 +9,0.48274046182632446,0.5,0.7060931899641577,0.46774561690641736,3.625594148026254e-07 +10,0.5737808346748352,0.45,0.7060931899641577,0.43718781316143707,3.585721492167595e-07 +11,0.7070144414901733,0.45,0.7060931899641577,0.43718781316143707,3.537385769364163e-07 +12,0.7299017906188965,0.55,0.7275985663082437,0.47581326344223984,3.480822466398767e-07 +13,0.574883222579956,0.725,0.7455197132616488,0.5837384718927753,3.4163071539977294e-07 +14,0.5160149931907654,0.725,0.7562724014336918,0.47022713568740326,3.344154144278013e-07 +15,0.5480642318725586,0.8,0.7706093189964157,0.5424262491103766,3.2647149594501757e-07 +16,0.6683022975921631,0.825,0.7813620071684588,0.5731287481249294,3.178376619237501e-07 +17,0.6420748829841614,0.8,0.7921146953405018,0.5495947078917386,3.0855597553548053e-07 +18,0.6943216919898987,0.775,0.7921146953405017,0.5637255363455206,2.986716562233006e-07 +19,0.4880589246749878,0.775,0.7956989247311828,0.6427396490272378,2.88232859397331e-07 +20,0.3831441402435303,0.65,0.7921146953405017,0.5936266972664387,2.7729044182640365e-07 +21,0.6317847371101379,0.5,0.7956989247311828,0.4976141951620923,2.6589771386899607e-07 +22,0.4736146926879883,0.45,0.7992831541218637,0.468251134547339,2.541101797505239e-07 +23,0.4527874290943146,0.525,0.8100358422939067,0.5178867761296154,2.419852671523334e-07 +24,0.5480691194534302,0.65,0.8243727598566307,0.6043793854384817,2.2958204742981102e-07 +25,0.4869018793106079,0.675,0.8387096774193548,0.6123735886642384,2.1696094782267776e-07 +26,0.4313433766365051,0.75,0.8458781362007168,0.6160309918749528,2.0418345705955085e-07 +27,0.4774608910083771,0.8,0.8422939068100358,0.6606929510155316,1.9131182579103953e-07 +28,0.5121029615402222,0.8,0.8422939068100358,0.6606929510155316,1.7840876331083206e-07 +29,0.32458382844924927,0.8,0.8387096774193548,0.6594982078853046,1.655371320423207e-07 +30,0.42480599880218506,0.8,0.8387096774193548,0.6594982078853046,1.5275964127919378e-07 +31,0.44482940435409546,0.8,0.8387096774193548,0.6594982078853046,1.4013854167206046e-07 +32,0.5201488733291626,0.8,0.8422939068100358,0.6606929510155316,1.2773532194953815e-07 +33,0.3977617621421814,0.775,0.8351254480286738,0.6346077319299411,1.1561040935134761e-07 +34,0.42279279232025146,0.75,0.8387096774193548,0.6136415056144989,1.0382287523287539e-07 +35,0.5137747526168823,0.75,0.8351254480286738,0.6124467624842719,9.243014727546792e-08 +36,0.42430582642555237,0.75,0.8351254480286738,0.6124467624842719,8.148772970454054e-08 +37,0.44768694043159485,0.75,0.8279569892473118,0.6100572762238179,7.104893287857089e-08 diff --git a/results/downsample/adam/025/vit/metrics.json b/results/downsample/adam/025/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..3cdf4a18fb779347700328f83080865f2fe1b700 --- /dev/null +++ b/results/downsample/adam/025/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8125, + "balanced_accuracy": 0.7804659498207885, + "precision_macro": 0.7387490465293669, + "recall_macro": 0.7804659498207885, + "f1_macro": 0.7540479606476738, + "precision_weighted": 0.8341914569031272, + "recall_weighted": 0.8125, + "f1_weighted": 0.8199938511990161, + "cohen_kappa": 0.5106035889070146, + "quadratic_weighted_kappa": 0.5106035889070146, + "mcc": 0.5175363875271632, + "auroc": 0.8351254480286737, + "auprc": 0.6520936912622649, + "sensitivity": 0.7222222222222222, + "specificity": 0.8387096774193549, + "precision_pos": 0.5652173913043478, + "f1_pos": 0.6341463414634146, + "per_class": { + "0": { + "precision": 0.9122807017543859, + "recall": 0.8387096774193549, + "f1-score": 0.8739495798319328, + "support": 62.0 + }, + "1": { + "precision": 0.5652173913043478, + "recall": 0.7222222222222222, + "f1-score": 0.6341463414634146, + "support": 18.0 + }, + "accuracy": 0.8125, + "macro avg": { + "precision": 0.7387490465293669, + "recall": 0.7804659498207885, + "f1-score": 0.7540479606476738, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8341914569031272, + "recall": 0.8125, + "f1-score": 0.8199938511990161, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/025/vit/pr.png b/results/downsample/adam/025/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..66dbb06b5168cd8ef4a7dab8569e0a3384612e48 --- /dev/null +++ b/results/downsample/adam/025/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35050e3bb2ffe5829edbf61be8c639007da93a0369f11332fd60245b8d971fa3 +size 54504 diff --git a/results/downsample/adam/025/vit/roc.png b/results/downsample/adam/025/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..36e8dd0c788650509cafaffa26f7d6f2a07981a0 --- /dev/null +++ b/results/downsample/adam/025/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b35e05841b3d45f02f35d7c127ac4cad8a5962a7672f5b563a907c1e2dc350e +size 57720 diff --git a/results/downsample/adam/025/vit/test_pred.npz b/results/downsample/adam/025/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..5a5d7a30ea5f1acb48638fabffd565d38e37dcdd --- /dev/null +++ b/results/downsample/adam/025/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42019f96c1452cf0c868fa86b48944e5ca09c4dbd7085900edcc34041ea17ff9 +size 1790 diff --git a/results/downsample/adam/025/vit/train.log b/results/downsample/adam/025/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1fce7537f1a01e766d47f5d42ef28325ddaba6dd --- /dev/null +++ b/results/downsample/adam/025/vit/train.log @@ -0,0 +1,199 @@ +[vit] train=70 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=1.0640 val_acc=0.3000 val_auc=0.3262 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7907 val_acc=0.2750 val_auc=0.3728 score=0.2183 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.9116 val_acc=0.3500 val_auc=0.4803 score=0.3011 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6448 val_acc=0.6750 val_auc=0.6272 score=0.4273 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6288 val_acc=0.7750 val_auc=0.6738 score=0.4856 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5497 val_acc=0.7750 val_auc=0.6738 score=0.4856 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5109 val_acc=0.7750 val_auc=0.7097 score=0.3821 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5584 val_acc=0.7000 val_auc=0.7061 score=0.3884 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.7031 val_acc=0.7000 val_auc=0.6989 score=0.5487 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4827 val_acc=0.5000 val_auc=0.7061 score=0.4677 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.5738 val_acc=0.4500 val_auc=0.7061 score=0.4372 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.7070 val_acc=0.4500 val_auc=0.7061 score=0.4372 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.7299 val_acc=0.5500 val_auc=0.7276 score=0.4758 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.5749 val_acc=0.7250 val_auc=0.7455 score=0.5837 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.5160 val_acc=0.7250 val_auc=0.7563 score=0.4702 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.5481 val_acc=0.8000 val_auc=0.7706 score=0.5424 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.6683 val_acc=0.8250 val_auc=0.7814 score=0.5731 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.6421 val_acc=0.8000 val_auc=0.7921 score=0.5496 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.6943 val_acc=0.7750 val_auc=0.7921 score=0.5637 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4881 val_acc=0.7750 val_auc=0.7957 score=0.6427 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3831 val_acc=0.6500 val_auc=0.7921 score=0.5936 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.6318 val_acc=0.5000 val_auc=0.7957 score=0.4976 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.4736 val_acc=0.4500 val_auc=0.7993 score=0.4683 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4528 val_acc=0.5250 val_auc=0.8100 score=0.5179 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.5481 val_acc=0.6500 val_auc=0.8244 score=0.6044 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.4869 val_acc=0.6750 val_auc=0.8387 score=0.6124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4313 val_acc=0.7500 val_auc=0.8459 score=0.6160 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.4775 val_acc=0.8000 val_auc=0.8423 score=0.6607 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.5121 val_acc=0.8000 val_auc=0.8423 score=0.6607 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3246 val_acc=0.8000 val_auc=0.8387 score=0.6595 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.4248 val_acc=0.8000 val_auc=0.8387 score=0.6595 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.4448 val_acc=0.8000 val_auc=0.8387 score=0.6595 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep32 loss=0.5201 val_acc=0.8000 val_auc=0.8423 score=0.6607 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep33 loss=0.3978 val_acc=0.7750 val_auc=0.8351 score=0.6346 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep34 loss=0.4228 val_acc=0.7500 val_auc=0.8387 score=0.6136 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep35 loss=0.5138 val_acc=0.7500 val_auc=0.8351 score=0.6124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep36 loss=0.4243 val_acc=0.7500 val_auc=0.8351 score=0.6124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep37 loss=0.4477 val_acc=0.7500 val_auc=0.8280 score=0.6101 +[vit] early stop at ep37 (best ep27 score=0.6607) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=27 best_val_score=0.6607 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/025/vit acc=0.8125 auroc=0.8351254480286737 f1_macro=0.7540 qwk=0.5106035889070146 diff --git a/results/downsample/adam/050/resnet/confusion_matrix.png b/results/downsample/adam/050/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..4fd5dd2b5eae0523c5a3a43fa2040e6fe842db1b --- /dev/null +++ b/results/downsample/adam/050/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0bf4261be52c0730615f3e022caa514d068f6c1a5770466c7d769fc6d02da0ff +size 69485 diff --git a/results/downsample/adam/050/resnet/log.csv b/results/downsample/adam/050/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..ec9bc8f7195caa4fa9731de87912d8320f2290f0 --- /dev/null +++ b/results/downsample/adam/050/resnet/log.csv @@ -0,0 +1,37 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6868610978126526,0.5,0.38351254480286734,0.20961805323656654,8.333333333333333e-05 +1,0.6840530633926392,0.6,0.33691756272401435,0.20549581839904416,0.00025 +2,0.6767300069332123,0.725,0.42652329749103945,0.25256808055747143,0.0004166666666666667 +3,0.648181140422821,0.7,0.6702508960573477,0.3184183461864729,0.0004998603909325636 +4,0.6379013359546661,0.725,0.7706093189964157,0.42776666217927645,0.0004987444537721261 +5,0.5991038680076599,0.75,0.7921146953405018,0.5027168108842783,0.0004965175634975071 +6,0.557020902633667,0.75,0.7311827956989246,0.43400652703910564,0.0004931896659412593 +7,0.528007298707962,0.8,0.7491039426523298,0.48475491681031285,0.0004887756243017281 +8,0.4937596023082733,0.8,0.7222222222222222,0.4757943433336103,0.00048329515276040064 +9,0.4438846558332443,0.825,0.7992831541218637,0.6176045263069817,0.0004767727284335852 +10,0.3952649235725403,0.8,0.8172043010752689,0.5579579098033277,0.0004692374820516679 +11,0.3369579613208771,0.8,0.8387096774193548,0.5651263685846896,0.00046072306785419927 +12,0.2719482183456421,0.825,0.8602150537634409,0.5994130969899235,0.0004512675132818908 +13,0.25056545436382294,0.825,0.8637992831541219,0.6006078401201504,0.00044091304913683303 +14,0.21361441910266876,0.825,0.8530465949820788,0.5970236107294694,0.0004297059209694824 +15,0.19061264395713806,0.8,0.8637992831541219,0.5229866969775769,0.0004176961825348059 +16,0.14823254197835922,0.775,0.8548387096774194,0.4964804334570905,0.0004049374722400612 +17,0.11817124485969543,0.825,0.8530465949820788,0.6355256732603868,0.0003914867735826488 +18,0.09893495589494705,0.8,0.8637992831541218,0.6122092910013294,0.00037740416064797937 +19,0.0881650261580944,0.825,0.8637992831541218,0.6391099026510677,0.00036275252980402545 +20,0.060780243948102,0.85,0.8315412186379928,0.6580144299630094,0.0003475973187908737 +21,0.08541093021631241,0.85,0.7921146953405017,0.6448722555305123,0.00033200621445989226 +22,0.05186849646270275,0.85,0.7813620071684588,0.6412880261398315,0.0003160488504678216 +23,0.035933734849095345,0.85,0.7885304659498207,0.6436775124002855,0.000299796496275959 +24,0.03296269103884697,0.85,0.8064516129032258,0.6496512280514204,0.00028332173884344477 +25,0.039373237639665604,0.85,0.7903225806451613,0.673314674344187,0.0002666981584362804 +26,0.018641382921487093,0.85,0.7777777777777778,0.6691330733883926,0.00025 +27,0.019445965066552162,0.825,0.7741935483870968,0.6092413243953928,0.0002333018415637196 +28,0.014585250522941351,0.8,0.7670250896057347,0.5412315059801496,0.00021667826115655538 +29,0.019004317931830883,0.8,0.7795698924731183,0.545413106935944,0.00020020350372404103 +30,0.011807670816779137,0.8,0.8136200716845878,0.5567631666731007,0.0001839511495321785 +31,0.013345627579838037,0.8,0.8530465949820789,0.5699053411055975,0.00016799378554010772 +32,0.01423738058656454,0.8,0.8566308243727598,0.5711000842358246,0.00015240268120912632 +33,0.010431832633912563,0.8,0.8709677419354839,0.5758790567567327,0.0001372474701959745 +34,0.01287631830200553,0.8,0.8853046594982079,0.5806580292776405,0.00012259583935202061 +35,0.01891399919986725,0.8,0.888888888888889,0.5818527724078676,0.00010851322641735117 diff --git a/results/downsample/adam/050/resnet/metrics.json b/results/downsample/adam/050/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..7235a4e9566b5a51aa9400a9b4da7c8910fa0600 --- /dev/null +++ b/results/downsample/adam/050/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.875, + "balanced_accuracy": 0.7616487455197133, + "precision_macro": 0.857843137254902, + "recall_macro": 0.7616487455197133, + "f1_macro": 0.7948717948717949, + "precision_weighted": 0.8713235294117647, + "recall_weighted": 0.875, + "f1_weighted": 0.8653846153846153, + "cohen_kappa": 0.5934959349593496, + "quadratic_weighted_kappa": 0.5934959349593496, + "mcc": 0.6119778033738925, + "auroc": 0.9247311827956989, + "auprc": 0.8385595763885239, + "sensitivity": 0.5555555555555556, + "specificity": 0.967741935483871, + "precision_pos": 0.8333333333333334, + "f1_pos": 0.6666666666666666, + "per_class": { + "0": { + "precision": 0.8823529411764706, + "recall": 0.967741935483871, + "f1-score": 0.9230769230769231, + "support": 62.0 + }, + "1": { + "precision": 0.8333333333333334, + "recall": 0.5555555555555556, + "f1-score": 0.6666666666666666, + "support": 18.0 + }, + "accuracy": 0.875, + "macro avg": { + "precision": 0.857843137254902, + "recall": 0.7616487455197133, + "f1-score": 0.7948717948717949, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8713235294117647, + "recall": 0.875, + "f1-score": 0.8653846153846153, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/050/resnet/pr.png b/results/downsample/adam/050/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..10c5663d9355fb32e665877e272dc30b232dcb18 --- /dev/null +++ b/results/downsample/adam/050/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95c414409158618c387a6baf4afa48f2fe44a30faed0abc145473894f7fd3cd7 +size 45585 diff --git a/results/downsample/adam/050/resnet/roc.png b/results/downsample/adam/050/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..8208fb351cce5584fd4a7bd8b2aacda268996bb7 --- /dev/null +++ b/results/downsample/adam/050/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd5852c1e7ba92dfc420c135621a7f774a3f297f2a88df1229a7acbfae4b5dcf +size 57440 diff --git a/results/downsample/adam/050/resnet/test_pred.npz b/results/downsample/adam/050/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..bfe1e34b3c08bbf627150dae99fb185502a6aaaa --- /dev/null +++ b/results/downsample/adam/050/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:155311571c0e8c6db50495520a68e274446b0cff27f2d4653f2a699a9ea0825c +size 1790 diff --git a/results/downsample/adam/050/resnet/train.log b/results/downsample/adam/050/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..cd23bb36163216e60ebfee2ea1c9e2ab44710f67 --- /dev/null +++ b/results/downsample/adam/050/resnet/train.log @@ -0,0 +1,189 @@ +[resnet] train=140 val=40 test=80 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6869 val_acc=0.5000 val_auc=0.3835 score=0.2096 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6841 val_acc=0.6000 val_auc=0.3369 score=0.2055 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6767 val_acc=0.7250 val_auc=0.4265 score=0.2526 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6482 val_acc=0.7000 val_auc=0.6703 score=0.3184 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6379 val_acc=0.7250 val_auc=0.7706 score=0.4278 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.5991 val_acc=0.7500 val_auc=0.7921 score=0.5027 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.5570 val_acc=0.7500 val_auc=0.7312 score=0.4340 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.5280 val_acc=0.8000 val_auc=0.7491 score=0.4848 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.4938 val_acc=0.8000 val_auc=0.7222 score=0.4758 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.4439 val_acc=0.8250 val_auc=0.7993 score=0.6176 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.3953 val_acc=0.8000 val_auc=0.8172 score=0.5580 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.3370 val_acc=0.8000 val_auc=0.8387 score=0.5651 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.2719 val_acc=0.8250 val_auc=0.8602 score=0.5994 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.2506 val_acc=0.8250 val_auc=0.8638 score=0.6006 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.2136 val_acc=0.8250 val_auc=0.8530 score=0.5970 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.1906 val_acc=0.8000 val_auc=0.8638 score=0.5230 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.1482 val_acc=0.7750 val_auc=0.8548 score=0.4965 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.1182 val_acc=0.8250 val_auc=0.8530 score=0.6355 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0989 val_acc=0.8000 val_auc=0.8638 score=0.6122 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0882 val_acc=0.8250 val_auc=0.8638 score=0.6391 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0608 val_acc=0.8500 val_auc=0.8315 score=0.6580 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0854 val_acc=0.8500 val_auc=0.7921 score=0.6449 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0519 val_acc=0.8500 val_auc=0.7814 score=0.6413 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0359 val_acc=0.8500 val_auc=0.7885 score=0.6437 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0330 val_acc=0.8500 val_auc=0.8065 score=0.6497 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0394 val_acc=0.8500 val_auc=0.7903 score=0.6733 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0186 val_acc=0.8500 val_auc=0.7778 score=0.6691 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0194 val_acc=0.8250 val_auc=0.7742 score=0.6092 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0146 val_acc=0.8000 val_auc=0.7670 score=0.5412 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0190 val_acc=0.8000 val_auc=0.7796 score=0.5454 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep30 loss=0.0118 val_acc=0.8000 val_auc=0.8136 score=0.5568 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep31 loss=0.0133 val_acc=0.8000 val_auc=0.8530 score=0.5699 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep32 loss=0.0142 val_acc=0.8000 val_auc=0.8566 score=0.5711 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep33 loss=0.0104 val_acc=0.8000 val_auc=0.8710 score=0.5759 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep34 loss=0.0129 val_acc=0.8000 val_auc=0.8853 score=0.5807 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep35 loss=0.0189 val_acc=0.8000 val_auc=0.8889 score=0.5819 +[resnet] early stop at ep35 (best ep25 score=0.6733) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=25 best_val_score=0.6733 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/050/resnet acc=0.8750 auroc=0.9247311827956989 f1_macro=0.7949 qwk=0.5934959349593496 diff --git a/results/downsample/adam/050/retfound/confusion_matrix.png b/results/downsample/adam/050/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..733b83ea877bc153647ac339aa6088dd1906b567 --- /dev/null +++ b/results/downsample/adam/050/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7142b7b7972e6151b7e2e8d6ac8a164aec7a1fea97226fe5dd05f1952e67a98 +size 68153 diff --git a/results/downsample/adam/050/retfound/confusion_matrix_test.jpg b/results/downsample/adam/050/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..dca75d74bd4dbe952a45d142f6b5e9f8f16695af --- /dev/null +++ b/results/downsample/adam/050/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32bf90581835c5cfef3b2cfa602abce96759dac4a4a2fe9e2c20659164953ba4 +size 257614 diff --git a/results/downsample/adam/050/retfound/log.txt b/results/downsample/adam/050/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..ad142c0512910425db0e3c1294ab3343ba6ef357 --- /dev/null +++ b/results/downsample/adam/050/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.34375e-05, "train_loss": 0.6916275024414062, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.593750000000001e-05, "train_loss": 0.67095947265625, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014843750000000002, "train_loss": 0.5983791351318359, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021093750000000002, "train_loss": 0.5122385025024414, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002734375, "train_loss": 0.5121245384216309, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003359375, "train_loss": 0.5894756317138672, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00039843750000000003, "train_loss": 0.4907093048095703, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004609375, "train_loss": 0.4813241958618164, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005234375, "train_loss": 0.49800825119018555, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005859375, "train_loss": 0.4729076623916626, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247895471787854, "train_loss": 0.4618467092514038, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006231077089086987, "train_loss": 0.4049675464630127, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006195139534919123, "train_loss": 0.3947642147541046, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006140304376256159, "train_loss": 0.43792739510536194, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006066909690085789, "train_loss": 0.3956857919692993, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005975407979054466, "train_loss": 0.33892762660980225, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005866363381638103, "train_loss": 0.36240649223327637, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005740448194040781, "train_loss": 0.30792927742004395, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005598438725265087, "train_loss": 0.36508120596408844, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005441210510908961, "train_loss": 0.4274558871984482, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005269732915197597, "train_loss": 0.30642037093639374, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005085063154530642, "train_loss": 0.38374583423137665, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004888339779391536, "train_loss": 0.3515627384185791, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00046807756548051713, "train_loss": 0.39791035652160645, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00044636504826216976, "train_loss": 0.3352440595626831, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004238302911729072, "train_loss": 0.35935504734516144, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004006122284837521, "train_loss": 0.3457023352384567, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003768540072719723, "train_loss": 0.2980170249938965, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035270210487174864, "train_loss": 0.2961459159851074, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.000328305425792701, "train_loss": 0.30029987543821335, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030381438367407094, "train_loss": 0.3299238309264183, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002793799739346176, "train_loss": 0.3516339436173439, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002551528428356505, "train_loss": 0.3470081090927124, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.0002312823586967332, "train_loss": 0.3307485803961754, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.0002079156909903239, "train_loss": 0.36081282049417496, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018519690299303795, "train_loss": 0.2946041226387024, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016326606358764225, "train_loss": 0.27281615138053894, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014225838369181528, "train_loss": 0.2497253492474556, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012230338263787712, "train_loss": 0.3009471744298935, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010352408964303667, "train_loss": 0.3267394080758095, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.60362852933556e-05, "train_loss": 0.2467283234000206, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.994778771793487e-05, "train_loss": 0.2131228744983673, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.535778785429517e-05, "train_loss": 0.2563367113471031, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.2356237903263196e-05, "train_loss": 0.33405549824237823, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.102329674374109e-05, "train_loss": 0.23964642360806465, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.1428835726562727e-05, "train_loss": 0.3255276530981064, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3632007894380988e-05, "train_loss": 0.26253604888916016, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.680883283489666e-06, "train_loss": 0.28621405735611916, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.6121525560646343e-06, "train_loss": 0.23660265654325485, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.4509007900307425e-06, "train_loss": 0.31956296041607857, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/adam/050/retfound/metrics.json b/results/downsample/adam/050/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..4052dde8c59aafd5c900dd785fa3ddf1cc942775 --- /dev/null +++ b/results/downsample/adam/050/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.925, + "balanced_accuracy": 0.89247311827957, + "precision_macro": 0.89247311827957, + "recall_macro": 0.89247311827957, + "f1_macro": 0.89247311827957, + "precision_weighted": 0.925, + "recall_weighted": 0.925, + "f1_weighted": 0.925, + "cohen_kappa": 0.7849462365591398, + "quadratic_weighted_kappa": 0.7849462365591398, + "mcc": 0.7849462365591398, + "auroc": 0.9444444444444444, + "auprc": 0.9235311447811447, + "sensitivity": 0.8333333333333334, + "specificity": 0.9516129032258065, + "precision_pos": 0.8333333333333334, + "f1_pos": 0.8333333333333334, + "per_class": { + "0": { + "precision": 0.9516129032258065, + "recall": 0.9516129032258065, + "f1-score": 0.9516129032258065, + "support": 62.0 + }, + "1": { + "precision": 0.8333333333333334, + "recall": 0.8333333333333334, + "f1-score": 0.8333333333333334, + "support": 18.0 + }, + "accuracy": 0.925, + "macro avg": { + "precision": 0.89247311827957, + "recall": 0.89247311827957, + "f1-score": 0.89247311827957, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.925, + "recall": 0.925, + "f1-score": 0.925, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/050/retfound/metrics_test.csv b/results/downsample/adam/050/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..3d17f6dc11583ad1fdda292f114109e9def95be8 --- /dev/null +++ b/results/downsample/adam/050/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.22319571177164713,0.925,0.89247311827957,0.944668458781362,0.075,0.810989010989011,0.89247311827957,0.89247311827957,0.9482573749323184,0.7849462365591398 diff --git a/results/downsample/adam/050/retfound/metrics_val.csv b/results/downsample/adam/050/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..935db11dc94b29013f05feb605f7f82335719136 --- /dev/null +++ b/results/downsample/adam/050/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6932296752929688,0.775,0.43661971830985913,0.7096774193548387,0.225,0.3875,0.3875,0.5,0.6026463560334528,0.0 +0.6982345581054688,0.775,0.43661971830985913,0.7329749103942653,0.225,0.3875,0.3875,0.5,0.7383121445844772,0.0 +0.7288017272949219,0.775,0.43661971830985913,0.739247311827957,0.225,0.3875,0.3875,0.5,0.7259327319676805,0.0 +0.8104629516601562,0.775,0.43661971830985913,0.7634408602150538,0.225,0.3875,0.3875,0.5,0.707076180500089,0.0 +0.8870391845703125,0.775,0.43661971830985913,0.825268817204301,0.225,0.3875,0.3875,0.5,0.7663250874570724,0.0 +0.8351268768310547,0.775,0.43661971830985913,0.8521505376344086,0.225,0.3875,0.3875,0.5,0.7832095713574962,0.0 +0.7682437896728516,0.775,0.43661971830985913,0.8646953405017921,0.225,0.3875,0.3875,0.5,0.7954671296974392,0.0 +0.7307825088500977,0.775,0.43661971830985913,0.8745519713261649,0.225,0.3875,0.3875,0.5,0.8181057574125019,0.0 +0.7333745956420898,0.775,0.43661971830985913,0.8637992831541219,0.225,0.3875,0.3875,0.5,0.7973422952175879,0.0 +0.7696075439453125,0.775,0.43661971830985913,0.8637992831541219,0.225,0.3875,0.3875,0.5,0.7973422952175879,0.0 +0.5585274696350098,0.775,0.5866819747416763,0.8637992831541218,0.225,0.47248803827751196,0.6527777777777778,0.578853046594982,0.7901035544262156,0.1964285714285714 +0.4624495506286621,0.75,0.6865203761755485,0.8602150537634408,0.25,0.5404411764705883,0.6752136752136753,0.7204301075268817,0.785944595779744,0.3808049535603715 +0.55164635181427,0.8,0.7132616487455197,0.8566308243727598,0.2,0.5780219780219781,0.7132616487455197,0.7132616487455197,0.7791640873412337,0.42652329749103934 +0.6862562894821167,0.85,0.7402597402597402,0.8637992831541219,0.15,0.6166666666666667,0.8285714285714285,0.7060931899641577,0.8009698371888624,0.4893617021276596 +0.7887794971466064,0.8,0.5428571428571428,0.8897849462365591,0.2,0.45299145299145294,0.8974358974358974,0.5555555555555556,0.8654985349045743,0.16230366492146608 +0.6461172103881836,0.825,0.6310935441370225,0.8853046594982079,0.175,0.5190058479532164,0.9078947368421053,0.6111111111111112,0.8633074449312262,0.306930693069307 +0.5030531883239746,0.875,0.7948717948717949,0.8853046594982079,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8619999696474241,0.5934959349593496 +0.5522456765174866,0.875,0.7948717948717949,0.8960573476702509,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8767294456672228,0.5934959349593496 +0.7561146914958954,0.825,0.6310935441370225,0.8996415770609318,0.175,0.5190058479532164,0.9078947368421053,0.6111111111111112,0.8802280109243729,0.306930693069307 +0.6209883987903595,0.875,0.7703788748564868,0.903225806451613,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.88248027742237,0.5535714285714286 +0.447883665561676,0.9,0.8268398268398269,0.9068100358422939,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.885489601760949,0.6595744680851063 +0.4562894105911255,0.875,0.7948717948717949,0.9068100358422939,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8799924757513359,0.5934959349593496 +0.5264500379562378,0.85,0.7058823529411764,0.9103942652329748,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.8845031038319028,0.43661971830985913 +0.47636842727661133,0.85,0.7402597402597402,0.913978494623656,0.15,0.6166666666666667,0.8285714285714285,0.7060931899641577,0.8875124281704819,0.4893617021276596 +0.337121844291687,0.875,0.8132586367880486,0.9103942652329748,0.125,0.6991978609625669,0.828125,0.8010752688172043,0.8740258283883144,0.6268656716417911 +0.3412959575653076,0.85,0.765625,0.9175627240143369,0.15,0.6415584415584416,0.7965367965367965,0.7455197132616487,0.8834935846112888,0.5330739299610895 +0.3782167434692383,0.85,0.765625,0.9175627240143369,0.15,0.6415584415584416,0.7965367965367965,0.7455197132616487,0.8834935846112889,0.5330739299610895 +0.39102625846862793,0.875,0.7948717948717949,0.9175627240143369,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8834935846112889,0.5934959349593496 +0.39682629704475403,0.875,0.7948717948717949,0.9175627240143369,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8858387009628654,0.5934959349593496 +0.3844609260559082,0.875,0.7948717948717949,0.9202508960573477,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8876905528147174,0.5934959349593496 +0.32598935067653656,0.875,0.8132586367880486,0.9283154121863799,0.125,0.6991978609625669,0.828125,0.8010752688172043,0.8969757547517221,0.6268656716417911 +0.30468180775642395,0.9,0.84375,0.931899641577061,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9084622412382085,0.688715953307393 +0.44658397138118744,0.875,0.7703788748564868,0.942652329749104,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9264611681903296,0.5535714285714286 +0.5571950972080231,0.85,0.7058823529411764,0.953405017921147,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.940212322774027,0.43661971830985913 +0.5383399128913879,0.85,0.7058823529411764,0.953405017921147,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.9402123227740269,0.43661971830985913 +0.46633538603782654,0.85,0.7058823529411764,0.9426523297491041,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.9323344476096241,0.43661971830985913 +0.4113213121891022,0.875,0.7703788748564868,0.9435483870967742,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9323344476096241,0.5535714285714286 +0.3553314805030823,0.9,0.8268398268398269,0.9426523297491041,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9323344476096241,0.6595744680851063 +0.3050922453403473,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096243,0.688715953307393 +0.28915050625801086,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096243,0.688715953307393 +0.28989332914352417,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096241,0.688715953307393 +0.3002821356058121,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096241,0.688715953307393 +0.3087248206138611,0.9,0.84375,0.9435483870967742,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096243,0.688715953307393 +0.31064677238464355,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096241,0.688715953307393 +0.3116318881511688,0.9,0.84375,0.9426523297491041,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.9323344476096243,0.688715953307393 +0.3142450749874115,0.925,0.8769230769230769,0.9435483870967742,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9323344476096243,0.7560975609756098 +0.3163122981786728,0.925,0.8769230769230769,0.9435483870967742,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9323344476096243,0.7560975609756098 +0.31609438359737396,0.925,0.8769230769230769,0.9444444444444444,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9323344476096243,0.7560975609756098 +0.3159617781639099,0.925,0.8769230769230769,0.9453405017921148,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.934172682903742,0.7560975609756098 +0.3158949166536331,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9348178441940644,0.7560975609756098 diff --git a/results/downsample/adam/050/retfound/pr.png b/results/downsample/adam/050/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..ee250dfca80155607c8f3e0b7117d7a8ea0c44b1 --- /dev/null +++ b/results/downsample/adam/050/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab0203335bb558ee33a06525912eaa266773e6dc9fad327b30c5dc2ad6ee0687 +size 41244 diff --git a/results/downsample/adam/050/retfound/roc.png b/results/downsample/adam/050/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..b92a460bf0e9c180a4e45f75338e9ec0485233aa --- /dev/null +++ b/results/downsample/adam/050/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:530ad11f9928ff1a836956547d309b85b27753f3532f80c11450c0f24a234da8 +size 56848 diff --git a/results/downsample/adam/050/retfound/test_pred.npz b/results/downsample/adam/050/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..a684a1a500c982bb1bfc9228e2a252edd1d74486 --- /dev/null +++ b/results/downsample/adam/050/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07b88803c8ce9da73a17475154da415cf1cc7e171cc375ac352c234673fb93f2 +size 1470 diff --git a/results/downsample/adam/050/retfound/train.log b/results/downsample/adam/050/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..066be768b5f90f4e8766202c93d0d62d929e11f0 --- /dev/null +++ b/results/downsample/adam/050/retfound/train.log @@ -0,0 +1,732 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:42:25.200321500 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:42:26.854596] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:42:26.854842] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/adam_50', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/050', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:42:41.214093] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:43.234907] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:47.838020] Sampler_train = +[14:42:48.045122] len of train_set: 128 +[14:42:48.334596] [Adaptation] Full fine-tuning: training all parameters. +[14:42:48.339767] number of trainable params (M): 303.30 +[14:42:48.343919] base lr: 5.00e-03 +[14:42:48.344026] actual lr: 6.25e-04 +[14:42:48.344074] accumulate grad iterations: 1 +[14:42:48.344121] effective batch size: 32 +[14:42:48.347680] criterion = CrossEntropyLoss() +[14:42:48.347774] Start training for 50 epochs +[14:42:48.350044] log_dir: ./output_logs/retfound +[14:42:50.838533] Epoch: [0] [0/4] eta: 0:00:09 lr: 0.000000 loss: 0.6926 (0.6926) time: 2.4876 data: 1.8809 max mem: 7340 +[14:42:51.204721] Epoch: [0] [3/4] eta: 0:00:00 lr: 0.000047 loss: 0.6926 (0.6916) time: 0.7132 data: 0.4703 max mem: 9671 +[14:42:51.264030] Epoch: [0] Total time: 0:00:02 (0.7285 s / it) +[14:42:51.265028] Averaged stats: lr: 0.000047 loss: 0.6926 (0.6916) +[14:42:53.466358] val: [0/2] eta: 0:00:04 loss: 0.6848 (0.6848) time: 2.1860 data: 2.1605 max mem: 9671 +[14:42:53.516051] val: [1/2] eta: 0:00:01 loss: 0.6848 (0.6932) time: 1.1176 data: 1.0803 max mem: 9671 +[14:42:53.585684] val: Total time: 0:00:02 (1.1529 s / it) +[14:42:53.597109] val loss: 0.6932296752929688 +[14:42:53.597304] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7097, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.6026, Kappa: 0.0000, Score: 0.3821 +[14:42:56.000940] Best epoch = 0, Best score = 0.3821 +[14:42:56.296645] log_dir: ./output_logs/retfound +[14:42:58.454689] Epoch: [1] [0/4] eta: 0:00:08 lr: 0.000063 loss: 0.6828 (0.6828) time: 2.1470 data: 1.9757 max mem: 9671 +[14:42:58.652342] Epoch: [1] [3/4] eta: 0:00:00 lr: 0.000109 loss: 0.6713 (0.6710) time: 0.5860 data: 0.4940 max mem: 9671 +[14:42:58.728850] Epoch: [1] Total time: 0:00:02 (0.6055 s / it) +[14:42:58.729563] Averaged stats: lr: 0.000109 loss: 0.6713 (0.6710) +[14:43:01.037354] val: [0/2] eta: 0:00:04 loss: 0.5991 (0.5991) time: 2.2966 data: 2.2797 max mem: 9671 +[14:43:01.047170] val: [1/2] eta: 0:00:01 loss: 0.5991 (0.6982) time: 1.1529 data: 1.1399 max mem: 9671 +[14:43:01.113096] val: Total time: 0:00:02 (1.1865 s / it) +[14:43:01.122372] val loss: 0.6982345581054688 +[14:43:01.122557] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7330, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7383, Kappa: 0.0000, Score: 0.3899 +[14:43:03.244141] Best epoch = 1, Best score = 0.3899 +[14:43:03.309150] log_dir: ./output_logs/retfound +[14:43:05.362315] Epoch: [2] [0/4] eta: 0:00:08 lr: 0.000125 loss: 0.6283 (0.6283) time: 2.0518 data: 1.9615 max mem: 9671 +[14:43:05.636789] Epoch: [2] [3/4] eta: 0:00:00 lr: 0.000172 loss: 0.5844 (0.5984) time: 0.5814 data: 0.5100 max mem: 9671 +[14:43:05.776543] Epoch: [2] Total time: 0:00:02 (0.6168 s / it) +[14:43:05.777498] Averaged stats: lr: 0.000172 loss: 0.5844 (0.5984) +[14:43:07.989837] val: [0/2] eta: 0:00:04 loss: 0.4654 (0.4654) time: 2.1962 data: 2.1791 max mem: 9671 +[14:43:07.999661] val: [1/2] eta: 0:00:01 loss: 0.4654 (0.7288) time: 1.1027 data: 1.0896 max mem: 9671 +[14:43:08.073769] val: Total time: 0:00:02 (1.1404 s / it) +[14:43:08.082722] val loss: 0.7288017272949219 +[14:43:08.082909] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7392, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7259, Kappa: 0.0000, Score: 0.3920 +[14:43:10.112673] Best epoch = 2, Best score = 0.3920 +[14:43:10.397250] log_dir: ./output_logs/retfound +[14:43:12.567397] Epoch: [3] [0/4] eta: 0:00:08 lr: 0.000188 loss: 0.4768 (0.4768) time: 2.1690 data: 2.0535 max mem: 9671 +[14:43:12.974549] Epoch: [3] [3/4] eta: 0:00:00 lr: 0.000234 loss: 0.4791 (0.5122) time: 0.6438 data: 0.5640 max mem: 9671 +[14:43:13.045139] Epoch: [3] Total time: 0:00:02 (0.6619 s / it) +[14:43:13.046402] Averaged stats: lr: 0.000234 loss: 0.4791 (0.5122) +[14:43:15.305210] val: [0/2] eta: 0:00:04 loss: 0.3223 (0.3223) time: 2.2429 data: 2.2259 max mem: 9671 +[14:43:15.314785] val: [1/2] eta: 0:00:01 loss: 0.3223 (0.8105) time: 1.1259 data: 1.1130 max mem: 9671 +[14:43:15.397580] val: Total time: 0:00:02 (1.1680 s / it) +[14:43:15.406922] val loss: 0.8104629516601562 +[14:43:15.407106] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7634, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7071, Kappa: 0.0000, Score: 0.4000 +[14:43:17.375953] Best epoch = 3, Best score = 0.4000 +[14:43:17.561928] log_dir: ./output_logs/retfound +[14:43:19.513409] Epoch: [4] [0/4] eta: 0:00:07 lr: 0.000250 loss: 0.4685 (0.4685) time: 1.9504 data: 1.8807 max mem: 9671 +[14:43:20.020671] Epoch: [4] [3/4] eta: 0:00:00 lr: 0.000297 loss: 0.4685 (0.5121) time: 0.6142 data: 0.5454 max mem: 9671 +[14:43:20.122008] Epoch: [4] Total time: 0:00:02 (0.6400 s / it) +[14:43:20.122738] Averaged stats: lr: 0.000297 loss: 0.4685 (0.5121) +[14:43:22.537208] val: [0/2] eta: 0:00:04 loss: 0.2409 (0.2409) time: 2.3985 data: 2.3814 max mem: 9671 +[14:43:22.547544] val: [1/2] eta: 0:00:01 loss: 0.2409 (0.8870) time: 1.2041 data: 1.1908 max mem: 9671 +[14:43:22.622493] val: Total time: 0:00:02 (1.2423 s / it) +[14:43:22.634219] val loss: 0.8870391845703125 +[14:43:22.634418] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8253, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7663, Kappa: 0.0000, Score: 0.4206 +[14:43:24.650048] Best epoch = 4, Best score = 0.4206 +[14:43:24.867741] log_dir: ./output_logs/retfound +[14:43:27.007401] Epoch: [5] [0/4] eta: 0:00:08 lr: 0.000313 loss: 0.4702 (0.4702) time: 2.1385 data: 2.0246 max mem: 9671 +[14:43:27.203568] Epoch: [5] [3/4] eta: 0:00:00 lr: 0.000359 loss: 0.5645 (0.5895) time: 0.5835 data: 0.5062 max mem: 9671 +[14:43:27.274240] Epoch: [5] Total time: 0:00:02 (0.6016 s / it) +[14:43:27.275017] Averaged stats: lr: 0.000359 loss: 0.5645 (0.5895) +[14:43:29.805300] val: [0/2] eta: 0:00:05 loss: 0.2534 (0.2534) time: 2.5076 data: 2.4737 max mem: 9671 +[14:43:29.820397] val: [1/2] eta: 0:00:01 loss: 0.2534 (0.8351) time: 1.2611 data: 1.2369 max mem: 9671 +[14:43:29.892678] val: Total time: 0:00:02 (1.2979 s / it) +[14:43:29.901689] val loss: 0.8351268768310547 +[14:43:29.901903] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8522, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7832, Kappa: 0.0000, Score: 0.4296 +[14:43:31.717522] Best epoch = 5, Best score = 0.4296 +[14:43:31.804642] log_dir: ./output_logs/retfound +[14:43:33.985502] Epoch: [6] [0/4] eta: 0:00:08 lr: 0.000375 loss: 0.3761 (0.3761) time: 2.1797 data: 2.1040 max mem: 9671 +[14:43:34.227225] Epoch: [6] [3/4] eta: 0:00:00 lr: 0.000422 loss: 0.4841 (0.4907) time: 0.6051 data: 0.5271 max mem: 9671 +[14:43:34.307006] Epoch: [6] Total time: 0:00:02 (0.6255 s / it) +[14:43:34.308044] Averaged stats: lr: 0.000422 loss: 0.4841 (0.4907) +[14:43:36.667221] val: [0/2] eta: 0:00:04 loss: 0.2822 (0.2822) time: 2.3435 data: 2.3267 max mem: 9671 +[14:43:36.676945] val: [1/2] eta: 0:00:01 loss: 0.2822 (0.7682) time: 1.1763 data: 1.1634 max mem: 9671 +[14:43:36.747461] val: Total time: 0:00:02 (1.2123 s / it) +[14:43:36.756652] val loss: 0.7682437896728516 +[14:43:36.756849] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8647, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7955, Kappa: 0.0000, Score: 0.4338 +[14:43:38.810097] Best epoch = 6, Best score = 0.4338 +[14:43:38.981331] log_dir: ./output_logs/retfound +[14:43:41.193358] Epoch: [7] [0/4] eta: 0:00:08 lr: 0.000438 loss: 0.3980 (0.3980) time: 2.2110 data: 2.1405 max mem: 9671 +[14:43:41.389995] Epoch: [7] [3/4] eta: 0:00:00 lr: 0.000484 loss: 0.4564 (0.4813) time: 0.6017 data: 0.5352 max mem: 9671 +[14:43:41.461843] Epoch: [7] Total time: 0:00:02 (0.6201 s / it) +[14:43:41.462889] Averaged stats: lr: 0.000484 loss: 0.4564 (0.4813) +[14:43:43.741191] val: [0/2] eta: 0:00:04 loss: 0.2704 (0.2704) time: 2.2635 data: 2.2466 max mem: 9671 +[14:43:43.750971] val: [1/2] eta: 0:00:01 loss: 0.2704 (0.7308) time: 1.1364 data: 1.1233 max mem: 9671 +[14:43:43.819185] val: Total time: 0:00:02 (1.1711 s / it) +[14:43:43.828195] val loss: 0.7307825088500977 +[14:43:43.828365] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8746, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8181, Kappa: 0.0000, Score: 0.4371 +[14:43:45.594577] Best epoch = 7, Best score = 0.4371 +[14:43:45.673693] log_dir: ./output_logs/retfound +[14:43:47.751058] Epoch: [8] [0/4] eta: 0:00:08 lr: 0.000500 loss: 0.4682 (0.4682) time: 2.0764 data: 2.0037 max mem: 9671 +[14:43:47.947247] Epoch: [8] [3/4] eta: 0:00:00 lr: 0.000547 loss: 0.5007 (0.4980) time: 0.5680 data: 0.5010 max mem: 9671 +[14:43:48.020101] Epoch: [8] Total time: 0:00:02 (0.5866 s / it) +[14:43:48.020833] Averaged stats: lr: 0.000547 loss: 0.5007 (0.4980) +[14:43:50.322182] val: [0/2] eta: 0:00:04 loss: 0.2104 (0.2104) time: 2.2863 data: 2.2689 max mem: 9671 +[14:43:50.333316] val: [1/2] eta: 0:00:01 loss: 0.2104 (0.7334) time: 1.1484 data: 1.1345 max mem: 9671 +[14:43:50.408775] val: Total time: 0:00:02 (1.1869 s / it) +[14:43:50.419014] val loss: 0.7333745956420898 +[14:43:50.419263] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8638, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7973, Kappa: 0.0000, Score: 0.4335 +[14:43:50.466070] Best epoch = 7, Best score = 0.4371 +[14:43:50.722087] log_dir: ./output_logs/retfound +[14:43:52.957338] Epoch: [9] [0/4] eta: 0:00:08 lr: 0.000562 loss: 0.4165 (0.4165) time: 2.2344 data: 2.1626 max mem: 9671 +[14:43:53.177636] Epoch: [9] [3/4] eta: 0:00:00 lr: 0.000609 loss: 0.4209 (0.4729) time: 0.6135 data: 0.5471 max mem: 9671 +[14:43:53.248952] Epoch: [9] Total time: 0:00:02 (0.6317 s / it) +[14:43:53.249768] Averaged stats: lr: 0.000609 loss: 0.4209 (0.4729) +[14:43:55.708035] val: [0/2] eta: 0:00:04 loss: 0.1643 (0.1643) time: 2.4471 data: 2.4290 max mem: 9671 +[14:43:55.719310] val: [1/2] eta: 0:00:01 loss: 0.1643 (0.7696) time: 1.2289 data: 1.2146 max mem: 9671 +[14:43:55.787267] val: Total time: 0:00:02 (1.2635 s / it) +[14:43:55.797542] val loss: 0.7696075439453125 +[14:43:55.797715] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8638, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7973, Kappa: 0.0000, Score: 0.4335 +[14:43:55.837481] Best epoch = 7, Best score = 0.4371 +[14:43:56.083694] log_dir: ./output_logs/retfound +[14:43:58.115787] Epoch: [10] [0/4] eta: 0:00:08 lr: 0.000625 loss: 0.4452 (0.4452) time: 2.0311 data: 1.9614 max mem: 9671 +[14:43:58.327399] Epoch: [10] [3/4] eta: 0:00:00 lr: 0.000624 loss: 0.4452 (0.4618) time: 0.5605 data: 0.4940 max mem: 9671 +[14:43:58.398486] Epoch: [10] Total time: 0:00:02 (0.5787 s / it) +[14:43:58.399287] Averaged stats: lr: 0.000624 loss: 0.4452 (0.4618) +[14:44:00.566158] val: [0/2] eta: 0:00:04 loss: 0.2169 (0.2169) time: 2.1549 data: 2.1381 max mem: 9671 +[14:44:00.575570] val: [1/2] eta: 0:00:01 loss: 0.2169 (0.5585) time: 1.0819 data: 1.0691 max mem: 9671 +[14:44:00.645421] val: Total time: 0:00:02 (1.1174 s / it) +[14:44:00.654305] val loss: 0.5585274696350098 +[14:44:00.654483] Accuracy: 0.7750, F1 Score: 0.5867, ROC AUC: 0.8638, Hamming Loss: 0.2250, + Jaccard Score: 0.4725, Precision: 0.6528, Recall: 0.5789, + Average Precision: 0.7901, Kappa: 0.1964, Score: 0.5490 +[14:44:02.328746] Best epoch = 10, Best score = 0.5490 +[14:44:02.396910] log_dir: ./output_logs/retfound +[14:44:04.540295] Epoch: [11] [0/4] eta: 0:00:08 lr: 0.000624 loss: 0.4455 (0.4455) time: 2.1425 data: 2.0702 max mem: 9671 +[14:44:04.747210] Epoch: [11] [3/4] eta: 0:00:00 lr: 0.000622 loss: 0.3801 (0.4050) time: 0.5872 data: 0.5176 max mem: 9671 +[14:44:04.814885] Epoch: [11] Total time: 0:00:02 (0.6045 s / it) +[14:44:04.815703] Averaged stats: lr: 0.000622 loss: 0.3801 (0.4050) +[14:44:07.120097] val: [0/2] eta: 0:00:04 loss: 0.3139 (0.3139) time: 2.2802 data: 2.2634 max mem: 9671 +[14:44:07.129392] val: [1/2] eta: 0:00:01 loss: 0.3139 (0.4624) time: 1.1445 data: 1.1318 max mem: 9671 +[14:44:07.200140] val: Total time: 0:00:02 (1.1805 s / it) +[14:44:07.208902] val loss: 0.4624495506286621 +[14:44:07.209069] Accuracy: 0.7500, F1 Score: 0.6865, ROC AUC: 0.8602, Hamming Loss: 0.2500, + Jaccard Score: 0.5404, Precision: 0.6752, Recall: 0.7204, + Average Precision: 0.7859, Kappa: 0.3808, Score: 0.6425 +[14:44:09.260806] Best epoch = 11, Best score = 0.6425 +[14:44:09.341941] log_dir: ./output_logs/retfound +[14:44:11.457249] Epoch: [12] [0/4] eta: 0:00:08 lr: 0.000621 loss: 0.3781 (0.3781) time: 2.1144 data: 2.0436 max mem: 9671 +[14:44:11.759619] Epoch: [12] [3/4] eta: 0:00:00 lr: 0.000618 loss: 0.3832 (0.3948) time: 0.6040 data: 0.5376 max mem: 9671 +[14:44:11.830780] Epoch: [12] Total time: 0:00:02 (0.6222 s / it) +[14:44:11.831526] Averaged stats: lr: 0.000618 loss: 0.3832 (0.3948) +[14:44:14.007372] val: [0/2] eta: 0:00:04 loss: 0.2492 (0.2492) time: 2.1595 data: 2.1426 max mem: 9671 +[14:44:14.016808] val: [1/2] eta: 0:00:01 loss: 0.2492 (0.5516) time: 1.0842 data: 1.0714 max mem: 9671 +[14:44:14.088836] val: Total time: 0:00:02 (1.1209 s / it) +[14:44:14.097771] val loss: 0.55164635181427 +[14:44:14.097956] Accuracy: 0.8000, F1 Score: 0.7133, ROC AUC: 0.8566, Hamming Loss: 0.2000, + Jaccard Score: 0.5780, Precision: 0.7133, Recall: 0.7133, + Average Precision: 0.7792, Kappa: 0.4265, Score: 0.6655 +[14:44:16.035920] Best epoch = 12, Best score = 0.6655 +[14:44:16.135278] log_dir: ./output_logs/retfound +[14:44:18.475120] Epoch: [13] [0/4] eta: 0:00:09 lr: 0.000616 loss: 0.5683 (0.5683) time: 2.3385 data: 2.2523 max mem: 9671 +[14:44:18.834612] Epoch: [13] [3/4] eta: 0:00:00 lr: 0.000612 loss: 0.3718 (0.4379) time: 0.6743 data: 0.5632 max mem: 9671 +[14:44:18.908375] Epoch: [13] Total time: 0:00:02 (0.6932 s / it) +[14:44:18.909219] Averaged stats: lr: 0.000612 loss: 0.3718 (0.4379) +[14:44:21.206362] val: [0/2] eta: 0:00:04 loss: 0.1853 (0.1853) time: 2.2820 data: 2.2652 max mem: 9671 +[14:44:21.215818] val: [1/2] eta: 0:00:01 loss: 0.1853 (0.6863) time: 1.1455 data: 1.1326 max mem: 9671 +[14:44:21.283702] val: Total time: 0:00:02 (1.1800 s / it) +[14:44:21.292687] val loss: 0.6862562894821167 +[14:44:21.292891] Accuracy: 0.8500, F1 Score: 0.7403, ROC AUC: 0.8638, Hamming Loss: 0.1500, + Jaccard Score: 0.6167, Precision: 0.8286, Recall: 0.7061, + Average Precision: 0.8010, Kappa: 0.4894, Score: 0.6978 +[14:44:23.087495] Best epoch = 13, Best score = 0.6978 +[14:44:23.151923] log_dir: ./output_logs/retfound +[14:44:25.192546] Epoch: [14] [0/4] eta: 0:00:08 lr: 0.000610 loss: 0.3714 (0.3714) time: 2.0396 data: 1.9689 max mem: 9671 +[14:44:25.469558] Epoch: [14] [3/4] eta: 0:00:00 lr: 0.000604 loss: 0.3714 (0.3957) time: 0.5790 data: 0.5126 max mem: 9671 +[14:44:25.540033] Epoch: [14] Total time: 0:00:02 (0.5970 s / it) +[14:44:25.540980] Averaged stats: lr: 0.000604 loss: 0.3714 (0.3957) +[14:44:27.695740] val: [0/2] eta: 0:00:04 loss: 0.1472 (0.1472) time: 2.1420 data: 2.1251 max mem: 9671 +[14:44:27.705007] val: [1/2] eta: 0:00:01 loss: 0.1472 (0.7888) time: 1.0753 data: 1.0626 max mem: 9671 +[14:44:27.778721] val: Total time: 0:00:02 (1.1129 s / it) +[14:44:27.787510] val loss: 0.7887794971466064 +[14:44:27.787689] Accuracy: 0.8000, F1 Score: 0.5429, ROC AUC: 0.8898, Hamming Loss: 0.2000, + Jaccard Score: 0.4530, Precision: 0.8974, Recall: 0.5556, + Average Precision: 0.8655, Kappa: 0.1623, Score: 0.5316 +[14:44:27.826999] Best epoch = 13, Best score = 0.6978 +[14:44:28.082650] log_dir: ./output_logs/retfound +[14:44:30.279036] Epoch: [15] [0/4] eta: 0:00:08 lr: 0.000601 loss: 0.3566 (0.3566) time: 2.1954 data: 2.1252 max mem: 9671 +[14:44:30.487897] Epoch: [15] [3/4] eta: 0:00:00 lr: 0.000594 loss: 0.3290 (0.3389) time: 0.6009 data: 0.5345 max mem: 9671 +[14:44:30.555984] Epoch: [15] Total time: 0:00:02 (0.6183 s / it) +[14:44:30.556678] Averaged stats: lr: 0.000594 loss: 0.3290 (0.3389) +[14:44:32.752984] val: [0/2] eta: 0:00:04 loss: 0.1695 (0.1695) time: 2.1849 data: 2.1677 max mem: 9671 +[14:44:32.762696] val: [1/2] eta: 0:00:01 loss: 0.1695 (0.6461) time: 1.0970 data: 1.0839 max mem: 9671 +[14:44:32.833633] val: Total time: 0:00:02 (1.1331 s / it) +[14:44:32.842484] val loss: 0.6461172103881836 +[14:44:32.842657] Accuracy: 0.8250, F1 Score: 0.6311, ROC AUC: 0.8853, Hamming Loss: 0.1750, + Jaccard Score: 0.5190, Precision: 0.9079, Recall: 0.6111, + Average Precision: 0.8633, Kappa: 0.3069, Score: 0.6078 +[14:44:32.877664] Best epoch = 13, Best score = 0.6978 +[14:44:33.139862] log_dir: ./output_logs/retfound +[14:44:35.200693] Epoch: [16] [0/4] eta: 0:00:08 lr: 0.000591 loss: 0.3439 (0.3439) time: 2.0599 data: 1.9908 max mem: 9671 +[14:44:35.502056] Epoch: [16] [3/4] eta: 0:00:00 lr: 0.000582 loss: 0.3440 (0.3624) time: 0.5901 data: 0.5243 max mem: 9671 +[14:44:35.574411] Epoch: [16] Total time: 0:00:02 (0.6086 s / it) +[14:44:35.575201] Averaged stats: lr: 0.000582 loss: 0.3440 (0.3624) +[14:44:37.709393] val: [0/2] eta: 0:00:04 loss: 0.2031 (0.2031) time: 2.1188 data: 2.1018 max mem: 9671 +[14:44:37.719050] val: [1/2] eta: 0:00:01 loss: 0.2031 (0.5031) time: 1.0639 data: 1.0510 max mem: 9671 +[14:44:37.785940] val: Total time: 0:00:02 (1.0980 s / it) +[14:44:37.794864] val loss: 0.5030531883239746 +[14:44:37.795046] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.8853, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8620, Kappa: 0.5935, Score: 0.7579 +[14:44:39.490885] Best epoch = 16, Best score = 0.7579 +[14:44:39.561302] log_dir: ./output_logs/retfound +[14:44:41.507083] Epoch: [17] [0/4] eta: 0:00:07 lr: 0.000579 loss: 0.2670 (0.2670) time: 1.9447 data: 1.8756 max mem: 9671 +[14:44:41.710361] Epoch: [17] [3/4] eta: 0:00:00 lr: 0.000569 loss: 0.2922 (0.3079) time: 0.5367 data: 0.4707 max mem: 9671 +[14:44:41.777247] Epoch: [17] Total time: 0:00:02 (0.5539 s / it) +[14:44:41.778113] Averaged stats: lr: 0.000569 loss: 0.2922 (0.3079) +[14:44:43.920758] val: [0/2] eta: 0:00:04 loss: 0.1707 (0.1707) time: 2.1307 data: 2.1125 max mem: 9671 +[14:44:43.930745] val: [1/2] eta: 0:00:01 loss: 0.1707 (0.5522) time: 1.0701 data: 1.0563 max mem: 9671 +[14:44:44.000558] val: Total time: 0:00:02 (1.1056 s / it) +[14:44:44.011950] val loss: 0.5522456765174866 +[14:44:44.012125] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.8961, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8767, Kappa: 0.5935, Score: 0.7615 +[14:44:46.037705] Best epoch = 17, Best score = 0.7615 +[14:44:46.110124] log_dir: ./output_logs/retfound +[14:44:47.987511] Epoch: [18] [0/4] eta: 0:00:07 lr: 0.000565 loss: 0.3370 (0.3370) time: 1.8764 data: 1.8053 max mem: 9671 +[14:44:48.332142] Epoch: [18] [3/4] eta: 0:00:00 lr: 0.000554 loss: 0.3402 (0.3651) time: 0.5550 data: 0.4882 max mem: 9671 +[14:44:48.409429] Epoch: [18] Total time: 0:00:02 (0.5748 s / it) +[14:44:48.410202] Averaged stats: lr: 0.000554 loss: 0.3402 (0.3651) +[14:44:50.567792] val: [0/2] eta: 0:00:04 loss: 0.1416 (0.1416) time: 2.1453 data: 2.1284 max mem: 9671 +[14:44:50.577403] val: [1/2] eta: 0:00:01 loss: 0.1416 (0.7561) time: 1.0772 data: 1.0642 max mem: 9671 +[14:44:50.649419] val: Total time: 0:00:02 (1.1138 s / it) +[14:44:50.658451] val loss: 0.7561146914958954 +[14:44:50.658642] Accuracy: 0.8250, F1 Score: 0.6311, ROC AUC: 0.8996, Hamming Loss: 0.1750, + Jaccard Score: 0.5190, Precision: 0.9079, Recall: 0.6111, + Average Precision: 0.8802, Kappa: 0.3069, Score: 0.6126 +[14:44:50.694709] Best epoch = 17, Best score = 0.7615 +[14:44:50.973417] log_dir: ./output_logs/retfound +[14:44:53.354653] Epoch: [19] [0/4] eta: 0:00:09 lr: 0.000550 loss: 0.6373 (0.6373) time: 2.3801 data: 2.3095 max mem: 9671 +[14:44:53.549775] Epoch: [19] [3/4] eta: 0:00:00 lr: 0.000538 loss: 0.4404 (0.4275) time: 0.6436 data: 0.5774 max mem: 9671 +[14:44:53.627486] Epoch: [19] Total time: 0:00:02 (0.6635 s / it) +[14:44:53.628290] Averaged stats: lr: 0.000538 loss: 0.4404 (0.4275) +[14:44:55.827058] val: [0/2] eta: 0:00:04 loss: 0.1415 (0.1415) time: 2.1872 data: 2.1701 max mem: 9671 +[14:44:55.836834] val: [1/2] eta: 0:00:01 loss: 0.1415 (0.6210) time: 1.0982 data: 1.0851 max mem: 9671 +[14:44:55.906993] val: Total time: 0:00:02 (1.1339 s / it) +[14:44:55.918824] val loss: 0.6209883987903595 +[14:44:55.919098] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9032, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.8825, Kappa: 0.5536, Score: 0.7424 +[14:44:55.958112] Best epoch = 17, Best score = 0.7615 +[14:44:56.225745] log_dir: ./output_logs/retfound +[14:44:58.469015] Epoch: [20] [0/4] eta: 0:00:08 lr: 0.000534 loss: 0.4004 (0.4004) time: 2.2422 data: 2.1715 max mem: 9671 +[14:44:58.664690] Epoch: [20] [3/4] eta: 0:00:00 lr: 0.000520 loss: 0.2475 (0.3064) time: 0.6093 data: 0.5429 max mem: 9671 +[14:44:58.739080] Epoch: [20] Total time: 0:00:02 (0.6283 s / it) +[14:44:58.739824] Averaged stats: lr: 0.000520 loss: 0.2475 (0.3064) +[14:45:01.005577] val: [0/2] eta: 0:00:04 loss: 0.1723 (0.1723) time: 2.2543 data: 2.2298 max mem: 9671 +[14:45:01.018877] val: [1/2] eta: 0:00:01 loss: 0.1723 (0.4479) time: 1.1334 data: 1.1150 max mem: 9671 +[14:45:01.089151] val: Total time: 0:00:02 (1.1694 s / it) +[14:45:01.098209] val loss: 0.447883665561676 +[14:45:01.098407] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9068, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8855, Kappa: 0.6596, Score: 0.7977 +[14:45:02.857823] Best epoch = 20, Best score = 0.7977 +[14:45:02.925659] log_dir: ./output_logs/retfound +[14:45:04.975537] Epoch: [21] [0/4] eta: 0:00:08 lr: 0.000516 loss: 0.4354 (0.4354) time: 2.0486 data: 1.9789 max mem: 9671 +[14:45:05.311809] Epoch: [21] [3/4] eta: 0:00:00 lr: 0.000501 loss: 0.3441 (0.3837) time: 0.5960 data: 0.5298 max mem: 9671 +[14:45:05.384912] Epoch: [21] Total time: 0:00:02 (0.6147 s / it) +[14:45:05.385691] Averaged stats: lr: 0.000501 loss: 0.3441 (0.3837) +[14:45:07.561238] val: [0/2] eta: 0:00:04 loss: 0.1764 (0.1764) time: 2.1561 data: 2.1384 max mem: 9671 +[14:45:07.571055] val: [1/2] eta: 0:00:01 loss: 0.1764 (0.4563) time: 1.0826 data: 1.0693 max mem: 9671 +[14:45:07.642133] val: Total time: 0:00:02 (1.1189 s / it) +[14:45:07.650901] val loss: 0.4562894105911255 +[14:45:07.651080] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.9068, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8800, Kappa: 0.5935, Score: 0.7651 +[14:45:07.688436] Best epoch = 20, Best score = 0.7977 +[14:45:07.950890] log_dir: ./output_logs/retfound +[14:45:10.145099] Epoch: [22] [0/4] eta: 0:00:08 lr: 0.000496 loss: 0.4153 (0.4153) time: 2.1933 data: 2.1240 max mem: 9671 +[14:45:10.341702] Epoch: [22] [3/4] eta: 0:00:00 lr: 0.000481 loss: 0.3878 (0.3516) time: 0.5973 data: 0.5311 max mem: 9671 +[14:45:10.413820] Epoch: [22] Total time: 0:00:02 (0.6157 s / it) +[14:45:10.414610] Averaged stats: lr: 0.000481 loss: 0.3878 (0.3516) +[14:45:12.547044] val: [0/2] eta: 0:00:04 loss: 0.1565 (0.1565) time: 2.1087 data: 2.0892 max mem: 9671 +[14:45:12.562650] val: [1/2] eta: 0:00:01 loss: 0.1565 (0.5265) time: 1.0618 data: 1.0447 max mem: 9671 +[14:45:12.631302] val: Total time: 0:00:02 (1.0970 s / it) +[14:45:12.644930] val loss: 0.5264500379562378 +[14:45:12.645171] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.9104, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.8845, Kappa: 0.4366, Score: 0.6843 +[14:45:12.682996] Best epoch = 20, Best score = 0.7977 +[14:45:12.967283] log_dir: ./output_logs/retfound +[14:45:14.821439] Epoch: [23] [0/4] eta: 0:00:07 lr: 0.000476 loss: 0.3927 (0.3927) time: 1.8532 data: 1.7816 max mem: 9671 +[14:45:15.244286] Epoch: [23] [3/4] eta: 0:00:00 lr: 0.000460 loss: 0.3269 (0.3979) time: 0.5688 data: 0.5019 max mem: 9671 +[14:45:15.318240] Epoch: [23] Total time: 0:00:02 (0.5877 s / it) +[14:45:15.319076] Averaged stats: lr: 0.000460 loss: 0.3269 (0.3979) +[14:45:17.462780] val: [0/2] eta: 0:00:04 loss: 0.1666 (0.1666) time: 2.1363 data: 2.1183 max mem: 9671 +[14:45:17.472267] val: [1/2] eta: 0:00:01 loss: 0.1666 (0.4764) time: 1.0726 data: 1.0592 max mem: 9671 +[14:45:17.545278] val: Total time: 0:00:02 (1.1097 s / it) +[14:45:17.554404] val loss: 0.47636842727661133 +[14:45:17.554572] Accuracy: 0.8500, F1 Score: 0.7403, ROC AUC: 0.9140, Hamming Loss: 0.1500, + Jaccard Score: 0.6167, Precision: 0.8286, Recall: 0.7061, + Average Precision: 0.8875, Kappa: 0.4894, Score: 0.7145 +[14:45:17.600043] Best epoch = 20, Best score = 0.7977 +[14:45:17.848955] log_dir: ./output_logs/retfound +[14:45:20.005558] Epoch: [24] [0/4] eta: 0:00:08 lr: 0.000455 loss: 0.5486 (0.5486) time: 2.1554 data: 2.0726 max mem: 9671 +[14:45:20.207010] Epoch: [24] [3/4] eta: 0:00:00 lr: 0.000438 loss: 0.2775 (0.3352) time: 0.5890 data: 0.5182 max mem: 9671 +[14:45:20.279596] Epoch: [24] Total time: 0:00:02 (0.6076 s / it) +[14:45:20.281274] Averaged stats: lr: 0.000438 loss: 0.2775 (0.3352) +[14:45:22.527634] val: [0/2] eta: 0:00:04 loss: 0.2953 (0.2953) time: 2.2213 data: 2.2021 max mem: 9671 +[14:45:22.542449] val: [1/2] eta: 0:00:01 loss: 0.2953 (0.3371) time: 1.1177 data: 1.1011 max mem: 9671 +[14:45:22.613392] val: Total time: 0:00:02 (1.1540 s / it) +[14:45:22.622348] val loss: 0.337121844291687 +[14:45:22.622525] Accuracy: 0.8750, F1 Score: 0.8133, ROC AUC: 0.9104, Hamming Loss: 0.1250, + Jaccard Score: 0.6992, Precision: 0.8281, Recall: 0.8011, + Average Precision: 0.8740, Kappa: 0.6269, Score: 0.7835 +[14:45:22.661046] Best epoch = 20, Best score = 0.7977 +[14:45:22.901063] log_dir: ./output_logs/retfound +[14:45:25.098115] Epoch: [25] [0/4] eta: 0:00:08 lr: 0.000432 loss: 0.3422 (0.3422) time: 2.1956 data: 2.1204 max mem: 9671 +[14:45:25.300854] Epoch: [25] [3/4] eta: 0:00:00 lr: 0.000415 loss: 0.3422 (0.3594) time: 0.5994 data: 0.5302 max mem: 9671 +[14:45:25.371898] Epoch: [25] Total time: 0:00:02 (0.6177 s / it) +[14:45:25.378521] Averaged stats: lr: 0.000415 loss: 0.3422 (0.3594) +[14:45:27.634990] val: [0/2] eta: 0:00:04 loss: 0.2672 (0.2672) time: 2.2440 data: 2.2218 max mem: 9671 +[14:45:27.648793] val: [1/2] eta: 0:00:01 loss: 0.2672 (0.3413) time: 1.1285 data: 1.1110 max mem: 9671 +[14:45:27.717682] val: Total time: 0:00:02 (1.1641 s / it) +[14:45:27.727037] val loss: 0.3412959575653076 +[14:45:27.727228] Accuracy: 0.8500, F1 Score: 0.7656, ROC AUC: 0.9176, Hamming Loss: 0.1500, + Jaccard Score: 0.6416, Precision: 0.7965, Recall: 0.7455, + Average Precision: 0.8835, Kappa: 0.5331, Score: 0.7388 +[14:45:27.764526] Best epoch = 20, Best score = 0.7977 +[14:45:28.036913] log_dir: ./output_logs/retfound +[14:45:30.046498] Epoch: [26] [0/4] eta: 0:00:08 lr: 0.000409 loss: 0.2011 (0.2011) time: 2.0087 data: 1.9312 max mem: 9671 +[14:45:30.314740] Epoch: [26] [3/4] eta: 0:00:00 lr: 0.000392 loss: 0.3076 (0.3457) time: 0.5690 data: 0.5004 max mem: 9671 +[14:45:30.387015] Epoch: [26] Total time: 0:00:02 (0.5875 s / it) +[14:45:30.387922] Averaged stats: lr: 0.000392 loss: 0.3076 (0.3457) +[14:45:32.515910] val: [0/2] eta: 0:00:04 loss: 0.2107 (0.2107) time: 2.1170 data: 2.0916 max mem: 9671 +[14:45:32.525717] val: [1/2] eta: 0:00:01 loss: 0.2107 (0.3782) time: 1.0631 data: 1.0459 max mem: 9671 +[14:45:32.600833] val: Total time: 0:00:02 (1.1013 s / it) +[14:45:32.609943] val loss: 0.3782167434692383 +[14:45:32.610190] Accuracy: 0.8500, F1 Score: 0.7656, ROC AUC: 0.9176, Hamming Loss: 0.1500, + Jaccard Score: 0.6416, Precision: 0.7965, Recall: 0.7455, + Average Precision: 0.8835, Kappa: 0.5331, Score: 0.7388 +[14:45:32.645051] Best epoch = 20, Best score = 0.7977 +[14:45:32.912384] log_dir: ./output_logs/retfound +[14:45:35.087631] Epoch: [27] [0/4] eta: 0:00:08 lr: 0.000386 loss: 0.3162 (0.3162) time: 2.1742 data: 2.1039 max mem: 9671 +[14:45:35.286414] Epoch: [27] [3/4] eta: 0:00:00 lr: 0.000368 loss: 0.3162 (0.2980) time: 0.5931 data: 0.5260 max mem: 9671 +[14:45:35.361224] Epoch: [27] Total time: 0:00:02 (0.6122 s / it) +[14:45:35.362041] Averaged stats: lr: 0.000368 loss: 0.3162 (0.2980) +[14:45:37.547448] val: [0/2] eta: 0:00:04 loss: 0.1918 (0.1918) time: 2.1764 data: 2.1595 max mem: 9671 +[14:45:37.557040] val: [1/2] eta: 0:00:01 loss: 0.1918 (0.3910) time: 1.0927 data: 1.0798 max mem: 9671 +[14:45:37.624821] val: Total time: 0:00:02 (1.1272 s / it) +[14:45:37.633641] val loss: 0.39102625846862793 +[14:45:37.633822] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.9176, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8835, Kappa: 0.5935, Score: 0.7686 +[14:45:37.673922] Best epoch = 20, Best score = 0.7977 +[14:45:37.939139] log_dir: ./output_logs/retfound +[14:45:40.269274] Epoch: [28] [0/4] eta: 0:00:09 lr: 0.000362 loss: 0.4803 (0.4803) time: 2.3292 data: 2.2600 max mem: 9671 +[14:45:40.464938] Epoch: [28] [3/4] eta: 0:00:00 lr: 0.000344 loss: 0.1406 (0.2961) time: 0.6310 data: 0.5650 max mem: 9671 +[14:45:40.535884] Epoch: [28] Total time: 0:00:02 (0.6491 s / it) +[14:45:40.536623] Averaged stats: lr: 0.000344 loss: 0.1406 (0.2961) +[14:45:42.672007] val: [0/2] eta: 0:00:04 loss: 0.1793 (0.1793) time: 2.1243 data: 2.1072 max mem: 9671 +[14:45:42.682290] val: [1/2] eta: 0:00:01 loss: 0.1793 (0.3968) time: 1.0670 data: 1.0537 max mem: 9671 +[14:45:42.764186] val: Total time: 0:00:02 (1.1086 s / it) +[14:45:42.773175] val loss: 0.39682629704475403 +[14:45:42.773346] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.9176, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8858, Kappa: 0.5935, Score: 0.7686 +[14:45:42.811083] Best epoch = 20, Best score = 0.7977 +[14:45:43.059803] log_dir: ./output_logs/retfound +[14:45:45.209917] Epoch: [29] [0/4] eta: 0:00:08 lr: 0.000337 loss: 0.2275 (0.2275) time: 2.1492 data: 2.0784 max mem: 9671 +[14:45:45.447415] Epoch: [29] [3/4] eta: 0:00:00 lr: 0.000319 loss: 0.2275 (0.3003) time: 0.5965 data: 0.5301 max mem: 9671 +[14:45:45.521592] Epoch: [29] Total time: 0:00:02 (0.6154 s / it) +[14:45:45.522321] Averaged stats: lr: 0.000319 loss: 0.2275 (0.3003) +[14:45:47.607316] val: [0/2] eta: 0:00:04 loss: 0.1852 (0.1852) time: 2.0735 data: 2.0562 max mem: 9671 +[14:45:47.617158] val: [1/2] eta: 0:00:01 loss: 0.1852 (0.3845) time: 1.0414 data: 1.0282 max mem: 9671 +[14:45:47.691682] val: Total time: 0:00:02 (1.0793 s / it) +[14:45:47.700774] val loss: 0.3844609260559082 +[14:45:47.700988] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.9203, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8877, Kappa: 0.5935, Score: 0.7695 +[14:45:47.742560] Best epoch = 20, Best score = 0.7977 +[14:45:48.007771] log_dir: ./output_logs/retfound +[14:45:50.305508] Epoch: [30] [0/4] eta: 0:00:09 lr: 0.000313 loss: 0.4005 (0.4005) time: 2.2968 data: 2.2274 max mem: 9671 +[14:45:50.501749] Epoch: [30] [3/4] eta: 0:00:00 lr: 0.000295 loss: 0.3263 (0.3299) time: 0.6231 data: 0.5569 max mem: 9671 +[14:45:50.571912] Epoch: [30] Total time: 0:00:02 (0.6410 s / it) +[14:45:50.572724] Averaged stats: lr: 0.000295 loss: 0.3263 (0.3299) +[14:45:52.750565] val: [0/2] eta: 0:00:04 loss: 0.2126 (0.2126) time: 2.1663 data: 2.1492 max mem: 9671 +[14:45:52.760253] val: [1/2] eta: 0:00:01 loss: 0.2126 (0.3260) time: 1.0877 data: 1.0747 max mem: 9671 +[14:45:52.828093] val: Total time: 0:00:02 (1.1223 s / it) +[14:45:52.837148] val loss: 0.32598935067653656 +[14:45:52.837355] Accuracy: 0.8750, F1 Score: 0.8133, ROC AUC: 0.9283, Hamming Loss: 0.1250, + Jaccard Score: 0.6992, Precision: 0.8281, Recall: 0.8011, + Average Precision: 0.8970, Kappa: 0.6269, Score: 0.7895 +[14:45:52.885912] Best epoch = 20, Best score = 0.7977 +[14:45:53.130243] log_dir: ./output_logs/retfound +[14:45:55.343631] Epoch: [31] [0/4] eta: 0:00:08 lr: 0.000289 loss: 0.4808 (0.4808) time: 2.2124 data: 2.1377 max mem: 9671 +[14:45:55.670124] Epoch: [31] [3/4] eta: 0:00:00 lr: 0.000270 loss: 0.2838 (0.3516) time: 0.6345 data: 0.5345 max mem: 9671 +[14:45:55.738166] Epoch: [31] Total time: 0:00:02 (0.6519 s / it) +[14:45:55.739088] Averaged stats: lr: 0.000270 loss: 0.2838 (0.3516) +[14:45:57.862499] val: [0/2] eta: 0:00:04 loss: 0.1971 (0.1971) time: 2.1120 data: 2.0929 max mem: 9671 +[14:45:57.877753] val: [1/2] eta: 0:00:01 loss: 0.1971 (0.3047) time: 1.0633 data: 1.0465 max mem: 9671 +[14:45:57.950357] val: Total time: 0:00:02 (1.1004 s / it) +[14:45:57.964055] val loss: 0.30468180775642395 +[14:45:57.964277] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9319, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9085, Kappa: 0.6887, Score: 0.8215 +[14:45:59.642582] Best epoch = 31, Best score = 0.8215 +[14:45:59.706962] log_dir: ./output_logs/retfound +[14:46:01.886309] Epoch: [32] [0/4] eta: 0:00:08 lr: 0.000264 loss: 0.3991 (0.3991) time: 2.1784 data: 2.1090 max mem: 9671 +[14:46:02.082081] Epoch: [32] [3/4] eta: 0:00:00 lr: 0.000246 loss: 0.3237 (0.3470) time: 0.5934 data: 0.5273 max mem: 9671 +[14:46:02.153210] Epoch: [32] Total time: 0:00:02 (0.6115 s / it) +[14:46:02.154113] Averaged stats: lr: 0.000246 loss: 0.3237 (0.3470) +[14:46:04.353203] val: [0/2] eta: 0:00:04 loss: 0.1240 (0.1240) time: 2.1873 data: 2.1702 max mem: 9671 +[14:46:04.373449] val: [1/2] eta: 0:00:01 loss: 0.1240 (0.4466) time: 1.0983 data: 1.0852 max mem: 9671 +[14:46:04.771899] val: Total time: 0:00:02 (1.3033 s / it) +[14:46:04.791164] val loss: 0.44658397138118744 +[14:46:04.791377] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9427, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9265, Kappa: 0.5536, Score: 0.7555 +[14:46:05.157825] Best epoch = 31, Best score = 0.8215 +[14:46:06.284077] log_dir: ./output_logs/retfound +[14:46:08.404149] Epoch: [33] [0/4] eta: 0:00:08 lr: 0.000240 loss: 0.1415 (0.1415) time: 2.1191 data: 2.0467 max mem: 9671 +[14:46:08.600353] Epoch: [33] [3/4] eta: 0:00:00 lr: 0.000222 loss: 0.2803 (0.3307) time: 0.5786 data: 0.5118 max mem: 9671 +[14:46:08.716491] Epoch: [33] Total time: 0:00:02 (0.6081 s / it) +[14:46:08.717294] Averaged stats: lr: 0.000222 loss: 0.2803 (0.3307) +[14:46:10.898689] val: [0/2] eta: 0:00:04 loss: 0.1132 (0.1132) time: 2.1685 data: 2.1472 max mem: 9671 +[14:46:10.911071] val: [1/2] eta: 0:00:01 loss: 0.1132 (0.5572) time: 1.0899 data: 1.0737 max mem: 9671 +[14:46:10.982957] val: Total time: 0:00:02 (1.1268 s / it) +[14:46:10.991873] val loss: 0.5571950972080231 +[14:46:10.992069] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.9534, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.9402, Kappa: 0.4366, Score: 0.6986 +[14:46:11.026475] Best epoch = 31, Best score = 0.8215 +[14:46:11.303078] log_dir: ./output_logs/retfound +[14:46:13.427755] Epoch: [34] [0/4] eta: 0:00:08 lr: 0.000217 loss: 0.2111 (0.2111) time: 2.1238 data: 2.0542 max mem: 9671 +[14:46:13.628287] Epoch: [34] [3/4] eta: 0:00:00 lr: 0.000199 loss: 0.2774 (0.3608) time: 0.5809 data: 0.5136 max mem: 9671 +[14:46:13.721765] Epoch: [34] Total time: 0:00:02 (0.6046 s / it) +[14:46:13.722541] Averaged stats: lr: 0.000199 loss: 0.2774 (0.3608) +[14:46:15.810600] val: [0/2] eta: 0:00:04 loss: 0.1126 (0.1126) time: 2.0760 data: 2.0589 max mem: 9671 +[14:46:15.820435] val: [1/2] eta: 0:00:01 loss: 0.1126 (0.5383) time: 1.0426 data: 1.0295 max mem: 9671 +[14:46:15.890747] val: Total time: 0:00:02 (1.0785 s / it) +[14:46:15.900565] val loss: 0.5383399128913879 +[14:46:15.900745] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.9534, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.9402, Kappa: 0.4366, Score: 0.6986 +[14:46:15.939682] Best epoch = 31, Best score = 0.8215 +[14:46:16.191097] log_dir: ./output_logs/retfound +[14:46:18.257053] Epoch: [35] [0/4] eta: 0:00:08 lr: 0.000194 loss: 0.1828 (0.1828) time: 2.0647 data: 1.9216 max mem: 9671 +[14:46:18.556124] Epoch: [35] [3/4] eta: 0:00:00 lr: 0.000177 loss: 0.2777 (0.2946) time: 0.5908 data: 0.5011 max mem: 9671 +[14:46:18.664031] Epoch: [35] Total time: 0:00:02 (0.6182 s / it) +[14:46:18.664872] Averaged stats: lr: 0.000177 loss: 0.2777 (0.2946) +[14:46:20.879859] val: [0/2] eta: 0:00:04 loss: 0.1185 (0.1185) time: 2.2067 data: 2.1865 max mem: 9671 +[14:46:20.894177] val: [1/2] eta: 0:00:01 loss: 0.1185 (0.4663) time: 1.1101 data: 1.0934 max mem: 9671 +[14:46:20.974779] val: Total time: 0:00:02 (1.1513 s / it) +[14:46:20.983833] val loss: 0.46633538603782654 +[14:46:20.984039] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.9427, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.9323, Kappa: 0.4366, Score: 0.6951 +[14:46:21.016741] Best epoch = 31, Best score = 0.8215 +[14:46:21.286866] log_dir: ./output_logs/retfound +[14:46:23.463039] Epoch: [36] [0/4] eta: 0:00:08 lr: 0.000171 loss: 0.1667 (0.1667) time: 2.1749 data: 2.0966 max mem: 9671 +[14:46:23.660045] Epoch: [36] [3/4] eta: 0:00:00 lr: 0.000155 loss: 0.2640 (0.2728) time: 0.5928 data: 0.5242 max mem: 9671 +[14:46:23.730152] Epoch: [36] Total time: 0:00:02 (0.6108 s / it) +[14:46:23.731043] Averaged stats: lr: 0.000155 loss: 0.2640 (0.2728) +[14:46:25.950014] val: [0/2] eta: 0:00:04 loss: 0.1266 (0.1266) time: 2.2075 data: 2.1905 max mem: 9671 +[14:46:25.966102] val: [1/2] eta: 0:00:01 loss: 0.1266 (0.4113) time: 1.1115 data: 1.0953 max mem: 9671 +[14:46:26.036745] val: Total time: 0:00:02 (1.1475 s / it) +[14:46:26.045676] val loss: 0.4113213121891022 +[14:46:26.045867] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9435, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9323, Kappa: 0.5536, Score: 0.7558 +[14:46:26.090298] Best epoch = 31, Best score = 0.8215 +[14:46:26.341159] log_dir: ./output_logs/retfound +[14:46:28.831379] Epoch: [37] [0/4] eta: 0:00:09 lr: 0.000150 loss: 0.1982 (0.1982) time: 2.4892 data: 2.4147 max mem: 9671 +[14:46:29.025761] Epoch: [37] [3/4] eta: 0:00:00 lr: 0.000135 loss: 0.2552 (0.2497) time: 0.6707 data: 0.6037 max mem: 9671 +[14:46:29.104336] Epoch: [37] Total time: 0:00:02 (0.6908 s / it) +[14:46:29.105389] Averaged stats: lr: 0.000135 loss: 0.2552 (0.2497) +[14:46:31.362296] val: [0/2] eta: 0:00:04 loss: 0.1392 (0.1392) time: 2.2443 data: 2.2245 max mem: 9671 +[14:46:31.371759] val: [1/2] eta: 0:00:01 loss: 0.1392 (0.3553) time: 1.1266 data: 1.1123 max mem: 9671 +[14:46:31.441689] val: Total time: 0:00:02 (1.1622 s / it) +[14:46:31.450552] val loss: 0.3553314805030823 +[14:46:31.450738] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9323, Kappa: 0.6596, Score: 0.8097 +[14:46:31.485935] Best epoch = 31, Best score = 0.8215 +[14:46:31.728664] log_dir: ./output_logs/retfound +[14:46:33.919141] Epoch: [38] [0/4] eta: 0:00:08 lr: 0.000130 loss: 0.3299 (0.3299) time: 2.1896 data: 2.1112 max mem: 9671 +[14:46:34.153385] Epoch: [38] [3/4] eta: 0:00:00 lr: 0.000115 loss: 0.3034 (0.3009) time: 0.6058 data: 0.5341 max mem: 9671 +[14:46:34.219031] Epoch: [38] Total time: 0:00:02 (0.6225 s / it) +[14:46:34.219816] Averaged stats: lr: 0.000115 loss: 0.3034 (0.3009) +[14:46:36.453301] val: [0/2] eta: 0:00:04 loss: 0.1592 (0.1592) time: 2.2263 data: 2.2094 max mem: 9671 +[14:46:36.462652] val: [1/2] eta: 0:00:01 loss: 0.1592 (0.3051) time: 1.1176 data: 1.1048 max mem: 9671 +[14:46:36.529088] val: Total time: 0:00:02 (1.1514 s / it) +[14:46:36.538166] val loss: 0.3050922453403473 +[14:46:36.538348] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:46:38.200019] Best epoch = 38, Best score = 0.8250 +[14:46:38.266459] log_dir: ./output_logs/retfound +[14:46:40.716470] Epoch: [39] [0/4] eta: 0:00:09 lr: 0.000110 loss: 0.3352 (0.3352) time: 2.4490 data: 2.3120 max mem: 9671 +[14:46:40.984736] Epoch: [39] [3/4] eta: 0:00:00 lr: 0.000097 loss: 0.2744 (0.3267) time: 0.6791 data: 0.5781 max mem: 9671 +[14:46:41.075615] Epoch: [39] Total time: 0:00:02 (0.7022 s / it) +[14:46:41.076418] Averaged stats: lr: 0.000097 loss: 0.2744 (0.3267) +[14:46:43.310765] val: [0/2] eta: 0:00:04 loss: 0.1645 (0.1645) time: 2.2227 data: 2.2057 max mem: 9671 +[14:46:43.320533] val: [1/2] eta: 0:00:01 loss: 0.1645 (0.2892) time: 1.1159 data: 1.1029 max mem: 9671 +[14:46:43.391751] val: Total time: 0:00:02 (1.1522 s / it) +[14:46:43.400512] val loss: 0.28915050625801086 +[14:46:43.400675] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:46:43.447460] Best epoch = 38, Best score = 0.8250 +[14:46:43.702566] log_dir: ./output_logs/retfound +[14:46:45.897273] Epoch: [40] [0/4] eta: 0:00:08 lr: 0.000092 loss: 0.3106 (0.3106) time: 2.1938 data: 2.1236 max mem: 9671 +[14:46:46.092409] Epoch: [40] [3/4] eta: 0:00:00 lr: 0.000080 loss: 0.2037 (0.2467) time: 0.5971 data: 0.5309 max mem: 9671 +[14:46:46.162555] Epoch: [40] Total time: 0:00:02 (0.6150 s / it) +[14:46:46.163318] Averaged stats: lr: 0.000080 loss: 0.2037 (0.2467) +[14:46:48.297366] val: [0/2] eta: 0:00:04 loss: 0.1596 (0.1596) time: 2.1231 data: 2.0984 max mem: 9671 +[14:46:48.310615] val: [1/2] eta: 0:00:01 loss: 0.1596 (0.2899) time: 1.0678 data: 1.0493 max mem: 9671 +[14:46:48.383561] val: Total time: 0:00:02 (1.1051 s / it) +[14:46:48.397137] val loss: 0.28989332914352417 +[14:46:48.397323] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:46:48.432201] Best epoch = 38, Best score = 0.8250 +[14:46:48.666967] log_dir: ./output_logs/retfound +[14:46:50.731162] Epoch: [41] [0/4] eta: 0:00:08 lr: 0.000076 loss: 0.1380 (0.1380) time: 2.0633 data: 1.9903 max mem: 9671 +[14:46:50.955238] Epoch: [41] [3/4] eta: 0:00:00 lr: 0.000064 loss: 0.2201 (0.2131) time: 0.5717 data: 0.5047 max mem: 9671 +[14:46:51.125442] Epoch: [41] Total time: 0:00:02 (0.6146 s / it) +[14:46:51.126324] Averaged stats: lr: 0.000064 loss: 0.2201 (0.2131) +[14:46:53.399303] val: [0/2] eta: 0:00:04 loss: 0.1499 (0.1499) time: 2.2615 data: 2.2443 max mem: 9671 +[14:46:53.409205] val: [1/2] eta: 0:00:01 loss: 0.1499 (0.3003) time: 1.1354 data: 1.1222 max mem: 9671 +[14:46:53.476656] val: Total time: 0:00:02 (1.1698 s / it) +[14:46:53.486140] val loss: 0.3002821356058121 +[14:46:53.486315] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:46:53.524571] Best epoch = 38, Best score = 0.8250 +[14:46:53.779228] log_dir: ./output_logs/retfound +[14:46:56.020427] Epoch: [42] [0/4] eta: 0:00:08 lr: 0.000061 loss: 0.2897 (0.2897) time: 2.2403 data: 2.1678 max mem: 9671 +[14:46:56.216406] Epoch: [42] [3/4] eta: 0:00:00 lr: 0.000050 loss: 0.2427 (0.2563) time: 0.6089 data: 0.5420 max mem: 9671 +[14:46:56.286875] Epoch: [42] Total time: 0:00:02 (0.6269 s / it) +[14:46:56.287854] Averaged stats: lr: 0.000050 loss: 0.2427 (0.2563) +[14:46:58.400325] val: [0/2] eta: 0:00:04 loss: 0.1438 (0.1438) time: 2.1004 data: 2.0836 max mem: 9671 +[14:46:58.409872] val: [1/2] eta: 0:00:01 loss: 0.1438 (0.3087) time: 1.0547 data: 1.0419 max mem: 9671 +[14:46:58.476524] val: Total time: 0:00:02 (1.0887 s / it) +[14:46:58.485477] val loss: 0.3087248206138611 +[14:46:58.485651] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9435, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8253 +[14:47:00.455467] Best epoch = 42, Best score = 0.8253 +[14:47:00.524574] log_dir: ./output_logs/retfound +[14:47:02.760018] Epoch: [43] [0/4] eta: 0:00:08 lr: 0.000047 loss: 0.4944 (0.4944) time: 2.2345 data: 2.1632 max mem: 9671 +[14:47:02.956148] Epoch: [43] [3/4] eta: 0:00:00 lr: 0.000038 loss: 0.2747 (0.3341) time: 0.6075 data: 0.5409 max mem: 9671 +[14:47:03.028521] Epoch: [43] Total time: 0:00:02 (0.6259 s / it) +[14:47:03.029338] Averaged stats: lr: 0.000038 loss: 0.2747 (0.3341) +[14:47:05.150063] val: [0/2] eta: 0:00:04 loss: 0.1420 (0.1420) time: 2.1101 data: 2.0927 max mem: 9671 +[14:47:05.160333] val: [1/2] eta: 0:00:01 loss: 0.1420 (0.3106) time: 1.0598 data: 1.0464 max mem: 9671 +[14:47:05.235857] val: Total time: 0:00:02 (1.0983 s / it) +[14:47:05.246641] val loss: 0.31064677238464355 +[14:47:05.246867] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:47:05.284661] Best epoch = 42, Best score = 0.8253 +[14:47:05.563034] log_dir: ./output_logs/retfound +[14:47:07.776119] Epoch: [44] [0/4] eta: 0:00:08 lr: 0.000035 loss: 0.2197 (0.2197) time: 2.2120 data: 2.1375 max mem: 9671 +[14:47:07.971594] Epoch: [44] [3/4] eta: 0:00:00 lr: 0.000027 loss: 0.2163 (0.2396) time: 0.6017 data: 0.5344 max mem: 9671 +[14:47:08.041706] Epoch: [44] Total time: 0:00:02 (0.6196 s / it) +[14:47:08.042507] Averaged stats: lr: 0.000027 loss: 0.2163 (0.2396) +[14:47:10.238771] val: [0/2] eta: 0:00:04 loss: 0.1412 (0.1412) time: 2.1769 data: 2.1589 max mem: 9671 +[14:47:10.250466] val: [1/2] eta: 0:00:01 loss: 0.1412 (0.3116) time: 1.0940 data: 1.0795 max mem: 9671 +[14:47:10.327961] val: Total time: 0:00:02 (1.1334 s / it) +[14:47:10.336855] val loss: 0.3116318881511688 +[14:47:10.337075] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9427, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9323, Kappa: 0.6887, Score: 0.8250 +[14:47:10.375045] Best epoch = 42, Best score = 0.8253 +[14:47:10.623698] log_dir: ./output_logs/retfound +[14:47:12.823497] Epoch: [45] [0/4] eta: 0:00:08 lr: 0.000025 loss: 0.3436 (0.3436) time: 2.1986 data: 2.1268 max mem: 9671 +[14:47:13.023670] Epoch: [45] [3/4] eta: 0:00:00 lr: 0.000018 loss: 0.2587 (0.3255) time: 0.5995 data: 0.5318 max mem: 9671 +[14:47:13.098554] Epoch: [45] Total time: 0:00:02 (0.6187 s / it) +[14:47:13.099330] Averaged stats: lr: 0.000018 loss: 0.2587 (0.3255) +[14:47:15.402245] val: [0/2] eta: 0:00:04 loss: 0.1397 (0.1397) time: 2.2910 data: 2.2730 max mem: 9671 +[14:47:15.414242] val: [1/2] eta: 0:00:01 loss: 0.1397 (0.3142) time: 1.1512 data: 1.1366 max mem: 9671 +[14:47:15.495171] val: Total time: 0:00:02 (1.1925 s / it) +[14:47:15.505184] val loss: 0.3142450749874115 +[14:47:15.505380] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9435, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9323, Kappa: 0.7561, Score: 0.8589 +[14:47:17.577503] Best epoch = 45, Best score = 0.8589 +[14:47:17.791473] log_dir: ./output_logs/retfound +[14:47:20.128635] Epoch: [46] [0/4] eta: 0:00:09 lr: 0.000016 loss: 0.3239 (0.3239) time: 2.3362 data: 2.2666 max mem: 9671 +[14:47:20.324282] Epoch: [46] [3/4] eta: 0:00:00 lr: 0.000011 loss: 0.2773 (0.2625) time: 0.6328 data: 0.5667 max mem: 9671 +[14:47:20.390640] Epoch: [46] Total time: 0:00:02 (0.6497 s / it) +[14:47:20.391384] Averaged stats: lr: 0.000011 loss: 0.2773 (0.2625) +[14:47:22.547990] val: [0/2] eta: 0:00:04 loss: 0.1387 (0.1387) time: 2.1444 data: 2.1274 max mem: 9671 +[14:47:22.557969] val: [1/2] eta: 0:00:01 loss: 0.1387 (0.3163) time: 1.0769 data: 1.0638 max mem: 9671 +[14:47:22.628086] val: Total time: 0:00:02 (1.1127 s / it) +[14:47:22.636995] val loss: 0.3163122981786728 +[14:47:22.637179] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9435, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9323, Kappa: 0.7561, Score: 0.8589 +[14:47:22.688178] Best epoch = 45, Best score = 0.8589 +[14:47:22.936110] log_dir: ./output_logs/retfound +[14:47:25.186526] Epoch: [47] [0/4] eta: 0:00:08 lr: 0.000010 loss: 0.3408 (0.3408) time: 2.2492 data: 2.1749 max mem: 9671 +[14:47:25.389337] Epoch: [47] [3/4] eta: 0:00:00 lr: 0.000006 loss: 0.2315 (0.2862) time: 0.6127 data: 0.5439 max mem: 9671 +[14:47:25.462167] Epoch: [47] Total time: 0:00:02 (0.6315 s / it) +[14:47:25.463136] Averaged stats: lr: 0.000006 loss: 0.2315 (0.2862) +[14:47:27.749545] val: [0/2] eta: 0:00:04 loss: 0.1387 (0.1387) time: 2.2734 data: 2.2563 max mem: 9671 +[14:47:27.759040] val: [1/2] eta: 0:00:01 loss: 0.1387 (0.3161) time: 1.1411 data: 1.1282 max mem: 9671 +[14:47:27.829830] val: Total time: 0:00:02 (1.1772 s / it) +[14:47:27.838637] val loss: 0.31609438359737396 +[14:47:27.838847] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9444, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9323, Kappa: 0.7561, Score: 0.8592 +[14:47:29.631203] Best epoch = 47, Best score = 0.8592 +[14:47:29.694591] log_dir: ./output_logs/retfound +[14:47:31.964333] Epoch: [48] [0/4] eta: 0:00:09 lr: 0.000005 loss: 0.2687 (0.2687) time: 2.2687 data: 2.1963 max mem: 9671 +[14:47:32.175621] Epoch: [48] [3/4] eta: 0:00:00 lr: 0.000003 loss: 0.2216 (0.2366) time: 0.6198 data: 0.5530 max mem: 9671 +[14:47:32.247823] Epoch: [48] Total time: 0:00:02 (0.6383 s / it) +[14:47:32.248655] Averaged stats: lr: 0.000003 loss: 0.2216 (0.2366) +[14:47:34.525545] val: [0/2] eta: 0:00:04 loss: 0.1387 (0.1387) time: 2.2619 data: 2.2440 max mem: 9671 +[14:47:34.535226] val: [1/2] eta: 0:00:01 loss: 0.1387 (0.3160) time: 1.1356 data: 1.1221 max mem: 9671 +[14:47:34.679867] val: Total time: 0:00:02 (1.2085 s / it) +[14:47:34.688977] val loss: 0.3159617781639099 +[14:47:34.689163] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9453, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9342, Kappa: 0.7561, Score: 0.8595 +[14:47:36.519085] Best epoch = 48, Best score = 0.8595 +[14:47:36.618534] log_dir: ./output_logs/retfound +[14:47:39.205642] Epoch: [49] [0/4] eta: 0:00:10 lr: 0.000002 loss: 0.2089 (0.2089) time: 2.5862 data: 2.5171 max mem: 9671 +[14:47:39.401483] Epoch: [49] [3/4] eta: 0:00:00 lr: 0.000001 loss: 0.3229 (0.3196) time: 0.6953 data: 0.6293 max mem: 9671 +[14:47:39.480545] Epoch: [49] Total time: 0:00:02 (0.7155 s / it) +[14:47:39.481638] Averaged stats: lr: 0.000001 loss: 0.3229 (0.3196) +[14:47:41.956686] val: [0/2] eta: 0:00:04 loss: 0.1387 (0.1387) time: 2.4524 data: 2.4356 max mem: 9671 +[14:47:41.966220] val: [1/2] eta: 0:00:01 loss: 0.1387 (0.3159) time: 1.2307 data: 1.2179 max mem: 9671 +[14:47:42.033997] val: Total time: 0:00:02 (1.2652 s / it) +[14:47:42.044973] val loss: 0.3158949166536331 +[14:47:42.045154] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9348, Kappa: 0.7561, Score: 0.8598 +[14:47:43.878652] Best epoch = 49, Best score = 0.8598 +[14:47:46.536813] Test with the best model, epoch = 49: +[14:47:49.462336] test: [0/3] eta: 0:00:08 loss: 0.1903 (0.1903) time: 2.8937 data: 2.8588 max mem: 9671 +[14:47:49.582925] test: [2/3] eta: 0:00:01 loss: 0.1903 (0.2232) time: 1.0046 data: 0.9530 max mem: 9671 +[14:47:49.646958] test: Total time: 0:00:03 (1.0264 s / it) +[14:47:49.658529] val loss: 0.22319571177164713 +[14:47:49.658639] Accuracy: 0.9250, F1 Score: 0.8925, ROC AUC: 0.9447, Hamming Loss: 0.0750, + Jaccard Score: 0.8110, Precision: 0.8925, Recall: 0.8925, + Average Precision: 0.9483, Kappa: 0.7849, Score: 0.8740 +[14:47:50.395256] Training time 0:05:02 +[rank0]:[W701 14:47:50.847841273 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) diff --git a/results/downsample/adam/050/vit/confusion_matrix.png b/results/downsample/adam/050/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..cc13fa8894510ab34b6a0aa269f830f7747a94b3 --- /dev/null +++ b/results/downsample/adam/050/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c06160d888e1429a22c292cff2c6895a8812af6c90432a3393a39d06502380cf +size 68150 diff --git a/results/downsample/adam/050/vit/log.csv b/results/downsample/adam/050/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..714ed57dc68b7cd610346c2a11f69d8ed471b47c --- /dev/null +++ b/results/downsample/adam/050/vit/log.csv @@ -0,0 +1,34 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8792800903320312,0.3,0.3261648745519713,0.16738866653636775,3.697205891018715e-08 +1,0.8709757924079895,0.625,0.5913978494623655,0.3818659622377501,1.1091617673056146e-07 +2,0.7610486745834351,0.6,0.7670250896057348,0.5371094666147933,1.8486029455093576e-07 +3,0.6221799552440643,0.65,0.8351254480286738,0.5769052396467605,2.5880441237131e-07 +4,0.6431924998760223,0.7,0.8602150537634408,0.621461252734875,3.327485301916844e-07 +5,0.5850805938243866,0.725,0.8817204301075269,0.6640775198291936,3.6960797720518913e-07 +6,0.5217376351356506,0.65,0.8853046594982079,0.6246900186523408,3.687079050660697e-07 +7,0.5671521425247192,0.8,0.9139784946236559,0.7226129194707974,3.6691214584881775e-07 +8,0.5217204391956329,0.85,0.9426523297491038,0.8054182714070511,3.6422944831183056e-07 +9,0.45959457755088806,0.65,0.9498207885304659,0.6461953949964268,3.6067288228784465e-07 +10,0.4830379784107208,0.725,0.9498207885304659,0.7007741596205902,3.5625977500902404e-07 +11,0.5449035167694092,0.85,0.953405017921147,0.7694145758661888,3.5101162669042713e-07 +12,0.4963052272796631,0.8,0.9569892473118279,0.7643672389797613,3.449540057831202e-07 +13,0.4861457943916321,0.775,0.9534050179211468,0.7418807382288818,3.381164244072535e-07 +14,0.44054560363292694,0.85,0.967741935483871,0.8027123777751409,3.3053219457197917e-07 +15,0.4229888617992401,0.925,0.974910394265233,0.8881587700618722,3.222382658826906e-07 +16,0.42020002007484436,0.75,0.967741935483871,0.7262926144259589,3.13275045526259e-07 +17,0.4250647723674774,0.775,0.9534050179211471,0.728656286191625,3.036862014112813e-07 +18,0.4067462384700775,0.925,0.946236559139785,0.8701037147325629,2.9351844942241844e-07 +19,0.4398604929447174,0.95,0.96415770609319,0.9101302275822954,2.8282132582530113e-07 +20,0.39352771639823914,0.825,0.974910394265233,0.7926718678875257,2.716469459308236e-07 +21,0.44525811076164246,0.775,0.982078853046595,0.751438683270698,2.60049750194587e-07 +22,0.4396931082010269,0.95,0.978494623655914,0.9149092001032035,2.4808623898847216e-07 +23,0.4040451943874359,0.95,0.978494623655914,0.9149092001032035,2.3581469733650727e-07 +24,0.3924393653869629,0.85,0.9749103942652328,0.805101864035595,2.232949109560898e-07 +25,0.4416560232639313,0.75,0.967741935483871,0.7262926144259589,2.1058787498798055e-07 +26,0.4181835204362869,0.8,0.9569892473118279,0.7517086310915202,1.9775549683410935e-07 +27,0.35921163856983185,0.9,0.9569892473118279,0.8422939068100358,1.8486029455093576e-07 +28,0.37670303881168365,0.95,0.9605734767025089,0.9089354844520684,1.7196509226776213e-07 +29,0.3816651701927185,0.875,0.9641577060931898,0.8274081283922726,1.5913271411389099e-07 +30,0.357041671872139,0.825,0.9641577060931898,0.7771395358770251,1.464256781457817e-07 +31,0.3465737998485565,0.825,0.9677419354838709,0.7902823816270718,1.339058917653642e-07 +32,0.3870619088411331,0.85,0.9677419354838709,0.8137814733186403,1.2163435011339938e-07 diff --git a/results/downsample/adam/050/vit/metrics.json b/results/downsample/adam/050/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..4b510224a8d96b5cfad23f82c475bad086489b5d --- /dev/null +++ b/results/downsample/adam/050/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8625, + "balanced_accuracy": 0.8127240143369175, + "precision_macro": 0.8011216566005177, + "recall_macro": 0.8127240143369175, + "f1_macro": 0.8066359041968798, + "precision_weighted": 0.8654227782571182, + "recall_weighted": 0.8625, + "f1_weighted": 0.8637991650186772, + "cohen_kappa": 0.6133567662565905, + "quadratic_weighted_kappa": 0.6133567662565905, + "mcc": 0.613736012487117, + "auroc": 0.9005376344086021, + "auprc": 0.8093657787862768, + "sensitivity": 0.7222222222222222, + "specificity": 0.9032258064516129, + "precision_pos": 0.6842105263157895, + "f1_pos": 0.7027027027027027, + "per_class": { + "0": { + "precision": 0.9180327868852459, + "recall": 0.9032258064516129, + "f1-score": 0.9105691056910569, + "support": 62.0 + }, + "1": { + "precision": 0.6842105263157895, + "recall": 0.7222222222222222, + "f1-score": 0.7027027027027027, + "support": 18.0 + }, + "accuracy": 0.8625, + "macro avg": { + "precision": 0.8011216566005177, + "recall": 0.8127240143369175, + "f1-score": 0.8066359041968798, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.8654227782571182, + "recall": 0.8625, + "f1-score": 0.8637991650186772, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/050/vit/pr.png b/results/downsample/adam/050/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..6ebae97356c88838529d27a1b45de6b9b14876d3 --- /dev/null +++ b/results/downsample/adam/050/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c80cd17d5edbfa55ac7e4baccf6361473679f98397a7375c356194b001cd106 +size 45786 diff --git a/results/downsample/adam/050/vit/roc.png b/results/downsample/adam/050/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..142e333ec26b321d5fdcbcadf977dda731a66112 --- /dev/null +++ b/results/downsample/adam/050/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:846b05e6c99c6d37a5a6e270e970b921d4aada656b235301657bb1d1dc7ab9ca +size 57172 diff --git a/results/downsample/adam/050/vit/test_pred.npz b/results/downsample/adam/050/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..9a227e76691c10bd55433b513b6bd881dcdc8bd1 --- /dev/null +++ b/results/downsample/adam/050/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8c821944d6900d95ebeeb3e5408a13bc7f4013a27b78e55216b9be5f86dcbbf +size 1790 diff --git a/results/downsample/adam/050/vit/train.log b/results/downsample/adam/050/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..76e78efcc5fd54a08df274e5610661bb27d83c1c --- /dev/null +++ b/results/downsample/adam/050/vit/train.log @@ -0,0 +1,174 @@ +[vit] train=140 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8793 val_acc=0.3000 val_auc=0.3262 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.8710 val_acc=0.6250 val_auc=0.5914 score=0.3819 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.7610 val_acc=0.6000 val_auc=0.7670 score=0.5371 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6222 val_acc=0.6500 val_auc=0.8351 score=0.5769 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6432 val_acc=0.7000 val_auc=0.8602 score=0.6215 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5851 val_acc=0.7250 val_auc=0.8817 score=0.6641 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5217 val_acc=0.6500 val_auc=0.8853 score=0.6247 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5672 val_acc=0.8000 val_auc=0.9140 score=0.7226 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5217 val_acc=0.8500 val_auc=0.9427 score=0.8054 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4596 val_acc=0.6500 val_auc=0.9498 score=0.6462 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4830 val_acc=0.7250 val_auc=0.9498 score=0.7008 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.5449 val_acc=0.8500 val_auc=0.9534 score=0.7694 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.4963 val_acc=0.8000 val_auc=0.9570 score=0.7644 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4861 val_acc=0.7750 val_auc=0.9534 score=0.7419 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4405 val_acc=0.8500 val_auc=0.9677 score=0.8027 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4230 val_acc=0.9250 val_auc=0.9749 score=0.8882 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4202 val_acc=0.7500 val_auc=0.9677 score=0.7263 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.4251 val_acc=0.7750 val_auc=0.9534 score=0.7287 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4067 val_acc=0.9250 val_auc=0.9462 score=0.8701 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4399 val_acc=0.9500 val_auc=0.9642 score=0.9101 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3935 val_acc=0.8250 val_auc=0.9749 score=0.7927 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4453 val_acc=0.7750 val_auc=0.9821 score=0.7514 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.4397 val_acc=0.9500 val_auc=0.9785 score=0.9149 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4040 val_acc=0.9500 val_auc=0.9785 score=0.9149 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3924 val_acc=0.8500 val_auc=0.9749 score=0.8051 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.4417 val_acc=0.7500 val_auc=0.9677 score=0.7263 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4182 val_acc=0.8000 val_auc=0.9570 score=0.7517 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3592 val_acc=0.9000 val_auc=0.9570 score=0.8423 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3767 val_acc=0.9500 val_auc=0.9606 score=0.9089 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3817 val_acc=0.8750 val_auc=0.9642 score=0.8274 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.3570 val_acc=0.8250 val_auc=0.9642 score=0.7771 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.3466 val_acc=0.8250 val_auc=0.9677 score=0.7903 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep32 loss=0.3871 val_acc=0.8500 val_auc=0.9677 score=0.8138 +[vit] early stop at ep32 (best ep22 score=0.9149) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=22 best_val_score=0.9149 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/050/vit acc=0.8625 auroc=0.9005376344086021 f1_macro=0.8066 qwk=0.6133567662565905 diff --git a/results/downsample/adam/100/resnet/confusion_matrix.png b/results/downsample/adam/100/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..506a3cf3d17837bf728c848d1ac92be80ba45170 --- /dev/null +++ b/results/downsample/adam/100/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c0bfdf6883acbdc7580fcb8c819a39991065f46ba88fda64621285a84be6e661 +size 67078 diff --git a/results/downsample/adam/100/resnet/log.csv b/results/downsample/adam/100/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..1e5ef9a59708211e810817fe45115e87f3bc596a --- /dev/null +++ b/results/downsample/adam/100/resnet/log.csv @@ -0,0 +1,14 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6891738623380661,0.475,0.3225806451612903,0.14145444992498205,0.000125 +1,0.6806204319000244,0.725,0.4946236559139785,0.3357714411517974,0.0002916666666666667 +2,0.6577800512313843,0.8,0.7921146953405018,0.6190155339850959,0.0004583333333333333 +3,0.6238041073083878,0.6,0.7347670250896057,0.5098371835456835,0.0004996859161456965 +4,0.5598906725645065,0.65,0.7025089605734767,0.5326997438283615,0.0004982915790812436 +5,0.4717961400747299,0.725,0.7526881720430108,0.5861279581532292,0.0004957883115509159 +6,0.41071654856204987,0.75,0.7455197132616488,0.5567502986857825,0.0004921872937551814 +7,0.3300529792904854,0.775,0.7311827956989247,0.5047644472897241,0.000487504608713676 +8,0.25707992538809776,0.775,0.7240143369175627,0.5023749610292702,0.00048176117043453436 +9,0.17822003364562988,0.775,0.8243727598566308,0.486325116850161,0.000474982630507352 +10,0.1229514330625534,0.8,0.8064516129032258,0.5038708068939449,0.00046719926353695914 +11,0.1245944295078516,0.8,0.8458781362007168,0.5675158548451437,0.00045844583192968674 +12,0.08867437392473221,0.8,0.7956989247311828,0.5507894510219655,0.00044876143063602076 diff --git a/results/downsample/adam/100/resnet/metrics.json b/results/downsample/adam/100/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..b596ef7f202ee0d40f294fd537cf43922d2c57f7 --- /dev/null +++ b/results/downsample/adam/100/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7375, + "balanced_accuracy": 0.7320788530465949, + "precision_macro": 0.6751183231913456, + "recall_macro": 0.7320788530465949, + "f1_macro": 0.6836753906985502, + "precision_weighted": 0.7998816768086545, + "recall_weighted": 0.7375, + "f1_weighted": 0.7554415364338166, + "cohen_kappa": 0.38144329896907225, + "quadratic_weighted_kappa": 0.38144329896907225, + "mcc": 0.4031935495202789, + "auroc": 0.7795698924731183, + "auprc": 0.596719534313656, + "sensitivity": 0.7222222222222222, + "specificity": 0.7419354838709677, + "precision_pos": 0.4482758620689655, + "f1_pos": 0.5531914893617021, + "per_class": { + "0": { + "precision": 0.9019607843137255, + "recall": 0.7419354838709677, + "f1-score": 0.8141592920353983, + "support": 62.0 + }, + "1": { + "precision": 0.4482758620689655, + "recall": 0.7222222222222222, + "f1-score": 0.5531914893617021, + "support": 18.0 + }, + "accuracy": 0.7375, + "macro avg": { + "precision": 0.6751183231913456, + "recall": 0.7320788530465949, + "f1-score": 0.6836753906985502, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.7998816768086545, + "recall": 0.7375, + "f1-score": 0.7554415364338166, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/100/resnet/pr.png b/results/downsample/adam/100/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a67c1ea0449f06addcd1a9729815ba876c771e9b --- /dev/null +++ b/results/downsample/adam/100/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32cae357856d21d3a77f9dfe1384d166a59d6f70f1782d7d8911f279db466158 +size 52872 diff --git a/results/downsample/adam/100/resnet/roc.png b/results/downsample/adam/100/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..e589d920a43d089fb37e7309292662ab0e2168fd --- /dev/null +++ b/results/downsample/adam/100/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a9ab007820fc10fd859f41d92d55f82ba309715a42a01107a9cd78095069fdf +size 57662 diff --git a/results/downsample/adam/100/resnet/test_pred.npz b/results/downsample/adam/100/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..7395a9c64aa88c869fb73df3080591b855d9cb99 --- /dev/null +++ b/results/downsample/adam/100/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85a0d409b513fa3b96719f13e032a6c35e3a412d2b53ae21d3f4c8fbee32b75c +size 1790 diff --git a/results/downsample/adam/100/resnet/train.log b/results/downsample/adam/100/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d7e42ab7a13882b7de5b09d4313ca3254ff05e2e --- /dev/null +++ b/results/downsample/adam/100/resnet/train.log @@ -0,0 +1,74 @@ +[resnet] train=280 val=40 test=80 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6892 val_acc=0.4750 val_auc=0.3226 score=0.1415 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6806 val_acc=0.7250 val_auc=0.4946 score=0.3358 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6578 val_acc=0.8000 val_auc=0.7921 score=0.6190 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6238 val_acc=0.6000 val_auc=0.7348 score=0.5098 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.5599 val_acc=0.6500 val_auc=0.7025 score=0.5327 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.4718 val_acc=0.7250 val_auc=0.7527 score=0.5861 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.4107 val_acc=0.7500 val_auc=0.7455 score=0.5568 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.3301 val_acc=0.7750 val_auc=0.7312 score=0.5048 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.2571 val_acc=0.7750 val_auc=0.7240 score=0.5024 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1782 val_acc=0.7750 val_auc=0.8244 score=0.4863 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1230 val_acc=0.8000 val_auc=0.8065 score=0.5039 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1246 val_acc=0.8000 val_auc=0.8459 score=0.5675 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.0887 val_acc=0.8000 val_auc=0.7957 score=0.5508 +[resnet] early stop at ep12 (best ep2 score=0.6190) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=2 best_val_score=0.6190 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/100/resnet acc=0.7375 auroc=0.7795698924731183 f1_macro=0.6837 qwk=0.38144329896907225 diff --git a/results/downsample/adam/100/retfound/confusion_matrix.png b/results/downsample/adam/100/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..733b83ea877bc153647ac339aa6088dd1906b567 --- /dev/null +++ b/results/downsample/adam/100/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7142b7b7972e6151b7e2e8d6ac8a164aec7a1fea97226fe5dd05f1952e67a98 +size 68153 diff --git a/results/downsample/adam/100/retfound/confusion_matrix_test.jpg b/results/downsample/adam/100/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..dca75d74bd4dbe952a45d142f6b5e9f8f16695af --- /dev/null +++ b/results/downsample/adam/100/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32bf90581835c5cfef3b2cfa602abce96759dac4a4a2fe9e2c20659164953ba4 +size 257614 diff --git a/results/downsample/adam/100/retfound/log.txt b/results/downsample/adam/100/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..642c68ea3f6fc466c7dacbf53b86f233ea9a17da --- /dev/null +++ b/results/downsample/adam/100/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.7343749999999997e-05, "train_loss": 0.6853790283203125, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.984375e-05, "train_loss": 0.600438117980957, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015234375, "train_loss": 0.5224623680114746, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021484375, "train_loss": 0.5393064022064209, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.00027734375000000003, "train_loss": 0.5149924755096436, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003398437499999999, "train_loss": 0.482379674911499, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040234375, "train_loss": 0.46404457092285156, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046484375000000003, "train_loss": 0.46213990449905396, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.00052734375, "train_loss": 0.41957637667655945, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.00058984375, "train_loss": 0.47334104776382446, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247369453830207, "train_loss": 0.4200931787490845, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006229352074360887, "train_loss": 0.4227827489376068, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006192226158713984, "train_loss": 0.3990233540534973, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006136220600505078, "train_loss": 0.39853784441947937, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006061680692624268, "train_loss": 0.37906284630298615, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005969065998390672, "train_loss": 0.3678862899541855, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005858947518191751, "train_loss": 0.3812587708234787, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005732004169076044, "train_loss": 0.37992827594280243, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005589018599003793, "train_loss": 0.29254330694675446, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005430872361562023, "train_loss": 0.33580098673701286, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005258540480893521, "train_loss": 0.3463967442512512, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005073085440348776, "train_loss": 0.31276238709688187, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004875650631922804, "train_loss": 0.31376905739307404, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00046674533068632887, "train_loss": 0.3495359867811203, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004449777070911855, "train_loss": 0.32266679406166077, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.00042239639704478405, "train_loss": 0.2978495806455612, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.00039914062183260795, "train_loss": 0.2612730134278536, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003753537610421703, "train_loss": 0.32327523455023766, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035118246858017804, "train_loss": 0.25920983776450157, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003267757685024298, "train_loss": 0.29856251925230026, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0003022841362309562, "train_loss": 0.27047010883688927, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002778585708230051, "train_loss": 0.2923644706606865, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002536496640116411, "train_loss": 0.28958597406744957, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022980667175763478, "train_loss": 0.2692524269223213, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00020647659403683159, "train_loss": 0.26890668272972107, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018380326853641835, "train_loss": 0.2811972163617611, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016192648384775085, "train_loss": 0.2884734384715557, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.000140981117623204, "train_loss": 0.2802345249801874, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012109630501059311, "train_loss": 0.28150836005806923, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010239464249204608, "train_loss": 0.2860140986740589, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.499143203592362e-05, "train_loss": 0.27556246146559715, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.899397022184248e-05, "train_loss": 0.2808167040348053, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.450088672158235e-05, "train_loss": 0.24387812614440918, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.1601536214361626e-05, "train_loss": 0.23795588314533234, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.0375447485526644e-05, "train_loss": 0.2574918810278177, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.0891833105144005e-05, "train_loss": 0.2669173013418913, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3209162709490563e-05, "train_loss": 0.2666049748659134, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.374802516302661e-06, "train_loss": 0.2570477966219187, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.4247232962929436e-06, "train_loss": 0.2644502613693476, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.3832786013874152e-06, "train_loss": 0.2804808057844639, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/adam/100/retfound/metrics.json b/results/downsample/adam/100/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..fbf6e32a18f31177fb01f58967bf87dd62ed25a4 --- /dev/null +++ b/results/downsample/adam/100/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.925, + "balanced_accuracy": 0.89247311827957, + "precision_macro": 0.89247311827957, + "recall_macro": 0.89247311827957, + "f1_macro": 0.89247311827957, + "precision_weighted": 0.925, + "recall_weighted": 0.925, + "f1_weighted": 0.925, + "cohen_kappa": 0.7849462365591398, + "quadratic_weighted_kappa": 0.7849462365591398, + "mcc": 0.7849462365591398, + "auroc": 0.9516129032258065, + "auprc": 0.9214129072681705, + "sensitivity": 0.8333333333333334, + "specificity": 0.9516129032258065, + "precision_pos": 0.8333333333333334, + "f1_pos": 0.8333333333333334, + "per_class": { + "0": { + "precision": 0.9516129032258065, + "recall": 0.9516129032258065, + "f1-score": 0.9516129032258065, + "support": 62.0 + }, + "1": { + "precision": 0.8333333333333334, + "recall": 0.8333333333333334, + "f1-score": 0.8333333333333334, + "support": 18.0 + }, + "accuracy": 0.925, + "macro avg": { + "precision": 0.89247311827957, + "recall": 0.89247311827957, + "f1-score": 0.89247311827957, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.925, + "recall": 0.925, + "f1-score": 0.925, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/100/retfound/metrics_test.csv b/results/downsample/adam/100/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..1da15a9eb6ee759ca1bac98d71e7cb6f2b9aaa58 --- /dev/null +++ b/results/downsample/adam/100/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.24544517199198404,0.925,0.89247311827957,0.9516129032258065,0.075,0.810989010989011,0.89247311827957,0.89247311827957,0.9511565422863844,0.7849462365591398 diff --git a/results/downsample/adam/100/retfound/metrics_val.csv b/results/downsample/adam/100/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..1504c8554fabdb27101c5f803242490b41a375e0 --- /dev/null +++ b/results/downsample/adam/100/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6949386596679688,0.775,0.43661971830985913,0.7526881720430108,0.225,0.3875,0.3875,0.5,0.702140345769078,0.0 +0.7357521057128906,0.775,0.43661971830985913,0.7535842293906809,0.225,0.3875,0.3875,0.5,0.7023472297814682,0.0 +0.8536128997802734,0.775,0.43661971830985913,0.7974910394265233,0.225,0.3875,0.3875,0.5,0.7664684375686055,0.0 +0.8364524841308594,0.775,0.43661971830985913,0.8279569892473118,0.225,0.3875,0.3875,0.5,0.8171861361036572,0.0 +0.726959228515625,0.775,0.43661971830985913,0.85752688172043,0.225,0.3875,0.3875,0.5,0.8267116528482032,0.0 +0.6651697158813477,0.775,0.43661971830985913,0.870967741935484,0.225,0.3875,0.3875,0.5,0.8399610377402678,0.0 +0.5009684562683105,0.825,0.7584124245038826,0.8611111111111112,0.175,0.6278280542986425,0.75,0.7688172043010753,0.7919264893748371,0.5172413793103448 +0.6560912132263184,0.85,0.7058823529411764,0.870967741935484,0.15,0.5855855855855856,0.9189189189189189,0.6666666666666666,0.8455413148826051,0.43661971830985913 +0.4889563322067261,0.85,0.7849462365591398,0.881720430107527,0.15,0.6617647058823529,0.7849462365591398,0.7849462365591398,0.8786670749538394,0.5698924731182795 +0.5198326110839844,0.9,0.8268398268398269,0.8530465949820789,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8587794130254447,0.6595744680851063 +0.7482777833938599,0.8,0.5428571428571428,0.8996415770609318,0.2,0.45299145299145294,0.8974358974358974,0.5555555555555556,0.8843861452756653,0.16230366492146608 +0.45795178413391113,0.9,0.8268398268398269,0.9068100358422939,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8862751464020953,0.6595744680851063 +0.5212460160255432,0.9,0.8268398268398269,0.8996415770609318,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.881716849965241,0.6595744680851063 +0.5566548109054565,0.875,0.7948717948717949,0.8924731182795699,0.125,0.6785714285714286,0.857843137254902,0.7616487455197133,0.8804819825795989,0.5934959349593496 +0.6318426728248596,0.9,0.8268398268398269,0.8853046594982079,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.8765545253605149,0.6595744680851063 +0.6386967301368713,0.875,0.7703788748564868,0.8996415770609318,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.8778435406015122,0.5535714285714286 +0.5159331560134888,0.9,0.8268398268398269,0.9175627240143369,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.906900466227048,0.6595744680851063 +0.39144355058670044,0.95,0.921875,0.9283154121863799,0.05,0.8585858585858586,0.9696969696969697,0.8888888888888888,0.9285526168442524,0.8443579766536965 +0.5359691083431244,0.875,0.7703788748564868,0.9292114695340502,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9285066651853976,0.5535714285714286 +0.4337925612926483,0.95,0.921875,0.9283154121863799,0.05,0.8585858585858586,0.9696969696969697,0.8888888888888888,0.9284603359145964,0.8443579766536965 +0.6121203303337097,0.875,0.7703788748564868,0.9283154121863799,0.125,0.6527777777777778,0.9305555555555556,0.7222222222222222,0.9284603359145964,0.5535714285714286 +0.42638447880744934,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145964,0.7560975609756098 +0.3997582793235779,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145963,0.7560975609756098 +0.4086921811103821,0.925,0.8769230769230769,0.9283154121863799,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9284603359145964,0.7560975609756098 +0.41143083572387695,0.9,0.8268398268398269,0.9283154121863799,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9284603359145964,0.6595744680851063 +0.5094860941171646,0.9,0.8268398268398269,0.931899641577061,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9305555924889939,0.6595744680851063 +0.5649861097335815,0.9,0.8268398268398269,0.9292114695340502,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9233688544403764,0.6595744680851063 +0.43275587260723114,0.9,0.8268398268398269,0.9390681003584229,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9251959935525449,0.6595744680851063 +0.21898877620697021,0.9,0.8566308243727598,0.9498207885304659,0.1,0.7575757575757576,0.8566308243727598,0.8566308243727598,0.9370107576036646,0.7132616487455197 +0.28504690527915955,0.925,0.8879551820728291,0.9498207885304659,0.075,0.8045454545454545,0.90625,0.8727598566308243,0.9346306558217193,0.7761194029850746 +0.3365369886159897,0.9,0.84375,0.9498207885304659,0.1,0.7411764705882353,0.8831168831168831,0.8172043010752688,0.940824081249271,0.688715953307393 +0.3427543491125107,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9369379825912034,0.7560975609756098 +0.35773538053035736,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9357159201324631,0.7560975609756098 +0.42521847784519196,0.9,0.8268398268398269,0.946236559139785,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9409427467992981,0.6595744680851063 +0.4029083847999573,0.925,0.8769230769230769,0.9390681003584229,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9352552443376021,0.7560975609756098 +0.3561532497406006,0.925,0.8769230769230769,0.9390681003584229,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9352552443376021,0.7560975609756098 +0.2910751849412918,0.925,0.8769230769230769,0.946236559139785,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9409427467992981,0.7560975609756098 +0.3142062872648239,0.925,0.8769230769230769,0.9498207885304659,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9431059973754814,0.7560975609756098 +0.2977108359336853,0.925,0.8769230769230769,0.953405017921147,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9454120273143385,0.7560975609756098 +0.2965516448020935,0.925,0.8769230769230769,0.953405017921147,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9454120273143386,0.7560975609756098 +0.3348654955625534,0.925,0.8769230769230769,0.9498207885304659,0.075,0.7892156862745098,0.9558823529411764,0.8333333333333333,0.9424153227047307,0.7560975609756098 +0.3686094284057617,0.9,0.8268398268398269,0.953405017921147,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9454425747239183,0.6595744680851063 +0.4043458551168442,0.9,0.8268398268398269,0.953405017921147,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9454120273143387,0.6595744680851063 +0.416122242808342,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492624,0.6595744680851063 +0.4057278037071228,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492624,0.6595744680851063 +0.38761037588119507,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492622,0.6595744680851063 +0.3810284584760666,0.9,0.8268398268398269,0.956989247311828,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9487973403492622,0.6595744680851063 +0.3793119490146637,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 +0.3780297338962555,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 +0.37748883664608,0.9,0.8268398268398269,0.9605734767025089,0.1,0.7206349206349206,0.9428571428571428,0.7777777777777778,0.9526834390073299,0.6595744680851063 diff --git a/results/downsample/adam/100/retfound/pr.png b/results/downsample/adam/100/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..82d2b670c777139265363636bc117fa6727529db --- /dev/null +++ b/results/downsample/adam/100/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9ef8742fd307bb9273e0d952fdb9b2e45d58356db6b2c3f3d2bdf053b1a0e98 +size 41633 diff --git a/results/downsample/adam/100/retfound/roc.png b/results/downsample/adam/100/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..94d89a59442837c9bfb01a10c3cdf55dc732b2f9 --- /dev/null +++ b/results/downsample/adam/100/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9144441574909e862a8170b3e1e6474ce4e034938d1af53ad1199b41c2d6f5a0 +size 57049 diff --git a/results/downsample/adam/100/retfound/test_pred.npz b/results/downsample/adam/100/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..36ade77cb5ec6dbb32b22a2cc8775630b22f4512 --- /dev/null +++ b/results/downsample/adam/100/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e34dd007edfeca4baf2b9c331c60be066942258bbe5b6a3a35befa603b82ef51 +size 1470 diff --git a/results/downsample/adam/100/retfound/train.log b/results/downsample/adam/100/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..67dc548f9fcfbb10081a33674c1152a5cea1cdc5 --- /dev/null +++ b/results/downsample/adam/100/retfound/train.log @@ -0,0 +1,732 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:42:25.262726227 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:42:26.945337] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:42:26.945789] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/adam_100', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/100', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:42:42.399687] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:44.366335] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:42:49.190123] Sampler_train = +[14:42:49.236196] len of train_set: 256 +[14:42:49.442191] [Adaptation] Full fine-tuning: training all parameters. +[14:42:49.443226] number of trainable params (M): 303.30 +[14:42:49.443332] base lr: 5.00e-03 +[14:42:49.443413] actual lr: 6.25e-04 +[14:42:49.443478] accumulate grad iterations: 1 +[14:42:49.443547] effective batch size: 32 +[14:42:49.446480] criterion = CrossEntropyLoss() +[14:42:49.446566] Start training for 50 epochs +[14:42:49.448475] log_dir: ./output_logs/retfound +[14:42:52.425788] Epoch: [0] [0/8] eta: 0:00:23 lr: 0.000000 loss: 0.6928 (0.6928) time: 2.9764 data: 2.1421 max mem: 7340 +[14:42:53.444440] Epoch: [0] [7/8] eta: 0:00:00 lr: 0.000055 loss: 0.6831 (0.6854) time: 0.4993 data: 0.2679 max mem: 9671 +[14:42:53.508634] Epoch: [0] Total time: 0:00:04 (0.5075 s / it) +[14:42:53.517033] Averaged stats: lr: 0.000055 loss: 0.6831 (0.6854) +[14:42:55.760416] val: [0/2] eta: 0:00:04 loss: 0.6396 (0.6396) time: 2.2311 data: 2.1773 max mem: 9671 +[14:42:55.865613] val: [1/2] eta: 0:00:01 loss: 0.6396 (0.6949) time: 1.1679 data: 1.0887 max mem: 9671 +[14:42:56.040954] val: Total time: 0:00:02 (1.2561 s / it) +[14:42:56.053024] val loss: 0.6949386596679688 +[14:42:56.053211] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7527, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7021, Kappa: 0.0000, Score: 0.3964 +[14:42:58.599129] Best epoch = 0, Best score = 0.3964 +[14:42:58.661474] log_dir: ./output_logs/retfound +[14:43:00.852991] Epoch: [1] [0/8] eta: 0:00:17 lr: 0.000063 loss: 0.6628 (0.6628) time: 2.1905 data: 2.0168 max mem: 9671 +[14:43:01.851234] Epoch: [1] [7/8] eta: 0:00:00 lr: 0.000117 loss: 0.6032 (0.6004) time: 0.3985 data: 0.2522 max mem: 9671 +[14:43:02.036532] Epoch: [1] Total time: 0:00:03 (0.4219 s / it) +[14:43:02.044987] Averaged stats: lr: 0.000117 loss: 0.6032 (0.6004) +[14:43:04.235654] val: [0/2] eta: 0:00:04 loss: 0.4481 (0.4481) time: 2.1713 data: 2.1356 max mem: 9671 +[14:43:04.251485] val: [1/2] eta: 0:00:01 loss: 0.4481 (0.7358) time: 1.0933 data: 1.0679 max mem: 9671 +[14:43:04.400085] val: Total time: 0:00:02 (1.1682 s / it) +[14:43:04.409253] val loss: 0.7357521057128906 +[14:43:04.409421] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7536, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7023, Kappa: 0.0000, Score: 0.3967 +[14:43:06.679524] Best epoch = 1, Best score = 0.3967 +[14:43:06.751934] log_dir: ./output_logs/retfound +[14:43:09.023944] Epoch: [2] [0/8] eta: 0:00:18 lr: 0.000125 loss: 0.6224 (0.6224) time: 2.2710 data: 2.1225 max mem: 9671 +[14:43:10.018423] Epoch: [2] [7/8] eta: 0:00:00 lr: 0.000180 loss: 0.4999 (0.5225) time: 0.4081 data: 0.2654 max mem: 9671 +[14:43:10.205514] Epoch: [2] Total time: 0:00:03 (0.4317 s / it) +[14:43:10.214461] Averaged stats: lr: 0.000180 loss: 0.4999 (0.5225) +[14:43:12.416817] val: [0/2] eta: 0:00:04 loss: 0.2783 (0.2783) time: 2.1715 data: 2.1350 max mem: 9671 +[14:43:12.432579] val: [1/2] eta: 0:00:01 loss: 0.2783 (0.8536) time: 1.0933 data: 1.0676 max mem: 9671 +[14:43:12.501874] val: Total time: 0:00:02 (1.1287 s / it) +[14:43:12.511650] val loss: 0.8536128997802734 +[14:43:12.511855] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.7975, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.7665, Kappa: 0.0000, Score: 0.4114 +[14:43:14.403721] Best epoch = 2, Best score = 0.4114 +[14:43:14.472149] log_dir: ./output_logs/retfound +[14:43:16.650481] Epoch: [3] [0/8] eta: 0:00:17 lr: 0.000188 loss: 0.6936 (0.6936) time: 2.1773 data: 2.0283 max mem: 9671 +[14:43:17.620738] Epoch: [3] [7/8] eta: 0:00:00 lr: 0.000242 loss: 0.5377 (0.5393) time: 0.3934 data: 0.2536 max mem: 9671 +[14:43:17.694902] Epoch: [3] Total time: 0:00:03 (0.4028 s / it) +[14:43:17.703145] Averaged stats: lr: 0.000242 loss: 0.5377 (0.5393) +[14:43:20.040299] val: [0/2] eta: 0:00:04 loss: 0.2665 (0.2665) time: 2.3249 data: 2.2889 max mem: 9671 +[14:43:20.056394] val: [1/2] eta: 0:00:01 loss: 0.2665 (0.8365) time: 1.1702 data: 1.1445 max mem: 9671 +[14:43:20.141116] val: Total time: 0:00:02 (1.2132 s / it) +[14:43:20.151094] val loss: 0.8364524841308594 +[14:43:20.151280] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8280, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8172, Kappa: 0.0000, Score: 0.4215 +[14:43:22.085041] Best epoch = 3, Best score = 0.4215 +[14:43:22.148876] log_dir: ./output_logs/retfound +[14:43:24.362815] Epoch: [4] [0/8] eta: 0:00:17 lr: 0.000250 loss: 0.6812 (0.6812) time: 2.2129 data: 2.0754 max mem: 9671 +[14:43:25.347578] Epoch: [4] [7/8] eta: 0:00:00 lr: 0.000305 loss: 0.4633 (0.5150) time: 0.3996 data: 0.2595 max mem: 9671 +[14:43:25.418125] Epoch: [4] Total time: 0:00:03 (0.4086 s / it) +[14:43:25.426713] Averaged stats: lr: 0.000305 loss: 0.4633 (0.5150) +[14:43:27.600830] val: [0/2] eta: 0:00:04 loss: 0.3205 (0.3205) time: 2.1625 data: 2.1260 max mem: 9671 +[14:43:27.616525] val: [1/2] eta: 0:00:01 loss: 0.3205 (0.7270) time: 1.0888 data: 1.0631 max mem: 9671 +[14:43:27.689773] val: Total time: 0:00:02 (1.1261 s / it) +[14:43:27.701125] val loss: 0.726959228515625 +[14:43:27.701328] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8575, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8267, Kappa: 0.0000, Score: 0.4314 +[14:43:29.621863] Best epoch = 4, Best score = 0.4314 +[14:43:29.712048] log_dir: ./output_logs/retfound +[14:43:31.982298] Epoch: [5] [0/8] eta: 0:00:18 lr: 0.000313 loss: 0.4578 (0.4578) time: 2.2692 data: 2.1961 max mem: 9671 +[14:43:32.440991] Epoch: [5] [7/8] eta: 0:00:00 lr: 0.000367 loss: 0.4578 (0.4824) time: 0.3409 data: 0.2746 max mem: 9671 +[14:43:32.515150] Epoch: [5] Total time: 0:00:02 (0.3504 s / it) +[14:43:32.516005] Averaged stats: lr: 0.000367 loss: 0.4578 (0.4824) +[14:43:34.708285] val: [0/2] eta: 0:00:04 loss: 0.2633 (0.2633) time: 2.1776 data: 2.1575 max mem: 9671 +[14:43:34.721546] val: [1/2] eta: 0:00:01 loss: 0.2633 (0.6652) time: 1.0950 data: 1.0788 max mem: 9671 +[14:43:34.794201] val: Total time: 0:00:02 (1.1321 s / it) +[14:43:34.803975] val loss: 0.6651697158813477 +[14:43:34.804301] Accuracy: 0.7750, F1 Score: 0.4366, ROC AUC: 0.8710, Hamming Loss: 0.2250, + Jaccard Score: 0.3875, Precision: 0.3875, Recall: 0.5000, + Average Precision: 0.8400, Kappa: 0.0000, Score: 0.4359 +[14:43:36.729319] Best epoch = 5, Best score = 0.4359 +[14:43:36.807861] log_dir: ./output_logs/retfound +[14:43:39.044483] Epoch: [6] [0/8] eta: 0:00:17 lr: 0.000375 loss: 0.3839 (0.3839) time: 2.2356 data: 2.1614 max mem: 9671 +[14:43:39.505201] Epoch: [6] [7/8] eta: 0:00:00 lr: 0.000430 loss: 0.4587 (0.4640) time: 0.3369 data: 0.2702 max mem: 9671 +[14:43:39.575363] Epoch: [6] Total time: 0:00:02 (0.3459 s / it) +[14:43:39.576212] Averaged stats: lr: 0.000430 loss: 0.4587 (0.4640) +[14:43:41.851387] val: [0/2] eta: 0:00:04 loss: 0.3132 (0.3132) time: 2.2642 data: 2.2472 max mem: 9671 +[14:43:41.862426] val: [1/2] eta: 0:00:01 loss: 0.3132 (0.5010) time: 1.1374 data: 1.1236 max mem: 9671 +[14:43:41.929715] val: Total time: 0:00:02 (1.1716 s / it) +[14:43:41.939647] val loss: 0.5009684562683105 +[14:43:41.939859] Accuracy: 0.8250, F1 Score: 0.7584, ROC AUC: 0.8611, Hamming Loss: 0.1750, + Jaccard Score: 0.6278, Precision: 0.7500, Recall: 0.7688, + Average Precision: 0.7919, Kappa: 0.5172, Score: 0.7123 +[14:43:43.838897] Best epoch = 6, Best score = 0.7123 +[14:43:43.900909] log_dir: ./output_logs/retfound +[14:43:46.027929] Epoch: [7] [0/8] eta: 0:00:17 lr: 0.000438 loss: 0.5242 (0.5242) time: 2.1260 data: 2.0539 max mem: 9671 +[14:43:46.513438] Epoch: [7] [7/8] eta: 0:00:00 lr: 0.000492 loss: 0.4543 (0.4621) time: 0.3264 data: 0.2568 max mem: 9671 +[14:43:46.586492] Epoch: [7] Total time: 0:00:02 (0.3357 s / it) +[14:43:46.587445] Averaged stats: lr: 0.000492 loss: 0.4543 (0.4621) +[14:43:49.001533] val: [0/2] eta: 0:00:04 loss: 0.1982 (0.1982) time: 2.4031 data: 2.3861 max mem: 9671 +[14:43:49.011495] val: [1/2] eta: 0:00:01 loss: 0.1982 (0.6561) time: 1.2063 data: 1.1931 max mem: 9671 +[14:43:49.082691] val: Total time: 0:00:02 (1.2425 s / it) +[14:43:49.096852] val loss: 0.6560912132263184 +[14:43:49.097085] Accuracy: 0.8500, F1 Score: 0.7059, ROC AUC: 0.8710, Hamming Loss: 0.1500, + Jaccard Score: 0.5856, Precision: 0.9189, Recall: 0.6667, + Average Precision: 0.8455, Kappa: 0.4366, Score: 0.6712 +[14:43:49.141002] Best epoch = 6, Best score = 0.7123 +[14:43:49.371457] log_dir: ./output_logs/retfound +[14:43:51.516392] Epoch: [8] [0/8] eta: 0:00:17 lr: 0.000500 loss: 0.5544 (0.5544) time: 2.1440 data: 2.0697 max mem: 9671 +[14:43:51.980368] Epoch: [8] [7/8] eta: 0:00:00 lr: 0.000555 loss: 0.4068 (0.4196) time: 0.3259 data: 0.2588 max mem: 9671 +[14:43:52.048982] Epoch: [8] Total time: 0:00:02 (0.3347 s / it) +[14:43:52.049776] Averaged stats: lr: 0.000555 loss: 0.4068 (0.4196) +[14:43:54.497264] val: [0/2] eta: 0:00:04 loss: 0.2713 (0.2713) time: 2.4318 data: 2.4151 max mem: 9671 +[14:43:54.507190] val: [1/2] eta: 0:00:01 loss: 0.2713 (0.4890) time: 1.2206 data: 1.2076 max mem: 9671 +[14:43:54.581942] val: Total time: 0:00:02 (1.2586 s / it) +[14:43:54.590784] val loss: 0.4889563322067261 +[14:43:54.590984] Accuracy: 0.8500, F1 Score: 0.7849, ROC AUC: 0.8817, Hamming Loss: 0.1500, + Jaccard Score: 0.6618, Precision: 0.7849, Recall: 0.7849, + Average Precision: 0.8787, Kappa: 0.5699, Score: 0.7455 +[14:43:56.465335] Best epoch = 8, Best score = 0.7455 +[14:43:56.536912] log_dir: ./output_logs/retfound +[14:43:58.700719] Epoch: [9] [0/8] eta: 0:00:17 lr: 0.000562 loss: 0.5163 (0.5163) time: 2.1628 data: 2.0918 max mem: 9671 +[14:43:59.160118] Epoch: [9] [7/8] eta: 0:00:00 lr: 0.000617 loss: 0.4599 (0.4733) time: 0.3277 data: 0.2616 max mem: 9671 +[14:43:59.240528] Epoch: [9] Total time: 0:00:02 (0.3379 s / it) +[14:43:59.241414] Averaged stats: lr: 0.000617 loss: 0.4599 (0.4733) +[14:44:01.488430] val: [0/2] eta: 0:00:04 loss: 0.2701 (0.2701) time: 2.2354 data: 2.2180 max mem: 9671 +[14:44:01.498729] val: [1/2] eta: 0:00:01 loss: 0.2701 (0.5198) time: 1.1225 data: 1.1091 max mem: 9671 +[14:44:01.575642] val: Total time: 0:00:02 (1.1617 s / it) +[14:44:01.585385] val loss: 0.5198326110839844 +[14:44:01.585594] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8530, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8588, Kappa: 0.6596, Score: 0.7798 +[14:44:03.488059] Best epoch = 9, Best score = 0.7798 +[14:44:03.559202] log_dir: ./output_logs/retfound +[14:44:05.667931] Epoch: [10] [0/8] eta: 0:00:16 lr: 0.000625 loss: 0.4179 (0.4179) time: 2.1076 data: 2.0299 max mem: 9671 +[14:44:06.283849] Epoch: [10] [7/8] eta: 0:00:00 lr: 0.000624 loss: 0.4179 (0.4201) time: 0.3403 data: 0.2734 max mem: 9671 +[14:44:06.353611] Epoch: [10] Total time: 0:00:02 (0.3493 s / it) +[14:44:06.354435] Averaged stats: lr: 0.000624 loss: 0.4179 (0.4201) +[14:44:08.487122] val: [0/2] eta: 0:00:04 loss: 0.1527 (0.1527) time: 2.1207 data: 2.1033 max mem: 9671 +[14:44:08.498147] val: [1/2] eta: 0:00:01 loss: 0.1527 (0.7483) time: 1.0656 data: 1.0517 max mem: 9671 +[14:44:08.672539] val: Total time: 0:00:02 (1.1534 s / it) +[14:44:08.682310] val loss: 0.7482777833938599 +[14:44:08.682509] Accuracy: 0.8000, F1 Score: 0.5429, ROC AUC: 0.8996, Hamming Loss: 0.2000, + Jaccard Score: 0.4530, Precision: 0.8974, Recall: 0.5556, + Average Precision: 0.8844, Kappa: 0.1623, Score: 0.5349 +[14:44:08.837054] Best epoch = 9, Best score = 0.7798 +[14:44:09.447344] log_dir: ./output_logs/retfound +[14:44:11.606190] Epoch: [11] [0/8] eta: 0:00:17 lr: 0.000624 loss: 0.3514 (0.3514) time: 2.1578 data: 2.0874 max mem: 9671 +[14:44:12.550596] Epoch: [11] [7/8] eta: 0:00:00 lr: 0.000622 loss: 0.3698 (0.4228) time: 0.3877 data: 0.2611 max mem: 9671 +[14:44:12.622069] Epoch: [11] Total time: 0:00:03 (0.3968 s / it) +[14:44:12.631741] Averaged stats: lr: 0.000622 loss: 0.3698 (0.4228) +[14:44:14.779766] val: [0/2] eta: 0:00:04 loss: 0.2029 (0.2029) time: 2.1366 data: 2.1195 max mem: 9671 +[14:44:14.789712] val: [1/2] eta: 0:00:01 loss: 0.2029 (0.4580) time: 1.0730 data: 1.0598 max mem: 9671 +[14:44:14.856505] val: Total time: 0:00:02 (1.1070 s / it) +[14:44:14.865420] val loss: 0.45795178413391113 +[14:44:14.865621] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9068, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8863, Kappa: 0.6596, Score: 0.7977 +[14:44:16.810604] Best epoch = 11, Best score = 0.7977 +[14:44:16.889069] log_dir: ./output_logs/retfound +[14:44:19.303574] Epoch: [12] [0/8] eta: 0:00:19 lr: 0.000621 loss: 0.3098 (0.3098) time: 2.4135 data: 2.3400 max mem: 9671 +[14:44:19.764424] Epoch: [12] [7/8] eta: 0:00:00 lr: 0.000617 loss: 0.3706 (0.3990) time: 0.3592 data: 0.2926 max mem: 9671 +[14:44:19.837192] Epoch: [12] Total time: 0:00:02 (0.3685 s / it) +[14:44:19.838077] Averaged stats: lr: 0.000617 loss: 0.3706 (0.3990) +[14:44:22.020287] val: [0/2] eta: 0:00:04 loss: 0.1898 (0.1898) time: 2.1702 data: 2.1517 max mem: 9671 +[14:44:22.032504] val: [1/2] eta: 0:00:01 loss: 0.1898 (0.5212) time: 1.0909 data: 1.0759 max mem: 9671 +[14:44:22.122600] val: Total time: 0:00:02 (1.1366 s / it) +[14:44:22.133444] val loss: 0.5212460160255432 +[14:44:22.133631] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8996, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8817, Kappa: 0.6596, Score: 0.7954 +[14:44:22.176125] Best epoch = 11, Best score = 0.7977 +[14:44:22.426892] log_dir: ./output_logs/retfound +[14:44:24.674559] Epoch: [13] [0/8] eta: 0:00:17 lr: 0.000616 loss: 0.5697 (0.5697) time: 2.2466 data: 2.1766 max mem: 9671 +[14:44:25.218546] Epoch: [13] [7/8] eta: 0:00:00 lr: 0.000611 loss: 0.3551 (0.3985) time: 0.3487 data: 0.2828 max mem: 9671 +[14:44:25.294210] Epoch: [13] Total time: 0:00:02 (0.3584 s / it) +[14:44:25.295045] Averaged stats: lr: 0.000611 loss: 0.3551 (0.3985) +[14:44:27.381508] val: [0/2] eta: 0:00:04 loss: 0.1778 (0.1778) time: 2.0704 data: 2.0525 max mem: 9671 +[14:44:27.395707] val: [1/2] eta: 0:00:01 loss: 0.1778 (0.5567) time: 1.0419 data: 1.0263 max mem: 9671 +[14:44:27.464848] val: Total time: 0:00:02 (1.0773 s / it) +[14:44:27.474001] val loss: 0.5566548109054565 +[14:44:27.474224] Accuracy: 0.8750, F1 Score: 0.7949, ROC AUC: 0.8925, Hamming Loss: 0.1250, + Jaccard Score: 0.6786, Precision: 0.8578, Recall: 0.7616, + Average Precision: 0.8805, Kappa: 0.5935, Score: 0.7603 +[14:44:27.520148] Best epoch = 11, Best score = 0.7977 +[14:44:27.768495] log_dir: ./output_logs/retfound +[14:44:29.975094] Epoch: [14] [0/8] eta: 0:00:17 lr: 0.000610 loss: 0.4109 (0.4109) time: 2.2057 data: 2.1052 max mem: 9671 +[14:44:30.435911] Epoch: [14] [7/8] eta: 0:00:00 lr: 0.000602 loss: 0.3012 (0.3791) time: 0.3332 data: 0.2632 max mem: 9671 +[14:44:30.506430] Epoch: [14] Total time: 0:00:02 (0.3422 s / it) +[14:44:30.507206] Averaged stats: lr: 0.000602 loss: 0.3012 (0.3791) +[14:44:32.608835] val: [0/2] eta: 0:00:04 loss: 0.1636 (0.1636) time: 2.0858 data: 2.0689 max mem: 9671 +[14:44:32.618887] val: [1/2] eta: 0:00:01 loss: 0.1636 (0.6318) time: 1.0476 data: 1.0345 max mem: 9671 +[14:44:32.685448] val: Total time: 0:00:02 (1.0816 s / it) +[14:44:32.694392] val loss: 0.6318426728248596 +[14:44:32.694567] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.8853, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.8766, Kappa: 0.6596, Score: 0.7906 +[14:44:32.735089] Best epoch = 11, Best score = 0.7977 +[14:44:32.977932] log_dir: ./output_logs/retfound +[14:44:35.160570] Epoch: [15] [0/8] eta: 0:00:17 lr: 0.000601 loss: 0.4683 (0.4683) time: 2.1817 data: 2.1095 max mem: 9671 +[14:44:35.620639] Epoch: [15] [7/8] eta: 0:00:00 lr: 0.000592 loss: 0.3347 (0.3679) time: 0.3301 data: 0.2637 max mem: 9671 +[14:44:35.695110] Epoch: [15] Total time: 0:00:02 (0.3396 s / it) +[14:44:35.695956] Averaged stats: lr: 0.000592 loss: 0.3347 (0.3679) +[14:44:37.820579] val: [0/2] eta: 0:00:04 loss: 0.1511 (0.1511) time: 2.1134 data: 2.0958 max mem: 9671 +[14:44:37.830840] val: [1/2] eta: 0:00:01 loss: 0.1511 (0.6387) time: 1.0616 data: 1.0480 max mem: 9671 +[14:44:37.898511] val: Total time: 0:00:02 (1.0961 s / it) +[14:44:37.907467] val loss: 0.6386967301368713 +[14:44:37.907675] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.8996, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.8778, Kappa: 0.5536, Score: 0.7412 +[14:44:37.944278] Best epoch = 11, Best score = 0.7977 +[14:44:38.202770] log_dir: ./output_logs/retfound +[14:44:40.566721] Epoch: [16] [0/8] eta: 0:00:18 lr: 0.000591 loss: 0.3331 (0.3331) time: 2.3630 data: 2.2308 max mem: 9671 +[14:44:41.557422] Epoch: [16] [7/8] eta: 0:00:00 lr: 0.000581 loss: 0.2960 (0.3813) time: 0.4191 data: 0.2790 max mem: 9671 +[14:44:41.638078] Epoch: [16] Total time: 0:00:03 (0.4294 s / it) +[14:44:41.638972] Averaged stats: lr: 0.000581 loss: 0.2960 (0.3813) +[14:44:43.787545] val: [0/2] eta: 0:00:04 loss: 0.1563 (0.1563) time: 2.1262 data: 2.1092 max mem: 9671 +[14:44:43.798088] val: [1/2] eta: 0:00:01 loss: 0.1563 (0.5159) time: 1.0681 data: 1.0547 max mem: 9671 +[14:44:43.866735] val: Total time: 0:00:02 (1.1031 s / it) +[14:44:43.875617] val loss: 0.5159331560134888 +[14:44:43.875866] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9176, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9069, Kappa: 0.6596, Score: 0.8013 +[14:44:45.972030] Best epoch = 16, Best score = 0.8013 +[14:44:46.084016] log_dir: ./output_logs/retfound +[14:44:48.157568] Epoch: [17] [0/8] eta: 0:00:16 lr: 0.000579 loss: 0.4858 (0.4858) time: 2.0721 data: 1.9891 max mem: 9671 +[14:44:48.709017] Epoch: [17] [7/8] eta: 0:00:00 lr: 0.000567 loss: 0.3880 (0.3799) time: 0.3278 data: 0.2593 max mem: 9671 +[14:44:48.801568] Epoch: [17] Total time: 0:00:02 (0.3397 s / it) +[14:44:48.802347] Averaged stats: lr: 0.000567 loss: 0.3880 (0.3799) +[14:44:51.077905] val: [0/2] eta: 0:00:04 loss: 0.1928 (0.1928) time: 2.2598 data: 2.2424 max mem: 9671 +[14:44:51.088209] val: [1/2] eta: 0:00:01 loss: 0.1928 (0.3914) time: 1.1348 data: 1.1213 max mem: 9671 +[14:44:51.159667] val: Total time: 0:00:02 (1.1711 s / it) +[14:44:51.168578] val loss: 0.39144355058670044 +[14:44:51.168762] Accuracy: 0.9500, F1 Score: 0.9219, ROC AUC: 0.9283, Hamming Loss: 0.0500, + Jaccard Score: 0.8586, Precision: 0.9697, Recall: 0.8889, + Average Precision: 0.9286, Kappa: 0.8444, Score: 0.8982 +[14:44:53.155091] Best epoch = 17, Best score = 0.8982 +[14:44:53.206948] log_dir: ./output_logs/retfound +[14:44:55.287176] Epoch: [18] [0/8] eta: 0:00:16 lr: 0.000565 loss: 0.2349 (0.2349) time: 2.0793 data: 2.0056 max mem: 9671 +[14:44:55.746694] Epoch: [18] [7/8] eta: 0:00:00 lr: 0.000552 loss: 0.2442 (0.2925) time: 0.3173 data: 0.2508 max mem: 9671 +[14:44:55.820987] Epoch: [18] Total time: 0:00:02 (0.3267 s / it) +[14:44:55.821893] Averaged stats: lr: 0.000552 loss: 0.2442 (0.2925) +[14:44:57.964328] val: [0/2] eta: 0:00:04 loss: 0.1443 (0.1443) time: 2.1235 data: 2.1037 max mem: 9671 +[14:44:57.976304] val: [1/2] eta: 0:00:01 loss: 0.1443 (0.5360) time: 1.0674 data: 1.0519 max mem: 9671 +[14:44:58.043946] val: Total time: 0:00:02 (1.1020 s / it) +[14:44:58.053645] val loss: 0.5359691083431244 +[14:44:58.053843] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9292, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9285, Kappa: 0.5536, Score: 0.7511 +[14:44:58.089662] Best epoch = 17, Best score = 0.8982 +[14:44:58.337576] log_dir: ./output_logs/retfound +[14:45:00.782323] Epoch: [19] [0/8] eta: 0:00:19 lr: 0.000550 loss: 0.4059 (0.4059) time: 2.4438 data: 2.3576 max mem: 9671 +[14:45:01.275056] Epoch: [19] [7/8] eta: 0:00:00 lr: 0.000536 loss: 0.2734 (0.3358) time: 0.3670 data: 0.2948 max mem: 9671 +[14:45:01.360164] Epoch: [19] Total time: 0:00:03 (0.3778 s / it) +[14:45:01.361754] Averaged stats: lr: 0.000536 loss: 0.2734 (0.3358) +[14:45:03.767646] val: [0/2] eta: 0:00:04 loss: 0.1677 (0.1677) time: 2.3944 data: 2.3601 max mem: 9671 +[14:45:03.783214] val: [1/2] eta: 0:00:01 loss: 0.1677 (0.4338) time: 1.2047 data: 1.1801 max mem: 9671 +[14:45:03.851971] val: Total time: 0:00:02 (1.2398 s / it) +[14:45:03.860761] val loss: 0.4337925612926483 +[14:45:03.860976] Accuracy: 0.9500, F1 Score: 0.9219, ROC AUC: 0.9283, Hamming Loss: 0.0500, + Jaccard Score: 0.8586, Precision: 0.9697, Recall: 0.8889, + Average Precision: 0.9285, Kappa: 0.8444, Score: 0.8982 +[14:45:03.898101] Best epoch = 17, Best score = 0.8982 +[14:45:04.162187] log_dir: ./output_logs/retfound +[14:45:06.302256] Epoch: [20] [0/8] eta: 0:00:17 lr: 0.000534 loss: 0.3245 (0.3245) time: 2.1384 data: 2.0246 max mem: 9671 +[14:45:06.770292] Epoch: [20] [7/8] eta: 0:00:00 lr: 0.000518 loss: 0.3245 (0.3464) time: 0.3257 data: 0.2531 max mem: 9671 +[14:45:06.845260] Epoch: [20] Total time: 0:00:02 (0.3354 s / it) +[14:45:06.846050] Averaged stats: lr: 0.000518 loss: 0.3245 (0.3464) +[14:45:09.001810] val: [0/2] eta: 0:00:04 loss: 0.1523 (0.1523) time: 2.1451 data: 2.1278 max mem: 9671 +[14:45:09.012166] val: [1/2] eta: 0:00:01 loss: 0.1523 (0.6121) time: 1.0774 data: 1.0640 max mem: 9671 +[14:45:09.081043] val: Total time: 0:00:02 (1.1125 s / it) +[14:45:09.090036] val loss: 0.6121203303337097 +[14:45:09.090336] Accuracy: 0.8750, F1 Score: 0.7704, ROC AUC: 0.9283, Hamming Loss: 0.1250, + Jaccard Score: 0.6528, Precision: 0.9306, Recall: 0.7222, + Average Precision: 0.9285, Kappa: 0.5536, Score: 0.7508 +[14:45:09.135750] Best epoch = 17, Best score = 0.8982 +[14:45:09.385025] log_dir: ./output_logs/retfound +[14:45:11.602028] Epoch: [21] [0/8] eta: 0:00:17 lr: 0.000516 loss: 0.2419 (0.2419) time: 2.2160 data: 2.1468 max mem: 9671 +[14:45:12.080635] Epoch: [21] [7/8] eta: 0:00:00 lr: 0.000499 loss: 0.2997 (0.3128) time: 0.3367 data: 0.2684 max mem: 9671 +[14:45:12.154406] Epoch: [21] Total time: 0:00:02 (0.3461 s / it) +[14:45:12.155264] Averaged stats: lr: 0.000499 loss: 0.2997 (0.3128) +[14:45:14.295739] val: [0/2] eta: 0:00:04 loss: 0.1637 (0.1637) time: 2.1287 data: 2.1113 max mem: 9671 +[14:45:14.305979] val: [1/2] eta: 0:00:01 loss: 0.1637 (0.4264) time: 1.0692 data: 1.0557 max mem: 9671 +[14:45:14.373807] val: Total time: 0:00:02 (1.1038 s / it) +[14:45:14.382722] val loss: 0.42638447880744934 +[14:45:14.382923] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[14:45:14.417379] Best epoch = 17, Best score = 0.8982 +[14:45:14.673940] log_dir: ./output_logs/retfound +[14:45:16.899864] Epoch: [22] [0/8] eta: 0:00:17 lr: 0.000496 loss: 0.2969 (0.2969) time: 2.2250 data: 2.1550 max mem: 9671 +[14:45:17.358728] Epoch: [22] [7/8] eta: 0:00:00 lr: 0.000479 loss: 0.3222 (0.3138) time: 0.3354 data: 0.2694 max mem: 9671 +[14:45:17.431694] Epoch: [22] Total time: 0:00:02 (0.3447 s / it) +[14:45:17.432507] Averaged stats: lr: 0.000479 loss: 0.3222 (0.3138) +[14:45:19.542821] val: [0/2] eta: 0:00:04 loss: 0.1736 (0.1736) time: 2.0988 data: 2.0817 max mem: 9671 +[14:45:19.553153] val: [1/2] eta: 0:00:01 loss: 0.1736 (0.3998) time: 1.0542 data: 1.0409 max mem: 9671 +[14:45:19.625142] val: Total time: 0:00:02 (1.0909 s / it) +[14:45:19.634165] val loss: 0.3997582793235779 +[14:45:19.634350] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[14:45:19.672747] Best epoch = 17, Best score = 0.8982 +[14:45:19.911161] log_dir: ./output_logs/retfound +[14:45:22.168906] Epoch: [23] [0/8] eta: 0:00:18 lr: 0.000476 loss: 0.2342 (0.2342) time: 2.2566 data: 2.1146 max mem: 9671 +[14:45:23.177423] Epoch: [23] [7/8] eta: 0:00:00 lr: 0.000457 loss: 0.3115 (0.3495) time: 0.4080 data: 0.2644 max mem: 9671 +[14:45:23.249725] Epoch: [23] Total time: 0:00:03 (0.4173 s / it) +[14:45:23.250650] Averaged stats: lr: 0.000457 loss: 0.3115 (0.3495) +[14:45:25.355905] val: [0/2] eta: 0:00:04 loss: 0.1643 (0.1643) time: 2.0929 data: 2.0759 max mem: 9671 +[14:45:25.365785] val: [1/2] eta: 0:00:01 loss: 0.1643 (0.4087) time: 1.0511 data: 1.0380 max mem: 9671 +[14:45:25.436468] val: Total time: 0:00:02 (1.0871 s / it) +[14:45:25.445342] val loss: 0.4086921811103821 +[14:45:25.445588] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9283, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9285, Kappa: 0.7561, Score: 0.8538 +[14:45:25.480593] Best epoch = 17, Best score = 0.8982 +[14:45:25.739522] log_dir: ./output_logs/retfound +[14:45:27.892193] Epoch: [24] [0/8] eta: 0:00:17 lr: 0.000455 loss: 0.3266 (0.3266) time: 2.1514 data: 2.0723 max mem: 9671 +[14:45:28.352936] Epoch: [24] [7/8] eta: 0:00:00 lr: 0.000435 loss: 0.3266 (0.3227) time: 0.3264 data: 0.2593 max mem: 9671 +[14:45:28.427698] Epoch: [24] Total time: 0:00:02 (0.3360 s / it) +[14:45:28.428473] Averaged stats: lr: 0.000435 loss: 0.3266 (0.3227) +[14:45:30.689764] val: [0/2] eta: 0:00:04 loss: 0.1567 (0.1567) time: 2.2496 data: 2.2326 max mem: 9671 +[14:45:30.699934] val: [1/2] eta: 0:00:01 loss: 0.1567 (0.4114) time: 1.1296 data: 1.1164 max mem: 9671 +[14:45:30.768920] val: Total time: 0:00:02 (1.1647 s / it) +[14:45:30.777998] val loss: 0.41143083572387695 +[14:45:30.778243] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9283, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9285, Kappa: 0.6596, Score: 0.8049 +[14:45:30.813836] Best epoch = 17, Best score = 0.8982 +[14:45:31.075646] log_dir: ./output_logs/retfound +[14:45:33.339060] Epoch: [25] [0/8] eta: 0:00:18 lr: 0.000432 loss: 0.2123 (0.2123) time: 2.2623 data: 2.1915 max mem: 9671 +[14:45:33.801082] Epoch: [25] [7/8] eta: 0:00:00 lr: 0.000412 loss: 0.2637 (0.2978) time: 0.3405 data: 0.2740 max mem: 9671 +[14:45:33.976237] Epoch: [25] Total time: 0:00:02 (0.3626 s / it) +[14:45:33.977030] Averaged stats: lr: 0.000412 loss: 0.2637 (0.2978) +[14:45:36.178222] val: [0/2] eta: 0:00:04 loss: 0.1612 (0.1612) time: 2.1908 data: 2.1735 max mem: 9671 +[14:45:36.188588] val: [1/2] eta: 0:00:01 loss: 0.1612 (0.5095) time: 1.1003 data: 1.0868 max mem: 9671 +[14:45:36.258467] val: Total time: 0:00:02 (1.1359 s / it) +[14:45:36.267368] val loss: 0.5094860941171646 +[14:45:36.267558] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9319, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9306, Kappa: 0.6596, Score: 0.8061 +[14:45:36.312370] Best epoch = 17, Best score = 0.8982 +[14:45:36.563127] log_dir: ./output_logs/retfound +[14:45:38.763986] Epoch: [26] [0/8] eta: 0:00:17 lr: 0.000409 loss: 0.3678 (0.3678) time: 2.2001 data: 2.1278 max mem: 9671 +[14:45:39.224436] Epoch: [26] [7/8] eta: 0:00:00 lr: 0.000389 loss: 0.2022 (0.2613) time: 0.3325 data: 0.2660 max mem: 9671 +[14:45:39.290983] Epoch: [26] Total time: 0:00:02 (0.3410 s / it) +[14:45:39.291758] Averaged stats: lr: 0.000389 loss: 0.2022 (0.2613) +[14:45:41.500497] val: [0/2] eta: 0:00:04 loss: 0.1778 (0.1778) time: 2.1954 data: 2.1783 max mem: 9671 +[14:45:41.510634] val: [1/2] eta: 0:00:01 loss: 0.1778 (0.5650) time: 1.1025 data: 1.0892 max mem: 9671 +[14:45:41.578974] val: Total time: 0:00:02 (1.1373 s / it) +[14:45:41.587880] val loss: 0.5649861097335815 +[14:45:41.588162] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9292, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9234, Kappa: 0.6596, Score: 0.8052 +[14:45:41.623547] Best epoch = 17, Best score = 0.8982 +[14:45:41.896320] log_dir: ./output_logs/retfound +[14:45:44.361180] Epoch: [27] [0/8] eta: 0:00:19 lr: 0.000386 loss: 0.3435 (0.3435) time: 2.4637 data: 2.3919 max mem: 9671 +[14:45:44.821275] Epoch: [27] [7/8] eta: 0:00:00 lr: 0.000365 loss: 0.3040 (0.3233) time: 0.3654 data: 0.2991 max mem: 9671 +[14:45:44.891979] Epoch: [27] Total time: 0:00:02 (0.3744 s / it) +[14:45:44.892770] Averaged stats: lr: 0.000365 loss: 0.3040 (0.3233) +[14:45:47.170264] val: [0/2] eta: 0:00:04 loss: 0.1657 (0.1657) time: 2.2543 data: 2.2355 max mem: 9671 +[14:45:47.181642] val: [1/2] eta: 0:00:01 loss: 0.1657 (0.4328) time: 1.1326 data: 1.1178 max mem: 9671 +[14:45:47.251380] val: Total time: 0:00:02 (1.1681 s / it) +[14:45:47.260871] val loss: 0.43275587260723114 +[14:45:47.261054] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9391, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9252, Kappa: 0.6596, Score: 0.8085 +[14:45:47.295928] Best epoch = 17, Best score = 0.8982 +[14:45:47.700349] log_dir: ./output_logs/retfound +[14:45:49.862930] Epoch: [28] [0/8] eta: 0:00:17 lr: 0.000362 loss: 0.2153 (0.2153) time: 2.1613 data: 2.0903 max mem: 9671 +[14:45:50.383622] Epoch: [28] [7/8] eta: 0:00:00 lr: 0.000341 loss: 0.2252 (0.2592) time: 0.3352 data: 0.2689 max mem: 9671 +[14:45:50.452666] Epoch: [28] Total time: 0:00:02 (0.3440 s / it) +[14:45:50.453485] Averaged stats: lr: 0.000341 loss: 0.2252 (0.2592) +[14:45:52.556744] val: [0/2] eta: 0:00:04 loss: 0.2523 (0.2523) time: 2.0916 data: 2.0723 max mem: 9671 +[14:45:52.571561] val: [1/2] eta: 0:00:01 loss: 0.1857 (0.2190) time: 1.0528 data: 1.0362 max mem: 9671 +[14:45:52.639486] val: Total time: 0:00:02 (1.0875 s / it) +[14:45:52.652571] val loss: 0.21898877620697021 +[14:45:52.652781] Accuracy: 0.9000, F1 Score: 0.8566, ROC AUC: 0.9498, Hamming Loss: 0.1000, + Jaccard Score: 0.7576, Precision: 0.8566, Recall: 0.8566, + Average Precision: 0.9370, Kappa: 0.7133, Score: 0.8399 +[14:45:52.685928] Best epoch = 17, Best score = 0.8982 +[14:45:52.933238] log_dir: ./output_logs/retfound +[14:45:54.867341] Epoch: [29] [0/8] eta: 0:00:15 lr: 0.000337 loss: 0.2674 (0.2674) time: 1.9332 data: 1.8629 max mem: 9671 +[14:45:56.020131] Epoch: [29] [7/8] eta: 0:00:00 lr: 0.000316 loss: 0.2900 (0.2986) time: 0.3857 data: 0.2601 max mem: 9671 +[14:45:56.116877] Epoch: [29] Total time: 0:00:03 (0.3979 s / it) +[14:45:56.124747] Averaged stats: lr: 0.000316 loss: 0.2900 (0.2986) +[14:45:58.446539] val: [0/2] eta: 0:00:04 loss: 0.1922 (0.1922) time: 2.3098 data: 2.2922 max mem: 9671 +[14:45:58.456903] val: [1/2] eta: 0:00:01 loss: 0.1922 (0.2850) time: 1.1598 data: 1.1462 max mem: 9671 +[14:45:58.525578] val: Total time: 0:00:02 (1.1948 s / it) +[14:45:58.534514] val loss: 0.28504690527915955 +[14:45:58.534767] Accuracy: 0.9250, F1 Score: 0.8880, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.8045, Precision: 0.9062, Recall: 0.8728, + Average Precision: 0.9346, Kappa: 0.7761, Score: 0.8713 +[14:45:58.557897] Best epoch = 17, Best score = 0.8982 +[14:45:58.840333] log_dir: ./output_logs/retfound +[14:46:01.078321] Epoch: [30] [0/8] eta: 0:00:17 lr: 0.000313 loss: 0.2006 (0.2006) time: 2.2370 data: 2.1660 max mem: 9671 +[14:46:01.566756] Epoch: [30] [7/8] eta: 0:00:00 lr: 0.000292 loss: 0.2116 (0.2705) time: 0.3406 data: 0.2708 max mem: 9671 +[14:46:01.639144] Epoch: [30] Total time: 0:00:02 (0.3498 s / it) +[14:46:01.639969] Averaged stats: lr: 0.000292 loss: 0.2116 (0.2705) +[14:46:03.971311] val: [0/2] eta: 0:00:04 loss: 0.1967 (0.1967) time: 2.3107 data: 2.2938 max mem: 9671 +[14:46:03.981309] val: [1/2] eta: 0:00:01 loss: 0.1967 (0.3365) time: 1.1601 data: 1.1470 max mem: 9671 +[14:46:04.384002] val: Total time: 0:00:02 (1.3622 s / it) +[14:46:04.392933] val loss: 0.3365369886159897 +[14:46:04.393139] Accuracy: 0.9000, F1 Score: 0.8438, ROC AUC: 0.9498, Hamming Loss: 0.1000, + Jaccard Score: 0.7412, Precision: 0.8831, Recall: 0.8172, + Average Precision: 0.9408, Kappa: 0.6887, Score: 0.8274 +[14:46:04.776833] Best epoch = 17, Best score = 0.8982 +[14:46:06.240456] log_dir: ./output_logs/retfound +[14:46:08.426085] Epoch: [31] [0/8] eta: 0:00:17 lr: 0.000289 loss: 0.5076 (0.5076) time: 2.1846 data: 2.1148 max mem: 9671 +[14:46:09.237977] Epoch: [31] [7/8] eta: 0:00:00 lr: 0.000267 loss: 0.2211 (0.2924) time: 0.3744 data: 0.2644 max mem: 9671 +[14:46:09.358620] Epoch: [31] Total time: 0:00:03 (0.3897 s / it) +[14:46:09.365508] Averaged stats: lr: 0.000267 loss: 0.2211 (0.2924) +[14:46:11.575557] val: [0/2] eta: 0:00:04 loss: 0.1812 (0.1812) time: 2.1989 data: 2.1815 max mem: 9671 +[14:46:11.585736] val: [1/2] eta: 0:00:01 loss: 0.1812 (0.3428) time: 1.1042 data: 1.0908 max mem: 9671 +[14:46:11.659126] val: Total time: 0:00:02 (1.1416 s / it) +[14:46:11.668055] val loss: 0.3427543491125107 +[14:46:11.668289] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9369, Kappa: 0.7561, Score: 0.8598 +[14:46:11.704106] Best epoch = 17, Best score = 0.8982 +[14:46:11.964856] log_dir: ./output_logs/retfound +[14:46:14.139777] Epoch: [32] [0/8] eta: 0:00:17 lr: 0.000264 loss: 0.1763 (0.1763) time: 2.1735 data: 2.0361 max mem: 9671 +[14:46:15.165569] Epoch: [32] [7/8] eta: 0:00:00 lr: 0.000243 loss: 0.3075 (0.2896) time: 0.3998 data: 0.2546 max mem: 9671 +[14:46:15.241483] Epoch: [32] Total time: 0:00:03 (0.4096 s / it) +[14:46:15.255026] Averaged stats: lr: 0.000243 loss: 0.3075 (0.2896) +[14:46:17.508953] val: [0/2] eta: 0:00:04 loss: 0.1718 (0.1718) time: 2.2304 data: 2.1857 max mem: 9671 +[14:46:17.524502] val: [1/2] eta: 0:00:01 loss: 0.1718 (0.3577) time: 1.1226 data: 1.0929 max mem: 9671 +[14:46:17.606345] val: Total time: 0:00:02 (1.1643 s / it) +[14:46:17.615306] val loss: 0.35773538053035736 +[14:46:17.615601] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9357, Kappa: 0.7561, Score: 0.8598 +[14:46:17.653481] Best epoch = 17, Best score = 0.8982 +[14:46:17.900089] log_dir: ./output_logs/retfound +[14:46:20.224936] Epoch: [33] [0/8] eta: 0:00:18 lr: 0.000240 loss: 0.1372 (0.1372) time: 2.3237 data: 2.1773 max mem: 9671 +[14:46:21.012595] Epoch: [33] [7/8] eta: 0:00:00 lr: 0.000220 loss: 0.2494 (0.2693) time: 0.3888 data: 0.2722 max mem: 9671 +[14:46:21.080885] Epoch: [33] Total time: 0:00:03 (0.3976 s / it) +[14:46:21.081681] Averaged stats: lr: 0.000220 loss: 0.2494 (0.2693) +[14:46:23.260675] val: [0/2] eta: 0:00:04 loss: 0.1649 (0.1649) time: 2.1653 data: 2.1480 max mem: 9671 +[14:46:23.270912] val: [1/2] eta: 0:00:01 loss: 0.1649 (0.4252) time: 1.0875 data: 1.0741 max mem: 9671 +[14:46:23.340398] val: Total time: 0:00:02 (1.1229 s / it) +[14:46:23.350469] val loss: 0.42521847784519196 +[14:46:23.350678] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9462, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9409, Kappa: 0.6596, Score: 0.8109 +[14:46:23.388164] Best epoch = 17, Best score = 0.8982 +[14:46:23.643006] log_dir: ./output_logs/retfound +[14:46:25.897356] Epoch: [34] [0/8] eta: 0:00:18 lr: 0.000217 loss: 0.2293 (0.2293) time: 2.2529 data: 2.0959 max mem: 9671 +[14:46:26.923258] Epoch: [34] [7/8] eta: 0:00:00 lr: 0.000196 loss: 0.2456 (0.2689) time: 0.4097 data: 0.2621 max mem: 9671 +[14:46:26.992781] Epoch: [34] Total time: 0:00:03 (0.4187 s / it) +[14:46:27.001506] Averaged stats: lr: 0.000196 loss: 0.2456 (0.2689) +[14:46:29.298029] val: [0/2] eta: 0:00:04 loss: 0.1657 (0.1657) time: 2.2774 data: 2.2245 max mem: 9671 +[14:46:29.323101] val: [1/2] eta: 0:00:01 loss: 0.1657 (0.4029) time: 1.1509 data: 1.1123 max mem: 9671 +[14:46:29.390143] val: Total time: 0:00:02 (1.1851 s / it) +[14:46:29.401289] val loss: 0.4029083847999573 +[14:46:29.401475] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9391, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9353, Kappa: 0.7561, Score: 0.8574 +[14:46:29.448159] Best epoch = 17, Best score = 0.8982 +[14:46:29.738212] log_dir: ./output_logs/retfound +[14:46:32.029006] Epoch: [35] [0/8] eta: 0:00:18 lr: 0.000194 loss: 0.4496 (0.4496) time: 2.2896 data: 2.2069 max mem: 9671 +[14:46:32.612539] Epoch: [35] [7/8] eta: 0:00:00 lr: 0.000174 loss: 0.2822 (0.2812) time: 0.3590 data: 0.2759 max mem: 9671 +[14:46:32.679533] Epoch: [35] Total time: 0:00:02 (0.3676 s / it) +[14:46:32.688037] Averaged stats: lr: 0.000174 loss: 0.2822 (0.2812) +[14:46:34.991405] val: [0/2] eta: 0:00:04 loss: 0.1606 (0.1606) time: 2.2829 data: 2.2461 max mem: 9671 +[14:46:35.009470] val: [1/2] eta: 0:00:01 loss: 0.1606 (0.3562) time: 1.1501 data: 1.1231 max mem: 9671 +[14:46:35.090141] val: Total time: 0:00:02 (1.1913 s / it) +[14:46:35.101004] val loss: 0.3561532497406006 +[14:46:35.101244] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9391, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9353, Kappa: 0.7561, Score: 0.8574 +[14:46:35.139700] Best epoch = 17, Best score = 0.8982 +[14:46:35.415005] log_dir: ./output_logs/retfound +[14:46:37.705377] Epoch: [36] [0/8] eta: 0:00:18 lr: 0.000171 loss: 0.4470 (0.4470) time: 2.2894 data: 2.1373 max mem: 9671 +[14:46:38.673918] Epoch: [36] [7/8] eta: 0:00:00 lr: 0.000153 loss: 0.2518 (0.2885) time: 0.4071 data: 0.2673 max mem: 9671 +[14:46:38.748405] Epoch: [36] Total time: 0:00:03 (0.4167 s / it) +[14:46:38.757002] Averaged stats: lr: 0.000153 loss: 0.2518 (0.2885) +[14:46:40.976048] val: [0/2] eta: 0:00:04 loss: 0.1652 (0.1652) time: 2.1968 data: 2.1798 max mem: 9671 +[14:46:40.986329] val: [1/2] eta: 0:00:01 loss: 0.1652 (0.2911) time: 1.1033 data: 1.0900 max mem: 9671 +[14:46:41.076670] val: Total time: 0:00:02 (1.1491 s / it) +[14:46:41.086719] val loss: 0.2910751849412918 +[14:46:41.087028] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9462, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9409, Kappa: 0.7561, Score: 0.8598 +[14:46:41.135433] Best epoch = 17, Best score = 0.8982 +[14:46:41.426653] log_dir: ./output_logs/retfound +[14:46:43.892199] Epoch: [37] [0/8] eta: 0:00:19 lr: 0.000150 loss: 0.3017 (0.3017) time: 2.4645 data: 2.2298 max mem: 9671 +[14:46:44.907425] Epoch: [37] [7/8] eta: 0:00:00 lr: 0.000132 loss: 0.2246 (0.2802) time: 0.4349 data: 0.2789 max mem: 9671 +[14:46:44.982012] Epoch: [37] Total time: 0:00:03 (0.4444 s / it) +[14:46:44.991268] Averaged stats: lr: 0.000132 loss: 0.2246 (0.2802) +[14:46:47.203486] val: [0/2] eta: 0:00:04 loss: 0.1532 (0.1532) time: 2.2012 data: 2.1641 max mem: 9671 +[14:46:47.219052] val: [1/2] eta: 0:00:01 loss: 0.1532 (0.3142) time: 1.1081 data: 1.0821 max mem: 9671 +[14:46:47.285182] val: Total time: 0:00:02 (1.1419 s / it) +[14:46:47.294151] val loss: 0.3142062872648239 +[14:46:47.294340] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9431, Kappa: 0.7561, Score: 0.8609 +[14:46:47.332788] Best epoch = 17, Best score = 0.8982 +[14:46:47.602102] log_dir: ./output_logs/retfound +[14:46:49.624289] Epoch: [38] [0/8] eta: 0:00:16 lr: 0.000130 loss: 0.1331 (0.1331) time: 2.0211 data: 1.8784 max mem: 9671 +[14:46:50.695189] Epoch: [38] [7/8] eta: 0:00:00 lr: 0.000113 loss: 0.2899 (0.2815) time: 0.3864 data: 0.2393 max mem: 9671 +[14:46:50.804453] Epoch: [38] Total time: 0:00:03 (0.4003 s / it) +[14:46:50.812539] Averaged stats: lr: 0.000113 loss: 0.2899 (0.2815) +[14:46:52.931143] val: [0/2] eta: 0:00:04 loss: 0.1554 (0.1554) time: 2.1071 data: 2.0729 max mem: 9671 +[14:46:52.947169] val: [1/2] eta: 0:00:01 loss: 0.1554 (0.2977) time: 1.0612 data: 1.0365 max mem: 9671 +[14:46:53.016490] val: Total time: 0:00:02 (1.0966 s / it) +[14:46:53.025603] val loss: 0.2977108359336853 +[14:46:53.025868] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9534, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9454, Kappa: 0.7561, Score: 0.8621 +[14:46:53.075284] Best epoch = 17, Best score = 0.8982 +[14:46:53.314895] log_dir: ./output_logs/retfound +[14:46:55.800225] Epoch: [39] [0/8] eta: 0:00:19 lr: 0.000110 loss: 0.2601 (0.2601) time: 2.4842 data: 2.3375 max mem: 9671 +[14:46:56.635436] Epoch: [39] [7/8] eta: 0:00:00 lr: 0.000095 loss: 0.2684 (0.2860) time: 0.4148 data: 0.2923 max mem: 9671 +[14:46:56.721223] Epoch: [39] Total time: 0:00:03 (0.4258 s / it) +[14:46:56.730876] Averaged stats: lr: 0.000095 loss: 0.2684 (0.2860) +[14:46:58.908348] val: [0/2] eta: 0:00:04 loss: 0.1555 (0.1555) time: 2.1663 data: 2.1307 max mem: 9671 +[14:46:58.924337] val: [1/2] eta: 0:00:01 loss: 0.1555 (0.2966) time: 1.0908 data: 1.0654 max mem: 9671 +[14:46:59.006128] val: Total time: 0:00:02 (1.1324 s / it) +[14:46:59.015003] val loss: 0.2965516448020935 +[14:46:59.015198] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9534, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9454, Kappa: 0.7561, Score: 0.8621 +[14:46:59.057477] Best epoch = 17, Best score = 0.8982 +[14:46:59.334182] log_dir: ./output_logs/retfound +[14:47:01.655951] Epoch: [40] [0/8] eta: 0:00:18 lr: 0.000092 loss: 0.2017 (0.2017) time: 2.3206 data: 2.1756 max mem: 9671 +[14:47:02.763096] Epoch: [40] [7/8] eta: 0:00:00 lr: 0.000078 loss: 0.2769 (0.2756) time: 0.4284 data: 0.2833 max mem: 9671 +[14:47:02.845771] Epoch: [40] Total time: 0:00:03 (0.4389 s / it) +[14:47:02.863928] Averaged stats: lr: 0.000078 loss: 0.2769 (0.2756) +[14:47:05.010677] val: [0/2] eta: 0:00:04 loss: 0.1451 (0.1451) time: 2.1360 data: 2.1190 max mem: 9671 +[14:47:05.023038] val: [1/2] eta: 0:00:01 loss: 0.1451 (0.3349) time: 1.0736 data: 1.0596 max mem: 9671 +[14:47:05.092506] val: Total time: 0:00:02 (1.1092 s / it) +[14:47:05.101525] val loss: 0.3348654955625534 +[14:47:05.101811] Accuracy: 0.9250, F1 Score: 0.8769, ROC AUC: 0.9498, Hamming Loss: 0.0750, + Jaccard Score: 0.7892, Precision: 0.9559, Recall: 0.8333, + Average Precision: 0.9424, Kappa: 0.7561, Score: 0.8609 +[14:47:05.138345] Best epoch = 17, Best score = 0.8982 +[14:47:05.426226] log_dir: ./output_logs/retfound +[14:47:07.641722] Epoch: [41] [0/8] eta: 0:00:17 lr: 0.000076 loss: 0.2176 (0.2176) time: 2.2145 data: 2.0780 max mem: 9671 +[14:47:08.653024] Epoch: [41] [7/8] eta: 0:00:00 lr: 0.000062 loss: 0.2356 (0.2808) time: 0.4031 data: 0.2599 max mem: 9671 +[14:47:08.726807] Epoch: [41] Total time: 0:00:03 (0.4126 s / it) +[14:47:08.736122] Averaged stats: lr: 0.000062 loss: 0.2356 (0.2808) +[14:47:11.066478] val: [0/2] eta: 0:00:04 loss: 0.1402 (0.1402) time: 2.3191 data: 2.2739 max mem: 9671 +[14:47:11.086282] val: [1/2] eta: 0:00:01 loss: 0.1402 (0.3686) time: 1.1691 data: 1.1370 max mem: 9671 +[14:47:11.153655] val: Total time: 0:00:02 (1.2036 s / it) +[14:47:11.163271] val loss: 0.3686094284057617 +[14:47:11.163618] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9534, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9454, Kappa: 0.6596, Score: 0.8133 +[14:47:11.220873] Best epoch = 17, Best score = 0.8982 +[14:47:11.473395] log_dir: ./output_logs/retfound +[14:47:13.792479] Epoch: [42] [0/8] eta: 0:00:18 lr: 0.000061 loss: 0.2027 (0.2027) time: 2.3179 data: 2.1771 max mem: 9671 +[14:47:14.484152] Epoch: [42] [7/8] eta: 0:00:00 lr: 0.000049 loss: 0.2027 (0.2439) time: 0.3761 data: 0.2722 max mem: 9671 +[14:47:14.559265] Epoch: [42] Total time: 0:00:03 (0.3857 s / it) +[14:47:14.560044] Averaged stats: lr: 0.000049 loss: 0.2027 (0.2439) +[14:47:16.904069] val: [0/2] eta: 0:00:04 loss: 0.1391 (0.1391) time: 2.3285 data: 2.2918 max mem: 9671 +[14:47:16.919861] val: [1/2] eta: 0:00:01 loss: 0.1391 (0.4043) time: 1.1719 data: 1.1460 max mem: 9671 +[14:47:17.105574] val: Total time: 0:00:02 (1.2654 s / it) +[14:47:17.119156] val loss: 0.4043458551168442 +[14:47:17.119423] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9534, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9454, Kappa: 0.6596, Score: 0.8133 +[14:47:17.270751] Best epoch = 17, Best score = 0.8982 +[14:47:18.146875] log_dir: ./output_logs/retfound +[14:47:20.518369] Epoch: [43] [0/8] eta: 0:00:18 lr: 0.000047 loss: 0.1854 (0.1854) time: 2.3704 data: 2.2214 max mem: 9671 +[14:47:21.550348] Epoch: [43] [7/8] eta: 0:00:00 lr: 0.000036 loss: 0.2205 (0.2380) time: 0.4252 data: 0.2778 max mem: 9671 +[14:47:21.620536] Epoch: [43] Total time: 0:00:03 (0.4342 s / it) +[14:47:21.625947] Averaged stats: lr: 0.000036 loss: 0.2205 (0.2380) +[14:47:24.115284] val: [0/2] eta: 0:00:04 loss: 0.1402 (0.1402) time: 2.4777 data: 2.4605 max mem: 9671 +[14:47:24.125529] val: [1/2] eta: 0:00:01 loss: 0.1402 (0.4161) time: 1.2437 data: 1.2303 max mem: 9671 +[14:47:24.194464] val: Total time: 0:00:02 (1.2788 s / it) +[14:47:24.203427] val loss: 0.416122242808342 +[14:47:24.203674] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[14:47:24.243114] Best epoch = 17, Best score = 0.8982 +[14:47:24.540745] log_dir: ./output_logs/retfound +[14:47:26.958252] Epoch: [44] [0/8] eta: 0:00:19 lr: 0.000035 loss: 0.3020 (0.3020) time: 2.4164 data: 2.2806 max mem: 9671 +[14:47:27.990656] Epoch: [44] [7/8] eta: 0:00:00 lr: 0.000026 loss: 0.2097 (0.2575) time: 0.4310 data: 0.2852 max mem: 9671 +[14:47:28.061958] Epoch: [44] Total time: 0:00:03 (0.4401 s / it) +[14:47:28.071329] Averaged stats: lr: 0.000026 loss: 0.2097 (0.2575) +[14:47:30.329114] val: [0/2] eta: 0:00:04 loss: 0.1411 (0.1411) time: 2.2455 data: 2.2144 max mem: 9671 +[14:47:30.345465] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.4057) time: 1.1306 data: 1.1073 max mem: 9671 +[14:47:30.416689] val: Total time: 0:00:02 (1.1671 s / it) +[14:47:30.425611] val loss: 0.4057278037071228 +[14:47:30.425818] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[14:47:30.476483] Best epoch = 17, Best score = 0.8982 +[14:47:30.728248] log_dir: ./output_logs/retfound +[14:47:33.203316] Epoch: [45] [0/8] eta: 0:00:19 lr: 0.000025 loss: 0.2266 (0.2266) time: 2.4741 data: 2.3996 max mem: 9671 +[14:47:33.664194] Epoch: [45] [7/8] eta: 0:00:00 lr: 0.000017 loss: 0.2266 (0.2669) time: 0.3668 data: 0.3000 max mem: 9671 +[14:47:33.734015] Epoch: [45] Total time: 0:00:03 (0.3757 s / it) +[14:47:33.734739] Averaged stats: lr: 0.000017 loss: 0.2266 (0.2669) +[14:47:36.185405] val: [0/2] eta: 0:00:04 loss: 0.1414 (0.1414) time: 2.4338 data: 2.3823 max mem: 9671 +[14:47:36.208196] val: [1/2] eta: 0:00:01 loss: 0.1414 (0.3876) time: 1.2280 data: 1.1912 max mem: 9671 +[14:47:36.276833] val: Total time: 0:00:02 (1.2631 s / it) +[14:47:36.287291] val loss: 0.38761037588119507 +[14:47:36.287482] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[14:47:36.344050] Best epoch = 17, Best score = 0.8982 +[14:47:36.642337] log_dir: ./output_logs/retfound +[14:47:39.572306] Epoch: [46] [0/8] eta: 0:00:23 lr: 0.000016 loss: 0.2934 (0.2934) time: 2.9287 data: 2.7192 max mem: 9671 +[14:47:40.661938] Epoch: [46] [7/8] eta: 0:00:00 lr: 0.000010 loss: 0.2377 (0.2666) time: 0.5021 data: 0.3400 max mem: 9671 +[14:47:40.737608] Epoch: [46] Total time: 0:00:04 (0.5119 s / it) +[14:47:40.746983] Averaged stats: lr: 0.000010 loss: 0.2377 (0.2666) +[14:47:43.550484] val: [0/2] eta: 0:00:05 loss: 0.1412 (0.1412) time: 2.7854 data: 2.7667 max mem: 9671 +[14:47:43.560356] val: [1/2] eta: 0:00:01 loss: 0.1412 (0.3810) time: 1.3974 data: 1.3834 max mem: 9671 +[14:47:43.633804] val: Total time: 0:00:02 (1.4347 s / it) +[14:47:43.642807] val loss: 0.3810284584760666 +[14:47:43.642995] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9570, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9488, Kappa: 0.6596, Score: 0.8145 +[14:47:43.699965] Best epoch = 17, Best score = 0.8982 +[14:47:44.019851] log_dir: ./output_logs/retfound +[14:47:47.041404] Epoch: [47] [0/8] eta: 0:00:24 lr: 0.000010 loss: 0.2115 (0.2115) time: 3.0205 data: 2.8088 max mem: 9671 +[14:47:48.283500] Epoch: [47] [7/8] eta: 0:00:00 lr: 0.000005 loss: 0.2115 (0.2570) time: 0.5327 data: 0.3512 max mem: 9671 +[14:47:48.354461] Epoch: [47] Total time: 0:00:04 (0.5418 s / it) +[14:47:48.362639] Averaged stats: lr: 0.000005 loss: 0.2115 (0.2570) +[14:47:50.965930] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.5814 data: 2.5626 max mem: 9671 +[14:47:50.976091] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3793) time: 1.2955 data: 1.2813 max mem: 9671 +[14:47:51.047363] val: Total time: 0:00:02 (1.3319 s / it) +[14:47:51.056317] val loss: 0.3793119490146637 +[14:47:51.056499] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[14:47:51.114332] Best epoch = 17, Best score = 0.8982 +[14:47:51.408286] log_dir: ./output_logs/retfound +[14:47:54.175256] Epoch: [48] [0/8] eta: 0:00:22 lr: 0.000005 loss: 0.3700 (0.3700) time: 2.7657 data: 2.5923 max mem: 9671 +[14:47:55.690434] Epoch: [48] [7/8] eta: 0:00:00 lr: 0.000002 loss: 0.2148 (0.2645) time: 0.5350 data: 0.3242 max mem: 9671 +[14:47:55.773564] Epoch: [48] Total time: 0:00:04 (0.5456 s / it) +[14:47:55.791243] Averaged stats: lr: 0.000002 loss: 0.2148 (0.2645) +[14:47:58.260943] val: [0/2] eta: 0:00:04 loss: 0.1411 (0.1411) time: 2.4578 data: 2.4210 max mem: 9671 +[14:47:58.276828] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3780) time: 1.2365 data: 1.2106 max mem: 9671 +[14:47:58.354560] val: Total time: 0:00:02 (1.2761 s / it) +[14:47:58.363609] val loss: 0.3780297338962555 +[14:47:58.363788] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[14:47:58.416953] Best epoch = 17, Best score = 0.8982 +[14:47:58.711026] log_dir: ./output_logs/retfound +[14:48:01.068611] Epoch: [49] [0/8] eta: 0:00:18 lr: 0.000002 loss: 0.1158 (0.1158) time: 2.3565 data: 2.2856 max mem: 9671 +[14:48:02.367792] Epoch: [49] [7/8] eta: 0:00:00 lr: 0.000001 loss: 0.2527 (0.2805) time: 0.4568 data: 0.3179 max mem: 9671 +[14:48:02.442718] Epoch: [49] Total time: 0:00:03 (0.4664 s / it) +[14:48:02.456958] Averaged stats: lr: 0.000001 loss: 0.2527 (0.2805) +[14:48:05.126834] val: [0/2] eta: 0:00:05 loss: 0.1411 (0.1411) time: 2.5627 data: 2.5373 max mem: 9671 +[14:48:05.142610] val: [1/2] eta: 0:00:01 loss: 0.1411 (0.3775) time: 1.2889 data: 1.2687 max mem: 9671 +[14:48:05.211536] val: Total time: 0:00:02 (1.3241 s / it) +[14:48:05.220415] val loss: 0.37748883664608 +[14:48:05.220616] Accuracy: 0.9000, F1 Score: 0.8268, ROC AUC: 0.9606, Hamming Loss: 0.1000, + Jaccard Score: 0.7206, Precision: 0.9429, Recall: 0.7778, + Average Precision: 0.9527, Kappa: 0.6596, Score: 0.8157 +[14:48:05.331387] Best epoch = 17, Best score = 0.8982 +[14:48:08.855145] Test with the best model, epoch = 17: +[14:48:11.803810] test: [0/3] eta: 0:00:08 loss: 0.2582 (0.2582) time: 2.9031 data: 2.8685 max mem: 9671 +[14:48:12.043622] test: [2/3] eta: 0:00:01 loss: 0.2582 (0.2454) time: 1.0475 data: 0.9563 max mem: 9671 +[14:48:12.110251] test: Total time: 0:00:03 (1.0701 s / it) +[14:48:12.120181] val loss: 0.24544517199198404 +[14:48:12.120318] Accuracy: 0.9250, F1 Score: 0.8925, ROC AUC: 0.9516, Hamming Loss: 0.0750, + Jaccard Score: 0.8110, Precision: 0.8925, Recall: 0.8925, + Average Precision: 0.9512, Kappa: 0.7849, Score: 0.8763 +[14:48:12.895970] Training time 0:05:23 +[rank0]:[W701 14:48:13.415212597 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) diff --git a/results/downsample/adam/100/vit/confusion_matrix.png b/results/downsample/adam/100/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..e8a208602b1874377b6b866b683dc2f461be78b7 --- /dev/null +++ b/results/downsample/adam/100/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01b849ecc3b93121314108bdaf949eb826f6d28918114264525636aa6f71afa1 +size 68003 diff --git a/results/downsample/adam/100/vit/log.csv b/results/downsample/adam/100/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..108663060fd79a2a3e24e75f63a49ebfdb6b92bb --- /dev/null +++ b/results/downsample/adam/100/vit/log.csv @@ -0,0 +1,33 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8326617628335953,0.5,0.37275985663082434,0.20603382384588553,5.545808836528073e-08 +1,0.7702071815729141,0.525,0.7204301075268817,0.45818926807259003,1.29402206185655e-07 +2,0.6546564847230911,0.7,0.7849462365591398,0.5773837351677463,2.0334632400602932e-07 +3,0.6182350814342499,0.75,0.870967741935484,0.6940345499098299,2.7729044182640365e-07 +4,0.5662039071321487,0.75,0.8960573476702509,0.6544608924687236,3.512345596467779e-07 +5,0.5842887908220291,0.725,0.8745519713261649,0.6756845538858233,3.694672444929302e-07 +6,0.5582687333226204,0.825,0.899641577060932,0.7416449759260039,3.683426684987483e-07 +7,0.4713408350944519,0.775,0.9175627240143369,0.7299333069266117,3.663241848222764e-07 +8,0.5112659633159637,0.775,0.9175627240143369,0.7299333069266117,3.6342162731308893e-07 +9,0.465877428650856,0.9,0.9139784946236559,0.8279569892473119,3.5964913693960667e-07 +10,0.4645930454134941,0.75,0.9247311827956989,0.6824673300830383,3.550250928957199e-07 +11,0.4524441137909889,0.775,0.946236559139785,0.7109049537056887,3.495720230592029e-07 +12,0.4489521011710167,0.8,0.942652329749104,0.7321708645126134,3.4331649423815874e-07 +13,0.4123004525899887,0.825,0.946236559139785,0.7571766366189548,3.362889827402e-07 +14,0.412119522690773,0.9,0.9569892473118279,0.8422939068100358,3.285237258949416e-07 +15,0.38696958124637604,0.8,0.9641577060931898,0.7667567252402153,3.2005855525317277e-07 +16,0.3891528472304344,0.9,0.946236559139785,0.8387096774193549,3.1093471227534435e-07 +17,0.38670214265584946,0.775,0.9283154121863799,0.7049312380545537,3.0119664740731875e-07 +18,0.4111583083868027,0.925,0.935483870967742,0.866519485341882,2.908918035222662e-07 +19,0.40045975893735886,0.875,0.942652329749104,0.8084207298721364,2.800703847837553e-07 +20,0.38688723742961884,0.85,0.96415770609319,0.7884955365825584,2.687851120561149e-07 +21,0.3998527154326439,0.925,0.9605734767025089,0.8748826872534708,2.570909660536824e-07 +22,0.37270648032426834,0.875,0.9569892473118279,0.8131997023930445,2.450449194802903e-07 +23,0.36275216937065125,0.9,0.9569892473118279,0.8422939068100358,2.3270565946397963e-07 +24,0.39110884070396423,0.825,0.9498207885304659,0.7723605633561171,2.2013330163921197e-07 +25,0.39257583022117615,0.875,0.9283154121863799,0.8036417573512283,2.0738909726954043e-07 +26,0.3709259107708931,0.875,0.9283154121863799,0.8036417573512283,1.9453513483761127e-07 +27,0.35911235213279724,0.9,0.931899641577061,0.8339307048984468,1.8163403755631687e-07 +28,0.350225068628788,0.9,0.935483870967742,0.8351254480286738,1.6874865827479806e-07 +29,0.3573886975646019,0.875,0.9390681003584229,0.8072259867419094,1.5594177326568057e-07 +30,0.3484802022576332,0.925,0.9318996415770608,0.8653247422116549,1.4327577638538533e-07 +31,0.3666532337665558,0.875,0.942652329749104,0.8084207298721364,1.3081237509753143e-07 diff --git a/results/downsample/adam/100/vit/metrics.json b/results/downsample/adam/100/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..fa387bcb8fd7425d05b0b97e4264e6d762a6a6bf --- /dev/null +++ b/results/downsample/adam/100/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 80, + "n_classes": 2, + "task": "binary", + "accuracy": 0.9125, + "balanced_accuracy": 0.8844086021505376, + "precision_macro": 0.8701466781708369, + "recall_macro": 0.8844086021505376, + "f1_macro": 0.8769501208525599, + "precision_weighted": 0.9145168248490079, + "recall_weighted": 0.9125, + "f1_weighted": 0.9133267413755218, + "cohen_kappa": 0.7539543057996485, + "quadratic_weighted_kappa": 0.7539543057996485, + "mcc": 0.7544204852635336, + "auroc": 0.9318996415770608, + "auprc": 0.885447451927631, + "sensitivity": 0.8333333333333334, + "specificity": 0.9354838709677419, + "precision_pos": 0.7894736842105263, + "f1_pos": 0.8108108108108109, + "per_class": { + "0": { + "precision": 0.9508196721311475, + "recall": 0.9354838709677419, + "f1-score": 0.943089430894309, + "support": 62.0 + }, + "1": { + "precision": 0.7894736842105263, + "recall": 0.8333333333333334, + "f1-score": 0.8108108108108109, + "support": 18.0 + }, + "accuracy": 0.9125, + "macro avg": { + "precision": 0.8701466781708369, + "recall": 0.8844086021505376, + "f1-score": 0.8769501208525599, + "support": 80.0 + }, + "weighted avg": { + "precision": 0.9145168248490079, + "recall": 0.9125, + "f1-score": 0.9133267413755218, + "support": 80.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/adam/100/vit/pr.png b/results/downsample/adam/100/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..3c111f8607c642cf0cc3ca352ab966538d8de8d4 --- /dev/null +++ b/results/downsample/adam/100/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:822ccfdcbfbfa2cb1fcc682125bb453ffb6b3cfd7ca82926e12dbaf4afe3be8a +size 42694 diff --git a/results/downsample/adam/100/vit/roc.png b/results/downsample/adam/100/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..c10d1e438b63c7a1b8a483ece9aff81790fdd0c6 --- /dev/null +++ b/results/downsample/adam/100/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:56c1ee837c540ab36ab93e9471f702c9d8acbfe194d32fd75c878f8d0686c718 +size 57102 diff --git a/results/downsample/adam/100/vit/test_pred.npz b/results/downsample/adam/100/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..60e79dbf75d78b507400dc51b7bc8c48163de570 --- /dev/null +++ b/results/downsample/adam/100/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f70c97d3201c9d3359e705ebe55c3068c2b42b85f9d461b377bf407028d9864 +size 1790 diff --git a/results/downsample/adam/100/vit/train.log b/results/downsample/adam/100/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b1cf125d8dff06204a14a6fd714ae159daf85907 --- /dev/null +++ b/results/downsample/adam/100/vit/train.log @@ -0,0 +1,169 @@ +[vit] train=280 val=40 test=80 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8327 val_acc=0.5000 val_auc=0.3728 score=0.2060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7702 val_acc=0.5250 val_auc=0.7204 score=0.4582 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6547 val_acc=0.7000 val_auc=0.7849 score=0.5774 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6182 val_acc=0.7500 val_auc=0.8710 score=0.6940 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5662 val_acc=0.7500 val_auc=0.8961 score=0.6545 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5843 val_acc=0.7250 val_auc=0.8746 score=0.6757 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5583 val_acc=0.8250 val_auc=0.8996 score=0.7416 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4713 val_acc=0.7750 val_auc=0.9176 score=0.7299 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5113 val_acc=0.7750 val_auc=0.9176 score=0.7299 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4659 val_acc=0.9000 val_auc=0.9140 score=0.8280 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4646 val_acc=0.7500 val_auc=0.9247 score=0.6825 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4524 val_acc=0.7750 val_auc=0.9462 score=0.7109 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.4490 val_acc=0.8000 val_auc=0.9427 score=0.7322 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4123 val_acc=0.8250 val_auc=0.9462 score=0.7572 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4121 val_acc=0.9000 val_auc=0.9570 score=0.8423 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.3870 val_acc=0.8000 val_auc=0.9642 score=0.7668 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3892 val_acc=0.9000 val_auc=0.9462 score=0.8387 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.3867 val_acc=0.7750 val_auc=0.9283 score=0.7049 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4112 val_acc=0.9250 val_auc=0.9355 score=0.8665 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4005 val_acc=0.8750 val_auc=0.9427 score=0.8084 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3869 val_acc=0.8500 val_auc=0.9642 score=0.7885 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.3999 val_acc=0.9250 val_auc=0.9606 score=0.8749 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3727 val_acc=0.8750 val_auc=0.9570 score=0.8132 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.3628 val_acc=0.9000 val_auc=0.9570 score=0.8423 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3911 val_acc=0.8250 val_auc=0.9498 score=0.7724 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3926 val_acc=0.8750 val_auc=0.9283 score=0.8036 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.3709 val_acc=0.8750 val_auc=0.9283 score=0.8036 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3591 val_acc=0.9000 val_auc=0.9319 score=0.8339 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3502 val_acc=0.9000 val_auc=0.9355 score=0.8351 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3574 val_acc=0.8750 val_auc=0.9391 score=0.8072 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.3485 val_acc=0.9250 val_auc=0.9319 score=0.8653 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.3667 val_acc=0.8750 val_auc=0.9427 score=0.8084 +[vit] early stop at ep31 (best ep21 score=0.8749) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=21 best_val_score=0.8749 -> saved test_pred.npz (80 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/adam/100/vit acc=0.9125 auroc=0.9318996415770608 f1_macro=0.8770 qwk=0.7539543057996485 diff --git a/results/downsample/airogs/005/airogs_005pct/confusion_matrix_test.jpg b/results/downsample/airogs/005/airogs_005pct/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..fbbaaa4a09a1798f4b9c66ba821b80c284e3c4d9 --- /dev/null +++ b/results/downsample/airogs/005/airogs_005pct/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf484684d28e34fdb0453e82efd96e3957ce6dc0a82bb11920bd0c918cdba986 +size 254047 diff --git a/results/downsample/airogs/005/airogs_005pct/log.txt b/results/downsample/airogs/005/airogs_005pct/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..7bb0150b78124350fcdeac830d34cf4f436bea3e --- /dev/null +++ b/results/downsample/airogs/005/airogs_005pct/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 2.678571428571429e-05, "train_loss": 0.6924155099051339, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.928571428571429e-05, "train_loss": 0.6942640032087054, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.0001517857142857143, "train_loss": 0.689117431640625, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.0002142857142857143, "train_loss": 0.6861659458705357, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002767857142857143, "train_loss": 0.6811490740094867, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00033928571428571433, "train_loss": 0.6649998256138393, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040178571428571433, "train_loss": 0.6465181623186383, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046428571428571433, "train_loss": 0.6433522360665458, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005267857142857142, "train_loss": 0.6222888401576451, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005892857142857142, "train_loss": 0.6337502343314034, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006239798688517498, "train_loss": 0.6522892543247768, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006168712606453664, "train_loss": 0.5729876926967076, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006022803335109993, "train_loss": 0.5561528205871582, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005805663644992073, "train_loss": 0.6058855056762695, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005522640235915924, "train_loss": 0.5409190314156669, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005180702083513844, "train_loss": 0.5607307468141828, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00047882688399256346, "train_loss": 0.5483691351754325, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004355003514019706, "train_loss": 0.49768269062042236, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003891574536045035, "train_loss": 0.49430642809186665, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00034093930654719606, "train_loss": 0.4877063717160906, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002920332010374543, "train_loss": 0.4732391153063093, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024364336770298683, "train_loss": 0.45428013801574707, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019696132483713736, "train_loss": 0.4499066982950483, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015313653926408128, "train_loss": 0.4541608010019575, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011324812265075157, "train_loss": 0.46959154094968525, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.82782601962334e-05, "train_loss": 0.4867276208741324, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.908802597247657e-05, "train_loss": 0.4508460589817592, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.6396180422903234e-05, "train_loss": 0.4851775680269514, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.076147209486148e-05, "train_loss": 0.4269947239330837, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.5688793960098128e-06, "train_loss": 0.49186116456985474, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/005/airogs_005pct/metrics_test.csv b/results/downsample/airogs/005/airogs_005pct/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..b316e418dd518b2e7f2ed841e54b497e8dbe10f2 --- /dev/null +++ b/results/downsample/airogs/005/airogs_005pct/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.4487141091376543,0.834,0.8339574931182383,0.9092690000000001,0.166,0.7152122415949386,0.8343423665833813,0.834,0.9036981691740603,0.6679999999999999 diff --git a/results/downsample/airogs/005/airogs_005pct/metrics_val.csv b/results/downsample/airogs/005/airogs_005pct/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..6168fdfa37fb1efc9342368bf8e975e7f1d5b6e6 --- /dev/null +++ b/results/downsample/airogs/005/airogs_005pct/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6924146133310655,0.5,0.3333333333333333,0.6822016460905349,0.5,0.25,0.25,0.5,0.6521002697543876,0.0 +0.6892524536918191,0.5074074074074074,0.34958613319809456,0.7130212620027434,0.4925925925925926,0.25927307904919844,0.7518656716417911,0.5074074074074074,0.7110153304680054,0.014814814814814836 +0.6840392035596511,0.6462962962962963,0.6294925909295015,0.7180555555555556,0.3537037037037037,0.46416674342206254,0.678718056137411,0.6462962962962964,0.7195992183416331,0.2925925925925926 +0.6749616335420048,0.6555555555555556,0.63572267920094,0.7376097393689987,0.34444444444444444,0.47169014084507044,0.6988636363636364,0.6555555555555556,0.7435723625423136,0.3111111111111111 +0.6554630819488975,0.6888888888888889,0.6884615384615385,0.7709327846364884,0.3111111111111111,0.5250447227191413,0.6899310344827586,0.6888888888888889,0.7722443763470702,0.37777777777777777 +0.6128452209865346,0.7092592592592593,0.7047135080160635,0.803672839506173,0.29074074074074074,0.5452962685946979,0.7229904547514289,0.7092592592592593,0.8015167162225773,0.4185185185185185 +0.5681110988644993,0.7574074074074074,0.7536144557331486,0.8474142661179698,0.24259259259259258,0.6056058077137063,0.7742979930128195,0.7574074074074074,0.841787960953583,0.5148148148148148 +0.5174009011072271,0.8092592592592592,0.8092429049129727,0.8709979423868313,0.19074074074074074,0.679607376446457,0.8093653516295025,0.8092592592592593,0.8639768042509665,0.6185185185185185 +0.48041173815727234,0.7907407407407407,0.7892773892773892,0.8889437585733881,0.20925925925925926,0.6522536193635953,0.799047619047619,0.7907407407407407,0.8864789763261796,0.5814814814814815 +0.445953393683714,0.7925925925925926,0.7901719424360594,0.8992249657064473,0.2074074074074074,0.653701332574572,0.8067475839852738,0.7925925925925926,0.899917938417784,0.5851851851851853 +0.44857028302024393,0.8222222222222222,0.8220660113130869,0.8935493827160494,0.17777777777777778,0.6979220265188997,0.823357744617587,0.8222222222222222,0.8940204279166715,0.6444444444444444 +0.474556318100761,0.7759259259259259,0.7681326051547379,0.9019067215363511,0.22407407407407406,0.6254869825810077,0.8187847667950363,0.7759259259259259,0.9010987871988609,0.5518518518518518 +0.42532702053294463,0.812962962962963,0.8121739130434783,0.8900925925925925,0.18703703703703703,0.6839250493096647,0.8183118242064876,0.8129629629629629,0.8904308607502669,0.625925925925926 +0.4036855215535444,0.8074074074074075,0.8073413379073757,0.9009122085048011,0.1925925925925926,0.6769407383825814,0.8078296703296703,0.8074074074074074,0.900915006705773,0.6148148148148148 +0.43943337614045425,0.7962962962962963,0.7942386831275721,0.9039711934156379,0.2037037037037037,0.6591862416107382,0.808641975308642,0.7962962962962963,0.9005301045540812,0.5925925925925926 +0.41844501740792217,0.8222222222222222,0.8219780219780219,0.8972153635116598,0.17777777777777778,0.69781438074121,0.8240000000000001,0.8222222222222222,0.8926724618478821,0.6444444444444444 +0.4264329934821409,0.8166666666666667,0.8166509474406243,0.8969410150891632,0.18333333333333332,0.6901218723671063,0.8167753001715266,0.8166666666666667,0.8920939701660362,0.6333333333333333 +0.4286679892855532,0.8222222222222222,0.8215957271075381,0.9027091906721536,0.17777777777777778,0.6973470277987073,0.8268128443430353,0.8222222222222222,0.8971796581965663,0.6444444444444444 +0.4013425085474463,0.837037037037037,0.8369811320754716,0.9068724279835391,0.16296296296296298,0.719674143510769,0.8374999999999999,0.837037037037037,0.9004451326671328,0.674074074074074 +0.4122110434314784,0.8388888888888889,0.8385960813684634,0.9089849108367628,0.16111111111111112,0.7221141110545084,0.8413660167746749,0.8388888888888889,0.9017271445079654,0.6777777777777778 +0.41556279361248016,0.8425925925925926,0.8425488561637492,0.9090363511659809,0.1574074074074074,0.7279435863820722,0.842973674453096,0.8425925925925926,0.9019297934023629,0.6851851851851851 +0.4342325829407748,0.8333333333333334,0.8332212323784163,0.9092729766803842,0.16666666666666666,0.7141445511010729,0.8342319542253521,0.8333333333333334,0.9031963582329268,0.6666666666666667 +0.4333791496122585,0.8444444444444444,0.8444252376836646,0.9127469135802468,0.15555555555555556,0.7307443365695793,0.844614624505929,0.8444444444444444,0.9071689745730391,0.6888888888888889 +0.4288010969758034,0.8444444444444444,0.8444423105941097,0.9143518518518519,0.15555555555555556,0.730766464974369,0.8444633450395083,0.8444444444444444,0.9089499799720633,0.6888888888888889 +0.42460646085879383,0.8462962962962963,0.8461776061776062,0.9176200274348423,0.1537037037037037,0.7333927982181159,0.8473684210526315,0.8462962962962963,0.9124760876945562,0.6925925925925926 +0.44273925222018184,0.8407407407407408,0.8399481658142517,0.9189746227709192,0.15925925925925927,0.7242234243094828,0.8476265114196149,0.8407407407407408,0.9144371264109465,0.6814814814814815 +0.4483623088282697,0.8388888888888889,0.8379547390822955,0.9196742112482854,0.16111111111111112,0.7212963152193905,0.8468877687134051,0.8388888888888889,0.9150024285369017,0.6777777777777778 +0.44778599655803514,0.8407407407407408,0.8398620689655172,0.9201714677640604,0.15925925925925927,0.7241131916899834,0.8483870967741935,0.8407407407407408,0.9157843848635334,0.6814814814814815 +0.44352728939231706,0.8388888888888889,0.8381289386729881,0.9202160493827161,0.16111111111111112,0.7215182764958137,0.8453747326334038,0.8388888888888889,0.9155162553806266,0.6777777777777778 +0.44266577218385306,0.8388888888888889,0.8381289386729881,0.9201680384087791,0.16111111111111112,0.7215182764958137,0.8453747326334038,0.8388888888888889,0.915428978375372,0.6777777777777778 diff --git a/results/downsample/airogs/005/airogs_005pct/test_pred.npz b/results/downsample/airogs/005/airogs_005pct/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..53f07367407e92666de893690d62170fed42b8ba --- /dev/null +++ b/results/downsample/airogs/005/airogs_005pct/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d6bbc0b274d3de3fea7ece8997a523de09b2c88c4033593dc53d72b490273f8 +size 12510 diff --git a/results/downsample/airogs/005/resnet/confusion_matrix.png b/results/downsample/airogs/005/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..08be8364cd4ba6ce90b37c45f758cf05b5707d3a --- /dev/null +++ b/results/downsample/airogs/005/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60feffaaf8be85f98a68652ae48b7f9345b81074a131ff38420ccd0417055733 +size 72841 diff --git a/results/downsample/airogs/005/resnet/log.csv b/results/downsample/airogs/005/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..4a3970d51c90a22f3629f26b55febedc2b12335b --- /dev/null +++ b/results/downsample/airogs/005/resnet/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.696075439453125,0.5370370370370371,0.555480109739369,0.3874193313937293,0.0001111111111111111 +1,0.6909383138020834,0.6037037037037037,0.6346227709190672,0.48179522062618796,0.0002777777777777778 +2,0.6834971110026041,0.6462962962962963,0.6995541838134431,0.5461149314579966,0.0004444444444444444 +3,0.6681289672851562,0.6370370370370371,0.7257201646090535,0.5418297214956594,0.000499248235291948 +4,0.6497917175292969,0.65,0.7403223593964334,0.5608892052408917,0.0004953138276568462 +5,0.6296526590983073,0.6407407407407407,0.743312757201646,0.551099869184536,0.0004880619713346038 +6,0.5892232259114584,0.662962962962963,0.7514266117969821,0.5773192669602676,0.00047759073524153667 +7,0.5692329406738281,0.674074074074074,0.76019890260631,0.5927823502514861,0.0004640417248825666 +8,0.5488793055216471,0.7055555555555556,0.7645610425240054,0.626954245313457,0.0004475981673796898 +9,0.4997193018595378,0.7185185185185186,0.7728257887517147,0.6427886328865636,0.0004284824336394748 +10,0.4378053347269694,0.7333333333333333,0.7912894375857339,0.6636850404073437,0.0004069530311680247 +11,0.3849480946858724,0.7277777777777777,0.7962139917695474,0.658676232623664,0.00038330110820042286 +12,0.35981714725494385,0.7351851851851852,0.8036968449931412,0.6690179479826602,0.0003578465164203134 +13,0.3221238851547241,0.7462962962962963,0.8082098765432099,0.6818759159735222,0.0003309334855145803 +14,0.276999572912852,0.7296296296296296,0.8139711934156378,0.665385184629824,0.0003029259680573527 +15,0.25993892550468445,0.75,0.819122085048011,0.6878848036651087,0.0002742027176757948 +16,0.21072168151537576,0.7685185185185185,0.8295061728395061,0.7112826595258278,0.0002451521670570439 +17,0.18105709056059519,0.7574074074074074,0.8326543209876545,0.7016030458849452,0.0002161671750624672 +18,0.171821395556132,0.7555555555555555,0.8350548696844993,0.7005738454503886,0.0001876397139855047 +19,0.15844536821047464,0.7666666666666667,0.8324074074074075,0.7106731583303351,0.00015995556879882234 +20,0.1539833943049113,0.7666666666666667,0.8341015089163237,0.7110930780261002,0.00013348912007436536 +21,0.13029702007770538,0.762962962962963,0.8361934156378601,0.7081477617419193,0.00010859828112836532 +22,0.13638504097859064,0.7666666666666667,0.8405486968449931,0.7134093946748267,8.561965785773413e-05 +23,0.11042233804861705,0.7722222222222223,0.8463374485596709,0.720957342515704,6.486399672274882e-05 +24,0.12743479510148367,0.7777777777777778,0.8480315500685871,0.7270850299379553,4.661198243425813e-05 +25,0.10533674309651057,0.7796296296296297,0.8495473251028807,0.7293876859704519,3.1110442174477255e-05 +26,0.09973243375619252,0.7814814814814814,0.8503772290809328,0.7315262016984782,1.8569007682777416e-05 +27,0.08014421661694844,0.7833333333333333,0.8516255144032921,0.7337439129241249,9.15728034602989e-06 +28,0.09411432594060898,0.7833333333333333,0.8515843621399176,0.7337063242626755,3.002537630797747e-06 +29,0.12093828121821086,0.7833333333333333,0.8503840877914951,0.7331566799549879,1.8801187394248964e-07 diff --git a/results/downsample/airogs/005/resnet/metrics.json b/results/downsample/airogs/005/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..b6eb999e647aa9df31bed6b7c046d9c720eb07b1 --- /dev/null +++ b/results/downsample/airogs/005/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.774, + "balanced_accuracy": 0.774, + "precision_macro": 0.7747429048146187, + "recall_macro": 0.774, + "f1_macro": 0.7738471206535618, + "precision_weighted": 0.7747429048146187, + "recall_weighted": 0.774, + "f1_weighted": 0.7738471206535618, + "cohen_kappa": 0.548, + "quadratic_weighted_kappa": 0.548, + "mcc": 0.5487424019308351, + "auroc": 0.8552180000000001, + "auprc": 0.8551721093826545, + "sensitivity": 0.8, + "specificity": 0.748, + "precision_pos": 0.7604562737642585, + "f1_pos": 0.7797270955165692, + "per_class": { + "0": { + "precision": 0.7890295358649789, + "recall": 0.748, + "f1-score": 0.7679671457905544, + "support": 500.0 + }, + "1": { + "precision": 0.7604562737642585, + "recall": 0.8, + "f1-score": 0.7797270955165692, + "support": 500.0 + }, + "accuracy": 0.774, + "macro avg": { + "precision": 0.7747429048146187, + "recall": 0.774, + "f1-score": 0.7738471206535618, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.7747429048146187, + "recall": 0.774, + "f1-score": 0.7738471206535618, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/005/resnet/pr.png b/results/downsample/airogs/005/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..570a78dbe9c5fa8a12e81ef633ac5d35f315a7a2 --- /dev/null +++ b/results/downsample/airogs/005/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1153f45e4282312424ecb0c786ab5b4c0ad28f0964814f0309c448eb96bdbc1e +size 55120 diff --git a/results/downsample/airogs/005/resnet/roc.png b/results/downsample/airogs/005/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..9cd0e821c9c65e2cbc24281de85cfea528889843 --- /dev/null +++ b/results/downsample/airogs/005/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7f5a440a72c1809a006d1b18b9b15e45556bcdb295ea88ac952329a90711d02 +size 65594 diff --git a/results/downsample/airogs/005/resnet/test_pred.npz b/results/downsample/airogs/005/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..fa6727b500dc09a3088f0f8496d833946a9ebc5b --- /dev/null +++ b/results/downsample/airogs/005/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe5c3d9e92eb8b055f2c06b0a8ffa3e07b06c00cedc8a229cb0d4ef335dbbaab +size 16510 diff --git a/results/downsample/airogs/005/resnet/train.log b/results/downsample/airogs/005/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..93c7bf42fc36c5c5b9bb876fd4945bca2f04f643 --- /dev/null +++ b/results/downsample/airogs/005/resnet/train.log @@ -0,0 +1,158 @@ +[resnet] train=250 val=540 test=1000 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6961 val_acc=0.5370 val_auc=0.5555 score=0.3874 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6909 val_acc=0.6037 val_auc=0.6346 score=0.4818 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6835 val_acc=0.6463 val_auc=0.6996 score=0.5461 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6681 val_acc=0.6370 val_auc=0.7257 score=0.5418 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6498 val_acc=0.6500 val_auc=0.7403 score=0.5609 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6297 val_acc=0.6407 val_auc=0.7433 score=0.5511 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.5892 val_acc=0.6630 val_auc=0.7514 score=0.5773 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.5692 val_acc=0.6741 val_auc=0.7602 score=0.5928 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.5489 val_acc=0.7056 val_auc=0.7646 score=0.6270 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.4997 val_acc=0.7185 val_auc=0.7728 score=0.6428 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.4378 val_acc=0.7333 val_auc=0.7913 score=0.6637 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.3849 val_acc=0.7278 val_auc=0.7962 score=0.6587 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.3598 val_acc=0.7352 val_auc=0.8037 score=0.6690 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.3221 val_acc=0.7463 val_auc=0.8082 score=0.6819 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.2770 val_acc=0.7296 val_auc=0.8140 score=0.6654 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.2599 val_acc=0.7500 val_auc=0.8191 score=0.6879 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.2107 val_acc=0.7685 val_auc=0.8295 score=0.7113 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.1811 val_acc=0.7574 val_auc=0.8327 score=0.7016 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.1718 val_acc=0.7556 val_auc=0.8351 score=0.7006 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.1584 val_acc=0.7667 val_auc=0.8324 score=0.7107 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.1540 val_acc=0.7667 val_auc=0.8341 score=0.7111 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.1303 val_acc=0.7630 val_auc=0.8362 score=0.7081 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.1364 val_acc=0.7667 val_auc=0.8405 score=0.7134 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.1104 val_acc=0.7722 val_auc=0.8463 score=0.7210 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.1274 val_acc=0.7778 val_auc=0.8480 score=0.7271 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.1053 val_acc=0.7796 val_auc=0.8495 score=0.7294 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0997 val_acc=0.7815 val_auc=0.8504 score=0.7315 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0801 val_acc=0.7833 val_auc=0.8516 score=0.7337 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0941 val_acc=0.7833 val_auc=0.8516 score=0.7337 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.1209 val_acc=0.7833 val_auc=0.8504 score=0.7332 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=27 best_val_score=0.7337 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/005/resnet acc=0.7740 auroc=0.8552180000000001 f1_macro=0.7738 qwk=0.548 diff --git a/results/downsample/airogs/005/retfound/confusion_matrix.png b/results/downsample/airogs/005/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..29d92fff1f0dcee431471b26f7dd7f925ec6fdcb --- /dev/null +++ b/results/downsample/airogs/005/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deb4f99545e0862993d8ebef8223d1b40b90a0db8c572c1d343b4a355feba0e9 +size 71947 diff --git a/results/downsample/airogs/005/retfound/confusion_matrix_test.jpg b/results/downsample/airogs/005/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..fbbaaa4a09a1798f4b9c66ba821b80c284e3c4d9 --- /dev/null +++ b/results/downsample/airogs/005/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf484684d28e34fdb0453e82efd96e3957ce6dc0a82bb11920bd0c918cdba986 +size 254047 diff --git a/results/downsample/airogs/005/retfound/log.txt b/results/downsample/airogs/005/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..7bb0150b78124350fcdeac830d34cf4f436bea3e --- /dev/null +++ b/results/downsample/airogs/005/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 2.678571428571429e-05, "train_loss": 0.6924155099051339, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.928571428571429e-05, "train_loss": 0.6942640032087054, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.0001517857142857143, "train_loss": 0.689117431640625, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.0002142857142857143, "train_loss": 0.6861659458705357, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002767857142857143, "train_loss": 0.6811490740094867, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00033928571428571433, "train_loss": 0.6649998256138393, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040178571428571433, "train_loss": 0.6465181623186383, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046428571428571433, "train_loss": 0.6433522360665458, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005267857142857142, "train_loss": 0.6222888401576451, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005892857142857142, "train_loss": 0.6337502343314034, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006239798688517498, "train_loss": 0.6522892543247768, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006168712606453664, "train_loss": 0.5729876926967076, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006022803335109993, "train_loss": 0.5561528205871582, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005805663644992073, "train_loss": 0.6058855056762695, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005522640235915924, "train_loss": 0.5409190314156669, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005180702083513844, "train_loss": 0.5607307468141828, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00047882688399256346, "train_loss": 0.5483691351754325, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004355003514019706, "train_loss": 0.49768269062042236, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003891574536045035, "train_loss": 0.49430642809186665, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00034093930654719606, "train_loss": 0.4877063717160906, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002920332010374543, "train_loss": 0.4732391153063093, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024364336770298683, "train_loss": 0.45428013801574707, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019696132483713736, "train_loss": 0.4499066982950483, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015313653926408128, "train_loss": 0.4541608010019575, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011324812265075157, "train_loss": 0.46959154094968525, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.82782601962334e-05, "train_loss": 0.4867276208741324, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.908802597247657e-05, "train_loss": 0.4508460589817592, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.6396180422903234e-05, "train_loss": 0.4851775680269514, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.076147209486148e-05, "train_loss": 0.4269947239330837, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.5688793960098128e-06, "train_loss": 0.49186116456985474, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/005/retfound/metrics.json b/results/downsample/airogs/005/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..a236743ab64b67d1e94f6143718f2524d237f3c4 --- /dev/null +++ b/results/downsample/airogs/005/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.834, + "balanced_accuracy": 0.834, + "precision_macro": 0.8343423665833813, + "recall_macro": 0.834, + "f1_macro": 0.8339574931182383, + "precision_weighted": 0.8343423665833813, + "recall_weighted": 0.834, + "f1_weighted": 0.8339574931182383, + "cohen_kappa": 0.6679999999999999, + "quadratic_weighted_kappa": 0.6679999999999999, + "mcc": 0.6683422788926326, + "auroc": 0.9092500000000001, + "auprc": 0.9211760623684944, + "sensitivity": 0.85, + "specificity": 0.818, + "precision_pos": 0.8236434108527132, + "f1_pos": 0.8366141732283464, + "per_class": { + "0": { + "precision": 0.8450413223140496, + "recall": 0.818, + "f1-score": 0.8313008130081301, + "support": 500.0 + }, + "1": { + "precision": 0.8236434108527132, + "recall": 0.85, + "f1-score": 0.8366141732283464, + "support": 500.0 + }, + "accuracy": 0.834, + "macro avg": { + "precision": 0.8343423665833813, + "recall": 0.834, + "f1-score": 0.8339574931182383, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8343423665833813, + "recall": 0.834, + "f1-score": 0.8339574931182383, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/005/retfound/metrics_test.csv b/results/downsample/airogs/005/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..b316e418dd518b2e7f2ed841e54b497e8dbe10f2 --- /dev/null +++ b/results/downsample/airogs/005/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.4487141091376543,0.834,0.8339574931182383,0.9092690000000001,0.166,0.7152122415949386,0.8343423665833813,0.834,0.9036981691740603,0.6679999999999999 diff --git a/results/downsample/airogs/005/retfound/metrics_val.csv b/results/downsample/airogs/005/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..76088ddec249805a0d41a4f53c488f41dc49e47d --- /dev/null +++ b/results/downsample/airogs/005/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6924146133310655,0.5,0.3333333333333333,0.6822016460905349,0.5,0.25,0.25,0.5,0.6521002697543876,0.0 +0.6892514263882357,0.5074074074074074,0.34958613319809456,0.7130452674897119,0.4925925925925926,0.25927307904919844,0.7518656716417911,0.5074074074074074,0.711038983680687,0.014814814814814836 +0.6840392035596511,0.6462962962962963,0.6294925909295015,0.7180555555555556,0.3537037037037037,0.46416674342206254,0.678718056137411,0.6462962962962964,0.7195992183416331,0.2925925925925926 +0.6749616335420048,0.6555555555555556,0.63572267920094,0.7376097393689987,0.34444444444444444,0.47169014084507044,0.6988636363636364,0.6555555555555556,0.7435723625423136,0.3111111111111111 +0.6554641092524809,0.6888888888888889,0.6884615384615385,0.7709327846364884,0.3111111111111111,0.5250447227191413,0.6899310344827586,0.6888888888888889,0.7722443763470702,0.37777777777777777 +0.6128462447839624,0.7092592592592593,0.7047135080160635,0.803672839506173,0.29074074074074074,0.5452962685946979,0.7229904547514289,0.7092592592592593,0.8015167162225773,0.4185185185185185 +0.5681105869657853,0.7574074074074074,0.7536144557331486,0.8474176954732511,0.24259259259259258,0.6056058077137063,0.7742979930128195,0.7574074074074074,0.841787960953583,0.5148148148148148 +0.5174003892085132,0.8092592592592592,0.8092429049129727,0.8709979423868313,0.19074074074074074,0.679607376446457,0.8093653516295025,0.8092592592592593,0.8639754815648006,0.6185185185185185 +0.4804135333089268,0.7907407407407407,0.7892773892773892,0.8889437585733881,0.20925925925925926,0.6522536193635953,0.799047619047619,0.7907407407407407,0.8864789763261796,0.5814814814814815 +0.445953649633071,0.7925925925925926,0.7901719424360594,0.8992249657064473,0.2074074074074074,0.653701332574572,0.8067475839852738,0.7925925925925926,0.8999173292559907,0.5851851851851853 +0.44857079667203564,0.8222222222222222,0.8220660113130869,0.8935493827160494,0.17777777777777778,0.6979220265188997,0.823357744617587,0.8222222222222222,0.8940204279166715,0.6444444444444444 +0.4745547149111243,0.7759259259259259,0.7681326051547379,0.9019101508916323,0.22407407407407406,0.6254869825810077,0.8187847667950363,0.7759259259259259,0.9011274411058106,0.5518518518518518 +0.4253272764823016,0.812962962962963,0.8121739130434783,0.8900925925925927,0.18703703703703703,0.6839250493096647,0.8183118242064876,0.8129629629629629,0.8904299761780903,0.625925925925926 +0.4036833416013157,0.8074074074074075,0.8073413379073757,0.9009122085048011,0.1925925925925926,0.6769407383825814,0.8078296703296703,0.8074074074074074,0.9009129464829887,0.6148148148148148 +0.43943539568606543,0.7962962962962963,0.7942386831275721,0.9039677640603567,0.2037037037037037,0.6591862416107382,0.808641975308642,0.7962962962962963,0.9005216917455567,0.5925925925925926 +0.4184439348823884,0.8222222222222222,0.8219780219780219,0.8972119341563786,0.17777777777777778,0.69781438074121,0.8240000000000001,0.8222222222222222,0.8926732207591261,0.6444444444444444 +0.42643390683566823,0.8166666666666667,0.8166509474406243,0.8969410150891632,0.18333333333333332,0.6901218723671063,0.8167753001715266,0.8166666666666667,0.8920939701660362,0.6333333333333333 +0.4286675724913092,0.8222222222222222,0.8215957271075381,0.9027091906721536,0.17777777777777778,0.6973470277987073,0.8268128443430353,0.8222222222222222,0.8971796581965663,0.6444444444444444 +0.4013392381808337,0.837037037037037,0.8369811320754716,0.9068655692729766,0.16296296296296298,0.719674143510769,0.8374999999999999,0.837037037037037,0.9004451326671329,0.674074074074074 +0.412210370249608,0.8388888888888889,0.8385960813684634,0.9089849108367628,0.16111111111111112,0.7221141110545084,0.8413660167746749,0.8388888888888889,0.9017271445079654,0.6777777777777778 +0.4155636272009681,0.8425925925925926,0.8425488561637492,0.9090363511659809,0.1574074074074074,0.7279435863820722,0.842973674453096,0.8425925925925926,0.9019297934023629,0.6851851851851851 +0.43423646162537965,0.8333333333333334,0.8332212323784163,0.9092729766803842,0.16666666666666666,0.7141445511010729,0.8342319542253521,0.8333333333333334,0.9031963582329268,0.6666666666666667 +0.43337904968682456,0.8444444444444444,0.8444252376836646,0.9127572016460905,0.15555555555555556,0.7307443365695793,0.844614624505929,0.8444444444444444,0.9071870312543685,0.6888888888888889 +0.4288011767408427,0.8444444444444444,0.8444423105941097,0.9143518518518519,0.15555555555555556,0.730766464974369,0.8444633450395083,0.8444444444444444,0.9089514492614825,0.6888888888888889 +0.4246067654560594,0.8462962962962963,0.8461776061776062,0.9176165980795611,0.1537037037037037,0.7333927982181159,0.8473684210526315,0.8462962962962963,0.9124689503122527,0.6925925925925926 +0.4427410313750015,0.8407407407407408,0.8399481658142517,0.9189746227709192,0.15925925925925927,0.7242234243094828,0.8476265114196149,0.8407407407407408,0.9144371264109465,0.6814814814814815 +0.4483622448409305,0.8388888888888889,0.8379547390822955,0.9196742112482854,0.16111111111111112,0.7212963152193905,0.8468877687134051,0.8388888888888889,0.9150033717690879,0.6777777777777778 +0.4477856479146901,0.8407407407407408,0.8398620689655172,0.9201714677640604,0.15925925925925927,0.7241131916899834,0.8483870967741935,0.8407407407407408,0.9157843848635334,0.6814814814814815 +0.4435284231953761,0.8388888888888889,0.8381289386729881,0.9202160493827161,0.16111111111111112,0.7215182764958137,0.8453747326334038,0.8388888888888889,0.9155162553806268,0.6777777777777778 +0.4426653034546796,0.8388888888888889,0.8381289386729881,0.9201714677640604,0.16111111111111112,0.7215182764958137,0.8453747326334038,0.8388888888888889,0.9154273434043976,0.6777777777777778 diff --git a/results/downsample/airogs/005/retfound/pr.png b/results/downsample/airogs/005/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..274c05d4c4ca52474b157fc2221fcf9e1a33034f --- /dev/null +++ b/results/downsample/airogs/005/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:48dd9e548e99ce2c040d821bcac4426999635260d26341d0d5f80127ac807a57 +size 48594 diff --git a/results/downsample/airogs/005/retfound/roc.png b/results/downsample/airogs/005/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..fbd11a0369d9b81828aed0edc5dfbea539b09171 --- /dev/null +++ b/results/downsample/airogs/005/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d5d7a64e13512352f5f06495aecf5dc74fe09015172ff599a31ca9802f63701 +size 64203 diff --git a/results/downsample/airogs/005/retfound/test_pred.npz b/results/downsample/airogs/005/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..53f07367407e92666de893690d62170fed42b8ba --- /dev/null +++ b/results/downsample/airogs/005/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d6bbc0b274d3de3fea7ece8997a523de09b2c88c4033593dc53d72b490273f8 +size 12510 diff --git a/results/downsample/airogs/005/retfound/train.log b/results/downsample/airogs/005/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..bfc44e6b60b3ae911fa436f3769938481aa3bbc5 --- /dev/null +++ b/results/downsample/airogs/005/retfound/train.log @@ -0,0 +1,506 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:52:11.271160388 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:52:11.716282] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:52:11.716472] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/airogs_5', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/005', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:52:14.587792] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:52:16.109406] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:52:16.686689] Sampler_train = +[14:52:16.742233] len of train_set: 224 +[14:52:16.906915] [Adaptation] Full fine-tuning: training all parameters. +[14:52:16.907977] number of trainable params (M): 303.30 +[14:52:16.908073] base lr: 5.00e-03 +[14:52:16.908154] actual lr: 6.25e-04 +[14:52:16.908221] accumulate grad iterations: 1 +[14:52:16.908275] effective batch size: 32 +[14:52:16.910996] criterion = CrossEntropyLoss() +[14:52:16.911078] Start training for 30 epochs +[14:52:16.912866] log_dir: ./output_logs/retfound +[14:52:18.136217] Epoch: [0] [0/7] eta: 0:00:08 lr: 0.000000 loss: 0.6929 (0.6929) time: 1.2225 data: 0.6033 max mem: 7340 +[14:52:18.575303] Epoch: [0] [6/7] eta: 0:00:00 lr: 0.000054 loss: 0.6929 (0.6924) time: 0.2373 data: 0.0862 max mem: 9669 +[14:52:18.642952] Epoch: [0] Total time: 0:00:01 (0.2471 s / it) +[14:52:18.644093] Averaged stats: lr: 0.000054 loss: 0.6929 (0.6924) +[14:52:19.164596] val: [ 0/17] eta: 0:00:08 loss: 0.7147 (0.7147) time: 0.5078 data: 0.4817 max mem: 9669 +[14:52:19.569662] val: [10/17] eta: 0:00:00 loss: 0.7146 (0.7044) time: 0.0829 data: 0.0649 max mem: 9669 +[14:52:19.746321] val: [16/17] eta: 0:00:00 loss: 0.6897 (0.6924) time: 0.0640 data: 0.0420 max mem: 9669 +[14:52:19.826479] val: Total time: 0:00:01 (0.0688 s / it) +[14:52:19.842264] val loss: 0.6924146133310655 +[14:52:19.842524] Accuracy: 0.5000, F1 Score: 0.3333, ROC AUC: 0.6822, Hamming Loss: 0.5000, + Jaccard Score: 0.2500, Precision: 0.2500, Recall: 0.5000, + Average Precision: 0.6521, Kappa: 0.0000, Score: 0.3385 +[14:52:21.676448] Best epoch = 0, Best score = 0.3385 +[14:52:21.753817] log_dir: ./output_logs/retfound +[14:52:22.575926] Epoch: [1] [0/7] eta: 0:00:05 lr: 0.000063 loss: 0.6910 (0.6910) time: 0.8212 data: 0.6664 max mem: 9669 +[14:52:22.964853] Epoch: [1] [6/7] eta: 0:00:00 lr: 0.000116 loss: 0.6965 (0.6943) time: 0.1728 data: 0.0952 max mem: 9669 +[14:52:23.032728] Epoch: [1] Total time: 0:00:01 (0.1827 s / it) +[14:52:23.033507] Averaged stats: lr: 0.000116 loss: 0.6965 (0.6943) +[14:52:23.715458] val: [ 0/17] eta: 0:00:11 loss: 0.7103 (0.7103) time: 0.6579 data: 0.6410 max mem: 9669 +[14:52:23.963851] val: [10/17] eta: 0:00:00 loss: 0.7096 (0.7009) time: 0.0823 data: 0.0670 max mem: 9669 +[14:52:24.054852] val: [16/17] eta: 0:00:00 loss: 0.6876 (0.6893) time: 0.0586 data: 0.0434 max mem: 9669 +[14:52:24.139172] val: Total time: 0:00:01 (0.0637 s / it) +[14:52:24.153300] val loss: 0.6892514263882357 +[14:52:24.153485] Accuracy: 0.5074, F1 Score: 0.3496, ROC AUC: 0.7130, Hamming Loss: 0.4926, + Jaccard Score: 0.2593, Precision: 0.7519, Recall: 0.5074, + Average Precision: 0.7110, Kappa: 0.0148, Score: 0.3591 +[14:52:25.901744] Best epoch = 1, Best score = 0.3591 +[14:52:25.959445] log_dir: ./output_logs/retfound +[14:52:26.668186] Epoch: [2] [0/7] eta: 0:00:04 lr: 0.000125 loss: 0.6945 (0.6945) time: 0.7077 data: 0.6361 max mem: 9669 +[14:52:27.057247] Epoch: [2] [6/7] eta: 0:00:00 lr: 0.000179 loss: 0.6889 (0.6891) time: 0.1566 data: 0.0909 max mem: 9669 +[14:52:27.130913] Epoch: [2] Total time: 0:00:01 (0.1673 s / it) +[14:52:27.132053] Averaged stats: lr: 0.000179 loss: 0.6889 (0.6891) +[14:52:27.707304] val: [ 0/17] eta: 0:00:09 loss: 0.6644 (0.6644) time: 0.5622 data: 0.5452 max mem: 9669 +[14:52:27.874916] val: [10/17] eta: 0:00:00 loss: 0.6669 (0.6741) time: 0.0663 data: 0.0507 max mem: 9669 +[14:52:27.967664] val: [16/17] eta: 0:00:00 loss: 0.6895 (0.6840) time: 0.0483 data: 0.0328 max mem: 9669 +[14:52:28.052982] val: Total time: 0:00:00 (0.0534 s / it) +[14:52:28.066932] val loss: 0.6840392035596511 +[14:52:28.067164] Accuracy: 0.6463, F1 Score: 0.6295, ROC AUC: 0.7181, Hamming Loss: 0.3537, + Jaccard Score: 0.4642, Precision: 0.6787, Recall: 0.6463, + Average Precision: 0.7196, Kappa: 0.2926, Score: 0.5467 +[14:52:29.884418] Best epoch = 2, Best score = 0.5467 +[14:52:29.938139] log_dir: ./output_logs/retfound +[14:52:30.665050] Epoch: [3] [0/7] eta: 0:00:05 lr: 0.000188 loss: 0.6905 (0.6905) time: 0.7261 data: 0.6552 max mem: 9669 +[14:52:31.053570] Epoch: [3] [6/7] eta: 0:00:00 lr: 0.000241 loss: 0.6899 (0.6862) time: 0.1591 data: 0.0937 max mem: 9669 +[14:52:31.123065] Epoch: [3] Total time: 0:00:01 (0.1693 s / it) +[14:52:31.123843] Averaged stats: lr: 0.000241 loss: 0.6899 (0.6862) +[14:52:31.735959] val: [ 0/17] eta: 0:00:09 loss: 0.6356 (0.6356) time: 0.5874 data: 0.5674 max mem: 9669 +[14:52:31.933024] val: [10/17] eta: 0:00:00 loss: 0.6411 (0.6559) time: 0.0713 data: 0.0554 max mem: 9669 +[14:52:32.023818] val: [16/17] eta: 0:00:00 loss: 0.6869 (0.6750) time: 0.0514 data: 0.0359 max mem: 9669 +[14:52:32.104483] val: Total time: 0:00:00 (0.0563 s / it) +[14:52:32.119829] val loss: 0.6749616335420048 +[14:52:32.119998] Accuracy: 0.6556, F1 Score: 0.6357, ROC AUC: 0.7376, Hamming Loss: 0.3444, + Jaccard Score: 0.4717, Precision: 0.6989, Recall: 0.6556, + Average Precision: 0.7436, Kappa: 0.3111, Score: 0.5615 +[14:52:33.926469] Best epoch = 3, Best score = 0.5615 +[14:52:34.002668] log_dir: ./output_logs/retfound +[14:52:34.676903] Epoch: [4] [0/7] eta: 0:00:04 lr: 0.000250 loss: 0.6951 (0.6951) time: 0.6734 data: 0.6015 max mem: 9669 +[14:52:35.070150] Epoch: [4] [6/7] eta: 0:00:00 lr: 0.000304 loss: 0.6796 (0.6811) time: 0.1523 data: 0.0860 max mem: 9669 +[14:52:35.145867] Epoch: [4] Total time: 0:00:01 (0.1633 s / it) +[14:52:35.146659] Averaged stats: lr: 0.000304 loss: 0.6796 (0.6811) +[14:52:35.823818] val: [ 0/17] eta: 0:00:11 loss: 0.6353 (0.6353) time: 0.6529 data: 0.6359 max mem: 9669 +[14:52:36.084054] val: [10/17] eta: 0:00:00 loss: 0.6491 (0.6511) time: 0.0830 data: 0.0675 max mem: 9669 +[14:52:36.175602] val: [16/17] eta: 0:00:00 loss: 0.6556 (0.6555) time: 0.0590 data: 0.0437 max mem: 9669 +[14:52:36.250398] val: Total time: 0:00:01 (0.0635 s / it) +[14:52:36.264203] val loss: 0.6554641092524809 +[14:52:36.264391] Accuracy: 0.6889, F1 Score: 0.6885, ROC AUC: 0.7709, Hamming Loss: 0.3111, + Jaccard Score: 0.5250, Precision: 0.6899, Recall: 0.6889, + Average Precision: 0.7722, Kappa: 0.3778, Score: 0.6124 +[14:52:38.084981] Best epoch = 4, Best score = 0.6124 +[14:52:38.162627] log_dir: ./output_logs/retfound +[14:52:38.858975] Epoch: [5] [0/7] eta: 0:00:04 lr: 0.000313 loss: 0.6526 (0.6526) time: 0.6954 data: 0.6225 max mem: 9669 +[14:52:39.248262] Epoch: [5] [6/7] eta: 0:00:00 lr: 0.000366 loss: 0.6526 (0.6650) time: 0.1548 data: 0.0890 max mem: 9669 +[14:52:39.319804] Epoch: [5] Total time: 0:00:01 (0.1653 s / it) +[14:52:39.320652] Averaged stats: lr: 0.000366 loss: 0.6526 (0.6650) +[14:52:39.836285] val: [ 0/17] eta: 0:00:08 loss: 0.6521 (0.6521) time: 0.5032 data: 0.4862 max mem: 9669 +[14:52:40.103170] val: [10/17] eta: 0:00:00 loss: 0.6534 (0.6492) time: 0.0699 data: 0.0545 max mem: 9669 +[14:52:40.193705] val: [16/17] eta: 0:00:00 loss: 0.6252 (0.6128) time: 0.0505 data: 0.0353 max mem: 9669 +[14:52:40.280391] val: Total time: 0:00:00 (0.0558 s / it) +[14:52:40.295173] val loss: 0.6128462447839624 +[14:52:40.295359] Accuracy: 0.7093, F1 Score: 0.7047, ROC AUC: 0.8037, Hamming Loss: 0.2907, + Jaccard Score: 0.5453, Precision: 0.7230, Recall: 0.7093, + Average Precision: 0.8015, Kappa: 0.4185, Score: 0.6423 +[14:52:42.135897] Best epoch = 5, Best score = 0.6423 +[14:52:42.202142] log_dir: ./output_logs/retfound +[14:52:42.909944] Epoch: [6] [0/7] eta: 0:00:04 lr: 0.000375 loss: 0.6721 (0.6721) time: 0.7069 data: 0.6332 max mem: 9669 +[14:52:43.297310] Epoch: [6] [6/7] eta: 0:00:00 lr: 0.000429 loss: 0.6480 (0.6465) time: 0.1562 data: 0.0905 max mem: 9669 +[14:52:43.369040] Epoch: [6] Total time: 0:00:01 (0.1667 s / it) +[14:52:43.369873] Averaged stats: lr: 0.000429 loss: 0.6480 (0.6465) +[14:52:43.972551] val: [ 0/17] eta: 0:00:10 loss: 0.6177 (0.6177) time: 0.5902 data: 0.5732 max mem: 9669 +[14:52:44.165649] val: [10/17] eta: 0:00:00 loss: 0.6314 (0.6215) time: 0.0712 data: 0.0558 max mem: 9669 +[14:52:44.256533] val: [16/17] eta: 0:00:00 loss: 0.5793 (0.5681) time: 0.0514 data: 0.0361 max mem: 9669 +[14:52:44.331746] val: Total time: 0:00:00 (0.0559 s / it) +[14:52:44.346390] val loss: 0.5681105869657853 +[14:52:44.346562] Accuracy: 0.7574, F1 Score: 0.7536, ROC AUC: 0.8474, Hamming Loss: 0.2426, + Jaccard Score: 0.6056, Precision: 0.7743, Recall: 0.7574, + Average Precision: 0.8418, Kappa: 0.5148, Score: 0.7053 +[14:52:46.138727] Best epoch = 6, Best score = 0.7053 +[14:52:46.205114] log_dir: ./output_logs/retfound +[14:52:46.938787] Epoch: [7] [0/7] eta: 0:00:05 lr: 0.000438 loss: 0.5656 (0.5656) time: 0.7327 data: 0.6604 max mem: 9669 +[14:52:47.326903] Epoch: [7] [6/7] eta: 0:00:00 lr: 0.000491 loss: 0.6480 (0.6434) time: 0.1600 data: 0.0944 max mem: 9669 +[14:52:47.398206] Epoch: [7] Total time: 0:00:01 (0.1704 s / it) +[14:52:47.399029] Averaged stats: lr: 0.000491 loss: 0.6480 (0.6434) +[14:52:48.061013] val: [ 0/17] eta: 0:00:11 loss: 0.4783 (0.4783) time: 0.6498 data: 0.6329 max mem: 9669 +[14:52:48.215642] val: [10/17] eta: 0:00:00 loss: 0.5208 (0.5321) time: 0.0731 data: 0.0576 max mem: 9669 +[14:52:48.306371] val: [16/17] eta: 0:00:00 loss: 0.5136 (0.5174) time: 0.0526 data: 0.0373 max mem: 9669 +[14:52:48.385265] val: Total time: 0:00:00 (0.0573 s / it) +[14:52:48.400130] val loss: 0.5174003892085132 +[14:52:48.400456] Accuracy: 0.8093, F1 Score: 0.8092, ROC AUC: 0.8710, Hamming Loss: 0.1907, + Jaccard Score: 0.6796, Precision: 0.8094, Recall: 0.8093, + Average Precision: 0.8640, Kappa: 0.6185, Score: 0.7663 +[14:52:50.262720] Best epoch = 7, Best score = 0.7663 +[14:52:50.330179] log_dir: ./output_logs/retfound +[14:52:50.989702] Epoch: [8] [0/7] eta: 0:00:04 lr: 0.000500 loss: 0.5004 (0.5004) time: 0.6586 data: 0.5857 max mem: 9669 +[14:52:51.378815] Epoch: [8] [6/7] eta: 0:00:00 lr: 0.000554 loss: 0.6367 (0.6223) time: 0.1496 data: 0.0837 max mem: 9669 +[14:52:51.448358] Epoch: [8] Total time: 0:00:01 (0.1597 s / it) +[14:52:51.449153] Averaged stats: lr: 0.000554 loss: 0.6367 (0.6223) +[14:52:51.991254] val: [ 0/17] eta: 0:00:09 loss: 0.5055 (0.5055) time: 0.5298 data: 0.5102 max mem: 9669 +[14:52:52.151571] val: [10/17] eta: 0:00:00 loss: 0.5156 (0.5353) time: 0.0627 data: 0.0468 max mem: 9669 +[14:52:52.242851] val: [16/17] eta: 0:00:00 loss: 0.5027 (0.4804) time: 0.0459 data: 0.0303 max mem: 9669 +[14:52:52.317872] val: Total time: 0:00:00 (0.0504 s / it) +[14:52:52.332448] val loss: 0.4804135333089268 +[14:52:52.332625] Accuracy: 0.7907, F1 Score: 0.7893, ROC AUC: 0.8889, Hamming Loss: 0.2093, + Jaccard Score: 0.6523, Precision: 0.7990, Recall: 0.7907, + Average Precision: 0.8865, Kappa: 0.5815, Score: 0.7532 +[14:52:52.372183] Best epoch = 7, Best score = 0.7663 +[14:52:52.636472] log_dir: ./output_logs/retfound +[14:52:53.322050] Epoch: [9] [0/7] eta: 0:00:04 lr: 0.000562 loss: 0.6758 (0.6758) time: 0.6848 data: 0.6125 max mem: 9669 +[14:52:53.701237] Epoch: [9] [6/7] eta: 0:00:00 lr: 0.000616 loss: 0.6282 (0.6338) time: 0.1519 data: 0.0875 max mem: 9669 +[14:52:53.776888] Epoch: [9] Total time: 0:00:01 (0.1629 s / it) +[14:52:53.777635] Averaged stats: lr: 0.000616 loss: 0.6282 (0.6338) +[14:52:54.419462] val: [ 0/17] eta: 0:00:10 loss: 0.5093 (0.5093) time: 0.6296 data: 0.6126 max mem: 9669 +[14:52:54.664981] val: [10/17] eta: 0:00:00 loss: 0.5093 (0.5210) time: 0.0795 data: 0.0641 max mem: 9669 +[14:52:54.756099] val: [16/17] eta: 0:00:00 loss: 0.4874 (0.4460) time: 0.0568 data: 0.0415 max mem: 9669 +[14:52:54.835395] val: Total time: 0:00:01 (0.0615 s / it) +[14:52:54.849161] val loss: 0.445953649633071 +[14:52:54.849348] Accuracy: 0.7926, F1 Score: 0.7902, ROC AUC: 0.8992, Hamming Loss: 0.2074, + Jaccard Score: 0.6537, Precision: 0.8067, Recall: 0.7926, + Average Precision: 0.8999, Kappa: 0.5852, Score: 0.7582 +[14:52:54.904633] Best epoch = 7, Best score = 0.7663 +[14:52:55.125664] log_dir: ./output_logs/retfound +[14:52:55.819278] Epoch: [10] [0/7] eta: 0:00:04 lr: 0.000625 loss: 0.6074 (0.6074) time: 0.6927 data: 0.6213 max mem: 9669 +[14:52:56.208955] Epoch: [10] [6/7] eta: 0:00:00 lr: 0.000622 loss: 0.6291 (0.6523) time: 0.1545 data: 0.0888 max mem: 9669 +[14:52:56.284446] Epoch: [10] Total time: 0:00:01 (0.1655 s / it) +[14:52:56.285188] Averaged stats: lr: 0.000622 loss: 0.6291 (0.6523) +[14:52:56.946143] val: [ 0/17] eta: 0:00:11 loss: 0.3837 (0.3837) time: 0.6486 data: 0.6292 max mem: 9669 +[14:52:57.107051] val: [10/17] eta: 0:00:00 loss: 0.4128 (0.4441) time: 0.0735 data: 0.0573 max mem: 9669 +[14:52:57.244618] val: [16/17] eta: 0:00:00 loss: 0.4213 (0.4486) time: 0.0556 data: 0.0399 max mem: 9669 +[14:52:57.320788] val: Total time: 0:00:01 (0.0602 s / it) +[14:52:57.335352] val loss: 0.44857079667203564 +[14:52:57.335550] Accuracy: 0.8222, F1 Score: 0.8221, ROC AUC: 0.8935, Hamming Loss: 0.1778, + Jaccard Score: 0.6979, Precision: 0.8234, Recall: 0.8222, + Average Precision: 0.8940, Kappa: 0.6444, Score: 0.7867 +[14:52:59.159239] Best epoch = 10, Best score = 0.7867 +[14:52:59.223300] log_dir: ./output_logs/retfound +[14:52:59.974378] Epoch: [11] [0/7] eta: 0:00:05 lr: 0.000621 loss: 0.5871 (0.5871) time: 0.7502 data: 0.6767 max mem: 9669 +[14:53:00.367657] Epoch: [11] [6/7] eta: 0:00:00 lr: 0.000612 loss: 0.5802 (0.5730) time: 0.1632 data: 0.0967 max mem: 9669 +[14:53:00.447969] Epoch: [11] Total time: 0:00:01 (0.1749 s / it) +[14:53:00.448969] Averaged stats: lr: 0.000612 loss: 0.5802 (0.5730) +[14:53:01.135899] val: [ 0/17] eta: 0:00:11 loss: 0.7364 (0.7364) time: 0.6740 data: 0.6566 max mem: 9669 +[14:53:01.291184] val: [10/17] eta: 0:00:00 loss: 0.6833 (0.6373) time: 0.0753 data: 0.0598 max mem: 9669 +[14:53:01.382391] val: [16/17] eta: 0:00:00 loss: 0.5455 (0.4746) time: 0.0541 data: 0.0387 max mem: 9669 +[14:53:01.463841] val: Total time: 0:00:01 (0.0590 s / it) +[14:53:01.485525] val loss: 0.4745547149111243 +[14:53:01.485735] Accuracy: 0.7759, F1 Score: 0.7681, ROC AUC: 0.9019, Hamming Loss: 0.2241, + Jaccard Score: 0.6255, Precision: 0.8188, Recall: 0.7759, + Average Precision: 0.9011, Kappa: 0.5519, Score: 0.7406 +[14:53:01.525431] Best epoch = 10, Best score = 0.7867 +[14:53:01.783827] log_dir: ./output_logs/retfound +[14:53:02.547727] Epoch: [12] [0/7] eta: 0:00:05 lr: 0.000610 loss: 0.5265 (0.5265) time: 0.7631 data: 0.6946 max mem: 9669 +[14:53:02.935504] Epoch: [12] [6/7] eta: 0:00:00 lr: 0.000594 loss: 0.5265 (0.5562) time: 0.1643 data: 0.0993 max mem: 9669 +[14:53:03.006405] Epoch: [12] Total time: 0:00:01 (0.1746 s / it) +[14:53:03.007169] Averaged stats: lr: 0.000594 loss: 0.5265 (0.5562) +[14:53:03.652725] val: [ 0/17] eta: 0:00:10 loss: 0.2867 (0.2867) time: 0.6222 data: 0.6050 max mem: 9669 +[14:53:03.807732] val: [10/17] eta: 0:00:00 loss: 0.3244 (0.3965) time: 0.0706 data: 0.0551 max mem: 9669 +[14:53:03.983532] val: [16/17] eta: 0:00:00 loss: 0.3428 (0.4253) time: 0.0560 data: 0.0404 max mem: 9669 +[14:53:04.057926] val: Total time: 0:00:01 (0.0605 s / it) +[14:53:04.084530] val loss: 0.4253272764823016 +[14:53:04.084780] Accuracy: 0.8130, F1 Score: 0.8122, ROC AUC: 0.8901, Hamming Loss: 0.1870, + Jaccard Score: 0.6839, Precision: 0.8183, Recall: 0.8130, + Average Precision: 0.8904, Kappa: 0.6259, Score: 0.7761 +[14:53:04.133089] Best epoch = 10, Best score = 0.7867 +[14:53:04.371915] log_dir: ./output_logs/retfound +[14:53:05.026008] Epoch: [13] [0/7] eta: 0:00:04 lr: 0.000591 loss: 0.6344 (0.6344) time: 0.6532 data: 0.5816 max mem: 9669 +[14:53:05.416268] Epoch: [13] [6/7] eta: 0:00:00 lr: 0.000569 loss: 0.6099 (0.6059) time: 0.1490 data: 0.0832 max mem: 9669 +[14:53:05.487799] Epoch: [13] Total time: 0:00:01 (0.1594 s / it) +[14:53:05.488518] Averaged stats: lr: 0.000569 loss: 0.6099 (0.6059) +[14:53:05.999355] val: [ 0/17] eta: 0:00:08 loss: 0.4083 (0.4083) time: 0.4863 data: 0.4695 max mem: 9669 +[14:53:06.227153] val: [10/17] eta: 0:00:00 loss: 0.4479 (0.4468) time: 0.0649 data: 0.0495 max mem: 9669 +[14:53:06.317960] val: [16/17] eta: 0:00:00 loss: 0.3933 (0.4037) time: 0.0473 data: 0.0320 max mem: 9669 +[14:53:06.394675] val: Total time: 0:00:00 (0.0519 s / it) +[14:53:06.408413] val loss: 0.4036833416013157 +[14:53:06.408595] Accuracy: 0.8074, F1 Score: 0.8073, ROC AUC: 0.9009, Hamming Loss: 0.1926, + Jaccard Score: 0.6769, Precision: 0.8078, Recall: 0.8074, + Average Precision: 0.9009, Kappa: 0.6148, Score: 0.7744 +[14:53:06.434023] Best epoch = 10, Best score = 0.7867 +[14:53:06.710070] log_dir: ./output_logs/retfound +[14:53:07.569805] Epoch: [14] [0/7] eta: 0:00:06 lr: 0.000565 loss: 0.6178 (0.6178) time: 0.8588 data: 0.7884 max mem: 9669 +[14:53:07.962090] Epoch: [14] [6/7] eta: 0:00:00 lr: 0.000539 loss: 0.5448 (0.5409) time: 0.1786 data: 0.1127 max mem: 9669 +[14:53:08.051440] Epoch: [14] Total time: 0:00:01 (0.1916 s / it) +[14:53:08.052241] Averaged stats: lr: 0.000539 loss: 0.5448 (0.5409) +[14:53:08.695060] val: [ 0/17] eta: 0:00:10 loss: 0.6207 (0.6207) time: 0.6351 data: 0.6180 max mem: 9669 +[14:53:08.957690] val: [10/17] eta: 0:00:00 loss: 0.6178 (0.5752) time: 0.0816 data: 0.0662 max mem: 9669 +[14:53:09.047095] val: [16/17] eta: 0:00:00 loss: 0.5365 (0.4394) time: 0.0580 data: 0.0429 max mem: 9669 +[14:53:09.127344] val: Total time: 0:00:01 (0.0628 s / it) +[14:53:09.148157] val loss: 0.43943539568606543 +[14:53:09.148350] Accuracy: 0.7963, F1 Score: 0.7942, ROC AUC: 0.9040, Hamming Loss: 0.2037, + Jaccard Score: 0.6592, Precision: 0.8086, Recall: 0.7963, + Average Precision: 0.9005, Kappa: 0.5926, Score: 0.7636 +[14:53:09.196211] Best epoch = 10, Best score = 0.7867 +[14:53:09.429222] log_dir: ./output_logs/retfound +[14:53:10.179239] Epoch: [15] [0/7] eta: 0:00:05 lr: 0.000534 loss: 0.5302 (0.5302) time: 0.7484 data: 0.6742 max mem: 9669 +[14:53:10.569452] Epoch: [15] [6/7] eta: 0:00:00 lr: 0.000502 loss: 0.5829 (0.5607) time: 0.1625 data: 0.0964 max mem: 9669 +[14:53:10.646540] Epoch: [15] Total time: 0:00:01 (0.1739 s / it) +[14:53:10.647313] Averaged stats: lr: 0.000502 loss: 0.5829 (0.5607) +[14:53:11.156013] val: [ 0/17] eta: 0:00:08 loss: 0.3746 (0.3746) time: 0.4914 data: 0.4742 max mem: 9669 +[14:53:11.443600] val: [10/17] eta: 0:00:00 loss: 0.4573 (0.4575) time: 0.0708 data: 0.0552 max mem: 9669 +[14:53:11.534652] val: [16/17] eta: 0:00:00 loss: 0.3746 (0.4184) time: 0.0511 data: 0.0358 max mem: 9669 +[14:53:11.613721] val: Total time: 0:00:00 (0.0559 s / it) +[14:53:11.627592] val loss: 0.4184439348823884 +[14:53:11.627786] Accuracy: 0.8222, F1 Score: 0.8220, ROC AUC: 0.8972, Hamming Loss: 0.1778, + Jaccard Score: 0.6978, Precision: 0.8240, Recall: 0.8222, + Average Precision: 0.8927, Kappa: 0.6444, Score: 0.7879 +[14:53:13.496707] Best epoch = 15, Best score = 0.7879 +[14:53:13.563933] log_dir: ./output_logs/retfound +[14:53:14.360814] Epoch: [16] [0/7] eta: 0:00:05 lr: 0.000496 loss: 0.5976 (0.5976) time: 0.7959 data: 0.7229 max mem: 9669 +[14:53:14.749854] Epoch: [16] [6/7] eta: 0:00:00 lr: 0.000461 loss: 0.5206 (0.5484) time: 0.1692 data: 0.1033 max mem: 9669 +[14:53:14.818617] Epoch: [16] Total time: 0:00:01 (0.1792 s / it) +[14:53:14.819456] Averaged stats: lr: 0.000461 loss: 0.5206 (0.5484) +[14:53:15.417161] val: [ 0/17] eta: 0:00:09 loss: 0.4235 (0.4235) time: 0.5709 data: 0.5544 max mem: 9669 +[14:53:15.571888] val: [10/17] eta: 0:00:00 loss: 0.4692 (0.4942) time: 0.0659 data: 0.0505 max mem: 9669 +[14:53:15.663020] val: [16/17] eta: 0:00:00 loss: 0.3997 (0.4264) time: 0.0480 data: 0.0327 max mem: 9669 +[14:53:15.745548] val: Total time: 0:00:00 (0.0529 s / it) +[14:53:15.760168] val loss: 0.42643390683566823 +[14:53:15.760361] Accuracy: 0.8167, F1 Score: 0.8167, ROC AUC: 0.8969, Hamming Loss: 0.1833, + Jaccard Score: 0.6901, Precision: 0.8168, Recall: 0.8167, + Average Precision: 0.8921, Kappa: 0.6333, Score: 0.7823 +[14:53:15.788202] Best epoch = 15, Best score = 0.7879 +[14:53:16.069404] log_dir: ./output_logs/retfound +[14:53:16.736379] Epoch: [17] [0/7] eta: 0:00:04 lr: 0.000455 loss: 0.4945 (0.4945) time: 0.6660 data: 0.5966 max mem: 9669 +[14:53:17.125412] Epoch: [17] [6/7] eta: 0:00:00 lr: 0.000416 loss: 0.4945 (0.4977) time: 0.1506 data: 0.0853 max mem: 9669 +[14:53:17.205442] Epoch: [17] Total time: 0:00:01 (0.1623 s / it) +[14:53:17.206336] Averaged stats: lr: 0.000416 loss: 0.4945 (0.4977) +[14:53:17.868663] val: [ 0/17] eta: 0:00:10 loss: 0.5233 (0.5233) time: 0.6452 data: 0.6282 max mem: 9669 +[14:53:18.115664] val: [10/17] eta: 0:00:00 loss: 0.5233 (0.5460) time: 0.0811 data: 0.0656 max mem: 9669 +[14:53:18.207042] val: [16/17] eta: 0:00:00 loss: 0.4244 (0.4287) time: 0.0578 data: 0.0425 max mem: 9669 +[14:53:18.284525] val: Total time: 0:00:01 (0.0625 s / it) +[14:53:18.298607] val loss: 0.4286675724913092 +[14:53:18.298780] Accuracy: 0.8222, F1 Score: 0.8216, ROC AUC: 0.9027, Hamming Loss: 0.1778, + Jaccard Score: 0.6973, Precision: 0.8268, Recall: 0.8222, + Average Precision: 0.8972, Kappa: 0.6444, Score: 0.7896 +[14:53:20.129040] Best epoch = 17, Best score = 0.7896 +[14:53:20.188369] log_dir: ./output_logs/retfound +[14:53:20.866557] Epoch: [18] [0/7] eta: 0:00:04 lr: 0.000409 loss: 0.5717 (0.5717) time: 0.6772 data: 0.6021 max mem: 9669 +[14:53:21.256740] Epoch: [18] [6/7] eta: 0:00:00 lr: 0.000369 loss: 0.4912 (0.4943) time: 0.1524 data: 0.0861 max mem: 9669 +[14:53:21.329540] Epoch: [18] Total time: 0:00:01 (0.1630 s / it) +[14:53:21.330426] Averaged stats: lr: 0.000369 loss: 0.4912 (0.4943) +[14:53:21.937833] val: [ 0/17] eta: 0:00:10 loss: 0.4034 (0.4034) time: 0.6002 data: 0.5837 max mem: 9669 +[14:53:22.206234] val: [10/17] eta: 0:00:00 loss: 0.4508 (0.4709) time: 0.0789 data: 0.0635 max mem: 9669 +[14:53:22.392934] val: [16/17] eta: 0:00:00 loss: 0.3712 (0.4013) time: 0.0620 data: 0.0467 max mem: 9669 +[14:53:22.467316] val: Total time: 0:00:01 (0.0665 s / it) +[14:53:22.481962] val loss: 0.4013392381808337 +[14:53:22.482265] Accuracy: 0.8370, F1 Score: 0.8370, ROC AUC: 0.9069, Hamming Loss: 0.1630, + Jaccard Score: 0.7197, Precision: 0.8375, Recall: 0.8370, + Average Precision: 0.9004, Kappa: 0.6741, Score: 0.8060 +[14:53:24.304411] Best epoch = 18, Best score = 0.8060 +[14:53:24.377997] log_dir: ./output_logs/retfound +[14:53:25.058555] Epoch: [19] [0/7] eta: 0:00:04 lr: 0.000362 loss: 0.5548 (0.5548) time: 0.6797 data: 0.6072 max mem: 9669 +[14:53:25.448355] Epoch: [19] [6/7] eta: 0:00:00 lr: 0.000320 loss: 0.4624 (0.4877) time: 0.1527 data: 0.0868 max mem: 9669 +[14:53:25.520051] Epoch: [19] Total time: 0:00:01 (0.1631 s / it) +[14:53:25.520906] Averaged stats: lr: 0.000320 loss: 0.4624 (0.4877) +[14:53:26.116139] val: [ 0/17] eta: 0:00:09 loss: 0.4736 (0.4736) time: 0.5716 data: 0.5547 max mem: 9669 +[14:53:26.284031] val: [10/17] eta: 0:00:00 loss: 0.4736 (0.5099) time: 0.0672 data: 0.0517 max mem: 9669 +[14:53:26.375105] val: [16/17] eta: 0:00:00 loss: 0.4029 (0.4122) time: 0.0488 data: 0.0335 max mem: 9669 +[14:53:26.454731] val: Total time: 0:00:00 (0.0536 s / it) +[14:53:26.469416] val loss: 0.412210370249608 +[14:53:26.469604] Accuracy: 0.8389, F1 Score: 0.8386, ROC AUC: 0.9090, Hamming Loss: 0.1611, + Jaccard Score: 0.7221, Precision: 0.8414, Recall: 0.8389, + Average Precision: 0.9017, Kappa: 0.6778, Score: 0.8085 +[14:53:28.263453] Best epoch = 19, Best score = 0.8085 +[14:53:28.344256] log_dir: ./output_logs/retfound +[14:53:29.011820] Epoch: [20] [0/7] eta: 0:00:04 lr: 0.000313 loss: 0.4294 (0.4294) time: 0.6667 data: 0.5950 max mem: 9669 +[14:53:29.418532] Epoch: [20] [6/7] eta: 0:00:00 lr: 0.000271 loss: 0.4430 (0.4732) time: 0.1532 data: 0.0851 max mem: 9669 +[14:53:29.489194] Epoch: [20] Total time: 0:00:01 (0.1635 s / it) +[14:53:29.490189] Averaged stats: lr: 0.000271 loss: 0.4430 (0.4732) +[14:53:30.026561] val: [ 0/17] eta: 0:00:08 loss: 0.3810 (0.3810) time: 0.5117 data: 0.4890 max mem: 9669 +[14:53:30.196093] val: [10/17] eta: 0:00:00 loss: 0.4738 (0.4699) time: 0.0619 data: 0.0458 max mem: 9669 +[14:53:30.287502] val: [16/17] eta: 0:00:00 loss: 0.3462 (0.4156) time: 0.0454 data: 0.0297 max mem: 9669 +[14:53:30.370608] val: Total time: 0:00:00 (0.0504 s / it) +[14:53:30.384339] val loss: 0.4155636272009681 +[14:53:30.384525] Accuracy: 0.8426, F1 Score: 0.8425, ROC AUC: 0.9090, Hamming Loss: 0.1574, + Jaccard Score: 0.7279, Precision: 0.8430, Recall: 0.8426, + Average Precision: 0.9019, Kappa: 0.6852, Score: 0.8123 +[14:53:32.172724] Best epoch = 20, Best score = 0.8123 +[14:53:32.248076] log_dir: ./output_logs/retfound +[14:53:32.895252] Epoch: [21] [0/7] eta: 0:00:04 lr: 0.000264 loss: 0.4277 (0.4277) time: 0.6462 data: 0.5732 max mem: 9669 +[14:53:33.286833] Epoch: [21] [6/7] eta: 0:00:00 lr: 0.000223 loss: 0.4395 (0.4543) time: 0.1481 data: 0.0820 max mem: 9669 +[14:53:33.355085] Epoch: [21] Total time: 0:00:01 (0.1581 s / it) +[14:53:33.355915] Averaged stats: lr: 0.000223 loss: 0.4395 (0.4543) +[14:53:33.960611] val: [ 0/17] eta: 0:00:09 loss: 0.3704 (0.3704) time: 0.5801 data: 0.5636 max mem: 9669 +[14:53:34.115173] val: [10/17] eta: 0:00:00 loss: 0.4418 (0.4749) time: 0.0667 data: 0.0513 max mem: 9669 +[14:53:34.206601] val: [16/17] eta: 0:00:00 loss: 0.3704 (0.4342) time: 0.0485 data: 0.0333 max mem: 9669 +[14:53:34.285941] val: Total time: 0:00:00 (0.0533 s / it) +[14:53:34.300520] val loss: 0.43423646162537965 +[14:53:34.300722] Accuracy: 0.8333, F1 Score: 0.8332, ROC AUC: 0.9093, Hamming Loss: 0.1667, + Jaccard Score: 0.7141, Precision: 0.8342, Recall: 0.8333, + Average Precision: 0.9032, Kappa: 0.6667, Score: 0.8031 +[14:53:34.324026] Best epoch = 20, Best score = 0.8123 +[14:53:34.595924] log_dir: ./output_logs/retfound +[14:53:35.282867] Epoch: [22] [0/7] eta: 0:00:04 lr: 0.000217 loss: 0.5146 (0.5146) time: 0.6858 data: 0.6033 max mem: 9669 +[14:53:35.680837] Epoch: [22] [6/7] eta: 0:00:00 lr: 0.000178 loss: 0.4722 (0.4499) time: 0.1547 data: 0.0863 max mem: 9669 +[14:53:35.753053] Epoch: [22] Total time: 0:00:01 (0.1653 s / it) +[14:53:35.753813] Averaged stats: lr: 0.000178 loss: 0.4722 (0.4499) +[14:53:36.526767] val: [ 0/17] eta: 0:00:12 loss: 0.4352 (0.4352) time: 0.7516 data: 0.7348 max mem: 9669 +[14:53:36.776122] val: [10/17] eta: 0:00:00 loss: 0.4630 (0.5175) time: 0.0909 data: 0.0755 max mem: 9669 +[14:53:36.867454] val: [16/17] eta: 0:00:00 loss: 0.3808 (0.4334) time: 0.0642 data: 0.0489 max mem: 9669 +[14:53:36.947975] val: Total time: 0:00:01 (0.0690 s / it) +[14:53:36.967154] val loss: 0.43337904968682456 +[14:53:36.967380] Accuracy: 0.8444, F1 Score: 0.8444, ROC AUC: 0.9128, Hamming Loss: 0.1556, + Jaccard Score: 0.7307, Precision: 0.8446, Recall: 0.8444, + Average Precision: 0.9072, Kappa: 0.6889, Score: 0.8154 +[14:53:38.791982] Best epoch = 22, Best score = 0.8154 +[14:53:38.865829] log_dir: ./output_logs/retfound +[14:53:39.579413] Epoch: [23] [0/7] eta: 0:00:04 lr: 0.000171 loss: 0.4765 (0.4765) time: 0.7126 data: 0.6402 max mem: 9669 +[14:53:39.970205] Epoch: [23] [6/7] eta: 0:00:00 lr: 0.000135 loss: 0.4252 (0.4542) time: 0.1575 data: 0.0915 max mem: 9669 +[14:53:40.038344] Epoch: [23] Total time: 0:00:01 (0.1675 s / it) +[14:53:40.039180] Averaged stats: lr: 0.000135 loss: 0.4252 (0.4542) +[14:53:40.598867] val: [ 0/17] eta: 0:00:09 loss: 0.4179 (0.4179) time: 0.5350 data: 0.5184 max mem: 9669 +[14:53:40.869462] val: [10/17] eta: 0:00:00 loss: 0.4718 (0.5078) time: 0.0732 data: 0.0577 max mem: 9669 +[14:53:40.961404] val: [16/17] eta: 0:00:00 loss: 0.3612 (0.4288) time: 0.0527 data: 0.0374 max mem: 9669 +[14:53:41.042954] val: Total time: 0:00:00 (0.0576 s / it) +[14:53:41.059001] val loss: 0.4288011767408427 +[14:53:41.059202] Accuracy: 0.8444, F1 Score: 0.8444, ROC AUC: 0.9144, Hamming Loss: 0.1556, + Jaccard Score: 0.7308, Precision: 0.8445, Recall: 0.8444, + Average Precision: 0.9090, Kappa: 0.6889, Score: 0.8159 +[14:53:42.938412] Best epoch = 23, Best score = 0.8159 +[14:53:43.008057] log_dir: ./output_logs/retfound +[14:53:43.848713] Epoch: [24] [0/7] eta: 0:00:05 lr: 0.000130 loss: 0.4958 (0.4958) time: 0.8397 data: 0.7618 max mem: 9669 +[14:53:44.241480] Epoch: [24] [6/7] eta: 0:00:00 lr: 0.000097 loss: 0.4856 (0.4696) time: 0.1760 data: 0.1089 max mem: 9669 +[14:53:44.318924] Epoch: [24] Total time: 0:00:01 (0.1872 s / it) +[14:53:44.319935] Averaged stats: lr: 0.000097 loss: 0.4856 (0.4696) +[14:53:45.077188] val: [ 0/17] eta: 0:00:12 loss: 0.4888 (0.4888) time: 0.7268 data: 0.7080 max mem: 9669 +[14:53:45.337973] val: [10/17] eta: 0:00:00 loss: 0.4888 (0.5389) time: 0.0897 data: 0.0738 max mem: 9669 +[14:53:45.429173] val: [16/17] eta: 0:00:00 loss: 0.4037 (0.4246) time: 0.0634 data: 0.0478 max mem: 9669 +[14:53:45.510368] val: Total time: 0:00:01 (0.0686 s / it) +[14:53:45.527785] val loss: 0.4246067654560594 +[14:53:45.528079] Accuracy: 0.8463, F1 Score: 0.8462, ROC AUC: 0.9176, Hamming Loss: 0.1537, + Jaccard Score: 0.7334, Precision: 0.8474, Recall: 0.8463, + Average Precision: 0.9125, Kappa: 0.6926, Score: 0.8188 +[14:53:47.325542] Best epoch = 24, Best score = 0.8188 +[14:53:47.398308] log_dir: ./output_logs/retfound +[14:53:48.205548] Epoch: [25] [0/7] eta: 0:00:05 lr: 0.000092 loss: 0.4557 (0.4557) time: 0.8063 data: 0.7345 max mem: 9669 +[14:53:48.594591] Epoch: [25] [6/7] eta: 0:00:00 lr: 0.000065 loss: 0.5138 (0.4867) time: 0.1707 data: 0.1050 max mem: 9669 +[14:53:48.665208] Epoch: [25] Total time: 0:00:01 (0.1810 s / it) +[14:53:48.666071] Averaged stats: lr: 0.000065 loss: 0.5138 (0.4867) +[14:53:49.333337] val: [ 0/17] eta: 0:00:11 loss: 0.5922 (0.5922) time: 0.6517 data: 0.6346 max mem: 9669 +[14:53:49.633217] val: [10/17] eta: 0:00:00 loss: 0.5922 (0.5937) time: 0.0865 data: 0.0710 max mem: 9669 +[14:53:49.725094] val: [16/17] eta: 0:00:00 loss: 0.4581 (0.4427) time: 0.0613 data: 0.0460 max mem: 9669 +[14:53:49.804789] val: Total time: 0:00:01 (0.0661 s / it) +[14:53:49.820652] val loss: 0.4427410313750015 +[14:53:49.820897] Accuracy: 0.8407, F1 Score: 0.8399, ROC AUC: 0.9190, Hamming Loss: 0.1593, + Jaccard Score: 0.7242, Precision: 0.8476, Recall: 0.8407, + Average Precision: 0.9144, Kappa: 0.6815, Score: 0.8135 +[14:53:49.874662] Best epoch = 24, Best score = 0.8188 +[14:53:50.136369] log_dir: ./output_logs/retfound +[14:53:50.939516] Epoch: [26] [0/7] eta: 0:00:05 lr: 0.000061 loss: 0.4565 (0.4565) time: 0.8023 data: 0.7342 max mem: 9669 +[14:53:51.329879] Epoch: [26] [6/7] eta: 0:00:00 lr: 0.000038 loss: 0.4565 (0.4508) time: 0.1703 data: 0.1049 max mem: 9669 +[14:53:51.405034] Epoch: [26] Total time: 0:00:01 (0.1812 s / it) +[14:53:51.405806] Averaged stats: lr: 0.000038 loss: 0.4565 (0.4508) +[14:53:52.150391] val: [ 0/17] eta: 0:00:12 loss: 0.6236 (0.6236) time: 0.7319 data: 0.7154 max mem: 9669 +[14:53:52.373605] val: [10/17] eta: 0:00:00 loss: 0.6236 (0.6092) time: 0.0868 data: 0.0713 max mem: 9669 +[14:53:52.464952] val: [16/17] eta: 0:00:00 loss: 0.4918 (0.4484) time: 0.0615 data: 0.0462 max mem: 9669 +[14:53:52.542502] val: Total time: 0:00:01 (0.0662 s / it) +[14:53:52.556443] val loss: 0.4483622448409305 +[14:53:52.556668] Accuracy: 0.8389, F1 Score: 0.8380, ROC AUC: 0.9197, Hamming Loss: 0.1611, + Jaccard Score: 0.7213, Precision: 0.8469, Recall: 0.8389, + Average Precision: 0.9150, Kappa: 0.6778, Score: 0.8118 +[14:53:52.605519] Best epoch = 24, Best score = 0.8188 +[14:53:52.865284] log_dir: ./output_logs/retfound +[14:53:53.580675] Epoch: [27] [0/7] eta: 0:00:05 lr: 0.000035 loss: 0.4393 (0.4393) time: 0.7145 data: 0.6452 max mem: 9669 +[14:53:53.970803] Epoch: [27] [6/7] eta: 0:00:00 lr: 0.000019 loss: 0.4530 (0.4852) time: 0.1577 data: 0.0922 max mem: 9669 +[14:53:54.047005] Epoch: [27] Total time: 0:00:01 (0.1688 s / it) +[14:53:54.047693] Averaged stats: lr: 0.000019 loss: 0.4530 (0.4852) +[14:53:54.622350] val: [ 0/17] eta: 0:00:09 loss: 0.6208 (0.6208) time: 0.5580 data: 0.5411 max mem: 9669 +[14:53:54.891362] val: [10/17] eta: 0:00:00 loss: 0.6208 (0.6081) time: 0.0751 data: 0.0597 max mem: 9669 +[14:53:54.982353] val: [16/17] eta: 0:00:00 loss: 0.4903 (0.4478) time: 0.0539 data: 0.0387 max mem: 9669 +[14:53:55.200053] val: Total time: 0:00:01 (0.0669 s / it) +[14:53:55.214110] val loss: 0.4477856479146901 +[14:53:55.214343] Accuracy: 0.8407, F1 Score: 0.8399, ROC AUC: 0.9202, Hamming Loss: 0.1593, + Jaccard Score: 0.7241, Precision: 0.8484, Recall: 0.8407, + Average Precision: 0.9158, Kappa: 0.6815, Score: 0.8138 +[14:53:55.417440] Best epoch = 24, Best score = 0.8188 +[14:53:56.007662] log_dir: ./output_logs/retfound +[14:53:56.822789] Epoch: [28] [0/7] eta: 0:00:05 lr: 0.000016 loss: 0.4691 (0.4691) time: 0.8142 data: 0.7420 max mem: 9669 +[14:53:57.222635] Epoch: [28] [6/7] eta: 0:00:00 lr: 0.000006 loss: 0.3765 (0.4270) time: 0.1733 data: 0.1061 max mem: 9669 +[14:53:57.295254] Epoch: [28] Total time: 0:00:01 (0.1839 s / it) +[14:53:57.296170] Averaged stats: lr: 0.000006 loss: 0.3765 (0.4270) +[14:53:57.914866] val: [ 0/17] eta: 0:00:10 loss: 0.6065 (0.6065) time: 0.6046 data: 0.5876 max mem: 9669 +[14:53:58.069914] val: [10/17] eta: 0:00:00 loss: 0.6065 (0.5994) time: 0.0690 data: 0.0535 max mem: 9669 +[14:53:58.161224] val: [16/17] eta: 0:00:00 loss: 0.4781 (0.4435) time: 0.0500 data: 0.0347 max mem: 9669 +[14:53:58.247070] val: Total time: 0:00:00 (0.0552 s / it) +[14:53:58.260868] val loss: 0.4435284231953761 +[14:53:58.261096] Accuracy: 0.8389, F1 Score: 0.8381, ROC AUC: 0.9202, Hamming Loss: 0.1611, + Jaccard Score: 0.7215, Precision: 0.8454, Recall: 0.8389, + Average Precision: 0.9155, Kappa: 0.6778, Score: 0.8120 +[14:53:58.291609] Best epoch = 24, Best score = 0.8188 +[14:53:58.578060] log_dir: ./output_logs/retfound +[14:53:59.300310] Epoch: [29] [0/7] eta: 0:00:05 lr: 0.000005 loss: 0.5394 (0.5394) time: 0.7215 data: 0.6522 max mem: 9669 +[14:53:59.692210] Epoch: [29] [6/7] eta: 0:00:00 lr: 0.000001 loss: 0.4717 (0.4919) time: 0.1589 data: 0.0932 max mem: 9669 +[14:53:59.765605] Epoch: [29] Total time: 0:00:01 (0.1696 s / it) +[14:53:59.766529] Averaged stats: lr: 0.000001 loss: 0.4717 (0.4919) +[14:54:00.471375] val: [ 0/17] eta: 0:00:11 loss: 0.6036 (0.6036) time: 0.6798 data: 0.6632 max mem: 9669 +[14:54:00.717537] val: [10/17] eta: 0:00:00 loss: 0.6036 (0.5976) time: 0.0841 data: 0.0687 max mem: 9669 +[14:54:00.808409] val: [16/17] eta: 0:00:00 loss: 0.4754 (0.4427) time: 0.0598 data: 0.0445 max mem: 9669 +[14:54:00.888042] val: Total time: 0:00:01 (0.0645 s / it) +[14:54:00.901736] val loss: 0.4426653034546796 +[14:54:00.901955] Accuracy: 0.8389, F1 Score: 0.8381, ROC AUC: 0.9202, Hamming Loss: 0.1611, + Jaccard Score: 0.7215, Precision: 0.8454, Recall: 0.8389, + Average Precision: 0.9154, Kappa: 0.6778, Score: 0.8120 +[14:54:00.942973] Best epoch = 24, Best score = 0.8188 +[14:54:03.292488] Test with the best model, epoch = 24: +[14:54:04.028130] test: [ 0/32] eta: 0:00:23 loss: 0.3909 (0.3909) time: 0.7197 data: 0.7020 max mem: 9669 +[14:54:04.296408] test: [10/32] eta: 0:00:01 loss: 0.4804 (0.5666) time: 0.0898 data: 0.0742 max mem: 9669 +[14:54:04.517702] test: [20/32] eta: 0:00:00 loss: 0.4135 (0.4746) time: 0.0244 data: 0.0091 max mem: 9669 +[14:54:04.769694] test: [30/32] eta: 0:00:00 loss: 0.3382 (0.4376) time: 0.0236 data: 0.0082 max mem: 9669 +[14:54:04.811708] test: [31/32] eta: 0:00:00 loss: 0.3382 (0.4487) time: 0.0249 data: 0.0081 max mem: 9669 +[14:54:04.886921] test: Total time: 0:00:01 (0.0494 s / it) +[14:54:04.906166] val loss: 0.4487141091376543 +[14:54:04.906275] Accuracy: 0.8340, F1 Score: 0.8340, ROC AUC: 0.9093, Hamming Loss: 0.1660, + Jaccard Score: 0.7152, Precision: 0.8343, Recall: 0.8340, + Average Precision: 0.9037, Kappa: 0.6680, Score: 0.8037 +[14:54:05.641988] Training time 0:01:48 +[rank0]:[W701 14:54:06.090698911 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/005/retfound acc=0.8340 auroc=0.9092500000000001 f1_macro=0.8340 qwk=0.6679999999999999 diff --git a/results/downsample/airogs/005/vit/confusion_matrix.png b/results/downsample/airogs/005/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..f47ea43054fcea6c0fc404c0c843935cbce6fcea --- /dev/null +++ b/results/downsample/airogs/005/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20f24b9dfc5c9ed3cbafff519e10908dae0919d7112b5d31397359c53652d991 +size 71207 diff --git a/results/downsample/airogs/005/vit/log.csv b/results/downsample/airogs/005/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..8c9cd22c7a063bae851fdb44b654c05f06cedb0f --- /dev/null +++ b/results/downsample/airogs/005/vit/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.775609294573466,0.5537037037037037,0.5681893004115226,0.4089867672914475,4.9296078546916196e-08 +1,0.8097381989161173,0.5370370370370371,0.621570644718793,0.39012035520671784,1.232401963672905e-07 +2,0.7171350518862406,0.5462962962962963,0.6535116598079561,0.40230983994944425,1.9718431418766478e-07 +3,0.6705105702082316,0.6759259259259259,0.7256858710562415,0.5844875124883568,2.711284320080391e-07 +4,0.6414447029431661,0.7092592592592593,0.7563374485596709,0.6280221208383362,3.450725498284134e-07 +5,0.5507955551147461,0.7129629629629629,0.7752606310013718,0.6362687533119952,3.6907225802970327e-07 +6,0.5811153451601664,0.7,0.7821947873799725,0.6273310018870406,3.6568094813687817e-07 +7,0.5216130117575327,0.7314814814814815,0.8011591220850479,0.6640442705923163,3.594379885199801e-07 +8,0.49695590138435364,0.7425925925925926,0.8110493827160494,0.6788368286036617,3.504418343815308e-07 +9,0.4638751645882924,0.7481481481481481,0.8154183813443072,0.6863946225938161,3.388343604463356e-07 +10,0.4936785399913788,0.7333333333333333,0.8219478737997257,0.6715576078464999,3.2479862351232145e-07 +11,0.44932188590367633,0.7592592592592593,0.8279903978052127,0.7017362609851888,3.0855597553548053e-07 +12,0.3904678523540497,0.7592592592592593,0.8315912208504801,0.7019797411601747,2.9036257277729527e-07 +13,0.41381944219271344,0.7703703703703704,0.8341906721536352,0.7145890097775774,2.7050533606747414e-07 +14,0.37631357709566754,0.7592592592592593,0.8352812071330591,0.7031360659637356,2.4929742589106926e-07 +15,0.39292670289675397,0.7722222222222223,0.8369958847736626,0.7173079539593573,2.2707330366055404e-07 +16,0.3657362361749013,0.7685185185185185,0.836508916323731,0.7138281059301201,2.041834570595508e-07 +17,0.40728189547856647,0.7462962962962963,0.8386968449931413,0.6909543271784493,1.8098887264268902e-07 +18,0.3480362097422282,0.7722222222222223,0.8402057613168725,0.7185172444764634,1.5785534286199555e-07 +19,0.34865578015645343,0.7722222222222223,0.8403429355281207,0.7189278771338167,1.3514769730141294e-07 +20,0.359431525071462,0.75,0.839272976680384,0.6950411454486369,1.1322404909632138e-07 +21,0.3400772114594777,0.7425925925925926,0.8403566529492454,0.6874031506799655,9.243014727546784e-08 +22,0.3408369819323222,0.7518518518518519,0.8428532235939643,0.6986953898849264,7.309392409224211e-08 +23,0.35057754317919415,0.7648148148148148,0.8425240054869684,0.711992092750159,5.552032333714334e-08 +24,0.33229197065035504,0.7537037037037037,0.8420576131687243,0.7004295592808661,3.998649119203978e-08 +25,0.31483269731203717,0.7462962962962963,0.842235939643347,0.6928217894712366,2.6737405469321262e-08 +26,0.31975247462590534,0.7481481481481481,0.8423936899862826,0.6948269530408774,1.5982012165455135e-08 +27,0.3196420570214589,0.7481481481481481,0.8426680384087792,0.6949781957376194,7.889930257803989e-09 +28,0.3443579177061717,0.7481481481481481,0.8429012345679012,0.6950559277906602,2.588776712026834e-09 +29,0.3369385798772176,0.75,0.8429423868312756,0.6969556088013422,1.6215388629625446e-10 diff --git a/results/downsample/airogs/005/vit/metrics.json b/results/downsample/airogs/005/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..8b5a824b6434c98816eb470716149f3abb0d2550 --- /dev/null +++ b/results/downsample/airogs/005/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.766, + "balanced_accuracy": 0.766, + "precision_macro": 0.7662087076267794, + "recall_macro": 0.766, + "f1_macro": 0.7659541270088938, + "precision_weighted": 0.7662087076267794, + "recall_weighted": 0.766, + "f1_weighted": 0.7659541270088936, + "cohen_kappa": 0.532, + "quadratic_weighted_kappa": 0.532, + "mcc": 0.5322086667040412, + "auroc": 0.8468760000000001, + "auprc": 0.8482196596874365, + "sensitivity": 0.78, + "specificity": 0.752, + "precision_pos": 0.7587548638132295, + "f1_pos": 0.7692307692307693, + "per_class": { + "0": { + "precision": 0.7736625514403292, + "recall": 0.752, + "f1-score": 0.7626774847870182, + "support": 500.0 + }, + "1": { + "precision": 0.7587548638132295, + "recall": 0.78, + "f1-score": 0.7692307692307693, + "support": 500.0 + }, + "accuracy": 0.766, + "macro avg": { + "precision": 0.7662087076267794, + "recall": 0.766, + "f1-score": 0.7659541270088938, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.7662087076267794, + "recall": 0.766, + "f1-score": 0.7659541270088936, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/005/vit/pr.png b/results/downsample/airogs/005/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..275baa5b79589f823418115318f6add913c7e300 --- /dev/null +++ b/results/downsample/airogs/005/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89c9acea2e861b8661782f19dea1d7c633e6a34dbda88f1e6e18a78889232341 +size 54307 diff --git a/results/downsample/airogs/005/vit/roc.png b/results/downsample/airogs/005/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..6b356260cbe6235ed7cedfa4e60d9af7f0b3d725 --- /dev/null +++ b/results/downsample/airogs/005/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e360fa09b6d33865116ba97d0ee215808a7c5c95f63f893f16e21a6d267ce7ee +size 66267 diff --git a/results/downsample/airogs/005/vit/test_pred.npz b/results/downsample/airogs/005/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..cc2e03ca5cba91d3b95e46ad3613239726cfa8c5 --- /dev/null +++ b/results/downsample/airogs/005/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c37e8a77f3e583fb865cbfee7ec2192f74222f4637cc94977268c45e67c28c24 +size 16510 diff --git a/results/downsample/airogs/005/vit/train.log b/results/downsample/airogs/005/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d8638a647064d1f130db0f39170ef6ccfde4058e --- /dev/null +++ b/results/downsample/airogs/005/vit/train.log @@ -0,0 +1,159 @@ +[vit] train=250 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7756 val_acc=0.5537 val_auc=0.5682 score=0.4090 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.8097 val_acc=0.5370 val_auc=0.6216 score=0.3901 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.7171 val_acc=0.5463 val_auc=0.6535 score=0.4023 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6705 val_acc=0.6759 val_auc=0.7257 score=0.5845 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6414 val_acc=0.7093 val_auc=0.7563 score=0.6280 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5508 val_acc=0.7130 val_auc=0.7753 score=0.6363 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5811 val_acc=0.7000 val_auc=0.7822 score=0.6273 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5216 val_acc=0.7315 val_auc=0.8012 score=0.6640 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.4970 val_acc=0.7426 val_auc=0.8110 score=0.6788 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4639 val_acc=0.7481 val_auc=0.8154 score=0.6864 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4937 val_acc=0.7333 val_auc=0.8219 score=0.6716 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4493 val_acc=0.7593 val_auc=0.8280 score=0.7017 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3905 val_acc=0.7593 val_auc=0.8316 score=0.7020 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4138 val_acc=0.7704 val_auc=0.8342 score=0.7146 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.3763 val_acc=0.7593 val_auc=0.8353 score=0.7031 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.3929 val_acc=0.7722 val_auc=0.8370 score=0.7173 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3657 val_acc=0.7685 val_auc=0.8365 score=0.7138 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.4073 val_acc=0.7463 val_auc=0.8387 score=0.6910 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.3480 val_acc=0.7722 val_auc=0.8402 score=0.7185 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.3487 val_acc=0.7722 val_auc=0.8403 score=0.7189 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3594 val_acc=0.7500 val_auc=0.8393 score=0.6950 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.3401 val_acc=0.7426 val_auc=0.8404 score=0.6874 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3408 val_acc=0.7519 val_auc=0.8429 score=0.6987 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.3506 val_acc=0.7648 val_auc=0.8425 score=0.7120 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3323 val_acc=0.7537 val_auc=0.8421 score=0.7004 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3148 val_acc=0.7463 val_auc=0.8422 score=0.6928 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.3198 val_acc=0.7481 val_auc=0.8424 score=0.6948 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3196 val_acc=0.7481 val_auc=0.8427 score=0.6950 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3444 val_acc=0.7481 val_auc=0.8429 score=0.6951 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.3369 val_acc=0.7500 val_auc=0.8429 score=0.6970 +[vit] early stop at ep29 (best ep19 score=0.7189) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=19 best_val_score=0.7189 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/005/vit acc=0.7660 auroc=0.8468760000000001 f1_macro=0.7660 qwk=0.532 diff --git a/results/downsample/airogs/010/airogs_010pct/confusion_matrix_test.jpg b/results/downsample/airogs/010/airogs_010pct/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..79c937a632da31c1b263f3e4d5018f4ca2fb3fa0 --- /dev/null +++ b/results/downsample/airogs/010/airogs_010pct/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:375ae594b5f17caaa4da28079748793ef1ad059acf177ceae40f5a88ab1bc3cc +size 254261 diff --git a/results/downsample/airogs/010/airogs_010pct/log.txt b/results/downsample/airogs/010/airogs_010pct/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..9a1e401572f548bc96d2e1a17c6a9f64ba910241 --- /dev/null +++ b/results/downsample/airogs/010/airogs_010pct/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 2.916666666666667e-05, "train_loss": 0.6930694580078125, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.166666666666667e-05, "train_loss": 0.6926737467447917, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015416666666666668, "train_loss": 0.6894877115885417, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021666666666666663, "train_loss": 0.6831080118815104, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.00027916666666666666, "train_loss": 0.6410694122314453, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003416666666666667, "train_loss": 0.6245920817057292, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.0004041666666666666, "train_loss": 0.5560907046000163, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046666666666666666, "train_loss": 0.5844362258911133, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005291666666666667, "train_loss": 0.5735061963399252, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005916666666666667, "train_loss": 0.5443812370300293, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006238437351402742, "train_loss": 0.5428811073303222, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.000616445813265531, "train_loss": 0.5065884987513224, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006015760483892742, "train_loss": 0.5185736258824666, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005796005834726778, "train_loss": 0.5371347904205322, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005510605273943752, "train_loss": 0.5367340127627055, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005166586310540296, "train_loss": 0.5101562579472859, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00047724198331288874, "train_loss": 0.5326651255289714, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004337811528546931, "train_loss": 0.5179853280385335, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003873462895635502, "train_loss": 0.502229102452596, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00033908077388216404, "train_loss": 0.46000899076461793, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002901730629906729, "train_loss": 0.5129401405652364, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024182742704661195, "train_loss": 0.4756422201792399, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019523429605689067, "train_loss": 0.45967623392740886, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015154094754021218, "train_loss": 0.4458814938863119, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011182325674696714, "train_loss": 0.43889840443929035, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.705920503994361e-05, "train_loss": 0.460331384340922, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.8104798747441377e-05, "train_loss": 0.445121971766154, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5672991446547074e-05, "train_loss": 0.47218621969223024, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.0316128679908865e-05, "train_loss": 0.4452753702799479, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.4123473753469123e-06, "train_loss": 0.4721079548199972, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/010/airogs_010pct/metrics_test.csv b/results/downsample/airogs/010/airogs_010pct/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..9a2ad9eeb21055b453b1e77c5f18eb2184aa93de --- /dev/null +++ b/results/downsample/airogs/010/airogs_010pct/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.35092977061867714,0.851,0.850981968818227,0.93294,0.149,0.7406202654384583,0.8511699662636716,0.851,0.9354180474245168,0.702 diff --git a/results/downsample/airogs/010/airogs_010pct/metrics_val.csv b/results/downsample/airogs/010/airogs_010pct/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..3c5fa5ffeaf64fc9858cdd049951798c5447b8a3 --- /dev/null +++ b/results/downsample/airogs/010/airogs_010pct/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6911632629001841,0.6259259259259259,0.5981373141366657,0.7359979423868312,0.37407407407407406,0.43482654600301657,0.6740746359223301,0.6259259259259259,0.7150930878518238,0.2518518518518519 +0.6851682277286754,0.5351851851851852,0.42195249894446835,0.7583676268861455,0.4648148148148148,0.30160053468411424,0.6625578300272514,0.5351851851851852,0.7497576896332707,0.07037037037037042 +0.666696243426379,0.6777777777777778,0.6632933891405309,0.7883504801097394,0.32222222222222224,0.5003097021228673,0.7147259593081052,0.6777777777777778,0.7836422952565592,0.3555555555555555 +0.6354804144186132,0.7203703703703703,0.711046773284761,0.8127537722908094,0.2796296296296296,0.5541668803090123,0.7530279882188027,0.7203703703703703,0.8098056613082846,0.44074074074074077 +0.5088622307076174,0.8,0.7998654788540995,0.8703840877914952,0.2,0.6665110028877219,0.8008087588028169,0.8,0.86427274750956,0.6 +0.4414682791513555,0.8185185185185185,0.8177962016774318,0.9028395061728395,0.1814814814814815,0.6919150110375276,0.8236507582515611,0.8185185185185185,0.8998043365253571,0.6370370370370371 +0.41574880831381855,0.8,0.7975,0.9084053497942387,0.2,0.663784157730889,0.8155844155844156,0.8,0.9096725989660821,0.6 +0.4472920228453243,0.7703703703703704,0.763505877034358,0.9149725651577504,0.22962962962962963,0.6191956191956192,0.8058849090570488,0.7703703703703704,0.9141185126032487,0.5407407407407407 +0.35090961175806384,0.8629629629629629,0.8629328796443664,0.9246399176954733,0.13703703703703704,0.7589167276856303,0.8632818935691142,0.8629629629629629,0.9242505611165823,0.7259259259259259 +0.35795334419783426,0.8314814814814815,0.8306865910257691,0.9277674897119342,0.1685185185185185,0.7105738018911671,0.8378255581496135,0.8314814814814815,0.9260910332760703,0.662962962962963 +0.3373081123127657,0.8592592592592593,0.8591355922895272,0.9322633744855967,0.14074074074074075,0.7530801687763713,0.8605253014701834,0.8592592592592592,0.9319354267309927,0.7185185185185186 +0.3457377899219008,0.8722222222222222,0.8722007470391115,0.9341015089163236,0.12777777777777777,0.7733690727876774,0.8724725810215371,0.8722222222222222,0.9344763869703266,0.7444444444444445 +0.3471236675977707,0.8444444444444444,0.84367030149299,0.9385082304526748,0.15555555555555556,0.7297670725797305,0.8514050604567847,0.8444444444444444,0.9388250106525008,0.6888888888888889 +0.3746415477465181,0.8592592592592593,0.8586309523809523,0.9287997256515775,0.14074074074074075,0.7524011000634652,0.8657616892911011,0.8592592592592593,0.9285921612505711,0.7185185185185186 +0.3563896426383187,0.85,0.8499130154307537,0.9300685871056242,0.15,0.7390160634235431,0.8508132708198706,0.8500000000000001,0.9299398841987272,0.7 +0.34869736564510007,0.8629629629629629,0.8629460427213236,0.9297050754458163,0.13703703703703704,0.7589346349745331,0.8631422924901186,0.8629629629629629,0.9298228063950114,0.7259259259259259 +0.3465885931954664,0.8537037037037037,0.8535223647931767,0.9320233196159122,0.14629629629629629,0.7445088819226751,0.8554639573195109,0.8537037037037036,0.9327697570740787,0.7074074074074075 +0.32848908708376046,0.8648148148148148,0.8647364213141361,0.9365432098765432,0.13518518518518519,0.7617199148029818,0.8656625097963729,0.8648148148148148,0.9377488652017925,0.7296296296296296 +0.345769933260539,0.8666666666666667,0.8662538699690403,0.9416975308641975,0.13333333333333333,0.764139127083534,0.87125,0.8666666666666667,0.9431990334709781,0.7333333333333334 +0.3735544913831879,0.8611111111111112,0.8611068242847002,0.942218792866941,0.1388888888888889,0.756091757091186,0.8611556982343499,0.8611111111111112,0.9429574694430065,0.7222222222222222 +0.32237282614497575,0.8592592592592593,0.8592573286327658,0.9431069958847738,0.14074074074074075,0.7532441520930184,0.8592789727831431,0.8592592592592592,0.94430975560212,0.7185185185185186 +0.3205309624181074,0.8611111111111112,0.8611106348101332,0.9429492455418382,0.1388888888888889,0.7560969161132027,0.8611160646922454,0.8611111111111112,0.9442572637585742,0.7222222222222222 +0.34546037412741604,0.8629629629629629,0.8629554433713784,0.9429149519890261,0.13703703703703704,0.7589474242665393,0.8630426431041107,0.8629629629629629,0.9441533928336208,0.7259259259259259 +0.36230065454454985,0.8629629629629629,0.8629554433713784,0.9434190672153635,0.13703703703703704,0.7589474242665393,0.8630426431041107,0.8629629629629629,0.944393697312522,0.7259259259259259 +0.3588551642263637,0.8592592592592593,0.8592573286327658,0.9438717421124828,0.14074074074074075,0.7532441520930184,0.8592789727831431,0.8592592592592592,0.9448359980656468,0.7185185185185186 +0.34836164555128885,0.8574074074074074,0.8574030062656255,0.943810013717421,0.1425925925925926,0.7503992854891247,0.8574515372268181,0.8574074074074074,0.9446549764809626,0.7148148148148148 +0.34843237303635655,0.8611111111111112,0.8610877685207734,0.944221536351166,0.1388888888888889,0.7560659587070571,0.8613539965134316,0.8611111111111112,0.945226322842136,0.7222222222222222 +0.3486549998907482,0.8574074074074074,0.8573482137649711,0.9443106995884774,0.1425925925925926,0.7503258287612201,0.8580016213468171,0.8574074074074074,0.9449875403055257,0.7148148148148148 +0.3487606730092974,0.8574074074074074,0.8573482137649711,0.9442318244170096,0.1425925925925926,0.7503258287612201,0.8580016213468171,0.8574074074074074,0.9449357635372877,0.7148148148148148 +0.3486666948918034,0.8592592592592593,0.8592109777015438,0.944303840877915,0.14074074074074075,0.7531817079471526,0.8597527472527473,0.8592592592592593,0.9453097835253261,0.7185185185185186 diff --git a/results/downsample/airogs/010/airogs_010pct/test_pred.npz b/results/downsample/airogs/010/airogs_010pct/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..abfb3e9630e3ee6c9b0fcec560ded0b1d22573f5 --- /dev/null +++ b/results/downsample/airogs/010/airogs_010pct/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:694a752e3b3d3cd62e4a2628391d35a387c8ba49caf3edb846a9318b9392c8b1 +size 12510 diff --git a/results/downsample/airogs/010/resnet/confusion_matrix.png b/results/downsample/airogs/010/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..9196119c4acc4e2972082a1d4f99d38e6fddda08 --- /dev/null +++ b/results/downsample/airogs/010/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71c83478481a8fd78628a92ce247a10b3466876fc02859620b248b84b09c65d9 +size 72235 diff --git a/results/downsample/airogs/010/resnet/log.csv b/results/downsample/airogs/010/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..bd72aa3b38270ed376a046648d98ab19c50e2cb6 --- /dev/null +++ b/results/downsample/airogs/010/resnet/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6912351335797992,0.5666666666666667,0.6008984910836763,0.4310184512762581,0.00014285714285714284 +1,0.6859664916992188,0.6555555555555556,0.7055829903978053,0.5555665589626807,0.00030952380952380956 +2,0.6730782645089286,0.6907407407407408,0.759170096021948,0.6102057441271217,0.0004761904761904762 +3,0.6465944562639508,0.7,0.7749931412894376,0.6249304531901956,0.0004987576938413504 +4,0.5764168330601284,0.7185185185185186,0.7909739368998628,0.6470474478509569,0.0004941859029405353 +5,0.49954945700509207,0.7722222222222223,0.8328943758573388,0.7157244506842093,0.0004863119165994312 +6,0.40239274501800537,0.7833333333333333,0.8572359396433471,0.7350953109653062,0.00047524221697560476 +7,0.3181565148489816,0.8,0.8648696844993142,0.7543800597126159,0.000461126502766577 +8,0.23927946175847734,0.7851851851851852,0.8809876543209877,0.7448441975673306,0.0004441556647917446 +9,0.20716066019875662,0.7925925925925926,0.8679629629629629,0.7483050785931663,0.0004245592045215182 +10,0.1687560166631426,0.7574074074074074,0.8617901234567901,0.7112149687080738,0.0004026021304636408 +11,0.1326628518956048,0.7981481481481482,0.8913991769547325,0.7619476430583628,0.0003785813743777384 +12,0.1155527297939573,0.8166666666666667,0.8917969821673525,0.78046773039243,0.0003528217757826529 +13,0.10568499671561378,0.8074074074074075,0.8909327846364884,0.7710084893258243,0.0003256716890592065 +14,0.09707468322345189,0.7907407407407407,0.8872222222222222,0.7530942847609515,0.0002974982725547975 +15,0.0671453308314085,0.7944444444444444,0.8880932784636488,0.7571137602340139,0.000268682523396606 +16,0.05340523671891008,0.8037037037037037,0.8977297668038409,0.7692526212888069,0.00023961412515904334 +17,0.06415609642863274,0.8185185185185185,0.9020301783264746,0.7856988282432363,0.0002106861780619037 +18,0.04505604984504836,0.8222222222222222,0.900246913580247,0.788950864338705,0.00018228988296424876 +19,0.05491521315915244,0.8203703703703704,0.9040946502057614,0.7883770642341249,0.00015480925104388764 +20,0.040482928178140094,0.8222222222222222,0.9062688614540466,0.7909752576582912,0.00012861591070496194 +21,0.045258323795029094,0.8074074074074075,0.9027709190672154,0.7749756905964685,0.00010406408194130251 +22,0.036893069211925776,0.8185185185185185,0.9039300411522634,0.7864819188790818,8.148578611867113e-05 +23,0.038750575589282174,0.8185185185185185,0.9041358024691359,0.7865505059847059,6.118635595536626e-05 +24,0.03801685571670532,0.8222222222222222,0.9041083676268862,0.7902453357618812,4.344030642100133e-05 +25,0.025166826056582586,0.8222222222222222,0.9039437585733882,0.7901831460030854,2.8487622392466382e-05 +26,0.05620711350015232,0.812962962962963,0.9058573388203017,0.7815767303076754,1.6530513270159115e-05 +27,0.03389201393084867,0.8148148148148148,0.9059807956104252,0.783474233252284,7.73067844273051e-06 +28,0.029719898211104528,0.8148148148148148,0.9059670781893003,0.7834705075445817,2.2071205802468298e-06 +29,0.043793962470122745,0.8148148148148148,0.9069410150891632,0.7837943064118633,3.453632722355549e-08 diff --git a/results/downsample/airogs/010/resnet/metrics.json b/results/downsample/airogs/010/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..ee34da45735ce67af3cb5bf0daa552357b48d4ee --- /dev/null +++ b/results/downsample/airogs/010/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.819, + "balanced_accuracy": 0.819, + "precision_macro": 0.8203956434227495, + "recall_macro": 0.819, + "f1_macro": 0.8188026761142884, + "precision_weighted": 0.8203956434227496, + "recall_weighted": 0.819, + "f1_weighted": 0.8188026761142885, + "cohen_kappa": 0.638, + "quadratic_weighted_kappa": 0.638, + "mcc": 0.6393941202477768, + "auroc": 0.891534, + "auprc": 0.8783960392505922, + "sensitivity": 0.786, + "specificity": 0.852, + "precision_pos": 0.841541755888651, + "f1_pos": 0.81282316442606, + "per_class": { + "0": { + "precision": 0.799249530956848, + "recall": 0.852, + "f1-score": 0.8247821878025169, + "support": 500.0 + }, + "1": { + "precision": 0.841541755888651, + "recall": 0.786, + "f1-score": 0.81282316442606, + "support": 500.0 + }, + "accuracy": 0.819, + "macro avg": { + "precision": 0.8203956434227495, + "recall": 0.819, + "f1-score": 0.8188026761142884, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8203956434227496, + "recall": 0.819, + "f1-score": 0.8188026761142885, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/010/resnet/pr.png b/results/downsample/airogs/010/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..5e92f97526b930afc446179e8b9ac7beecc6de39 --- /dev/null +++ b/results/downsample/airogs/010/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ebead898902c140ea049671462993a1fa0ddb015137b7f90ac9fd529d75ab41 +size 52805 diff --git a/results/downsample/airogs/010/resnet/roc.png b/results/downsample/airogs/010/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..edb8beaef13cb1f6c683b99d8d4b30dd700175dd --- /dev/null +++ b/results/downsample/airogs/010/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1d7fdf234f9141b91f63726807ea53295122e11ca9ff41ea7508a9be5c36abd +size 64262 diff --git a/results/downsample/airogs/010/resnet/test_pred.npz b/results/downsample/airogs/010/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..bded4d020c6ce0070489edd9934b914d98fd054b --- /dev/null +++ b/results/downsample/airogs/010/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d4a4a604a7f64af05ff40a6407b1a040bb20cf7f90ce1bd0fbdc06fd65c437b3 +size 16510 diff --git a/results/downsample/airogs/010/resnet/train.log b/results/downsample/airogs/010/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1380e5c8ca37067ccd5e25e6edc94fae31567ce2 --- /dev/null +++ b/results/downsample/airogs/010/resnet/train.log @@ -0,0 +1,158 @@ +[resnet] train=500 val=540 test=1000 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6912 val_acc=0.5667 val_auc=0.6009 score=0.4310 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6860 val_acc=0.6556 val_auc=0.7056 score=0.5556 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6731 val_acc=0.6907 val_auc=0.7592 score=0.6102 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6466 val_acc=0.7000 val_auc=0.7750 score=0.6249 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.5764 val_acc=0.7185 val_auc=0.7910 score=0.6470 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.4995 val_acc=0.7722 val_auc=0.8329 score=0.7157 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.4024 val_acc=0.7833 val_auc=0.8572 score=0.7351 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.3182 val_acc=0.8000 val_auc=0.8649 score=0.7544 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.2393 val_acc=0.7852 val_auc=0.8810 score=0.7448 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.2072 val_acc=0.7926 val_auc=0.8680 score=0.7483 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1688 val_acc=0.7574 val_auc=0.8618 score=0.7112 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1327 val_acc=0.7981 val_auc=0.8914 score=0.7619 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.1156 val_acc=0.8167 val_auc=0.8918 score=0.7805 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.1057 val_acc=0.8074 val_auc=0.8909 score=0.7710 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0971 val_acc=0.7907 val_auc=0.8872 score=0.7531 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0671 val_acc=0.7944 val_auc=0.8881 score=0.7571 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0534 val_acc=0.8037 val_auc=0.8977 score=0.7693 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0642 val_acc=0.8185 val_auc=0.9020 score=0.7857 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0451 val_acc=0.8222 val_auc=0.9002 score=0.7890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0549 val_acc=0.8204 val_auc=0.9041 score=0.7884 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0405 val_acc=0.8222 val_auc=0.9063 score=0.7910 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0453 val_acc=0.8074 val_auc=0.9028 score=0.7750 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0369 val_acc=0.8185 val_auc=0.9039 score=0.7865 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0388 val_acc=0.8185 val_auc=0.9041 score=0.7866 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0380 val_acc=0.8222 val_auc=0.9041 score=0.7902 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0252 val_acc=0.8222 val_auc=0.9039 score=0.7902 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0562 val_acc=0.8130 val_auc=0.9059 score=0.7816 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0339 val_acc=0.8148 val_auc=0.9060 score=0.7835 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0297 val_acc=0.8148 val_auc=0.9060 score=0.7835 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0438 val_acc=0.8148 val_auc=0.9069 score=0.7838 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=20 best_val_score=0.7910 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/010/resnet acc=0.8190 auroc=0.891534 f1_macro=0.8188 qwk=0.638 diff --git a/results/downsample/airogs/010/retfound/confusion_matrix.png b/results/downsample/airogs/010/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..dc780c6d4db27d7ae5746f2e2b0286968308ba2e --- /dev/null +++ b/results/downsample/airogs/010/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66cfad3c81bde8c85a68005dccf92bf6974bdc6bdb03620b8d3098cbf7d27480 +size 72443 diff --git a/results/downsample/airogs/010/retfound/confusion_matrix_test.jpg b/results/downsample/airogs/010/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..79c937a632da31c1b263f3e4d5018f4ca2fb3fa0 --- /dev/null +++ b/results/downsample/airogs/010/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:375ae594b5f17caaa4da28079748793ef1ad059acf177ceae40f5a88ab1bc3cc +size 254261 diff --git a/results/downsample/airogs/010/retfound/log.txt b/results/downsample/airogs/010/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..05a6588985b44bfe36a3f0c11622f5e6db055077 --- /dev/null +++ b/results/downsample/airogs/010/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 2.916666666666667e-05, "train_loss": 0.6930694580078125, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.166666666666667e-05, "train_loss": 0.6926727294921875, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015416666666666668, "train_loss": 0.6894887288411459, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021666666666666663, "train_loss": 0.6831095377604167, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.00027916666666666666, "train_loss": 0.6410821278889974, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003416666666666667, "train_loss": 0.6246017456054688, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.0004041666666666666, "train_loss": 0.5560922304789225, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046666666666666666, "train_loss": 0.5844363848368327, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005291666666666667, "train_loss": 0.573507563273112, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005916666666666667, "train_loss": 0.5443623542785645, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006238437351402742, "train_loss": 0.5429067770640056, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.000616445813265531, "train_loss": 0.5066028038660685, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006015760483892742, "train_loss": 0.5185762882232666, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005796005834726778, "train_loss": 0.53712158203125, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005510605273943752, "train_loss": 0.5367470701535543, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005166586310540296, "train_loss": 0.5101282993952433, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00047724198331288874, "train_loss": 0.5326646963755289, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004337811528546931, "train_loss": 0.518016004562378, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003873462895635502, "train_loss": 0.5022597948710124, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00033908077388216404, "train_loss": 0.46001251935958865, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002901730629906729, "train_loss": 0.512948191165924, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024182742704661195, "train_loss": 0.4756464004516602, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019523429605689067, "train_loss": 0.45968577861785886, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015154094754021218, "train_loss": 0.44586576223373414, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011182325674696714, "train_loss": 0.43884477615356443, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.705920503994361e-05, "train_loss": 0.46032923460006714, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.8104798747441377e-05, "train_loss": 0.4451125979423523, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5672991446547074e-05, "train_loss": 0.4721834739049276, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.0316128679908865e-05, "train_loss": 0.4452788631121318, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.4123473753469123e-06, "train_loss": 0.4721410592397054, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/010/retfound/metrics.json b/results/downsample/airogs/010/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..06a23144bd52dd15fb430b38f300f2dbfc2185be --- /dev/null +++ b/results/downsample/airogs/010/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.851, + "balanced_accuracy": 0.851, + "precision_macro": 0.8511699662636716, + "recall_macro": 0.851, + "f1_macro": 0.850981968818227, + "precision_weighted": 0.8511699662636717, + "recall_weighted": 0.851, + "f1_weighted": 0.850981968818227, + "cohen_kappa": 0.702, + "quadratic_weighted_kappa": 0.702, + "mcc": 0.7021699456927752, + "auroc": 0.9329420000000002, + "auprc": 0.9378266633292438, + "sensitivity": 0.84, + "specificity": 0.862, + "precision_pos": 0.8588957055214724, + "f1_pos": 0.8493427704752275, + "per_class": { + "0": { + "precision": 0.8434442270058709, + "recall": 0.862, + "f1-score": 0.8526211671612265, + "support": 500.0 + }, + "1": { + "precision": 0.8588957055214724, + "recall": 0.84, + "f1-score": 0.8493427704752275, + "support": 500.0 + }, + "accuracy": 0.851, + "macro avg": { + "precision": 0.8511699662636716, + "recall": 0.851, + "f1-score": 0.850981968818227, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8511699662636717, + "recall": 0.851, + "f1-score": 0.850981968818227, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/010/retfound/metrics_test.csv b/results/downsample/airogs/010/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..43357d83076ac3a82eedb1675cee33cf1dfda0f6 --- /dev/null +++ b/results/downsample/airogs/010/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.3509616069495678,0.851,0.850981968818227,0.9329370000000001,0.149,0.7406202654384583,0.8511699662636716,0.851,0.9354072847765578,0.702 diff --git a/results/downsample/airogs/010/retfound/metrics_val.csv b/results/downsample/airogs/010/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..6ccd92673e70d643a675d20c9b89849b079686c5 --- /dev/null +++ b/results/downsample/airogs/010/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6911632629001841,0.6259259259259259,0.5981373141366657,0.7359979423868312,0.37407407407407406,0.43482654600301657,0.6740746359223301,0.6259259259259259,0.7150930878518238,0.2518518518518519 +0.685170920456157,0.5351851851851852,0.42195249894446835,0.758395061728395,0.4648148148148148,0.30160053468411424,0.6625578300272514,0.5351851851851852,0.7498420694218307,0.07037037037037042 +0.6666971410022062,0.6777777777777778,0.6632933891405309,0.7883573388203018,0.32222222222222224,0.5003097021228673,0.7147259593081052,0.6777777777777778,0.7838012587674041,0.3555555555555555 +0.6354802881970125,0.7203703703703703,0.711046773284761,0.8127674897119341,0.2796296296296296,0.5541668803090123,0.7530279882188027,0.7203703703703703,0.80979597590046,0.44074074074074077 +0.5088641871424282,0.8,0.7998654788540995,0.8703703703703705,0.2,0.6665110028877219,0.8008087588028169,0.8,0.8642583733977296,0.6 +0.4414311892846051,0.8185185185185185,0.8177962016774318,0.9028532235939644,0.1814814814814815,0.6919150110375276,0.8236507582515611,0.8185185185185185,0.8998183843102401,0.6370370370370371 +0.4157573475557215,0.8,0.7975,0.9083950617283951,0.2,0.663784157730889,0.8155844155844156,0.8,0.9096440825955378,0.6 +0.4472531325676862,0.7703703703703704,0.763505877034358,0.9149691358024692,0.22962962962962963,0.6191956191956192,0.8058849090570488,0.7703703703703704,0.9140467251375909,0.5407407407407407 +0.35088413077242236,0.8629629629629629,0.8629328796443664,0.9246467764060358,0.13703703703703704,0.7589167276856303,0.8632818935691142,0.8629629629629629,0.9242620878438041,0.7259259259259259 +0.35796982137595906,0.8333333333333334,0.8325892857142857,0.927746913580247,0.16666666666666666,0.7133497133497133,0.839366515837104,0.8333333333333334,0.9260300218604687,0.6666666666666667 +0.33727068848469677,0.8592592592592593,0.8591355922895272,0.9322565157750343,0.14074074074074075,0.7530801687763713,0.8605253014701834,0.8592592592592592,0.9319480032417642,0.7185185185185186 +0.34569788176347227,0.8722222222222222,0.8722007470391115,0.9340946502057612,0.12777777777777777,0.7733690727876774,0.8724725810215371,0.8722222222222222,0.9344818497593984,0.7444444444444445 +0.34715402652235594,0.8444444444444444,0.84367030149299,0.9385116598079561,0.15555555555555556,0.7297670725797305,0.8514050604567847,0.8444444444444444,0.9388278105872173,0.6888888888888889 +0.3745741515475161,0.8592592592592593,0.8586309523809523,0.9288134430727023,0.14074074074074075,0.7524011000634652,0.8657616892911011,0.8592592592592593,0.9286026645252983,0.7185185185185186 +0.35650092100395875,0.85,0.8499130154307537,0.9300308641975309,0.15,0.7390160634235431,0.8508132708198706,0.8500000000000001,0.9299132660741336,0.7 +0.34871040284633636,0.8629629629629629,0.8629460427213236,0.9296879286694102,0.13703703703703704,0.7589346349745331,0.8631422924901186,0.8629629629629629,0.9298129343603234,0.7259259259259259 +0.34657756195348854,0.8537037037037037,0.8535223647931767,0.9320507544581618,0.14629629629629629,0.7445088819226751,0.8554639573195109,0.8537037037037036,0.9327571097145066,0.7074074074074075 +0.3285454340717372,0.8648148148148148,0.8647364213141361,0.9365192043895747,0.13518518518518519,0.7617199148029818,0.8656625097963729,0.8648148148148148,0.9377117515330697,0.7296296296296296 +0.34580644943258343,0.8666666666666667,0.8662538699690403,0.9416838134430727,0.13333333333333333,0.764139127083534,0.87125,0.8666666666666667,0.9431902055926968,0.7333333333333334 +0.3736278914353427,0.8611111111111112,0.8611068242847002,0.9422016460905349,0.1388888888888889,0.756091757091186,0.8611556982343499,0.8611111111111112,0.9429857539479114,0.7222222222222222 +0.3224005668478854,0.8592592592592593,0.8592573286327658,0.9431412894375857,0.14074074074074075,0.7532441520930184,0.8592789727831431,0.8592592592592592,0.9443783310201082,0.7185185185185186 +0.32057176705668955,0.8611111111111112,0.8611106348101332,0.9429458161865569,0.1388888888888889,0.7560969161132027,0.8611160646922454,0.8611111111111112,0.9442113364773395,0.7222222222222222 +0.34548356121077256,0.8629629629629629,0.8629554433713784,0.9429080932784637,0.13703703703703704,0.7589474242665393,0.8630426431041107,0.8629629629629629,0.9441175023274324,0.7259259259259259 +0.3623477816581726,0.8629629629629629,0.8629554433713784,0.9434122085048011,0.13703703703703704,0.7589474242665393,0.8630426431041107,0.8629629629629629,0.9443846657391708,0.7259259259259259 +0.3589090209673433,0.8592592592592593,0.8592573286327658,0.9438854595336077,0.14074074074074075,0.7532441520930184,0.8592789727831431,0.8592592592592592,0.9448711459024962,0.7185185185185186 +0.3483984185492291,0.8574074074074074,0.8574030062656255,0.9437448559670782,0.1425925925925926,0.7503992854891247,0.8574515372268181,0.8574074074074074,0.9446179158764834,0.7148148148148148 +0.34846346671966943,0.8611111111111112,0.8610877685207734,0.9442043895747599,0.1388888888888889,0.7560659587070571,0.8613539965134316,0.8611111111111112,0.9452021702522139,0.7222222222222222 +0.3486983846215641,0.8574074074074074,0.8573482137649711,0.9442798353909465,0.1425925925925926,0.7503258287612201,0.8580016213468171,0.8574074074074074,0.9449713671296528,0.7148148148148148 +0.3487990077804117,0.8574074074074074,0.8573482137649711,0.9442626886145404,0.1425925925925926,0.7503258287612201,0.8580016213468171,0.8574074074074074,0.9449564359929981,0.7148148148148148 +0.34871361141695695,0.8592592592592593,0.8592109777015438,0.9441735253772291,0.14074074074074075,0.7531817079471526,0.8597527472527473,0.8592592592592593,0.9448724258973051,0.7185185185185186 diff --git a/results/downsample/airogs/010/retfound/pr.png b/results/downsample/airogs/010/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..6d61b9727ef9383d190c071eb1b03631d6d9eb60 --- /dev/null +++ b/results/downsample/airogs/010/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94b2999f67f5aa6278d888d308db502b6937672e2ff7792ac48aafc78cd255b9 +size 48221 diff --git a/results/downsample/airogs/010/retfound/roc.png b/results/downsample/airogs/010/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..eb4879a9b9817005c29692b722ba91cf8b319456 --- /dev/null +++ b/results/downsample/airogs/010/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:36c64e15ea79afec6721e54c35cd02364efe37e46094ce229297012253bbcc2d +size 63097 diff --git a/results/downsample/airogs/010/retfound/test_pred.npz b/results/downsample/airogs/010/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..0a05717dcf871392ef98a576ef9fde2b9ce5115a --- /dev/null +++ b/results/downsample/airogs/010/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c014e05e1558d54e3dd70a78d4956e247b6d0e574ff29845c3ca28d11fe1900c +size 12510 diff --git a/results/downsample/airogs/010/retfound/train.log b/results/downsample/airogs/010/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..919f83632399f2446f83f3eed70029fe3cf34dbe --- /dev/null +++ b/results/downsample/airogs/010/retfound/train.log @@ -0,0 +1,506 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:50:56.033191062 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:50:56.516586] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:50:56.516842] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/airogs_10', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/010', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:50:59.464253] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:51:01.010472] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:51:04.347644] Sampler_train = +[14:51:04.414548] len of train_set: 480 +[14:51:04.627637] [Adaptation] Full fine-tuning: training all parameters. +[14:51:04.628679] number of trainable params (M): 303.30 +[14:51:04.628781] base lr: 5.00e-03 +[14:51:04.628866] actual lr: 6.25e-04 +[14:51:04.628929] accumulate grad iterations: 1 +[14:51:04.628991] effective batch size: 32 +[14:51:04.631927] criterion = CrossEntropyLoss() +[14:51:04.632020] Start training for 30 epochs +[14:51:04.634168] log_dir: ./output_logs/retfound +[14:51:06.323753] Epoch: [0] [ 0/15] eta: 0:00:25 lr: 0.000000 loss: 0.6928 (0.6928) time: 1.6888 data: 1.1019 max mem: 7340 +[14:51:07.290978] Epoch: [0] [14/15] eta: 0:00:00 lr: 0.000058 loss: 0.6931 (0.6931) time: 0.1770 data: 0.0735 max mem: 9671 +[14:51:07.384895] Epoch: [0] Total time: 0:00:02 (0.1834 s / it) +[14:51:07.385987] Averaged stats: lr: 0.000058 loss: 0.6931 (0.6931) +[14:51:08.109319] val: [ 0/17] eta: 0:00:11 loss: 0.6951 (0.6951) time: 0.6756 data: 0.6485 max mem: 9671 +[14:51:08.639062] val: [10/17] eta: 0:00:00 loss: 0.6951 (0.6937) time: 0.1095 data: 0.0930 max mem: 9671 +[14:51:08.763943] val: [16/17] eta: 0:00:00 loss: 0.6909 (0.6912) time: 0.0782 data: 0.0602 max mem: 9671 +[14:51:08.841813] val: Total time: 0:00:01 (0.0829 s / it) +[14:51:08.863075] val loss: 0.6911632629001841 +[14:51:08.863273] Accuracy: 0.6259, F1 Score: 0.5981, ROC AUC: 0.7360, Hamming Loss: 0.3741, + Jaccard Score: 0.4348, Precision: 0.6741, Recall: 0.6259, + Average Precision: 0.7151, Kappa: 0.2519, Score: 0.5287 +[14:51:10.715449] Best epoch = 0, Best score = 0.5287 +[14:51:10.801887] log_dir: ./output_logs/retfound +[14:51:11.728566] Epoch: [1] [ 0/15] eta: 0:00:13 lr: 0.000063 loss: 0.6879 (0.6879) time: 0.9256 data: 0.8295 max mem: 9671 +[14:51:12.642278] Epoch: [1] [14/15] eta: 0:00:00 lr: 0.000121 loss: 0.6939 (0.6927) time: 0.1226 data: 0.0553 max mem: 9671 +[14:51:12.716012] Epoch: [1] Total time: 0:00:01 (0.1276 s / it) +[14:51:12.716921] Averaged stats: lr: 0.000121 loss: 0.6939 (0.6927) +[14:51:13.969570] val: [ 0/17] eta: 0:00:21 loss: 0.6505 (0.6505) time: 1.2412 data: 1.2240 max mem: 9671 +[14:51:14.654213] val: [10/17] eta: 0:00:01 loss: 0.6535 (0.6679) time: 0.1750 data: 0.1594 max mem: 9671 +[14:51:14.763437] val: [16/17] eta: 0:00:00 loss: 0.6915 (0.6852) time: 0.1196 data: 0.1042 max mem: 9671 +[14:51:14.838703] val: Total time: 0:00:02 (0.1242 s / it) +[14:51:14.858008] val loss: 0.685170920456157 +[14:51:14.858192] Accuracy: 0.5352, F1 Score: 0.4220, ROC AUC: 0.7584, Hamming Loss: 0.4648, + Jaccard Score: 0.3016, Precision: 0.6626, Recall: 0.5352, + Average Precision: 0.7498, Kappa: 0.0704, Score: 0.4169 +[14:51:14.923817] Best epoch = 0, Best score = 0.5287 +[14:51:15.201477] log_dir: ./output_logs/retfound +[14:51:16.401162] Epoch: [2] [ 0/15] eta: 0:00:17 lr: 0.000125 loss: 0.6877 (0.6877) time: 1.1988 data: 1.1246 max mem: 9671 +[14:51:17.316835] Epoch: [2] [14/15] eta: 0:00:00 lr: 0.000183 loss: 0.6880 (0.6895) time: 0.1409 data: 0.0750 max mem: 9671 +[14:51:17.390711] Epoch: [2] Total time: 0:00:02 (0.1459 s / it) +[14:51:17.391504] Averaged stats: lr: 0.000183 loss: 0.6880 (0.6895) +[14:51:18.432657] val: [ 0/17] eta: 0:00:17 loss: 0.6231 (0.6231) time: 1.0226 data: 1.0059 max mem: 9671 +[14:51:19.323430] val: [10/17] eta: 0:00:01 loss: 0.6354 (0.6488) time: 0.1739 data: 0.1584 max mem: 9671 +[14:51:19.429150] val: [16/17] eta: 0:00:00 loss: 0.6755 (0.6667) time: 0.1187 data: 0.1034 max mem: 9671 +[14:51:19.506402] val: Total time: 0:00:02 (0.1234 s / it) +[14:51:19.520254] val loss: 0.6666971410022062 +[14:51:19.520441] Accuracy: 0.6778, F1 Score: 0.6633, ROC AUC: 0.7884, Hamming Loss: 0.3222, + Jaccard Score: 0.5003, Precision: 0.7147, Recall: 0.6778, + Average Precision: 0.7838, Kappa: 0.3556, Score: 0.6024 +[14:51:21.441289] Best epoch = 2, Best score = 0.6024 +[14:51:21.579266] log_dir: ./output_logs/retfound +[14:51:22.844711] Epoch: [3] [ 0/15] eta: 0:00:18 lr: 0.000188 loss: 0.6774 (0.6774) time: 1.2646 data: 1.1908 max mem: 9671 +[14:51:23.756148] Epoch: [3] [14/15] eta: 0:00:00 lr: 0.000246 loss: 0.6872 (0.6831) time: 0.1450 data: 0.0794 max mem: 9671 +[14:51:23.834221] Epoch: [3] Total time: 0:00:02 (0.1503 s / it) +[14:51:23.835072] Averaged stats: lr: 0.000246 loss: 0.6872 (0.6831) +[14:51:24.515970] val: [ 0/17] eta: 0:00:11 loss: 0.5528 (0.5528) time: 0.6577 data: 0.6405 max mem: 9671 +[14:51:24.809762] val: [10/17] eta: 0:00:00 loss: 0.5766 (0.6030) time: 0.0864 data: 0.0709 max mem: 9671 +[14:51:24.902302] val: [16/17] eta: 0:00:00 loss: 0.6581 (0.6355) time: 0.0613 data: 0.0459 max mem: 9671 +[14:51:24.976829] val: Total time: 0:00:01 (0.0658 s / it) +[14:51:24.991782] val loss: 0.6354802881970125 +[14:51:24.992035] Accuracy: 0.7204, F1 Score: 0.7110, ROC AUC: 0.8128, Hamming Loss: 0.2796, + Jaccard Score: 0.5542, Precision: 0.7530, Recall: 0.7204, + Average Precision: 0.8098, Kappa: 0.4407, Score: 0.6549 +[14:51:27.221874] Best epoch = 3, Best score = 0.6549 +[14:51:27.421494] log_dir: ./output_logs/retfound +[14:51:28.292177] Epoch: [4] [ 0/15] eta: 0:00:13 lr: 0.000250 loss: 0.6338 (0.6338) time: 0.8694 data: 0.7986 max mem: 9671 +[14:51:29.206518] Epoch: [4] [14/15] eta: 0:00:00 lr: 0.000308 loss: 0.6369 (0.6411) time: 0.1189 data: 0.0533 max mem: 9671 +[14:51:29.281453] Epoch: [4] Total time: 0:00:01 (0.1240 s / it) +[14:51:29.282230] Averaged stats: lr: 0.000308 loss: 0.6369 (0.6411) +[14:51:30.573131] val: [ 0/17] eta: 0:00:20 loss: 0.4913 (0.4913) time: 1.2323 data: 1.2157 max mem: 9671 +[14:51:31.198249] val: [10/17] eta: 0:00:01 loss: 0.5369 (0.5401) time: 0.1688 data: 0.1532 max mem: 9671 +[14:51:31.290963] val: [16/17] eta: 0:00:00 loss: 0.4913 (0.5089) time: 0.1146 data: 0.0992 max mem: 9671 +[14:51:31.371333] val: Total time: 0:00:02 (0.1195 s / it) +[14:51:31.385373] val loss: 0.5088641871424282 +[14:51:31.385557] Accuracy: 0.8000, F1 Score: 0.7999, ROC AUC: 0.8704, Hamming Loss: 0.2000, + Jaccard Score: 0.6665, Precision: 0.8008, Recall: 0.8000, + Average Precision: 0.8643, Kappa: 0.6000, Score: 0.7567 +[14:51:33.197626] Best epoch = 4, Best score = 0.7567 +[14:51:33.280723] log_dir: ./output_logs/retfound +[14:51:33.974671] Epoch: [5] [ 0/15] eta: 0:00:10 lr: 0.000313 loss: 0.6411 (0.6411) time: 0.6931 data: 0.6223 max mem: 9671 +[14:51:34.887001] Epoch: [5] [14/15] eta: 0:00:00 lr: 0.000371 loss: 0.6249 (0.6246) time: 0.1070 data: 0.0415 max mem: 9671 +[14:51:34.964389] Epoch: [5] Total time: 0:00:01 (0.1122 s / it) +[14:51:34.965143] Averaged stats: lr: 0.000371 loss: 0.6249 (0.6246) +[14:51:35.818475] val: [ 0/17] eta: 0:00:13 loss: 0.3436 (0.3436) time: 0.7882 data: 0.7716 max mem: 9671 +[14:51:36.461407] val: [10/17] eta: 0:00:00 loss: 0.3914 (0.4284) time: 0.1300 data: 0.1146 max mem: 9671 +[14:51:36.553445] val: [16/17] eta: 0:00:00 loss: 0.4300 (0.4414) time: 0.0895 data: 0.0742 max mem: 9671 +[14:51:36.630223] val: Total time: 0:00:01 (0.0942 s / it) +[14:51:36.644059] val loss: 0.4414311892846051 +[14:51:36.644251] Accuracy: 0.8185, F1 Score: 0.8178, ROC AUC: 0.9029, Hamming Loss: 0.1815, + Jaccard Score: 0.6919, Precision: 0.8237, Recall: 0.8185, + Average Precision: 0.8998, Kappa: 0.6370, Score: 0.7859 +[14:51:38.468677] Best epoch = 5, Best score = 0.7859 +[14:51:38.608569] log_dir: ./output_logs/retfound +[14:51:39.422952] Epoch: [6] [ 0/15] eta: 0:00:12 lr: 0.000375 loss: 0.5262 (0.5262) time: 0.8136 data: 0.7441 max mem: 9671 +[14:51:40.343680] Epoch: [6] [14/15] eta: 0:00:00 lr: 0.000433 loss: 0.5699 (0.5561) time: 0.1156 data: 0.0497 max mem: 9671 +[14:51:40.427038] Epoch: [6] Total time: 0:00:01 (0.1212 s / it) +[14:51:40.427780] Averaged stats: lr: 0.000433 loss: 0.5699 (0.5561) +[14:51:41.122442] val: [ 0/17] eta: 0:00:11 loss: 0.2094 (0.2094) time: 0.6824 data: 0.6651 max mem: 9671 +[14:51:41.379258] val: [10/17] eta: 0:00:00 loss: 0.2682 (0.3501) time: 0.0853 data: 0.0697 max mem: 9671 +[14:51:41.472241] val: [16/17] eta: 0:00:00 loss: 0.3520 (0.4158) time: 0.0606 data: 0.0452 max mem: 9671 +[14:51:41.547782] val: Total time: 0:00:01 (0.0652 s / it) +[14:51:41.561699] val loss: 0.4157573475557215 +[14:51:41.561894] Accuracy: 0.8000, F1 Score: 0.7975, ROC AUC: 0.9084, Hamming Loss: 0.2000, + Jaccard Score: 0.6638, Precision: 0.8156, Recall: 0.8000, + Average Precision: 0.9096, Kappa: 0.6000, Score: 0.7686 +[14:51:41.595150] Best epoch = 5, Best score = 0.7859 +[14:51:41.866064] log_dir: ./output_logs/retfound +[14:51:42.605923] Epoch: [7] [ 0/15] eta: 0:00:11 lr: 0.000438 loss: 0.5579 (0.5579) time: 0.7391 data: 0.6708 max mem: 9671 +[14:51:43.519909] Epoch: [7] [14/15] eta: 0:00:00 lr: 0.000496 loss: 0.5763 (0.5844) time: 0.1102 data: 0.0448 max mem: 9671 +[14:51:43.595769] Epoch: [7] Total time: 0:00:01 (0.1153 s / it) +[14:51:43.596507] Averaged stats: lr: 0.000496 loss: 0.5763 (0.5844) +[14:51:44.118903] val: [ 0/17] eta: 0:00:08 loss: 0.1969 (0.1969) time: 0.5148 data: 0.4977 max mem: 9671 +[14:51:44.381408] val: [10/17] eta: 0:00:00 loss: 0.2461 (0.3439) time: 0.0706 data: 0.0550 max mem: 9671 +[14:51:44.473164] val: [16/17] eta: 0:00:00 loss: 0.5229 (0.4473) time: 0.0511 data: 0.0356 max mem: 9671 +[14:51:44.546242] val: Total time: 0:00:00 (0.0555 s / it) +[14:51:44.560011] val loss: 0.4472531325676862 +[14:51:44.560179] Accuracy: 0.7704, F1 Score: 0.7635, ROC AUC: 0.9150, Hamming Loss: 0.2296, + Jaccard Score: 0.6192, Precision: 0.8059, Recall: 0.7704, + Average Precision: 0.9140, Kappa: 0.5407, Score: 0.7397 +[14:51:44.602561] Best epoch = 5, Best score = 0.7859 +[14:51:44.865956] log_dir: ./output_logs/retfound +[14:51:45.540506] Epoch: [8] [ 0/15] eta: 0:00:10 lr: 0.000500 loss: 0.6166 (0.6166) time: 0.6738 data: 0.6040 max mem: 9671 +[14:51:46.451686] Epoch: [8] [14/15] eta: 0:00:00 lr: 0.000558 loss: 0.5341 (0.5735) time: 0.1056 data: 0.0403 max mem: 9671 +[14:51:46.527352] Epoch: [8] Total time: 0:00:01 (0.1107 s / it) +[14:51:46.528078] Averaged stats: lr: 0.000558 loss: 0.5341 (0.5735) +[14:51:47.054997] val: [ 0/17] eta: 0:00:08 loss: 0.3690 (0.3690) time: 0.5195 data: 0.5024 max mem: 9671 +[14:51:47.218551] val: [10/17] eta: 0:00:00 loss: 0.4045 (0.3899) time: 0.0620 data: 0.0465 max mem: 9671 +[14:51:47.310357] val: [16/17] eta: 0:00:00 loss: 0.3538 (0.3509) time: 0.0455 data: 0.0301 max mem: 9671 +[14:51:47.392041] val: Total time: 0:00:00 (0.0504 s / it) +[14:51:47.406204] val loss: 0.35088413077242236 +[14:51:47.406437] Accuracy: 0.8630, F1 Score: 0.8629, ROC AUC: 0.9246, Hamming Loss: 0.1370, + Jaccard Score: 0.7589, Precision: 0.8633, Recall: 0.8630, + Average Precision: 0.9243, Kappa: 0.7259, Score: 0.8378 +[14:51:49.158619] Best epoch = 8, Best score = 0.8378 +[14:51:49.256113] log_dir: ./output_logs/retfound +[14:51:50.072661] Epoch: [9] [ 0/15] eta: 0:00:12 lr: 0.000562 loss: 0.3939 (0.3939) time: 0.8157 data: 0.7453 max mem: 9671 +[14:51:50.988484] Epoch: [9] [14/15] eta: 0:00:00 lr: 0.000621 loss: 0.4960 (0.5444) time: 0.1154 data: 0.0497 max mem: 9671 +[14:51:51.062707] Epoch: [9] Total time: 0:00:01 (0.1204 s / it) +[14:51:51.063533] Averaged stats: lr: 0.000621 loss: 0.4960 (0.5444) +[14:51:51.587932] val: [ 0/17] eta: 0:00:08 loss: 0.2581 (0.2581) time: 0.5168 data: 0.4994 max mem: 9671 +[14:51:51.810713] val: [10/17] eta: 0:00:00 loss: 0.3545 (0.3452) time: 0.0672 data: 0.0511 max mem: 9671 +[14:51:51.902753] val: [16/17] eta: 0:00:00 loss: 0.3545 (0.3580) time: 0.0488 data: 0.0331 max mem: 9671 +[14:51:51.975949] val: Total time: 0:00:00 (0.0533 s / it) +[14:51:51.990349] val loss: 0.35796982137595906 +[14:51:51.990610] Accuracy: 0.8333, F1 Score: 0.8326, ROC AUC: 0.9277, Hamming Loss: 0.1667, + Jaccard Score: 0.7133, Precision: 0.8394, Recall: 0.8333, + Average Precision: 0.9260, Kappa: 0.6667, Score: 0.8090 +[14:51:52.017702] Best epoch = 8, Best score = 0.8378 +[14:51:52.298669] log_dir: ./output_logs/retfound +[14:51:52.996772] Epoch: [10] [ 0/15] eta: 0:00:10 lr: 0.000625 loss: 0.5211 (0.5211) time: 0.6972 data: 0.6277 max mem: 9671 +[14:51:53.911405] Epoch: [10] [14/15] eta: 0:00:00 lr: 0.000622 loss: 0.5204 (0.5429) time: 0.1074 data: 0.0419 max mem: 9671 +[14:51:53.989498] Epoch: [10] Total time: 0:00:01 (0.1127 s / it) +[14:51:53.990219] Averaged stats: lr: 0.000622 loss: 0.5204 (0.5429) +[14:51:54.707975] val: [ 0/17] eta: 0:00:11 loss: 0.2500 (0.2500) time: 0.6993 data: 0.6828 max mem: 9671 +[14:51:54.861817] val: [10/17] eta: 0:00:00 loss: 0.3414 (0.3365) time: 0.0775 data: 0.0621 max mem: 9671 +[14:51:54.953311] val: [16/17] eta: 0:00:00 loss: 0.3414 (0.3373) time: 0.0555 data: 0.0402 max mem: 9671 +[14:51:55.034016] val: Total time: 0:00:01 (0.0603 s / it) +[14:51:55.047851] val loss: 0.33727068848469677 +[14:51:55.048045] Accuracy: 0.8593, F1 Score: 0.8591, ROC AUC: 0.9323, Hamming Loss: 0.1407, + Jaccard Score: 0.7531, Precision: 0.8605, Recall: 0.8593, + Average Precision: 0.9319, Kappa: 0.7185, Score: 0.8366 +[14:51:55.096367] Best epoch = 8, Best score = 0.8378 +[14:51:55.330789] log_dir: ./output_logs/retfound +[14:51:56.061486] Epoch: [11] [ 0/15] eta: 0:00:10 lr: 0.000621 loss: 0.4216 (0.4216) time: 0.7295 data: 0.6566 max mem: 9671 +[14:51:56.991565] Epoch: [11] [14/15] eta: 0:00:00 lr: 0.000611 loss: 0.5098 (0.5066) time: 0.1106 data: 0.0440 max mem: 9671 +[14:51:57.067723] Epoch: [11] Total time: 0:00:01 (0.1158 s / it) +[14:51:57.068502] Averaged stats: lr: 0.000611 loss: 0.5098 (0.5066) +[14:51:57.583396] val: [ 0/17] eta: 0:00:08 loss: 0.3611 (0.3611) time: 0.5076 data: 0.4881 max mem: 9671 +[14:51:57.738968] val: [10/17] eta: 0:00:00 loss: 0.4099 (0.4203) time: 0.0602 data: 0.0445 max mem: 9671 +[14:51:57.830949] val: [16/17] eta: 0:00:00 loss: 0.3611 (0.3457) time: 0.0444 data: 0.0288 max mem: 9671 +[14:51:57.907844] val: Total time: 0:00:00 (0.0490 s / it) +[14:51:57.921723] val loss: 0.34569788176347227 +[14:51:57.921914] Accuracy: 0.8722, F1 Score: 0.8722, ROC AUC: 0.9341, Hamming Loss: 0.1278, + Jaccard Score: 0.7734, Precision: 0.8725, Recall: 0.8722, + Average Precision: 0.9345, Kappa: 0.7444, Score: 0.8502 +[14:51:59.692075] Best epoch = 11, Best score = 0.8502 +[14:51:59.752184] log_dir: ./output_logs/retfound +[14:52:00.414135] Epoch: [12] [ 0/15] eta: 0:00:09 lr: 0.000610 loss: 0.6368 (0.6368) time: 0.6611 data: 0.5835 max mem: 9671 +[14:52:01.331969] Epoch: [12] [14/15] eta: 0:00:00 lr: 0.000592 loss: 0.5032 (0.5186) time: 0.1052 data: 0.0390 max mem: 9671 +[14:52:01.413547] Epoch: [12] Total time: 0:00:01 (0.1107 s / it) +[14:52:01.414304] Averaged stats: lr: 0.000592 loss: 0.5032 (0.5186) +[14:52:01.900529] val: [ 0/17] eta: 0:00:07 loss: 0.2625 (0.2625) time: 0.4670 data: 0.4496 max mem: 9671 +[14:52:02.172935] val: [10/17] eta: 0:00:00 loss: 0.3075 (0.3485) time: 0.0672 data: 0.0515 max mem: 9671 +[14:52:02.265005] val: [16/17] eta: 0:00:00 loss: 0.3530 (0.3472) time: 0.0488 data: 0.0334 max mem: 9671 +[14:52:02.345077] val: Total time: 0:00:00 (0.0537 s / it) +[14:52:02.358934] val loss: 0.34715402652235594 +[14:52:02.359156] Accuracy: 0.8444, F1 Score: 0.8437, ROC AUC: 0.9385, Hamming Loss: 0.1556, + Jaccard Score: 0.7298, Precision: 0.8514, Recall: 0.8444, + Average Precision: 0.9388, Kappa: 0.6889, Score: 0.8237 +[14:52:02.392181] Best epoch = 11, Best score = 0.8502 +[14:52:02.680961] log_dir: ./output_logs/retfound +[14:52:03.384234] Epoch: [13] [ 0/15] eta: 0:00:10 lr: 0.000591 loss: 0.5069 (0.5069) time: 0.7024 data: 0.6301 max mem: 9671 +[14:52:04.301188] Epoch: [13] [14/15] eta: 0:00:00 lr: 0.000567 loss: 0.5505 (0.5371) time: 0.1079 data: 0.0421 max mem: 9671 +[14:52:04.374750] Epoch: [13] Total time: 0:00:01 (0.1129 s / it) +[14:52:04.375516] Averaged stats: lr: 0.000567 loss: 0.5505 (0.5371) +[14:52:04.925607] val: [ 0/17] eta: 0:00:09 loss: 0.4529 (0.4529) time: 0.5386 data: 0.5211 max mem: 9671 +[14:52:05.123888] val: [10/17] eta: 0:00:00 loss: 0.4872 (0.4877) time: 0.0669 data: 0.0512 max mem: 9671 +[14:52:05.216936] val: [16/17] eta: 0:00:00 loss: 0.4161 (0.3746) time: 0.0487 data: 0.0332 max mem: 9671 +[14:52:05.297887] val: Total time: 0:00:00 (0.0536 s / it) +[14:52:05.311749] val loss: 0.3745741515475161 +[14:52:05.312277] Accuracy: 0.8593, F1 Score: 0.8586, ROC AUC: 0.9288, Hamming Loss: 0.1407, + Jaccard Score: 0.7524, Precision: 0.8658, Recall: 0.8593, + Average Precision: 0.9286, Kappa: 0.7185, Score: 0.8353 +[14:52:05.350686] Best epoch = 11, Best score = 0.8502 +[14:52:05.606769] log_dir: ./output_logs/retfound +[14:52:06.287715] Epoch: [14] [ 0/15] eta: 0:00:10 lr: 0.000565 loss: 0.6727 (0.6727) time: 0.6800 data: 0.6097 max mem: 9671 +[14:52:07.207409] Epoch: [14] [14/15] eta: 0:00:00 lr: 0.000536 loss: 0.5125 (0.5367) time: 0.1066 data: 0.0408 max mem: 9671 +[14:52:07.287914] Epoch: [14] Total time: 0:00:01 (0.1121 s / it) +[14:52:07.288709] Averaged stats: lr: 0.000536 loss: 0.5125 (0.5367) +[14:52:07.836606] val: [ 0/17] eta: 0:00:09 loss: 0.2374 (0.2374) time: 0.5408 data: 0.5238 max mem: 9671 +[14:52:08.074050] val: [10/17] eta: 0:00:00 loss: 0.3191 (0.3709) time: 0.0707 data: 0.0552 max mem: 9671 +[14:52:08.178211] val: [16/17] eta: 0:00:00 loss: 0.3130 (0.3565) time: 0.0518 data: 0.0364 max mem: 9671 +[14:52:08.257373] val: Total time: 0:00:00 (0.0566 s / it) +[14:52:08.271436] val loss: 0.35650092100395875 +[14:52:08.271634] Accuracy: 0.8500, F1 Score: 0.8499, ROC AUC: 0.9300, Hamming Loss: 0.1500, + Jaccard Score: 0.7390, Precision: 0.8508, Recall: 0.8500, + Average Precision: 0.9299, Kappa: 0.7000, Score: 0.8266 +[14:52:08.312909] Best epoch = 11, Best score = 0.8502 +[14:52:08.557719] log_dir: ./output_logs/retfound +[14:52:09.260990] Epoch: [15] [ 0/15] eta: 0:00:10 lr: 0.000534 loss: 0.5130 (0.5130) time: 0.7024 data: 0.6320 max mem: 9671 +[14:52:10.175833] Epoch: [15] [14/15] eta: 0:00:00 lr: 0.000499 loss: 0.5362 (0.5101) time: 0.1078 data: 0.0422 max mem: 9671 +[14:52:10.262080] Epoch: [15] Total time: 0:00:01 (0.1136 s / it) +[14:52:10.262881] Averaged stats: lr: 0.000499 loss: 0.5362 (0.5101) +[14:52:10.775018] val: [ 0/17] eta: 0:00:08 loss: 0.2764 (0.2764) time: 0.5007 data: 0.4834 max mem: 9671 +[14:52:10.946751] val: [10/17] eta: 0:00:00 loss: 0.3700 (0.3982) time: 0.0611 data: 0.0455 max mem: 9671 +[14:52:11.038525] val: [16/17] eta: 0:00:00 loss: 0.3463 (0.3487) time: 0.0449 data: 0.0295 max mem: 9671 +[14:52:11.116775] val: Total time: 0:00:00 (0.0496 s / it) +[14:52:11.133054] val loss: 0.34871040284633636 +[14:52:11.133375] Accuracy: 0.8630, F1 Score: 0.8629, ROC AUC: 0.9297, Hamming Loss: 0.1370, + Jaccard Score: 0.7589, Precision: 0.8631, Recall: 0.8630, + Average Precision: 0.9298, Kappa: 0.7259, Score: 0.8395 +[14:52:11.162262] Best epoch = 11, Best score = 0.8502 +[14:52:11.427407] log_dir: ./output_logs/retfound +[14:52:12.221491] Epoch: [16] [ 0/15] eta: 0:00:11 lr: 0.000496 loss: 0.6445 (0.6445) time: 0.7932 data: 0.7229 max mem: 9671 +[14:52:13.139230] Epoch: [16] [14/15] eta: 0:00:00 lr: 0.000458 loss: 0.4837 (0.5327) time: 0.1140 data: 0.0482 max mem: 9671 +[14:52:13.217343] Epoch: [16] Total time: 0:00:01 (0.1193 s / it) +[14:52:13.218191] Averaged stats: lr: 0.000458 loss: 0.4837 (0.5327) +[14:52:13.862092] val: [ 0/17] eta: 0:00:10 loss: 0.1990 (0.1990) time: 0.6194 data: 0.6024 max mem: 9671 +[14:52:14.111594] val: [10/17] eta: 0:00:00 loss: 0.2775 (0.3318) time: 0.0789 data: 0.0630 max mem: 9671 +[14:52:14.204302] val: [16/17] eta: 0:00:00 loss: 0.3232 (0.3466) time: 0.0565 data: 0.0408 max mem: 9671 +[14:52:14.280617] val: Total time: 0:00:01 (0.0611 s / it) +[14:52:14.295388] val loss: 0.34657756195348854 +[14:52:14.295740] Accuracy: 0.8537, F1 Score: 0.8535, ROC AUC: 0.9321, Hamming Loss: 0.1463, + Jaccard Score: 0.7445, Precision: 0.8555, Recall: 0.8537, + Average Precision: 0.9328, Kappa: 0.7074, Score: 0.8310 +[14:52:14.343384] Best epoch = 11, Best score = 0.8502 +[14:52:14.582071] log_dir: ./output_logs/retfound +[14:52:15.302117] Epoch: [17] [ 0/15] eta: 0:00:10 lr: 0.000455 loss: 0.4270 (0.4270) time: 0.7192 data: 0.6472 max mem: 9671 +[14:52:16.220304] Epoch: [17] [14/15] eta: 0:00:00 lr: 0.000413 loss: 0.5237 (0.5180) time: 0.1091 data: 0.0432 max mem: 9671 +[14:52:16.295022] Epoch: [17] Total time: 0:00:01 (0.1142 s / it) +[14:52:16.295781] Averaged stats: lr: 0.000413 loss: 0.5237 (0.5180) +[14:52:16.841734] val: [ 0/17] eta: 0:00:09 loss: 0.3123 (0.3123) time: 0.5345 data: 0.5176 max mem: 9671 +[14:52:16.996581] val: [10/17] eta: 0:00:00 loss: 0.3612 (0.3834) time: 0.0626 data: 0.0472 max mem: 9671 +[14:52:17.088417] val: [16/17] eta: 0:00:00 loss: 0.3505 (0.3285) time: 0.0459 data: 0.0305 max mem: 9671 +[14:52:17.166983] val: Total time: 0:00:00 (0.0506 s / it) +[14:52:17.182266] val loss: 0.3285454340717372 +[14:52:17.182464] Accuracy: 0.8648, F1 Score: 0.8647, ROC AUC: 0.9365, Hamming Loss: 0.1352, + Jaccard Score: 0.7617, Precision: 0.8657, Recall: 0.8648, + Average Precision: 0.9377, Kappa: 0.7296, Score: 0.8436 +[14:52:17.219870] Best epoch = 11, Best score = 0.8502 +[14:52:17.489000] log_dir: ./output_logs/retfound +[14:52:18.157904] Epoch: [18] [ 0/15] eta: 0:00:10 lr: 0.000409 loss: 0.5128 (0.5128) time: 0.6681 data: 0.5988 max mem: 9671 +[14:52:19.072973] Epoch: [18] [14/15] eta: 0:00:00 lr: 0.000365 loss: 0.4966 (0.5023) time: 0.1055 data: 0.0400 max mem: 9671 +[14:52:19.146532] Epoch: [18] Total time: 0:00:01 (0.1105 s / it) +[14:52:19.147263] Averaged stats: lr: 0.000365 loss: 0.4966 (0.5023) +[14:52:19.812836] val: [ 0/17] eta: 0:00:10 loss: 0.4386 (0.4386) time: 0.6469 data: 0.6297 max mem: 9671 +[14:52:20.012740] val: [10/17] eta: 0:00:00 loss: 0.4454 (0.4557) time: 0.0769 data: 0.0614 max mem: 9671 +[14:52:20.127120] val: [16/17] eta: 0:00:00 loss: 0.3695 (0.3458) time: 0.0565 data: 0.0411 max mem: 9671 +[14:52:20.213434] val: Total time: 0:00:01 (0.0617 s / it) +[14:52:20.229385] val loss: 0.34580644943258343 +[14:52:20.229617] Accuracy: 0.8667, F1 Score: 0.8663, ROC AUC: 0.9417, Hamming Loss: 0.1333, + Jaccard Score: 0.7641, Precision: 0.8712, Recall: 0.8667, + Average Precision: 0.9432, Kappa: 0.7333, Score: 0.8471 +[14:52:20.279466] Best epoch = 11, Best score = 0.8502 +[14:52:20.529505] log_dir: ./output_logs/retfound +[14:52:21.312035] Epoch: [19] [ 0/15] eta: 0:00:11 lr: 0.000362 loss: 0.4583 (0.4583) time: 0.7816 data: 0.7100 max mem: 9671 +[14:52:22.228428] Epoch: [19] [14/15] eta: 0:00:00 lr: 0.000316 loss: 0.4538 (0.4600) time: 0.1132 data: 0.0474 max mem: 9671 +[14:52:22.300658] Epoch: [19] Total time: 0:00:01 (0.1181 s / it) +[14:52:22.301416] Averaged stats: lr: 0.000316 loss: 0.4538 (0.4600) +[14:52:22.944687] val: [ 0/17] eta: 0:00:10 loss: 0.3119 (0.3119) time: 0.6312 data: 0.6131 max mem: 9671 +[14:52:23.105835] val: [10/17] eta: 0:00:00 loss: 0.4260 (0.4355) time: 0.0720 data: 0.0563 max mem: 9671 +[14:52:23.197595] val: [16/17] eta: 0:00:00 loss: 0.3119 (0.3736) time: 0.0519 data: 0.0364 max mem: 9671 +[14:52:23.275129] val: Total time: 0:00:00 (0.0566 s / it) +[14:52:23.288994] val loss: 0.3736278914353427 +[14:52:23.289173] Accuracy: 0.8611, F1 Score: 0.8611, ROC AUC: 0.9422, Hamming Loss: 0.1389, + Jaccard Score: 0.7561, Precision: 0.8612, Recall: 0.8611, + Average Precision: 0.9430, Kappa: 0.7222, Score: 0.8418 +[14:52:23.314576] Best epoch = 11, Best score = 0.8502 +[14:52:23.593157] log_dir: ./output_logs/retfound +[14:52:24.251510] Epoch: [20] [ 0/15] eta: 0:00:09 lr: 0.000313 loss: 0.5011 (0.5011) time: 0.6571 data: 0.5842 max mem: 9671 +[14:52:25.166496] Epoch: [20] [14/15] eta: 0:00:00 lr: 0.000267 loss: 0.5011 (0.5129) time: 0.1048 data: 0.0390 max mem: 9671 +[14:52:25.244041] Epoch: [20] Total time: 0:00:01 (0.1100 s / it) +[14:52:25.244730] Averaged stats: lr: 0.000267 loss: 0.5011 (0.5129) +[14:52:25.956825] val: [ 0/17] eta: 0:00:11 loss: 0.2624 (0.2624) time: 0.7006 data: 0.6834 max mem: 9671 +[14:52:26.184180] val: [10/17] eta: 0:00:00 loss: 0.3386 (0.3577) time: 0.0843 data: 0.0686 max mem: 9671 +[14:52:26.279011] val: [16/17] eta: 0:00:00 loss: 0.2912 (0.3224) time: 0.0601 data: 0.0446 max mem: 9671 +[14:52:26.354751] val: Total time: 0:00:01 (0.0647 s / it) +[14:52:26.368805] val loss: 0.3224005668478854 +[14:52:26.368981] Accuracy: 0.8593, F1 Score: 0.8593, ROC AUC: 0.9431, Hamming Loss: 0.1407, + Jaccard Score: 0.7532, Precision: 0.8593, Recall: 0.8593, + Average Precision: 0.9444, Kappa: 0.7185, Score: 0.8403 +[14:52:26.419403] Best epoch = 11, Best score = 0.8502 +[14:52:26.672629] log_dir: ./output_logs/retfound +[14:52:27.347399] Epoch: [21] [ 0/15] eta: 0:00:10 lr: 0.000264 loss: 0.5864 (0.5864) time: 0.6738 data: 0.6040 max mem: 9671 +[14:52:28.272533] Epoch: [21] [14/15] eta: 0:00:00 lr: 0.000220 loss: 0.4658 (0.4756) time: 0.1066 data: 0.0403 max mem: 9671 +[14:52:28.353148] Epoch: [21] Total time: 0:00:01 (0.1120 s / it) +[14:52:28.353922] Averaged stats: lr: 0.000220 loss: 0.4658 (0.4756) +[14:52:29.037464] val: [ 0/17] eta: 0:00:11 loss: 0.2773 (0.2773) time: 0.6725 data: 0.6555 max mem: 9671 +[14:52:29.289225] val: [10/17] eta: 0:00:00 loss: 0.3456 (0.3589) time: 0.0840 data: 0.0684 max mem: 9671 +[14:52:29.381329] val: [16/17] eta: 0:00:00 loss: 0.2773 (0.3206) time: 0.0597 data: 0.0443 max mem: 9671 +[14:52:29.463510] val: Total time: 0:00:01 (0.0647 s / it) +[14:52:29.477301] val loss: 0.32057176705668955 +[14:52:29.477600] Accuracy: 0.8611, F1 Score: 0.8611, ROC AUC: 0.9429, Hamming Loss: 0.1389, + Jaccard Score: 0.7561, Precision: 0.8611, Recall: 0.8611, + Average Precision: 0.9442, Kappa: 0.7222, Score: 0.8421 +[14:52:29.522964] Best epoch = 11, Best score = 0.8502 +[14:52:29.802012] log_dir: ./output_logs/retfound +[14:52:30.445710] Epoch: [22] [ 0/15] eta: 0:00:09 lr: 0.000217 loss: 0.3436 (0.3436) time: 0.6427 data: 0.5722 max mem: 9671 +[14:52:31.363103] Epoch: [22] [14/15] eta: 0:00:00 lr: 0.000174 loss: 0.4529 (0.4597) time: 0.1039 data: 0.0382 max mem: 9671 +[14:52:31.441504] Epoch: [22] Total time: 0:00:01 (0.1093 s / it) +[14:52:31.442316] Averaged stats: lr: 0.000174 loss: 0.4529 (0.4597) +[14:52:32.040118] val: [ 0/17] eta: 0:00:09 loss: 0.3551 (0.3551) time: 0.5861 data: 0.5696 max mem: 9671 +[14:52:32.290537] val: [10/17] eta: 0:00:00 loss: 0.4240 (0.4223) time: 0.0760 data: 0.0605 max mem: 9671 +[14:52:32.403335] val: [16/17] eta: 0:00:00 loss: 0.3156 (0.3455) time: 0.0558 data: 0.0404 max mem: 9671 +[14:52:32.483244] val: Total time: 0:00:01 (0.0606 s / it) +[14:52:32.497054] val loss: 0.34548356121077256 +[14:52:32.497271] Accuracy: 0.8630, F1 Score: 0.8630, ROC AUC: 0.9429, Hamming Loss: 0.1370, + Jaccard Score: 0.7589, Precision: 0.8630, Recall: 0.8630, + Average Precision: 0.9441, Kappa: 0.7259, Score: 0.8439 +[14:52:32.532669] Best epoch = 11, Best score = 0.8502 +[14:52:32.806896] log_dir: ./output_logs/retfound +[14:52:33.607481] Epoch: [23] [ 0/15] eta: 0:00:11 lr: 0.000171 loss: 0.4566 (0.4566) time: 0.7997 data: 0.7291 max mem: 9671 +[14:52:34.521275] Epoch: [23] [14/15] eta: 0:00:00 lr: 0.000132 loss: 0.4491 (0.4459) time: 0.1142 data: 0.0487 max mem: 9671 +[14:52:34.594509] Epoch: [23] Total time: 0:00:01 (0.1192 s / it) +[14:52:34.595252] Averaged stats: lr: 0.000132 loss: 0.4491 (0.4459) +[14:52:35.081863] val: [ 0/17] eta: 0:00:08 loss: 0.3678 (0.3678) time: 0.4754 data: 0.4569 max mem: 9671 +[14:52:35.282058] val: [10/17] eta: 0:00:00 loss: 0.4448 (0.4414) time: 0.0614 data: 0.0456 max mem: 9671 +[14:52:35.373837] val: [16/17] eta: 0:00:00 loss: 0.3291 (0.3623) time: 0.0451 data: 0.0295 max mem: 9671 +[14:52:35.456806] val: Total time: 0:00:00 (0.0501 s / it) +[14:52:35.470711] val loss: 0.3623477816581726 +[14:52:35.470940] Accuracy: 0.8630, F1 Score: 0.8630, ROC AUC: 0.9434, Hamming Loss: 0.1370, + Jaccard Score: 0.7589, Precision: 0.8630, Recall: 0.8630, + Average Precision: 0.9444, Kappa: 0.7259, Score: 0.8441 +[14:52:35.497148] Best epoch = 11, Best score = 0.8502 +[14:52:35.776516] log_dir: ./output_logs/retfound +[14:52:36.474017] Epoch: [24] [ 0/15] eta: 0:00:10 lr: 0.000130 loss: 0.4917 (0.4917) time: 0.6966 data: 0.6214 max mem: 9671 +[14:52:37.386406] Epoch: [24] [14/15] eta: 0:00:00 lr: 0.000095 loss: 0.4204 (0.4388) time: 0.1072 data: 0.0415 max mem: 9671 +[14:52:37.465290] Epoch: [24] Total time: 0:00:01 (0.1126 s / it) +[14:52:37.466034] Averaged stats: lr: 0.000095 loss: 0.4204 (0.4388) +[14:52:37.963931] val: [ 0/17] eta: 0:00:08 loss: 0.3430 (0.3430) time: 0.4909 data: 0.4739 max mem: 9671 +[14:52:38.158573] val: [10/17] eta: 0:00:00 loss: 0.4299 (0.4205) time: 0.0623 data: 0.0467 max mem: 9671 +[14:52:38.250328] val: [16/17] eta: 0:00:00 loss: 0.2915 (0.3589) time: 0.0457 data: 0.0303 max mem: 9671 +[14:52:38.328524] val: Total time: 0:00:00 (0.0504 s / it) +[14:52:38.342306] val loss: 0.3589090209673433 +[14:52:38.342552] Accuracy: 0.8593, F1 Score: 0.8593, ROC AUC: 0.9439, Hamming Loss: 0.1407, + Jaccard Score: 0.7532, Precision: 0.8593, Recall: 0.8593, + Average Precision: 0.9449, Kappa: 0.7185, Score: 0.8406 +[14:52:38.384401] Best epoch = 11, Best score = 0.8502 +[14:52:38.651179] log_dir: ./output_logs/retfound +[14:52:39.290944] Epoch: [25] [ 0/15] eta: 0:00:09 lr: 0.000092 loss: 0.4097 (0.4097) time: 0.6387 data: 0.5554 max mem: 9671 +[14:52:40.209478] Epoch: [25] [14/15] eta: 0:00:00 lr: 0.000063 loss: 0.4781 (0.4603) time: 0.1038 data: 0.0371 max mem: 9671 +[14:52:40.286469] Epoch: [25] Total time: 0:00:01 (0.1090 s / it) +[14:52:40.287295] Averaged stats: lr: 0.000063 loss: 0.4781 (0.4603) +[14:52:40.899116] val: [ 0/17] eta: 0:00:10 loss: 0.3457 (0.3457) time: 0.6051 data: 0.5885 max mem: 9671 +[14:52:41.054836] val: [10/17] eta: 0:00:00 loss: 0.4143 (0.4095) time: 0.0691 data: 0.0536 max mem: 9671 +[14:52:41.146726] val: [16/17] eta: 0:00:00 loss: 0.3005 (0.3484) time: 0.0501 data: 0.0347 max mem: 9671 +[14:52:41.244696] val: Total time: 0:00:00 (0.0560 s / it) +[14:52:41.260720] val loss: 0.3483984185492291 +[14:52:41.260961] Accuracy: 0.8574, F1 Score: 0.8574, ROC AUC: 0.9437, Hamming Loss: 0.1426, + Jaccard Score: 0.7504, Precision: 0.8575, Recall: 0.8574, + Average Precision: 0.9446, Kappa: 0.7148, Score: 0.8387 +[14:52:41.305099] Best epoch = 11, Best score = 0.8502 +[14:52:41.575479] log_dir: ./output_logs/retfound +[14:52:42.273111] Epoch: [26] [ 0/15] eta: 0:00:10 lr: 0.000061 loss: 0.4052 (0.4052) time: 0.6968 data: 0.6272 max mem: 9671 +[14:52:43.185044] Epoch: [26] [14/15] eta: 0:00:00 lr: 0.000037 loss: 0.4377 (0.4451) time: 0.1072 data: 0.0419 max mem: 9671 +[14:52:43.274556] Epoch: [26] Total time: 0:00:01 (0.1133 s / it) +[14:52:43.275269] Averaged stats: lr: 0.000037 loss: 0.4377 (0.4451) +[14:52:43.853836] val: [ 0/17] eta: 0:00:09 loss: 0.3701 (0.3701) time: 0.5674 data: 0.5501 max mem: 9671 +[14:52:44.133236] val: [10/17] eta: 0:00:00 loss: 0.4211 (0.4204) time: 0.0769 data: 0.0613 max mem: 9671 +[14:52:44.224984] val: [16/17] eta: 0:00:00 loss: 0.3356 (0.3485) time: 0.0551 data: 0.0397 max mem: 9671 +[14:52:44.308253] val: Total time: 0:00:01 (0.0602 s / it) +[14:52:44.322099] val loss: 0.34846346671966943 +[14:52:44.322285] Accuracy: 0.8611, F1 Score: 0.8611, ROC AUC: 0.9442, Hamming Loss: 0.1389, + Jaccard Score: 0.7561, Precision: 0.8614, Recall: 0.8611, + Average Precision: 0.9452, Kappa: 0.7222, Score: 0.8425 +[14:52:44.358492] Best epoch = 11, Best score = 0.8502 +[14:52:44.606015] log_dir: ./output_logs/retfound +[14:52:45.253763] Epoch: [27] [ 0/15] eta: 0:00:09 lr: 0.000035 loss: 0.4714 (0.4714) time: 0.6469 data: 0.5792 max mem: 9671 +[14:52:46.166877] Epoch: [27] [14/15] eta: 0:00:00 lr: 0.000017 loss: 0.4569 (0.4722) time: 0.1040 data: 0.0387 max mem: 9671 +[14:52:46.242345] Epoch: [27] Total time: 0:00:01 (0.1091 s / it) +[14:52:46.243089] Averaged stats: lr: 0.000017 loss: 0.4569 (0.4722) +[14:52:46.882075] val: [ 0/17] eta: 0:00:10 loss: 0.4043 (0.4043) time: 0.6276 data: 0.6101 max mem: 9671 +[14:52:47.037810] val: [10/17] eta: 0:00:00 loss: 0.4043 (0.4338) time: 0.0712 data: 0.0556 max mem: 9671 +[14:52:47.129534] val: [16/17] eta: 0:00:00 loss: 0.3222 (0.3487) time: 0.0514 data: 0.0360 max mem: 9671 +[14:52:47.203645] val: Total time: 0:00:00 (0.0559 s / it) +[14:52:47.217612] val loss: 0.3486983846215641 +[14:52:47.217830] Accuracy: 0.8574, F1 Score: 0.8573, ROC AUC: 0.9443, Hamming Loss: 0.1426, + Jaccard Score: 0.7503, Precision: 0.8580, Recall: 0.8574, + Average Precision: 0.9450, Kappa: 0.7148, Score: 0.8388 +[14:52:47.244709] Best epoch = 11, Best score = 0.8502 +[14:52:47.546845] log_dir: ./output_logs/retfound +[14:52:48.280148] Epoch: [28] [ 0/15] eta: 0:00:10 lr: 0.000016 loss: 0.3572 (0.3572) time: 0.7324 data: 0.6615 max mem: 9671 +[14:52:49.193316] Epoch: [28] [14/15] eta: 0:00:00 lr: 0.000005 loss: 0.4081 (0.4453) time: 0.1097 data: 0.0442 max mem: 9671 +[14:52:49.274655] Epoch: [28] Total time: 0:00:01 (0.1152 s / it) +[14:52:49.275445] Averaged stats: lr: 0.000005 loss: 0.4081 (0.4453) +[14:52:49.833774] val: [ 0/17] eta: 0:00:09 loss: 0.4012 (0.4012) time: 0.5462 data: 0.5292 max mem: 9671 +[14:52:49.989689] val: [10/17] eta: 0:00:00 loss: 0.4012 (0.4330) time: 0.0638 data: 0.0482 max mem: 9671 +[14:52:50.081694] val: [16/17] eta: 0:00:00 loss: 0.3215 (0.3488) time: 0.0466 data: 0.0312 max mem: 9671 +[14:52:50.176624] val: Total time: 0:00:00 (0.0523 s / it) +[14:52:50.190787] val loss: 0.3487990077804117 +[14:52:50.191086] Accuracy: 0.8574, F1 Score: 0.8573, ROC AUC: 0.9443, Hamming Loss: 0.1426, + Jaccard Score: 0.7503, Precision: 0.8580, Recall: 0.8574, + Average Precision: 0.9450, Kappa: 0.7148, Score: 0.8388 +[14:52:50.226805] Best epoch = 11, Best score = 0.8502 +[14:52:50.501348] log_dir: ./output_logs/retfound +[14:52:51.167016] Epoch: [29] [ 0/15] eta: 0:00:09 lr: 0.000005 loss: 0.6518 (0.6518) time: 0.6648 data: 0.5951 max mem: 9671 +[14:52:52.080707] Epoch: [29] [14/15] eta: 0:00:00 lr: 0.000001 loss: 0.4509 (0.4721) time: 0.1052 data: 0.0397 max mem: 9671 +[14:52:52.153958] Epoch: [29] Total time: 0:00:01 (0.1102 s / it) +[14:52:52.154681] Averaged stats: lr: 0.000001 loss: 0.4509 (0.4721) +[14:52:52.854544] val: [ 0/17] eta: 0:00:11 loss: 0.3984 (0.3984) time: 0.6877 data: 0.6705 max mem: 9671 +[14:52:53.096542] val: [10/17] eta: 0:00:00 loss: 0.3984 (0.4320) time: 0.0845 data: 0.0689 max mem: 9671 +[14:52:53.188708] val: [16/17] eta: 0:00:00 loss: 0.3223 (0.3487) time: 0.0600 data: 0.0446 max mem: 9671 +[14:52:53.268562] val: Total time: 0:00:01 (0.0649 s / it) +[14:52:53.282403] val loss: 0.34871361141695695 +[14:52:53.282719] Accuracy: 0.8593, F1 Score: 0.8592, ROC AUC: 0.9442, Hamming Loss: 0.1407, + Jaccard Score: 0.7532, Precision: 0.8598, Recall: 0.8593, + Average Precision: 0.9449, Kappa: 0.7185, Score: 0.8406 +[14:52:53.317319] Best epoch = 11, Best score = 0.8502 +[14:52:56.384697] Test with the best model, epoch = 11: +[14:52:57.044331] test: [ 0/32] eta: 0:00:20 loss: 0.1509 (0.1509) time: 0.6496 data: 0.6331 max mem: 9671 +[14:52:57.285002] test: [10/32] eta: 0:00:01 loss: 0.3301 (0.3758) time: 0.0809 data: 0.0653 max mem: 9671 +[14:52:57.536249] test: [20/32] eta: 0:00:00 loss: 0.3301 (0.3419) time: 0.0245 data: 0.0087 max mem: 9671 +[14:52:57.762260] test: [30/32] eta: 0:00:00 loss: 0.3389 (0.3379) time: 0.0238 data: 0.0080 max mem: 9671 +[14:52:57.809988] test: [31/32] eta: 0:00:00 loss: 0.3393 (0.3510) time: 0.0254 data: 0.0080 max mem: 9671 +[14:52:57.884576] test: Total time: 0:00:01 (0.0466 s / it) +[14:52:57.902954] val loss: 0.3509616069495678 +[14:52:57.903102] Accuracy: 0.8510, F1 Score: 0.8510, ROC AUC: 0.9329, Hamming Loss: 0.1490, + Jaccard Score: 0.7406, Precision: 0.8512, Recall: 0.8510, + Average Precision: 0.9354, Kappa: 0.7020, Score: 0.8286 +[14:52:58.629753] Training time 0:01:53 +[rank0]:[W701 14:52:59.983011891 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/010/retfound acc=0.8510 auroc=0.9329420000000002 f1_macro=0.8510 qwk=0.702 diff --git a/results/downsample/airogs/010/vit/confusion_matrix.png b/results/downsample/airogs/010/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..9c6077e2f1d6af2a99e672cbc6ea95d8eb80a776 --- /dev/null +++ b/results/downsample/airogs/010/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76cff98a8adbd189a5d5887208a9fbdca7d67194cae6693e9f25e78bd3437b93 +size 73450 diff --git a/results/downsample/airogs/010/vit/log.csv b/results/downsample/airogs/010/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..10a922437fc80276f980b93fa9a7a399da341da3 --- /dev/null +++ b/results/downsample/airogs/010/vit/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7695331743785313,0.5388888888888889,0.6052949245541839,0.3923244121450487,6.338067241746367e-08 +1,0.7003041250365121,0.6185185185185185,0.678113854595336,0.5046173967778906,1.3732479023783798e-07 +2,0.6559353981699262,0.6962962962962963,0.7442524005486969,0.610811864707663,2.112689080582123e-07 +3,0.6137891667229789,0.7129629629629629,0.7930589849108367,0.64392714061541,2.852130258785866e-07 +4,0.550925863640649,0.7481481481481481,0.8237928669410152,0.6885644510953154,3.5915714369896085e-07 +5,0.5285699708121163,0.7592592592592593,0.8419958847736625,0.7065372463166311,3.6864926748197445e-07 +6,0.47314724751881193,0.7851851851851852,0.8581344307270232,0.7377983036625011,3.647092708908402e-07 +7,0.46333701269967215,0.7962962962962963,0.8699931412894376,0.7528674057995045,3.579329485062548e-07 +8,0.4202831642968314,0.8,0.8717626886145404,0.7572395944102143,3.48427166980202e-07 +9,0.4173589178494045,0.7944444444444444,0.8689368998628257,0.7503974744527165,3.3634183816582624e-07 +10,0.39630875842911856,0.7814814814814814,0.8701577503429356,0.7376683070228921,3.218675549179558e-07 +11,0.3730865035738264,0.7666666666666667,0.8652537722908092,0.7200939072292302,3.0523258532992716e-07 +12,0.3701561519077846,0.7888888888888889,0.869917695473251,0.7450577928954306,2.866992728093928e-07 +13,0.3604859582015446,0.762962962962963,0.8712757201646091,0.7181899213829728,2.665598987660658e-07 +14,0.35709546293531147,0.762962962962963,0.8663374485596709,0.716449464909504,2.4513207315928833e-07 +15,0.3510198252541678,0.8055555555555556,0.8764197530864197,0.7643601393987037,2.2275372559924176e-07 +16,0.3377582515989031,0.8,0.873079561042524,0.7576703168026252,1.9977777599512494e-07 +17,0.3333524635859898,0.8092592592592592,0.8812482853223593,0.769516005376954,1.765665687973618e-07 +18,0.32063582113810946,0.8055555555555556,0.8826406035665294,0.7663981708194157,1.5348615860916415e-07 +19,0.3232965384210859,0.812962962962963,0.8831344307270234,0.773990450149174,1.3090053728676854e-07 +20,0.3125927065099989,0.8,0.8813991769547326,0.7600208185960465,1.0916589357042543e-07 +21,0.2963191568851471,0.8037037037037037,0.879039780521262,0.7633836305441243,8.862499577518838e-08 +22,0.29086572783333914,0.7981481481481482,0.8793689986282579,0.757508452510011,6.960178612982571e-08 +23,0.29345465132168363,0.7925925925925926,0.878278463648834,0.7520035699724614,5.23962720143595e-08 +24,0.29507732817104887,0.7981481481481482,0.8788683127572017,0.7577189623485919,3.727979466446665e-08 +25,0.29290648869105745,0.7981481481481482,0.8786556927297668,0.7575314142695561,2.4490749958100682e-08 +26,0.2828697349343981,0.8,0.8785459533607681,0.7593604090593283,1.4230828770086085e-08 +27,0.28560382979256765,0.7962962962962963,0.8785665294924555,0.7557588089048757,6.661836186537674e-09 +28,0.2916159544672285,0.7962962962962963,0.8786145404663923,0.7557748125628546,1.9031397419949733e-09 +29,0.28352344036102295,0.7962962962962963,0.8785665294924554,0.7557588089048757,2.978692209590643e-11 diff --git a/results/downsample/airogs/010/vit/metrics.json b/results/downsample/airogs/010/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..3d7eb19b3e1ce27c878450b31225ca7be9715230 --- /dev/null +++ b/results/downsample/airogs/010/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8, + "balanced_accuracy": 0.8, + "precision_macro": 0.8, + "recall_macro": 0.8, + "f1_macro": 0.8, + "precision_weighted": 0.8, + "recall_weighted": 0.8, + "f1_weighted": 0.8, + "cohen_kappa": 0.6, + "quadratic_weighted_kappa": 0.6, + "mcc": 0.6, + "auroc": 0.890196, + "auprc": 0.8906712988441068, + "sensitivity": 0.8, + "specificity": 0.8, + "precision_pos": 0.8, + "f1_pos": 0.8, + "per_class": { + "0": { + "precision": 0.8, + "recall": 0.8, + "f1-score": 0.8, + "support": 500.0 + }, + "1": { + "precision": 0.8, + "recall": 0.8, + "f1-score": 0.8, + "support": 500.0 + }, + "accuracy": 0.8, + "macro avg": { + "precision": 0.8, + "recall": 0.8, + "f1-score": 0.8, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8, + "recall": 0.8, + "f1-score": 0.8, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/010/vit/pr.png b/results/downsample/airogs/010/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..97c6be9619eeab75874e203393c8272b8821eab9 --- /dev/null +++ b/results/downsample/airogs/010/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9499f672d32d8fb02c033ccf0e61f031fbac39a0c20ab7de2c36d446d62b3d81 +size 52252 diff --git a/results/downsample/airogs/010/vit/roc.png b/results/downsample/airogs/010/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..2d4c521f4a6554271b5e8b3e6f1d45346db2149a --- /dev/null +++ b/results/downsample/airogs/010/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ebc749f583a589b28da5f439d407b45dd2e78d3ffc721d8ba25492f9ef92dd4 +size 64794 diff --git a/results/downsample/airogs/010/vit/test_pred.npz b/results/downsample/airogs/010/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..ecf8cc6dcf4f2ce7de3080e56babe304c683f607 --- /dev/null +++ b/results/downsample/airogs/010/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3637a9c0de84627a80ee6c2f335687535afc8fa5d7d0cf79b48eeea16da1ec1c +size 16510 diff --git a/results/downsample/airogs/010/vit/train.log b/results/downsample/airogs/010/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..cdd46bc164bb30ed464381dbf3e7268e4e592425 --- /dev/null +++ b/results/downsample/airogs/010/vit/train.log @@ -0,0 +1,159 @@ +[vit] train=500 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7695 val_acc=0.5389 val_auc=0.6053 score=0.3923 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7003 val_acc=0.6185 val_auc=0.6781 score=0.5046 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6559 val_acc=0.6963 val_auc=0.7443 score=0.6108 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6138 val_acc=0.7130 val_auc=0.7931 score=0.6439 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5509 val_acc=0.7481 val_auc=0.8238 score=0.6886 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.5286 val_acc=0.7593 val_auc=0.8420 score=0.7065 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.4731 val_acc=0.7852 val_auc=0.8581 score=0.7378 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4633 val_acc=0.7963 val_auc=0.8700 score=0.7529 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.4203 val_acc=0.8000 val_auc=0.8718 score=0.7572 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4174 val_acc=0.7944 val_auc=0.8689 score=0.7504 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3963 val_acc=0.7815 val_auc=0.8702 score=0.7377 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3731 val_acc=0.7667 val_auc=0.8653 score=0.7201 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3702 val_acc=0.7889 val_auc=0.8699 score=0.7451 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.3605 val_acc=0.7630 val_auc=0.8713 score=0.7182 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.3571 val_acc=0.7630 val_auc=0.8663 score=0.7164 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.3510 val_acc=0.8056 val_auc=0.8764 score=0.7644 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3378 val_acc=0.8000 val_auc=0.8731 score=0.7577 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.3334 val_acc=0.8093 val_auc=0.8812 score=0.7695 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.3206 val_acc=0.8056 val_auc=0.8826 score=0.7664 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.3233 val_acc=0.8130 val_auc=0.8831 score=0.7740 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.3126 val_acc=0.8000 val_auc=0.8814 score=0.7600 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.2963 val_acc=0.8037 val_auc=0.8790 score=0.7634 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.2909 val_acc=0.7981 val_auc=0.8794 score=0.7575 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.2935 val_acc=0.7926 val_auc=0.8783 score=0.7520 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.2951 val_acc=0.7981 val_auc=0.8789 score=0.7577 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.2929 val_acc=0.7981 val_auc=0.8787 score=0.7575 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.2829 val_acc=0.8000 val_auc=0.8785 score=0.7594 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.2856 val_acc=0.7963 val_auc=0.8786 score=0.7558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.2916 val_acc=0.7963 val_auc=0.8786 score=0.7558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.2835 val_acc=0.7963 val_auc=0.8786 score=0.7558 +[vit] early stop at ep29 (best ep19 score=0.7740) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=19 best_val_score=0.7740 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/010/vit acc=0.8000 auroc=0.890196 f1_macro=0.8000 qwk=0.6 diff --git a/results/downsample/airogs/025/resnet/confusion_matrix.png b/results/downsample/airogs/025/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..405542a41a244413aef0734f97677e7487a8337e --- /dev/null +++ b/results/downsample/airogs/025/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30ce00b6f25b75270aea47b7b8a01a947a6bb8ba4414f6a05158bb26c2fa2dbc +size 71578 diff --git a/results/downsample/airogs/025/resnet/log.csv b/results/downsample/airogs/025/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..61bc28503bbc0d901a39fc30b11d3012cab389ef --- /dev/null +++ b/results/downsample/airogs/025/resnet/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6925321880139803,0.6111111111111112,0.6373319615912209,0.4898204348521591,0.00015789473684210527 +1,0.6734117206774259,0.6592592592592592,0.7344101508916325,0.5696662246136038,0.0003245614035087719 +2,0.6028906671624434,0.687037037037037,0.7899451303155007,0.6120626804633412,0.0004912280701754386 +3,0.4381817610640275,0.7962962962962963,0.8855555555555555,0.7577735206470839,0.0004984826693294874 +4,0.3505786625962508,0.7925925925925926,0.8841632373113855,0.7537051700426405,0.0004936097246739206 +5,0.3222995133776414,0.812962962962963,0.9156035665294925,0.7845039151885927,0.0004854423764006961 +6,0.2528576733250367,0.6981481481481482,0.8679835390946502,0.6470028283493665,0.00047409107388425505 +7,0.2311326641785471,0.7888888888888889,0.9095404663923183,0.7572362036949117,0.00045970932401815566 +8,0.17976600206211993,0.8425925925925926,0.9174759945130315,0.8150628095531828,0.0004424916152976768 +9,0.13190069559373355,0.812962962962963,0.9117283950617283,0.7829326733547163,0.00042267078769656107 +10,0.13150812803130402,0.8296296296296296,0.9183196159122085,0.8024020559089736,0.00040051488390580335 +11,0.10336887503140851,0.7796296296296297,0.9049382716049382,0.7459848735864383,0.00037632352451607746 +12,0.09445429446273729,0.8314814814814815,0.9217009602194788,0.8051168380694036,0.00035042385616324234 +13,0.08471647238260821,0.8388888888888889,0.9180246913580246,0.8115105083415086,0.0003231661274313064 +14,0.057033879388319816,0.8166666666666667,0.9215226337448561,0.7894932124998585,0.0002949189523411747 +15,0.08317540968327146,0.8629629629629629,0.9246227709190673,0.8378271921631198,0.00026606432547836756 +16,0.051500341119734866,0.825925925925926,0.9130041152263374,0.7968475934216674,0.00023699245617155696 +17,0.04175317052163576,0.837037037037037,0.9147873799725652,0.8085066084350374,0.0002080964915807942 +18,0.053221303577485835,0.8481481481481481,0.9161454046639231,0.820190366550472,0.00017976720005660773 +19,0.041184560640862115,0.8425925925925926,0.9231344307270233,0.8169662373998711,0.00015238768666841907 +20,0.027553899801875417,0.8611111111111112,0.9266529492455418,0.8365281854053191,0.00012632821236568945 +21,0.03255317860135907,0.8574074074074074,0.9280109739368998,0.8333521493690298,0.00010194118683375503 +22,0.03271028389664073,0.8555555555555555,0.9285939643347052,0.8316734941208823,7.955640275738594e-05 +23,0.026222456932852144,0.8555555555555555,0.9281412894375858,0.831472851239237,5.947657594047864e-05 +24,0.020991868682597812,0.8592592592592593,0.9281138545953361,0.8352450266083385,4.197325159411258e-05 +25,0.023985265018908603,0.8629629629629629,0.926090534979424,0.8382756637080372,2.7283132153401252e-05 +26,0.02559774025882545,0.8555555555555555,0.9272839506172841,0.8312633149773183,1.560487628311738e-05 +27,0.031105483637044306,0.8611111111111112,0.927750342935528,0.8369921420538704,7.096412360046545e-06 +28,0.017009241996627105,0.8611111111111112,0.9282235939643348,0.837149892396806,1.8728027626157273e-06 +29,0.016364647075533867,0.85,0.9270096021947873,0.8255624698555183,4.687849611939577e-09 diff --git a/results/downsample/airogs/025/resnet/metrics.json b/results/downsample/airogs/025/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..6defc997c8b26d1bff47f31460c537768c904a6f --- /dev/null +++ b/results/downsample/airogs/025/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.855, + "balanced_accuracy": 0.855, + "precision_macro": 0.8550127804600965, + "recall_macro": 0.855, + "f1_macro": 0.8549986949882549, + "precision_weighted": 0.8550127804600965, + "recall_weighted": 0.855, + "f1_weighted": 0.8549986949882549, + "cohen_kappa": 0.71, + "quadratic_weighted_kappa": 0.71, + "mcc": 0.7100127803450703, + "auroc": 0.939054, + "auprc": 0.9379741079636663, + "sensitivity": 0.852, + "specificity": 0.858, + "precision_pos": 0.8571428571428571, + "f1_pos": 0.8545636910732196, + "per_class": { + "0": { + "precision": 0.852882703777336, + "recall": 0.858, + "f1-score": 0.8554336989032901, + "support": 500.0 + }, + "1": { + "precision": 0.8571428571428571, + "recall": 0.852, + "f1-score": 0.8545636910732196, + "support": 500.0 + }, + "accuracy": 0.855, + "macro avg": { + "precision": 0.8550127804600965, + "recall": 0.855, + "f1-score": 0.8549986949882549, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8550127804600965, + "recall": 0.855, + "f1-score": 0.8549986949882549, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/025/resnet/pr.png b/results/downsample/airogs/025/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..62ab3d916e27836ac560d3bf99a7124e4bb50bbb --- /dev/null +++ b/results/downsample/airogs/025/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74db94f86a0f4bd15bd6d818983b6cf9cba87ce6a59218ec10ee7bcc1914dd8a +size 48079 diff --git a/results/downsample/airogs/025/resnet/roc.png b/results/downsample/airogs/025/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..67790eaad64fc090e91bb2921bd52336dadf922c --- /dev/null +++ b/results/downsample/airogs/025/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87003e95ddbd9003f7e21138fdf13b0436c6ea2cecbbce82621ec73857d75a38 +size 62638 diff --git a/results/downsample/airogs/025/resnet/test_pred.npz b/results/downsample/airogs/025/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..116205df868c8d879d2f3d608e6ce41759860bfb --- /dev/null +++ b/results/downsample/airogs/025/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b9841895de4ca2bf95775fb43e3b44a5eeeb757345d645a6bf5ea2cb41fefd2 +size 16510 diff --git a/results/downsample/airogs/025/resnet/train.log b/results/downsample/airogs/025/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b0827c10f650ededed03fcc17384023a8426ccbc --- /dev/null +++ b/results/downsample/airogs/025/resnet/train.log @@ -0,0 +1,158 @@ +[resnet] train=1250 val=540 test=1000 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6925 val_acc=0.6111 val_auc=0.6373 score=0.4898 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6734 val_acc=0.6593 val_auc=0.7344 score=0.5697 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6029 val_acc=0.6870 val_auc=0.7899 score=0.6121 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.4382 val_acc=0.7963 val_auc=0.8856 score=0.7578 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.3506 val_acc=0.7926 val_auc=0.8842 score=0.7537 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.3223 val_acc=0.8130 val_auc=0.9156 score=0.7845 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.2529 val_acc=0.6981 val_auc=0.8680 score=0.6470 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.2311 val_acc=0.7889 val_auc=0.9095 score=0.7572 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.1798 val_acc=0.8426 val_auc=0.9175 score=0.8151 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1319 val_acc=0.8130 val_auc=0.9117 score=0.7829 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1315 val_acc=0.8296 val_auc=0.9183 score=0.8024 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1034 val_acc=0.7796 val_auc=0.9049 score=0.7460 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.0945 val_acc=0.8315 val_auc=0.9217 score=0.8051 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.0847 val_acc=0.8389 val_auc=0.9180 score=0.8115 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0570 val_acc=0.8167 val_auc=0.9215 score=0.7895 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0832 val_acc=0.8630 val_auc=0.9246 score=0.8378 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0515 val_acc=0.8259 val_auc=0.9130 score=0.7968 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0418 val_acc=0.8370 val_auc=0.9148 score=0.8085 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0532 val_acc=0.8481 val_auc=0.9161 score=0.8202 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0412 val_acc=0.8426 val_auc=0.9231 score=0.8170 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0276 val_acc=0.8611 val_auc=0.9267 score=0.8365 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0326 val_acc=0.8574 val_auc=0.9280 score=0.8334 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0327 val_acc=0.8556 val_auc=0.9286 score=0.8317 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0262 val_acc=0.8556 val_auc=0.9281 score=0.8315 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0210 val_acc=0.8593 val_auc=0.9281 score=0.8352 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0240 val_acc=0.8630 val_auc=0.9261 score=0.8383 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0256 val_acc=0.8556 val_auc=0.9273 score=0.8313 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0311 val_acc=0.8611 val_auc=0.9278 score=0.8370 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0170 val_acc=0.8611 val_auc=0.9282 score=0.8371 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0164 val_acc=0.8500 val_auc=0.9270 score=0.8256 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=25 best_val_score=0.8383 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/025/resnet acc=0.8550 auroc=0.939054 f1_macro=0.8550 qwk=0.71 diff --git a/results/downsample/airogs/025/retfound/confusion_matrix.png b/results/downsample/airogs/025/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..38bf8a0ab114819b6c034abeb63396c5c36df236 --- /dev/null +++ b/results/downsample/airogs/025/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30cc4c275018ba2022b923779a7dc836c2b1618d98bd6ec4195c51e9ddd97775 +size 71791 diff --git a/results/downsample/airogs/025/retfound/confusion_matrix_test.jpg b/results/downsample/airogs/025/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..1e6fc7ae30478f6798fcfd5e5d0bb317f8b0565e --- /dev/null +++ b/results/downsample/airogs/025/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:21b1babf38594f16ce12098714b9c2de844a7629b00aad59802855f0839631fc +size 253800 diff --git a/results/downsample/airogs/025/retfound/log.txt b/results/downsample/airogs/025/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..7e8ef0807d03973a11b982df24fde4f735e1e9d7 --- /dev/null +++ b/results/downsample/airogs/025/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 3.0448717948717936e-05, "train_loss": 0.692771715995593, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.294871794871792e-05, "train_loss": 0.6863047281901041, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015544871794871798, "train_loss": 0.6545914380978315, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021794871794871795, "train_loss": 0.6095814827160958, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.00028044871794871793, "train_loss": 0.5815156300862631, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003429487179487179, "train_loss": 0.5631498862535526, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040544871794871804, "train_loss": 0.5339683233163296, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00046794871794871793, "train_loss": 0.5374089112648597, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005304487179487179, "train_loss": 0.570064290975913, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000592948717948718, "train_loss": 0.5338524778683981, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.000623767359987885, "train_loss": 0.5372465436275189, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006162137294824859, "train_loss": 0.5516729171459491, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006011939706485536, "train_loss": 0.5101183438912417, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005790779197963018, "train_loss": 0.49924285289568776, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005504101474966774, "train_loss": 0.5006361901760101, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005158965494465786, "train_loss": 0.4818159953141824, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0004763869649740636, "train_loss": 0.5029546725444305, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.000432854251173639, "train_loss": 0.4893035506590819, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.000386370327935865, "train_loss": 0.47119541733692855, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0003380797837221521, "train_loss": 0.45695631549908566, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.00028917169199816977, "train_loss": 0.4579616273060823, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024085033229881287, "train_loss": 0.45704332299721545, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019430553686927788, "train_loss": 0.4564189131443317, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0001506833930463599, "train_loss": 0.4419672496807881, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011105802278457405, "train_loss": 0.4272960707163199, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.640513420882977e-05, "train_loss": 0.43100996430103594, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.757799644220549e-05, "train_loss": 0.42728843368016756, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5286429288300625e-05, "train_loss": 0.4143176865883363, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.0079325111914166e-05, "train_loss": 0.4151816765467326, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.3311332873539566e-06, "train_loss": 0.40696347371125835, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/025/retfound/metrics.json b/results/downsample/airogs/025/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..6c329813eddafe7bb72aa591faeb0372ca692f99 --- /dev/null +++ b/results/downsample/airogs/025/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.881, + "balanced_accuracy": 0.881, + "precision_macro": 0.8811234840088189, + "recall_macro": 0.881, + "f1_macro": 0.8809903602191778, + "precision_weighted": 0.8811234840088189, + "recall_weighted": 0.881, + "f1_weighted": 0.8809903602191778, + "cohen_kappa": 0.762, + "quadratic_weighted_kappa": 0.762, + "mcc": 0.7621234740049935, + "auroc": 0.9543219999999999, + "auprc": 0.9571367736701105, + "sensitivity": 0.872, + "specificity": 0.89, + "precision_pos": 0.8879837067209776, + "f1_pos": 0.8799192734611504, + "per_class": { + "0": { + "precision": 0.8742632612966601, + "recall": 0.89, + "f1-score": 0.8820614469772051, + "support": 500.0 + }, + "1": { + "precision": 0.8879837067209776, + "recall": 0.872, + "f1-score": 0.8799192734611504, + "support": 500.0 + }, + "accuracy": 0.881, + "macro avg": { + "precision": 0.8811234840088189, + "recall": 0.881, + "f1-score": 0.8809903602191778, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8811234840088189, + "recall": 0.881, + "f1-score": 0.8809903602191778, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/025/retfound/metrics_test.csv b/results/downsample/airogs/025/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..6209ab3e1dfbde9802324557de07a22318b79b74 --- /dev/null +++ b/results/downsample/airogs/025/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.30853155441582203,0.881,0.8809903602191778,0.954302,0.119,0.7872963388920835,0.8811234840088189,0.881,0.9559373115608853,0.762 diff --git a/results/downsample/airogs/025/retfound/metrics_val.csv b/results/downsample/airogs/025/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..cf84133c9699854a1f5edbe9f058d9b1eb20df7e --- /dev/null +++ b/results/downsample/airogs/025/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6879917418255526,0.5907407407407408,0.5308058549012978,0.7905418381344307,0.40925925925925927,0.37925930847137335,0.6855487924602396,0.5907407407407408,0.7780161682540836,0.18148148148148147 +0.6493283440085018,0.7518518518518519,0.7514392888058368,0.8193998628257887,0.24814814814814815,0.6019497919751781,0.753535130357932,0.7518518518518519,0.8135057356289473,0.5037037037037038 +0.49212639997987184,0.8074074074074075,0.8068111455108359,0.8955589849108367,0.1925925925925926,0.6763162218849978,0.81125,0.8074074074074074,0.8953235635652623,0.6148148148148148 +0.3842074923655566,0.8388888888888889,0.8388750750235789,0.9241117969821673,0.16111111111111112,0.7224703895152428,0.8390051457975987,0.8388888888888888,0.9251434397934954,0.6777777777777778 +0.3863231715034036,0.8240740740740741,0.8220918255875651,0.9340192043895748,0.17592592592592593,0.6983569014140849,0.8391911099625275,0.8240740740740741,0.9335160479126594,0.6481481481481481 +0.33729660423362956,0.8722222222222222,0.872063837604167,0.9324862825788751,0.12777777777777777,0.773178235992659,0.8740746357132025,0.8722222222222222,0.9312264787181967,0.7444444444444445 +0.29537995247279897,0.8722222222222222,0.872028685357485,0.9496982167352539,0.12777777777777777,0.7731292517006803,0.8744876412867967,0.8722222222222222,0.9509894393868257,0.7444444444444445 +0.3060581701643327,0.8703703703703703,0.8703419131771033,0.9438614540466392,0.12962962962962962,0.7704523218183187,0.8706958097644022,0.8703703703703703,0.9436743351852045,0.7407407407407407 +0.28698721440399394,0.8814814814814815,0.8814814814814815,0.9514334705075446,0.11851851851851852,0.7880794701986755,0.8814814814814815,0.8814814814814815,0.9511779729888934,0.762962962962963 +0.29499851254855886,0.8777777777777778,0.8776418242491657,0.9507716049382715,0.12222222222222222,0.7819858712715855,0.8794642857142857,0.8777777777777778,0.9493791270863817,0.7555555555555555 +0.31338465476737304,0.8777777777777778,0.8775358733300347,0.9493278463648834,0.12222222222222222,0.7818360333824251,0.8807864609258338,0.8777777777777778,0.9482601799549586,0.7555555555555555 +0.29561132587054195,0.8777777777777778,0.8777358490566038,0.9515706447187929,0.12222222222222222,0.7821188878235859,0.8782967032967033,0.8777777777777778,0.9508431963016859,0.7555555555555555 +0.2779875792124692,0.8888888888888888,0.8887652947719689,0.9561556927297667,0.1111111111111111,0.7998198378540686,0.890625,0.8888888888888888,0.9541770778104626,0.7777777777777778 +0.28465022059047923,0.8777777777777778,0.8775745750834719,0.9575994513031549,0.12222222222222222,0.7818907599685891,0.8803026955368979,0.8777777777777778,0.9571678927325629,0.7555555555555555 +0.28252949390341253,0.8759259259259259,0.8757379988253842,0.9564780521262004,0.12407407407407407,0.778977630921067,0.8782138864737299,0.875925925925926,0.9568811800133976,0.7518518518518519 +0.29472022810403037,0.8722222222222222,0.872028685357485,0.955281207133059,0.12777777777777777,0.7731292517006803,0.8744876412867967,0.8722222222222222,0.9553212817746658,0.7444444444444445 +0.3056120175649138,0.8888888888888888,0.8885893485915493,0.9565294924554184,0.1111111111111111,0.7995634938310098,0.8931166454046259,0.8888888888888888,0.9571199714055642,0.7777777777777778 +0.28048206361777644,0.8907407407407407,0.8907103825136612,0.9599451303155008,0.10925925925925926,0.8029605263157895,0.8911753800519095,0.8907407407407408,0.961358179663048,0.7814814814814814 +0.2980305048472741,0.8907407407407407,0.8906563706563706,0.9599931412894376,0.10925925925925926,0.8028813975279996,0.8919504643962848,0.8907407407407407,0.961472259694544,0.7814814814814814 +0.32333332913763385,0.8740740740740741,0.8737814686997292,0.95940329218107,0.1259259259259259,0.7759059915227595,0.8775753212228622,0.8740740740740741,0.9611763765061714,0.7481481481481482 +0.2785748817464885,0.8814814814814815,0.8814749780509219,0.9596913580246913,0.11851851851851852,0.7880701754385965,0.8815652269359531,0.8814814814814815,0.9613264862537925,0.762962962962963 +0.2948901925016852,0.8740740740740741,0.873864713722365,0.9611454046639232,0.1259259259259259,0.7760225364320927,0.8765742377375165,0.8740740740740741,0.9627241181422468,0.7481481481481482 +0.25946737299947176,0.8870370370370371,0.8870335504182227,0.9617798353909465,0.11296296296296296,0.7969999335533455,0.8870848252870724,0.8870370370370371,0.9630748031054641,0.7740740740740741 +0.28577351131859946,0.8777777777777778,0.8776955704108386,0.9611111111111111,0.12222222222222222,0.7820619006102877,0.8787962147887324,0.8777777777777778,0.962467929427453,0.7555555555555555 +0.2844938118668163,0.8888888888888888,0.8888873647100783,0.9607818930041153,0.1111111111111111,0.7999977777530862,0.8889102282704127,0.8888888888888888,0.961762380623068,0.7777777777777778 +0.28866353192750144,0.8870370370370371,0.8870273514533139,0.9610356652949246,0.11296296296296296,0.7969909407048088,0.8871698113207547,0.8870370370370371,0.9618688630894078,0.7740740740740741 +0.2906857409021434,0.8888888888888888,0.8888751697740461,0.9613545953360768,0.1111111111111111,0.7999799979998,0.8890810276679841,0.8888888888888888,0.9623354558658312,0.7777777777777778 +0.28937938414952336,0.8888888888888888,0.8888751697740461,0.9615329218106996,0.1111111111111111,0.7999799979998,0.8890810276679841,0.8888888888888888,0.9623972205237801,0.7777777777777778 +0.2899962359053247,0.8814814814814815,0.8814798556907502,0.9615672153635115,0.11851851851851852,0.7880771465850904,0.8815024143985952,0.8814814814814815,0.962167544378913,0.762962962962963 +0.2906530749271898,0.8814814814814815,0.8814798556907502,0.9616186556927298,0.11851851851851852,0.7880771465850904,0.8815024143985952,0.8814814814814815,0.9623103858595858,0.762962962962963 diff --git a/results/downsample/airogs/025/retfound/pr.png b/results/downsample/airogs/025/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..368e1af4998920b7e9e5b7b00dae3d129934fb4f --- /dev/null +++ b/results/downsample/airogs/025/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c09cd8801ffc20178732b02abbdbfce65bf92bbc0e5fa4854d19b2ef023ade7 +size 45233 diff --git a/results/downsample/airogs/025/retfound/roc.png b/results/downsample/airogs/025/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..955b777990f45f291942a01adfc858a3c65e136a --- /dev/null +++ b/results/downsample/airogs/025/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:958a8ee9f07a44db1df16759fb150ab50813d9d3a1ab4b37f8dbb87b6d32104b +size 62001 diff --git a/results/downsample/airogs/025/retfound/test_pred.npz b/results/downsample/airogs/025/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e1a18832517979d993422c58df4f0be4ea1f255a --- /dev/null +++ b/results/downsample/airogs/025/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3f09dafde525693baaf36e2ba26cf2d528086b5584151825efb9623c69653b4e +size 12510 diff --git a/results/downsample/airogs/025/retfound/train.log b/results/downsample/airogs/025/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..76b8cc6da06b9b9ebd73a473958af5bb434f05cd --- /dev/null +++ b/results/downsample/airogs/025/retfound/train.log @@ -0,0 +1,536 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:50:11.215186010 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:50:12.350568] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:50:12.350802] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/airogs_25', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/025', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:50:15.254812] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:50:16.923495] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:50:25.292370] Sampler_train = +[14:50:25.356736] len of train_set: 1248 +[14:50:26.063970] [Adaptation] Full fine-tuning: training all parameters. +[14:50:26.065284] number of trainable params (M): 303.30 +[14:50:26.065374] base lr: 5.00e-03 +[14:50:26.065438] actual lr: 6.25e-04 +[14:50:26.065508] accumulate grad iterations: 1 +[14:50:26.065572] effective batch size: 32 +[14:50:26.068547] criterion = CrossEntropyLoss() +[14:50:26.068643] Start training for 30 epochs +[14:50:26.070890] log_dir: ./output_logs/retfound +[14:50:28.492567] Epoch: [0] [ 0/39] eta: 0:01:34 lr: 0.000000 loss: 0.6933 (0.6933) time: 2.4207 data: 1.0496 max mem: 7340 +[14:50:29.867099] Epoch: [0] [20/39] eta: 0:00:03 lr: 0.000032 loss: 0.6929 (0.6931) time: 0.0687 data: 0.0001 max mem: 9669 +[14:50:31.787319] Epoch: [0] [38/39] eta: 0:00:00 lr: 0.000061 loss: 0.6928 (0.6928) time: 0.1025 data: 0.0001 max mem: 9669 +[14:50:31.856381] Epoch: [0] Total time: 0:00:05 (0.1483 s / it) +[14:50:31.862245] Averaged stats: lr: 0.000061 loss: 0.6928 (0.6928) +[14:50:32.738982] val: [ 0/17] eta: 0:00:14 loss: 0.6982 (0.6982) time: 0.8575 data: 0.8100 max mem: 9669 +[14:50:33.141310] val: [10/17] eta: 0:00:00 loss: 0.6992 (0.6946) time: 0.1145 data: 0.0798 max mem: 9669 +[14:50:33.693300] val: [16/17] eta: 0:00:00 loss: 0.6861 (0.6880) time: 0.1065 data: 0.0517 max mem: 9669 +[14:50:33.795563] val: Total time: 0:00:01 (0.1126 s / it) +[14:50:33.821688] val loss: 0.6879917418255526 +[14:50:33.821915] Accuracy: 0.5907, F1 Score: 0.5308, ROC AUC: 0.7905, Hamming Loss: 0.4093, + Jaccard Score: 0.3793, Precision: 0.6855, Recall: 0.5907, + Average Precision: 0.7780, Kappa: 0.1815, Score: 0.5009 +[14:50:35.649497] Best epoch = 0, Best score = 0.5009 +[14:50:35.789368] log_dir: ./output_logs/retfound +[14:50:36.836609] Epoch: [1] [ 0/39] eta: 0:00:40 lr: 0.000063 loss: 0.6990 (0.6990) time: 1.0462 data: 0.9420 max mem: 9669 +[14:50:38.160370] Epoch: [1] [20/39] eta: 0:00:02 lr: 0.000095 loss: 0.6895 (0.6887) time: 0.0661 data: 0.0001 max mem: 9669 +[14:50:39.386845] Epoch: [1] [38/39] eta: 0:00:00 lr: 0.000123 loss: 0.6812 (0.6863) time: 0.0681 data: 0.0001 max mem: 9669 +[14:50:39.472729] Epoch: [1] Total time: 0:00:03 (0.0944 s / it) +[14:50:39.482688] Averaged stats: lr: 0.000123 loss: 0.6812 (0.6863) +[14:50:40.152718] val: [ 0/17] eta: 0:00:11 loss: 0.6445 (0.6445) time: 0.6529 data: 0.6163 max mem: 9669 +[14:50:40.489437] val: [10/17] eta: 0:00:00 loss: 0.6552 (0.6550) time: 0.0899 data: 0.0562 max mem: 9669 +[14:50:40.686374] val: [16/17] eta: 0:00:00 loss: 0.6470 (0.6493) time: 0.0697 data: 0.0364 max mem: 9669 +[14:50:40.769871] val: Total time: 0:00:01 (0.0747 s / it) +[14:50:40.784612] val loss: 0.6493283440085018 +[14:50:40.784790] Accuracy: 0.7519, F1 Score: 0.7514, ROC AUC: 0.8194, Hamming Loss: 0.2481, + Jaccard Score: 0.6019, Precision: 0.7535, Recall: 0.7519, + Average Precision: 0.8135, Kappa: 0.5037, Score: 0.6915 +[14:50:42.515017] Best epoch = 1, Best score = 0.6915 +[14:50:42.624393] log_dir: ./output_logs/retfound +[14:50:43.672043] Epoch: [2] [ 0/39] eta: 0:00:40 lr: 0.000125 loss: 0.7187 (0.7187) time: 1.0464 data: 0.9080 max mem: 9669 +[14:50:46.168952] Epoch: [2] [20/39] eta: 0:00:03 lr: 0.000157 loss: 0.6608 (0.6635) time: 0.1248 data: 0.0002 max mem: 9669 +[14:50:47.343681] Epoch: [2] [38/39] eta: 0:00:00 lr: 0.000186 loss: 0.6343 (0.6546) time: 0.0652 data: 0.0001 max mem: 9669 +[14:50:47.438613] Epoch: [2] Total time: 0:00:04 (0.1234 s / it) +[14:50:47.439458] Averaged stats: lr: 0.000186 loss: 0.6343 (0.6546) +[14:50:48.440408] val: [ 0/17] eta: 0:00:16 loss: 0.5074 (0.5074) time: 0.9762 data: 0.9587 max mem: 9669 +[14:50:48.960217] val: [10/17] eta: 0:00:00 loss: 0.5373 (0.5258) time: 0.1359 data: 0.1187 max mem: 9669 +[14:50:49.163478] val: [16/17] eta: 0:00:00 loss: 0.4876 (0.4921) time: 0.0999 data: 0.0768 max mem: 9669 +[14:50:49.240668] val: Total time: 0:00:01 (0.1045 s / it) +[14:50:49.256417] val loss: 0.49212639997987184 +[14:50:49.256634] Accuracy: 0.8074, F1 Score: 0.8068, ROC AUC: 0.8956, Hamming Loss: 0.1926, + Jaccard Score: 0.6763, Precision: 0.8113, Recall: 0.8074, + Average Precision: 0.8953, Kappa: 0.6148, Score: 0.7724 +[14:50:51.374989] Best epoch = 2, Best score = 0.7724 +[14:50:51.442867] log_dir: ./output_logs/retfound +[14:50:52.277934] Epoch: [3] [ 0/39] eta: 0:00:32 lr: 0.000188 loss: 0.6224 (0.6224) time: 0.8340 data: 0.6972 max mem: 9669 +[14:50:55.069896] Epoch: [3] [20/39] eta: 0:00:03 lr: 0.000220 loss: 0.6205 (0.6096) time: 0.1395 data: 0.0002 max mem: 9669 +[14:50:56.654242] Epoch: [3] [38/39] eta: 0:00:00 lr: 0.000248 loss: 0.6058 (0.6096) time: 0.0932 data: 0.0001 max mem: 9669 +[14:50:56.737041] Epoch: [3] Total time: 0:00:05 (0.1357 s / it) +[14:50:56.737845] Averaged stats: lr: 0.000248 loss: 0.6058 (0.6096) +[14:50:57.478196] val: [ 0/17] eta: 0:00:12 loss: 0.3311 (0.3311) time: 0.7070 data: 0.6884 max mem: 9669 +[14:50:57.658461] val: [10/17] eta: 0:00:00 loss: 0.3657 (0.3917) time: 0.0806 data: 0.0648 max mem: 9669 +[14:50:57.801299] val: [16/17] eta: 0:00:00 loss: 0.3657 (0.3842) time: 0.0605 data: 0.0449 max mem: 9669 +[14:50:57.884813] val: Total time: 0:00:01 (0.0655 s / it) +[14:50:57.900654] val loss: 0.3842074923655566 +[14:50:57.900843] Accuracy: 0.8389, F1 Score: 0.8389, ROC AUC: 0.9241, Hamming Loss: 0.1611, + Jaccard Score: 0.7225, Precision: 0.8390, Recall: 0.8389, + Average Precision: 0.9251, Kappa: 0.6778, Score: 0.8136 +[14:50:59.756195] Best epoch = 3, Best score = 0.8136 +[14:50:59.824802] log_dir: ./output_logs/retfound +[14:51:00.878134] Epoch: [4] [ 0/39] eta: 0:00:41 lr: 0.000250 loss: 0.5630 (0.5630) time: 1.0522 data: 0.9145 max mem: 9669 +[14:51:03.674026] Epoch: [4] [20/39] eta: 0:00:03 lr: 0.000282 loss: 0.5983 (0.6030) time: 0.1398 data: 0.0002 max mem: 9669 +[14:51:06.176841] Epoch: [4] [38/39] eta: 0:00:00 lr: 0.000311 loss: 0.5659 (0.5815) time: 0.1390 data: 0.0001 max mem: 9669 +[14:51:06.253515] Epoch: [4] Total time: 0:00:06 (0.1648 s / it) +[14:51:06.261466] Averaged stats: lr: 0.000311 loss: 0.5659 (0.5815) +[14:51:07.016738] val: [ 0/17] eta: 0:00:12 loss: 0.1708 (0.1708) time: 0.7121 data: 0.6956 max mem: 9669 +[14:51:07.517624] val: [10/17] eta: 0:00:00 loss: 0.2212 (0.3052) time: 0.1102 data: 0.0947 max mem: 9669 +[14:51:07.609475] val: [16/17] eta: 0:00:00 loss: 0.3426 (0.3863) time: 0.0767 data: 0.0613 max mem: 9669 +[14:51:07.692298] val: Total time: 0:00:01 (0.0817 s / it) +[14:51:07.710525] val loss: 0.3863231715034036 +[14:51:07.710705] Accuracy: 0.8241, F1 Score: 0.8221, ROC AUC: 0.9340, Hamming Loss: 0.1759, + Jaccard Score: 0.6984, Precision: 0.8392, Recall: 0.8241, + Average Precision: 0.9335, Kappa: 0.6481, Score: 0.8014 +[14:51:07.756140] Best epoch = 3, Best score = 0.8136 +[14:51:08.105321] log_dir: ./output_logs/retfound +[14:51:08.987109] Epoch: [5] [ 0/39] eta: 0:00:34 lr: 0.000313 loss: 0.5900 (0.5900) time: 0.8808 data: 0.8112 max mem: 9669 +[14:51:11.443762] Epoch: [5] [20/39] eta: 0:00:03 lr: 0.000345 loss: 0.5519 (0.5766) time: 0.1228 data: 0.0001 max mem: 9669 +[14:51:13.954282] Epoch: [5] [38/39] eta: 0:00:00 lr: 0.000373 loss: 0.5345 (0.5631) time: 0.1396 data: 0.0001 max mem: 9669 +[14:51:14.040729] Epoch: [5] Total time: 0:00:05 (0.1522 s / it) +[14:51:14.048973] Averaged stats: lr: 0.000373 loss: 0.5345 (0.5631) +[14:51:15.203379] val: [ 0/17] eta: 0:00:18 loss: 0.4294 (0.4294) time: 1.0865 data: 1.0518 max mem: 9669 +[14:51:15.872762] val: [10/17] eta: 0:00:01 loss: 0.4294 (0.3989) time: 0.1596 data: 0.1252 max mem: 9669 +[14:51:16.077440] val: [16/17] eta: 0:00:00 loss: 0.2884 (0.3373) time: 0.1153 data: 0.0811 max mem: 9669 +[14:51:16.159533] val: Total time: 0:00:02 (0.1202 s / it) +[14:51:16.173355] val loss: 0.33729660423362956 +[14:51:16.173569] Accuracy: 0.8722, F1 Score: 0.8721, ROC AUC: 0.9325, Hamming Loss: 0.1278, + Jaccard Score: 0.7732, Precision: 0.8741, Recall: 0.8722, + Average Precision: 0.9312, Kappa: 0.7444, Score: 0.8497 +[14:51:18.043253] Best epoch = 5, Best score = 0.8497 +[14:51:18.130751] log_dir: ./output_logs/retfound +[14:51:19.402161] Epoch: [6] [ 0/39] eta: 0:00:49 lr: 0.000375 loss: 0.4667 (0.4667) time: 1.2705 data: 1.1977 max mem: 9669 +[14:51:21.551992] Epoch: [6] [20/39] eta: 0:00:03 lr: 0.000407 loss: 0.5314 (0.5523) time: 0.1075 data: 0.0009 max mem: 9669 +[14:51:24.063343] Epoch: [6] [38/39] eta: 0:00:00 lr: 0.000436 loss: 0.4974 (0.5340) time: 0.1391 data: 0.0001 max mem: 9669 +[14:51:24.149657] Epoch: [6] Total time: 0:00:06 (0.1543 s / it) +[14:51:24.158221] Averaged stats: lr: 0.000436 loss: 0.4974 (0.5340) +[14:51:24.726550] val: [ 0/17] eta: 0:00:09 loss: 0.1959 (0.1959) time: 0.5610 data: 0.5349 max mem: 9669 +[14:51:25.421818] val: [10/17] eta: 0:00:00 loss: 0.2802 (0.2742) time: 0.1141 data: 0.0812 max mem: 9669 +[14:51:25.620921] val: [16/17] eta: 0:00:00 loss: 0.2989 (0.2954) time: 0.0855 data: 0.0526 max mem: 9669 +[14:51:25.710166] val: Total time: 0:00:01 (0.0909 s / it) +[14:51:25.724481] val loss: 0.29537995247279897 +[14:51:25.724662] Accuracy: 0.8722, F1 Score: 0.8720, ROC AUC: 0.9497, Hamming Loss: 0.1278, + Jaccard Score: 0.7731, Precision: 0.8745, Recall: 0.8722, + Average Precision: 0.9510, Kappa: 0.7444, Score: 0.8554 +[14:51:27.750479] Best epoch = 6, Best score = 0.8554 +[14:51:27.847790] log_dir: ./output_logs/retfound +[14:51:28.771245] Epoch: [7] [ 0/39] eta: 0:00:35 lr: 0.000438 loss: 0.5884 (0.5884) time: 0.9226 data: 0.8531 max mem: 9669 +[14:51:30.095243] Epoch: [7] [20/39] eta: 0:00:02 lr: 0.000470 loss: 0.5219 (0.5192) time: 0.0662 data: 0.0001 max mem: 9669 +[14:51:32.387375] Epoch: [7] [38/39] eta: 0:00:00 lr: 0.000498 loss: 0.5605 (0.5374) time: 0.1211 data: 0.0001 max mem: 9669 +[14:51:32.471474] Epoch: [7] Total time: 0:00:04 (0.1186 s / it) +[14:51:32.480963] Averaged stats: lr: 0.000498 loss: 0.5605 (0.5374) +[14:51:33.205702] val: [ 0/17] eta: 0:00:11 loss: 0.1930 (0.1930) time: 0.6790 data: 0.6428 max mem: 9669 +[14:51:33.560726] val: [10/17] eta: 0:00:00 loss: 0.3144 (0.3019) time: 0.0939 data: 0.0597 max mem: 9669 +[14:51:33.757078] val: [16/17] eta: 0:00:00 loss: 0.3144 (0.3061) time: 0.0723 data: 0.0387 max mem: 9669 +[14:51:33.838186] val: Total time: 0:00:01 (0.0772 s / it) +[14:51:33.852073] val loss: 0.3060581701643327 +[14:51:33.852248] Accuracy: 0.8704, F1 Score: 0.8703, ROC AUC: 0.9439, Hamming Loss: 0.1296, + Jaccard Score: 0.7705, Precision: 0.8707, Recall: 0.8704, + Average Precision: 0.9437, Kappa: 0.7407, Score: 0.8516 +[14:51:33.887852] Best epoch = 6, Best score = 0.8554 +[14:51:34.147898] log_dir: ./output_logs/retfound +[14:51:35.112669] Epoch: [8] [ 0/39] eta: 0:00:37 lr: 0.000500 loss: 0.6171 (0.6171) time: 0.9639 data: 0.8286 max mem: 9669 +[14:51:37.872007] Epoch: [8] [20/39] eta: 0:00:03 lr: 0.000532 loss: 0.5403 (0.5717) time: 0.1379 data: 0.0002 max mem: 9669 +[14:51:39.588815] Epoch: [8] [38/39] eta: 0:00:00 lr: 0.000561 loss: 0.5439 (0.5701) time: 0.1000 data: 0.0001 max mem: 9669 +[14:51:39.671783] Epoch: [8] Total time: 0:00:05 (0.1416 s / it) +[14:51:39.672640] Averaged stats: lr: 0.000561 loss: 0.5439 (0.5701) +[14:51:40.286162] val: [ 0/17] eta: 0:00:10 loss: 0.2300 (0.2300) time: 0.5893 data: 0.5716 max mem: 9669 +[14:51:40.442304] val: [10/17] eta: 0:00:00 loss: 0.3496 (0.3092) time: 0.0677 data: 0.0521 max mem: 9669 +[14:51:40.534125] val: [16/17] eta: 0:00:00 loss: 0.2937 (0.2870) time: 0.0492 data: 0.0337 max mem: 9669 +[14:51:40.620027] val: Total time: 0:00:00 (0.0543 s / it) +[14:51:40.633941] val loss: 0.28698721440399394 +[14:51:40.634130] Accuracy: 0.8815, F1 Score: 0.8815, ROC AUC: 0.9514, Hamming Loss: 0.1185, + Jaccard Score: 0.7881, Precision: 0.8815, Recall: 0.8815, + Average Precision: 0.9512, Kappa: 0.7630, Score: 0.8653 +[14:51:42.503725] Best epoch = 8, Best score = 0.8653 +[14:51:42.570550] log_dir: ./output_logs/retfound +[14:51:43.308084] Epoch: [9] [ 0/39] eta: 0:00:28 lr: 0.000562 loss: 0.4692 (0.4692) time: 0.7367 data: 0.5938 max mem: 9669 +[14:51:46.096449] Epoch: [9] [20/39] eta: 0:00:03 lr: 0.000595 loss: 0.5408 (0.5350) time: 0.1394 data: 0.0001 max mem: 9669 +[14:51:48.602870] Epoch: [9] [38/39] eta: 0:00:00 lr: 0.000623 loss: 0.5292 (0.5339) time: 0.1390 data: 0.0001 max mem: 9669 +[14:51:48.686471] Epoch: [9] Total time: 0:00:06 (0.1568 s / it) +[14:51:48.695248] Averaged stats: lr: 0.000623 loss: 0.5292 (0.5339) +[14:51:49.332990] val: [ 0/17] eta: 0:00:10 loss: 0.1691 (0.1691) time: 0.6233 data: 0.5879 max mem: 9669 +[14:51:49.696702] val: [10/17] eta: 0:00:00 loss: 0.2465 (0.2748) time: 0.0897 data: 0.0558 max mem: 9669 +[14:51:49.893594] val: [16/17] eta: 0:00:00 loss: 0.2896 (0.2950) time: 0.0696 data: 0.0361 max mem: 9669 +[14:51:49.978113] val: Total time: 0:00:01 (0.0747 s / it) +[14:51:49.992573] val loss: 0.29499851254855886 +[14:51:49.992780] Accuracy: 0.8778, F1 Score: 0.8776, ROC AUC: 0.9508, Hamming Loss: 0.1222, + Jaccard Score: 0.7820, Precision: 0.8795, Recall: 0.8778, + Average Precision: 0.9494, Kappa: 0.7556, Score: 0.8613 +[14:51:50.031973] Best epoch = 8, Best score = 0.8653 +[14:51:50.321176] log_dir: ./output_logs/retfound +[14:51:51.121209] Epoch: [10] [ 0/39] eta: 0:00:31 lr: 0.000625 loss: 0.5395 (0.5395) time: 0.7989 data: 0.7276 max mem: 9669 +[14:51:52.437278] Epoch: [10] [20/39] eta: 0:00:01 lr: 0.000624 loss: 0.5553 (0.5468) time: 0.0658 data: 0.0001 max mem: 9669 +[14:51:54.524171] Epoch: [10] [38/39] eta: 0:00:00 lr: 0.000621 loss: 0.5261 (0.5372) time: 0.1109 data: 0.0001 max mem: 9669 +[14:51:54.603262] Epoch: [10] Total time: 0:00:04 (0.1098 s / it) +[14:51:54.611433] Averaged stats: lr: 0.000621 loss: 0.5261 (0.5372) +[14:51:55.149488] val: [ 0/17] eta: 0:00:08 loss: 0.1817 (0.1817) time: 0.5264 data: 0.4919 max mem: 9669 +[14:51:55.513322] val: [10/17] eta: 0:00:00 loss: 0.2945 (0.2994) time: 0.0809 data: 0.0466 max mem: 9669 +[14:51:55.717862] val: [16/17] eta: 0:00:00 loss: 0.2945 (0.3134) time: 0.0643 data: 0.0302 max mem: 9669 +[14:51:55.801982] val: Total time: 0:00:01 (0.0694 s / it) +[14:51:55.815801] val loss: 0.31338465476737304 +[14:51:55.816016] Accuracy: 0.8778, F1 Score: 0.8775, ROC AUC: 0.9493, Hamming Loss: 0.1222, + Jaccard Score: 0.7818, Precision: 0.8808, Recall: 0.8778, + Average Precision: 0.9483, Kappa: 0.7556, Score: 0.8608 +[14:51:55.855496] Best epoch = 8, Best score = 0.8653 +[14:51:56.125095] log_dir: ./output_logs/retfound +[14:51:56.942592] Epoch: [11] [ 0/39] eta: 0:00:31 lr: 0.000621 loss: 0.7767 (0.7767) time: 0.8165 data: 0.6791 max mem: 9669 +[14:51:59.732377] Epoch: [11] [20/39] eta: 0:00:03 lr: 0.000616 loss: 0.5540 (0.5749) time: 0.1394 data: 0.0002 max mem: 9669 +[14:52:02.015353] Epoch: [11] [38/39] eta: 0:00:00 lr: 0.000610 loss: 0.5305 (0.5517) time: 0.1281 data: 0.0001 max mem: 9669 +[14:52:02.098238] Epoch: [11] Total time: 0:00:05 (0.1532 s / it) +[14:52:02.099019] Averaged stats: lr: 0.000610 loss: 0.5305 (0.5517) +[14:52:02.885247] val: [ 0/17] eta: 0:00:13 loss: 0.3476 (0.3476) time: 0.7702 data: 0.7526 max mem: 9669 +[14:52:03.041020] val: [10/17] eta: 0:00:00 loss: 0.3476 (0.3392) time: 0.0841 data: 0.0685 max mem: 9669 +[14:52:03.178033] val: [16/17] eta: 0:00:00 loss: 0.2886 (0.2956) time: 0.0625 data: 0.0470 max mem: 9669 +[14:52:03.263540] val: Total time: 0:00:01 (0.0676 s / it) +[14:52:03.281579] val loss: 0.29561132587054195 +[14:52:03.281757] Accuracy: 0.8778, F1 Score: 0.8777, ROC AUC: 0.9516, Hamming Loss: 0.1222, + Jaccard Score: 0.7821, Precision: 0.8783, Recall: 0.8778, + Average Precision: 0.9508, Kappa: 0.7556, Score: 0.8616 +[14:52:03.321427] Best epoch = 8, Best score = 0.8653 +[14:52:03.588901] log_dir: ./output_logs/retfound +[14:52:04.319926] Epoch: [12] [ 0/39] eta: 0:00:28 lr: 0.000610 loss: 0.4776 (0.4776) time: 0.7301 data: 0.6577 max mem: 9669 +[14:52:06.832687] Epoch: [12] [20/39] eta: 0:00:02 lr: 0.000601 loss: 0.4861 (0.4961) time: 0.1256 data: 0.0001 max mem: 9669 +[14:52:09.338409] Epoch: [12] [38/39] eta: 0:00:00 lr: 0.000592 loss: 0.5303 (0.5101) time: 0.1388 data: 0.0001 max mem: 9669 +[14:52:09.422287] Epoch: [12] Total time: 0:00:05 (0.1496 s / it) +[14:52:09.431363] Averaged stats: lr: 0.000592 loss: 0.5303 (0.5101) +[14:52:10.124284] val: [ 0/17] eta: 0:00:11 loss: 0.4053 (0.4053) time: 0.6700 data: 0.6342 max mem: 9669 +[14:52:10.530189] val: [10/17] eta: 0:00:00 loss: 0.3083 (0.3445) time: 0.0978 data: 0.0642 max mem: 9669 +[14:52:10.726670] val: [16/17] eta: 0:00:00 loss: 0.2044 (0.2780) time: 0.0748 data: 0.0416 max mem: 9669 +[14:52:10.811929] val: Total time: 0:00:01 (0.0799 s / it) +[14:52:10.825807] val loss: 0.2779875792124692 +[14:52:10.825984] Accuracy: 0.8889, F1 Score: 0.8888, ROC AUC: 0.9562, Hamming Loss: 0.1111, + Jaccard Score: 0.7998, Precision: 0.8906, Recall: 0.8889, + Average Precision: 0.9542, Kappa: 0.7778, Score: 0.8742 +[14:52:12.542319] Best epoch = 12, Best score = 0.8742 +[14:52:12.604997] log_dir: ./output_logs/retfound +[14:52:13.339619] Epoch: [13] [ 0/39] eta: 0:00:28 lr: 0.000591 loss: 0.6335 (0.6335) time: 0.7337 data: 0.6606 max mem: 9669 +[14:52:14.656399] Epoch: [13] [20/39] eta: 0:00:01 lr: 0.000579 loss: 0.4932 (0.5111) time: 0.0658 data: 0.0001 max mem: 9669 +[14:52:16.711728] Epoch: [13] [38/39] eta: 0:00:00 lr: 0.000566 loss: 0.4362 (0.4992) time: 0.1093 data: 0.0001 max mem: 9669 +[14:52:16.790204] Epoch: [13] Total time: 0:00:04 (0.1073 s / it) +[14:52:16.798532] Averaged stats: lr: 0.000566 loss: 0.4362 (0.4992) +[14:52:17.400514] val: [ 0/17] eta: 0:00:09 loss: 0.1810 (0.1810) time: 0.5822 data: 0.5479 max mem: 9669 +[14:52:17.743558] val: [10/17] eta: 0:00:00 loss: 0.2786 (0.2702) time: 0.0841 data: 0.0499 max mem: 9669 +[14:52:17.946951] val: [16/17] eta: 0:00:00 loss: 0.2872 (0.2847) time: 0.0663 data: 0.0323 max mem: 9669 +[14:52:18.027713] val: Total time: 0:00:01 (0.0712 s / it) +[14:52:18.041543] val loss: 0.28465022059047923 +[14:52:18.041743] Accuracy: 0.8778, F1 Score: 0.8776, ROC AUC: 0.9576, Hamming Loss: 0.1222, + Jaccard Score: 0.7819, Precision: 0.8803, Recall: 0.8778, + Average Precision: 0.9572, Kappa: 0.7556, Score: 0.8636 +[14:52:18.095290] Best epoch = 12, Best score = 0.8742 +[14:52:18.345960] log_dir: ./output_logs/retfound +[14:52:19.187029] Epoch: [14] [ 0/39] eta: 0:00:32 lr: 0.000565 loss: 0.6196 (0.6196) time: 0.8397 data: 0.6913 max mem: 9669 +[14:52:21.965642] Epoch: [14] [20/39] eta: 0:00:03 lr: 0.000550 loss: 0.4751 (0.5105) time: 0.1389 data: 0.0002 max mem: 9669 +[14:52:24.221253] Epoch: [14] [38/39] eta: 0:00:00 lr: 0.000535 loss: 0.4896 (0.5006) time: 0.1265 data: 0.0001 max mem: 9669 +[14:52:24.313279] Epoch: [14] Total time: 0:00:05 (0.1530 s / it) +[14:52:24.314144] Averaged stats: lr: 0.000535 loss: 0.4896 (0.5006) +[14:52:24.880538] val: [ 0/17] eta: 0:00:09 loss: 0.1679 (0.1679) time: 0.5594 data: 0.5424 max mem: 9669 +[14:52:25.035496] val: [10/17] eta: 0:00:00 loss: 0.2642 (0.2776) time: 0.0649 data: 0.0494 max mem: 9669 +[14:52:25.127191] val: [16/17] eta: 0:00:00 loss: 0.2642 (0.2825) time: 0.0473 data: 0.0320 max mem: 9669 +[14:52:25.203290] val: Total time: 0:00:00 (0.0519 s / it) +[14:52:25.217723] val loss: 0.28252949390341253 +[14:52:25.217910] Accuracy: 0.8759, F1 Score: 0.8757, ROC AUC: 0.9565, Hamming Loss: 0.1241, + Jaccard Score: 0.7790, Precision: 0.8782, Recall: 0.8759, + Average Precision: 0.9569, Kappa: 0.7519, Score: 0.8614 +[14:52:25.260602] Best epoch = 12, Best score = 0.8742 +[14:52:25.512960] log_dir: ./output_logs/retfound +[14:52:26.294673] Epoch: [15] [ 0/39] eta: 0:00:30 lr: 0.000534 loss: 0.4916 (0.4916) time: 0.7809 data: 0.7107 max mem: 9669 +[14:52:28.999855] Epoch: [15] [20/39] eta: 0:00:03 lr: 0.000515 loss: 0.4697 (0.4712) time: 0.1352 data: 0.0001 max mem: 9669 +[14:52:31.506683] Epoch: [15] [38/39] eta: 0:00:00 lr: 0.000497 loss: 0.4602 (0.4818) time: 0.1394 data: 0.0001 max mem: 9669 +[14:52:31.592516] Epoch: [15] Total time: 0:00:06 (0.1559 s / it) +[14:52:31.600511] Averaged stats: lr: 0.000497 loss: 0.4602 (0.4818) +[14:52:32.196025] val: [ 0/17] eta: 0:00:09 loss: 0.1762 (0.1762) time: 0.5835 data: 0.5485 max mem: 9669 +[14:52:32.532252] val: [10/17] eta: 0:00:00 loss: 0.2740 (0.2929) time: 0.0836 data: 0.0500 max mem: 9669 +[14:52:32.731343] val: [16/17] eta: 0:00:00 loss: 0.2900 (0.2947) time: 0.0657 data: 0.0324 max mem: 9669 +[14:52:32.813073] val: Total time: 0:00:01 (0.0707 s / it) +[14:52:32.826885] val loss: 0.29472022810403037 +[14:52:32.827129] Accuracy: 0.8722, F1 Score: 0.8720, ROC AUC: 0.9553, Hamming Loss: 0.1278, + Jaccard Score: 0.7731, Precision: 0.8745, Recall: 0.8722, + Average Precision: 0.9553, Kappa: 0.7444, Score: 0.8573 +[14:52:32.871124] Best epoch = 12, Best score = 0.8742 +[14:52:33.131904] log_dir: ./output_logs/retfound +[14:52:33.929279] Epoch: [16] [ 0/39] eta: 0:00:31 lr: 0.000496 loss: 0.5173 (0.5173) time: 0.7965 data: 0.6490 max mem: 9669 +[14:52:35.999239] Epoch: [16] [20/39] eta: 0:00:02 lr: 0.000475 loss: 0.5013 (0.5015) time: 0.1035 data: 0.0001 max mem: 9669 +[14:52:37.181650] Epoch: [16] [38/39] eta: 0:00:00 lr: 0.000456 loss: 0.4838 (0.5030) time: 0.0656 data: 0.0001 max mem: 9669 +[14:52:37.264622] Epoch: [16] Total time: 0:00:04 (0.1060 s / it) +[14:52:37.265429] Averaged stats: lr: 0.000456 loss: 0.4838 (0.5030) +[14:52:37.849870] val: [ 0/17] eta: 0:00:09 loss: 0.4444 (0.4444) time: 0.5591 data: 0.5248 max mem: 9669 +[14:52:38.260710] val: [10/17] eta: 0:00:00 loss: 0.3437 (0.4088) time: 0.0881 data: 0.0547 max mem: 9669 +[14:52:38.457757] val: [16/17] eta: 0:00:00 loss: 0.2361 (0.3056) time: 0.0686 data: 0.0355 max mem: 9669 +[14:52:38.535364] val: Total time: 0:00:01 (0.0732 s / it) +[14:52:38.549177] val loss: 0.3056120175649138 +[14:52:38.549377] Accuracy: 0.8889, F1 Score: 0.8886, ROC AUC: 0.9565, Hamming Loss: 0.1111, + Jaccard Score: 0.7996, Precision: 0.8931, Recall: 0.8889, + Average Precision: 0.9571, Kappa: 0.7778, Score: 0.8743 +[14:52:40.208376] Best epoch = 16, Best score = 0.8743 +[14:52:40.298613] log_dir: ./output_logs/retfound +[14:52:41.167835] Epoch: [17] [ 0/39] eta: 0:00:33 lr: 0.000455 loss: 0.3515 (0.3515) time: 0.8682 data: 0.7315 max mem: 9669 +[14:52:43.954655] Epoch: [17] [20/39] eta: 0:00:03 lr: 0.000432 loss: 0.4930 (0.5020) time: 0.1393 data: 0.0001 max mem: 9669 +[14:52:45.909955] Epoch: [17] [38/39] eta: 0:00:00 lr: 0.000411 loss: 0.4820 (0.4893) time: 0.1119 data: 0.0001 max mem: 9669 +[14:52:45.993031] Epoch: [17] Total time: 0:00:05 (0.1460 s / it) +[14:52:45.993866] Averaged stats: lr: 0.000411 loss: 0.4820 (0.4893) +[14:52:46.522184] val: [ 0/17] eta: 0:00:08 loss: 0.2934 (0.2934) time: 0.5050 data: 0.4880 max mem: 9669 +[14:52:46.769096] val: [10/17] eta: 0:00:00 loss: 0.3269 (0.3327) time: 0.0683 data: 0.0528 max mem: 9669 +[14:52:46.860277] val: [16/17] eta: 0:00:00 loss: 0.2640 (0.2805) time: 0.0495 data: 0.0342 max mem: 9669 +[14:52:46.944194] val: Total time: 0:00:00 (0.0546 s / it) +[14:52:46.958048] val loss: 0.28048206361777644 +[14:52:46.958238] Accuracy: 0.8907, F1 Score: 0.8907, ROC AUC: 0.9599, Hamming Loss: 0.1093, + Jaccard Score: 0.8030, Precision: 0.8912, Recall: 0.8907, + Average Precision: 0.9614, Kappa: 0.7815, Score: 0.8774 +[14:52:48.591189] Best epoch = 17, Best score = 0.8774 +[14:52:48.655630] log_dir: ./output_logs/retfound +[14:52:49.465394] Epoch: [18] [ 0/39] eta: 0:00:31 lr: 0.000409 loss: 0.3480 (0.3480) time: 0.8087 data: 0.6614 max mem: 9669 +[14:52:52.253526] Epoch: [18] [20/39] eta: 0:00:03 lr: 0.000385 loss: 0.4685 (0.4737) time: 0.1394 data: 0.0001 max mem: 9669 +[14:52:54.763229] Epoch: [18] [38/39] eta: 0:00:00 lr: 0.000363 loss: 0.4484 (0.4712) time: 0.1391 data: 0.0001 max mem: 9669 +[14:52:54.846399] Epoch: [18] Total time: 0:00:06 (0.1587 s / it) +[14:52:54.855626] Averaged stats: lr: 0.000363 loss: 0.4484 (0.4712) +[14:52:55.616103] val: [ 0/17] eta: 0:00:12 loss: 0.3733 (0.3733) time: 0.7393 data: 0.7045 max mem: 9669 +[14:52:55.954919] val: [10/17] eta: 0:00:00 loss: 0.3550 (0.3729) time: 0.0979 data: 0.0642 max mem: 9669 +[14:52:56.160707] val: [16/17] eta: 0:00:00 loss: 0.2403 (0.2980) time: 0.0754 data: 0.0416 max mem: 9669 +[14:52:56.243174] val: Total time: 0:00:01 (0.0804 s / it) +[14:52:56.257232] val loss: 0.2980305048472741 +[14:52:56.257430] Accuracy: 0.8907, F1 Score: 0.8907, ROC AUC: 0.9600, Hamming Loss: 0.1093, + Jaccard Score: 0.8029, Precision: 0.8920, Recall: 0.8907, + Average Precision: 0.9615, Kappa: 0.7815, Score: 0.8774 +[14:52:56.302098] Best epoch = 17, Best score = 0.8774 +[14:52:56.548192] log_dir: ./output_logs/retfound +[14:52:57.315817] Epoch: [19] [ 0/39] eta: 0:00:29 lr: 0.000362 loss: 0.6767 (0.6767) time: 0.7667 data: 0.6958 max mem: 9669 +[14:52:58.629245] Epoch: [19] [20/39] eta: 0:00:01 lr: 0.000337 loss: 0.4367 (0.4671) time: 0.0656 data: 0.0001 max mem: 9669 +[14:53:00.986894] Epoch: [19] [38/39] eta: 0:00:00 lr: 0.000314 loss: 0.4540 (0.4570) time: 0.1244 data: 0.0001 max mem: 9669 +[14:53:01.078017] Epoch: [19] Total time: 0:00:04 (0.1161 s / it) +[14:53:01.088122] Averaged stats: lr: 0.000314 loss: 0.4540 (0.4570) +[14:53:01.873159] val: [ 0/17] eta: 0:00:13 loss: 0.2100 (0.2100) time: 0.7720 data: 0.7356 max mem: 9669 +[14:53:02.206993] val: [10/17] eta: 0:00:00 loss: 0.2100 (0.2735) time: 0.1005 data: 0.0670 max mem: 9669 +[14:53:02.403279] val: [16/17] eta: 0:00:00 loss: 0.3133 (0.3233) time: 0.0765 data: 0.0434 max mem: 9669 +[14:53:02.485780] val: Total time: 0:00:01 (0.0815 s / it) +[14:53:02.499735] val loss: 0.32333332913763385 +[14:53:02.500014] Accuracy: 0.8741, F1 Score: 0.8738, ROC AUC: 0.9594, Hamming Loss: 0.1259, + Jaccard Score: 0.7759, Precision: 0.8776, Recall: 0.8741, + Average Precision: 0.9612, Kappa: 0.7481, Score: 0.8604 +[14:53:02.539710] Best epoch = 17, Best score = 0.8774 +[14:53:02.789159] log_dir: ./output_logs/retfound +[14:53:03.615568] Epoch: [20] [ 0/39] eta: 0:00:32 lr: 0.000313 loss: 0.4074 (0.4074) time: 0.8254 data: 0.6755 max mem: 9669 +[14:53:06.409486] Epoch: [20] [20/39] eta: 0:00:03 lr: 0.000288 loss: 0.4564 (0.4611) time: 0.1397 data: 0.0001 max mem: 9669 +[14:53:08.157182] Epoch: [20] [38/39] eta: 0:00:00 lr: 0.000265 loss: 0.4584 (0.4580) time: 0.1013 data: 0.0001 max mem: 9669 +[14:53:08.237012] Epoch: [20] Total time: 0:00:05 (0.1397 s / it) +[14:53:08.237811] Averaged stats: lr: 0.000265 loss: 0.4584 (0.4580) +[14:53:08.814213] val: [ 0/17] eta: 0:00:09 loss: 0.3813 (0.3813) time: 0.5638 data: 0.5466 max mem: 9669 +[14:53:08.969887] val: [10/17] eta: 0:00:00 loss: 0.3276 (0.3064) time: 0.0653 data: 0.0498 max mem: 9669 +[14:53:09.062059] val: [16/17] eta: 0:00:00 loss: 0.2967 (0.2786) time: 0.0477 data: 0.0322 max mem: 9669 +[14:53:09.144744] val: Total time: 0:00:00 (0.0527 s / it) +[14:53:09.165256] val loss: 0.2785748817464885 +[14:53:09.165433] Accuracy: 0.8815, F1 Score: 0.8815, ROC AUC: 0.9597, Hamming Loss: 0.1185, + Jaccard Score: 0.7881, Precision: 0.8816, Recall: 0.8815, + Average Precision: 0.9613, Kappa: 0.7630, Score: 0.8680 +[14:53:09.192954] Best epoch = 17, Best score = 0.8774 +[14:53:09.464832] log_dir: ./output_logs/retfound +[14:53:10.248858] Epoch: [21] [ 0/39] eta: 0:00:30 lr: 0.000264 loss: 0.2669 (0.2669) time: 0.7831 data: 0.6468 max mem: 9669 +[14:53:13.054503] Epoch: [21] [20/39] eta: 0:00:03 lr: 0.000240 loss: 0.4765 (0.4589) time: 0.1402 data: 0.0001 max mem: 9669 +[14:53:15.559551] Epoch: [21] [38/39] eta: 0:00:00 lr: 0.000218 loss: 0.4362 (0.4570) time: 0.1394 data: 0.0001 max mem: 9669 +[14:53:15.638382] Epoch: [21] Total time: 0:00:06 (0.1583 s / it) +[14:53:15.646826] Averaged stats: lr: 0.000218 loss: 0.4362 (0.4570) +[14:53:16.358841] val: [ 0/17] eta: 0:00:11 loss: 0.2345 (0.2345) time: 0.7003 data: 0.6674 max mem: 9669 +[14:53:16.707192] val: [10/17] eta: 0:00:00 loss: 0.2381 (0.2572) time: 0.0953 data: 0.0608 max mem: 9669 +[14:53:16.904308] val: [16/17] eta: 0:00:00 loss: 0.2781 (0.2949) time: 0.0732 data: 0.0394 max mem: 9669 +[14:53:16.988678] val: Total time: 0:00:01 (0.0783 s / it) +[14:53:17.002561] val loss: 0.2948901925016852 +[14:53:17.002784] Accuracy: 0.8741, F1 Score: 0.8739, ROC AUC: 0.9611, Hamming Loss: 0.1259, + Jaccard Score: 0.7760, Precision: 0.8766, Recall: 0.8741, + Average Precision: 0.9627, Kappa: 0.7481, Score: 0.8611 +[14:53:17.041364] Best epoch = 17, Best score = 0.8774 +[14:53:17.292914] log_dir: ./output_logs/retfound +[14:53:18.150254] Epoch: [22] [ 0/39] eta: 0:00:33 lr: 0.000217 loss: 0.4475 (0.4475) time: 0.8564 data: 0.7122 max mem: 9669 +[14:53:19.666101] Epoch: [22] [20/39] eta: 0:00:02 lr: 0.000193 loss: 0.4125 (0.4513) time: 0.0758 data: 0.0001 max mem: 9669 +[14:53:20.844485] Epoch: [22] [38/39] eta: 0:00:00 lr: 0.000172 loss: 0.4503 (0.4564) time: 0.0654 data: 0.0001 max mem: 9669 +[14:53:20.924040] Epoch: [22] Total time: 0:00:03 (0.0931 s / it) +[14:53:20.924808] Averaged stats: lr: 0.000172 loss: 0.4503 (0.4564) +[14:53:21.742745] val: [ 0/17] eta: 0:00:13 loss: 0.3372 (0.3372) time: 0.7941 data: 0.7598 max mem: 9669 +[14:53:22.077324] val: [10/17] eta: 0:00:00 loss: 0.3262 (0.2918) time: 0.1025 data: 0.0692 max mem: 9669 +[14:53:22.276529] val: [16/17] eta: 0:00:00 loss: 0.2525 (0.2595) time: 0.0780 data: 0.0448 max mem: 9669 +[14:53:22.357293] val: Total time: 0:00:01 (0.0829 s / it) +[14:53:22.371116] val loss: 0.25946737299947176 +[14:53:22.371291] Accuracy: 0.8870, F1 Score: 0.8870, ROC AUC: 0.9618, Hamming Loss: 0.1130, + Jaccard Score: 0.7970, Precision: 0.8871, Recall: 0.8870, + Average Precision: 0.9631, Kappa: 0.7741, Score: 0.8743 +[14:53:22.422358] Best epoch = 17, Best score = 0.8774 +[14:53:22.680414] log_dir: ./output_logs/retfound +[14:53:23.523374] Epoch: [23] [ 0/39] eta: 0:00:32 lr: 0.000171 loss: 0.4917 (0.4917) time: 0.8419 data: 0.6992 max mem: 9669 +[14:53:26.315855] Epoch: [23] [20/39] eta: 0:00:03 lr: 0.000149 loss: 0.4580 (0.4523) time: 0.1396 data: 0.0001 max mem: 9669 +[14:53:28.832576] Epoch: [23] [38/39] eta: 0:00:00 lr: 0.000131 loss: 0.4260 (0.4420) time: 0.1400 data: 0.0001 max mem: 9669 +[14:53:28.914115] Epoch: [23] Total time: 0:00:06 (0.1598 s / it) +[14:53:28.923371] Averaged stats: lr: 0.000131 loss: 0.4260 (0.4420) +[14:53:29.623780] val: [ 0/17] eta: 0:00:11 loss: 0.3241 (0.3241) time: 0.6865 data: 0.6693 max mem: 9669 +[14:53:29.787313] val: [10/17] eta: 0:00:00 loss: 0.3325 (0.2925) time: 0.0772 data: 0.0617 max mem: 9669 +[14:53:29.936436] val: [16/17] eta: 0:00:00 loss: 0.3241 (0.2858) time: 0.0587 data: 0.0433 max mem: 9669 +[14:53:30.017630] val: Total time: 0:00:01 (0.0636 s / it) +[14:53:30.031522] val loss: 0.28577351131859946 +[14:53:30.031694] Accuracy: 0.8778, F1 Score: 0.8777, ROC AUC: 0.9611, Hamming Loss: 0.1222, + Jaccard Score: 0.7821, Precision: 0.8788, Recall: 0.8778, + Average Precision: 0.9625, Kappa: 0.7556, Score: 0.8648 +[14:53:30.069982] Best epoch = 17, Best score = 0.8774 +[14:53:30.325813] log_dir: ./output_logs/retfound +[14:53:31.065627] Epoch: [24] [ 0/39] eta: 0:00:28 lr: 0.000130 loss: 0.6041 (0.6041) time: 0.7390 data: 0.6679 max mem: 9669 +[14:53:33.215095] Epoch: [24] [20/39] eta: 0:00:02 lr: 0.000110 loss: 0.4166 (0.4375) time: 0.1074 data: 0.0001 max mem: 9669 +[14:53:35.731164] Epoch: [24] [38/39] eta: 0:00:00 lr: 0.000093 loss: 0.3806 (0.4273) time: 0.1395 data: 0.0001 max mem: 9669 +[14:53:35.815948] Epoch: [24] Total time: 0:00:05 (0.1408 s / it) +[14:53:35.825955] Averaged stats: lr: 0.000093 loss: 0.3806 (0.4273) +[14:53:36.550588] val: [ 0/17] eta: 0:00:12 loss: 0.4340 (0.4340) time: 0.7121 data: 0.6769 max mem: 9669 +[14:53:36.887658] val: [10/17] eta: 0:00:00 loss: 0.3599 (0.3318) time: 0.0953 data: 0.0617 max mem: 9669 +[14:53:37.095756] val: [16/17] eta: 0:00:00 loss: 0.2625 (0.2845) time: 0.0739 data: 0.0399 max mem: 9669 +[14:53:37.178549] val: Total time: 0:00:01 (0.0789 s / it) +[14:53:37.192727] val loss: 0.2844938118668163 +[14:53:37.192950] Accuracy: 0.8889, F1 Score: 0.8889, ROC AUC: 0.9608, Hamming Loss: 0.1111, + Jaccard Score: 0.8000, Precision: 0.8889, Recall: 0.8889, + Average Precision: 0.9618, Kappa: 0.7778, Score: 0.8758 +[14:53:37.232528] Best epoch = 17, Best score = 0.8774 +[14:53:37.488888] log_dir: ./output_logs/retfound +[14:53:38.364436] Epoch: [25] [ 0/39] eta: 0:00:34 lr: 0.000092 loss: 0.5360 (0.5360) time: 0.8746 data: 0.7380 max mem: 9669 +[14:53:40.749472] Epoch: [25] [20/39] eta: 0:00:02 lr: 0.000075 loss: 0.4507 (0.4427) time: 0.1192 data: 0.0001 max mem: 9669 +[14:53:41.928803] Epoch: [25] [38/39] eta: 0:00:00 lr: 0.000061 loss: 0.4309 (0.4310) time: 0.0655 data: 0.0001 max mem: 9669 +[14:53:42.012494] Epoch: [25] Total time: 0:00:04 (0.1160 s / it) +[14:53:42.013263] Averaged stats: lr: 0.000061 loss: 0.4309 (0.4310) +[14:53:42.802660] val: [ 0/17] eta: 0:00:13 loss: 0.4042 (0.4042) time: 0.7768 data: 0.7403 max mem: 9669 +[14:53:43.168860] val: [10/17] eta: 0:00:00 loss: 0.3812 (0.3178) time: 0.1039 data: 0.0699 max mem: 9669 +[14:53:43.365964] val: [16/17] eta: 0:00:00 loss: 0.2804 (0.2887) time: 0.0788 data: 0.0453 max mem: 9669 +[14:53:43.448161] val: Total time: 0:00:01 (0.0837 s / it) +[14:53:43.461914] val loss: 0.28866353192750144 +[14:53:43.462084] Accuracy: 0.8870, F1 Score: 0.8870, ROC AUC: 0.9610, Hamming Loss: 0.1130, + Jaccard Score: 0.7970, Precision: 0.8872, Recall: 0.8870, + Average Precision: 0.9619, Kappa: 0.7741, Score: 0.8740 +[14:53:43.500064] Best epoch = 17, Best score = 0.8774 +[14:53:43.750692] log_dir: ./output_logs/retfound +[14:53:44.601064] Epoch: [26] [ 0/39] eta: 0:00:33 lr: 0.000061 loss: 0.4748 (0.4748) time: 0.8494 data: 0.7121 max mem: 9669 +[14:53:47.405749] Epoch: [26] [20/39] eta: 0:00:03 lr: 0.000047 loss: 0.4162 (0.4293) time: 0.1402 data: 0.0002 max mem: 9669 +[14:53:49.909682] Epoch: [26] [38/39] eta: 0:00:00 lr: 0.000036 loss: 0.3856 (0.4273) time: 0.1394 data: 0.0001 max mem: 9669 +[14:53:49.996068] Epoch: [26] Total time: 0:00:06 (0.1601 s / it) +[14:53:50.005503] Averaged stats: lr: 0.000036 loss: 0.3856 (0.4273) +[14:53:50.728912] val: [ 0/17] eta: 0:00:12 loss: 0.3968 (0.3968) time: 0.7109 data: 0.6755 max mem: 9669 +[14:53:51.077376] val: [10/17] eta: 0:00:00 loss: 0.3837 (0.3141) time: 0.0962 data: 0.0616 max mem: 9669 +[14:53:51.281021] val: [16/17] eta: 0:00:00 loss: 0.3050 (0.2907) time: 0.0742 data: 0.0399 max mem: 9669 +[14:53:51.365843] val: Total time: 0:00:01 (0.0793 s / it) +[14:53:51.380116] val loss: 0.2906857409021434 +[14:53:51.380311] Accuracy: 0.8889, F1 Score: 0.8889, ROC AUC: 0.9614, Hamming Loss: 0.1111, + Jaccard Score: 0.8000, Precision: 0.8891, Recall: 0.8889, + Average Precision: 0.9623, Kappa: 0.7778, Score: 0.8760 +[14:53:51.427736] Best epoch = 17, Best score = 0.8774 +[14:53:51.667460] log_dir: ./output_logs/retfound +[14:53:52.534903] Epoch: [27] [ 0/39] eta: 0:00:33 lr: 0.000035 loss: 0.3469 (0.3469) time: 0.8665 data: 0.7946 max mem: 9669 +[14:53:53.894620] Epoch: [27] [20/39] eta: 0:00:02 lr: 0.000025 loss: 0.4090 (0.4239) time: 0.0679 data: 0.0001 max mem: 9669 +[14:53:56.391849] Epoch: [27] [38/39] eta: 0:00:00 lr: 0.000017 loss: 0.4251 (0.4143) time: 0.1314 data: 0.0001 max mem: 9669 +[14:53:56.475382] Epoch: [27] Total time: 0:00:04 (0.1233 s / it) +[14:53:56.476789] Averaged stats: lr: 0.000017 loss: 0.4251 (0.4143) +[14:53:57.223507] val: [ 0/17] eta: 0:00:12 loss: 0.4120 (0.4120) time: 0.7229 data: 0.6855 max mem: 9669 +[14:53:57.576018] val: [10/17] eta: 0:00:00 loss: 0.3947 (0.3234) time: 0.0977 data: 0.0641 max mem: 9669 +[14:53:57.779595] val: [16/17] eta: 0:00:00 loss: 0.2828 (0.2894) time: 0.0752 data: 0.0415 max mem: 9669 +[14:53:57.857944] val: Total time: 0:00:01 (0.0799 s / it) +[14:53:57.872482] val loss: 0.28937938414952336 +[14:53:57.872684] Accuracy: 0.8889, F1 Score: 0.8889, ROC AUC: 0.9615, Hamming Loss: 0.1111, + Jaccard Score: 0.8000, Precision: 0.8891, Recall: 0.8889, + Average Precision: 0.9624, Kappa: 0.7778, Score: 0.8761 +[14:53:57.916399] Best epoch = 17, Best score = 0.8774 +[14:53:58.203128] log_dir: ./output_logs/retfound +[14:53:59.047750] Epoch: [28] [ 0/39] eta: 0:00:32 lr: 0.000016 loss: 0.4600 (0.4600) time: 0.8435 data: 0.7064 max mem: 9669 +[14:54:01.831498] Epoch: [28] [20/39] eta: 0:00:03 lr: 0.000009 loss: 0.3928 (0.4312) time: 0.1391 data: 0.0001 max mem: 9669 +[14:54:03.419860] Epoch: [28] [38/39] eta: 0:00:00 lr: 0.000005 loss: 0.3613 (0.4152) time: 0.0934 data: 0.0001 max mem: 9669 +[14:54:03.502198] Epoch: [28] Total time: 0:00:05 (0.1359 s / it) +[14:54:03.503031] Averaged stats: lr: 0.000005 loss: 0.3613 (0.4152) +[14:54:04.221624] val: [ 0/17] eta: 0:00:12 loss: 0.4240 (0.4240) time: 0.7062 data: 0.6890 max mem: 9669 +[14:54:04.472994] val: [10/17] eta: 0:00:00 loss: 0.3828 (0.3299) time: 0.0870 data: 0.0715 max mem: 9669 +[14:54:04.565052] val: [16/17] eta: 0:00:00 loss: 0.2712 (0.2900) time: 0.0617 data: 0.0463 max mem: 9669 +[14:54:04.645221] val: Total time: 0:00:01 (0.0665 s / it) +[14:54:04.659347] val loss: 0.2899962359053247 +[14:54:04.659540] Accuracy: 0.8815, F1 Score: 0.8815, ROC AUC: 0.9616, Hamming Loss: 0.1185, + Jaccard Score: 0.7881, Precision: 0.8815, Recall: 0.8815, + Average Precision: 0.9622, Kappa: 0.7630, Score: 0.8687 +[14:54:04.701314] Best epoch = 17, Best score = 0.8774 +[14:54:04.960015] log_dir: ./output_logs/retfound +[14:54:05.827983] Epoch: [29] [ 0/39] eta: 0:00:33 lr: 0.000005 loss: 0.4883 (0.4883) time: 0.8671 data: 0.7206 max mem: 9669 +[14:54:08.617553] Epoch: [29] [20/39] eta: 0:00:03 lr: 0.000002 loss: 0.3964 (0.4090) time: 0.1394 data: 0.0001 max mem: 9669 +[14:54:11.130158] Epoch: [29] [38/39] eta: 0:00:00 lr: 0.000001 loss: 0.4146 (0.4070) time: 0.1395 data: 0.0001 max mem: 9669 +[14:54:11.217550] Epoch: [29] Total time: 0:00:06 (0.1604 s / it) +[14:54:11.226074] Averaged stats: lr: 0.000001 loss: 0.4146 (0.4070) +[14:54:11.979859] val: [ 0/17] eta: 0:00:12 loss: 0.4257 (0.4257) time: 0.7373 data: 0.7021 max mem: 9669 +[14:54:12.325074] val: [10/17] eta: 0:00:00 loss: 0.3829 (0.3309) time: 0.0984 data: 0.0640 max mem: 9669 +[14:54:12.521283] val: [16/17] eta: 0:00:00 loss: 0.2713 (0.2907) time: 0.0751 data: 0.0414 max mem: 9669 +[14:54:12.600665] val: Total time: 0:00:01 (0.0799 s / it) +[14:54:12.614594] val loss: 0.2906530749271898 +[14:54:12.614806] Accuracy: 0.8815, F1 Score: 0.8815, ROC AUC: 0.9616, Hamming Loss: 0.1185, + Jaccard Score: 0.7881, Precision: 0.8815, Recall: 0.8815, + Average Precision: 0.9623, Kappa: 0.7630, Score: 0.8687 +[14:54:12.669728] Best epoch = 17, Best score = 0.8774 +[14:54:15.420806] Test with the best model, epoch = 17: +[14:54:16.167293] test: [ 0/32] eta: 0:00:23 loss: 0.1150 (0.1150) time: 0.7194 data: 0.6844 max mem: 9669 +[14:54:16.510929] test: [10/32] eta: 0:00:02 loss: 0.3645 (0.2926) time: 0.0966 data: 0.0624 max mem: 9669 +[14:54:16.849278] test: [20/32] eta: 0:00:00 loss: 0.3243 (0.2870) time: 0.0340 data: 0.0001 max mem: 9669 +[14:54:17.191179] test: [30/32] eta: 0:00:00 loss: 0.2711 (0.2895) time: 0.0339 data: 0.0001 max mem: 9669 +[14:54:17.323245] test: [31/32] eta: 0:00:00 loss: 0.2711 (0.3085) time: 0.0388 data: 0.0001 max mem: 9669 +[14:54:17.397350] test: Total time: 0:00:01 (0.0609 s / it) +[14:54:17.415977] val loss: 0.30853155441582203 +[14:54:17.416102] Accuracy: 0.8810, F1 Score: 0.8810, ROC AUC: 0.9543, Hamming Loss: 0.1190, + Jaccard Score: 0.7873, Precision: 0.8811, Recall: 0.8810, + Average Precision: 0.9559, Kappa: 0.7620, Score: 0.8658 +[14:54:18.136933] Training time 0:03:52 +[rank0]:[W701 14:54:18.517832940 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/025/retfound acc=0.8810 auroc=0.9543219999999999 f1_macro=0.8810 qwk=0.762 diff --git a/results/downsample/airogs/025/vit/confusion_matrix.png b/results/downsample/airogs/025/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..6ff563772068e48683f31202bfd75c74f3427ed2 --- /dev/null +++ b/results/downsample/airogs/025/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82ef350cc1eb2c974c143aa5f6bf1a19efc36ed74e003572a9e30588eaeaef6d +size 72375 diff --git a/results/downsample/airogs/025/vit/log.csv b/results/downsample/airogs/025/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..089bb7aa793223f3ef94d20d8d62c0cb3631ee4d --- /dev/null +++ b/results/downsample/airogs/025/vit/log.csv @@ -0,0 +1,26 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7405031135207728,0.5981481481481481,0.6554732510288066,0.47258502672539926,7.005232214561775e-08 +1,0.6655782210199457,0.7037037037037037,0.7552949245541838,0.6221339903961556,1.4399643996599205e-07 +2,0.6032569659383673,0.7314814814814815,0.8132853223593964,0.6683726507711222,2.1794055778636637e-07 +3,0.5574340083097157,0.7777777777777778,0.8572427983539095,0.7301757826424522,2.9188467560674067e-07 +4,0.5209903654299284,0.7814814814814814,0.8608230452674897,0.7345567386644101,3.65828793427115e-07 +5,0.47900784329364177,0.7833333333333333,0.8844238683127572,0.7433090095694079,3.684121359194458e-07 +6,0.4595828291616942,0.7925925925925926,0.8910973936899863,0.7562680065661663,3.6421300374415756e-07 +7,0.4272685882292296,0.812962962962963,0.9039780521262003,0.7808783675537082,3.571853722047769e-07 +8,0.39398610121325445,0.8,0.8973799725651579,0.7649027548634487,3.474400712487837e-07 +9,0.39296953458535044,0.8166666666666667,0.9057407407407408,0.7851862872698394,3.351307900938144e-07 +10,0.3591029298932929,0.8333333333333334,0.9079149519890262,0.8026108689326618,3.2045165345689106e-07 +11,0.35168259081087616,0.8148148148148148,0.906687242798354,0.7833699919587532,3.036341600916691e-07 +12,0.34110256088407415,0.812962962962963,0.9026268861454046,0.780349003510147,2.849435319148016e-07 +13,0.34274184390118245,0.8296296296296296,0.9112551440329217,0.7998478805900794,2.64674531297878e-07 +14,0.3153679182654933,0.8388888888888889,0.916426611796982,0.8106211578837735,2.431468124887412e-07 +15,0.32378437801411275,0.8148148148148148,0.9106790123456789,0.7847655790711346,2.2069988047304712e-07 +16,0.30459294350523697,0.7981481481481482,0.8984019204389575,0.7628105747839916,1.9768773677783045e-07 +17,0.31535907638700383,0.8277777777777777,0.9093278463648834,0.7973639085495,1.7447329665595308e-07 +18,0.2961108072807914,0.8240740740740741,0.9104663923182441,0.7942277281867819,1.5142266569577133e-07 +19,0.285906068588558,0.8222222222222222,0.9097873799725651,0.7917552833489016,1.2889936611730676e-07 +20,0.28111287245624944,0.8277777777777777,0.9101783264746227,0.7976220244126947,1.0725860380967895e-07 +21,0.28549550944253016,0.825925925925926,0.9071810699588477,0.7948066395715748,8.684166652204363e-08 +22,0.28284644296294764,0.8222222222222222,0.908244170096022,0.7912408800567204,6.797054155191418e-08 +23,0.2687452851157439,0.8296296296296296,0.9075857338820302,0.79829333178554,5.094283781313315e-08 +24,0.2804084937823446,0.8277777777777777,0.9091152263374487,0.7973785195094193,3.602709236551822e-08 diff --git a/results/downsample/airogs/025/vit/metrics.json b/results/downsample/airogs/025/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..ef209e121287540e03928d18b33ea8a7b63ce170 --- /dev/null +++ b/results/downsample/airogs/025/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.832, + "balanced_accuracy": 0.8320000000000001, + "precision_macro": 0.8343592387888944, + "recall_macro": 0.8320000000000001, + "f1_macro": 0.8317031243112851, + "precision_weighted": 0.8343592387888944, + "recall_weighted": 0.832, + "f1_weighted": 0.8317031243112851, + "cohen_kappa": 0.6639999999999999, + "quadratic_weighted_kappa": 0.6639999999999999, + "mcc": 0.6663550623441318, + "auroc": 0.9150260000000001, + "auprc": 0.9193396212166045, + "sensitivity": 0.874, + "specificity": 0.79, + "precision_pos": 0.8062730627306273, + "f1_pos": 0.8387715930902111, + "per_class": { + "0": { + "precision": 0.8624454148471615, + "recall": 0.79, + "f1-score": 0.824634655532359, + "support": 500.0 + }, + "1": { + "precision": 0.8062730627306273, + "recall": 0.874, + "f1-score": 0.8387715930902111, + "support": 500.0 + }, + "accuracy": 0.832, + "macro avg": { + "precision": 0.8343592387888944, + "recall": 0.8320000000000001, + "f1-score": 0.8317031243112851, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8343592387888944, + "recall": 0.832, + "f1-score": 0.8317031243112851, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/025/vit/pr.png b/results/downsample/airogs/025/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..1530e19ff63e57969f9a70e291f47239363dcc4b --- /dev/null +++ b/results/downsample/airogs/025/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2f9fe1be752c9cbac8202b65856a0aca25e42f3f11f8fe7288a725d544c2090 +size 48774 diff --git a/results/downsample/airogs/025/vit/roc.png b/results/downsample/airogs/025/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..df4743eb394bf4113f2518ff93e7de567951a53d --- /dev/null +++ b/results/downsample/airogs/025/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cceaa66aac1a7cf7c7e254a90b5ed7e16cdd4a3c259e7968e748d1275e22a78b +size 63166 diff --git a/results/downsample/airogs/025/vit/test_pred.npz b/results/downsample/airogs/025/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..1e86b74d15be56c660390a6db07c72aeba126b5b --- /dev/null +++ b/results/downsample/airogs/025/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:371d61b3155097b5504f5d911077b5a27b94d3f8bfbdd6bd294263ea49dcbfc4 +size 16510 diff --git a/results/downsample/airogs/025/vit/train.log b/results/downsample/airogs/025/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..47c0e38120ae8e237b2e02c5dcb008460488c641 --- /dev/null +++ b/results/downsample/airogs/025/vit/train.log @@ -0,0 +1,134 @@ +[vit] train=1250 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7405 val_acc=0.5981 val_auc=0.6555 score=0.4726 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.6656 val_acc=0.7037 val_auc=0.7553 score=0.6221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6033 val_acc=0.7315 val_auc=0.8133 score=0.6684 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.5574 val_acc=0.7778 val_auc=0.8572 score=0.7302 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5210 val_acc=0.7815 val_auc=0.8608 score=0.7346 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.4790 val_acc=0.7833 val_auc=0.8844 score=0.7433 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.4596 val_acc=0.7926 val_auc=0.8911 score=0.7563 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4273 val_acc=0.8130 val_auc=0.9040 score=0.7809 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3940 val_acc=0.8000 val_auc=0.8974 score=0.7649 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.3930 val_acc=0.8167 val_auc=0.9057 score=0.7852 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3591 val_acc=0.8333 val_auc=0.9079 score=0.8026 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3517 val_acc=0.8148 val_auc=0.9067 score=0.7834 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3411 val_acc=0.8130 val_auc=0.9026 score=0.7803 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.3427 val_acc=0.8296 val_auc=0.9113 score=0.7998 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.3154 val_acc=0.8389 val_auc=0.9164 score=0.8106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.3238 val_acc=0.8148 val_auc=0.9107 score=0.7848 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.3046 val_acc=0.7981 val_auc=0.8984 score=0.7628 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.3154 val_acc=0.8278 val_auc=0.9093 score=0.7974 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.2961 val_acc=0.8241 val_auc=0.9105 score=0.7942 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.2859 val_acc=0.8222 val_auc=0.9098 score=0.7918 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.2811 val_acc=0.8278 val_auc=0.9102 score=0.7976 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.2855 val_acc=0.8259 val_auc=0.9072 score=0.7948 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.2828 val_acc=0.8222 val_auc=0.9082 score=0.7912 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.2687 val_acc=0.8296 val_auc=0.9076 score=0.7983 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.2804 val_acc=0.8278 val_auc=0.9091 score=0.7974 +[vit] early stop at ep24 (best ep14 score=0.8106) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=14 best_val_score=0.8106 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/025/vit acc=0.8320 auroc=0.9150260000000001 f1_macro=0.8317 qwk=0.6639999999999999 diff --git a/results/downsample/airogs/050/resnet/confusion_matrix.png b/results/downsample/airogs/050/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..2b2730c2e3ab7798ec7815e8e0b3f1ae5cce215d --- /dev/null +++ b/results/downsample/airogs/050/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0e2bca39d204e8326831369a52b91558492507d9271fea541c7b0aaf026ae81 +size 72417 diff --git a/results/downsample/airogs/050/resnet/log.csv b/results/downsample/airogs/050/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..6b6dcc8bf8c0577ca569afd58026c1fcc93daf64 --- /dev/null +++ b/results/downsample/airogs/050/resnet/log.csv @@ -0,0 +1,26 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6889433249449118,0.6333333333333333,0.6948422496570643,0.5315536967455915,0.0001623931623931624 +1,0.6324744591346154,0.737037037037037,0.8257407407407408,0.6789458074550992,0.00032905982905982907 +2,0.458308430818411,0.7962962962962963,0.8971262002743484,0.7608445272816556,0.0004957264957264957 +3,0.3474043791110699,0.8240740740740741,0.9201851851851851,0.7969413561984643,0.0004983950740923246 +4,0.296918721535267,0.8537037037037037,0.9444650205761317,0.8350308006343582,0.0004934321435602374 +5,0.24417829513549805,0.85,0.9363031550068587,0.8287634312739716,0.0004851772108904175 +6,0.1993879265128038,0.8240740740740741,0.9253017832647462,0.7988671852085364,0.0004737419098860204 +7,0.1856185155801284,0.8629629629629629,0.9430109739368998,0.8437145988048499,0.00045928088337605585 +8,0.16747403068420214,0.8462962962962963,0.9368998628257887,0.8247258486498131,0.0004419896919364133 +9,0.14580732870560426,0.8481481481481481,0.9368038408779149,0.8269986071229426,0.00042210216926855343 +10,0.1213865112990905,0.8629629629629629,0.9402263374485597,0.8428344846168919,0.00039988725999983457 +11,0.13326815945597795,0.8666666666666667,0.9421947873799725,0.8473243705711523,0.0003756453826687972 +12,0.10761880511656785,0.8333333333333334,0.9388545953360767,0.8123475665621583,0.0003497043670797936 +13,0.09573271789420874,0.8648148148148148,0.9504183813443072,0.8482862177592372,0.00032241502096727094 +14,0.059399234369779244,0.8777777777777778,0.9495884773662552,0.860971701345608,0.0002941463859229675 +15,0.06410680601421075,0.8703703703703703,0.9416255144032921,0.850911615768597,0.0002652807467414696 +16,0.05539266488108879,0.8648148148148148,0.9432373113854595,0.8457809819473238,0.00023620846167416815 +17,0.038721806871203274,0.8722222222222222,0.9488957475994514,0.8551803130276691,0.0002073226835035618 +18,0.04192125830703821,0.8759259259259259,0.9443827160493827,0.8573355666892,0.00017901404282632093 +19,0.036214300216390535,0.8722222222222222,0.9473113854595338,0.8545185179867606,0.00015166536544458856 +20,0.04071318047742049,0.8722222222222222,0.9422496570644718,0.8529649495160093,0.000125646495303737 +21,0.031986239581153944,0.8722222222222222,0.9438340192043896,0.8534930702293152,0.00010130929298746907 +22,0.034003888471768454,0.8703703703703703,0.9433813443072703,0.8514494207606583,7.898287740703355e-05 +23,0.021123187807508003,0.8740740740740741,0.9452606310013717,0.8557144580733143,5.896917503255006e-05 +24,0.023565021271889027,0.8629629629629629,0.9433676268861453,0.8439950965253497,4.153883685544921e-05 diff --git a/results/downsample/airogs/050/resnet/metrics.json b/results/downsample/airogs/050/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..fa2467af9c1031e794d820fbd5498199f390ee8e --- /dev/null +++ b/results/downsample/airogs/050/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.879, + "balanced_accuracy": 0.879, + "precision_macro": 0.879001516006064, + "recall_macro": 0.879, + "f1_macro": 0.878999878999879, + "precision_weighted": 0.8790015160060641, + "recall_weighted": 0.879, + "f1_weighted": 0.878999878999879, + "cohen_kappa": 0.758, + "quadratic_weighted_kappa": 0.758, + "mcc": 0.758001516004548, + "auroc": 0.9442860000000001, + "auprc": 0.9391568960913146, + "sensitivity": 0.878, + "specificity": 0.88, + "precision_pos": 0.8797595190380761, + "f1_pos": 0.8788788788788788, + "per_class": { + "0": { + "precision": 0.8782435129740519, + "recall": 0.88, + "f1-score": 0.8791208791208791, + "support": 500.0 + }, + "1": { + "precision": 0.8797595190380761, + "recall": 0.878, + "f1-score": 0.8788788788788788, + "support": 500.0 + }, + "accuracy": 0.879, + "macro avg": { + "precision": 0.879001516006064, + "recall": 0.879, + "f1-score": 0.878999878999879, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8790015160060641, + "recall": 0.879, + "f1-score": 0.878999878999879, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/050/resnet/pr.png b/results/downsample/airogs/050/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..fbc1d9416ca8b8c5e6747fff309612208c71fff7 --- /dev/null +++ b/results/downsample/airogs/050/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e70c24fe7012c137334f659dfc80276b39559cfa5c7a8ef8bbf3a3979682ab10 +size 48668 diff --git a/results/downsample/airogs/050/resnet/roc.png b/results/downsample/airogs/050/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..3180c5ee21acddcfb329b57561ccd9d84a8f5f76 --- /dev/null +++ b/results/downsample/airogs/050/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d62f8501cebf719de36ae9bd269cd1116fd943fe49b9229c6eb9512661e759f4 +size 62150 diff --git a/results/downsample/airogs/050/resnet/test_pred.npz b/results/downsample/airogs/050/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..39074ac891011104502f4ac848cc17ad5a20c9f3 --- /dev/null +++ b/results/downsample/airogs/050/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ece41ed8bf23231cdaeebdd5a2388bcc6a1c5028bd133847fa4c51c5b4bc438 +size 16510 diff --git a/results/downsample/airogs/050/resnet/train.log b/results/downsample/airogs/050/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..33ffda3eb10fc36f6c2f7c6c0261cc4b31d914e8 --- /dev/null +++ b/results/downsample/airogs/050/resnet/train.log @@ -0,0 +1,134 @@ +[resnet] train=2500 val=540 test=1000 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6889 val_acc=0.6333 val_auc=0.6948 score=0.5316 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6325 val_acc=0.7370 val_auc=0.8257 score=0.6789 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.4583 val_acc=0.7963 val_auc=0.8971 score=0.7608 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.3474 val_acc=0.8241 val_auc=0.9202 score=0.7969 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.2969 val_acc=0.8537 val_auc=0.9445 score=0.8350 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.2442 val_acc=0.8500 val_auc=0.9363 score=0.8288 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.1994 val_acc=0.8241 val_auc=0.9253 score=0.7989 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.1856 val_acc=0.8630 val_auc=0.9430 score=0.8437 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.1675 val_acc=0.8463 val_auc=0.9369 score=0.8247 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1458 val_acc=0.8481 val_auc=0.9368 score=0.8270 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1214 val_acc=0.8630 val_auc=0.9402 score=0.8428 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1333 val_acc=0.8667 val_auc=0.9422 score=0.8473 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.1076 val_acc=0.8333 val_auc=0.9389 score=0.8123 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.0957 val_acc=0.8648 val_auc=0.9504 score=0.8483 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0594 val_acc=0.8778 val_auc=0.9496 score=0.8610 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0641 val_acc=0.8704 val_auc=0.9416 score=0.8509 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0554 val_acc=0.8648 val_auc=0.9432 score=0.8458 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0387 val_acc=0.8722 val_auc=0.9489 score=0.8552 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0419 val_acc=0.8759 val_auc=0.9444 score=0.8573 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0362 val_acc=0.8722 val_auc=0.9473 score=0.8545 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0407 val_acc=0.8722 val_auc=0.9422 score=0.8530 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0320 val_acc=0.8722 val_auc=0.9438 score=0.8535 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0340 val_acc=0.8704 val_auc=0.9434 score=0.8514 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0211 val_acc=0.8741 val_auc=0.9453 score=0.8557 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0236 val_acc=0.8630 val_auc=0.9434 score=0.8440 +[resnet] early stop at ep24 (best ep14 score=0.8610) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=14 best_val_score=0.8610 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/050/resnet acc=0.8790 auroc=0.9442860000000001 f1_macro=0.8790 qwk=0.758 diff --git a/results/downsample/airogs/050/retfound/confusion_matrix.png b/results/downsample/airogs/050/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..66e51159cee05d91836568d7c4a52445933d17f2 --- /dev/null +++ b/results/downsample/airogs/050/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:742a359255e94f207b6a51dc11e683bfd05a31981e456c7a2670f166235e8199 +size 73715 diff --git a/results/downsample/airogs/050/retfound/confusion_matrix_test.jpg b/results/downsample/airogs/050/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..215792eacf3dac3fecb21df2a55571dece694959 --- /dev/null +++ b/results/downsample/airogs/050/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e4568d41f227b5ce510c42d267fe0c5ea28e62b9a2f920d3dcd0877cec48755 +size 256341 diff --git a/results/downsample/airogs/050/retfound/log.txt b/results/downsample/airogs/050/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..94188645d9b6756a806c6216c239af0356376ea3 --- /dev/null +++ b/results/downsample/airogs/050/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 3.084935897435896e-05, "train_loss": 0.6916361099634415, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.334935897435898e-05, "train_loss": 0.6667044223883213, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015584935897435895, "train_loss": 0.5854436923296024, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021834935897435898, "train_loss": 0.5640196311168182, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.000280849358974359, "train_loss": 0.5314944065534152, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00034334935897435904, "train_loss": 0.5366591245700152, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040584935897435904, "train_loss": 0.5363526084484198, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.000468349358974359, "train_loss": 0.5425754502797738, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005308493589743588, "train_loss": 0.533293352677272, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000593349358974359, "train_loss": 0.498906624622834, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006237430517033904, "train_loss": 0.5028205674428207, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006161407731110107, "train_loss": 0.4940442893749628, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006010741626180924, "train_loss": 0.4864236276883345, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005789142101781839, "train_loss": 0.4741158577112051, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.000550206567365022, "train_loss": 0.46852376636786336, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005156581116198706, "train_loss": 0.4538854215389643, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0004761195405829266, "train_loss": 0.469632999255107, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004325644250941134, "train_loss": 0.45292354585268557, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.00038606523664790624, "train_loss": 0.469928672680488, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00033776693958581, "train_loss": 0.4448081977092303, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.000288858798274351, "train_loss": 0.4446002382498521, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024054509346943255, "train_loss": 0.42469956018985844, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00019401546892734494, "train_loss": 0.42910681588527483, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015041563842728504, "train_loss": 0.42125809001616943, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011081917449570559, "train_loss": 0.4205930855793831, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.620107348773163e-05, "train_loss": 0.4059202743646426, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.741374794106982e-05, "train_loss": 0.41685784550813526, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5166037350334925e-05, "train_loss": 0.3930121111946228, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 1.0005754186189052e-05, "train_loss": 0.39902972067013764, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.3061949342286973e-06, "train_loss": 0.4167448219198447, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/050/retfound/metrics.json b/results/downsample/airogs/050/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..7fe6cda39c1812e56c4035a5be7820e1271155d5 --- /dev/null +++ b/results/downsample/airogs/050/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.898, + "balanced_accuracy": 0.898, + "precision_macro": 0.8980063681018896, + "recall_macro": 0.898, + "f1_macro": 0.897999591998368, + "precision_weighted": 0.8980063681018896, + "recall_weighted": 0.898, + "f1_weighted": 0.897999591998368, + "cohen_kappa": 0.796, + "quadratic_weighted_kappa": 0.796, + "mcc": 0.796006368076417, + "auroc": 0.9645199999999999, + "auprc": 0.966132104050851, + "sensitivity": 0.896, + "specificity": 0.9, + "precision_pos": 0.8995983935742972, + "f1_pos": 0.8977955911823647, + "per_class": { + "0": { + "precision": 0.896414342629482, + "recall": 0.9, + "f1-score": 0.8982035928143712, + "support": 500.0 + }, + "1": { + "precision": 0.8995983935742972, + "recall": 0.896, + "f1-score": 0.8977955911823647, + "support": 500.0 + }, + "accuracy": 0.898, + "macro avg": { + "precision": 0.8980063681018896, + "recall": 0.898, + "f1-score": 0.897999591998368, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8980063681018896, + "recall": 0.898, + "f1-score": 0.897999591998368, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/050/retfound/metrics_test.csv b/results/downsample/airogs/050/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..1717ead4ea5f55332dba1462c588f437cc911e18 --- /dev/null +++ b/results/downsample/airogs/050/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.27743660961277783,0.898,0.897999591998368,0.9645729999999999,0.102,0.8148814229249012,0.8980063681018896,0.898,0.9641816623474131,0.796 diff --git a/results/downsample/airogs/050/retfound/metrics_val.csv b/results/downsample/airogs/050/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..3ad6df944dc7ee0558d09df8335852b2ad0d5b3b --- /dev/null +++ b/results/downsample/airogs/050/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6806979144320768,0.5481481481481482,0.43222559293987867,0.789866255144033,0.45185185185185184,0.31079406254503533,0.7626459143968871,0.5481481481481482,0.7875259949728106,0.09629629629629632 +0.489725402172874,0.8277777777777777,0.8276779066056803,0.8932304526748972,0.17222222222222222,0.7060371517027864,0.8285394123551169,0.8277777777777777,0.8901044315185644,0.6555555555555556 +0.3463934528477052,0.8685185185185185,0.8681889909960085,0.9359499314128944,0.13148148148148148,0.7671393910765663,0.8722409277964833,0.8685185185185185,0.9352080337247233,0.737037037037037 +0.3252469073323643,0.8648148148148148,0.8648032238703593,0.9388820301783265,0.13518518518518519,0.7618112331800375,0.8649399656946827,0.8648148148148148,0.9379325486367616,0.7296296296296296 +0.2923526299350402,0.8777777777777778,0.8777761011810862,0.9483916323731139,0.12222222222222222,0.7821758452422447,0.8777985074626866,0.8777777777777778,0.9491771441404628,0.7555555555555555 +0.2861848245648777,0.8833333333333333,0.883329732399148,0.9492455418381345,0.11666666666666667,0.7910396039603961,0.8833806642795407,0.8833333333333333,0.9496235895962679,0.7666666666666666 +0.29206473687115836,0.8703703703703703,0.8702261772339637,0.9492969821673525,0.12962962962962962,0.7702917886038905,0.8720238095238095,0.8703703703703703,0.9482820530310468,0.7407407407407407 +0.38481829972828135,0.85,0.8491302743179991,0.9472085048010974,0.15,0.737988332117929,0.8582611381794184,0.85,0.9459783382890903,0.7 +0.31191692632787366,0.8814814814814815,0.8814408233276159,0.9487654320987654,0.11851851851851852,0.7880213645685958,0.8820054945054945,0.8814814814814815,0.948846418375243,0.762962962962963 +0.2915133987279499,0.8925925925925926,0.8924982151683234,0.9528737997256516,0.10740740740740741,0.8058811018950716,0.8939761026375199,0.8925925925925926,0.9526898268231357,0.7851851851851852 +0.2874247786753318,0.8870370370370371,0.8870180517302291,0.9535905349794238,0.11296296296296296,0.7969774499379763,0.887297360365678,0.8870370370370371,0.9546795232001959,0.7740740740740741 +0.28243404363884644,0.8833333333333333,0.8832849021713399,0.9549108367626886,0.11666666666666667,0.7909752178888987,0.8839706508745655,0.8833333333333333,0.9550862608949127,0.7666666666666666 +0.3211606368422508,0.8796296296296297,0.878546292547725,0.954320987654321,0.12037037037037036,0.7835861142907459,0.8936755857124568,0.8796296296296297,0.9532398432193828,0.7592592592592593 +0.29891377117703943,0.8907407407407407,0.8907223779030085,0.9543449931412895,0.10925925925925926,0.8029781018642405,0.891003555201713,0.8907407407407407,0.9540240454645186,0.7814814814814814 +0.2983887769720134,0.9055555555555556,0.9054125065685759,0.9579561042524005,0.09444444444444444,0.8271929824561404,0.9080238479691964,0.9055555555555556,0.9575689674323488,0.8111111111111111 +0.3011535397347282,0.8944444444444445,0.8944151153098083,0.9544958847736627,0.10555555555555556,0.8090018185492018,0.8948832035595106,0.8944444444444444,0.9540939042711736,0.7888888888888889 +0.2757409114171477,0.9055555555555556,0.9054826254826255,0.9588305898491085,0.09444444444444444,0.8272999174917491,0.9068111455108359,0.9055555555555556,0.9587396267902231,0.8111111111111111 +0.29369787696529837,0.8925925925925926,0.8925690137753142,0.9572702331961591,0.10740740740740741,0.8059853442971416,0.8929375583502663,0.8925925925925926,0.9575053279991639,0.7851851851851852 +0.2757764432360144,0.8962962962962963,0.8959308408127306,0.9600240054869684,0.1037037037037037,0.8115372453358631,0.9019422338471812,0.8962962962962964,0.9602589906592075,0.7925925925925926 +0.2659992505522335,0.8944444444444445,0.8944353939809655,0.9593827160493826,0.10555555555555556,0.8090318308341564,0.8945797598627787,0.8944444444444444,0.9601556374653952,0.7888888888888889 +0.2602909899809781,0.9,0.8999945127304654,0.9614849108367627,0.1,0.8181735729267048,0.9000878107677954,0.9,0.9612259403458715,0.8 +0.27147211572703195,0.9018518518518519,0.9018245809021024,0.9627469135802469,0.09814814814814815,0.8212067082328312,0.9022988505747126,0.9018518518518519,0.9636766589251713,0.8037037037037037 +0.2723842714639271,0.9092592592592592,0.9090643526076123,0.9633641975308642,0.09074074074074075,0.8333160071153003,0.9127983396748529,0.9092592592592592,0.9637447001879904,0.8185185185185185 +0.279085652793155,0.9166666666666666,0.9166183419059744,0.9635390946502058,0.08333333333333333,0.8460778352082701,0.9176348462141315,0.9166666666666667,0.9632653292322255,0.8333333333333334 +0.26972537093302784,0.8925925925925926,0.8925911192197424,0.9625617283950617,0.10740740740740741,0.8060178970917226,0.8926141352063213,0.8925925925925926,0.9631256823116747,0.7851851851851852 +0.2818167503265774,0.8944444444444445,0.8944440824557012,0.9620473251028807,0.10555555555555556,0.8090446903548743,0.8944498552792219,0.8944444444444444,0.9625740831113985,0.7888888888888889 +0.2663208666969748,0.9055555555555556,0.9055396825941259,0.9629663923182443,0.09444444444444444,0.8273869519402575,0.9058283345458538,0.9055555555555556,0.9634235839219636,0.8111111111111111 +0.2758449465036392,0.9,0.8999945127304654,0.9627777777777777,0.1,0.8181735729267048,0.9000878107677954,0.9,0.9633054025910985,0.8 +0.27622606123194976,0.8981481481481481,0.8981450044754469,0.9625788751714678,0.10185185185185185,0.8151213504474374,0.8981973083096679,0.8981481481481481,0.9630246126920161,0.7962962962962963 +0.27518679902834053,0.9037037037037037,0.9036918138041734,0.9627743484224967,0.0962962962962963,0.8243062768956819,0.9039031620553359,0.9037037037037037,0.9632762101790555,0.8074074074074074 diff --git a/results/downsample/airogs/050/retfound/pr.png b/results/downsample/airogs/050/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..9b5f1560992514a8e46652f6b02ec603b5bab7c7 --- /dev/null +++ b/results/downsample/airogs/050/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c6f7521908837d8c38d034a347a97c13af3c01d7981609f932c3d48dd4b0a87b +size 44754 diff --git a/results/downsample/airogs/050/retfound/roc.png b/results/downsample/airogs/050/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..22c47bb4dfaf63d0ef2e9f59ea31e5af0d4cd055 --- /dev/null +++ b/results/downsample/airogs/050/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6549487a4cef457835640a13fb396e14dc40c488e160fc82f0642c7e244306f8 +size 61402 diff --git a/results/downsample/airogs/050/retfound/test_pred.npz b/results/downsample/airogs/050/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e69ca498e10c18ec690db96b823af9ff5f5dbb82 --- /dev/null +++ b/results/downsample/airogs/050/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4aee2e9477a6b40637ffd8905c40a608c5ce3b1166875cb5cab52df0532a9a2b +size 12510 diff --git a/results/downsample/airogs/050/retfound/train.log b/results/downsample/airogs/050/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..56c6d1b35a9b8dbc6efee2060d308d72d5d466ab --- /dev/null +++ b/results/downsample/airogs/050/retfound/train.log @@ -0,0 +1,596 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:47:10.679342221 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:47:11.143604] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:47:11.143879] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/airogs_50', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/050', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:47:14.111946] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:47:15.661168] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:47:23.979168] Sampler_train = +[14:47:24.024690] len of train_set: 2496 +[14:47:24.201907] [Adaptation] Full fine-tuning: training all parameters. +[14:47:24.202931] number of trainable params (M): 303.30 +[14:47:24.203025] base lr: 5.00e-03 +[14:47:24.203091] actual lr: 6.25e-04 +[14:47:24.203154] accumulate grad iterations: 1 +[14:47:24.203218] effective batch size: 32 +[14:47:24.206115] criterion = CrossEntropyLoss() +[14:47:24.206212] Start training for 30 epochs +[14:47:24.208113] log_dir: ./output_logs/retfound +[14:47:25.663091] Epoch: [0] [ 0/78] eta: 0:01:53 lr: 0.000000 loss: 0.6931 (0.6931) time: 1.4540 data: 0.7087 max mem: 7340 +[14:47:27.148190] Epoch: [0] [20/78] eta: 0:00:08 lr: 0.000016 loss: 0.6930 (0.6927) time: 0.0742 data: 0.0001 max mem: 9669 +[14:47:28.463243] Epoch: [0] [40/78] eta: 0:00:03 lr: 0.000032 loss: 0.6932 (0.6932) time: 0.0657 data: 0.0001 max mem: 9669 +[14:47:29.847965] Epoch: [0] [60/78] eta: 0:00:01 lr: 0.000048 loss: 0.6922 (0.6928) time: 0.0692 data: 0.0001 max mem: 9669 +[14:47:30.967162] Epoch: [0] [77/78] eta: 0:00:00 lr: 0.000062 loss: 0.6881 (0.6916) time: 0.0659 data: 0.0001 max mem: 9669 +[14:47:31.037593] Epoch: [0] Total time: 0:00:06 (0.0876 s / it) +[14:47:31.038896] Averaged stats: lr: 0.000062 loss: 0.6881 (0.6916) +[14:47:31.802649] val: [ 0/17] eta: 0:00:12 loss: 0.6344 (0.6344) time: 0.7481 data: 0.7210 max mem: 9669 +[14:47:32.215386] val: [10/17] eta: 0:00:00 loss: 0.6373 (0.6574) time: 0.1055 data: 0.0889 max mem: 9669 +[14:47:32.472367] val: [16/17] eta: 0:00:00 loss: 0.6867 (0.6807) time: 0.0833 data: 0.0591 max mem: 9669 +[14:47:32.554704] val: Total time: 0:00:01 (0.0883 s / it) +[14:47:32.572378] val loss: 0.6806979144320768 +[14:47:32.572637] Accuracy: 0.5481, F1 Score: 0.4322, ROC AUC: 0.7899, Hamming Loss: 0.4519, + Jaccard Score: 0.3108, Precision: 0.7626, Recall: 0.5481, + Average Precision: 0.7875, Kappa: 0.0963, Score: 0.4395 +[14:47:34.548332] Best epoch = 0, Best score = 0.4395 +[14:47:34.748702] log_dir: ./output_logs/retfound +[14:47:35.549456] Epoch: [1] [ 0/78] eta: 0:01:02 lr: 0.000063 loss: 0.7014 (0.7014) time: 0.7998 data: 0.6952 max mem: 9669 +[14:47:36.874331] Epoch: [1] [20/78] eta: 0:00:05 lr: 0.000079 loss: 0.6859 (0.6839) time: 0.0662 data: 0.0001 max mem: 9669 +[14:47:38.191482] Epoch: [1] [40/78] eta: 0:00:03 lr: 0.000095 loss: 0.6728 (0.6786) time: 0.0658 data: 0.0001 max mem: 9669 +[14:47:39.521935] Epoch: [1] [60/78] eta: 0:00:01 lr: 0.000111 loss: 0.6620 (0.6778) time: 0.0665 data: 0.0001 max mem: 9669 +[14:47:40.635027] Epoch: [1] [77/78] eta: 0:00:00 lr: 0.000124 loss: 0.6340 (0.6667) time: 0.0657 data: 0.0001 max mem: 9669 +[14:47:40.718063] Epoch: [1] Total time: 0:00:05 (0.0765 s / it) +[14:47:40.719162] Averaged stats: lr: 0.000124 loss: 0.6340 (0.6667) +[14:47:41.816211] val: [ 0/17] eta: 0:00:18 loss: 0.4816 (0.4816) time: 1.0855 data: 1.0679 max mem: 9669 +[14:47:42.634424] val: [10/17] eta: 0:00:01 loss: 0.5055 (0.5196) time: 0.1730 data: 0.1573 max mem: 9669 +[14:47:42.727415] val: [16/17] eta: 0:00:00 loss: 0.4845 (0.4897) time: 0.1174 data: 0.1018 max mem: 9669 +[14:47:42.808143] val: Total time: 0:00:02 (0.1222 s / it) +[14:47:42.830778] val loss: 0.489725402172874 +[14:47:42.831147] Accuracy: 0.8278, F1 Score: 0.8277, ROC AUC: 0.8932, Hamming Loss: 0.1722, + Jaccard Score: 0.7060, Precision: 0.8285, Recall: 0.8278, + Average Precision: 0.8901, Kappa: 0.6556, Score: 0.7922 +[14:47:44.830411] Best epoch = 1, Best score = 0.7922 +[14:47:44.977902] log_dir: ./output_logs/retfound +[14:47:46.127466] Epoch: [2] [ 0/78] eta: 0:01:29 lr: 0.000125 loss: 0.7033 (0.7033) time: 1.1484 data: 1.0650 max mem: 9669 +[14:47:47.454350] Epoch: [2] [20/78] eta: 0:00:06 lr: 0.000141 loss: 0.6039 (0.6110) time: 0.0663 data: 0.0001 max mem: 9669 +[14:47:48.775159] Epoch: [2] [40/78] eta: 0:00:03 lr: 0.000157 loss: 0.5983 (0.6089) time: 0.0660 data: 0.0001 max mem: 9669 +[14:47:50.101047] Epoch: [2] [60/78] eta: 0:00:01 lr: 0.000173 loss: 0.5463 (0.5934) time: 0.0662 data: 0.0001 max mem: 9669 +[14:47:51.287100] Epoch: [2] [77/78] eta: 0:00:00 lr: 0.000187 loss: 0.5639 (0.5854) time: 0.0693 data: 0.0001 max mem: 9669 +[14:47:51.393540] Epoch: [2] Total time: 0:00:06 (0.0822 s / it) +[14:47:51.394281] Averaged stats: lr: 0.000187 loss: 0.5639 (0.5854) +[14:47:52.448415] val: [ 0/17] eta: 0:00:17 loss: 0.3415 (0.3415) time: 1.0293 data: 1.0119 max mem: 9669 +[14:47:53.192559] val: [10/17] eta: 0:00:01 loss: 0.3883 (0.4077) time: 0.1612 data: 0.1455 max mem: 9669 +[14:47:53.290324] val: [16/17] eta: 0:00:00 loss: 0.3415 (0.3464) time: 0.1100 data: 0.0945 max mem: 9669 +[14:47:53.365649] val: Total time: 0:00:01 (0.1145 s / it) +[14:47:53.379955] val loss: 0.3463934528477052 +[14:47:53.380228] Accuracy: 0.8685, F1 Score: 0.8682, ROC AUC: 0.9359, Hamming Loss: 0.1315, + Jaccard Score: 0.7671, Precision: 0.8722, Recall: 0.8685, + Average Precision: 0.9352, Kappa: 0.7370, Score: 0.8471 +[14:47:55.516636] Best epoch = 2, Best score = 0.8471 +[14:47:55.583454] log_dir: ./output_logs/retfound +[14:47:56.664419] Epoch: [3] [ 0/78] eta: 0:01:24 lr: 0.000188 loss: 0.7470 (0.7470) time: 1.0801 data: 1.0081 max mem: 9669 +[14:47:58.003035] Epoch: [3] [20/78] eta: 0:00:06 lr: 0.000204 loss: 0.5535 (0.5625) time: 0.0669 data: 0.0012 max mem: 9669 +[14:47:59.324698] Epoch: [3] [40/78] eta: 0:00:03 lr: 0.000220 loss: 0.5687 (0.5619) time: 0.0660 data: 0.0001 max mem: 9669 +[14:48:00.643964] Epoch: [3] [60/78] eta: 0:00:01 lr: 0.000236 loss: 0.5471 (0.5646) time: 0.0659 data: 0.0001 max mem: 9669 +[14:48:01.756443] Epoch: [3] [77/78] eta: 0:00:00 lr: 0.000249 loss: 0.5611 (0.5640) time: 0.0654 data: 0.0001 max mem: 9669 +[14:48:01.842265] Epoch: [3] Total time: 0:00:06 (0.0802 s / it) +[14:48:01.843103] Averaged stats: lr: 0.000249 loss: 0.5611 (0.5640) +[14:48:03.058805] val: [ 0/17] eta: 0:00:20 loss: 0.2461 (0.2461) time: 1.1822 data: 1.1612 max mem: 9669 +[14:48:03.751626] val: [10/17] eta: 0:00:01 loss: 0.2996 (0.3237) time: 0.1704 data: 0.1544 max mem: 9669 +[14:48:03.900412] val: [16/17] eta: 0:00:00 loss: 0.3184 (0.3252) time: 0.1190 data: 0.1033 max mem: 9669 +[14:48:03.972581] val: Total time: 0:00:02 (0.1233 s / it) +[14:48:03.987786] val loss: 0.3252469073323643 +[14:48:03.987995] Accuracy: 0.8648, F1 Score: 0.8648, ROC AUC: 0.9389, Hamming Loss: 0.1352, + Jaccard Score: 0.7618, Precision: 0.8649, Recall: 0.8648, + Average Precision: 0.9379, Kappa: 0.7296, Score: 0.8444 +[14:48:04.054711] Best epoch = 2, Best score = 0.8471 +[14:48:04.374071] log_dir: ./output_logs/retfound +[14:48:05.476359] Epoch: [4] [ 0/78] eta: 0:01:25 lr: 0.000250 loss: 0.5457 (0.5457) time: 1.1014 data: 1.0323 max mem: 9669 +[14:48:06.890862] Epoch: [4] [20/78] eta: 0:00:06 lr: 0.000266 loss: 0.5444 (0.5538) time: 0.0707 data: 0.0049 max mem: 9669 +[14:48:08.205665] Epoch: [4] [40/78] eta: 0:00:03 lr: 0.000282 loss: 0.5047 (0.5394) time: 0.0657 data: 0.0001 max mem: 9669 +[14:48:09.524280] Epoch: [4] [60/78] eta: 0:00:01 lr: 0.000298 loss: 0.5058 (0.5290) time: 0.0659 data: 0.0001 max mem: 9669 +[14:48:10.639909] Epoch: [4] [77/78] eta: 0:00:00 lr: 0.000312 loss: 0.5259 (0.5315) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:10.722412] Epoch: [4] Total time: 0:00:06 (0.0814 s / it) +[14:48:10.723206] Averaged stats: lr: 0.000312 loss: 0.5259 (0.5315) +[14:48:11.880278] val: [ 0/17] eta: 0:00:19 loss: 0.2064 (0.2064) time: 1.1234 data: 1.1068 max mem: 9669 +[14:48:12.520789] val: [10/17] eta: 0:00:01 loss: 0.2951 (0.3045) time: 0.1603 data: 0.1447 max mem: 9669 +[14:48:12.640976] val: [16/17] eta: 0:00:00 loss: 0.2933 (0.2924) time: 0.1108 data: 0.0953 max mem: 9669 +[14:48:12.729745] val: Total time: 0:00:01 (0.1161 s / it) +[14:48:12.743652] val loss: 0.2923526299350402 +[14:48:12.743900] Accuracy: 0.8778, F1 Score: 0.8778, ROC AUC: 0.9484, Hamming Loss: 0.1222, + Jaccard Score: 0.7822, Precision: 0.8778, Recall: 0.8778, + Average Precision: 0.9492, Kappa: 0.7556, Score: 0.8606 +[14:48:14.815563] Best epoch = 4, Best score = 0.8606 +[14:48:14.947922] log_dir: ./output_logs/retfound +[14:48:15.724455] Epoch: [5] [ 0/78] eta: 0:01:00 lr: 0.000313 loss: 0.4954 (0.4954) time: 0.7756 data: 0.7006 max mem: 9669 +[14:48:17.050654] Epoch: [5] [20/78] eta: 0:00:05 lr: 0.000329 loss: 0.5459 (0.5428) time: 0.0663 data: 0.0001 max mem: 9669 +[14:48:18.389352] Epoch: [5] [40/78] eta: 0:00:03 lr: 0.000345 loss: 0.5357 (0.5408) time: 0.0669 data: 0.0001 max mem: 9669 +[14:48:19.701663] Epoch: [5] [60/78] eta: 0:00:01 lr: 0.000361 loss: 0.4968 (0.5369) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:20.815996] Epoch: [5] [77/78] eta: 0:00:00 lr: 0.000374 loss: 0.5219 (0.5367) time: 0.0655 data: 0.0001 max mem: 9669 +[14:48:20.894582] Epoch: [5] Total time: 0:00:05 (0.0762 s / it) +[14:48:20.895468] Averaged stats: lr: 0.000374 loss: 0.5219 (0.5367) +[14:48:21.960080] val: [ 0/17] eta: 0:00:17 loss: 0.1987 (0.1987) time: 1.0409 data: 1.0239 max mem: 9669 +[14:48:22.138222] val: [10/17] eta: 0:00:00 loss: 0.2900 (0.2880) time: 0.1108 data: 0.0953 max mem: 9669 +[14:48:22.501395] val: [16/17] eta: 0:00:00 loss: 0.2900 (0.2862) time: 0.0930 data: 0.0776 max mem: 9669 +[14:48:22.583884] val: Total time: 0:00:01 (0.0980 s / it) +[14:48:22.598855] val loss: 0.2861848245648777 +[14:48:22.599061] Accuracy: 0.8833, F1 Score: 0.8833, ROC AUC: 0.9492, Hamming Loss: 0.1167, + Jaccard Score: 0.7910, Precision: 0.8834, Recall: 0.8833, + Average Precision: 0.9496, Kappa: 0.7667, Score: 0.8664 +[14:48:24.506701] Best epoch = 5, Best score = 0.8664 +[14:48:24.579597] log_dir: ./output_logs/retfound +[14:48:25.694044] Epoch: [6] [ 0/78] eta: 0:01:26 lr: 0.000375 loss: 0.4515 (0.4515) time: 1.1135 data: 1.0416 max mem: 9669 +[14:48:27.044469] Epoch: [6] [20/78] eta: 0:00:06 lr: 0.000391 loss: 0.4922 (0.5016) time: 0.0675 data: 0.0010 max mem: 9669 +[14:48:28.375059] Epoch: [6] [40/78] eta: 0:00:03 lr: 0.000407 loss: 0.5091 (0.5218) time: 0.0665 data: 0.0001 max mem: 9669 +[14:48:29.729430] Epoch: [6] [60/78] eta: 0:00:01 lr: 0.000423 loss: 0.5576 (0.5355) time: 0.0677 data: 0.0001 max mem: 9669 +[14:48:30.849277] Epoch: [6] [77/78] eta: 0:00:00 lr: 0.000437 loss: 0.5231 (0.5364) time: 0.0672 data: 0.0001 max mem: 9669 +[14:48:30.950288] Epoch: [6] Total time: 0:00:06 (0.0817 s / it) +[14:48:30.951116] Averaged stats: lr: 0.000437 loss: 0.5231 (0.5364) +[14:48:32.225322] val: [ 0/17] eta: 0:00:21 loss: 0.1959 (0.1959) time: 1.2617 data: 1.2447 max mem: 9669 +[14:48:32.909153] val: [10/17] eta: 0:00:01 loss: 0.2828 (0.2952) time: 0.1768 data: 0.1611 max mem: 9669 +[14:48:33.030471] val: [16/17] eta: 0:00:00 loss: 0.2700 (0.2921) time: 0.1215 data: 0.1060 max mem: 9669 +[14:48:33.109144] val: Total time: 0:00:02 (0.1263 s / it) +[14:48:33.129481] val loss: 0.29206473687115836 +[14:48:33.129731] Accuracy: 0.8704, F1 Score: 0.8702, ROC AUC: 0.9493, Hamming Loss: 0.1296, + Jaccard Score: 0.7703, Precision: 0.8720, Recall: 0.8704, + Average Precision: 0.9483, Kappa: 0.7407, Score: 0.8534 +[14:48:33.212739] Best epoch = 5, Best score = 0.8664 +[14:48:33.509520] log_dir: ./output_logs/retfound +[14:48:34.595359] Epoch: [7] [ 0/78] eta: 0:01:24 lr: 0.000438 loss: 0.4191 (0.4191) time: 1.0849 data: 1.0093 max mem: 9669 +[14:48:35.904932] Epoch: [7] [20/78] eta: 0:00:06 lr: 0.000454 loss: 0.5139 (0.5283) time: 0.0654 data: 0.0001 max mem: 9669 +[14:48:37.217268] Epoch: [7] [40/78] eta: 0:00:03 lr: 0.000470 loss: 0.5396 (0.5338) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:38.530935] Epoch: [7] [60/78] eta: 0:00:01 lr: 0.000486 loss: 0.5727 (0.5420) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:39.656553] Epoch: [7] [77/78] eta: 0:00:00 lr: 0.000499 loss: 0.5207 (0.5426) time: 0.0662 data: 0.0001 max mem: 9669 +[14:48:39.733508] Epoch: [7] Total time: 0:00:06 (0.0798 s / it) +[14:48:39.734301] Averaged stats: lr: 0.000499 loss: 0.5207 (0.5426) +[14:48:40.870693] val: [ 0/17] eta: 0:00:18 loss: 0.1041 (0.1041) time: 1.1123 data: 1.0944 max mem: 9669 +[14:48:41.586549] val: [10/17] eta: 0:00:01 loss: 0.2932 (0.3171) time: 0.1661 data: 0.1504 max mem: 9669 +[14:48:41.678688] val: [16/17] eta: 0:00:00 loss: 0.3368 (0.3848) time: 0.1129 data: 0.0974 max mem: 9669 +[14:48:41.755605] val: Total time: 0:00:01 (0.1175 s / it) +[14:48:41.769864] val loss: 0.38481829972828135 +[14:48:41.770067] Accuracy: 0.8500, F1 Score: 0.8491, ROC AUC: 0.9472, Hamming Loss: 0.1500, + Jaccard Score: 0.7380, Precision: 0.8583, Recall: 0.8500, + Average Precision: 0.9460, Kappa: 0.7000, Score: 0.8321 +[14:48:41.831822] Best epoch = 5, Best score = 0.8664 +[14:48:42.141376] log_dir: ./output_logs/retfound +[14:48:43.195055] Epoch: [8] [ 0/78] eta: 0:01:22 lr: 0.000500 loss: 0.3723 (0.3723) time: 1.0528 data: 0.9802 max mem: 9669 +[14:48:44.514884] Epoch: [8] [20/78] eta: 0:00:06 lr: 0.000516 loss: 0.5376 (0.5402) time: 0.0660 data: 0.0001 max mem: 9669 +[14:48:45.824853] Epoch: [8] [40/78] eta: 0:00:03 lr: 0.000532 loss: 0.5123 (0.5332) time: 0.0655 data: 0.0001 max mem: 9669 +[14:48:47.133142] Epoch: [8] [60/78] eta: 0:00:01 lr: 0.000548 loss: 0.5489 (0.5364) time: 0.0654 data: 0.0001 max mem: 9669 +[14:48:48.249356] Epoch: [8] [77/78] eta: 0:00:00 lr: 0.000562 loss: 0.5211 (0.5333) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:48.325392] Epoch: [8] Total time: 0:00:06 (0.0793 s / it) +[14:48:48.326200] Averaged stats: lr: 0.000562 loss: 0.5211 (0.5333) +[14:48:49.437647] val: [ 0/17] eta: 0:00:18 loss: 0.3283 (0.3283) time: 1.0675 data: 1.0503 max mem: 9669 +[14:48:50.010844] val: [10/17] eta: 0:00:01 loss: 0.3407 (0.3345) time: 0.1491 data: 0.1335 max mem: 9669 +[14:48:50.108663] val: [16/17] eta: 0:00:00 loss: 0.3264 (0.3119) time: 0.1022 data: 0.0867 max mem: 9669 +[14:48:50.190310] val: Total time: 0:00:01 (0.1071 s / it) +[14:48:50.206398] val loss: 0.31191692632787366 +[14:48:50.206674] Accuracy: 0.8815, F1 Score: 0.8814, ROC AUC: 0.9488, Hamming Loss: 0.1185, + Jaccard Score: 0.7880, Precision: 0.8820, Recall: 0.8815, + Average Precision: 0.9488, Kappa: 0.7630, Score: 0.8644 +[14:48:50.261310] Best epoch = 5, Best score = 0.8664 +[14:48:50.540561] log_dir: ./output_logs/retfound +[14:48:51.492302] Epoch: [9] [ 0/78] eta: 0:01:14 lr: 0.000562 loss: 0.4863 (0.4863) time: 0.9508 data: 0.8785 max mem: 9669 +[14:48:52.814128] Epoch: [9] [20/78] eta: 0:00:06 lr: 0.000579 loss: 0.4959 (0.5111) time: 0.0661 data: 0.0001 max mem: 9669 +[14:48:54.136515] Epoch: [9] [40/78] eta: 0:00:03 lr: 0.000595 loss: 0.4739 (0.5028) time: 0.0661 data: 0.0001 max mem: 9669 +[14:48:55.446295] Epoch: [9] [60/78] eta: 0:00:01 lr: 0.000611 loss: 0.4768 (0.5002) time: 0.0655 data: 0.0001 max mem: 9669 +[14:48:56.559146] Epoch: [9] [77/78] eta: 0:00:00 lr: 0.000624 loss: 0.4926 (0.4989) time: 0.0654 data: 0.0001 max mem: 9669 +[14:48:56.635869] Epoch: [9] Total time: 0:00:06 (0.0781 s / it) +[14:48:56.636696] Averaged stats: lr: 0.000624 loss: 0.4926 (0.4989) +[14:48:57.851616] val: [ 0/17] eta: 0:00:19 loss: 0.1961 (0.1961) time: 1.1631 data: 1.1455 max mem: 9669 +[14:48:58.445128] val: [10/17] eta: 0:00:01 loss: 0.2767 (0.2969) time: 0.1596 data: 0.1439 max mem: 9669 +[14:48:58.539617] val: [16/17] eta: 0:00:00 loss: 0.2777 (0.2915) time: 0.1088 data: 0.0932 max mem: 9669 +[14:48:58.626747] val: Total time: 0:00:01 (0.1141 s / it) +[14:48:58.640979] val loss: 0.2915133987279499 +[14:48:58.641241] Accuracy: 0.8926, F1 Score: 0.8925, ROC AUC: 0.9529, Hamming Loss: 0.1074, + Jaccard Score: 0.8059, Precision: 0.8940, Recall: 0.8926, + Average Precision: 0.9527, Kappa: 0.7852, Score: 0.8769 +[14:49:00.626422] Best epoch = 9, Best score = 0.8769 +[14:49:00.706931] log_dir: ./output_logs/retfound +[14:49:01.662361] Epoch: [10] [ 0/78] eta: 0:01:14 lr: 0.000625 loss: 0.3736 (0.3736) time: 0.9543 data: 0.8837 max mem: 9669 +[14:49:02.979471] Epoch: [10] [20/78] eta: 0:00:06 lr: 0.000625 loss: 0.4897 (0.4931) time: 0.0658 data: 0.0001 max mem: 9669 +[14:49:04.293561] Epoch: [10] [40/78] eta: 0:00:03 lr: 0.000624 loss: 0.4804 (0.4858) time: 0.0657 data: 0.0001 max mem: 9669 +[14:49:05.610395] Epoch: [10] [60/78] eta: 0:00:01 lr: 0.000623 loss: 0.5070 (0.4989) time: 0.0658 data: 0.0001 max mem: 9669 +[14:49:06.724897] Epoch: [10] [77/78] eta: 0:00:00 lr: 0.000621 loss: 0.5171 (0.5028) time: 0.0657 data: 0.0001 max mem: 9669 +[14:49:06.805987] Epoch: [10] Total time: 0:00:06 (0.0782 s / it) +[14:49:06.806735] Averaged stats: lr: 0.000621 loss: 0.5171 (0.5028) +[14:49:07.429468] val: [ 0/17] eta: 0:00:09 loss: 0.2817 (0.2817) time: 0.5870 data: 0.5697 max mem: 9669 +[14:49:07.927909] val: [10/17] eta: 0:00:00 loss: 0.3026 (0.3349) time: 0.0986 data: 0.0830 max mem: 9669 +[14:49:08.057011] val: [16/17] eta: 0:00:00 loss: 0.2817 (0.2874) time: 0.0714 data: 0.0559 max mem: 9669 +[14:49:08.133724] val: Total time: 0:00:01 (0.0760 s / it) +[14:49:08.147646] val loss: 0.2874247786753318 +[14:49:08.147910] Accuracy: 0.8870, F1 Score: 0.8870, ROC AUC: 0.9536, Hamming Loss: 0.1130, + Jaccard Score: 0.7970, Precision: 0.8873, Recall: 0.8870, + Average Precision: 0.9547, Kappa: 0.7741, Score: 0.8716 +[14:49:08.205185] Best epoch = 9, Best score = 0.8769 +[14:49:08.585610] log_dir: ./output_logs/retfound +[14:49:09.628603] Epoch: [11] [ 0/78] eta: 0:01:21 lr: 0.000621 loss: 0.4325 (0.4325) time: 1.0421 data: 0.9720 max mem: 9669 +[14:49:10.938659] Epoch: [11] [20/78] eta: 0:00:06 lr: 0.000619 loss: 0.5009 (0.5191) time: 0.0655 data: 0.0001 max mem: 9669 +[14:49:12.258258] Epoch: [11] [40/78] eta: 0:00:03 lr: 0.000616 loss: 0.4954 (0.5143) time: 0.0659 data: 0.0001 max mem: 9669 +[14:49:13.564998] Epoch: [11] [60/78] eta: 0:00:01 lr: 0.000613 loss: 0.4424 (0.4981) time: 0.0653 data: 0.0001 max mem: 9669 +[14:49:14.675096] Epoch: [11] [77/78] eta: 0:00:00 lr: 0.000610 loss: 0.4253 (0.4940) time: 0.0653 data: 0.0001 max mem: 9669 +[14:49:14.758971] Epoch: [11] Total time: 0:00:06 (0.0791 s / it) +[14:49:14.759763] Averaged stats: lr: 0.000610 loss: 0.4253 (0.4940) +[14:49:15.452257] val: [ 0/17] eta: 0:00:11 loss: 0.1808 (0.1808) time: 0.6623 data: 0.6451 max mem: 9669 +[14:49:16.017942] val: [10/17] eta: 0:00:00 loss: 0.2404 (0.2922) time: 0.1116 data: 0.0960 max mem: 9669 +[14:49:16.150088] val: [16/17] eta: 0:00:00 loss: 0.2404 (0.2824) time: 0.0799 data: 0.0645 max mem: 9669 +[14:49:16.229542] val: Total time: 0:00:01 (0.0847 s / it) +[14:49:16.243726] val loss: 0.28243404363884644 +[14:49:16.243916] Accuracy: 0.8833, F1 Score: 0.8833, ROC AUC: 0.9549, Hamming Loss: 0.1167, + Jaccard Score: 0.7910, Precision: 0.8840, Recall: 0.8833, + Average Precision: 0.9551, Kappa: 0.7667, Score: 0.8683 +[14:49:16.273576] Best epoch = 9, Best score = 0.8769 +[14:49:16.546498] log_dir: ./output_logs/retfound +[14:49:17.475315] Epoch: [12] [ 0/78] eta: 0:01:12 lr: 0.000610 loss: 0.4902 (0.4902) time: 0.9276 data: 0.8536 max mem: 9669 +[14:49:18.799056] Epoch: [12] [20/78] eta: 0:00:06 lr: 0.000606 loss: 0.4685 (0.4882) time: 0.0661 data: 0.0001 max mem: 9669 +[14:49:20.115118] Epoch: [12] [40/78] eta: 0:00:03 lr: 0.000601 loss: 0.4625 (0.4909) time: 0.0658 data: 0.0001 max mem: 9669 +[14:49:21.444279] Epoch: [12] [60/78] eta: 0:00:01 lr: 0.000596 loss: 0.4706 (0.4850) time: 0.0664 data: 0.0001 max mem: 9669 +[14:49:22.555760] Epoch: [12] [77/78] eta: 0:00:00 lr: 0.000591 loss: 0.4780 (0.4864) time: 0.0653 data: 0.0001 max mem: 9669 +[14:49:22.641246] Epoch: [12] Total time: 0:00:06 (0.0781 s / it) +[14:49:22.642179] Averaged stats: lr: 0.000591 loss: 0.4780 (0.4864) +[14:49:23.555952] val: [ 0/17] eta: 0:00:14 loss: 0.4624 (0.4624) time: 0.8741 data: 0.8569 max mem: 9669 +[14:49:24.005500] val: [10/17] eta: 0:00:00 loss: 0.4390 (0.4325) time: 0.1203 data: 0.1046 max mem: 9669 +[14:49:24.097849] val: [16/17] eta: 0:00:00 loss: 0.3557 (0.3212) time: 0.0832 data: 0.0677 max mem: 9669 +[14:49:24.176540] val: Total time: 0:00:01 (0.0880 s / it) +[14:49:24.190503] val loss: 0.3211606368422508 +[14:49:24.190695] Accuracy: 0.8796, F1 Score: 0.8785, ROC AUC: 0.9543, Hamming Loss: 0.1204, + Jaccard Score: 0.7836, Precision: 0.8937, Recall: 0.8796, + Average Precision: 0.9532, Kappa: 0.7593, Score: 0.8640 +[14:49:24.237197] Best epoch = 9, Best score = 0.8769 +[14:49:24.504554] log_dir: ./output_logs/retfound +[14:49:25.448958] Epoch: [13] [ 0/78] eta: 0:01:13 lr: 0.000591 loss: 0.5024 (0.5024) time: 0.9432 data: 0.8706 max mem: 9669 +[14:49:26.788680] Epoch: [13] [20/78] eta: 0:00:06 lr: 0.000585 loss: 0.4747 (0.4779) time: 0.0669 data: 0.0007 max mem: 9669 +[14:49:28.110036] Epoch: [13] [40/78] eta: 0:00:03 lr: 0.000579 loss: 0.4591 (0.4709) time: 0.0660 data: 0.0001 max mem: 9669 +[14:49:29.427009] Epoch: [13] [60/78] eta: 0:00:01 lr: 0.000572 loss: 0.4601 (0.4795) time: 0.0658 data: 0.0001 max mem: 9669 +[14:49:30.540729] Epoch: [13] [77/78] eta: 0:00:00 lr: 0.000566 loss: 0.4395 (0.4741) time: 0.0655 data: 0.0001 max mem: 9669 +[14:49:30.634694] Epoch: [13] Total time: 0:00:06 (0.0786 s / it) +[14:49:30.635519] Averaged stats: lr: 0.000566 loss: 0.4395 (0.4741) +[14:49:31.941997] val: [ 0/17] eta: 0:00:21 loss: 0.2944 (0.2944) time: 1.2686 data: 1.2515 max mem: 9669 +[14:49:32.501608] val: [10/17] eta: 0:00:01 loss: 0.3161 (0.3577) time: 0.1661 data: 0.1506 max mem: 9669 +[14:49:32.594288] val: [16/17] eta: 0:00:00 loss: 0.2791 (0.2989) time: 0.1129 data: 0.0975 max mem: 9669 +[14:49:32.684913] val: Total time: 0:00:02 (0.1184 s / it) +[14:49:32.699380] val loss: 0.29891377117703943 +[14:49:32.699615] Accuracy: 0.8907, F1 Score: 0.8907, ROC AUC: 0.9543, Hamming Loss: 0.1093, + Jaccard Score: 0.8030, Precision: 0.8910, Recall: 0.8907, + Average Precision: 0.9540, Kappa: 0.7815, Score: 0.8755 +[14:49:32.793712] Best epoch = 9, Best score = 0.8769 +[14:49:33.038586] log_dir: ./output_logs/retfound +[14:49:34.120573] Epoch: [14] [ 0/78] eta: 0:01:24 lr: 0.000565 loss: 0.3791 (0.3791) time: 1.0812 data: 1.0103 max mem: 9669 +[14:49:35.443437] Epoch: [14] [20/78] eta: 0:00:06 lr: 0.000558 loss: 0.4476 (0.4461) time: 0.0661 data: 0.0001 max mem: 9669 +[14:49:36.756864] Epoch: [14] [40/78] eta: 0:00:03 lr: 0.000550 loss: 0.4389 (0.4539) time: 0.0656 data: 0.0001 max mem: 9669 +[14:49:38.071856] Epoch: [14] [60/78] eta: 0:00:01 lr: 0.000541 loss: 0.5117 (0.4754) time: 0.0657 data: 0.0001 max mem: 9669 +[14:49:39.185362] Epoch: [14] [77/78] eta: 0:00:00 lr: 0.000534 loss: 0.4114 (0.4685) time: 0.0655 data: 0.0001 max mem: 9669 +[14:49:39.283156] Epoch: [14] Total time: 0:00:06 (0.0801 s / it) +[14:49:39.283981] Averaged stats: lr: 0.000534 loss: 0.4114 (0.4685) +[14:49:40.401973] val: [ 0/17] eta: 0:00:18 loss: 0.3620 (0.3620) time: 1.0893 data: 1.0720 max mem: 9669 +[14:49:40.596814] val: [10/17] eta: 0:00:00 loss: 0.3550 (0.3905) time: 0.1167 data: 0.1011 max mem: 9669 +[14:49:40.908547] val: [16/17] eta: 0:00:00 loss: 0.2580 (0.2984) time: 0.0938 data: 0.0781 max mem: 9669 +[14:49:40.996034] val: Total time: 0:00:01 (0.0991 s / it) +[14:49:41.011433] val loss: 0.2983887769720134 +[14:49:41.011629] Accuracy: 0.9056, F1 Score: 0.9054, ROC AUC: 0.9580, Hamming Loss: 0.0944, + Jaccard Score: 0.8272, Precision: 0.9080, Recall: 0.9056, + Average Precision: 0.9576, Kappa: 0.8111, Score: 0.8915 +[14:49:42.809407] Best epoch = 14, Best score = 0.8915 +[14:49:42.894691] log_dir: ./output_logs/retfound +[14:49:43.955432] Epoch: [15] [ 0/78] eta: 0:01:22 lr: 0.000534 loss: 0.5945 (0.5945) time: 1.0598 data: 0.9856 max mem: 9669 +[14:49:45.289931] Epoch: [15] [20/78] eta: 0:00:06 lr: 0.000525 loss: 0.4535 (0.4753) time: 0.0667 data: 0.0001 max mem: 9669 +[14:49:46.604613] Epoch: [15] [40/78] eta: 0:00:03 lr: 0.000515 loss: 0.4250 (0.4631) time: 0.0657 data: 0.0001 max mem: 9669 +[14:49:47.922001] Epoch: [15] [60/78] eta: 0:00:01 lr: 0.000505 loss: 0.4288 (0.4520) time: 0.0658 data: 0.0001 max mem: 9669 +[14:49:49.057607] Epoch: [15] [77/78] eta: 0:00:00 lr: 0.000497 loss: 0.4419 (0.4539) time: 0.0666 data: 0.0001 max mem: 9669 +[14:49:49.143306] Epoch: [15] Total time: 0:00:06 (0.0801 s / it) +[14:49:49.144109] Averaged stats: lr: 0.000497 loss: 0.4419 (0.4539) +[14:49:50.411756] val: [ 0/17] eta: 0:00:20 loss: 0.2084 (0.2084) time: 1.2289 data: 1.2106 max mem: 9669 +[14:49:50.611410] val: [10/17] eta: 0:00:00 loss: 0.2781 (0.2980) time: 0.1298 data: 0.1141 max mem: 9669 +[14:49:51.050121] val: [16/17] eta: 0:00:00 loss: 0.2933 (0.3012) time: 0.1098 data: 0.0942 max mem: 9669 +[14:49:51.135691] val: Total time: 0:00:01 (0.1149 s / it) +[14:49:51.154071] val loss: 0.3011535397347282 +[14:49:51.154294] Accuracy: 0.8944, F1 Score: 0.8944, ROC AUC: 0.9545, Hamming Loss: 0.1056, + Jaccard Score: 0.8090, Precision: 0.8949, Recall: 0.8944, + Average Precision: 0.9541, Kappa: 0.7889, Score: 0.8793 +[14:49:51.199841] Best epoch = 14, Best score = 0.8915 +[14:49:51.492729] log_dir: ./output_logs/retfound +[14:49:52.464549] Epoch: [16] [ 0/78] eta: 0:01:15 lr: 0.000496 loss: 0.6196 (0.6196) time: 0.9708 data: 0.8977 max mem: 9669 +[14:49:53.790060] Epoch: [16] [20/78] eta: 0:00:06 lr: 0.000486 loss: 0.5082 (0.5256) time: 0.0662 data: 0.0001 max mem: 9669 +[14:49:55.113587] Epoch: [16] [40/78] eta: 0:00:03 lr: 0.000475 loss: 0.4164 (0.4742) time: 0.0661 data: 0.0001 max mem: 9669 +[14:49:56.428842] Epoch: [16] [60/78] eta: 0:00:01 lr: 0.000465 loss: 0.4189 (0.4676) time: 0.0657 data: 0.0001 max mem: 9669 +[14:49:57.545118] Epoch: [16] [77/78] eta: 0:00:00 lr: 0.000455 loss: 0.4765 (0.4696) time: 0.0656 data: 0.0001 max mem: 9669 +[14:49:57.633930] Epoch: [16] Total time: 0:00:06 (0.0787 s / it) +[14:49:57.634876] Averaged stats: lr: 0.000455 loss: 0.4765 (0.4696) +[14:49:58.729807] val: [ 0/17] eta: 0:00:18 loss: 0.3044 (0.3044) time: 1.0668 data: 1.0482 max mem: 9669 +[14:49:59.157927] val: [10/17] eta: 0:00:00 loss: 0.3514 (0.3371) time: 0.1358 data: 0.1201 max mem: 9669 +[14:49:59.250834] val: [16/17] eta: 0:00:00 loss: 0.2322 (0.2757) time: 0.0933 data: 0.0778 max mem: 9669 +[14:49:59.336235] val: Total time: 0:00:01 (0.0985 s / it) +[14:49:59.352546] val loss: 0.2757409114171477 +[14:49:59.352760] Accuracy: 0.9056, F1 Score: 0.9055, ROC AUC: 0.9588, Hamming Loss: 0.0944, + Jaccard Score: 0.8273, Precision: 0.9068, Recall: 0.9056, + Average Precision: 0.9587, Kappa: 0.8111, Score: 0.8918 +[14:50:01.100593] Best epoch = 16, Best score = 0.8918 +[14:50:01.181828] log_dir: ./output_logs/retfound +[14:50:02.047575] Epoch: [17] [ 0/78] eta: 0:01:07 lr: 0.000455 loss: 0.4216 (0.4216) time: 0.8647 data: 0.7922 max mem: 9669 +[14:50:03.367232] Epoch: [17] [20/78] eta: 0:00:06 lr: 0.000443 loss: 0.4466 (0.4667) time: 0.0659 data: 0.0001 max mem: 9669 +[14:50:04.679230] Epoch: [17] [40/78] eta: 0:00:03 lr: 0.000432 loss: 0.4271 (0.4565) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:06.001418] Epoch: [17] [60/78] eta: 0:00:01 lr: 0.000420 loss: 0.4834 (0.4585) time: 0.0661 data: 0.0001 max mem: 9669 +[14:50:07.118725] Epoch: [17] [77/78] eta: 0:00:00 lr: 0.000410 loss: 0.4194 (0.4529) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:07.207033] Epoch: [17] Total time: 0:00:06 (0.0772 s / it) +[14:50:07.208063] Averaged stats: lr: 0.000410 loss: 0.4194 (0.4529) +[14:50:08.328195] val: [ 0/17] eta: 0:00:18 loss: 0.1845 (0.1845) time: 1.1054 data: 1.0838 max mem: 9669 +[14:50:08.511660] val: [10/17] eta: 0:00:00 loss: 0.2583 (0.2702) time: 0.1171 data: 0.1011 max mem: 9669 +[14:50:08.820360] val: [16/17] eta: 0:00:00 loss: 0.2926 (0.2937) time: 0.0939 data: 0.0781 max mem: 9669 +[14:50:08.901557] val: Total time: 0:00:01 (0.0988 s / it) +[14:50:08.919585] val loss: 0.29369787696529837 +[14:50:08.919784] Accuracy: 0.8926, F1 Score: 0.8926, ROC AUC: 0.9573, Hamming Loss: 0.1074, + Jaccard Score: 0.8060, Precision: 0.8929, Recall: 0.8926, + Average Precision: 0.9575, Kappa: 0.7852, Score: 0.8783 +[14:50:08.971528] Best epoch = 16, Best score = 0.8918 +[14:50:09.299849] log_dir: ./output_logs/retfound +[14:50:10.330910] Epoch: [18] [ 0/78] eta: 0:01:20 lr: 0.000409 loss: 0.5990 (0.5990) time: 1.0301 data: 0.9610 max mem: 9669 +[14:50:11.645049] Epoch: [18] [20/78] eta: 0:00:06 lr: 0.000397 loss: 0.4210 (0.4623) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:12.957010] Epoch: [18] [40/78] eta: 0:00:03 lr: 0.000385 loss: 0.4597 (0.4660) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:14.271507] Epoch: [18] [60/78] eta: 0:00:01 lr: 0.000373 loss: 0.4560 (0.4597) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:15.392297] Epoch: [18] [77/78] eta: 0:00:00 lr: 0.000362 loss: 0.4569 (0.4699) time: 0.0658 data: 0.0001 max mem: 9669 +[14:50:15.484059] Epoch: [18] Total time: 0:00:06 (0.0793 s / it) +[14:50:15.484945] Averaged stats: lr: 0.000362 loss: 0.4569 (0.4699) +[14:50:16.407519] val: [ 0/17] eta: 0:00:14 loss: 0.3818 (0.3818) time: 0.8699 data: 0.8526 max mem: 9669 +[14:50:16.588455] val: [10/17] eta: 0:00:00 loss: 0.3599 (0.3574) time: 0.0955 data: 0.0799 max mem: 9669 +[14:50:16.680559] val: [16/17] eta: 0:00:00 loss: 0.2016 (0.2758) time: 0.0672 data: 0.0517 max mem: 9669 +[14:50:16.768600] val: Total time: 0:00:01 (0.0724 s / it) +[14:50:16.782685] val loss: 0.2757764432360144 +[14:50:16.782890] Accuracy: 0.8963, F1 Score: 0.8959, ROC AUC: 0.9600, Hamming Loss: 0.1037, + Jaccard Score: 0.8115, Precision: 0.9019, Recall: 0.8963, + Average Precision: 0.9603, Kappa: 0.7926, Score: 0.8828 +[14:50:16.840256] Best epoch = 16, Best score = 0.8918 +[14:50:17.160916] log_dir: ./output_logs/retfound +[14:50:18.359161] Epoch: [19] [ 0/78] eta: 0:01:33 lr: 0.000362 loss: 0.4375 (0.4375) time: 1.1971 data: 1.1253 max mem: 9669 +[14:50:19.676989] Epoch: [19] [20/78] eta: 0:00:06 lr: 0.000349 loss: 0.4443 (0.4609) time: 0.0659 data: 0.0001 max mem: 9669 +[14:50:20.990897] Epoch: [19] [40/78] eta: 0:00:03 lr: 0.000337 loss: 0.3835 (0.4406) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:22.305710] Epoch: [19] [60/78] eta: 0:00:01 lr: 0.000324 loss: 0.4349 (0.4452) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:23.418161] Epoch: [19] [77/78] eta: 0:00:00 lr: 0.000314 loss: 0.4531 (0.4448) time: 0.0654 data: 0.0001 max mem: 9669 +[14:50:23.507012] Epoch: [19] Total time: 0:00:06 (0.0814 s / it) +[14:50:23.507859] Averaged stats: lr: 0.000314 loss: 0.4531 (0.4448) +[14:50:24.673454] val: [ 0/17] eta: 0:00:19 loss: 0.2530 (0.2530) time: 1.1272 data: 1.1090 max mem: 9669 +[14:50:25.288489] val: [10/17] eta: 0:00:01 loss: 0.2881 (0.2829) time: 0.1583 data: 0.1426 max mem: 9669 +[14:50:25.382152] val: [16/17] eta: 0:00:00 loss: 0.2815 (0.2660) time: 0.1079 data: 0.0923 max mem: 9669 +[14:50:25.467258] val: Total time: 0:00:01 (0.1130 s / it) +[14:50:25.484172] val loss: 0.2659992505522335 +[14:50:25.484375] Accuracy: 0.8944, F1 Score: 0.8944, ROC AUC: 0.9594, Hamming Loss: 0.1056, + Jaccard Score: 0.8090, Precision: 0.8946, Recall: 0.8944, + Average Precision: 0.9602, Kappa: 0.7889, Score: 0.8809 +[14:50:25.525474] Best epoch = 16, Best score = 0.8918 +[14:50:25.867476] log_dir: ./output_logs/retfound +[14:50:26.835302] Epoch: [20] [ 0/78] eta: 0:01:15 lr: 0.000313 loss: 0.5070 (0.5070) time: 0.9658 data: 0.8948 max mem: 9669 +[14:50:28.401585] Epoch: [20] [20/78] eta: 0:00:06 lr: 0.000300 loss: 0.3958 (0.4229) time: 0.0783 data: 0.0126 max mem: 9669 +[14:50:29.713707] Epoch: [20] [40/78] eta: 0:00:03 lr: 0.000288 loss: 0.4507 (0.4404) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:31.045517] Epoch: [20] [60/78] eta: 0:00:01 lr: 0.000275 loss: 0.4168 (0.4373) time: 0.0666 data: 0.0002 max mem: 9669 +[14:50:32.158248] Epoch: [20] [77/78] eta: 0:00:00 lr: 0.000265 loss: 0.4511 (0.4446) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:32.245166] Epoch: [20] Total time: 0:00:06 (0.0818 s / it) +[14:50:32.246088] Averaged stats: lr: 0.000265 loss: 0.4511 (0.4446) +[14:50:32.985494] val: [ 0/17] eta: 0:00:12 loss: 0.2032 (0.2032) time: 0.7324 data: 0.7151 max mem: 9669 +[14:50:33.157372] val: [10/17] eta: 0:00:00 loss: 0.3213 (0.2785) time: 0.0821 data: 0.0666 max mem: 9669 +[14:50:33.252405] val: [16/17] eta: 0:00:00 loss: 0.2715 (0.2603) time: 0.0587 data: 0.0433 max mem: 9669 +[14:50:33.337081] val: Total time: 0:00:01 (0.0638 s / it) +[14:50:33.350990] val loss: 0.2602909899809781 +[14:50:33.351176] Accuracy: 0.9000, F1 Score: 0.9000, ROC AUC: 0.9615, Hamming Loss: 0.1000, + Jaccard Score: 0.8182, Precision: 0.9001, Recall: 0.9000, + Average Precision: 0.9612, Kappa: 0.8000, Score: 0.8872 +[14:50:33.392557] Best epoch = 16, Best score = 0.8918 +[14:50:33.665173] log_dir: ./output_logs/retfound +[14:50:34.613194] Epoch: [21] [ 0/78] eta: 0:01:13 lr: 0.000264 loss: 0.5421 (0.5421) time: 0.9470 data: 0.8740 max mem: 9669 +[14:50:35.950237] Epoch: [21] [20/78] eta: 0:00:06 lr: 0.000252 loss: 0.4251 (0.4567) time: 0.0668 data: 0.0001 max mem: 9669 +[14:50:37.260158] Epoch: [21] [40/78] eta: 0:00:03 lr: 0.000240 loss: 0.4189 (0.4429) time: 0.0655 data: 0.0001 max mem: 9669 +[14:50:38.573137] Epoch: [21] [60/78] eta: 0:00:01 lr: 0.000227 loss: 0.3688 (0.4218) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:39.688747] Epoch: [21] [77/78] eta: 0:00:00 lr: 0.000217 loss: 0.4216 (0.4247) time: 0.0655 data: 0.0001 max mem: 9669 +[14:50:39.774763] Epoch: [21] Total time: 0:00:06 (0.0783 s / it) +[14:50:39.775598] Averaged stats: lr: 0.000217 loss: 0.4216 (0.4247) +[14:50:40.469977] val: [ 0/17] eta: 0:00:11 loss: 0.2748 (0.2748) time: 0.6758 data: 0.6528 max mem: 9669 +[14:50:40.780492] val: [10/17] eta: 0:00:00 loss: 0.3254 (0.3128) time: 0.0896 data: 0.0735 max mem: 9669 +[14:50:40.872592] val: [16/17] eta: 0:00:00 loss: 0.2549 (0.2715) time: 0.0634 data: 0.0476 max mem: 9669 +[14:50:40.949919] val: Total time: 0:00:01 (0.0680 s / it) +[14:50:40.965639] val loss: 0.27147211572703195 +[14:50:40.965840] Accuracy: 0.9019, F1 Score: 0.9018, ROC AUC: 0.9627, Hamming Loss: 0.0981, + Jaccard Score: 0.8212, Precision: 0.9023, Recall: 0.9019, + Average Precision: 0.9637, Kappa: 0.8037, Score: 0.8894 +[14:50:41.008052] Best epoch = 16, Best score = 0.8918 +[14:50:41.282889] log_dir: ./output_logs/retfound +[14:50:42.266998] Epoch: [22] [ 0/78] eta: 0:01:16 lr: 0.000217 loss: 0.3375 (0.3375) time: 0.9832 data: 0.9065 max mem: 9669 +[14:50:43.588406] Epoch: [22] [20/78] eta: 0:00:06 lr: 0.000205 loss: 0.3641 (0.3946) time: 0.0660 data: 0.0001 max mem: 9669 +[14:50:44.906910] Epoch: [22] [40/78] eta: 0:00:03 lr: 0.000193 loss: 0.3990 (0.4051) time: 0.0659 data: 0.0001 max mem: 9669 +[14:50:46.229365] Epoch: [22] [60/78] eta: 0:00:01 lr: 0.000182 loss: 0.4305 (0.4234) time: 0.0661 data: 0.0001 max mem: 9669 +[14:50:47.349029] Epoch: [22] [77/78] eta: 0:00:00 lr: 0.000172 loss: 0.4409 (0.4291) time: 0.0659 data: 0.0001 max mem: 9669 +[14:50:47.445241] Epoch: [22] Total time: 0:00:06 (0.0790 s / it) +[14:50:47.445985] Averaged stats: lr: 0.000172 loss: 0.4409 (0.4291) +[14:50:48.433109] val: [ 0/17] eta: 0:00:16 loss: 0.3709 (0.3709) time: 0.9673 data: 0.9498 max mem: 9669 +[14:50:48.937444] val: [10/17] eta: 0:00:00 loss: 0.3709 (0.3596) time: 0.1337 data: 0.1181 max mem: 9669 +[14:50:49.029603] val: [16/17] eta: 0:00:00 loss: 0.1637 (0.2724) time: 0.0919 data: 0.0765 max mem: 9669 +[14:50:49.112452] val: Total time: 0:00:01 (0.0969 s / it) +[14:50:49.126831] val loss: 0.2723842714639271 +[14:50:49.127019] Accuracy: 0.9093, F1 Score: 0.9091, ROC AUC: 0.9634, Hamming Loss: 0.0907, + Jaccard Score: 0.8333, Precision: 0.9128, Recall: 0.9093, + Average Precision: 0.9637, Kappa: 0.8185, Score: 0.8970 +[14:50:51.207431] Best epoch = 22, Best score = 0.8970 +[14:50:51.294257] log_dir: ./output_logs/retfound +[14:50:52.099254] Epoch: [23] [ 0/78] eta: 0:01:02 lr: 0.000171 loss: 0.4316 (0.4316) time: 0.8039 data: 0.7208 max mem: 9669 +[14:50:53.415934] Epoch: [23] [20/78] eta: 0:00:05 lr: 0.000160 loss: 0.4234 (0.4163) time: 0.0658 data: 0.0001 max mem: 9669 +[14:50:54.735551] Epoch: [23] [40/78] eta: 0:00:03 lr: 0.000149 loss: 0.4082 (0.4143) time: 0.0659 data: 0.0001 max mem: 9669 +[14:50:56.050407] Epoch: [23] [60/78] eta: 0:00:01 lr: 0.000139 loss: 0.4038 (0.4195) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:57.168547] Epoch: [23] [77/78] eta: 0:00:00 lr: 0.000130 loss: 0.4251 (0.4213) time: 0.0657 data: 0.0001 max mem: 9669 +[14:50:57.255301] Epoch: [23] Total time: 0:00:05 (0.0764 s / it) +[14:50:57.256426] Averaged stats: lr: 0.000130 loss: 0.4251 (0.4213) +[14:50:58.030562] val: [ 0/17] eta: 0:00:12 loss: 0.3098 (0.3098) time: 0.7514 data: 0.7324 max mem: 9669 +[14:50:58.295855] val: [10/17] eta: 0:00:00 loss: 0.3098 (0.3497) time: 0.0924 data: 0.0766 max mem: 9669 +[14:50:58.388497] val: [16/17] eta: 0:00:00 loss: 0.2291 (0.2791) time: 0.0652 data: 0.0496 max mem: 9669 +[14:50:58.485867] val: Total time: 0:00:01 (0.0710 s / it) +[14:50:58.504249] val loss: 0.279085652793155 +[14:50:58.504439] Accuracy: 0.9167, F1 Score: 0.9166, ROC AUC: 0.9635, Hamming Loss: 0.0833, + Jaccard Score: 0.8461, Precision: 0.9176, Recall: 0.9167, + Average Precision: 0.9633, Kappa: 0.8333, Score: 0.9045 +[14:51:00.293207] Best epoch = 23, Best score = 0.9045 +[14:51:00.395191] log_dir: ./output_logs/retfound +[14:51:01.693338] Epoch: [24] [ 0/78] eta: 0:01:41 lr: 0.000130 loss: 0.6623 (0.6623) time: 1.2972 data: 1.2283 max mem: 9669 +[14:51:03.006921] Epoch: [24] [20/78] eta: 0:00:07 lr: 0.000120 loss: 0.3732 (0.4278) time: 0.0656 data: 0.0001 max mem: 9669 +[14:51:04.317690] Epoch: [24] [40/78] eta: 0:00:03 lr: 0.000110 loss: 0.4038 (0.4267) time: 0.0655 data: 0.0001 max mem: 9669 +[14:51:05.632843] Epoch: [24] [60/78] eta: 0:00:01 lr: 0.000101 loss: 0.3981 (0.4207) time: 0.0657 data: 0.0001 max mem: 9669 +[14:51:06.746080] Epoch: [24] [77/78] eta: 0:00:00 lr: 0.000093 loss: 0.4013 (0.4206) time: 0.0654 data: 0.0001 max mem: 9669 +[14:51:06.828894] Epoch: [24] Total time: 0:00:06 (0.0825 s / it) +[14:51:06.829697] Averaged stats: lr: 0.000093 loss: 0.4013 (0.4206) +[14:51:07.961505] val: [ 0/17] eta: 0:00:18 loss: 0.2105 (0.2105) time: 1.1014 data: 1.0822 max mem: 9669 +[14:51:08.218434] val: [10/17] eta: 0:00:00 loss: 0.3270 (0.2843) time: 0.1234 data: 0.1076 max mem: 9669 +[14:51:08.310849] val: [16/17] eta: 0:00:00 loss: 0.2778 (0.2697) time: 0.0853 data: 0.0696 max mem: 9669 +[14:51:08.391442] val: Total time: 0:00:01 (0.0901 s / it) +[14:51:08.408147] val loss: 0.26972537093302784 +[14:51:08.408329] Accuracy: 0.8926, F1 Score: 0.8926, ROC AUC: 0.9626, Hamming Loss: 0.1074, + Jaccard Score: 0.8060, Precision: 0.8926, Recall: 0.8926, + Average Precision: 0.9631, Kappa: 0.7852, Score: 0.8801 +[14:51:08.491865] Best epoch = 23, Best score = 0.9045 +[14:51:08.785853] log_dir: ./output_logs/retfound +[14:51:09.925488] Epoch: [25] [ 0/78] eta: 0:01:28 lr: 0.000092 loss: 0.2670 (0.2670) time: 1.1388 data: 1.0680 max mem: 9669 +[14:51:11.248244] Epoch: [25] [20/78] eta: 0:00:06 lr: 0.000084 loss: 0.4132 (0.4202) time: 0.0661 data: 0.0001 max mem: 9669 +[14:51:12.575345] Epoch: [25] [40/78] eta: 0:00:03 lr: 0.000075 loss: 0.4054 (0.4168) time: 0.0663 data: 0.0001 max mem: 9669 +[14:51:13.892979] Epoch: [25] [60/78] eta: 0:00:01 lr: 0.000067 loss: 0.3792 (0.4125) time: 0.0658 data: 0.0001 max mem: 9669 +[14:51:15.007124] Epoch: [25] [77/78] eta: 0:00:00 lr: 0.000061 loss: 0.3720 (0.4059) time: 0.0655 data: 0.0001 max mem: 9669 +[14:51:15.095251] Epoch: [25] Total time: 0:00:06 (0.0809 s / it) +[14:51:15.096070] Averaged stats: lr: 0.000061 loss: 0.3720 (0.4059) +[14:51:16.411356] val: [ 0/17] eta: 0:00:21 loss: 0.2106 (0.2106) time: 1.2555 data: 1.2374 max mem: 9669 +[14:51:17.017518] val: [10/17] eta: 0:00:01 loss: 0.3230 (0.2925) time: 0.1692 data: 0.1535 max mem: 9669 +[14:51:17.112169] val: [16/17] eta: 0:00:00 loss: 0.3024 (0.2818) time: 0.1150 data: 0.0995 max mem: 9669 +[14:51:17.199109] val: Total time: 0:00:02 (0.1202 s / it) +[14:51:17.213294] val loss: 0.2818167503265774 +[14:51:17.213531] Accuracy: 0.8944, F1 Score: 0.8944, ROC AUC: 0.9620, Hamming Loss: 0.1056, + Jaccard Score: 0.8090, Precision: 0.8944, Recall: 0.8944, + Average Precision: 0.9626, Kappa: 0.7889, Score: 0.8818 +[14:51:17.294625] Best epoch = 23, Best score = 0.9045 +[14:51:17.558590] log_dir: ./output_logs/retfound +[14:51:18.891768] Epoch: [26] [ 0/78] eta: 0:01:43 lr: 0.000061 loss: 0.4083 (0.4083) time: 1.3322 data: 1.2604 max mem: 9669 +[14:51:20.275017] Epoch: [26] [20/78] eta: 0:00:07 lr: 0.000053 loss: 0.4188 (0.4210) time: 0.0691 data: 0.0033 max mem: 9669 +[14:51:21.607694] Epoch: [26] [40/78] eta: 0:00:03 lr: 0.000047 loss: 0.3950 (0.4119) time: 0.0666 data: 0.0008 max mem: 9669 +[14:51:22.934905] Epoch: [26] [60/78] eta: 0:00:01 lr: 0.000040 loss: 0.4428 (0.4201) time: 0.0663 data: 0.0010 max mem: 9669 +[14:51:24.050830] Epoch: [26] [77/78] eta: 0:00:00 lr: 0.000035 loss: 0.4156 (0.4169) time: 0.0664 data: 0.0010 max mem: 9669 +[14:51:24.147255] Epoch: [26] Total time: 0:00:06 (0.0845 s / it) +[14:51:24.148107] Averaged stats: lr: 0.000035 loss: 0.4156 (0.4169) +[14:51:24.936130] val: [ 0/17] eta: 0:00:13 loss: 0.2425 (0.2425) time: 0.7762 data: 0.7589 max mem: 9669 +[14:51:25.243014] val: [10/17] eta: 0:00:00 loss: 0.3265 (0.3008) time: 0.0984 data: 0.0828 max mem: 9669 +[14:51:25.374561] val: [16/17] eta: 0:00:00 loss: 0.2425 (0.2663) time: 0.0714 data: 0.0560 max mem: 9669 +[14:51:25.455941] val: Total time: 0:00:01 (0.0763 s / it) +[14:51:25.470673] val loss: 0.2663208666969748 +[14:51:25.470882] Accuracy: 0.9056, F1 Score: 0.9055, ROC AUC: 0.9630, Hamming Loss: 0.0944, + Jaccard Score: 0.8274, Precision: 0.9058, Recall: 0.9056, + Average Precision: 0.9634, Kappa: 0.8111, Score: 0.8932 +[14:51:25.507117] Best epoch = 23, Best score = 0.9045 +[14:51:25.782694] log_dir: ./output_logs/retfound +[14:51:26.629037] Epoch: [27] [ 0/78] eta: 0:01:05 lr: 0.000035 loss: 0.3796 (0.3796) time: 0.8454 data: 0.7733 max mem: 9669 +[14:51:28.237934] Epoch: [27] [20/78] eta: 0:00:06 lr: 0.000030 loss: 0.3763 (0.3915) time: 0.0804 data: 0.0001 max mem: 9669 +[14:51:29.550258] Epoch: [27] [40/78] eta: 0:00:03 lr: 0.000025 loss: 0.3293 (0.3750) time: 0.0656 data: 0.0001 max mem: 9669 +[14:51:30.871484] Epoch: [27] [60/78] eta: 0:00:01 lr: 0.000020 loss: 0.3856 (0.3828) time: 0.0660 data: 0.0001 max mem: 9669 +[14:51:32.011372] Epoch: [27] [77/78] eta: 0:00:00 lr: 0.000016 loss: 0.4332 (0.3930) time: 0.0668 data: 0.0001 max mem: 9669 +[14:51:32.105912] Epoch: [27] Total time: 0:00:06 (0.0811 s / it) +[14:51:32.106707] Averaged stats: lr: 0.000016 loss: 0.4332 (0.3930) +[14:51:32.808277] val: [ 0/17] eta: 0:00:11 loss: 0.2318 (0.2318) time: 0.6840 data: 0.6666 max mem: 9669 +[14:51:33.074709] val: [10/17] eta: 0:00:00 loss: 0.3333 (0.3026) time: 0.0863 data: 0.0707 max mem: 9669 +[14:51:33.241615] val: [16/17] eta: 0:00:00 loss: 0.2767 (0.2758) time: 0.0657 data: 0.0502 max mem: 9669 +[14:51:33.328314] val: Total time: 0:00:01 (0.0709 s / it) +[14:51:33.342466] val loss: 0.2758449465036392 +[14:51:33.342654] Accuracy: 0.9000, F1 Score: 0.9000, ROC AUC: 0.9628, Hamming Loss: 0.1000, + Jaccard Score: 0.8182, Precision: 0.9001, Recall: 0.9000, + Average Precision: 0.9633, Kappa: 0.8000, Score: 0.8876 +[14:51:33.399761] Best epoch = 23, Best score = 0.9045 +[14:51:33.702845] log_dir: ./output_logs/retfound +[14:51:34.556738] Epoch: [28] [ 0/78] eta: 0:01:06 lr: 0.000016 loss: 0.3553 (0.3553) time: 0.8530 data: 0.7800 max mem: 9669 +[14:51:35.871165] Epoch: [28] [20/78] eta: 0:00:05 lr: 0.000013 loss: 0.4099 (0.4038) time: 0.0657 data: 0.0001 max mem: 9669 +[14:51:37.190809] Epoch: [28] [40/78] eta: 0:00:03 lr: 0.000009 loss: 0.3861 (0.4038) time: 0.0659 data: 0.0001 max mem: 9669 +[14:51:38.506612] Epoch: [28] [60/78] eta: 0:00:01 lr: 0.000007 loss: 0.3833 (0.4040) time: 0.0658 data: 0.0001 max mem: 9669 +[14:51:39.618724] Epoch: [28] [77/78] eta: 0:00:00 lr: 0.000005 loss: 0.3475 (0.3990) time: 0.0654 data: 0.0001 max mem: 9669 +[14:51:39.712592] Epoch: [28] Total time: 0:00:06 (0.0770 s / it) +[14:51:39.713386] Averaged stats: lr: 0.000005 loss: 0.3475 (0.3990) +[14:51:40.306888] val: [ 0/17] eta: 0:00:09 loss: 0.2289 (0.2289) time: 0.5821 data: 0.5619 max mem: 9669 +[14:51:40.465658] val: [10/17] eta: 0:00:00 loss: 0.3318 (0.3012) time: 0.0673 data: 0.0513 max mem: 9669 +[14:51:40.557411] val: [16/17] eta: 0:00:00 loss: 0.2787 (0.2762) time: 0.0489 data: 0.0332 max mem: 9669 +[14:51:40.639575] val: Total time: 0:00:00 (0.0539 s / it) +[14:51:40.653857] val loss: 0.27622606123194976 +[14:51:40.654029] Accuracy: 0.8981, F1 Score: 0.8981, ROC AUC: 0.9626, Hamming Loss: 0.1019, + Jaccard Score: 0.8151, Precision: 0.8982, Recall: 0.8981, + Average Precision: 0.9630, Kappa: 0.7963, Score: 0.8857 +[14:51:40.683159] Best epoch = 23, Best score = 0.9045 +[14:51:40.955355] log_dir: ./output_logs/retfound +[14:51:41.778252] Epoch: [29] [ 0/78] eta: 0:01:04 lr: 0.000005 loss: 0.5878 (0.5878) time: 0.8219 data: 0.7514 max mem: 9669 +[14:51:43.094151] Epoch: [29] [20/78] eta: 0:00:05 lr: 0.000003 loss: 0.3694 (0.4034) time: 0.0658 data: 0.0001 max mem: 9669 +[14:51:44.434987] Epoch: [29] [40/78] eta: 0:00:03 lr: 0.000002 loss: 0.4345 (0.4129) time: 0.0670 data: 0.0001 max mem: 9669 +[14:51:45.761184] Epoch: [29] [60/78] eta: 0:00:01 lr: 0.000001 loss: 0.4415 (0.4238) time: 0.0663 data: 0.0001 max mem: 9669 +[14:51:46.887122] Epoch: [29] [77/78] eta: 0:00:00 lr: 0.000001 loss: 0.3984 (0.4167) time: 0.0661 data: 0.0001 max mem: 9669 +[14:51:46.975837] Epoch: [29] Total time: 0:00:06 (0.0772 s / it) +[14:51:46.976670] Averaged stats: lr: 0.000001 loss: 0.3984 (0.4167) +[14:51:47.762010] val: [ 0/17] eta: 0:00:12 loss: 0.2329 (0.2329) time: 0.7619 data: 0.7443 max mem: 9669 +[14:51:47.986490] val: [10/17] eta: 0:00:00 loss: 0.3294 (0.3025) time: 0.0896 data: 0.0740 max mem: 9669 +[14:51:48.078530] val: [16/17] eta: 0:00:00 loss: 0.2721 (0.2752) time: 0.0634 data: 0.0479 max mem: 9669 +[14:51:48.160586] val: Total time: 0:00:01 (0.0683 s / it) +[14:51:48.175714] val loss: 0.27518679902834053 +[14:51:48.175912] Accuracy: 0.9037, F1 Score: 0.9037, ROC AUC: 0.9628, Hamming Loss: 0.0963, + Jaccard Score: 0.8243, Precision: 0.9039, Recall: 0.9037, + Average Precision: 0.9633, Kappa: 0.8074, Score: 0.8913 +[14:51:48.222472] Best epoch = 23, Best score = 0.9045 +[14:51:51.253807] Test with the best model, epoch = 23: +[14:51:51.999463] test: [ 0/32] eta: 0:00:23 loss: 0.1566 (0.1566) time: 0.7355 data: 0.7172 max mem: 9669 +[14:51:52.259743] test: [10/32] eta: 0:00:01 loss: 0.2990 (0.2927) time: 0.0905 data: 0.0746 max mem: 9669 +[14:51:52.502473] test: [20/32] eta: 0:00:00 loss: 0.2658 (0.2841) time: 0.0251 data: 0.0095 max mem: 9669 +[14:51:52.750830] test: [30/32] eta: 0:00:00 loss: 0.2383 (0.2634) time: 0.0245 data: 0.0089 max mem: 9669 +[14:51:52.805721] test: [31/32] eta: 0:00:00 loss: 0.2547 (0.2774) time: 0.0264 data: 0.0089 max mem: 9669 +[14:51:52.884255] test: Total time: 0:00:01 (0.0507 s / it) +[14:51:52.904628] val loss: 0.27743660961277783 +[14:51:52.904750] Accuracy: 0.8980, F1 Score: 0.8980, ROC AUC: 0.9646, Hamming Loss: 0.1020, + Jaccard Score: 0.8149, Precision: 0.8980, Recall: 0.8980, + Average Precision: 0.9642, Kappa: 0.7960, Score: 0.8862 +[14:51:53.653624] Training time 0:04:29 +[rank0]:[W701 14:51:54.195152194 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/050/retfound acc=0.8980 auroc=0.9645199999999999 f1_macro=0.8980 qwk=0.796 diff --git a/results/downsample/airogs/050/vit/confusion_matrix.png b/results/downsample/airogs/050/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..25a12e87e12ebc7248451bab650283cdf2eb29b5 --- /dev/null +++ b/results/downsample/airogs/050/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9f4bf3292afcb2227e0566d3e5d50e5e19ebc1abdef9513906ebd79701b3569 +size 72624 diff --git a/results/downsample/airogs/050/vit/log.csv b/results/downsample/airogs/050/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..bc8374fc67c5146c3b8ef0580eb53da6a8f3dd8a --- /dev/null +++ b/results/downsample/airogs/050/vit/log.csv @@ -0,0 +1,24 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7207764219015073,0.6648148148148149,0.7178395061728396,0.5689068309451696,7.204811479933906e-08 +1,0.6228443796818073,0.7222222222222222,0.8066186556927297,0.6568951287859404,1.4599223261971336e-07 +2,0.5467187074514536,0.774074074074074,0.8618038408779151,0.7278596354645677,2.1993635044008766e-07 +3,0.5197362991479727,0.7648148148148148,0.8538477366255145,0.7150840928771749,2.93880468260462e-07 +4,0.505318514811687,0.7925925925925926,0.8893072702331961,0.7554513790146142,3.6782458608083625e-07 +5,0.4496602843969296,0.8148148148148148,0.9031687242798354,0.7824690595098044,3.683366126261768e-07 +6,0.4332921122893309,0.85,0.9109259259259259,0.8200727953565808,3.6406006464092253e-07 +7,0.4220382617070125,0.8444444444444444,0.9181961591220851,0.8165713349460061,3.5695742923259533e-07 +8,0.3877863563024081,0.8481481481481481,0.924931412894376,0.8227221615669588,3.471407192044933e-07 +9,0.37152315485171783,0.837037037037037,0.9277229080932785,0.8129178346354943,3.347647499379779e-07 +10,0.3655512776130285,0.8555555555555555,0.9299314128943759,0.8321458024030156,3.200246978614278e-07 +11,0.3579256916657472,0.8537037037037037,0.9274074074074075,0.8293449295781169,3.031530224013705e-07 +12,0.335525618149684,0.8666666666666667,0.9290603566529493,0.8428634951056454,2.844157999584614e-07 +13,0.3179920399800325,0.8537037037037037,0.9235253772290809,0.8278711457723803,2.6410852772369945e-07 +14,0.3128989063776456,0.8555555555555555,0.9288408779149517,0.8317935410726487,2.4255146351120325e-07 +15,0.2988759336563257,0.8462962962962963,0.92781207133059,0.8216146469165905,2.200845751011696e-07 +16,0.29602126738963985,0.8555555555555555,0.9295473251028806,0.8318563834999021,1.970621787448937e-07 +17,0.2860643454851248,0.8629629629629629,0.9289506172839506,0.8390525076988528,1.7384735138582672e-07 +18,0.2848845506325746,0.8592592592592593,0.9278737997256515,0.8349583589549532,1.508062047192868e-07 +19,0.2829464028278987,0.8592592592592593,0.9273799725651577,0.8346468970048321,1.2830211139231923e-07 +20,0.28747899524676496,0.8555555555555555,0.9298216735253773,0.8320966424978771,1.0668997439998408e-07 +21,0.2718046410725667,0.8574074074074074,0.927565157750343,0.8330382601187312,8.631063005311509e-08 +22,0.2614507529980097,0.8592592592592593,0.9269410150891633,0.8346727926052084,6.748547278609277e-08 diff --git a/results/downsample/airogs/050/vit/metrics.json b/results/downsample/airogs/050/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..914a0c60f262ab8033f8f3c260320973622a142b --- /dev/null +++ b/results/downsample/airogs/050/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.863, + "balanced_accuracy": 0.863, + "precision_macro": 0.863001452005808, + "recall_macro": 0.863, + "f1_macro": 0.862999862999863, + "precision_weighted": 0.863001452005808, + "recall_weighted": 0.863, + "f1_weighted": 0.8629998629998629, + "cohen_kappa": 0.726, + "quadratic_weighted_kappa": 0.726, + "mcc": 0.7260014520043561, + "auroc": 0.939084, + "auprc": 0.9395471636357733, + "sensitivity": 0.862, + "specificity": 0.864, + "precision_pos": 0.8637274549098196, + "f1_pos": 0.8628628628628628, + "per_class": { + "0": { + "precision": 0.8622754491017964, + "recall": 0.864, + "f1-score": 0.8631368631368631, + "support": 500.0 + }, + "1": { + "precision": 0.8637274549098196, + "recall": 0.862, + "f1-score": 0.8628628628628628, + "support": 500.0 + }, + "accuracy": 0.863, + "macro avg": { + "precision": 0.863001452005808, + "recall": 0.863, + "f1-score": 0.862999862999863, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.863001452005808, + "recall": 0.863, + "f1-score": 0.8629998629998629, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/050/vit/pr.png b/results/downsample/airogs/050/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..858ac4eefbdd40b7f52af75fd4b336a2eeefd1a5 --- /dev/null +++ b/results/downsample/airogs/050/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86b2b8e45a7b14bdf34601fa5dfd063b377b57e93cb4b0b242795f63df21c693 +size 47198 diff --git a/results/downsample/airogs/050/vit/roc.png b/results/downsample/airogs/050/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..b072542e06bec4ff3d204292dc8a1c043db3a3e9 --- /dev/null +++ b/results/downsample/airogs/050/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66530a0eb51b6f5020c4ee3d3be0ace85f1c497f19e7c3fa2a79343dbd0676ba +size 62739 diff --git a/results/downsample/airogs/050/vit/test_pred.npz b/results/downsample/airogs/050/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..6bd11ef37b4d1a42bd90d25a86e94af7d9d4e630 --- /dev/null +++ b/results/downsample/airogs/050/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5d13c674136606990149de3d471435e9ea25e1ae40b6b0cfb8d1d9eade04501 +size 16510 diff --git a/results/downsample/airogs/050/vit/train.log b/results/downsample/airogs/050/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..11d1094a4bb32b36ee32eda53f87edff407e4f63 --- /dev/null +++ b/results/downsample/airogs/050/vit/train.log @@ -0,0 +1,124 @@ +[vit] train=2500 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7208 val_acc=0.6648 val_auc=0.7178 score=0.5689 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.6228 val_acc=0.7222 val_auc=0.8066 score=0.6569 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.5467 val_acc=0.7741 val_auc=0.8618 score=0.7279 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.5197 val_acc=0.7648 val_auc=0.8538 score=0.7151 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5053 val_acc=0.7926 val_auc=0.8893 score=0.7555 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.4497 val_acc=0.8148 val_auc=0.9032 score=0.7825 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.4333 val_acc=0.8500 val_auc=0.9109 score=0.8201 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4220 val_acc=0.8444 val_auc=0.9182 score=0.8166 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3878 val_acc=0.8481 val_auc=0.9249 score=0.8227 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.3715 val_acc=0.8370 val_auc=0.9277 score=0.8129 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3656 val_acc=0.8556 val_auc=0.9299 score=0.8321 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3579 val_acc=0.8537 val_auc=0.9274 score=0.8293 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3355 val_acc=0.8667 val_auc=0.9291 score=0.8429 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.3180 val_acc=0.8537 val_auc=0.9235 score=0.8279 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.3129 val_acc=0.8556 val_auc=0.9288 score=0.8318 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.2989 val_acc=0.8463 val_auc=0.9278 score=0.8216 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.2960 val_acc=0.8556 val_auc=0.9295 score=0.8319 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.2861 val_acc=0.8630 val_auc=0.9290 score=0.8391 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.2849 val_acc=0.8593 val_auc=0.9279 score=0.8350 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.2829 val_acc=0.8593 val_auc=0.9274 score=0.8346 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.2875 val_acc=0.8556 val_auc=0.9298 score=0.8321 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.2718 val_acc=0.8574 val_auc=0.9276 score=0.8330 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.2615 val_acc=0.8593 val_auc=0.9269 score=0.8347 +[vit] early stop at ep22 (best ep12 score=0.8429) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=12 best_val_score=0.8429 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/050/vit acc=0.8630 auroc=0.939084 f1_macro=0.8630 qwk=0.726 diff --git a/results/downsample/airogs/100/resnet/confusion_matrix.png b/results/downsample/airogs/100/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..e926c17f33dfeb2c429c9445d4766b2c22fa72ec --- /dev/null +++ b/results/downsample/airogs/100/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bcd957f91ca76757f8a8a25870be1913cc7dedd56c7bb349ef867a9d6632e491 +size 65340 diff --git a/results/downsample/airogs/100/resnet/log.csv b/results/downsample/airogs/100/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..72ece0428a6b431d4409615e9cb5eda689e9ce66 --- /dev/null +++ b/results/downsample/airogs/100/resnet/log.csv @@ -0,0 +1,25 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6806929661677434,0.6611111111111111,0.7466803840877915,0.5753984269543008,0.00016452991452991454 +1,0.49043761079127973,0.8277777777777777,0.900212620027435,0.7943492708902502,0.0003311965811965812 +2,0.3247737426024217,0.8574074074074074,0.9353840877914952,0.835796778030733,0.0004978632478632479 +3,0.26725975137490493,0.8611111111111112,0.9449725651577504,0.8426526020125761,0.0004983526080898481 +4,0.23663267607872301,0.85,0.9502812071330591,0.8327371225504075,0.0004933469513590679 +5,0.21497210936668593,0.85,0.960034293552812,0.8357473364395475,0.00048505044456893006 +6,0.19433321918432528,0.8981481481481481,0.9619958847736626,0.8854538930464857,0.00047357528374124746 +7,0.17887980132721937,0.9092592592592592,0.9660219478737997,0.8979157017855338,0.00045907665074076113 +8,0.15023378086968875,0.8870370370370371,0.9568312757201648,0.8724291360770836,0.0004417506147066312 +9,0.13541560491117147,0.8851851851851852,0.9554938271604939,0.8703492692437594,0.0004218314805536838 +10,0.13837236132568273,0.8944444444444445,0.9600480109739369,0.8810389313475734,0.00039958862040038377 +11,0.10960012932236378,0.8925925925925926,0.9630178326474622,0.8802411225292088,0.0003753228307730364 +12,0.09871907835491957,0.8981481481481481,0.9675994513031551,0.8872965424107431,0.0003493622648487805 +13,0.08899849906372718,0.9203703703703704,0.970761316872428,0.910619681632804,0.00032205799474680896 +14,0.07103978868764944,0.9203703703703704,0.9648765432098766,0.9086515327103518,0.00029377926388021573 +15,0.06878597263055734,0.9037037037037037,0.9654115226337447,0.8921671647051905,0.0002649084935722644 +16,0.060439678792579055,0.9055555555555556,0.9682235939643348,0.894915737214674,0.00023583611146402853 +17,0.054122735579044394,0.9,0.966275720164609,0.8887512558242201,0.00020695527165031925 +18,0.047834626685541407,0.9,0.9669135802469135,0.8889707361619946,0.00017865653794500142 +19,0.03888522723737436,0.9166666666666666,0.9695404663923182,0.9065133935372437,0.0001513226021754179 +20,0.03414216463286907,0.9111111111111111,0.9685048010973938,0.9005720114362091,0.00012532310893192616 +21,0.027899769230339773,0.9018518518518519,0.9663100137174211,0.8906163579474904,0.00010100965675893233 +22,0.02533931263650839,0.9055555555555556,0.9656652949245542,0.8941079546027026,7.871104338774113e-05 +23,0.02096657249598931,0.9074074074074074,0.9622839506172839,0.8948201418370502,5.872881931128568e-05 diff --git a/results/downsample/airogs/100/resnet/metrics.json b/results/downsample/airogs/100/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..b286d060153767706ecae6aca76ed9c2f7cb42f7 --- /dev/null +++ b/results/downsample/airogs/100/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.898, + "balanced_accuracy": 0.898, + "precision_macro": 0.8980254736303124, + "recall_macro": 0.898, + "f1_macro": 0.8979983679738875, + "precision_weighted": 0.8980254736303124, + "recall_weighted": 0.898, + "f1_weighted": 0.8979983679738875, + "cohen_kappa": 0.796, + "quadratic_weighted_kappa": 0.796, + "mcc": 0.7960254732227212, + "auroc": 0.963978, + "auprc": 0.9632216437603003, + "sensitivity": 0.894, + "specificity": 0.902, + "precision_pos": 0.9012096774193549, + "f1_pos": 0.8975903614457831, + "per_class": { + "0": { + "precision": 0.8948412698412699, + "recall": 0.902, + "f1-score": 0.898406374501992, + "support": 500.0 + }, + "1": { + "precision": 0.9012096774193549, + "recall": 0.894, + "f1-score": 0.8975903614457831, + "support": 500.0 + }, + "accuracy": 0.898, + "macro avg": { + "precision": 0.8980254736303124, + "recall": 0.898, + "f1-score": 0.8979983679738875, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8980254736303124, + "recall": 0.898, + "f1-score": 0.8979983679738875, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/100/resnet/pr.png b/results/downsample/airogs/100/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..9febc6615b640097baf9d8b2a4b35be8cc73e537 --- /dev/null +++ b/results/downsample/airogs/100/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:62bfa75030b149e57d05feea536e6605578c429410bed73a70754ec81376750e +size 44818 diff --git a/results/downsample/airogs/100/resnet/roc.png b/results/downsample/airogs/100/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..378f32c89e63e4103cfda15dce863cfaff510fa1 --- /dev/null +++ b/results/downsample/airogs/100/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5bfbe08e16ed3ff3a3a8863337918061dfa6a317fe12797c0726aa985789945 +size 60701 diff --git a/results/downsample/airogs/100/resnet/test_pred.npz b/results/downsample/airogs/100/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..f97737c38806dc58d6757974b0443ada7c99a5b2 --- /dev/null +++ b/results/downsample/airogs/100/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c08f1270a1965507d5c29f5b97d8debea4bc8884084e983fef71c72eab21e3c +size 16510 diff --git a/results/downsample/airogs/100/resnet/train.log b/results/downsample/airogs/100/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e9fa38e0ae1a9dd6a9161dfe318e1ceec9de0662 --- /dev/null +++ b/results/downsample/airogs/100/resnet/train.log @@ -0,0 +1,129 @@ +[resnet] train=5000 val=540 test=1000 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6807 val_acc=0.6611 val_auc=0.7467 score=0.5754 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.4904 val_acc=0.8278 val_auc=0.9002 score=0.7943 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.3248 val_acc=0.8574 val_auc=0.9354 score=0.8358 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.2673 val_acc=0.8611 val_auc=0.9450 score=0.8427 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.2366 val_acc=0.8500 val_auc=0.9503 score=0.8327 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.2150 val_acc=0.8500 val_auc=0.9600 score=0.8357 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.1943 val_acc=0.8981 val_auc=0.9620 score=0.8855 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.1789 val_acc=0.9093 val_auc=0.9660 score=0.8979 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.1502 val_acc=0.8870 val_auc=0.9568 score=0.8724 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.1354 val_acc=0.8852 val_auc=0.9555 score=0.8703 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1384 val_acc=0.8944 val_auc=0.9600 score=0.8810 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.1096 val_acc=0.8926 val_auc=0.9630 score=0.8802 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.0987 val_acc=0.8981 val_auc=0.9676 score=0.8873 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.0890 val_acc=0.9204 val_auc=0.9708 score=0.9106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.0710 val_acc=0.9204 val_auc=0.9649 score=0.9087 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0688 val_acc=0.9037 val_auc=0.9654 score=0.8922 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0604 val_acc=0.9056 val_auc=0.9682 score=0.8949 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0541 val_acc=0.9000 val_auc=0.9663 score=0.8888 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0478 val_acc=0.9000 val_auc=0.9669 score=0.8890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0389 val_acc=0.9167 val_auc=0.9695 score=0.9065 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0341 val_acc=0.9111 val_auc=0.9685 score=0.9006 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0279 val_acc=0.9019 val_auc=0.9663 score=0.8906 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0253 val_acc=0.9056 val_auc=0.9657 score=0.8941 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0210 val_acc=0.9074 val_auc=0.9623 score=0.8948 +[resnet] early stop at ep23 (best ep13 score=0.9106) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=13 best_val_score=0.9106 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/100/resnet acc=0.8980 auroc=0.963978 f1_macro=0.8980 qwk=0.796 diff --git a/results/downsample/airogs/100/retfound/confusion_matrix.png b/results/downsample/airogs/100/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7a69b6a5f996d8983b1615ac77be7eff01ed2cf7 --- /dev/null +++ b/results/downsample/airogs/100/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b42880b2ed890eff5ed141947fe5dbe728465da48a286e573325f6b941269d0f +size 66364 diff --git a/results/downsample/airogs/100/retfound/confusion_matrix_test.jpg b/results/downsample/airogs/100/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..25d3f0b897147fab7387e85f28310c7b7f1eac96 --- /dev/null +++ b/results/downsample/airogs/100/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d518ea0035d9db32a50fcbb1896fa44a1e98fc6d69671c646a28a12caeb6a51 +size 260464 diff --git a/results/downsample/airogs/100/retfound/log.txt b/results/downsample/airogs/100/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..0c8d66950394ee444a80aec02df2b748b374444c --- /dev/null +++ b/results/downsample/airogs/100/retfound/log.txt @@ -0,0 +1,30 @@ +{"train_lr": 3.104967948717949e-05, "train_loss": 0.6869127811529697, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.354967948717946e-05, "train_loss": 0.6098002164791791, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.0001560496794871795, "train_loss": 0.5515275139075059, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021854967948717953, "train_loss": 0.5289115004050426, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002810496794871795, "train_loss": 0.5009684100365027, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00034354967948717954, "train_loss": 0.5024133649392005, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00040604967948717954, "train_loss": 0.49966024053402436, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004685496794871794, "train_loss": 0.4900955099325914, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005310496794871795, "train_loss": 0.48253656178712845, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005935496794871795, "train_loss": 0.4739962577437743, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006237308188025447, "train_loss": 0.4772948145102232, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006161042181059726, "train_loss": 0.4518326301223192, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006010141856144047, "train_loss": 0.46381485901581937, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0005788322880087305, "train_loss": 0.4572136507202417, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0005501047172254988, "train_loss": 0.4376392687360446, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005155388413987338, "train_loss": 0.4237176972704056, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0004759857871088555, "train_loss": 0.43037567211267275, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0004324194818215422, "train_loss": 0.4216204109864357, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0003859126725612448, "train_loss": 0.42659685015678406, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0003376105113192029, "train_loss": 0.40269207667845947, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.000288702357610878, "train_loss": 0.4008018556886759, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00024039249249742462, "train_loss": 0.3933946737685265, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0001938704651891872, "train_loss": 0.38881010562181473, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00015028180239625032, "train_loss": 0.381559801980471, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00011069980165905447, "train_loss": 0.38073944586973923, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 7.609910320087817e-05, "train_loss": 0.38250331026621354, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 4.733169105089629e-05, "train_loss": 0.37621724328551537, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 2.5105914369808592e-05, "train_loss": 0.37985354212996286, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 9.969045542627247e-06, "train_loss": 0.37229995333995575, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 2.2938045162649166e-06, "train_loss": 0.36475676097548926, "epoch": 29, "n_parameters": 303303682} diff --git a/results/downsample/airogs/100/retfound/metrics.json b/results/downsample/airogs/100/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..80482b59cdf02c496623c42cd326f98b872d3481 --- /dev/null +++ b/results/downsample/airogs/100/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.908, + "balanced_accuracy": 0.908, + "precision_macro": 0.9084182202575437, + "recall_macro": 0.908, + "f1_macro": 0.9079764419691441, + "precision_weighted": 0.9084182202575437, + "recall_weighted": 0.908, + "f1_weighted": 0.9079764419691442, + "cohen_kappa": 0.8160000000000001, + "quadratic_weighted_kappa": 0.8160000000000001, + "mcc": 0.8164181131383057, + "auroc": 0.970758, + "auprc": 0.9715458663728014, + "sensitivity": 0.892, + "specificity": 0.924, + "precision_pos": 0.9214876033057852, + "f1_pos": 0.9065040650406504, + "per_class": { + "0": { + "precision": 0.8953488372093024, + "recall": 0.924, + "f1-score": 0.9094488188976378, + "support": 500.0 + }, + "1": { + "precision": 0.9214876033057852, + "recall": 0.892, + "f1-score": 0.9065040650406504, + "support": 500.0 + }, + "accuracy": 0.908, + "macro avg": { + "precision": 0.9084182202575437, + "recall": 0.908, + "f1-score": 0.9079764419691441, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.9084182202575437, + "recall": 0.908, + "f1-score": 0.9079764419691442, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/100/retfound/metrics_test.csv b/results/downsample/airogs/100/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..e45cb3eb84acd227645a97cd8b0ba8b686ffe152 --- /dev/null +++ b/results/downsample/airogs/100/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.2516249555628747,0.908,0.9079764419691441,0.9707290000000001,0.092,0.8314656502892113,0.9084182202575437,0.908,0.9703819077045409,0.8160000000000001 diff --git a/results/downsample/airogs/100/retfound/metrics_val.csv b/results/downsample/airogs/100/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..9a223402cb90c004db9232e2f4c8bcfb5b279736 --- /dev/null +++ b/results/downsample/airogs/100/retfound/metrics_val.csv @@ -0,0 +1,31 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6267830974915448,0.7444444444444445,0.7425764463666387,0.8363888888888888,0.25555555555555554,0.5910378348136234,0.7517518083182639,0.7444444444444445,0.8347821851878624,0.48888888888888893 +0.3606555602129768,0.85,0.8499377313631513,0.923576817558299,0.15,0.739048557792073,0.8505818986246032,0.85,0.923376540042333,0.7 +0.3017797531450496,0.8796296296296297,0.8794473122932831,0.9466872427983539,0.12037037037037036,0.7848647643222898,0.8819401316606632,0.8796296296296297,0.9464058847540735,0.7592592592592593 +0.262339744497748,0.8888888888888888,0.8888873647100783,0.9587654320987654,0.1111111111111111,0.7999977777530862,0.8889102282704127,0.8888888888888888,0.9599995369798717,0.7777777777777778 +0.2907406424774843,0.8814814814814815,0.8813186813186813,0.9527469135802469,0.11851851851851852,0.787846856340007,0.8835862068965518,0.8814814814814815,0.9541414159026794,0.762962962962963 +0.27061635694083047,0.8944444444444445,0.8941798941798942,0.9617283950617284,0.10555555555555556,0.8086538461538462,0.8984287317620651,0.8944444444444444,0.962772345352085,0.7888888888888889 +0.25615958650322523,0.8925925925925926,0.8925557461406518,0.9634087791495198,0.10740740740740741,0.8059658073755818,0.8931318681318681,0.8925925925925926,0.9640760453258947,0.7851851851851852 +0.24167559558854385,0.8907407407407407,0.8905752526969799,0.9655418381344307,0.10925925925925926,0.8027625851099454,0.8931188672214632,0.8907407407407407,0.9665968636192291,0.7814814814814814 +0.2527696148437612,0.9018518518518519,0.9017031931006769,0.9657064471879286,0.09814814814814815,0.821023439101615,0.9042976027822631,0.9018518518518519,0.9664421166860533,0.8037037037037037 +0.2629097079967751,0.8907407407407407,0.8905060572214107,0.9655521262002743,0.10925925925925926,0.8026612615027249,0.8941196817710135,0.8907407407407407,0.9669754289687871,0.7814814814814814 +0.2516040574101841,0.9018518518518519,0.9018353564213465,0.9649142661179699,0.09814814814814815,0.8212229806598408,0.9021221397098187,0.9018518518518519,0.9658003944161853,0.8037037037037037 +0.23272023919750662,0.912962962962963,0.9129483349396846,0.9687551440329218,0.08703703703703704,0.8398409381168002,0.9132407242179242,0.912962962962963,0.9692950646828473,0.825925925925926 +0.22826781693626852,0.9037037037037037,0.903638934263085,0.9681035665294924,0.0962962962962963,0.8242260212180388,0.9047920334507042,0.9037037037037037,0.9682865231177928,0.8074074074074074 +0.23409806586363735,0.912962962962963,0.9129483349396846,0.9672187928669411,0.08703703703703704,0.8398409381168002,0.9132407242179242,0.912962962962963,0.9680735164326913,0.825925925925926 +0.23617563791134777,0.9,0.8999876527966415,0.9679526748971193,0.1,0.8181632653061224,0.900197628458498,0.8999999999999999,0.9691229672185688,0.8 +0.22935042880913792,0.9185185185185185,0.9185084578343006,0.9698319615912209,0.08148148148148149,0.8492991613395109,0.9187252964426877,0.9185185185185185,0.9709637428818148,0.837037037037037 +0.23483087297748118,0.9203703703703704,0.9203241933768199,0.968923182441701,0.07962962962962963,0.8524137525020605,0.9213471559582571,0.9203703703703703,0.9700474303549396,0.8407407407407408 +0.2510683304246734,0.9074074074074074,0.9074061372583986,0.967681755829904,0.09259259259259259,0.8305065269350984,0.9074297629499561,0.9074074074074074,0.9684479216576163,0.8148148148148149 +0.23531346899621627,0.9203703703703704,0.9202913724507484,0.9707510288065844,0.07962962962962963,0.8523616018845701,0.9220434920328875,0.9203703703703703,0.9716494849123236,0.8407407407407408 +0.21684648622484767,0.9148148148148149,0.9148148148148149,0.9712688614540466,0.08518518518518518,0.8430034129692833,0.9148148148148149,0.9148148148148149,0.9723009834191337,0.8296296296296296 +0.23225475234143875,0.9111111111111111,0.9111098917680627,0.9698285322359397,0.08888888888888889,0.8367328049979754,0.9111336698858647,0.9111111111111112,0.9704452446619845,0.8222222222222222 +0.23779110676225493,0.9166666666666666,0.9165404469722729,0.9706310013717421,0.08333333333333333,0.8459553402148725,0.9192025835299962,0.9166666666666667,0.9709067553336702,0.8333333333333334 +0.2386372685432434,0.9185185185185185,0.9185084578343006,0.9701268861454047,0.08148148148148149,0.8492991613395109,0.9187252964426877,0.9185185185185185,0.9704915941773309,0.837037037037037 +0.23999586043988957,0.9222222222222223,0.9221699084432609,0.9707338820301783,0.07777777777777778,0.8555865393704509,0.9233604753521127,0.9222222222222223,0.9711377497439626,0.8444444444444444 +0.23967786308597117,0.9203703703703704,0.9203088803088804,0.9703326474622771,0.07962962962962963,0.8523894201328371,0.9216718266253869,0.9203703703703704,0.9701346387665579,0.8407407407407408 +0.2330227932509254,0.9166666666666666,0.9166663808860798,0.9717661179698217,0.08333333333333333,0.846153396605732,0.9166723823372063,0.9166666666666667,0.9719504025999759,0.8333333333333334 +0.2302675979102359,0.924074074074074,0.9240529776789849,0.9718689986282579,0.07592592592592592,0.8588301528979495,0.9245457916203189,0.924074074074074,0.9721188556183242,0.8481481481481481 +0.22829260633272283,0.9277777777777778,0.927757710475132,0.9717798353909465,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.9718743015411218,0.8555555555555556 +0.22767536298317068,0.9277777777777778,0.927757710475132,0.971786694101509,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.9718428216570703,0.8555555555555556 +0.22825184420627706,0.9277777777777778,0.927757710475132,0.9718518518518517,0.07222222222222222,0.8652524167561761,0.9282536151279199,0.9277777777777778,0.9719820208335395,0.8555555555555556 diff --git a/results/downsample/airogs/100/retfound/pr.png b/results/downsample/airogs/100/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a7b8722b6e008818786512019ce060492f13009a --- /dev/null +++ b/results/downsample/airogs/100/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0814fce5a15fb592055797d516f1aae0378941c53a5d9790107042caa8eaf745 +size 43330 diff --git a/results/downsample/airogs/100/retfound/roc.png b/results/downsample/airogs/100/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..d44878c06bb7944fb573ac6929feec5044d0a255 --- /dev/null +++ b/results/downsample/airogs/100/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:96cf71dccc3e6d1955819e049dad85405cec7cab0b5e796170e7e06c8d26b874 +size 60366 diff --git a/results/downsample/airogs/100/retfound/test_pred.npz b/results/downsample/airogs/100/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..63f31414cdb2125dff4da4fb0d029def414291ff --- /dev/null +++ b/results/downsample/airogs/100/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:465d825381cbf129f91200b2a0d6f9cd0c6592b691865e383f9f66e830b2ab0c +size 12510 diff --git a/results/downsample/airogs/100/retfound/train.log b/results/downsample/airogs/100/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..2887255704120f3ba526191059b10cc7e2f816c1 --- /dev/null +++ b/results/downsample/airogs/100/retfound/train.log @@ -0,0 +1,716 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:46:25.257680898 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:46:25.709802] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:46:25.710012] Namespace(batch_size=32, +epochs=30, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/airogs_100', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/100', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:46:28.802603] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:46:30.509871] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:46:52.649510] Sampler_train = +[14:46:52.692177] len of train_set: 4992 +[14:46:52.871832] [Adaptation] Full fine-tuning: training all parameters. +[14:46:52.872899] number of trainable params (M): 303.30 +[14:46:52.872978] base lr: 5.00e-03 +[14:46:52.873022] actual lr: 6.25e-04 +[14:46:52.873065] accumulate grad iterations: 1 +[14:46:52.873107] effective batch size: 32 +[14:46:52.876257] criterion = CrossEntropyLoss() +[14:46:52.876350] Start training for 30 epochs +[14:46:52.878351] log_dir: ./output_logs/retfound +[14:46:54.494115] Epoch: [0] [ 0/156] eta: 0:04:11 lr: 0.000000 loss: 0.6929 (0.6929) time: 1.6148 data: 0.7882 max mem: 7340 +[14:46:55.880475] Epoch: [0] [ 20/156] eta: 0:00:19 lr: 0.000008 loss: 0.6930 (0.6929) time: 0.0693 data: 0.0002 max mem: 9669 +[14:46:57.189051] Epoch: [0] [ 40/156] eta: 0:00:12 lr: 0.000016 loss: 0.6920 (0.6928) time: 0.0654 data: 0.0001 max mem: 9669 +[14:46:58.517394] Epoch: [0] [ 60/156] eta: 0:00:08 lr: 0.000024 loss: 0.6920 (0.6928) time: 0.0664 data: 0.0001 max mem: 9669 +[14:46:59.829153] Epoch: [0] [ 80/156] eta: 0:00:06 lr: 0.000032 loss: 0.6909 (0.6920) time: 0.0655 data: 0.0001 max mem: 9669 +[14:47:01.445636] Epoch: [0] [100/156] eta: 0:00:04 lr: 0.000040 loss: 0.6899 (0.6917) time: 0.0808 data: 0.0002 max mem: 9669 +[14:47:02.759475] Epoch: [0] [120/156] eta: 0:00:02 lr: 0.000048 loss: 0.6860 (0.6908) time: 0.0657 data: 0.0001 max mem: 9669 +[14:47:04.091752] Epoch: [0] [140/156] eta: 0:00:01 lr: 0.000056 loss: 0.6767 (0.6888) time: 0.0666 data: 0.0001 max mem: 9669 +[14:47:05.072720] Epoch: [0] [155/156] eta: 0:00:00 lr: 0.000062 loss: 0.6643 (0.6869) time: 0.0653 data: 0.0001 max mem: 9669 +[14:47:05.148770] Epoch: [0] Total time: 0:00:12 (0.0787 s / it) +[14:47:05.150038] Averaged stats: lr: 0.000062 loss: 0.6643 (0.6869) +[14:47:05.874597] val: [ 0/17] eta: 0:00:12 loss: 0.6420 (0.6420) time: 0.7109 data: 0.6766 max mem: 9669 +[14:47:06.163064] val: [10/17] eta: 0:00:00 loss: 0.6479 (0.6470) time: 0.0908 data: 0.0736 max mem: 9669 +[14:47:06.298247] val: [16/17] eta: 0:00:00 loss: 0.6328 (0.6268) time: 0.0667 data: 0.0477 max mem: 9669 +[14:47:06.401132] val: Total time: 0:00:01 (0.0728 s / it) +[14:47:06.423673] val loss: 0.6267830974915448 +[14:47:06.423937] Accuracy: 0.7444, F1 Score: 0.7426, ROC AUC: 0.8364, Hamming Loss: 0.2556, + Jaccard Score: 0.5910, Precision: 0.7518, Recall: 0.7444, + Average Precision: 0.8348, Kappa: 0.4889, Score: 0.6893 +[14:47:08.256238] Best epoch = 0, Best score = 0.6893 +[14:47:08.329867] log_dir: ./output_logs/retfound +[14:47:09.200875] Epoch: [1] [ 0/156] eta: 0:02:15 lr: 0.000063 loss: 0.6670 (0.6670) time: 0.8700 data: 0.7541 max mem: 9669 +[14:47:10.542839] Epoch: [1] [ 20/156] eta: 0:00:14 lr: 0.000071 loss: 0.6449 (0.6479) time: 0.0671 data: 0.0001 max mem: 9669 +[14:47:11.849760] Epoch: [1] [ 40/156] eta: 0:00:09 lr: 0.000079 loss: 0.6459 (0.6417) time: 0.0653 data: 0.0001 max mem: 9669 +[14:47:13.465570] Epoch: [1] [ 60/156] eta: 0:00:08 lr: 0.000087 loss: 0.6346 (0.6419) time: 0.0807 data: 0.0001 max mem: 9669 +[14:47:14.787226] Epoch: [1] [ 80/156] eta: 0:00:06 lr: 0.000095 loss: 0.6075 (0.6338) time: 0.0660 data: 0.0002 max mem: 9669 +[14:47:16.119922] Epoch: [1] [100/156] eta: 0:00:04 lr: 0.000103 loss: 0.5754 (0.6276) time: 0.0666 data: 0.0001 max mem: 9669 +[14:47:18.434802] Epoch: [1] [120/156] eta: 0:00:03 lr: 0.000111 loss: 0.5869 (0.6222) time: 0.1157 data: 0.0002 max mem: 9669 +[14:47:20.043664] Epoch: [1] [140/156] eta: 0:00:01 lr: 0.000119 loss: 0.5824 (0.6155) time: 0.0804 data: 0.0001 max mem: 9669 +[14:47:21.026810] Epoch: [1] [155/156] eta: 0:00:00 lr: 0.000125 loss: 0.5344 (0.6098) time: 0.0806 data: 0.0001 max mem: 9669 +[14:47:21.120459] Epoch: [1] Total time: 0:00:12 (0.0820 s / it) +[14:47:21.121334] Averaged stats: lr: 0.000125 loss: 0.5344 (0.6098) +[14:47:21.830051] val: [ 0/17] eta: 0:00:11 loss: 0.2565 (0.2565) time: 0.6832 data: 0.6654 max mem: 9669 +[14:47:22.087871] val: [10/17] eta: 0:00:00 loss: 0.3011 (0.3592) time: 0.0855 data: 0.0700 max mem: 9669 +[14:47:22.189742] val: [16/17] eta: 0:00:00 loss: 0.3327 (0.3607) time: 0.0613 data: 0.0460 max mem: 9669 +[14:47:22.279023] val: Total time: 0:00:01 (0.0666 s / it) +[14:47:22.294862] val loss: 0.3606555602129768 +[14:47:22.295083] Accuracy: 0.8500, F1 Score: 0.8499, ROC AUC: 0.9236, Hamming Loss: 0.1500, + Jaccard Score: 0.7390, Precision: 0.8506, Recall: 0.8500, + Average Precision: 0.9234, Kappa: 0.7000, Score: 0.8245 +[14:47:24.142599] Best epoch = 1, Best score = 0.8245 +[14:47:24.224409] log_dir: ./output_logs/retfound +[14:47:25.002564] Epoch: [2] [ 0/156] eta: 0:02:01 lr: 0.000125 loss: 0.5732 (0.5732) time: 0.7769 data: 0.6915 max mem: 9669 +[14:47:26.613640] Epoch: [2] [ 20/156] eta: 0:00:15 lr: 0.000133 loss: 0.5497 (0.5636) time: 0.0805 data: 0.0002 max mem: 9669 +[14:47:27.927989] Epoch: [2] [ 40/156] eta: 0:00:10 lr: 0.000141 loss: 0.5511 (0.5642) time: 0.0657 data: 0.0002 max mem: 9669 +[14:47:29.235198] Epoch: [2] [ 60/156] eta: 0:00:07 lr: 0.000149 loss: 0.5691 (0.5663) time: 0.0653 data: 0.0002 max mem: 9669 +[14:47:30.589944] Epoch: [2] [ 80/156] eta: 0:00:05 lr: 0.000157 loss: 0.5262 (0.5577) time: 0.0677 data: 0.0002 max mem: 9669 +[14:47:31.940686] Epoch: [2] [100/156] eta: 0:00:04 lr: 0.000165 loss: 0.5398 (0.5551) time: 0.0675 data: 0.0002 max mem: 9669 +[14:47:33.562342] Epoch: [2] [120/156] eta: 0:00:02 lr: 0.000173 loss: 0.5339 (0.5563) time: 0.0810 data: 0.0002 max mem: 9669 +[14:47:34.885822] Epoch: [2] [140/156] eta: 0:00:01 lr: 0.000181 loss: 0.5275 (0.5525) time: 0.0661 data: 0.0001 max mem: 9669 +[14:47:35.900068] Epoch: [2] [155/156] eta: 0:00:00 lr: 0.000187 loss: 0.5123 (0.5515) time: 0.0671 data: 0.0001 max mem: 9669 +[14:47:35.991959] Epoch: [2] Total time: 0:00:11 (0.0754 s / it) +[14:47:35.992789] Averaged stats: lr: 0.000187 loss: 0.5123 (0.5515) +[14:47:37.228626] val: [ 0/17] eta: 0:00:20 loss: 0.1746 (0.1746) time: 1.2079 data: 1.1908 max mem: 9669 +[14:47:37.718118] val: [10/17] eta: 0:00:01 loss: 0.2856 (0.2840) time: 0.1542 data: 0.1389 max mem: 9669 +[14:47:38.023956] val: [16/17] eta: 0:00:00 loss: 0.3181 (0.3018) time: 0.1178 data: 0.1025 max mem: 9669 +[14:47:38.117884] val: Total time: 0:00:02 (0.1234 s / it) +[14:47:38.134563] val loss: 0.3017797531450496 +[14:47:38.134771] Accuracy: 0.8796, F1 Score: 0.8794, ROC AUC: 0.9467, Hamming Loss: 0.1204, + Jaccard Score: 0.7849, Precision: 0.8819, Recall: 0.8796, + Average Precision: 0.9464, Kappa: 0.7593, Score: 0.8618 +[14:47:40.038555] Best epoch = 2, Best score = 0.8618 +[14:47:40.119698] log_dir: ./output_logs/retfound +[14:47:41.063626] Epoch: [3] [ 0/156] eta: 0:02:27 lr: 0.000188 loss: 0.5378 (0.5378) time: 0.9430 data: 0.8684 max mem: 9669 +[14:47:42.474562] Epoch: [3] [ 20/156] eta: 0:00:15 lr: 0.000196 loss: 0.5189 (0.5252) time: 0.0705 data: 0.0021 max mem: 9669 +[14:47:43.787699] Epoch: [3] [ 40/156] eta: 0:00:10 lr: 0.000204 loss: 0.4852 (0.5192) time: 0.0656 data: 0.0001 max mem: 9669 +[14:47:45.138250] Epoch: [3] [ 60/156] eta: 0:00:07 lr: 0.000212 loss: 0.5251 (0.5225) time: 0.0675 data: 0.0001 max mem: 9669 +[14:47:46.794601] Epoch: [3] [ 80/156] eta: 0:00:06 lr: 0.000220 loss: 0.5494 (0.5297) time: 0.0828 data: 0.0001 max mem: 9669 +[14:47:48.124114] Epoch: [3] [100/156] eta: 0:00:04 lr: 0.000228 loss: 0.4800 (0.5266) time: 0.0664 data: 0.0001 max mem: 9669 +[14:47:49.467346] Epoch: [3] [120/156] eta: 0:00:02 lr: 0.000236 loss: 0.4913 (0.5219) time: 0.0671 data: 0.0001 max mem: 9669 +[14:47:50.777065] Epoch: [3] [140/156] eta: 0:00:01 lr: 0.000244 loss: 0.5545 (0.5290) time: 0.0654 data: 0.0001 max mem: 9669 +[14:47:51.762947] Epoch: [3] [155/156] eta: 0:00:00 lr: 0.000250 loss: 0.5226 (0.5289) time: 0.0655 data: 0.0001 max mem: 9669 +[14:47:51.848075] Epoch: [3] Total time: 0:00:11 (0.0752 s / it) +[14:47:51.849390] Averaged stats: lr: 0.000250 loss: 0.5226 (0.5289) +[14:47:52.590577] val: [ 0/17] eta: 0:00:12 loss: 0.1953 (0.1953) time: 0.7169 data: 0.6999 max mem: 9669 +[14:47:53.184242] val: [10/17] eta: 0:00:00 loss: 0.2957 (0.2924) time: 0.1191 data: 0.1035 max mem: 9669 +[14:47:53.296157] val: [16/17] eta: 0:00:00 loss: 0.2227 (0.2623) time: 0.0836 data: 0.0682 max mem: 9669 +[14:47:53.371457] val: Total time: 0:00:01 (0.0881 s / it) +[14:47:53.388532] val loss: 0.262339744497748 +[14:47:53.388747] Accuracy: 0.8889, F1 Score: 0.8889, ROC AUC: 0.9588, Hamming Loss: 0.1111, + Jaccard Score: 0.8000, Precision: 0.8889, Recall: 0.8889, + Average Precision: 0.9600, Kappa: 0.7778, Score: 0.8751 +[14:47:55.419214] Best epoch = 3, Best score = 0.8751 +[14:47:55.542471] log_dir: ./output_logs/retfound +[14:47:56.667540] Epoch: [4] [ 0/156] eta: 0:02:55 lr: 0.000250 loss: 0.5731 (0.5731) time: 1.1238 data: 1.0495 max mem: 9669 +[14:47:58.046539] Epoch: [4] [ 20/156] eta: 0:00:16 lr: 0.000258 loss: 0.5041 (0.5161) time: 0.0689 data: 0.0036 max mem: 9669 +[14:47:59.668930] Epoch: [4] [ 40/156] eta: 0:00:11 lr: 0.000266 loss: 0.5131 (0.5237) time: 0.0811 data: 0.0002 max mem: 9669 +[14:48:00.982170] Epoch: [4] [ 60/156] eta: 0:00:08 lr: 0.000274 loss: 0.5255 (0.5249) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:02.297181] Epoch: [4] [ 80/156] eta: 0:00:06 lr: 0.000282 loss: 0.4897 (0.5184) time: 0.0657 data: 0.0001 max mem: 9669 +[14:48:03.627316] Epoch: [4] [100/156] eta: 0:00:04 lr: 0.000290 loss: 0.4656 (0.5108) time: 0.0665 data: 0.0001 max mem: 9669 +[14:48:04.932086] Epoch: [4] [120/156] eta: 0:00:02 lr: 0.000298 loss: 0.4011 (0.4958) time: 0.0652 data: 0.0001 max mem: 9669 +[14:48:06.482110] Epoch: [4] [140/156] eta: 0:00:01 lr: 0.000306 loss: 0.5319 (0.5011) time: 0.0775 data: 0.0001 max mem: 9669 +[14:48:07.543578] Epoch: [4] [155/156] eta: 0:00:00 lr: 0.000312 loss: 0.4962 (0.5010) time: 0.0815 data: 0.0001 max mem: 9669 +[14:48:07.637455] Epoch: [4] Total time: 0:00:12 (0.0775 s / it) +[14:48:07.638332] Averaged stats: lr: 0.000312 loss: 0.4962 (0.5010) +[14:48:08.486619] val: [ 0/17] eta: 0:00:13 loss: 0.1446 (0.1446) time: 0.7844 data: 0.7665 max mem: 9669 +[14:48:09.184972] val: [10/17] eta: 0:00:00 loss: 0.2819 (0.2813) time: 0.1347 data: 0.1193 max mem: 9669 +[14:48:09.275565] val: [16/17] eta: 0:00:00 loss: 0.2819 (0.2907) time: 0.0925 data: 0.0772 max mem: 9669 +[14:48:09.356266] val: Total time: 0:00:01 (0.0973 s / it) +[14:48:09.371420] val loss: 0.2907406424774843 +[14:48:09.371614] Accuracy: 0.8815, F1 Score: 0.8813, ROC AUC: 0.9527, Hamming Loss: 0.1185, + Jaccard Score: 0.7878, Precision: 0.8836, Recall: 0.8815, + Average Precision: 0.9541, Kappa: 0.7630, Score: 0.8657 +[14:48:09.432439] Best epoch = 3, Best score = 0.8751 +[14:48:09.736998] log_dir: ./output_logs/retfound +[14:48:10.790542] Epoch: [5] [ 0/156] eta: 0:02:44 lr: 0.000313 loss: 0.5227 (0.5227) time: 1.0527 data: 0.9816 max mem: 9669 +[14:48:12.160278] Epoch: [5] [ 20/156] eta: 0:00:15 lr: 0.000321 loss: 0.4854 (0.4888) time: 0.0684 data: 0.0032 max mem: 9669 +[14:48:13.777061] Epoch: [5] [ 40/156] eta: 0:00:11 lr: 0.000329 loss: 0.4325 (0.4803) time: 0.0808 data: 0.0001 max mem: 9669 +[14:48:15.098777] Epoch: [5] [ 60/156] eta: 0:00:08 lr: 0.000337 loss: 0.4913 (0.4887) time: 0.0661 data: 0.0001 max mem: 9669 +[14:48:16.406372] Epoch: [5] [ 80/156] eta: 0:00:06 lr: 0.000345 loss: 0.5478 (0.5029) time: 0.0653 data: 0.0001 max mem: 9669 +[14:48:17.744541] Epoch: [5] [100/156] eta: 0:00:04 lr: 0.000353 loss: 0.4967 (0.4999) time: 0.0669 data: 0.0001 max mem: 9669 +[14:48:19.062659] Epoch: [5] [120/156] eta: 0:00:02 lr: 0.000361 loss: 0.5006 (0.5023) time: 0.0659 data: 0.0001 max mem: 9669 +[14:48:20.374994] Epoch: [5] [140/156] eta: 0:00:01 lr: 0.000369 loss: 0.4839 (0.5014) time: 0.0656 data: 0.0001 max mem: 9669 +[14:48:21.654166] Epoch: [5] [155/156] eta: 0:00:00 lr: 0.000375 loss: 0.5141 (0.5024) time: 0.0802 data: 0.0001 max mem: 9669 +[14:48:21.738048] Epoch: [5] Total time: 0:00:12 (0.0769 s / it) +[14:48:21.738788] Averaged stats: lr: 0.000375 loss: 0.5141 (0.5024) +[14:48:22.365513] val: [ 0/17] eta: 0:00:10 loss: 0.3622 (0.3622) time: 0.5997 data: 0.5825 max mem: 9669 +[14:48:22.576074] val: [10/17] eta: 0:00:00 loss: 0.3603 (0.3440) time: 0.0736 data: 0.0581 max mem: 9669 +[14:48:22.667185] val: [16/17] eta: 0:00:00 loss: 0.2352 (0.2706) time: 0.0530 data: 0.0377 max mem: 9669 +[14:48:22.742694] val: Total time: 0:00:00 (0.0575 s / it) +[14:48:22.756995] val loss: 0.27061635694083047 +[14:48:22.757175] Accuracy: 0.8944, F1 Score: 0.8942, ROC AUC: 0.9617, Hamming Loss: 0.1056, + Jaccard Score: 0.8087, Precision: 0.8984, Recall: 0.8944, + Average Precision: 0.9628, Kappa: 0.7889, Score: 0.8816 +[14:48:24.580768] Best epoch = 5, Best score = 0.8816 +[14:48:24.648572] log_dir: ./output_logs/retfound +[14:48:25.659077] Epoch: [6] [ 0/156] eta: 0:02:37 lr: 0.000375 loss: 0.5117 (0.5117) time: 1.0096 data: 0.9380 max mem: 9669 +[14:48:27.062101] Epoch: [6] [ 20/156] eta: 0:00:15 lr: 0.000383 loss: 0.4926 (0.5112) time: 0.0701 data: 0.0047 max mem: 9669 +[14:48:28.699424] Epoch: [6] [ 40/156] eta: 0:00:11 lr: 0.000391 loss: 0.4482 (0.4945) time: 0.0818 data: 0.0001 max mem: 9669 +[14:48:29.997748] Epoch: [6] [ 60/156] eta: 0:00:08 lr: 0.000399 loss: 0.5284 (0.5101) time: 0.0649 data: 0.0001 max mem: 9669 +[14:48:31.335556] Epoch: [6] [ 80/156] eta: 0:00:06 lr: 0.000407 loss: 0.4940 (0.5119) time: 0.0669 data: 0.0001 max mem: 9669 +[14:48:32.642753] Epoch: [6] [100/156] eta: 0:00:04 lr: 0.000415 loss: 0.4631 (0.5055) time: 0.0653 data: 0.0001 max mem: 9669 +[14:48:33.949869] Epoch: [6] [120/156] eta: 0:00:02 lr: 0.000423 loss: 0.4742 (0.5046) time: 0.0653 data: 0.0001 max mem: 9669 +[14:48:35.558726] Epoch: [6] [140/156] eta: 0:00:01 lr: 0.000431 loss: 0.4733 (0.5021) time: 0.0804 data: 0.0001 max mem: 9669 +[14:48:36.531988] Epoch: [6] [155/156] eta: 0:00:00 lr: 0.000437 loss: 0.4729 (0.4997) time: 0.0648 data: 0.0001 max mem: 9669 +[14:48:36.612955] Epoch: [6] Total time: 0:00:11 (0.0767 s / it) +[14:48:36.613705] Averaged stats: lr: 0.000437 loss: 0.4729 (0.4997) +[14:48:37.453279] val: [ 0/17] eta: 0:00:13 loss: 0.3536 (0.3536) time: 0.8144 data: 0.7932 max mem: 9669 +[14:48:37.835081] val: [10/17] eta: 0:00:00 loss: 0.2742 (0.3143) time: 0.1087 data: 0.0929 max mem: 9669 +[14:48:38.131457] val: [16/17] eta: 0:00:00 loss: 0.2401 (0.2562) time: 0.0877 data: 0.0719 max mem: 9669 +[14:48:38.224577] val: Total time: 0:00:01 (0.0933 s / it) +[14:48:38.238817] val loss: 0.25615958650322523 +[14:48:38.238981] Accuracy: 0.8926, F1 Score: 0.8926, ROC AUC: 0.9634, Hamming Loss: 0.1074, + Jaccard Score: 0.8060, Precision: 0.8931, Recall: 0.8926, + Average Precision: 0.9641, Kappa: 0.7852, Score: 0.8804 +[14:48:38.278404] Best epoch = 5, Best score = 0.8816 +[14:48:38.578360] log_dir: ./output_logs/retfound +[14:48:39.597566] Epoch: [7] [ 0/156] eta: 0:02:38 lr: 0.000438 loss: 0.5038 (0.5038) time: 1.0183 data: 0.9474 max mem: 9669 +[14:48:40.899691] Epoch: [7] [ 20/156] eta: 0:00:15 lr: 0.000446 loss: 0.5144 (0.5279) time: 0.0651 data: 0.0001 max mem: 9669 +[14:48:42.503600] Epoch: [7] [ 40/156] eta: 0:00:11 lr: 0.000454 loss: 0.4662 (0.5053) time: 0.0802 data: 0.0001 max mem: 9669 +[14:48:43.817671] Epoch: [7] [ 60/156] eta: 0:00:08 lr: 0.000462 loss: 0.5224 (0.5106) time: 0.0657 data: 0.0001 max mem: 9669 +[14:48:45.153463] Epoch: [7] [ 80/156] eta: 0:00:06 lr: 0.000470 loss: 0.4508 (0.5010) time: 0.0668 data: 0.0001 max mem: 9669 +[14:48:46.461895] Epoch: [7] [100/156] eta: 0:00:04 lr: 0.000478 loss: 0.4589 (0.4949) time: 0.0654 data: 0.0001 max mem: 9669 +[14:48:47.775740] Epoch: [7] [120/156] eta: 0:00:02 lr: 0.000486 loss: 0.4859 (0.4906) time: 0.0657 data: 0.0001 max mem: 9669 +[14:48:49.384068] Epoch: [7] [140/156] eta: 0:00:01 lr: 0.000494 loss: 0.4952 (0.4913) time: 0.0804 data: 0.0001 max mem: 9669 +[14:48:50.358906] Epoch: [7] [155/156] eta: 0:00:00 lr: 0.000500 loss: 0.4568 (0.4901) time: 0.0649 data: 0.0001 max mem: 9669 +[14:48:50.439143] Epoch: [7] Total time: 0:00:11 (0.0760 s / it) +[14:48:50.439902] Averaged stats: lr: 0.000500 loss: 0.4568 (0.4901) +[14:48:51.461778] val: [ 0/17] eta: 0:00:17 loss: 0.3089 (0.3089) time: 1.0001 data: 0.9782 max mem: 9669 +[14:48:51.890994] val: [10/17] eta: 0:00:00 loss: 0.2991 (0.2929) time: 0.1299 data: 0.1140 max mem: 9669 +[14:48:51.993418] val: [16/17] eta: 0:00:00 loss: 0.2143 (0.2417) time: 0.0900 data: 0.0744 max mem: 9669 +[14:48:52.073139] val: Total time: 0:00:01 (0.0948 s / it) +[14:48:52.090634] val loss: 0.24167559558854385 +[14:48:52.090812] Accuracy: 0.8907, F1 Score: 0.8906, ROC AUC: 0.9655, Hamming Loss: 0.1093, + Jaccard Score: 0.8028, Precision: 0.8931, Recall: 0.8907, + Average Precision: 0.9666, Kappa: 0.7815, Score: 0.8792 +[14:48:52.135191] Best epoch = 5, Best score = 0.8816 +[14:48:52.485426] log_dir: ./output_logs/retfound +[14:48:53.308145] Epoch: [8] [ 0/156] eta: 0:02:08 lr: 0.000500 loss: 0.3634 (0.3634) time: 0.8218 data: 0.7519 max mem: 9669 +[14:48:54.615707] Epoch: [8] [ 20/156] eta: 0:00:13 lr: 0.000508 loss: 0.4905 (0.4941) time: 0.0653 data: 0.0001 max mem: 9669 +[14:48:55.913540] Epoch: [8] [ 40/156] eta: 0:00:09 lr: 0.000516 loss: 0.4808 (0.4790) time: 0.0649 data: 0.0001 max mem: 9669 +[14:48:57.518772] Epoch: [8] [ 60/156] eta: 0:00:07 lr: 0.000524 loss: 0.4981 (0.4807) time: 0.0802 data: 0.0001 max mem: 9669 +[14:48:58.817939] Epoch: [8] [ 80/156] eta: 0:00:05 lr: 0.000532 loss: 0.5331 (0.4941) time: 0.0649 data: 0.0001 max mem: 9669 +[14:49:00.157872] Epoch: [8] [100/156] eta: 0:00:04 lr: 0.000540 loss: 0.4574 (0.4899) time: 0.0670 data: 0.0001 max mem: 9669 +[14:49:01.460144] Epoch: [8] [120/156] eta: 0:00:02 lr: 0.000548 loss: 0.4651 (0.4857) time: 0.0651 data: 0.0001 max mem: 9669 +[14:49:02.769834] Epoch: [8] [140/156] eta: 0:00:01 lr: 0.000556 loss: 0.4696 (0.4866) time: 0.0654 data: 0.0001 max mem: 9669 +[14:49:04.047922] Epoch: [8] [155/156] eta: 0:00:00 lr: 0.000562 loss: 0.4585 (0.4825) time: 0.0800 data: 0.0001 max mem: 9669 +[14:49:04.125199] Epoch: [8] Total time: 0:00:11 (0.0746 s / it) +[14:49:04.126046] Averaged stats: lr: 0.000562 loss: 0.4585 (0.4825) +[14:49:05.130574] val: [ 0/17] eta: 0:00:16 loss: 0.2019 (0.2019) time: 0.9575 data: 0.9401 max mem: 9669 +[14:49:05.365217] val: [10/17] eta: 0:00:00 loss: 0.2579 (0.3103) time: 0.1083 data: 0.0929 max mem: 9669 +[14:49:05.599085] val: [16/17] eta: 0:00:00 loss: 0.2253 (0.2528) time: 0.0838 data: 0.0685 max mem: 9669 +[14:49:05.676498] val: Total time: 0:00:01 (0.0885 s / it) +[14:49:05.690412] val loss: 0.2527696148437612 +[14:49:05.690565] Accuracy: 0.9019, F1 Score: 0.9017, ROC AUC: 0.9657, Hamming Loss: 0.0981, + Jaccard Score: 0.8210, Precision: 0.9043, Recall: 0.9019, + Average Precision: 0.9664, Kappa: 0.8037, Score: 0.8904 +[14:49:07.439000] Best epoch = 8, Best score = 0.8904 +[14:49:07.581022] log_dir: ./output_logs/retfound +[14:49:08.461066] Epoch: [9] [ 0/156] eta: 0:02:17 lr: 0.000562 loss: 0.4432 (0.4432) time: 0.8792 data: 0.8021 max mem: 9669 +[14:49:09.980920] Epoch: [9] [ 20/156] eta: 0:00:15 lr: 0.000571 loss: 0.4576 (0.4668) time: 0.0760 data: 0.0001 max mem: 9669 +[14:49:11.371256] Epoch: [9] [ 40/156] eta: 0:00:10 lr: 0.000579 loss: 0.4368 (0.4555) time: 0.0695 data: 0.0001 max mem: 9669 +[14:49:12.671039] Epoch: [9] [ 60/156] eta: 0:00:08 lr: 0.000587 loss: 0.4287 (0.4578) time: 0.0650 data: 0.0001 max mem: 9669 +[14:49:14.011189] Epoch: [9] [ 80/156] eta: 0:00:06 lr: 0.000595 loss: 0.4820 (0.4673) time: 0.0670 data: 0.0001 max mem: 9669 +[14:49:15.317814] Epoch: [9] [100/156] eta: 0:00:04 lr: 0.000603 loss: 0.4720 (0.4682) time: 0.0653 data: 0.0001 max mem: 9669 +[14:49:16.624403] Epoch: [9] [120/156] eta: 0:00:02 lr: 0.000611 loss: 0.5268 (0.4759) time: 0.0653 data: 0.0001 max mem: 9669 +[14:49:18.230749] Epoch: [9] [140/156] eta: 0:00:01 lr: 0.000619 loss: 0.4779 (0.4769) time: 0.0803 data: 0.0001 max mem: 9669 +[14:49:28.195929] Epoch: [9] [155/156] eta: 0:00:00 lr: 0.000625 loss: 0.4439 (0.4740) time: 0.5146 data: 0.0001 max mem: 9669 +[14:49:28.273203] Epoch: [9] Total time: 0:00:20 (0.1326 s / it) +[14:49:28.273998] Averaged stats: lr: 0.000625 loss: 0.4439 (0.4740) +[14:49:29.209379] val: [ 0/17] eta: 0:00:15 loss: 0.3370 (0.3370) time: 0.9272 data: 0.8933 max mem: 9669 +[14:49:29.750524] val: [10/17] eta: 0:00:00 loss: 0.2942 (0.3432) time: 0.1334 data: 0.1013 max mem: 9669 +[14:49:29.842265] val: [16/17] eta: 0:00:00 loss: 0.2737 (0.2629) time: 0.0917 data: 0.0656 max mem: 9669 +[14:49:29.917328] val: Total time: 0:00:01 (0.0962 s / it) +[14:49:29.931267] val loss: 0.2629097079967751 +[14:49:29.931413] Accuracy: 0.8907, F1 Score: 0.8905, ROC AUC: 0.9656, Hamming Loss: 0.1093, + Jaccard Score: 0.8027, Precision: 0.8941, Recall: 0.8907, + Average Precision: 0.9670, Kappa: 0.7815, Score: 0.8792 +[14:49:29.990553] Best epoch = 8, Best score = 0.8904 +[14:49:30.283682] log_dir: ./output_logs/retfound +[14:49:31.190348] Epoch: [10] [ 0/156] eta: 0:02:21 lr: 0.000625 loss: 0.4573 (0.4573) time: 0.9051 data: 0.8262 max mem: 9669 +[14:49:32.600733] Epoch: [10] [ 20/156] eta: 0:00:14 lr: 0.000625 loss: 0.4759 (0.5153) time: 0.0705 data: 0.0043 max mem: 9669 +[14:49:33.942479] Epoch: [10] [ 40/156] eta: 0:00:10 lr: 0.000625 loss: 0.4536 (0.4997) time: 0.0671 data: 0.0001 max mem: 9669 +[14:49:35.253155] Epoch: [10] [ 60/156] eta: 0:00:07 lr: 0.000624 loss: 0.4371 (0.4820) time: 0.0655 data: 0.0001 max mem: 9669 +[14:49:36.856114] Epoch: [10] [ 80/156] eta: 0:00:06 lr: 0.000624 loss: 0.4576 (0.4800) time: 0.0801 data: 0.0001 max mem: 9669 +[14:49:38.158116] Epoch: [10] [100/156] eta: 0:00:04 lr: 0.000623 loss: 0.4665 (0.4773) time: 0.0651 data: 0.0001 max mem: 9669 +[14:49:39.484705] Epoch: [10] [120/156] eta: 0:00:02 lr: 0.000623 loss: 0.4646 (0.4758) time: 0.0663 data: 0.0001 max mem: 9669 +[14:49:40.783204] Epoch: [10] [140/156] eta: 0:00:01 lr: 0.000622 loss: 0.4452 (0.4737) time: 0.0649 data: 0.0001 max mem: 9669 +[14:49:41.760161] Epoch: [10] [155/156] eta: 0:00:00 lr: 0.000621 loss: 0.4679 (0.4773) time: 0.0650 data: 0.0001 max mem: 9669 +[14:49:41.838964] Epoch: [10] Total time: 0:00:11 (0.0741 s / it) +[14:49:41.839701] Averaged stats: lr: 0.000621 loss: 0.4679 (0.4773) +[14:49:42.641779] val: [ 0/17] eta: 0:00:13 loss: 0.1854 (0.1854) time: 0.7936 data: 0.7750 max mem: 9669 +[14:49:42.896708] val: [10/17] eta: 0:00:00 loss: 0.2088 (0.2325) time: 0.0953 data: 0.0796 max mem: 9669 +[14:49:43.194646] val: [16/17] eta: 0:00:00 loss: 0.2786 (0.2516) time: 0.0791 data: 0.0637 max mem: 9669 +[14:49:43.269330] val: Total time: 0:00:01 (0.0836 s / it) +[14:49:43.286771] val loss: 0.2516040574101841 +[14:49:43.287002] Accuracy: 0.9019, F1 Score: 0.9018, ROC AUC: 0.9649, Hamming Loss: 0.0981, + Jaccard Score: 0.8212, Precision: 0.9021, Recall: 0.9019, + Average Precision: 0.9658, Kappa: 0.8037, Score: 0.8902 +[14:49:43.337055] Best epoch = 8, Best score = 0.8904 +[14:49:43.670705] log_dir: ./output_logs/retfound +[14:49:44.660595] Epoch: [11] [ 0/156] eta: 0:02:34 lr: 0.000621 loss: 0.4665 (0.4665) time: 0.9889 data: 0.9189 max mem: 9669 +[14:49:46.038848] Epoch: [11] [ 20/156] eta: 0:00:15 lr: 0.000620 loss: 0.4409 (0.4736) time: 0.0689 data: 0.0024 max mem: 9669 +[14:49:47.338707] Epoch: [11] [ 40/156] eta: 0:00:10 lr: 0.000619 loss: 0.4341 (0.4579) time: 0.0650 data: 0.0001 max mem: 9669 +[14:49:48.762970] Epoch: [11] [ 60/156] eta: 0:00:08 lr: 0.000618 loss: 0.4596 (0.4576) time: 0.0712 data: 0.0028 max mem: 9669 +[14:49:50.306219] Epoch: [11] [ 80/156] eta: 0:00:06 lr: 0.000616 loss: 0.4355 (0.4541) time: 0.0771 data: 0.0001 max mem: 9669 +[14:49:51.616595] Epoch: [11] [100/156] eta: 0:00:04 lr: 0.000615 loss: 0.4711 (0.4532) time: 0.0655 data: 0.0001 max mem: 9669 +[14:49:52.954117] Epoch: [11] [120/156] eta: 0:00:02 lr: 0.000613 loss: 0.4498 (0.4599) time: 0.0668 data: 0.0001 max mem: 9669 +[14:50:04.197522] Epoch: [11] [140/156] eta: 0:00:02 lr: 0.000611 loss: 0.3991 (0.4530) time: 0.5621 data: 0.0001 max mem: 9669 +[14:50:05.169150] Epoch: [11] [155/156] eta: 0:00:00 lr: 0.000610 loss: 0.3987 (0.4518) time: 0.5616 data: 0.0001 max mem: 9669 +[14:50:05.252464] Epoch: [11] Total time: 0:00:21 (0.1383 s / it) +[14:50:05.253230] Averaged stats: lr: 0.000610 loss: 0.3987 (0.4518) +[14:50:05.941260] val: [ 0/17] eta: 0:00:10 loss: 0.2339 (0.2339) time: 0.6463 data: 0.6295 max mem: 9669 +[14:50:06.465244] val: [10/17] eta: 0:00:00 loss: 0.2979 (0.2607) time: 0.1063 data: 0.0909 max mem: 9669 +[14:50:06.556223] val: [16/17] eta: 0:00:00 loss: 0.2054 (0.2327) time: 0.0741 data: 0.0589 max mem: 9669 +[14:50:06.626024] val: Total time: 0:00:01 (0.0783 s / it) +[14:50:06.640167] val loss: 0.23272023919750662 +[14:50:06.640332] Accuracy: 0.9130, F1 Score: 0.9129, ROC AUC: 0.9688, Hamming Loss: 0.0870, + Jaccard Score: 0.8398, Precision: 0.9132, Recall: 0.9130, + Average Precision: 0.9693, Kappa: 0.8259, Score: 0.9025 +[14:50:08.431753] Best epoch = 11, Best score = 0.9025 +[14:50:08.524142] log_dir: ./output_logs/retfound +[14:50:09.496569] Epoch: [12] [ 0/156] eta: 0:02:31 lr: 0.000610 loss: 0.4175 (0.4175) time: 0.9716 data: 0.9021 max mem: 9669 +[14:50:10.795063] Epoch: [12] [ 20/156] eta: 0:00:14 lr: 0.000608 loss: 0.4299 (0.4135) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:12.100242] Epoch: [12] [ 40/156] eta: 0:00:10 lr: 0.000606 loss: 0.4877 (0.4603) time: 0.0652 data: 0.0001 max mem: 9669 +[14:50:13.404368] Epoch: [12] [ 60/156] eta: 0:00:07 lr: 0.000603 loss: 0.4499 (0.4613) time: 0.0652 data: 0.0001 max mem: 9669 +[14:50:14.711135] Epoch: [12] [ 80/156] eta: 0:00:05 lr: 0.000601 loss: 0.4070 (0.4535) time: 0.0653 data: 0.0001 max mem: 9669 +[14:50:16.016078] Epoch: [12] [100/156] eta: 0:00:04 lr: 0.000599 loss: 0.4636 (0.4610) time: 0.0652 data: 0.0001 max mem: 9669 +[14:50:17.316262] Epoch: [12] [120/156] eta: 0:00:02 lr: 0.000596 loss: 0.4434 (0.4627) time: 0.0650 data: 0.0001 max mem: 9669 +[14:50:18.615706] Epoch: [12] [140/156] eta: 0:00:01 lr: 0.000593 loss: 0.4253 (0.4616) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:19.599171] Epoch: [12] [155/156] eta: 0:00:00 lr: 0.000591 loss: 0.4490 (0.4638) time: 0.0654 data: 0.0001 max mem: 9669 +[14:50:19.676276] Epoch: [12] Total time: 0:00:11 (0.0715 s / it) +[14:50:19.677074] Averaged stats: lr: 0.000591 loss: 0.4490 (0.4638) +[14:50:20.732868] val: [ 0/17] eta: 0:00:17 loss: 0.2895 (0.2895) time: 1.0142 data: 0.9972 max mem: 9669 +[14:50:21.315687] val: [10/17] eta: 0:00:01 loss: 0.2840 (0.2778) time: 0.1451 data: 0.1298 max mem: 9669 +[14:50:21.407005] val: [16/17] eta: 0:00:00 loss: 0.1948 (0.2283) time: 0.0992 data: 0.0840 max mem: 9669 +[14:50:21.482535] val: Total time: 0:00:01 (0.1038 s / it) +[14:50:21.497601] val loss: 0.22826781693626852 +[14:50:21.497769] Accuracy: 0.9037, F1 Score: 0.9036, ROC AUC: 0.9681, Hamming Loss: 0.0963, + Jaccard Score: 0.8242, Precision: 0.9048, Recall: 0.9037, + Average Precision: 0.9683, Kappa: 0.8074, Score: 0.8930 +[14:50:21.597665] Best epoch = 11, Best score = 0.9025 +[14:50:21.860202] log_dir: ./output_logs/retfound +[14:50:22.887873] Epoch: [13] [ 0/156] eta: 0:02:40 lr: 0.000591 loss: 0.3888 (0.3888) time: 1.0268 data: 0.9575 max mem: 9669 +[14:50:24.184592] Epoch: [13] [ 20/156] eta: 0:00:15 lr: 0.000588 loss: 0.4828 (0.4831) time: 0.0648 data: 0.0001 max mem: 9669 +[14:50:25.490096] Epoch: [13] [ 40/156] eta: 0:00:10 lr: 0.000585 loss: 0.4745 (0.4803) time: 0.0652 data: 0.0001 max mem: 9669 +[14:50:26.799240] Epoch: [13] [ 60/156] eta: 0:00:07 lr: 0.000582 loss: 0.4478 (0.4713) time: 0.0654 data: 0.0001 max mem: 9669 +[14:50:28.097924] Epoch: [13] [ 80/156] eta: 0:00:05 lr: 0.000579 loss: 0.4615 (0.4669) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:29.399956] Epoch: [13] [100/156] eta: 0:00:04 lr: 0.000575 loss: 0.4323 (0.4634) time: 0.0651 data: 0.0001 max mem: 9669 +[14:50:30.712824] Epoch: [13] [120/156] eta: 0:00:02 lr: 0.000572 loss: 0.3835 (0.4549) time: 0.0656 data: 0.0001 max mem: 9669 +[14:50:32.011805] Epoch: [13] [140/156] eta: 0:00:01 lr: 0.000568 loss: 0.4419 (0.4550) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:32.986572] Epoch: [13] [155/156] eta: 0:00:00 lr: 0.000566 loss: 0.4427 (0.4572) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:33.066310] Epoch: [13] Total time: 0:00:11 (0.0718 s / it) +[14:50:33.067050] Averaged stats: lr: 0.000566 loss: 0.4427 (0.4572) +[14:50:33.738824] val: [ 0/17] eta: 0:00:11 loss: 0.2312 (0.2312) time: 0.6523 data: 0.6352 max mem: 9669 +[14:50:33.931808] val: [10/17] eta: 0:00:00 loss: 0.2646 (0.2644) time: 0.0768 data: 0.0614 max mem: 9669 +[14:50:34.044247] val: [16/17] eta: 0:00:00 loss: 0.2106 (0.2341) time: 0.0563 data: 0.0410 max mem: 9669 +[14:50:34.120644] val: Total time: 0:00:01 (0.0609 s / it) +[14:50:34.135092] val loss: 0.23409806586363735 +[14:50:34.135268] Accuracy: 0.9130, F1 Score: 0.9129, ROC AUC: 0.9672, Hamming Loss: 0.0870, + Jaccard Score: 0.8398, Precision: 0.9132, Recall: 0.9130, + Average Precision: 0.9681, Kappa: 0.8259, Score: 0.9020 +[14:50:34.180997] Best epoch = 11, Best score = 0.9025 +[14:50:34.495308] log_dir: ./output_logs/retfound +[14:50:35.368947] Epoch: [14] [ 0/156] eta: 0:02:16 lr: 0.000565 loss: 0.4599 (0.4599) time: 0.8727 data: 0.8023 max mem: 9669 +[14:50:36.667750] Epoch: [14] [ 20/156] eta: 0:00:14 lr: 0.000562 loss: 0.4364 (0.4398) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:37.965565] Epoch: [14] [ 40/156] eta: 0:00:09 lr: 0.000558 loss: 0.4232 (0.4337) time: 0.0648 data: 0.0001 max mem: 9669 +[14:50:39.264700] Epoch: [14] [ 60/156] eta: 0:00:07 lr: 0.000554 loss: 0.4154 (0.4344) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:40.571314] Epoch: [14] [ 80/156] eta: 0:00:05 lr: 0.000550 loss: 0.4435 (0.4407) time: 0.0653 data: 0.0001 max mem: 9669 +[14:50:41.872151] Epoch: [14] [100/156] eta: 0:00:04 lr: 0.000546 loss: 0.4026 (0.4389) time: 0.0650 data: 0.0001 max mem: 9669 +[14:50:43.171387] Epoch: [14] [120/156] eta: 0:00:02 lr: 0.000541 loss: 0.4247 (0.4368) time: 0.0649 data: 0.0001 max mem: 9669 +[14:50:44.472503] Epoch: [14] [140/156] eta: 0:00:01 lr: 0.000537 loss: 0.3900 (0.4370) time: 0.0650 data: 0.0001 max mem: 9669 +[14:50:56.708730] Epoch: [14] [155/156] eta: 0:00:00 lr: 0.000534 loss: 0.4216 (0.4376) time: 0.6281 data: 0.0001 max mem: 9669 +[14:50:56.789349] Epoch: [14] Total time: 0:00:22 (0.1429 s / it) +[14:50:56.790064] Averaged stats: lr: 0.000534 loss: 0.4216 (0.4376) +[14:50:57.486176] val: [ 0/17] eta: 0:00:11 loss: 0.2281 (0.2281) time: 0.6656 data: 0.6460 max mem: 9669 +[14:50:57.642983] val: [10/17] eta: 0:00:00 loss: 0.2639 (0.2634) time: 0.0747 data: 0.0591 max mem: 9669 +[14:50:57.734058] val: [16/17] eta: 0:00:00 loss: 0.2171 (0.2362) time: 0.0537 data: 0.0383 max mem: 9669 +[14:50:57.811166] val: Total time: 0:00:00 (0.0583 s / it) +[14:50:57.825220] val loss: 0.23617563791134777 +[14:50:57.825388] Accuracy: 0.9000, F1 Score: 0.9000, ROC AUC: 0.9680, Hamming Loss: 0.1000, + Jaccard Score: 0.8182, Precision: 0.9002, Recall: 0.9000, + Average Precision: 0.9691, Kappa: 0.8000, Score: 0.8893 +[14:50:57.865862] Best epoch = 11, Best score = 0.9025 +[14:50:58.170708] log_dir: ./output_logs/retfound +[14:50:58.934321] Epoch: [15] [ 0/156] eta: 0:01:58 lr: 0.000534 loss: 0.5035 (0.5035) time: 0.7624 data: 0.6913 max mem: 9669 +[14:51:00.255314] Epoch: [15] [ 20/156] eta: 0:00:13 lr: 0.000529 loss: 0.4108 (0.4235) time: 0.0660 data: 0.0001 max mem: 9669 +[14:51:01.559494] Epoch: [15] [ 40/156] eta: 0:00:09 lr: 0.000525 loss: 0.4453 (0.4318) time: 0.0652 data: 0.0001 max mem: 9669 +[14:51:02.866984] Epoch: [15] [ 60/156] eta: 0:00:07 lr: 0.000520 loss: 0.4378 (0.4427) time: 0.0653 data: 0.0001 max mem: 9669 +[14:51:04.163395] Epoch: [15] [ 80/156] eta: 0:00:05 lr: 0.000515 loss: 0.3557 (0.4318) time: 0.0648 data: 0.0001 max mem: 9669 +[14:51:05.460661] Epoch: [15] [100/156] eta: 0:00:04 lr: 0.000510 loss: 0.3944 (0.4249) time: 0.0648 data: 0.0001 max mem: 9669 +[14:51:06.756680] Epoch: [15] [120/156] eta: 0:00:02 lr: 0.000505 loss: 0.4038 (0.4225) time: 0.0648 data: 0.0001 max mem: 9669 +[14:51:08.056565] Epoch: [15] [140/156] eta: 0:00:01 lr: 0.000500 loss: 0.4266 (0.4221) time: 0.0650 data: 0.0001 max mem: 9669 +[14:51:09.030404] Epoch: [15] [155/156] eta: 0:00:00 lr: 0.000497 loss: 0.4199 (0.4237) time: 0.0648 data: 0.0001 max mem: 9669 +[14:51:09.109069] Epoch: [15] Total time: 0:00:10 (0.0701 s / it) +[14:51:09.109790] Averaged stats: lr: 0.000497 loss: 0.4199 (0.4237) +[14:51:10.208630] val: [ 0/17] eta: 0:00:18 loss: 0.2239 (0.2239) time: 1.0775 data: 1.0606 max mem: 9669 +[14:51:10.734240] val: [10/17] eta: 0:00:01 loss: 0.2239 (0.2526) time: 0.1457 data: 0.1303 max mem: 9669 +[14:51:10.826103] val: [16/17] eta: 0:00:00 loss: 0.2040 (0.2294) time: 0.0996 data: 0.0843 max mem: 9669 +[14:51:10.901263] val: Total time: 0:00:01 (0.1042 s / it) +[14:51:10.916721] val loss: 0.22935042880913792 +[14:51:10.916965] Accuracy: 0.9185, F1 Score: 0.9185, ROC AUC: 0.9698, Hamming Loss: 0.0815, + Jaccard Score: 0.8493, Precision: 0.9187, Recall: 0.9185, + Average Precision: 0.9710, Kappa: 0.8370, Score: 0.9085 +[14:51:12.635418] Best epoch = 15, Best score = 0.9085 +[14:51:12.705895] log_dir: ./output_logs/retfound +[14:51:13.884061] Epoch: [16] [ 0/156] eta: 0:03:03 lr: 0.000496 loss: 0.3148 (0.3148) time: 1.1770 data: 1.1047 max mem: 9669 +[14:51:15.223778] Epoch: [16] [ 20/156] eta: 0:00:16 lr: 0.000491 loss: 0.3946 (0.4188) time: 0.0670 data: 0.0023 max mem: 9669 +[14:51:16.741024] Epoch: [16] [ 40/156] eta: 0:00:11 lr: 0.000486 loss: 0.4368 (0.4331) time: 0.0758 data: 0.0108 max mem: 9669 +[14:51:18.046547] Epoch: [16] [ 60/156] eta: 0:00:08 lr: 0.000481 loss: 0.4131 (0.4302) time: 0.0652 data: 0.0001 max mem: 9669 +[14:51:19.347211] Epoch: [16] [ 80/156] eta: 0:00:06 lr: 0.000475 loss: 0.4221 (0.4310) time: 0.0650 data: 0.0001 max mem: 9669 +[14:51:20.709958] Epoch: [16] [100/156] eta: 0:00:04 lr: 0.000470 loss: 0.3896 (0.4337) time: 0.0681 data: 0.0031 max mem: 9669 +[14:51:22.011801] Epoch: [16] [120/156] eta: 0:00:02 lr: 0.000465 loss: 0.4177 (0.4348) time: 0.0651 data: 0.0001 max mem: 9669 +[14:51:23.329726] Epoch: [16] [140/156] eta: 0:00:01 lr: 0.000459 loss: 0.3996 (0.4352) time: 0.0659 data: 0.0011 max mem: 9669 +[14:51:24.303697] Epoch: [16] [155/156] eta: 0:00:00 lr: 0.000455 loss: 0.3996 (0.4304) time: 0.0659 data: 0.0011 max mem: 9669 +[14:51:24.388113] Epoch: [16] Total time: 0:00:11 (0.0749 s / it) +[14:51:24.388993] Averaged stats: lr: 0.000455 loss: 0.3996 (0.4304) +[14:51:25.125639] val: [ 0/17] eta: 0:00:12 loss: 0.3264 (0.3264) time: 0.7177 data: 0.7005 max mem: 9669 +[14:51:25.291155] val: [10/17] eta: 0:00:00 loss: 0.2887 (0.2838) time: 0.0802 data: 0.0648 max mem: 9669 +[14:51:25.384932] val: [16/17] eta: 0:00:00 loss: 0.1784 (0.2348) time: 0.0574 data: 0.0421 max mem: 9669 +[14:51:25.464175] val: Total time: 0:00:01 (0.0622 s / it) +[14:51:25.478395] val loss: 0.23483087297748118 +[14:51:25.478614] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9689, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9213, Recall: 0.9204, + Average Precision: 0.9700, Kappa: 0.8407, Score: 0.9100 +[14:51:27.546862] Best epoch = 16, Best score = 0.9100 +[14:51:27.636908] log_dir: ./output_logs/retfound +[14:51:28.566743] Epoch: [17] [ 0/156] eta: 0:02:24 lr: 0.000455 loss: 0.3511 (0.3511) time: 0.9289 data: 0.8558 max mem: 9669 +[14:51:29.955698] Epoch: [17] [ 20/156] eta: 0:00:15 lr: 0.000449 loss: 0.3710 (0.3823) time: 0.0694 data: 0.0040 max mem: 9669 +[14:51:31.320027] Epoch: [17] [ 40/156] eta: 0:00:10 lr: 0.000443 loss: 0.3868 (0.3839) time: 0.0682 data: 0.0033 max mem: 9669 +[14:51:32.620119] Epoch: [17] [ 60/156] eta: 0:00:07 lr: 0.000438 loss: 0.4111 (0.4096) time: 0.0650 data: 0.0001 max mem: 9669 +[14:51:33.931758] Epoch: [17] [ 80/156] eta: 0:00:05 lr: 0.000432 loss: 0.4047 (0.4131) time: 0.0655 data: 0.0001 max mem: 9669 +[14:51:35.232002] Epoch: [17] [100/156] eta: 0:00:04 lr: 0.000426 loss: 0.4097 (0.4126) time: 0.0650 data: 0.0001 max mem: 9669 +[14:51:36.536239] Epoch: [17] [120/156] eta: 0:00:02 lr: 0.000420 loss: 0.3984 (0.4150) time: 0.0652 data: 0.0001 max mem: 9669 +[14:51:37.844661] Epoch: [17] [140/156] eta: 0:00:01 lr: 0.000414 loss: 0.4486 (0.4206) time: 0.0654 data: 0.0001 max mem: 9669 +[14:51:38.819133] Epoch: [17] [155/156] eta: 0:00:00 lr: 0.000410 loss: 0.4067 (0.4216) time: 0.0650 data: 0.0001 max mem: 9669 +[14:51:38.900328] Epoch: [17] Total time: 0:00:11 (0.0722 s / it) +[14:51:38.901226] Averaged stats: lr: 0.000410 loss: 0.4067 (0.4216) +[14:51:39.596434] val: [ 0/17] eta: 0:00:11 loss: 0.1918 (0.1918) time: 0.6821 data: 0.6639 max mem: 9669 +[14:51:39.750725] val: [10/17] eta: 0:00:00 loss: 0.2945 (0.2638) time: 0.0760 data: 0.0605 max mem: 9669 +[14:51:39.841837] val: [16/17] eta: 0:00:00 loss: 0.2452 (0.2511) time: 0.0545 data: 0.0392 max mem: 9669 +[14:51:39.922221] val: Total time: 0:00:01 (0.0593 s / it) +[14:51:39.936999] val loss: 0.2510683304246734 +[14:51:39.937183] Accuracy: 0.9074, F1 Score: 0.9074, ROC AUC: 0.9677, Hamming Loss: 0.0926, + Jaccard Score: 0.8305, Precision: 0.9074, Recall: 0.9074, + Average Precision: 0.9684, Kappa: 0.8148, Score: 0.8966 +[14:51:39.981607] Best epoch = 16, Best score = 0.9100 +[14:51:40.276034] log_dir: ./output_logs/retfound +[14:51:41.118491] Epoch: [18] [ 0/156] eta: 0:02:11 lr: 0.000409 loss: 0.6675 (0.6675) time: 0.8416 data: 0.7704 max mem: 9669 +[14:51:52.122186] Epoch: [18] [ 20/156] eta: 0:01:16 lr: 0.000403 loss: 0.4165 (0.4052) time: 0.5501 data: 0.0001 max mem: 9669 +[14:51:53.425558] Epoch: [18] [ 40/156] eta: 0:00:37 lr: 0.000397 loss: 0.4269 (0.4197) time: 0.0651 data: 0.0001 max mem: 9669 +[14:51:54.723823] Epoch: [18] [ 60/156] eta: 0:00:22 lr: 0.000391 loss: 0.4218 (0.4280) time: 0.0649 data: 0.0001 max mem: 9669 +[14:51:56.026053] Epoch: [18] [ 80/156] eta: 0:00:14 lr: 0.000385 loss: 0.3917 (0.4246) time: 0.0651 data: 0.0001 max mem: 9669 +[14:51:57.347057] Epoch: [18] [100/156] eta: 0:00:09 lr: 0.000379 loss: 0.4344 (0.4275) time: 0.0660 data: 0.0001 max mem: 9669 +[14:51:58.649082] Epoch: [18] [120/156] eta: 0:00:05 lr: 0.000373 loss: 0.4358 (0.4303) time: 0.0651 data: 0.0001 max mem: 9669 +[14:51:59.952108] Epoch: [18] [140/156] eta: 0:00:02 lr: 0.000367 loss: 0.4121 (0.4294) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:00.922317] Epoch: [18] [155/156] eta: 0:00:00 lr: 0.000362 loss: 0.3924 (0.4266) time: 0.0648 data: 0.0001 max mem: 9669 +[14:52:01.003208] Epoch: [18] Total time: 0:00:20 (0.1329 s / it) +[14:52:01.003948] Averaged stats: lr: 0.000362 loss: 0.3924 (0.4266) +[14:52:01.501383] val: [ 0/17] eta: 0:00:08 loss: 0.2619 (0.2619) time: 0.4830 data: 0.4649 max mem: 9669 +[14:52:01.706702] val: [10/17] eta: 0:00:00 loss: 0.2619 (0.2848) time: 0.0625 data: 0.0470 max mem: 9669 +[14:52:01.798076] val: [16/17] eta: 0:00:00 loss: 0.1932 (0.2353) time: 0.0458 data: 0.0305 max mem: 9669 +[14:52:01.879929] val: Total time: 0:00:00 (0.0507 s / it) +[14:52:01.896472] val loss: 0.23531346899621627 +[14:52:01.896700] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9708, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9220, Recall: 0.9204, + Average Precision: 0.9716, Kappa: 0.8407, Score: 0.9106 +[14:52:03.609336] Best epoch = 18, Best score = 0.9106 +[14:52:03.671364] log_dir: ./output_logs/retfound +[14:52:04.378816] Epoch: [19] [ 0/156] eta: 0:01:50 lr: 0.000362 loss: 0.4672 (0.4672) time: 0.7065 data: 0.6349 max mem: 9669 +[14:52:05.691040] Epoch: [19] [ 20/156] eta: 0:00:13 lr: 0.000356 loss: 0.4151 (0.4028) time: 0.0656 data: 0.0007 max mem: 9669 +[14:52:06.998690] Epoch: [19] [ 40/156] eta: 0:00:09 lr: 0.000349 loss: 0.3841 (0.3963) time: 0.0653 data: 0.0001 max mem: 9669 +[14:52:08.303651] Epoch: [19] [ 60/156] eta: 0:00:07 lr: 0.000343 loss: 0.4096 (0.4003) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:09.603073] Epoch: [19] [ 80/156] eta: 0:00:05 lr: 0.000337 loss: 0.3970 (0.3987) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:10.898749] Epoch: [19] [100/156] eta: 0:00:04 lr: 0.000331 loss: 0.4069 (0.3998) time: 0.0647 data: 0.0001 max mem: 9669 +[14:52:12.202973] Epoch: [19] [120/156] eta: 0:00:02 lr: 0.000324 loss: 0.3498 (0.3965) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:13.505837] Epoch: [19] [140/156] eta: 0:00:01 lr: 0.000318 loss: 0.4158 (0.3991) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:14.498509] Epoch: [19] [155/156] eta: 0:00:00 lr: 0.000313 loss: 0.4382 (0.4027) time: 0.0660 data: 0.0001 max mem: 9669 +[14:52:14.577675] Epoch: [19] Total time: 0:00:10 (0.0699 s / it) +[14:52:14.578400] Averaged stats: lr: 0.000313 loss: 0.4382 (0.4027) +[14:52:15.221146] val: [ 0/17] eta: 0:00:10 loss: 0.1684 (0.1684) time: 0.6299 data: 0.6129 max mem: 9669 +[14:52:15.416518] val: [10/17] eta: 0:00:00 loss: 0.2131 (0.2261) time: 0.0750 data: 0.0596 max mem: 9669 +[14:52:15.582330] val: [16/17] eta: 0:00:00 loss: 0.2131 (0.2168) time: 0.0582 data: 0.0430 max mem: 9669 +[14:52:15.660194] val: Total time: 0:00:01 (0.0629 s / it) +[14:52:15.674414] val loss: 0.21684648622484767 +[14:52:15.674573] Accuracy: 0.9148, F1 Score: 0.9148, ROC AUC: 0.9713, Hamming Loss: 0.0852, + Jaccard Score: 0.8430, Precision: 0.9148, Recall: 0.9148, + Average Precision: 0.9723, Kappa: 0.8296, Score: 0.9052 +[14:52:15.710103] Best epoch = 18, Best score = 0.9106 +[14:52:16.021690] log_dir: ./output_logs/retfound +[14:52:16.798872] Epoch: [20] [ 0/156] eta: 0:02:01 lr: 0.000313 loss: 0.4803 (0.4803) time: 0.7763 data: 0.7018 max mem: 9669 +[14:52:18.094249] Epoch: [20] [ 20/156] eta: 0:00:13 lr: 0.000307 loss: 0.3682 (0.3869) time: 0.0647 data: 0.0001 max mem: 9669 +[14:52:19.398567] Epoch: [20] [ 40/156] eta: 0:00:09 lr: 0.000300 loss: 0.3869 (0.3848) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:20.706480] Epoch: [20] [ 60/156] eta: 0:00:07 lr: 0.000294 loss: 0.3871 (0.3932) time: 0.0654 data: 0.0001 max mem: 9669 +[14:52:22.008676] Epoch: [20] [ 80/156] eta: 0:00:05 lr: 0.000288 loss: 0.4182 (0.3983) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:23.312460] Epoch: [20] [100/156] eta: 0:00:04 lr: 0.000282 loss: 0.3950 (0.4008) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:24.611818] Epoch: [20] [120/156] eta: 0:00:02 lr: 0.000275 loss: 0.4196 (0.4051) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:25.916609] Epoch: [20] [140/156] eta: 0:00:01 lr: 0.000269 loss: 0.3612 (0.4010) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:26.893439] Epoch: [20] [155/156] eta: 0:00:00 lr: 0.000265 loss: 0.4231 (0.4008) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:26.974010] Epoch: [20] Total time: 0:00:10 (0.0702 s / it) +[14:52:26.974725] Averaged stats: lr: 0.000265 loss: 0.4231 (0.4008) +[14:52:27.507695] val: [ 0/17] eta: 0:00:08 loss: 0.1891 (0.1891) time: 0.5200 data: 0.5031 max mem: 9669 +[14:52:27.763559] val: [10/17] eta: 0:00:00 loss: 0.2415 (0.2510) time: 0.0705 data: 0.0551 max mem: 9669 +[14:52:27.855656] val: [16/17] eta: 0:00:00 loss: 0.2240 (0.2323) time: 0.0510 data: 0.0358 max mem: 9669 +[14:52:27.933363] val: Total time: 0:00:00 (0.0557 s / it) +[14:52:27.947919] val loss: 0.23225475234143875 +[14:52:27.948087] Accuracy: 0.9111, F1 Score: 0.9111, ROC AUC: 0.9698, Hamming Loss: 0.0889, + Jaccard Score: 0.8367, Precision: 0.9111, Recall: 0.9111, + Average Precision: 0.9704, Kappa: 0.8222, Score: 0.9011 +[14:52:27.985721] Best epoch = 18, Best score = 0.9106 +[14:52:28.320884] log_dir: ./output_logs/retfound +[14:52:29.119434] Epoch: [21] [ 0/156] eta: 0:02:04 lr: 0.000264 loss: 0.3076 (0.3076) time: 0.7977 data: 0.7280 max mem: 9669 +[14:52:30.417533] Epoch: [21] [ 20/156] eta: 0:00:13 lr: 0.000258 loss: 0.3947 (0.3811) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:39.887080] Epoch: [21] [ 40/156] eta: 0:00:32 lr: 0.000252 loss: 0.3580 (0.3706) time: 0.4734 data: 0.0001 max mem: 9669 +[14:52:41.190629] Epoch: [21] [ 60/156] eta: 0:00:20 lr: 0.000246 loss: 0.3954 (0.3747) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:42.509380] Epoch: [21] [ 80/156] eta: 0:00:13 lr: 0.000240 loss: 0.4262 (0.3865) time: 0.0659 data: 0.0001 max mem: 9669 +[14:52:43.806098] Epoch: [21] [100/156] eta: 0:00:08 lr: 0.000233 loss: 0.4166 (0.3932) time: 0.0648 data: 0.0001 max mem: 9669 +[14:52:45.105759] Epoch: [21] [120/156] eta: 0:00:04 lr: 0.000227 loss: 0.3974 (0.3968) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:46.404058] Epoch: [21] [140/156] eta: 0:00:02 lr: 0.000221 loss: 0.3612 (0.3919) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:47.375442] Epoch: [21] [155/156] eta: 0:00:00 lr: 0.000217 loss: 0.3752 (0.3934) time: 0.0647 data: 0.0001 max mem: 9669 +[14:52:47.454075] Epoch: [21] Total time: 0:00:19 (0.1226 s / it) +[14:52:47.454832] Averaged stats: lr: 0.000217 loss: 0.3752 (0.3934) +[14:52:48.056962] val: [ 0/17] eta: 0:00:10 loss: 0.3412 (0.3412) time: 0.5892 data: 0.5696 max mem: 9669 +[14:52:48.210779] val: [10/17] eta: 0:00:00 loss: 0.2534 (0.3023) time: 0.0675 data: 0.0519 max mem: 9669 +[14:52:48.301421] val: [16/17] eta: 0:00:00 loss: 0.1760 (0.2378) time: 0.0490 data: 0.0336 max mem: 9669 +[14:52:48.373334] val: Total time: 0:00:00 (0.0533 s / it) +[14:52:48.387263] val loss: 0.23779110676225493 +[14:52:48.387422] Accuracy: 0.9167, F1 Score: 0.9165, ROC AUC: 0.9706, Hamming Loss: 0.0833, + Jaccard Score: 0.8460, Precision: 0.9192, Recall: 0.9167, + Average Precision: 0.9709, Kappa: 0.8333, Score: 0.9068 +[14:52:48.427761] Best epoch = 18, Best score = 0.9106 +[14:52:48.741168] log_dir: ./output_logs/retfound +[14:52:49.460327] Epoch: [22] [ 0/156] eta: 0:01:52 lr: 0.000217 loss: 0.2743 (0.2743) time: 0.7183 data: 0.6431 max mem: 9669 +[14:52:50.757214] Epoch: [22] [ 20/156] eta: 0:00:13 lr: 0.000211 loss: 0.3164 (0.3383) time: 0.0648 data: 0.0001 max mem: 9669 +[14:52:52.055392] Epoch: [22] [ 40/156] eta: 0:00:09 lr: 0.000205 loss: 0.3849 (0.3751) time: 0.0649 data: 0.0001 max mem: 9669 +[14:52:53.368276] Epoch: [22] [ 60/156] eta: 0:00:07 lr: 0.000199 loss: 0.3804 (0.3815) time: 0.0656 data: 0.0001 max mem: 9669 +[14:52:54.676231] Epoch: [22] [ 80/156] eta: 0:00:05 lr: 0.000193 loss: 0.3572 (0.3854) time: 0.0654 data: 0.0001 max mem: 9669 +[14:52:55.987221] Epoch: [22] [100/156] eta: 0:00:04 lr: 0.000187 loss: 0.4054 (0.3910) time: 0.0655 data: 0.0001 max mem: 9669 +[14:52:57.290737] Epoch: [22] [120/156] eta: 0:00:02 lr: 0.000182 loss: 0.3655 (0.3888) time: 0.0651 data: 0.0001 max mem: 9669 +[14:52:58.595435] Epoch: [22] [140/156] eta: 0:00:01 lr: 0.000176 loss: 0.3798 (0.3888) time: 0.0652 data: 0.0001 max mem: 9669 +[14:52:59.572485] Epoch: [22] [155/156] eta: 0:00:00 lr: 0.000172 loss: 0.3645 (0.3888) time: 0.0650 data: 0.0001 max mem: 9669 +[14:52:59.658019] Epoch: [22] Total time: 0:00:10 (0.0700 s / it) +[14:52:59.658870] Averaged stats: lr: 0.000172 loss: 0.3645 (0.3888) +[14:53:00.366644] val: [ 0/17] eta: 0:00:11 loss: 0.2359 (0.2359) time: 0.6837 data: 0.6669 max mem: 9669 +[14:53:00.631586] val: [10/17] eta: 0:00:00 loss: 0.2614 (0.2715) time: 0.0862 data: 0.0709 max mem: 9669 +[14:53:00.729338] val: [16/17] eta: 0:00:00 loss: 0.2143 (0.2386) time: 0.0615 data: 0.0463 max mem: 9669 +[14:53:00.808959] val: Total time: 0:00:01 (0.0663 s / it) +[14:53:00.822822] val loss: 0.2386372685432434 +[14:53:00.822985] Accuracy: 0.9185, F1 Score: 0.9185, ROC AUC: 0.9701, Hamming Loss: 0.0815, + Jaccard Score: 0.8493, Precision: 0.9187, Recall: 0.9185, + Average Precision: 0.9705, Kappa: 0.8370, Score: 0.9086 +[14:53:00.871846] Best epoch = 18, Best score = 0.9106 +[14:53:01.151066] log_dir: ./output_logs/retfound +[14:53:01.936895] Epoch: [23] [ 0/156] eta: 0:02:02 lr: 0.000171 loss: 0.3711 (0.3711) time: 0.7850 data: 0.7124 max mem: 9669 +[14:53:03.248819] Epoch: [23] [ 20/156] eta: 0:00:13 lr: 0.000166 loss: 0.3567 (0.3692) time: 0.0656 data: 0.0001 max mem: 9669 +[14:53:04.553188] Epoch: [23] [ 40/156] eta: 0:00:09 lr: 0.000160 loss: 0.3816 (0.3839) time: 0.0652 data: 0.0001 max mem: 9669 +[14:53:05.851073] Epoch: [23] [ 60/156] eta: 0:00:07 lr: 0.000155 loss: 0.3849 (0.3865) time: 0.0649 data: 0.0001 max mem: 9669 +[14:53:07.148704] Epoch: [23] [ 80/156] eta: 0:00:05 lr: 0.000149 loss: 0.3736 (0.3917) time: 0.0648 data: 0.0001 max mem: 9669 +[14:53:08.447676] Epoch: [23] [100/156] eta: 0:00:04 lr: 0.000144 loss: 0.3504 (0.3864) time: 0.0649 data: 0.0001 max mem: 9669 +[14:53:09.748308] Epoch: [23] [120/156] eta: 0:00:02 lr: 0.000139 loss: 0.3553 (0.3820) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:11.048792] Epoch: [23] [140/156] eta: 0:00:01 lr: 0.000134 loss: 0.3917 (0.3831) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:12.023881] Epoch: [23] [155/156] eta: 0:00:00 lr: 0.000130 loss: 0.3600 (0.3816) time: 0.0649 data: 0.0001 max mem: 9669 +[14:53:12.100494] Epoch: [23] Total time: 0:00:10 (0.0702 s / it) +[14:53:12.101275] Averaged stats: lr: 0.000130 loss: 0.3600 (0.3816) +[14:53:12.782209] val: [ 0/17] eta: 0:00:11 loss: 0.3133 (0.3133) time: 0.6680 data: 0.6510 max mem: 9669 +[14:53:13.041152] val: [10/17] eta: 0:00:00 loss: 0.2837 (0.2940) time: 0.0842 data: 0.0688 max mem: 9669 +[14:53:13.137230] val: [16/17] eta: 0:00:00 loss: 0.1864 (0.2400) time: 0.0601 data: 0.0448 max mem: 9669 +[14:53:13.221681] val: Total time: 0:00:01 (0.0652 s / it) +[14:53:13.235937] val loss: 0.23999586043988957 +[14:53:13.236143] Accuracy: 0.9222, F1 Score: 0.9222, ROC AUC: 0.9707, Hamming Loss: 0.0778, + Jaccard Score: 0.8556, Precision: 0.9234, Recall: 0.9222, + Average Precision: 0.9711, Kappa: 0.8444, Score: 0.9124 +[14:53:14.879977] Best epoch = 23, Best score = 0.9124 +[14:53:14.949838] log_dir: ./output_logs/retfound +[14:53:15.721693] Epoch: [24] [ 0/156] eta: 0:02:00 lr: 0.000130 loss: 0.4154 (0.4154) time: 0.7710 data: 0.7006 max mem: 9669 +[14:53:17.021739] Epoch: [24] [ 20/156] eta: 0:00:13 lr: 0.000125 loss: 0.3271 (0.3633) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:26.436569] Epoch: [24] [ 40/156] eta: 0:00:32 lr: 0.000120 loss: 0.3947 (0.3858) time: 0.4707 data: 0.0002 max mem: 9669 +[14:53:27.743526] Epoch: [24] [ 60/156] eta: 0:00:20 lr: 0.000115 loss: 0.3738 (0.3892) time: 0.0653 data: 0.0001 max mem: 9669 +[14:53:29.047690] Epoch: [24] [ 80/156] eta: 0:00:13 lr: 0.000110 loss: 0.3735 (0.3839) time: 0.0652 data: 0.0001 max mem: 9669 +[14:53:30.354988] Epoch: [24] [100/156] eta: 0:00:08 lr: 0.000105 loss: 0.3391 (0.3819) time: 0.0653 data: 0.0001 max mem: 9669 +[14:53:31.659591] Epoch: [24] [120/156] eta: 0:00:04 lr: 0.000101 loss: 0.3435 (0.3804) time: 0.0652 data: 0.0001 max mem: 9669 +[14:53:32.956742] Epoch: [24] [140/156] eta: 0:00:02 lr: 0.000096 loss: 0.3509 (0.3794) time: 0.0648 data: 0.0001 max mem: 9669 +[14:53:33.926465] Epoch: [24] [155/156] eta: 0:00:00 lr: 0.000093 loss: 0.3651 (0.3807) time: 0.0646 data: 0.0001 max mem: 9669 +[14:53:34.008911] Epoch: [24] Total time: 0:00:19 (0.1222 s / it) +[14:53:34.009625] Averaged stats: lr: 0.000093 loss: 0.3651 (0.3807) +[14:53:34.779203] val: [ 0/17] eta: 0:00:12 loss: 0.2879 (0.2879) time: 0.7573 data: 0.7406 max mem: 9669 +[14:53:35.003023] val: [10/17] eta: 0:00:00 loss: 0.2879 (0.2952) time: 0.0891 data: 0.0738 max mem: 9669 +[14:53:35.094234] val: [16/17] eta: 0:00:00 loss: 0.1953 (0.2397) time: 0.0630 data: 0.0478 max mem: 9669 +[14:53:35.168344] val: Total time: 0:00:01 (0.0675 s / it) +[14:53:35.182675] val loss: 0.23967786308597117 +[14:53:35.182876] Accuracy: 0.9204, F1 Score: 0.9203, ROC AUC: 0.9703, Hamming Loss: 0.0796, + Jaccard Score: 0.8524, Precision: 0.9217, Recall: 0.9204, + Average Precision: 0.9701, Kappa: 0.8407, Score: 0.9105 +[14:53:35.224631] Best epoch = 23, Best score = 0.9124 +[14:53:35.492527] log_dir: ./output_logs/retfound +[14:53:36.293119] Epoch: [25] [ 0/156] eta: 0:02:04 lr: 0.000092 loss: 0.2706 (0.2706) time: 0.7997 data: 0.7302 max mem: 9669 +[14:53:37.597048] Epoch: [25] [ 20/156] eta: 0:00:13 lr: 0.000088 loss: 0.4050 (0.3961) time: 0.0652 data: 0.0001 max mem: 9669 +[14:53:38.907290] Epoch: [25] [ 40/156] eta: 0:00:09 lr: 0.000084 loss: 0.3465 (0.3772) time: 0.0655 data: 0.0001 max mem: 9669 +[14:53:40.206056] Epoch: [25] [ 60/156] eta: 0:00:07 lr: 0.000079 loss: 0.4074 (0.3937) time: 0.0649 data: 0.0001 max mem: 9669 +[14:53:41.509247] Epoch: [25] [ 80/156] eta: 0:00:05 lr: 0.000075 loss: 0.4057 (0.3941) time: 0.0651 data: 0.0001 max mem: 9669 +[14:53:42.817429] Epoch: [25] [100/156] eta: 0:00:04 lr: 0.000071 loss: 0.3420 (0.3892) time: 0.0654 data: 0.0001 max mem: 9669 +[14:53:44.119636] Epoch: [25] [120/156] eta: 0:00:02 lr: 0.000067 loss: 0.3155 (0.3838) time: 0.0651 data: 0.0001 max mem: 9669 +[14:53:45.430373] Epoch: [25] [140/156] eta: 0:00:01 lr: 0.000064 loss: 0.3565 (0.3815) time: 0.0655 data: 0.0001 max mem: 9669 +[14:53:46.406263] Epoch: [25] [155/156] eta: 0:00:00 lr: 0.000061 loss: 0.3476 (0.3825) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:46.483472] Epoch: [25] Total time: 0:00:10 (0.0705 s / it) +[14:53:46.484203] Averaged stats: lr: 0.000061 loss: 0.3476 (0.3825) +[14:53:47.450148] val: [ 0/17] eta: 0:00:16 loss: 0.1976 (0.1976) time: 0.9451 data: 0.9282 max mem: 9669 +[14:53:47.740911] val: [10/17] eta: 0:00:00 loss: 0.2580 (0.2567) time: 0.1123 data: 0.0969 max mem: 9669 +[14:53:47.832096] val: [16/17] eta: 0:00:00 loss: 0.2172 (0.2330) time: 0.0780 data: 0.0627 max mem: 9669 +[14:53:47.914330] val: Total time: 0:00:01 (0.0829 s / it) +[14:53:47.927855] val loss: 0.2330227932509254 +[14:53:47.928005] Accuracy: 0.9167, F1 Score: 0.9167, ROC AUC: 0.9718, Hamming Loss: 0.0833, + Jaccard Score: 0.8462, Precision: 0.9167, Recall: 0.9167, + Average Precision: 0.9720, Kappa: 0.8333, Score: 0.9073 +[14:53:47.974894] Best epoch = 23, Best score = 0.9124 +[14:53:48.252575] log_dir: ./output_logs/retfound +[14:53:49.029428] Epoch: [26] [ 0/156] eta: 0:02:01 lr: 0.000061 loss: 0.5383 (0.5383) time: 0.7760 data: 0.7071 max mem: 9669 +[14:53:50.329350] Epoch: [26] [ 20/156] eta: 0:00:13 lr: 0.000057 loss: 0.3742 (0.3995) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:51.630545] Epoch: [26] [ 40/156] eta: 0:00:09 lr: 0.000053 loss: 0.3442 (0.3819) time: 0.0650 data: 0.0001 max mem: 9669 +[14:53:52.933436] Epoch: [26] [ 60/156] eta: 0:00:07 lr: 0.000050 loss: 0.3598 (0.3880) time: 0.0651 data: 0.0001 max mem: 9669 +[14:53:54.230805] Epoch: [26] [ 80/156] eta: 0:00:05 lr: 0.000047 loss: 0.3191 (0.3707) time: 0.0648 data: 0.0001 max mem: 9669 +[14:53:56.401863] Epoch: [26] [100/156] eta: 0:00:04 lr: 0.000043 loss: 0.3691 (0.3724) time: 0.1085 data: 0.0001 max mem: 9669 +[14:53:57.720687] Epoch: [26] [120/156] eta: 0:00:02 lr: 0.000040 loss: 0.3316 (0.3727) time: 0.0659 data: 0.0001 max mem: 9669 +[14:53:59.023273] Epoch: [26] [140/156] eta: 0:00:01 lr: 0.000037 loss: 0.3928 (0.3753) time: 0.0651 data: 0.0001 max mem: 9669 +[14:54:00.001760] Epoch: [26] [155/156] eta: 0:00:00 lr: 0.000035 loss: 0.3924 (0.3762) time: 0.0652 data: 0.0001 max mem: 9669 +[14:54:00.078743] Epoch: [26] Total time: 0:00:11 (0.0758 s / it) +[14:54:00.079499] Averaged stats: lr: 0.000035 loss: 0.3924 (0.3762) +[14:54:00.589842] val: [ 0/17] eta: 0:00:08 loss: 0.2516 (0.2516) time: 0.4982 data: 0.4814 max mem: 9669 +[14:54:00.779110] val: [10/17] eta: 0:00:00 loss: 0.2528 (0.2728) time: 0.0624 data: 0.0470 max mem: 9669 +[14:54:00.869972] val: [16/17] eta: 0:00:00 loss: 0.2173 (0.2303) time: 0.0457 data: 0.0305 max mem: 9669 +[14:54:00.946484] val: Total time: 0:00:00 (0.0503 s / it) +[14:54:00.961236] val loss: 0.2302675979102359 +[14:54:00.961396] Accuracy: 0.9241, F1 Score: 0.9241, ROC AUC: 0.9719, Hamming Loss: 0.0759, + Jaccard Score: 0.8588, Precision: 0.9245, Recall: 0.9241, + Average Precision: 0.9721, Kappa: 0.8481, Score: 0.9147 +[14:54:02.739811] Best epoch = 26, Best score = 0.9147 +[14:54:02.816585] log_dir: ./output_logs/retfound +[14:54:03.695518] Epoch: [27] [ 0/156] eta: 0:02:16 lr: 0.000035 loss: 0.3272 (0.3272) time: 0.8780 data: 0.8068 max mem: 9669 +[14:54:04.999924] Epoch: [27] [ 20/156] eta: 0:00:14 lr: 0.000032 loss: 0.3556 (0.3730) time: 0.0652 data: 0.0001 max mem: 9669 +[14:54:06.311144] Epoch: [27] [ 40/156] eta: 0:00:09 lr: 0.000030 loss: 0.3624 (0.3686) time: 0.0655 data: 0.0001 max mem: 9669 +[14:54:16.400033] Epoch: [27] [ 60/156] eta: 0:00:21 lr: 0.000027 loss: 0.3282 (0.3590) time: 0.5044 data: 0.0001 max mem: 9669 +[14:54:17.697718] Epoch: [27] [ 80/156] eta: 0:00:13 lr: 0.000025 loss: 0.3779 (0.3642) time: 0.0648 data: 0.0001 max mem: 9669 +[14:54:18.995739] Epoch: [27] [100/156] eta: 0:00:08 lr: 0.000022 loss: 0.3637 (0.3665) time: 0.0649 data: 0.0001 max mem: 9669 +[14:54:20.295129] Epoch: [27] [120/156] eta: 0:00:05 lr: 0.000020 loss: 0.3731 (0.3708) time: 0.0649 data: 0.0001 max mem: 9669 +[14:54:21.595503] Epoch: [27] [140/156] eta: 0:00:02 lr: 0.000018 loss: 0.3775 (0.3725) time: 0.0650 data: 0.0001 max mem: 9669 +[14:54:22.572905] Epoch: [27] [155/156] eta: 0:00:00 lr: 0.000016 loss: 0.4017 (0.3799) time: 0.0651 data: 0.0001 max mem: 9669 +[14:54:22.652246] Epoch: [27] Total time: 0:00:19 (0.1272 s / it) +[14:54:22.653145] Averaged stats: lr: 0.000016 loss: 0.4017 (0.3799) +[14:54:23.306295] val: [ 0/17] eta: 0:00:10 loss: 0.2363 (0.2363) time: 0.6284 data: 0.6114 max mem: 9669 +[14:54:23.550652] val: [10/17] eta: 0:00:00 loss: 0.2602 (0.2659) time: 0.0793 data: 0.0638 max mem: 9669 +[14:54:23.647470] val: [16/17] eta: 0:00:00 loss: 0.2213 (0.2283) time: 0.0570 data: 0.0417 max mem: 9669 +[14:54:23.721244] val: Total time: 0:00:01 (0.0614 s / it) +[14:54:23.736078] val loss: 0.22829260633272283 +[14:54:23.736248] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9718, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9719, Kappa: 0.8556, Score: 0.9184 +[14:54:25.345224] Best epoch = 27, Best score = 0.9184 +[14:54:25.427711] log_dir: ./output_logs/retfound +[14:54:26.184892] Epoch: [28] [ 0/156] eta: 0:01:57 lr: 0.000016 loss: 0.5803 (0.5803) time: 0.7563 data: 0.6812 max mem: 9669 +[14:54:27.494326] Epoch: [28] [ 20/156] eta: 0:00:13 lr: 0.000014 loss: 0.3818 (0.3711) time: 0.0654 data: 0.0001 max mem: 9669 +[14:54:28.794664] Epoch: [28] [ 40/156] eta: 0:00:09 lr: 0.000013 loss: 0.3826 (0.3896) time: 0.0650 data: 0.0001 max mem: 9669 +[14:54:30.093623] Epoch: [28] [ 60/156] eta: 0:00:07 lr: 0.000011 loss: 0.3227 (0.3742) time: 0.0649 data: 0.0001 max mem: 9669 +[14:54:31.394858] Epoch: [28] [ 80/156] eta: 0:00:05 lr: 0.000009 loss: 0.3472 (0.3726) time: 0.0650 data: 0.0001 max mem: 9669 +[14:54:32.697595] Epoch: [28] [100/156] eta: 0:00:04 lr: 0.000008 loss: 0.3824 (0.3784) time: 0.0651 data: 0.0001 max mem: 9669 +[14:54:34.007381] Epoch: [28] [120/156] eta: 0:00:02 lr: 0.000007 loss: 0.3396 (0.3717) time: 0.0654 data: 0.0001 max mem: 9669 +[14:54:35.303757] Epoch: [28] [140/156] eta: 0:00:01 lr: 0.000006 loss: 0.3599 (0.3713) time: 0.0648 data: 0.0001 max mem: 9669 +[14:54:36.276411] Epoch: [28] [155/156] eta: 0:00:00 lr: 0.000005 loss: 0.3999 (0.3723) time: 0.0647 data: 0.0001 max mem: 9669 +[14:54:36.356836] Epoch: [28] Total time: 0:00:10 (0.0701 s / it) +[14:54:36.357611] Averaged stats: lr: 0.000005 loss: 0.3999 (0.3723) +[14:54:37.059553] val: [ 0/17] eta: 0:00:11 loss: 0.2405 (0.2405) time: 0.6770 data: 0.6600 max mem: 9669 +[14:54:37.318419] val: [10/17] eta: 0:00:00 loss: 0.2569 (0.2667) time: 0.0850 data: 0.0696 max mem: 9669 +[14:54:37.409430] val: [16/17] eta: 0:00:00 loss: 0.2159 (0.2277) time: 0.0603 data: 0.0451 max mem: 9669 +[14:54:37.481298] val: Total time: 0:00:01 (0.0647 s / it) +[14:54:37.497934] val loss: 0.22767536298317068 +[14:54:37.498118] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9718, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9718, Kappa: 0.8556, Score: 0.9184 +[14:54:39.156392] Best epoch = 28, Best score = 0.9184 +[14:54:39.247752] log_dir: ./output_logs/retfound +[14:54:40.054711] Epoch: [29] [ 0/156] eta: 0:02:05 lr: 0.000005 loss: 0.3711 (0.3711) time: 0.8061 data: 0.7372 max mem: 9669 +[14:54:41.358863] Epoch: [29] [ 20/156] eta: 0:00:13 lr: 0.000004 loss: 0.3350 (0.3650) time: 0.0652 data: 0.0001 max mem: 9669 +[14:54:42.662988] Epoch: [29] [ 40/156] eta: 0:00:09 lr: 0.000003 loss: 0.3628 (0.3587) time: 0.0652 data: 0.0001 max mem: 9669 +[14:54:43.965893] Epoch: [29] [ 60/156] eta: 0:00:07 lr: 0.000002 loss: 0.3660 (0.3623) time: 0.0651 data: 0.0001 max mem: 9669 +[14:54:45.267131] Epoch: [29] [ 80/156] eta: 0:00:05 lr: 0.000002 loss: 0.3133 (0.3579) time: 0.0650 data: 0.0001 max mem: 9669 +[14:54:46.577414] Epoch: [29] [100/156] eta: 0:00:04 lr: 0.000001 loss: 0.3825 (0.3635) time: 0.0655 data: 0.0001 max mem: 9669 +[14:54:47.886994] Epoch: [29] [120/156] eta: 0:00:02 lr: 0.000001 loss: 0.4103 (0.3692) time: 0.0654 data: 0.0002 max mem: 9669 +[14:54:49.191514] Epoch: [29] [140/156] eta: 0:00:01 lr: 0.000001 loss: 0.3245 (0.3648) time: 0.0652 data: 0.0001 max mem: 9669 +[14:54:50.164135] Epoch: [29] [155/156] eta: 0:00:00 lr: 0.000001 loss: 0.3686 (0.3648) time: 0.0649 data: 0.0001 max mem: 9669 +[14:54:50.239639] Epoch: [29] Total time: 0:00:10 (0.0705 s / it) +[14:54:50.240394] Averaged stats: lr: 0.000001 loss: 0.3686 (0.3648) +[14:54:50.900098] val: [ 0/17] eta: 0:00:10 loss: 0.2422 (0.2422) time: 0.6342 data: 0.6172 max mem: 9669 +[14:54:51.164995] val: [10/17] eta: 0:00:00 loss: 0.2575 (0.2679) time: 0.0817 data: 0.0662 max mem: 9669 +[14:54:51.256000] val: [16/17] eta: 0:00:00 loss: 0.2153 (0.2283) time: 0.0582 data: 0.0429 max mem: 9669 +[14:54:51.330006] val: Total time: 0:00:01 (0.0626 s / it) +[14:54:51.344338] val loss: 0.22825184420627706 +[14:54:51.344640] Accuracy: 0.9278, F1 Score: 0.9278, ROC AUC: 0.9719, Hamming Loss: 0.0722, + Jaccard Score: 0.8653, Precision: 0.9283, Recall: 0.9278, + Average Precision: 0.9720, Kappa: 0.8556, Score: 0.9184 +[14:54:52.994703] Best epoch = 29, Best score = 0.9184 +[14:54:55.370917] Test with the best model, epoch = 29: +[14:54:55.997484] test: [ 0/32] eta: 0:00:19 loss: 0.1497 (0.1497) time: 0.6109 data: 0.5935 max mem: 9669 +[14:54:56.258732] test: [10/32] eta: 0:00:01 loss: 0.2625 (0.2276) time: 0.0792 data: 0.0637 max mem: 9669 +[14:54:56.491569] test: [20/32] eta: 0:00:00 loss: 0.2398 (0.2350) time: 0.0246 data: 0.0087 max mem: 9669 +[14:54:56.744292] test: [30/32] eta: 0:00:00 loss: 0.2097 (0.2370) time: 0.0242 data: 0.0075 max mem: 9669 +[14:54:56.854703] test: [31/32] eta: 0:00:00 loss: 0.2398 (0.2516) time: 0.0287 data: 0.0075 max mem: 9669 +[14:54:56.923648] test: Total time: 0:00:01 (0.0481 s / it) +[14:54:56.941643] val loss: 0.2516249555628747 +[14:54:56.941753] Accuracy: 0.9080, F1 Score: 0.9080, ROC AUC: 0.9707, Hamming Loss: 0.0920, + Jaccard Score: 0.8315, Precision: 0.9084, Recall: 0.9080, + Average Precision: 0.9704, Kappa: 0.8160, Score: 0.8982 +[14:54:57.701478] Training time 0:08:04 +[rank0]:[W701 14:54:58.093967911 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/100/retfound acc=0.9080 auroc=0.970758 f1_macro=0.9080 qwk=0.8160000000000001 diff --git a/results/downsample/airogs/100/vit/confusion_matrix.png b/results/downsample/airogs/100/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..afe999d377cf4d3b3d6cd13aa77e07d14c471f9f --- /dev/null +++ b/results/downsample/airogs/100/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:217a9cd7e265081bfe83af03a270d65560e7c0124557a654b5700030f56a93ef +size 72148 diff --git a/results/downsample/airogs/100/vit/log.csv b/results/downsample/airogs/100/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..0396721977e0521518f00b444563927c3f5e29a4 --- /dev/null +++ b/results/downsample/airogs/100/vit/log.csv @@ -0,0 +1,24 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6942946353019812,0.7092592592592593,0.768326474622771,0.6320078228812479,7.299611630985668e-08 +1,0.5653198212385178,0.7851851851851852,0.8664334705075445,0.7404699716071924,1.46940234130231e-07 +2,0.5144480100044837,0.8018518518518518,0.8898696844993141,0.7643975340952848,2.208843519506053e-07 +3,0.48243222213708437,0.8277777777777777,0.9118244170096022,0.7983699658481062,2.948284697709796e-07 +4,0.45127938305720305,0.8481481481481481,0.9218998628257887,0.8220974046701975,3.6877258759135393e-07 +5,0.4164263606071472,0.8462962962962963,0.9283401920438957,0.8220827483841955,3.6829999960923774e-07 +6,0.3999119840371303,0.8537037037037037,0.9364540466392318,0.8319722538801697,3.639866962939037e-07 +7,0.3715562627483637,0.8703703703703703,0.9364334705075447,0.8488927294233127,3.568484626181574e-07 +8,0.3723203562773191,0.8611111111111112,0.9361316872427983,0.839371360682192,3.4699787279123945e-07 +9,0.353521853303298,0.8611111111111112,0.9354869684499314,0.8395939036055461,3.3459027649917746e-07 +10,0.33366694817176235,0.8703703703703703,0.929190672153635,0.8464514160026294,3.19821348947436e-07 +11,0.32187017187094075,0.8555555555555555,0.9335116598079561,0.8327068204619942,3.0292400494603573e-07 +12,0.315995030487195,0.8814814814814815,0.9400068587105623,0.8614055763816914,2.841647257038628e-07 +13,0.30471921616639847,0.8759259259259259,0.938079561042524,0.8550689229726932,2.638393562608164e-07 +14,0.29201855835242146,0.8740740740740741,0.93960219478738,0.853871685552631,2.4226843983480064e-07 +15,0.29112845812088406,0.8685185185185185,0.9397050754458162,0.8484128444497284,2.1979216266369462e-07 +16,0.2847846393019725,0.8759259259259259,0.9393964334705076,0.8556837071555163,1.9676498906516377e-07 +17,0.2747403339315683,0.8722222222222222,0.9327709190672154,0.8496893324933327,1.7355007132262385e-07 +18,0.26897891629964876,0.8759259259259259,0.9358916323731139,0.8544014367163136,1.505135225567943e-07 +19,0.27152149780438495,0.8759259259259259,0.9285733882030178,0.8519620219929483,1.2801864290307935e-07 +20,0.2643142025440167,0.8777777777777778,0.9327846364883401,0.8553525277827477,1.0642019005140791e-07 +21,0.2558365009534053,0.8759259259259259,0.9285528120713307,0.8520350314674481,8.605878450554121e-08 +22,0.2579246640969545,0.8833333333333333,0.9299519890260631,0.8599004649198866,6.725553779425159e-08 diff --git a/results/downsample/airogs/100/vit/metrics.json b/results/downsample/airogs/100/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..0eb1b1b66e8d455c1c5a705557322c81490dd1dd --- /dev/null +++ b/results/downsample/airogs/100/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 1000, + "n_classes": 2, + "task": "binary", + "accuracy": 0.873, + "balanced_accuracy": 0.873, + "precision_macro": 0.8731806194197991, + "recall_macro": 0.873, + "f1_macro": 0.872984631140368, + "precision_weighted": 0.8731806194197991, + "recall_weighted": 0.873, + "f1_weighted": 0.8729846311403681, + "cohen_kappa": 0.746, + "quadratic_weighted_kappa": 0.746, + "mcc": 0.7461805975595589, + "auroc": 0.945154, + "auprc": 0.9413931705313047, + "sensitivity": 0.884, + "specificity": 0.862, + "precision_pos": 0.8649706457925636, + "f1_pos": 0.874381800197824, + "per_class": { + "0": { + "precision": 0.8813905930470347, + "recall": 0.862, + "f1-score": 0.871587462082912, + "support": 500.0 + }, + "1": { + "precision": 0.8649706457925636, + "recall": 0.884, + "f1-score": 0.874381800197824, + "support": 500.0 + }, + "accuracy": 0.873, + "macro avg": { + "precision": 0.8731806194197991, + "recall": 0.873, + "f1-score": 0.872984631140368, + "support": 1000.0 + }, + "weighted avg": { + "precision": 0.8731806194197991, + "recall": 0.873, + "f1-score": 0.8729846311403681, + "support": 1000.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/airogs/100/vit/pr.png b/results/downsample/airogs/100/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a69b318aefe1fba67c15c810d8bd57e3bf3abac4 --- /dev/null +++ b/results/downsample/airogs/100/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:44d9f03d36982781d81a50ffdab843d21f26f95eeeef002b6eb268c9525988eb +size 47379 diff --git a/results/downsample/airogs/100/vit/roc.png b/results/downsample/airogs/100/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..19d8145af0d68f8da73a901e90806b1b1da26db7 --- /dev/null +++ b/results/downsample/airogs/100/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f45ca6663c1f5cafd2b23eb904a600fa944e1ad92785f4e77cebd911a3762209 +size 62075 diff --git a/results/downsample/airogs/100/vit/test_pred.npz b/results/downsample/airogs/100/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..b6c8ee5e5914280afc8fd9b8948093cd6c685682 --- /dev/null +++ b/results/downsample/airogs/100/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a77a100181926043000b8bfce92b35bad5818ccf1bea32877c364823039b0a33 +size 16510 diff --git a/results/downsample/airogs/100/vit/train.log b/results/downsample/airogs/100/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d6835b671fdc431273ba3b3a74adb6f057d20e9b --- /dev/null +++ b/results/downsample/airogs/100/vit/train.log @@ -0,0 +1,124 @@ +[vit] train=5000 val=540 test=1000 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.6943 val_acc=0.7093 val_auc=0.7683 score=0.6320 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.5653 val_acc=0.7852 val_auc=0.8664 score=0.7405 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.5144 val_acc=0.8019 val_auc=0.8899 score=0.7644 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.4824 val_acc=0.8278 val_auc=0.9118 score=0.7984 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.4513 val_acc=0.8481 val_auc=0.9219 score=0.8221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.4164 val_acc=0.8463 val_auc=0.9283 score=0.8221 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.3999 val_acc=0.8537 val_auc=0.9365 score=0.8320 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.3716 val_acc=0.8704 val_auc=0.9364 score=0.8489 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3723 val_acc=0.8611 val_auc=0.9361 score=0.8394 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.3535 val_acc=0.8611 val_auc=0.9355 score=0.8396 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3337 val_acc=0.8704 val_auc=0.9292 score=0.8465 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.3219 val_acc=0.8556 val_auc=0.9335 score=0.8327 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.3160 val_acc=0.8815 val_auc=0.9400 score=0.8614 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.3047 val_acc=0.8759 val_auc=0.9381 score=0.8551 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.2920 val_acc=0.8741 val_auc=0.9396 score=0.8539 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.2911 val_acc=0.8685 val_auc=0.9397 score=0.8484 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.2848 val_acc=0.8759 val_auc=0.9394 score=0.8557 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.2747 val_acc=0.8722 val_auc=0.9328 score=0.8497 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.2690 val_acc=0.8759 val_auc=0.9359 score=0.8544 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.2715 val_acc=0.8759 val_auc=0.9286 score=0.8520 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.2643 val_acc=0.8778 val_auc=0.9328 score=0.8554 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.2558 val_acc=0.8759 val_auc=0.9286 score=0.8520 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.2579 val_acc=0.8833 val_auc=0.9300 score=0.8599 +[vit] early stop at ep22 (best ep12 score=0.8614) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=12 best_val_score=0.8614 -> saved test_pred.npz (1000 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/airogs/100/vit acc=0.8730 auroc=0.945154 f1_macro=0.8730 qwk=0.746 diff --git a/results/downsample/curve_adam_combined.png b/results/downsample/curve_adam_combined.png new file mode 100644 index 0000000000000000000000000000000000000000..3aeea651625c9c2298436dce26b2dcfb7efec7ef --- /dev/null +++ b/results/downsample/curve_adam_combined.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5fade09e10f4a304d52f72c2fe617eaa8eba6925d8474b3cf76d32d061c40d38 +size 74557 diff --git a/results/downsample/curve_resnet.png b/results/downsample/curve_resnet.png new file mode 100644 index 0000000000000000000000000000000000000000..9be2ec5ef6a4914f38b503cd4e4b3bb510bc8ed6 --- /dev/null +++ b/results/downsample/curve_resnet.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05c179594f5b2849830a86f10ccb58709b2c1b5732c46bb03f4941f5b6a43014 +size 106751 diff --git a/results/downsample/curve_retfound.png b/results/downsample/curve_retfound.png new file mode 100644 index 0000000000000000000000000000000000000000..0ffd664d1bd510d0a7b7fa8c423041902c1f5f05 --- /dev/null +++ b/results/downsample/curve_retfound.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b5b73977468bc5dad38fdd50689b4aaa7db1e9227338ab96e54ce0a1a38deabf +size 103950 diff --git a/results/downsample/curve_vit.png b/results/downsample/curve_vit.png new file mode 100644 index 0000000000000000000000000000000000000000..11885d818e185134f83435253693917b2d0426a3 --- /dev/null +++ b/results/downsample/curve_vit.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3a17e0ea1f0b06b8188b1c9fda593aa2f1a45acd2c0cbd425b86ffe9d6a866a +size 104196 diff --git a/results/downsample/papila/005/resnet/confusion_matrix.png b/results/downsample/papila/005/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..feb6460711b7f169000380c14ca79d4b676bebe6 --- /dev/null +++ b/results/downsample/papila/005/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5d6e24e265177fd840f774e30afc01f1724e1eebf1cb1cf32c27464021e786f2 +size 73829 diff --git a/results/downsample/papila/005/resnet/log.csv b/results/downsample/papila/005/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..9f34db3380c2661505bb026d2c0d7f126f52079f --- /dev/null +++ b/results/downsample/papila/005/resnet/log.csv @@ -0,0 +1,16 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6958895325660706,0.2857142857142857,0.6312500000000001,0.30694758672699846,0.0 +1,0.6953569054603577,0.30952380952380953,0.540625,0.2743594980727334,0.00016666666666666666 +2,0.6739612817764282,0.5476190476190477,0.546875,0.3854720664386672,0.0003333333333333333 +3,0.6804917454719543,0.6666666666666666,0.546875,0.3551161079474838,0.0005 +4,0.6596331000328064,0.7619047619047619,0.503125,0.4159273059617548,0.0004997919498361457 +5,0.64990234375,0.7380952380952381,0.490625,0.290011221822765,0.0004991681456235483 +6,0.6881318688392639,0.7380952380952381,0.45,0.27646955515609833,0.000498129625622757 +7,0.6505301594734192,0.7380952380952381,0.384375,0.25459455515609836,0.0004966781183478222 +8,0.62603759765625,0.7142857142857143,0.328125,0.21952825670498086,0.0004948160396893552 +9,0.64306640625,0.7380952380952381,0.3,0.22646955515609832,0.0004925464888935161 +10,0.6254185438156128,0.7619047619047619,0.22812499999999997,0.2201858108108108,0.0004898732434036243 +11,0.6194957494735718,0.7619047619047619,0.25625,0.2295608108108108,0.00048680075257297753 +12,0.5725319385528564,0.7619047619047619,0.309375,0.24726914414414414,0.0004833341302593417 +13,0.5772007703781128,0.7619047619047619,0.30625,0.2462274774774775,0.0004794791463134399 +14,0.6087757349014282,0.7619047619047619,0.29375,0.2420608108108108,0.00047524221697560476 diff --git a/results/downsample/papila/005/resnet/metrics.json b/results/downsample/papila/005/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..45c14d5c1ebe0e3fd345b80661c72231a84872fe --- /dev/null +++ b/results/downsample/papila/005/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8214285714285714, + "balanced_accuracy": 0.6029411764705883, + "precision_macro": 0.7077922077922078, + "recall_macro": 0.6029411764705883, + "f1_macro": 0.6221889055472264, + "precision_weighted": 0.7922077922077922, + "recall_weighted": 0.8214285714285714, + "f1_weighted": 0.7920325551509959, + "cohen_kappa": 0.2622950819672131, + "quadratic_weighted_kappa": 0.2622950819672131, + "mcc": 0.29250896965085227, + "auroc": 0.5647977941176471, + "auprc": 0.351946785352831, + "sensitivity": 0.25, + "specificity": 0.9558823529411765, + "precision_pos": 0.5714285714285714, + "f1_pos": 0.34782608695652173, + "per_class": { + "0": { + "precision": 0.8441558441558441, + "recall": 0.9558823529411765, + "f1-score": 0.896551724137931, + "support": 68.0 + }, + "1": { + "precision": 0.5714285714285714, + "recall": 0.25, + "f1-score": 0.34782608695652173, + "support": 16.0 + }, + "accuracy": 0.8214285714285714, + "macro avg": { + "precision": 0.7077922077922078, + "recall": 0.6029411764705883, + "f1-score": 0.6221889055472264, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7922077922077922, + "recall": 0.8214285714285714, + "f1-score": 0.7920325551509959, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/005/resnet/pr.png b/results/downsample/papila/005/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..9332fd0881c456b2586595fa0e4db0aba233c55d --- /dev/null +++ b/results/downsample/papila/005/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e5253806da37eb0a43bbeb67c0e8d43b193be04360b64172470db6ec9bb0687 +size 50793 diff --git a/results/downsample/papila/005/resnet/roc.png b/results/downsample/papila/005/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..68d2600ea1a0434e0c852074e47691355234e453 --- /dev/null +++ b/results/downsample/papila/005/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b9d205a2d9294073234109ca9a5b83880338dea8d500ede817d0d3471c44f8c +size 58032 diff --git a/results/downsample/papila/005/resnet/test_pred.npz b/results/downsample/papila/005/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..b3138bcaf20f287b891b0e5a9f9e0d17f7670a80 --- /dev/null +++ b/results/downsample/papila/005/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5efcd5dfc892b96440d5c9d207aa93dac2b5050ab1194b2e9d5bf87ee6bcb1ac +size 1854 diff --git a/results/downsample/papila/005/resnet/train.log b/results/downsample/papila/005/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..51913586d9e7cb25df51c969514ff976a3d63dd3 --- /dev/null +++ b/results/downsample/papila/005/resnet/train.log @@ -0,0 +1,84 @@ +[resnet] train=15 val=42 test=84 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6959 val_acc=0.2857 val_auc=0.6313 score=0.3069 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6954 val_acc=0.3095 val_auc=0.5406 score=0.2744 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6740 val_acc=0.5476 val_auc=0.5469 score=0.3855 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6805 val_acc=0.6667 val_auc=0.5469 score=0.3551 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6596 val_acc=0.7619 val_auc=0.5031 score=0.4159 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6499 val_acc=0.7381 val_auc=0.4906 score=0.2900 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6881 val_acc=0.7381 val_auc=0.4500 score=0.2765 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6505 val_acc=0.7381 val_auc=0.3844 score=0.2546 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6260 val_acc=0.7143 val_auc=0.3281 score=0.2195 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6431 val_acc=0.7381 val_auc=0.3000 score=0.2265 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.6254 val_acc=0.7619 val_auc=0.2281 score=0.2202 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.6195 val_acc=0.7619 val_auc=0.2562 score=0.2296 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.5725 val_acc=0.7619 val_auc=0.3094 score=0.2473 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.5772 val_acc=0.7619 val_auc=0.3063 score=0.2462 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.6088 val_acc=0.7619 val_auc=0.2938 score=0.2421 +[resnet] early stop at ep14 (best ep4 score=0.4159) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=4 best_val_score=0.4159 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/005/resnet acc=0.8214 auroc=0.5647977941176471 f1_macro=0.6222 qwk=0.2622950819672131 diff --git a/results/downsample/papila/005/retfound/confusion_matrix.png b/results/downsample/papila/005/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7627f40aad23385b9ca50fc133405e269ee87449 --- /dev/null +++ b/results/downsample/papila/005/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7c220a1f8fb4a561324744530bb09bbf8cda1e520112fbfcbb7517b6f9ee185 +size 73143 diff --git a/results/downsample/papila/005/retfound/confusion_matrix_test.jpg b/results/downsample/papila/005/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a44063cdd273cf01b218e16dedbcc5eb44589de0 --- /dev/null +++ b/results/downsample/papila/005/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbf6a7f569061967235b05e90273681d0d6dd737bd7ced46cbededfec57d2cd +size 241221 diff --git a/results/downsample/papila/005/retfound/log.txt b/results/downsample/papila/005/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..03f0193030178ce66c432f8e16ad6066a5cab823 --- /dev/null +++ b/results/downsample/papila/005/retfound/log.txt @@ -0,0 +1,80 @@ +{"train_lr": 0.0, "train_loss": 0.69293212890625, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 1.5625e-05, "train_loss": 0.69244384765625, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 3.125e-05, "train_loss": 0.69281005859375, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 4.6875e-05, "train_loss": 0.69293212890625, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 6.25e-05, "train_loss": 0.68408203125, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 7.8125e-05, "train_loss": 0.669189453125, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 9.375e-05, "train_loss": 0.661376953125, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00010937500000000002, "train_loss": 0.63909912109375, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000125, "train_loss": 0.63385009765625, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000140625, "train_loss": 0.56292724609375, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.00015625, "train_loss": 0.55078125, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.00015617183679026957, "train_loss": 0.62762451171875, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.00015593750457138045, "train_loss": 0.599365234375, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.00015554747525723593, "train_loss": 0.5745849609375, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.00015500253431496863, "train_loss": 0.438934326171875, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.00015430377918311406, "train_loss": 0.4346923828125, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.00015345261706151562, "train_loss": 0.544097900390625, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0001524507620774113, "train_loss": 0.714996337890625, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0001513002318334096, "train_loss": 0.7147064208984375, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00015000334334430636, "train_loss": 0.5786895751953125, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0001485627083709253, "train_loss": 0.3713531494140625, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.00014698122816037928, "train_loss": 0.241729736328125, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00014526208760334486, "train_loss": 0.410125732421875, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0001434087488201161, "train_loss": 0.402008056640625, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0001414249441883553, "train_loss": 0.799072265625, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.00013931466882658082, "train_loss": 0.618743896484375, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0001370821725485303, "train_loss": 0.60369873046875, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.00013473195130460128, "train_loss": 0.42083740234375, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00013226873812760537, "train_loss": 0.3929595947265625, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0001296974936010702, "train_loss": 0.178863525390625, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0001270233958692842, "train_loss": 0.72174072265625, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.00012425183020920322, "train_loss": 0.160675048828125, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00012138837818521947, "train_loss": 0.6429595947265625, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00011843880640863348, "train_loss": 0.5572052001953125, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00011540905492446652, "train_loss": 0.6110382080078125, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00011230522524900044, "train_loss": 0.3163604736328125, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00010913356808213583, "train_loss": 0.1724090576171875, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00010590047071931423, "train_loss": 0.3562164306640625, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.0001026124441883553, "train_loss": 0.363037109375, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 9.927611013711315e-05, "train_loss": 0.574371337890625, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 9.58981874983589e-05, "train_loss": 0.7217254638671875, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 9.248547895874417e-05, "train_loss": 0.5898895263671875, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 8.904485725909551e-05, "train_loss": 0.5131683349609375, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 8.558325135362903e-05, "train_loss": 0.3763885498046875, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 8.210763245595873e-05, "train_loss": 0.5915679931640625, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 7.8625e-05, "train_loss": 0.540679931640625, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 7.514236754404128e-05, "train_loss": 0.39801025390625, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.166674864637099e-05, "train_loss": 0.6185302734375, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 6.82051427409045e-05, "train_loss": 0.4056549072265625, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 6.476452104125584e-05, "train_loss": 0.2036590576171875, "epoch": 49, "n_parameters": 303303682} +{"train_lr": 6.13518125016411e-05, "train_loss": 0.3745880126953125, "epoch": 50, "n_parameters": 303303682} +{"train_lr": 5.7973889862886864e-05, "train_loss": 0.76910400390625, "epoch": 51, "n_parameters": 303303682} +{"train_lr": 5.4637555811644714e-05, "train_loss": 0.7091064453125, "epoch": 52, "n_parameters": 303303682} +{"train_lr": 5.134952928068578e-05, "train_loss": 0.535400390625, "epoch": 53, "n_parameters": 303303682} +{"train_lr": 4.81164319178642e-05, "train_loss": 0.571685791015625, "epoch": 54, "n_parameters": 303303682} +{"train_lr": 4.4944774750999556e-05, "train_loss": 0.3514556884765625, "epoch": 55, "n_parameters": 303303682} +{"train_lr": 4.184094507553348e-05, "train_loss": 0.35092926025390625, "epoch": 56, "n_parameters": 303303682} +{"train_lr": 3.881119359136654e-05, "train_loss": 0.5604705810546875, "epoch": 57, "n_parameters": 303303682} +{"train_lr": 3.586162181478055e-05, "train_loss": 0.662078857421875, "epoch": 58, "n_parameters": 303303682} +{"train_lr": 3.299816979079678e-05, "train_loss": 0.5900726318359375, "epoch": 59, "n_parameters": 303303682} +{"train_lr": 3.0226604130715816e-05, "train_loss": 0.38507080078125, "epoch": 60, "n_parameters": 303303682} +{"train_lr": 2.755250639892981e-05, "train_loss": 0.5032501220703125, "epoch": 61, "n_parameters": 303303682} +{"train_lr": 2.4981261872394632e-05, "train_loss": 0.523040771484375, "epoch": 62, "n_parameters": 303303682} +{"train_lr": 2.251804869539874e-05, "train_loss": 0.4280242919921875, "epoch": 63, "n_parameters": 303303682} +{"train_lr": 2.016782745146971e-05, "train_loss": 0.396331787109375, "epoch": 64, "n_parameters": 303303682} +{"train_lr": 1.7935331173419187e-05, "train_loss": 0.581207275390625, "epoch": 65, "n_parameters": 303303682} +{"train_lr": 1.5825055811644713e-05, "train_loss": 0.2144775390625, "epoch": 66, "n_parameters": 303303682} +{"train_lr": 1.3841251179883884e-05, "train_loss": 0.6783447265625, "epoch": 67, "n_parameters": 303303682} +{"train_lr": 1.1987912396655145e-05, "train_loss": 0.5520477294921875, "epoch": 68, "n_parameters": 303303682} +{"train_lr": 1.0268771839620714e-05, "train_loss": 0.4133453369140625, "epoch": 69, "n_parameters": 303303682} +{"train_lr": 8.687291629074723e-06, "train_loss": 0.51300048828125, "epoch": 70, "n_parameters": 303303682} +{"train_lr": 7.246656655693649e-06, "train_loss": 0.54205322265625, "epoch": 71, "n_parameters": 303303682} +{"train_lr": 5.949768166590399e-06, "train_loss": 0.509307861328125, "epoch": 72, "n_parameters": 303303682} +{"train_lr": 4.799237922588707e-06, "train_loss": 0.2194976806640625, "epoch": 73, "n_parameters": 303303682} +{"train_lr": 3.797382938484392e-06, "train_loss": 0.637542724609375, "epoch": 74, "n_parameters": 303303682} +{"train_lr": 2.9462208168859414e-06, "train_loss": 0.55511474609375, "epoch": 75, "n_parameters": 303303682} +{"train_lr": 2.247465685031372e-06, "train_loss": 0.5967254638671875, "epoch": 76, "n_parameters": 303303682} +{"train_lr": 1.702524742764069e-06, "train_loss": 0.5454254150390625, "epoch": 77, "n_parameters": 303303682} +{"train_lr": 1.3124954286195674e-06, "train_loss": 0.5909576416015625, "epoch": 78, "n_parameters": 303303682} +{"train_lr": 1.078163209730455e-06, "train_loss": 0.3428955078125, "epoch": 79, "n_parameters": 303303682} diff --git a/results/downsample/papila/005/retfound/metrics.json b/results/downsample/papila/005/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..ce48fa79a7d9525dbf04acbf3418e5e9c7850577 --- /dev/null +++ b/results/downsample/papila/005/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8095238095238095, + "balanced_accuracy": 0.5, + "precision_macro": 0.40476190476190477, + "recall_macro": 0.5, + "f1_macro": 0.4473684210526316, + "precision_weighted": 0.655328798185941, + "recall_weighted": 0.8095238095238095, + "f1_weighted": 0.7243107769423559, + "cohen_kappa": 0.0, + "quadratic_weighted_kappa": 0.0, + "mcc": 0.0, + "auroc": 0.6806066176470589, + "auprc": 0.3678665444945403, + "sensitivity": 0.0, + "specificity": 1.0, + "precision_pos": null, + "f1_pos": 0.0, + "per_class": { + "0": { + "precision": 0.8095238095238095, + "recall": 1.0, + "f1-score": 0.8947368421052632, + "support": 68.0 + }, + "1": { + "precision": 0.0, + "recall": 0.0, + "f1-score": 0.0, + "support": 16.0 + }, + "accuracy": 0.8095238095238095, + "macro avg": { + "precision": 0.40476190476190477, + "recall": 0.5, + "f1-score": 0.4473684210526316, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.655328798185941, + "recall": 0.8095238095238095, + "f1-score": 0.7243107769423559, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/005/retfound/metrics_test.csv b/results/downsample/papila/005/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..8d39a02033c5767977491979fd76d8dc16d68383 --- /dev/null +++ b/results/downsample/papila/005/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.53875732421875,0.8095238095238095,0.4473684210526316,0.6810661764705883,0.19047619047619047,0.40476190476190477,0.40476190476190477,0.5,0.6268273255396831,0.0 diff --git a/results/downsample/papila/005/retfound/metrics_val.csv b/results/downsample/papila/005/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..843865eb42cb562f7d097df81d0554ab37544422 --- /dev/null +++ b/results/downsample/papila/005/retfound/metrics_val.csv @@ -0,0 +1,81 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6928914388020834,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6928914388020834,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6928914388020834,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6876627604166666,0.7619047619047619,0.43243243243243246,0.59765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.5817494833839406,0.0 +0.6818339029947916,0.7619047619047619,0.43243243243243246,0.66875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6415219153080075,0.0 +0.6751810709635416,0.7619047619047619,0.43243243243243246,0.66328125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6364903593344503,0.0 +0.6675618489583334,0.7619047619047619,0.43243243243243246,0.67578125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6521439271466601,0.0 +0.6597493489583334,0.7619047619047619,0.43243243243243246,0.6820312500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6613980767116134,0.0 +0.651519775390625,0.7619047619047619,0.43243243243243246,0.70625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6815876633934559,0.0 +0.643096923828125,0.7619047619047619,0.43243243243243246,0.72890625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6925687377355045,0.0 +0.635162353515625,0.7619047619047619,0.43243243243243246,0.7453125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6999056486927258,0.0 +0.629302978515625,0.7619047619047619,0.43243243243243246,0.7734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7191306545519696,0.0 +0.6263631184895834,0.7619047619047619,0.43243243243243246,0.7859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7261844444862844,0.0 +0.6251169840494791,0.7619047619047619,0.43243243243243246,0.79765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7464141177418806,0.0 +0.625579833984375,0.7619047619047619,0.43243243243243246,0.8046875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7449708390080166,0.0 +0.6277618408203125,0.7619047619047619,0.43243243243243246,0.8109375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7503945099812745,0.0 +0.6307932535807291,0.7619047619047619,0.43243243243243246,0.82421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7805686713592427,0.0 +0.6336669921875,0.7619047619047619,0.43243243243243246,0.83125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8028101365900171,0.0 +0.6357320149739584,0.7619047619047619,0.43243243243243246,0.83125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8026420158729655,0.0 +0.6382191975911459,0.7619047619047619,0.43243243243243246,0.834375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8035433434057269,0.0 +0.6411463419596354,0.7619047619047619,0.43243243243243246,0.8359375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8058111478459102,0.0 +0.6460622151692709,0.7619047619047619,0.43243243243243246,0.8421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.811012149567936,0.0 +0.6516876220703125,0.7619047619047619,0.43243243243243246,0.8500000000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8256077267462337,0.0 +0.65875244140625,0.7619047619047619,0.43243243243243246,0.85,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8280590039592608,0.0 +0.6633809407552084,0.7619047619047619,0.43243243243243246,0.85546875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.831139751336883,0.0 +0.6670939127604166,0.7619047619047619,0.43243243243243246,0.8640625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8560478158386009,0.0 +0.669830322265625,0.7619047619047619,0.43243243243243246,0.8648437499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8569320542934715,0.0 +0.6723505655924479,0.7619047619047619,0.43243243243243246,0.8656250000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8630960985479341,0.0 +0.6755549112955729,0.7619047619047619,0.43243243243243246,0.8656250000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8684505739663735,0.0 +0.6805419921875,0.7619047619047619,0.43243243243243246,0.86171875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8658956821670918,0.0 +0.6794102986653646,0.7619047619047619,0.43243243243243246,0.8671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8699278306774632,0.0 +0.6801274617513021,0.7619047619047619,0.43243243243243246,0.8656250000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8699408029061083,0.0 +0.6799952189127604,0.7619047619047619,0.43243243243243246,0.87109375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8745037643019817,0.0 +0.6791865030924479,0.7619047619047619,0.43243243243243246,0.87421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8781053232600908,0.0 +0.6781794230143229,0.7619047619047619,0.43243243243243246,0.87421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8782399617298586,0.0 +0.6786982218424479,0.7619047619047619,0.43243243243243246,0.87421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8780726895925887,0.0 +0.6808090209960938,0.7619047619047619,0.43243243243243246,0.8734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8775663399448319,0.0 +0.68316650390625,0.7619047619047619,0.43243243243243246,0.8726562499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8762671183331103,0.0 +0.6862208048502604,0.7619047619047619,0.43243243243243246,0.86875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8719493535075603,0.0 +0.6876093546549479,0.7619047619047619,0.43243243243243246,0.8671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8715862249060244,0.0 +0.6847763061523438,0.7619047619047619,0.43243243243243246,0.86484375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8723956691079908,0.0 +0.681915283203125,0.7619047619047619,0.43243243243243246,0.865625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8679639979782224,0.0 +0.6786982218424479,0.7619047619047619,0.43243243243243246,0.8648437499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8670696776401166,0.0 +0.6762619018554688,0.7619047619047619,0.43243243243243246,0.86328125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.866547673912225,0.0 +0.6721598307291666,0.7619047619047619,0.43243243243243246,0.8625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.866393467018419,0.0 +0.6679662068684896,0.7619047619047619,0.43243243243243246,0.8609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8646673465153589,0.0 +0.6648025512695312,0.7619047619047619,0.43243243243243246,0.85859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8618825485683683,0.0 +0.6610209147135416,0.7619047619047619,0.43243243243243246,0.8570312499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8613442135551292,0.0 +0.6582107543945312,0.7619047619047619,0.43243243243243246,0.85703125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8618749111561393,0.0 +0.6567026774088541,0.7619047619047619,0.43243243243243246,0.85625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8618192070563175,0.0 +0.65576171875,0.7619047619047619,0.43243243243243246,0.8531249999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8561804607514529,0.0 +0.6542078653971354,0.7619047619047619,0.43243243243243246,0.8546875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8562455649181195,0.0 +0.6512120564778646,0.7619047619047619,0.43243243243243246,0.85390625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8562108990631412,0.0 +0.648681640625,0.7619047619047619,0.43243243243243246,0.853125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8562666031629629,0.0 +0.6461156209309896,0.7619047619047619,0.43243243243243246,0.85078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8473935757033207,0.0 +0.6440480550130209,0.7619047619047619,0.43243243243243246,0.8499999999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8413565677131878,0.0 +0.6425145467122396,0.7619047619047619,0.43243243243243246,0.846875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8341489206453434,0.0 +0.6408055623372396,0.7619047619047619,0.43243243243243246,0.8468749999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8341489206453434,0.0 +0.6387939453125,0.7619047619047619,0.43243243243243246,0.846875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8342793872222712,0.0 +0.6368230183919271,0.7619047619047619,0.43243243243243246,0.84453125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8217793872222713,0.0 +0.6356124877929688,0.7619047619047619,0.43243243243243246,0.84375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8217575034687699,0.0 +0.6343485514322916,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8213635959057446,0.0 +0.6331812540690104,0.7619047619047619,0.43243243243243246,0.8421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8170717289185652,0.0 +0.6324615478515625,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8169871139533145,0.0 +0.6317698160807291,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8170328010878174,0.0 +0.6313095092773438,0.7619047619047619,0.43243243243243246,0.840625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8170471096958761,0.0 +0.631103515625,0.7619047619047619,0.43243243243243246,0.840625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8169532783036701,0.0 +0.6308085123697916,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.817047109695876,0.0 +0.6305618286132812,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8171430921447997,0.0 +0.6303456624348959,0.7619047619047619,0.43243243243243246,0.840625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8171507514585252,0.0 +0.6302820841471354,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8171507514585252,0.0 +0.6301396687825521,0.7619047619047619,0.43243243243243246,0.8421875000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8169322199899938,0.0 +0.6301040649414062,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8228899758993751,0.0 +0.6300532023111979,0.7619047619047619,0.43243243243243246,0.8421875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8169651064979432,0.0 +0.6299972534179688,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8231623236326604,0.0 +0.6299362182617188,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8232536979016662,0.0 +0.6299692789713541,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8232871570089932,0.0 +0.6299616495768229,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8232871570089932,0.0 +0.6299540201822916,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8232871570089932,0.0 +0.6299387613932291,0.7619047619047619,0.43243243243243246,0.84296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8232871570089932,0.0 diff --git a/results/downsample/papila/005/retfound/pr.png b/results/downsample/papila/005/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..e2c08b21d12c2185cd65cb4678cef7668a319c54 --- /dev/null +++ b/results/downsample/papila/005/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d18720bb4a49ea7614690503aaa31dcc71076db8a26ab6d49217efddaa8cffcd +size 53680 diff --git a/results/downsample/papila/005/retfound/roc.png b/results/downsample/papila/005/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..adbd7f2c072d4f7ef2a8a50a59fec70e97cb82ec --- /dev/null +++ b/results/downsample/papila/005/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92e20d0443de29866e177da8e83dcd81671bedd6a5d9e5980eb7fa8e89b1a3e0 +size 59674 diff --git a/results/downsample/papila/005/retfound/test_pred.npz b/results/downsample/papila/005/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..2cbb193d3a9c43b4c17d91ef68110bc69b28a442 --- /dev/null +++ b/results/downsample/papila/005/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:098296d515c8dcb27a983975faac1e0e3759ccd8ab8e2ec1d4afe38d3035b9ca +size 1518 diff --git a/results/downsample/papila/005/retfound/train.log b/results/downsample/papila/005/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0757f5cda4979c19e3798dc683caac27a0f1e48a --- /dev/null +++ b/results/downsample/papila/005/retfound/train.log @@ -0,0 +1,1043 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:57:58.094874146 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:57:58.587751] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:57:58.588075] Namespace(batch_size=8, +epochs=80, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/papila_5', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/005', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:58:01.617916] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:58:03.226524] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:58:03.388342] Sampler_train = +[14:58:03.433893] len of train_set: 8 +[14:58:03.608592] [Adaptation] Full fine-tuning: training all parameters. +[14:58:03.609634] number of trainable params (M): 303.30 +[14:58:03.609701] base lr: 5.00e-03 +[14:58:03.609747] actual lr: 1.56e-04 +[14:58:03.609799] accumulate grad iterations: 1 +[14:58:03.609849] effective batch size: 8 +[14:58:03.612806] criterion = CrossEntropyLoss() +[14:58:03.612878] Start training for 80 epochs +[14:58:03.615104] log_dir: ./output_logs/retfound +[14:58:04.987642] Epoch: [0] [0/1] eta: 0:00:01 lr: 0.000000 loss: 0.6929 (0.6929) time: 1.3716 data: 0.7039 max mem: 3223 +[14:58:05.046354] Epoch: [0] Total time: 0:00:01 (1.4311 s / it) +[14:58:05.047500] Averaged stats: lr: 0.000000 loss: 0.6929 (0.6929) +[14:58:06.034932] val: [0/6] eta: 0:00:05 loss: 0.6924 (0.6924) time: 0.9651 data: 0.9052 max mem: 3223 +[14:58:08.068867] val: [5/6] eta: 0:00:00 loss: 0.6924 (0.6929) time: 0.4997 data: 0.1510 max mem: 3223 +[14:58:08.137938] val: Total time: 0:00:03 (0.5114 s / it) +[14:58:08.151327] val loss: 0.6928914388020834 +[14:58:08.151565] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:58:09.678411] Best epoch = 0, Best score = 0.2994 +[14:58:09.764148] log_dir: ./output_logs/retfound +[14:58:10.665019] Epoch: [1] [0/1] eta: 0:00:00 lr: 0.000016 loss: 0.6924 (0.6924) time: 0.9000 data: 0.8484 max mem: 3225 +[14:58:10.733499] Epoch: [1] Total time: 0:00:00 (0.9692 s / it) +[14:58:10.734408] Averaged stats: lr: 0.000016 loss: 0.6924 (0.6924) +[14:58:11.819127] val: [0/6] eta: 0:00:06 loss: 0.6924 (0.6924) time: 1.0618 data: 1.0467 max mem: 3225 +[14:58:11.865424] val: [5/6] eta: 0:00:00 loss: 0.6924 (0.6929) time: 0.1845 data: 0.1746 max mem: 3225 +[14:58:11.941517] val: Total time: 0:00:01 (0.1975 s / it) +[14:58:11.958116] val loss: 0.6928914388020834 +[14:58:11.958323] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:58:12.003558] Best epoch = 0, Best score = 0.2994 +[14:58:12.274289] log_dir: ./output_logs/retfound +[14:58:13.117361] Epoch: [2] [0/1] eta: 0:00:00 lr: 0.000031 loss: 0.6928 (0.6928) time: 0.8421 data: 0.8048 max mem: 3230 +[14:58:13.184306] Epoch: [2] Total time: 0:00:00 (0.9098 s / it) +[14:58:13.185128] Averaged stats: lr: 0.000031 loss: 0.6928 (0.6928) +[14:58:14.145538] val: [0/6] eta: 0:00:05 loss: 0.6924 (0.6924) time: 0.9416 data: 0.9273 max mem: 3230 +[14:58:14.185242] val: [5/6] eta: 0:00:00 loss: 0.6924 (0.6929) time: 0.1635 data: 0.1547 max mem: 3230 +[14:58:14.256714] val: Total time: 0:00:01 (0.1755 s / it) +[14:58:14.266435] val loss: 0.6928914388020834 +[14:58:14.266691] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:58:14.315396] Best epoch = 0, Best score = 0.2994 +[14:58:14.563335] log_dir: ./output_logs/retfound +[14:58:15.450042] Epoch: [3] [0/1] eta: 0:00:00 lr: 0.000047 loss: 0.6929 (0.6929) time: 0.8859 data: 0.7504 max mem: 4749 +[14:58:15.518756] Epoch: [3] Total time: 0:00:00 (0.9553 s / it) +[14:58:15.519594] Averaged stats: lr: 0.000047 loss: 0.6929 (0.6929) +[14:58:16.426864] val: [0/6] eta: 0:00:05 loss: 0.6771 (0.6771) time: 0.8847 data: 0.8711 max mem: 4749 +[14:58:16.468900] val: [5/6] eta: 0:00:00 loss: 0.6771 (0.6877) time: 0.1544 data: 0.1453 max mem: 4749 +[14:58:16.537245] val: Total time: 0:00:00 (0.1659 s / it) +[14:58:16.546878] val loss: 0.6876627604166666 +[14:58:16.547115] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.5977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.5817, Kappa: 0.0000, Score: 0.3434 +[14:58:18.152674] Best epoch = 3, Best score = 0.3434 +[14:58:18.239945] log_dir: ./output_logs/retfound +[14:58:19.071129] Epoch: [4] [0/1] eta: 0:00:00 lr: 0.000063 loss: 0.6841 (0.6841) time: 0.8302 data: 0.7682 max mem: 5539 +[14:58:19.148440] Epoch: [4] Total time: 0:00:00 (0.9084 s / it) +[14:58:19.149486] Averaged stats: lr: 0.000063 loss: 0.6841 (0.6841) +[14:58:20.201841] val: [0/6] eta: 0:00:06 loss: 0.6594 (0.6594) time: 1.0449 data: 1.0298 max mem: 5539 +[14:58:20.249898] val: [5/6] eta: 0:00:00 loss: 0.6599 (0.6818) time: 0.1820 data: 0.1718 max mem: 5539 +[14:58:20.322841] val: Total time: 0:00:01 (0.1944 s / it) +[14:58:20.331756] val loss: 0.6818339029947916 +[14:58:20.332005] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6687, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6415, Kappa: 0.0000, Score: 0.3671 +[14:58:22.148027] Best epoch = 4, Best score = 0.3671 +[14:58:22.211175] log_dir: ./output_logs/retfound +[14:58:23.161958] Epoch: [5] [0/1] eta: 0:00:00 lr: 0.000078 loss: 0.6692 (0.6692) time: 0.9499 data: 0.8988 max mem: 5539 +[14:58:23.234597] Epoch: [5] Total time: 0:00:01 (1.0232 s / it) +[14:58:23.235484] Averaged stats: lr: 0.000078 loss: 0.6692 (0.6692) +[14:58:24.156338] val: [0/6] eta: 0:00:05 loss: 0.6385 (0.6385) time: 0.8984 data: 0.8830 max mem: 5539 +[14:58:24.198661] val: [5/6] eta: 0:00:00 loss: 0.6392 (0.6752) time: 0.1567 data: 0.1473 max mem: 5539 +[14:58:24.270550] val: Total time: 0:00:01 (0.1689 s / it) +[14:58:24.279554] val loss: 0.6751810709635416 +[14:58:24.279758] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6633, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6365, Kappa: 0.0000, Score: 0.3652 +[14:58:24.324856] Best epoch = 4, Best score = 0.3671 +[14:58:24.600542] log_dir: ./output_logs/retfound +[14:58:25.439241] Epoch: [6] [0/1] eta: 0:00:00 lr: 0.000094 loss: 0.6614 (0.6614) time: 0.8379 data: 0.7866 max mem: 5539 +[14:58:25.505103] Epoch: [6] Total time: 0:00:00 (0.9044 s / it) +[14:58:25.505985] Averaged stats: lr: 0.000094 loss: 0.6614 (0.6614) +[14:58:26.456080] val: [0/6] eta: 0:00:05 loss: 0.6132 (0.6132) time: 0.9190 data: 0.8962 max mem: 5539 +[14:58:26.517044] val: [5/6] eta: 0:00:00 loss: 0.6143 (0.6676) time: 0.1632 data: 0.1495 max mem: 5539 +[14:58:26.586760] val: Total time: 0:00:01 (0.1751 s / it) +[14:58:26.595636] val loss: 0.6675618489583334 +[14:58:26.595819] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6758, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6521, Kappa: 0.0000, Score: 0.3694 +[14:58:28.247646] Best epoch = 6, Best score = 0.3694 +[14:58:28.318916] log_dir: ./output_logs/retfound +[14:58:29.197802] Epoch: [7] [0/1] eta: 0:00:00 lr: 0.000109 loss: 0.6391 (0.6391) time: 0.8779 data: 0.8173 max mem: 5539 +[14:58:29.277643] Epoch: [7] Total time: 0:00:00 (0.9586 s / it) +[14:58:29.278534] Averaged stats: lr: 0.000109 loss: 0.6391 (0.6391) +[14:58:30.365685] val: [0/6] eta: 0:00:06 loss: 0.5857 (0.5857) time: 1.0799 data: 1.0655 max mem: 5539 +[14:58:30.410370] val: [5/6] eta: 0:00:00 loss: 0.5873 (0.6597) time: 0.1873 data: 0.1777 max mem: 5539 +[14:58:30.481461] val: Total time: 0:00:01 (0.1994 s / it) +[14:58:30.491105] val loss: 0.6597493489583334 +[14:58:30.491310] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6820, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6614, Kappa: 0.0000, Score: 0.3715 +[14:58:32.064389] Best epoch = 7, Best score = 0.3715 +[14:58:32.129404] log_dir: ./output_logs/retfound +[14:58:32.950941] Epoch: [8] [0/1] eta: 0:00:00 lr: 0.000125 loss: 0.6339 (0.6339) time: 0.8206 data: 0.7621 max mem: 5539 +[14:58:33.017240] Epoch: [8] Total time: 0:00:00 (0.8877 s / it) +[14:58:33.018136] Averaged stats: lr: 0.000125 loss: 0.6339 (0.6339) +[14:58:34.079347] val: [0/6] eta: 0:00:06 loss: 0.5538 (0.5538) time: 1.0423 data: 1.0287 max mem: 5539 +[14:58:34.121139] val: [5/6] eta: 0:00:00 loss: 0.5558 (0.6515) time: 0.1806 data: 0.1715 max mem: 5539 +[14:58:34.193371] val: Total time: 0:00:01 (0.1928 s / it) +[14:58:34.202425] val loss: 0.651519775390625 +[14:58:34.202654] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7063, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6816, Kappa: 0.0000, Score: 0.3796 +[14:58:35.903099] Best epoch = 8, Best score = 0.3796 +[14:58:35.965041] log_dir: ./output_logs/retfound +[14:58:36.785574] Epoch: [9] [0/1] eta: 0:00:00 lr: 0.000141 loss: 0.5629 (0.5629) time: 0.8197 data: 0.7695 max mem: 5539 +[14:58:36.854020] Epoch: [9] Total time: 0:00:00 (0.8888 s / it) +[14:58:36.854967] Averaged stats: lr: 0.000141 loss: 0.5629 (0.5629) +[14:58:37.782952] val: [0/6] eta: 0:00:05 loss: 0.5173 (0.5173) time: 0.9086 data: 0.8932 max mem: 5539 +[14:58:37.823607] val: [5/6] eta: 0:00:00 loss: 0.5192 (0.6431) time: 0.1581 data: 0.1490 max mem: 5539 +[14:58:37.898744] val: Total time: 0:00:01 (0.1708 s / it) +[14:58:37.907838] val loss: 0.643096923828125 +[14:58:37.908008] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7289, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6926, Kappa: 0.0000, Score: 0.3871 +[14:58:39.479572] Best epoch = 9, Best score = 0.3871 +[14:58:39.542095] log_dir: ./output_logs/retfound +[14:58:40.402197] Epoch: [10] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5508 (0.5508) time: 0.8591 data: 0.8054 max mem: 5539 +[14:58:40.469391] Epoch: [10] Total time: 0:00:00 (0.9271 s / it) +[14:58:40.470200] Averaged stats: lr: 0.000156 loss: 0.5508 (0.5508) +[14:58:41.489025] val: [0/6] eta: 0:00:06 loss: 0.4760 (0.4760) time: 1.0117 data: 0.9988 max mem: 5539 +[14:58:41.529270] val: [5/6] eta: 0:00:00 loss: 0.4781 (0.6352) time: 0.1752 data: 0.1666 max mem: 5539 +[14:58:41.599812] val: Total time: 0:00:01 (0.1872 s / it) +[14:58:41.608669] val loss: 0.635162353515625 +[14:58:41.608863] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7453, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6999, Kappa: 0.0000, Score: 0.3926 +[14:58:43.204249] Best epoch = 10, Best score = 0.3926 +[14:58:43.299672] log_dir: ./output_logs/retfound +[14:58:44.129512] Epoch: [11] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.6276 (0.6276) time: 0.8290 data: 0.7795 max mem: 5539 +[14:58:44.205535] Epoch: [11] Total time: 0:00:00 (0.9057 s / it) +[14:58:44.206453] Averaged stats: lr: 0.000156 loss: 0.6276 (0.6276) +[14:58:45.193353] val: [0/6] eta: 0:00:05 loss: 0.4362 (0.4362) time: 0.9796 data: 0.9653 max mem: 5539 +[14:58:45.286474] val: [5/6] eta: 0:00:00 loss: 0.4381 (0.6293) time: 0.1787 data: 0.1696 max mem: 5539 +[14:58:45.355094] val: Total time: 0:00:01 (0.1903 s / it) +[14:58:45.364711] val loss: 0.629302978515625 +[14:58:45.364898] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7734, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7191, Kappa: 0.0000, Score: 0.4020 +[14:58:46.926357] Best epoch = 11, Best score = 0.4020 +[14:58:47.013280] log_dir: ./output_logs/retfound +[14:58:47.900925] Epoch: [12] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5994 (0.5994) time: 0.8868 data: 0.8387 max mem: 5539 +[14:58:47.965595] Epoch: [12] Total time: 0:00:00 (0.9522 s / it) +[14:58:47.966395] Averaged stats: lr: 0.000156 loss: 0.5994 (0.5994) +[14:58:48.926711] val: [0/6] eta: 0:00:05 loss: 0.4032 (0.4032) time: 0.9380 data: 0.9255 max mem: 5539 +[14:58:48.965836] val: [5/6] eta: 0:00:00 loss: 0.4049 (0.6264) time: 0.1628 data: 0.1543 max mem: 5539 +[14:58:49.037869] val: Total time: 0:00:01 (0.1750 s / it) +[14:58:49.046606] val loss: 0.6263631184895834 +[14:58:49.046774] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7859, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7262, Kappa: 0.0000, Score: 0.4061 +[14:58:50.574890] Best epoch = 12, Best score = 0.4061 +[14:58:50.648892] log_dir: ./output_logs/retfound +[14:58:51.500306] Epoch: [13] [0/1] eta: 0:00:00 lr: 0.000156 loss: 0.5746 (0.5746) time: 0.8506 data: 0.8028 max mem: 5539 +[14:58:51.566537] Epoch: [13] Total time: 0:00:00 (0.9175 s / it) +[14:58:51.567619] Averaged stats: lr: 0.000156 loss: 0.5746 (0.5746) +[14:58:52.639595] val: [0/6] eta: 0:00:06 loss: 0.3736 (0.3736) time: 1.0623 data: 1.0482 max mem: 5539 +[14:58:52.679188] val: [5/6] eta: 0:00:00 loss: 0.3750 (0.6251) time: 0.1836 data: 0.1748 max mem: 5539 +[14:58:52.762238] val: Total time: 0:00:01 (0.1976 s / it) +[14:58:52.770904] val loss: 0.6251169840494791 +[14:58:52.771064] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7464, Kappa: 0.0000, Score: 0.4100 +[14:58:54.280258] Best epoch = 13, Best score = 0.4100 +[14:58:54.353817] log_dir: ./output_logs/retfound +[14:58:55.269831] Epoch: [14] [0/1] eta: 0:00:00 lr: 0.000155 loss: 0.4389 (0.4389) time: 0.9152 data: 0.8658 max mem: 5539 +[14:58:55.336754] Epoch: [14] Total time: 0:00:00 (0.9828 s / it) +[14:58:55.337636] Averaged stats: lr: 0.000155 loss: 0.4389 (0.4389) +[14:58:56.265090] val: [0/6] eta: 0:00:05 loss: 0.3447 (0.3447) time: 0.9158 data: 0.9016 max mem: 5539 +[14:58:56.311591] val: [5/6] eta: 0:00:00 loss: 0.3462 (0.6256) time: 0.1603 data: 0.1514 max mem: 5539 +[14:58:56.384129] val: Total time: 0:00:01 (0.1726 s / it) +[14:58:56.393149] val loss: 0.625579833984375 +[14:58:56.393323] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8047, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7450, Kappa: 0.0000, Score: 0.4124 +[14:58:57.900501] Best epoch = 14, Best score = 0.4124 +[14:58:57.985257] log_dir: ./output_logs/retfound +[14:58:58.806016] Epoch: [15] [0/1] eta: 0:00:00 lr: 0.000154 loss: 0.4347 (0.4347) time: 0.8199 data: 0.7706 max mem: 5539 +[14:58:58.875020] Epoch: [15] Total time: 0:00:00 (0.8896 s / it) +[14:58:58.875858] Averaged stats: lr: 0.000154 loss: 0.4347 (0.4347) +[14:58:59.785524] val: [0/6] eta: 0:00:05 loss: 0.3163 (0.3163) time: 0.9027 data: 0.8895 max mem: 5539 +[14:58:59.826068] val: [5/6] eta: 0:00:00 loss: 0.3180 (0.6278) time: 0.1571 data: 0.1483 max mem: 5539 +[14:58:59.893494] val: Total time: 0:00:01 (0.1686 s / it) +[14:58:59.902243] val loss: 0.6277618408203125 +[14:58:59.902411] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8109, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7504, Kappa: 0.0000, Score: 0.4145 +[14:59:01.450179] Best epoch = 15, Best score = 0.4145 +[14:59:01.546590] log_dir: ./output_logs/retfound +[14:59:02.245583] Epoch: [16] [0/1] eta: 0:00:00 lr: 0.000153 loss: 0.5441 (0.5441) time: 0.6983 data: 0.6502 max mem: 5539 +[14:59:02.315548] Epoch: [16] Total time: 0:00:00 (0.7688 s / it) +[14:59:02.316333] Averaged stats: lr: 0.000153 loss: 0.5441 (0.5441) +[14:59:03.216634] val: [0/6] eta: 0:00:05 loss: 0.2938 (0.2938) time: 0.8928 data: 0.8776 max mem: 5539 +[14:59:03.256531] val: [5/6] eta: 0:00:00 loss: 0.2955 (0.6308) time: 0.1553 data: 0.1464 max mem: 5539 +[14:59:03.329466] val: Total time: 0:00:01 (0.1677 s / it) +[14:59:03.338184] val loss: 0.6307932535807291 +[14:59:03.338360] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8242, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7806, Kappa: 0.0000, Score: 0.4189 +[14:59:04.856764] Best epoch = 16, Best score = 0.4189 +[14:59:04.917772] log_dir: ./output_logs/retfound +[14:59:05.703714] Epoch: [17] [0/1] eta: 0:00:00 lr: 0.000152 loss: 0.7150 (0.7150) time: 0.7852 data: 0.7374 max mem: 5539 +[14:59:05.775674] Epoch: [17] Total time: 0:00:00 (0.8578 s / it) +[14:59:05.776451] Averaged stats: lr: 0.000152 loss: 0.7150 (0.7150) +[14:59:06.677038] val: [0/6] eta: 0:00:05 loss: 0.2807 (0.2807) time: 0.8824 data: 0.8690 max mem: 5539 +[14:59:06.717689] val: [5/6] eta: 0:00:00 loss: 0.2823 (0.6337) time: 0.1538 data: 0.1449 max mem: 5539 +[14:59:06.790582] val: Total time: 0:00:00 (0.1661 s / it) +[14:59:06.799707] val loss: 0.6336669921875 +[14:59:06.799904] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8313, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8028, Kappa: 0.0000, Score: 0.4212 +[14:59:08.341764] Best epoch = 17, Best score = 0.4212 +[14:59:08.401178] log_dir: ./output_logs/retfound +[14:59:09.172987] Epoch: [18] [0/1] eta: 0:00:00 lr: 0.000151 loss: 0.7147 (0.7147) time: 0.7710 data: 0.7207 max mem: 5539 +[14:59:09.238632] Epoch: [18] Total time: 0:00:00 (0.8373 s / it) +[14:59:09.239420] Averaged stats: lr: 0.000151 loss: 0.7147 (0.7147) +[14:59:10.167562] val: [0/6] eta: 0:00:05 loss: 0.2725 (0.2725) time: 0.9166 data: 0.9039 max mem: 5539 +[14:59:10.208297] val: [5/6] eta: 0:00:00 loss: 0.2737 (0.6357) time: 0.1595 data: 0.1508 max mem: 5539 +[14:59:10.280606] val: Total time: 0:00:01 (0.1717 s / it) +[14:59:10.289742] val loss: 0.6357320149739584 +[14:59:10.289946] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8313, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8026, Kappa: 0.0000, Score: 0.4212 +[14:59:10.336915] Best epoch = 17, Best score = 0.4212 +[14:59:10.595186] log_dir: ./output_logs/retfound +[14:59:11.485409] Epoch: [19] [0/1] eta: 0:00:00 lr: 0.000150 loss: 0.5787 (0.5787) time: 0.8895 data: 0.8416 max mem: 5539 +[14:59:11.552880] Epoch: [19] Total time: 0:00:00 (0.9576 s / it) +[14:59:11.553662] Averaged stats: lr: 0.000150 loss: 0.5787 (0.5787) +[14:59:12.410726] val: [0/6] eta: 0:00:05 loss: 0.2642 (0.2642) time: 0.8461 data: 0.8324 max mem: 5539 +[14:59:12.473221] val: [5/6] eta: 0:00:00 loss: 0.2649 (0.6382) time: 0.1513 data: 0.1424 max mem: 5539 +[14:59:12.543747] val: Total time: 0:00:00 (0.1633 s / it) +[14:59:12.552430] val loss: 0.6382191975911459 +[14:59:12.552590] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8344, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8035, Kappa: 0.0000, Score: 0.4223 +[14:59:14.050513] Best epoch = 19, Best score = 0.4223 +[14:59:14.112883] log_dir: ./output_logs/retfound +[14:59:14.878877] Epoch: [20] [0/1] eta: 0:00:00 lr: 0.000149 loss: 0.3714 (0.3714) time: 0.7652 data: 0.7174 max mem: 5539 +[14:59:14.946568] Epoch: [20] Total time: 0:00:00 (0.8336 s / it) +[14:59:14.947431] Averaged stats: lr: 0.000149 loss: 0.3714 (0.3714) +[14:59:15.866446] val: [0/6] eta: 0:00:05 loss: 0.2548 (0.2548) time: 0.8951 data: 0.8819 max mem: 5539 +[14:59:15.906099] val: [5/6] eta: 0:00:00 loss: 0.2552 (0.6411) time: 0.1557 data: 0.1471 max mem: 5539 +[14:59:15.978469] val: Total time: 0:00:01 (0.1679 s / it) +[14:59:15.987247] val loss: 0.6411463419596354 +[14:59:15.987432] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8359, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8058, Kappa: 0.0000, Score: 0.4228 +[14:59:17.508370] Best epoch = 20, Best score = 0.4228 +[14:59:17.577737] log_dir: ./output_logs/retfound +[14:59:18.337539] Epoch: [21] [0/1] eta: 0:00:00 lr: 0.000147 loss: 0.2417 (0.2417) time: 0.7590 data: 0.7112 max mem: 5539 +[14:59:18.405551] Epoch: [21] Total time: 0:00:00 (0.8277 s / it) +[14:59:18.406250] Averaged stats: lr: 0.000147 loss: 0.2417 (0.2417) +[14:59:19.308928] val: [0/6] eta: 0:00:05 loss: 0.2422 (0.2422) time: 0.8838 data: 0.8706 max mem: 5539 +[14:59:19.348744] val: [5/6] eta: 0:00:00 loss: 0.2422 (0.6461) time: 0.1538 data: 0.1452 max mem: 5539 +[14:59:19.419835] val: Total time: 0:00:00 (0.1659 s / it) +[14:59:19.428579] val loss: 0.6460622151692709 +[14:59:19.428742] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8422, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8110, Kappa: 0.0000, Score: 0.4249 +[14:59:20.935088] Best epoch = 21, Best score = 0.4249 +[14:59:21.021714] log_dir: ./output_logs/retfound +[14:59:21.838136] Epoch: [22] [0/1] eta: 0:00:00 lr: 0.000145 loss: 0.4101 (0.4101) time: 0.8156 data: 0.7676 max mem: 5539 +[14:59:21.909806] Epoch: [22] Total time: 0:00:00 (0.8879 s / it) +[14:59:21.910650] Averaged stats: lr: 0.000145 loss: 0.4101 (0.4101) +[14:59:22.803616] val: [0/6] eta: 0:00:05 loss: 0.2300 (0.2300) time: 0.8862 data: 0.8737 max mem: 5539 +[14:59:22.866510] val: [5/6] eta: 0:00:00 loss: 0.2300 (0.6517) time: 0.1581 data: 0.1495 max mem: 5539 +[14:59:22.937295] val: Total time: 0:00:01 (0.1701 s / it) +[14:59:22.946034] val loss: 0.6516876220703125 +[14:59:22.946202] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8256, Kappa: 0.0000, Score: 0.4275 +[14:59:24.441952] Best epoch = 22, Best score = 0.4275 +[14:59:24.544992] log_dir: ./output_logs/retfound +[14:59:25.348939] Epoch: [23] [0/1] eta: 0:00:00 lr: 0.000143 loss: 0.4020 (0.4020) time: 0.8032 data: 0.7554 max mem: 5539 +[14:59:25.413581] Epoch: [23] Total time: 0:00:00 (0.8685 s / it) +[14:59:25.414347] Averaged stats: lr: 0.000143 loss: 0.4020 (0.4020) +[14:59:26.300544] val: [0/6] eta: 0:00:05 loss: 0.2169 (0.2169) time: 0.8794 data: 0.8636 max mem: 5539 +[14:59:26.341607] val: [5/6] eta: 0:00:00 loss: 0.2169 (0.6588) time: 0.1533 data: 0.1440 max mem: 5539 +[14:59:26.410875] val: Total time: 0:00:00 (0.1650 s / it) +[14:59:26.419668] val loss: 0.65875244140625 +[14:59:26.419855] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8281, Kappa: 0.0000, Score: 0.4275 +[14:59:26.454153] Best epoch = 22, Best score = 0.4275 +[14:59:26.715424] log_dir: ./output_logs/retfound +[14:59:27.497403] Epoch: [24] [0/1] eta: 0:00:00 lr: 0.000141 loss: 0.7991 (0.7991) time: 0.7812 data: 0.7332 max mem: 5539 +[14:59:27.565882] Epoch: [24] Total time: 0:00:00 (0.8503 s / it) +[14:59:27.566695] Averaged stats: lr: 0.000141 loss: 0.7991 (0.7991) +[14:59:28.458463] val: [0/6] eta: 0:00:05 loss: 0.2093 (0.2093) time: 0.8664 data: 0.8530 max mem: 5539 +[14:59:28.514226] val: [5/6] eta: 0:00:00 loss: 0.2093 (0.6634) time: 0.1536 data: 0.1448 max mem: 5539 +[14:59:28.583022] val: Total time: 0:00:00 (0.1653 s / it) +[14:59:28.591747] val loss: 0.6633809407552084 +[14:59:28.591940] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8555, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8311, Kappa: 0.0000, Score: 0.4293 +[14:59:30.081962] Best epoch = 24, Best score = 0.4293 +[14:59:30.149011] log_dir: ./output_logs/retfound +[14:59:31.036660] Epoch: [25] [0/1] eta: 0:00:00 lr: 0.000139 loss: 0.6187 (0.6187) time: 0.8868 data: 0.8381 max mem: 5539 +[14:59:31.104546] Epoch: [25] Total time: 0:00:00 (0.9554 s / it) +[14:59:31.105473] Averaged stats: lr: 0.000139 loss: 0.6187 (0.6187) +[14:59:32.000929] val: [0/6] eta: 0:00:05 loss: 0.2035 (0.2035) time: 0.8885 data: 0.8749 max mem: 5539 +[14:59:32.052381] val: [5/6] eta: 0:00:00 loss: 0.2035 (0.6671) time: 0.1566 data: 0.1475 max mem: 5539 +[14:59:32.124696] val: Total time: 0:00:01 (0.1688 s / it) +[14:59:32.133774] val loss: 0.6670939127604166 +[14:59:32.133966] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8641, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8560, Kappa: 0.0000, Score: 0.4322 +[14:59:33.645981] Best epoch = 25, Best score = 0.4322 +[14:59:33.731425] log_dir: ./output_logs/retfound +[14:59:34.506194] Epoch: [26] [0/1] eta: 0:00:00 lr: 0.000137 loss: 0.6037 (0.6037) time: 0.7740 data: 0.7258 max mem: 5539 +[14:59:34.574057] Epoch: [26] Total time: 0:00:00 (0.8425 s / it) +[14:59:34.574884] Averaged stats: lr: 0.000137 loss: 0.6037 (0.6037) +[14:59:35.486028] val: [0/6] eta: 0:00:05 loss: 0.1994 (0.1994) time: 0.8964 data: 0.8824 max mem: 5539 +[14:59:35.542324] val: [5/6] eta: 0:00:00 loss: 0.1994 (0.6698) time: 0.1587 data: 0.1500 max mem: 5539 +[14:59:35.617134] val: Total time: 0:00:01 (0.1713 s / it) +[14:59:35.625763] val loss: 0.669830322265625 +[14:59:35.625925] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8648, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8569, Kappa: 0.0000, Score: 0.4324 +[14:59:37.116606] Best epoch = 26, Best score = 0.4324 +[14:59:37.181030] log_dir: ./output_logs/retfound +[14:59:38.034498] Epoch: [27] [0/1] eta: 0:00:00 lr: 0.000135 loss: 0.4208 (0.4208) time: 0.8526 data: 0.8038 max mem: 5539 +[14:59:38.107366] Epoch: [27] Total time: 0:00:00 (0.9262 s / it) +[14:59:38.108167] Averaged stats: lr: 0.000135 loss: 0.4208 (0.4208) +[14:59:39.045547] val: [0/6] eta: 0:00:05 loss: 0.1960 (0.1960) time: 0.9191 data: 0.9067 max mem: 5539 +[14:59:39.085048] val: [5/6] eta: 0:00:00 loss: 0.1960 (0.6724) time: 0.1597 data: 0.1512 max mem: 5539 +[14:59:39.153663] val: Total time: 0:00:01 (0.1713 s / it) +[14:59:39.162431] val loss: 0.6723505655924479 +[14:59:39.162603] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8631, Kappa: 0.0000, Score: 0.4327 +[14:59:40.655916] Best epoch = 27, Best score = 0.4327 +[14:59:40.721523] log_dir: ./output_logs/retfound +[14:59:41.539986] Epoch: [28] [0/1] eta: 0:00:00 lr: 0.000132 loss: 0.3930 (0.3930) time: 0.8177 data: 0.7704 max mem: 5539 +[14:59:41.606576] Epoch: [28] Total time: 0:00:00 (0.8849 s / it) +[14:59:41.607353] Averaged stats: lr: 0.000132 loss: 0.3930 (0.3930) +[14:59:42.525400] val: [0/6] eta: 0:00:05 loss: 0.1913 (0.1913) time: 0.8995 data: 0.8864 max mem: 5539 +[14:59:42.566259] val: [5/6] eta: 0:00:00 loss: 0.1913 (0.6756) time: 0.1566 data: 0.1478 max mem: 5539 +[14:59:42.639414] val: Total time: 0:00:01 (0.1690 s / it) +[14:59:42.648368] val loss: 0.6755549112955729 +[14:59:42.648547] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8685, Kappa: 0.0000, Score: 0.4327 +[14:59:42.696180] Best epoch = 27, Best score = 0.4327 +[14:59:42.952990] log_dir: ./output_logs/retfound +[14:59:43.764076] Epoch: [29] [0/1] eta: 0:00:00 lr: 0.000130 loss: 0.1789 (0.1789) time: 0.8103 data: 0.7621 max mem: 5539 +[14:59:43.837280] Epoch: [29] Total time: 0:00:00 (0.8841 s / it) +[14:59:43.838065] Averaged stats: lr: 0.000130 loss: 0.1789 (0.1789) +[14:59:44.819942] val: [0/6] eta: 0:00:05 loss: 0.1848 (0.1848) time: 0.9590 data: 0.9469 max mem: 5539 +[14:59:44.859262] val: [5/6] eta: 0:00:00 loss: 0.1848 (0.6805) time: 0.1663 data: 0.1579 max mem: 5539 +[14:59:44.932121] val: Total time: 0:00:01 (0.1786 s / it) +[14:59:44.940728] val loss: 0.6805419921875 +[14:59:44.940916] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8617, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8659, Kappa: 0.0000, Score: 0.4314 +[14:59:44.976591] Best epoch = 27, Best score = 0.4327 +[14:59:45.233853] log_dir: ./output_logs/retfound +[14:59:46.018921] Epoch: [30] [0/1] eta: 0:00:00 lr: 0.000127 loss: 0.7217 (0.7217) time: 0.7843 data: 0.7365 max mem: 5539 +[14:59:46.085835] Epoch: [30] Total time: 0:00:00 (0.8518 s / it) +[14:59:46.086823] Averaged stats: lr: 0.000127 loss: 0.7217 (0.7217) +[14:59:47.039644] val: [0/6] eta: 0:00:05 loss: 0.1850 (0.1850) time: 0.9298 data: 0.9168 max mem: 5539 +[14:59:47.078875] val: [5/6] eta: 0:00:00 loss: 0.1850 (0.6794) time: 0.1614 data: 0.1529 max mem: 5539 +[14:59:47.152493] val: Total time: 0:00:01 (0.1739 s / it) +[14:59:47.161459] val loss: 0.6794102986653646 +[14:59:47.161710] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8672, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8699, Kappa: 0.0000, Score: 0.4332 +[14:59:48.652496] Best epoch = 30, Best score = 0.4332 +[14:59:48.735071] log_dir: ./output_logs/retfound +[14:59:49.499982] Epoch: [31] [0/1] eta: 0:00:00 lr: 0.000124 loss: 0.1607 (0.1607) time: 0.7641 data: 0.7157 max mem: 5539 +[14:59:49.566638] Epoch: [31] Total time: 0:00:00 (0.8314 s / it) +[14:59:49.567496] Averaged stats: lr: 0.000124 loss: 0.1607 (0.1607) +[14:59:50.546378] val: [0/6] eta: 0:00:05 loss: 0.1829 (0.1829) time: 0.9719 data: 0.9576 max mem: 5539 +[14:59:50.594193] val: [5/6] eta: 0:00:00 loss: 0.1829 (0.6801) time: 0.1699 data: 0.1611 max mem: 5539 +[14:59:50.665762] val: Total time: 0:00:01 (0.1820 s / it) +[14:59:50.674513] val loss: 0.6801274617513021 +[14:59:50.674672] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8699, Kappa: 0.0000, Score: 0.4327 +[14:59:50.709045] Best epoch = 30, Best score = 0.4332 +[14:59:50.961054] log_dir: ./output_logs/retfound +[14:59:51.733221] Epoch: [32] [0/1] eta: 0:00:00 lr: 0.000121 loss: 0.6430 (0.6430) time: 0.7714 data: 0.7230 max mem: 5539 +[14:59:51.817260] Epoch: [32] Total time: 0:00:00 (0.8560 s / it) +[14:59:51.818098] Averaged stats: lr: 0.000121 loss: 0.6430 (0.6430) +[14:59:52.839895] val: [0/6] eta: 0:00:05 loss: 0.1820 (0.1820) time: 0.9995 data: 0.9864 max mem: 5539 +[14:59:52.918812] val: [5/6] eta: 0:00:00 loss: 0.1820 (0.6800) time: 0.1796 data: 0.1707 max mem: 5539 +[14:59:52.989277] val: Total time: 0:00:01 (0.1916 s / it) +[14:59:52.998083] val loss: 0.6799952189127604 +[14:59:52.998279] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8711, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8745, Kappa: 0.0000, Score: 0.4345 +[14:59:54.505965] Best epoch = 32, Best score = 0.4345 +[14:59:54.565549] log_dir: ./output_logs/retfound +[14:59:55.353132] Epoch: [33] [0/1] eta: 0:00:00 lr: 0.000118 loss: 0.5572 (0.5572) time: 0.7868 data: 0.7388 max mem: 5539 +[14:59:55.420587] Epoch: [33] Total time: 0:00:00 (0.8549 s / it) +[14:59:55.421451] Averaged stats: lr: 0.000118 loss: 0.5572 (0.5572) +[14:59:56.333351] val: [0/6] eta: 0:00:05 loss: 0.1819 (0.1819) time: 0.8930 data: 0.8805 max mem: 5539 +[14:59:56.372463] val: [5/6] eta: 0:00:00 loss: 0.1819 (0.6792) time: 0.1553 data: 0.1468 max mem: 5539 +[14:59:56.442314] val: Total time: 0:00:01 (0.1671 s / it) +[14:59:56.451301] val loss: 0.6791865030924479 +[14:59:56.451474] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8742, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8781, Kappa: 0.0000, Score: 0.4356 +[14:59:57.946436] Best epoch = 33, Best score = 0.4356 +[14:59:58.030193] log_dir: ./output_logs/retfound +[14:59:58.814065] Epoch: [34] [0/1] eta: 0:00:00 lr: 0.000115 loss: 0.6110 (0.6110) time: 0.7830 data: 0.7348 max mem: 5539 +[14:59:58.878244] Epoch: [34] Total time: 0:00:00 (0.8479 s / it) +[14:59:58.879047] Averaged stats: lr: 0.000115 loss: 0.6110 (0.6110) +[14:59:59.759992] val: [0/6] eta: 0:00:05 loss: 0.1822 (0.1822) time: 0.8738 data: 0.8610 max mem: 5539 +[14:59:59.812366] val: [5/6] eta: 0:00:00 loss: 0.1822 (0.6782) time: 0.1543 data: 0.1456 max mem: 5539 +[14:59:59.883121] val: Total time: 0:00:00 (0.1663 s / it) +[14:59:59.891830] val loss: 0.6781794230143229 +[14:59:59.892007] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8742, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8782, Kappa: 0.0000, Score: 0.4356 +[14:59:59.940176] Best epoch = 33, Best score = 0.4356 +[15:00:00.207837] log_dir: ./output_logs/retfound +[15:00:00.926736] Epoch: [35] [0/1] eta: 0:00:00 lr: 0.000112 loss: 0.3164 (0.3164) time: 0.7181 data: 0.6700 max mem: 5539 +[15:00:00.992504] Epoch: [35] Total time: 0:00:00 (0.7845 s / it) +[15:00:00.993291] Averaged stats: lr: 0.000112 loss: 0.3164 (0.3164) +[15:00:01.918607] val: [0/6] eta: 0:00:05 loss: 0.1809 (0.1809) time: 0.9025 data: 0.8882 max mem: 5539 +[15:00:01.960640] val: [5/6] eta: 0:00:00 loss: 0.1809 (0.6787) time: 0.1573 data: 0.1481 max mem: 5539 +[15:00:02.032111] val: Total time: 0:00:01 (0.1694 s / it) +[15:00:02.040926] val loss: 0.6786982218424479 +[15:00:02.041115] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8742, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8781, Kappa: 0.0000, Score: 0.4356 +[15:00:02.090589] Best epoch = 33, Best score = 0.4356 +[15:00:02.363981] log_dir: ./output_logs/retfound +[15:00:03.330541] Epoch: [36] [0/1] eta: 0:00:00 lr: 0.000109 loss: 0.1724 (0.1724) time: 0.9656 data: 0.9131 max mem: 5539 +[15:00:03.398180] Epoch: [36] Total time: 0:00:01 (1.0340 s / it) +[15:00:03.399031] Averaged stats: lr: 0.000109 loss: 0.1724 (0.1724) +[15:00:04.341657] val: [0/6] eta: 0:00:05 loss: 0.1777 (0.1777) time: 0.9328 data: 0.9166 max mem: 5539 +[15:00:04.380946] val: [5/6] eta: 0:00:00 loss: 0.1777 (0.6808) time: 0.1619 data: 0.1529 max mem: 5539 +[15:00:04.455819] val: Total time: 0:00:01 (0.1746 s / it) +[15:00:04.464594] val loss: 0.6808090209960938 +[15:00:04.464759] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8734, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8776, Kappa: 0.0000, Score: 0.4353 +[15:00:04.502381] Best epoch = 33, Best score = 0.4356 +[15:00:04.785546] log_dir: ./output_logs/retfound +[15:00:05.610338] Epoch: [37] [0/1] eta: 0:00:00 lr: 0.000106 loss: 0.3562 (0.3562) time: 0.8240 data: 0.7763 max mem: 5539 +[15:00:05.676949] Epoch: [37] Total time: 0:00:00 (0.8913 s / it) +[15:00:05.677746] Averaged stats: lr: 0.000106 loss: 0.3562 (0.3562) +[15:00:06.595710] val: [0/6] eta: 0:00:05 loss: 0.1741 (0.1741) time: 0.8945 data: 0.8812 max mem: 5539 +[15:00:06.635931] val: [5/6] eta: 0:00:00 loss: 0.1741 (0.6832) time: 0.1557 data: 0.1469 max mem: 5539 +[15:00:06.707418] val: Total time: 0:00:01 (0.1678 s / it) +[15:00:06.716244] val loss: 0.68316650390625 +[15:00:06.716400] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8727, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8763, Kappa: 0.0000, Score: 0.4350 +[15:00:06.763116] Best epoch = 33, Best score = 0.4356 +[15:00:07.030807] log_dir: ./output_logs/retfound +[15:00:07.799177] Epoch: [38] [0/1] eta: 0:00:00 lr: 0.000103 loss: 0.3630 (0.3630) time: 0.7676 data: 0.7187 max mem: 5539 +[15:00:07.872532] Epoch: [38] Total time: 0:00:00 (0.8416 s / it) +[15:00:07.873259] Averaged stats: lr: 0.000103 loss: 0.3630 (0.3630) +[15:00:08.757598] val: [0/6] eta: 0:00:05 loss: 0.1699 (0.1699) time: 0.8733 data: 0.8605 max mem: 5539 +[15:00:08.812122] val: [5/6] eta: 0:00:00 loss: 0.1699 (0.6862) time: 0.1545 data: 0.1459 max mem: 5539 +[15:00:08.881409] val: Total time: 0:00:00 (0.1663 s / it) +[15:00:08.890243] val loss: 0.6862208048502604 +[15:00:08.890444] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8688, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8719, Kappa: 0.0000, Score: 0.4337 +[15:00:08.941561] Best epoch = 33, Best score = 0.4356 +[15:00:09.218629] log_dir: ./output_logs/retfound +[15:00:10.033694] Epoch: [39] [0/1] eta: 0:00:00 lr: 0.000099 loss: 0.5744 (0.5744) time: 0.8143 data: 0.7664 max mem: 5539 +[15:00:10.099212] Epoch: [39] Total time: 0:00:00 (0.8805 s / it) +[15:00:10.099990] Averaged stats: lr: 0.000099 loss: 0.5744 (0.5744) +[15:00:11.043124] val: [0/6] eta: 0:00:05 loss: 0.1677 (0.1677) time: 0.9204 data: 0.9068 max mem: 5539 +[15:00:11.082947] val: [5/6] eta: 0:00:00 loss: 0.1677 (0.6876) time: 0.1600 data: 0.1512 max mem: 5539 +[15:00:11.153225] val: Total time: 0:00:01 (0.1719 s / it) +[15:00:11.161944] val loss: 0.6876093546549479 +[15:00:11.162113] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8672, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8716, Kappa: 0.0000, Score: 0.4332 +[15:00:11.209520] Best epoch = 33, Best score = 0.4356 +[15:00:11.468797] log_dir: ./output_logs/retfound +[15:00:12.390673] Epoch: [40] [0/1] eta: 0:00:00 lr: 0.000096 loss: 0.7217 (0.7217) time: 0.9211 data: 0.8727 max mem: 5539 +[15:00:12.458806] Epoch: [40] Total time: 0:00:00 (0.9899 s / it) +[15:00:12.459551] Averaged stats: lr: 0.000096 loss: 0.7217 (0.7217) +[15:00:13.383883] val: [0/6] eta: 0:00:05 loss: 0.1693 (0.1693) time: 0.8908 data: 0.8785 max mem: 5539 +[15:00:13.423572] val: [5/6] eta: 0:00:00 loss: 0.1693 (0.6848) time: 0.1550 data: 0.1465 max mem: 5539 +[15:00:13.494134] val: Total time: 0:00:01 (0.1670 s / it) +[15:00:13.502850] val loss: 0.6847763061523438 +[15:00:13.503017] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8648, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8724, Kappa: 0.0000, Score: 0.4324 +[15:00:13.550919] Best epoch = 33, Best score = 0.4356 +[15:00:13.832181] log_dir: ./output_logs/retfound +[15:00:14.719891] Epoch: [41] [0/1] eta: 0:00:00 lr: 0.000092 loss: 0.5899 (0.5899) time: 0.8869 data: 0.8381 max mem: 5539 +[15:00:14.784973] Epoch: [41] Total time: 0:00:00 (0.9526 s / it) +[15:00:14.785757] Averaged stats: lr: 0.000092 loss: 0.5899 (0.5899) +[15:00:15.679347] val: [0/6] eta: 0:00:05 loss: 0.1709 (0.1709) time: 0.8710 data: 0.8559 max mem: 5539 +[15:00:15.718870] val: [5/6] eta: 0:00:00 loss: 0.1709 (0.6819) time: 0.1517 data: 0.1427 max mem: 5539 +[15:00:15.789837] val: Total time: 0:00:00 (0.1637 s / it) +[15:00:15.798504] val loss: 0.681915283203125 +[15:00:15.798661] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8680, Kappa: 0.0000, Score: 0.4327 +[15:00:15.843837] Best epoch = 33, Best score = 0.4356 +[15:00:16.098458] log_dir: ./output_logs/retfound +[15:00:16.917921] Epoch: [42] [0/1] eta: 0:00:00 lr: 0.000089 loss: 0.5132 (0.5132) time: 0.8187 data: 0.7708 max mem: 5539 +[15:00:16.982724] Epoch: [42] Total time: 0:00:00 (0.8841 s / it) +[15:00:16.983551] Averaged stats: lr: 0.000089 loss: 0.5132 (0.5132) +[15:00:17.921147] val: [0/6] eta: 0:00:05 loss: 0.1729 (0.1729) time: 0.9147 data: 0.8953 max mem: 5539 +[15:00:17.977677] val: [5/6] eta: 0:00:00 loss: 0.1729 (0.6787) time: 0.1618 data: 0.1493 max mem: 5539 +[15:00:18.046096] val: Total time: 0:00:01 (0.1734 s / it) +[15:00:18.054897] val loss: 0.6786982218424479 +[15:00:18.055070] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8648, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8671, Kappa: 0.0000, Score: 0.4324 +[15:00:18.101894] Best epoch = 33, Best score = 0.4356 +[15:00:18.363222] log_dir: ./output_logs/retfound +[15:00:19.146886] Epoch: [43] [0/1] eta: 0:00:00 lr: 0.000086 loss: 0.3764 (0.3764) time: 0.7828 data: 0.7342 max mem: 5539 +[15:00:19.215261] Epoch: [43] Total time: 0:00:00 (0.8519 s / it) +[15:00:19.216044] Averaged stats: lr: 0.000086 loss: 0.3764 (0.3764) +[15:00:20.161507] val: [0/6] eta: 0:00:05 loss: 0.1742 (0.1742) time: 0.9222 data: 0.9082 max mem: 5539 +[15:00:20.201069] val: [5/6] eta: 0:00:00 loss: 0.1742 (0.6763) time: 0.1602 data: 0.1514 max mem: 5539 +[15:00:20.277113] val: Total time: 0:00:01 (0.1731 s / it) +[15:00:20.285830] val loss: 0.6762619018554688 +[15:00:20.286025] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8633, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8665, Kappa: 0.0000, Score: 0.4319 +[15:00:20.333141] Best epoch = 33, Best score = 0.4356 +[15:00:20.617055] log_dir: ./output_logs/retfound +[15:00:21.527626] Epoch: [44] [0/1] eta: 0:00:00 lr: 0.000082 loss: 0.5916 (0.5916) time: 0.9098 data: 0.8614 max mem: 5539 +[15:00:21.598624] Epoch: [44] Total time: 0:00:00 (0.9814 s / it) +[15:00:21.599455] Averaged stats: lr: 0.000082 loss: 0.5916 (0.5916) +[15:00:22.525718] val: [0/6] eta: 0:00:05 loss: 0.1774 (0.1774) time: 0.8926 data: 0.8792 max mem: 5539 +[15:00:22.565442] val: [5/6] eta: 0:00:00 loss: 0.1774 (0.6722) time: 0.1553 data: 0.1466 max mem: 5539 +[15:00:22.636821] val: Total time: 0:00:01 (0.1674 s / it) +[15:00:22.645530] val loss: 0.6721598307291666 +[15:00:22.645689] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8625, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8664, Kappa: 0.0000, Score: 0.4316 +[15:00:22.701077] Best epoch = 33, Best score = 0.4356 +[15:00:22.965892] log_dir: ./output_logs/retfound +[15:00:23.746505] Epoch: [45] [0/1] eta: 0:00:00 lr: 0.000079 loss: 0.5407 (0.5407) time: 0.7798 data: 0.7317 max mem: 5539 +[15:00:23.813592] Epoch: [45] Total time: 0:00:00 (0.8476 s / it) +[15:00:23.814433] Averaged stats: lr: 0.000079 loss: 0.5407 (0.5407) +[15:00:24.730649] val: [0/6] eta: 0:00:05 loss: 0.1812 (0.1812) time: 0.8935 data: 0.8782 max mem: 5539 +[15:00:24.770086] val: [5/6] eta: 0:00:00 loss: 0.1812 (0.6680) time: 0.1554 data: 0.1464 max mem: 5539 +[15:00:24.898017] val: Total time: 0:00:01 (0.1769 s / it) +[15:00:24.906670] val loss: 0.6679662068684896 +[15:00:24.906851] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8609, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8647, Kappa: 0.0000, Score: 0.4311 +[15:00:24.954729] Best epoch = 33, Best score = 0.4356 +[15:00:25.249006] log_dir: ./output_logs/retfound +[15:00:26.034453] Epoch: [46] [0/1] eta: 0:00:00 lr: 0.000075 loss: 0.3980 (0.3980) time: 0.7847 data: 0.7356 max mem: 5539 +[15:00:26.100598] Epoch: [46] Total time: 0:00:00 (0.8515 s / it) +[15:00:26.101403] Averaged stats: lr: 0.000075 loss: 0.3980 (0.3980) +[15:00:27.040017] val: [0/6] eta: 0:00:05 loss: 0.1842 (0.1842) time: 0.9157 data: 0.9025 max mem: 5539 +[15:00:27.079364] val: [5/6] eta: 0:00:00 loss: 0.1842 (0.6648) time: 0.1591 data: 0.1505 max mem: 5539 +[15:00:27.149553] val: Total time: 0:00:01 (0.1710 s / it) +[15:00:27.158365] val loss: 0.6648025512695312 +[15:00:27.158546] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8586, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8619, Kappa: 0.0000, Score: 0.4303 +[15:00:27.204030] Best epoch = 33, Best score = 0.4356 +[15:00:27.519091] log_dir: ./output_logs/retfound +[15:00:28.348886] Epoch: [47] [0/1] eta: 0:00:00 lr: 0.000072 loss: 0.6185 (0.6185) time: 0.8290 data: 0.7803 max mem: 5539 +[15:00:28.414395] Epoch: [47] Total time: 0:00:00 (0.8952 s / it) +[15:00:28.415235] Averaged stats: lr: 0.000072 loss: 0.6185 (0.6185) +[15:00:29.327957] val: [0/6] eta: 0:00:05 loss: 0.1876 (0.1876) time: 0.8898 data: 0.8758 max mem: 5539 +[15:00:29.368542] val: [5/6] eta: 0:00:00 loss: 0.1876 (0.6610) time: 0.1550 data: 0.1461 max mem: 5539 +[15:00:29.436902] val: Total time: 0:00:00 (0.1666 s / it) +[15:00:29.445604] val loss: 0.6610209147135416 +[15:00:29.445771] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8570, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8613, Kappa: 0.0000, Score: 0.4298 +[15:00:29.491698] Best epoch = 33, Best score = 0.4356 +[15:00:29.797681] log_dir: ./output_logs/retfound +[15:00:30.611530] Epoch: [48] [0/1] eta: 0:00:00 lr: 0.000068 loss: 0.4057 (0.4057) time: 0.8130 data: 0.7653 max mem: 5539 +[15:00:30.678574] Epoch: [48] Total time: 0:00:00 (0.8807 s / it) +[15:00:30.679399] Averaged stats: lr: 0.000068 loss: 0.4057 (0.4057) +[15:00:31.577671] val: [0/6] eta: 0:00:05 loss: 0.1903 (0.1903) time: 0.8752 data: 0.8610 max mem: 5539 +[15:00:31.618686] val: [5/6] eta: 0:00:00 loss: 0.1903 (0.6582) time: 0.1526 data: 0.1436 max mem: 5539 +[15:00:31.690425] val: Total time: 0:00:00 (0.1647 s / it) +[15:00:31.699164] val loss: 0.6582107543945312 +[15:00:31.699329] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8570, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8619, Kappa: 0.0000, Score: 0.4298 +[15:00:31.744367] Best epoch = 33, Best score = 0.4356 +[15:00:32.007412] log_dir: ./output_logs/retfound +[15:00:32.832436] Epoch: [49] [0/1] eta: 0:00:00 lr: 0.000065 loss: 0.2037 (0.2037) time: 0.8242 data: 0.7761 max mem: 5539 +[15:00:32.899090] Epoch: [49] Total time: 0:00:00 (0.8915 s / it) +[15:00:32.899914] Averaged stats: lr: 0.000065 loss: 0.2037 (0.2037) +[15:00:33.792255] val: [0/6] eta: 0:00:05 loss: 0.1911 (0.1911) time: 0.8593 data: 0.8467 max mem: 5539 +[15:00:33.851894] val: [5/6] eta: 0:00:00 loss: 0.1911 (0.6567) time: 0.1531 data: 0.1445 max mem: 5539 +[15:00:33.922772] val: Total time: 0:00:00 (0.1651 s / it) +[15:00:33.931528] val loss: 0.6567026774088541 +[15:00:33.931760] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8562, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8618, Kappa: 0.0000, Score: 0.4296 +[15:00:33.987991] Best epoch = 33, Best score = 0.4356 +[15:00:34.243996] log_dir: ./output_logs/retfound +[15:00:35.003810] Epoch: [50] [0/1] eta: 0:00:00 lr: 0.000061 loss: 0.3746 (0.3746) time: 0.7590 data: 0.7111 max mem: 5539 +[15:00:35.074845] Epoch: [50] Total time: 0:00:00 (0.8307 s / it) +[15:00:35.075645] Averaged stats: lr: 0.000061 loss: 0.3746 (0.3746) +[15:00:36.014885] val: [0/6] eta: 0:00:05 loss: 0.1912 (0.1912) time: 0.9203 data: 0.9040 max mem: 5539 +[15:00:36.054287] val: [5/6] eta: 0:00:00 loss: 0.1912 (0.6558) time: 0.1599 data: 0.1507 max mem: 5539 +[15:00:36.124873] val: Total time: 0:00:01 (0.1718 s / it) +[15:00:36.133680] val loss: 0.65576171875 +[15:00:36.133891] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8531, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8562, Kappa: 0.0000, Score: 0.4285 +[15:00:36.194748] Best epoch = 33, Best score = 0.4356 +[15:00:36.423408] log_dir: ./output_logs/retfound +[15:00:37.226944] Epoch: [51] [0/1] eta: 0:00:00 lr: 0.000058 loss: 0.7691 (0.7691) time: 0.8027 data: 0.7541 max mem: 5539 +[15:00:37.299610] Epoch: [51] Total time: 0:00:00 (0.8760 s / it) +[15:00:37.300330] Averaged stats: lr: 0.000058 loss: 0.7691 (0.7691) +[15:00:38.201087] val: [0/6] eta: 0:00:05 loss: 0.1926 (0.1926) time: 0.8778 data: 0.8630 max mem: 5539 +[15:00:38.240545] val: [5/6] eta: 0:00:00 loss: 0.1926 (0.6542) time: 0.1528 data: 0.1439 max mem: 5539 +[15:00:38.311702] val: Total time: 0:00:00 (0.1648 s / it) +[15:00:38.320505] val loss: 0.6542078653971354 +[15:00:38.320735] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8547, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8562, Kappa: 0.0000, Score: 0.4290 +[15:00:38.371311] Best epoch = 33, Best score = 0.4356 +[15:00:38.640517] log_dir: ./output_logs/retfound +[15:00:39.443596] Epoch: [52] [0/1] eta: 0:00:00 lr: 0.000055 loss: 0.7091 (0.7091) time: 0.8012 data: 0.7527 max mem: 5539 +[15:00:39.511523] Epoch: [52] Total time: 0:00:00 (0.8709 s / it) +[15:00:39.512321] Averaged stats: lr: 0.000055 loss: 0.7091 (0.7091) +[15:00:40.399223] val: [0/6] eta: 0:00:05 loss: 0.1958 (0.1958) time: 0.8721 data: 0.8596 max mem: 5539 +[15:00:40.451383] val: [5/6] eta: 0:00:00 loss: 0.1958 (0.6512) time: 0.1540 data: 0.1454 max mem: 5539 +[15:00:40.520981] val: Total time: 0:00:00 (0.1657 s / it) +[15:00:40.529572] val loss: 0.6512120564778646 +[15:00:40.529823] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8539, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8562, Kappa: 0.0000, Score: 0.4288 +[15:00:40.564363] Best epoch = 33, Best score = 0.4356 +[15:00:40.846470] log_dir: ./output_logs/retfound +[15:00:41.631529] Epoch: [53] [0/1] eta: 0:00:00 lr: 0.000051 loss: 0.5354 (0.5354) time: 0.7843 data: 0.7364 max mem: 5539 +[15:00:41.702693] Epoch: [53] Total time: 0:00:00 (0.8561 s / it) +[15:00:41.703462] Averaged stats: lr: 0.000051 loss: 0.5354 (0.5354) +[15:00:42.649556] val: [0/6] eta: 0:00:05 loss: 0.1985 (0.1985) time: 0.9235 data: 0.9109 max mem: 5539 +[15:00:42.688974] val: [5/6] eta: 0:00:00 loss: 0.1985 (0.6487) time: 0.1604 data: 0.1519 max mem: 5539 +[15:00:42.759940] val: Total time: 0:00:01 (0.1724 s / it) +[15:00:42.768712] val loss: 0.648681640625 +[15:00:42.768930] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8531, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8563, Kappa: 0.0000, Score: 0.4285 +[15:00:42.815576] Best epoch = 33, Best score = 0.4356 +[15:00:43.065744] log_dir: ./output_logs/retfound +[15:00:43.903052] Epoch: [54] [0/1] eta: 0:00:00 lr: 0.000048 loss: 0.5717 (0.5717) time: 0.8364 data: 0.7878 max mem: 5539 +[15:00:43.968491] Epoch: [54] Total time: 0:00:00 (0.9026 s / it) +[15:00:43.969256] Averaged stats: lr: 0.000048 loss: 0.5717 (0.5717) +[15:00:44.867204] val: [0/6] eta: 0:00:05 loss: 0.2018 (0.2018) time: 0.8864 data: 0.8733 max mem: 5539 +[15:00:44.906226] val: [5/6] eta: 0:00:00 loss: 0.2018 (0.6461) time: 0.1541 data: 0.1456 max mem: 5539 +[15:00:44.975445] val: Total time: 0:00:00 (0.1659 s / it) +[15:00:44.984208] val loss: 0.6461156209309896 +[15:00:44.984472] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8508, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8474, Kappa: 0.0000, Score: 0.4277 +[15:00:45.031021] Best epoch = 33, Best score = 0.4356 +[15:00:45.311767] log_dir: ./output_logs/retfound +[15:00:46.126388] Epoch: [55] [0/1] eta: 0:00:00 lr: 0.000045 loss: 0.3515 (0.3515) time: 0.8138 data: 0.7656 max mem: 5539 +[15:00:46.191821] Epoch: [55] Total time: 0:00:00 (0.8799 s / it) +[15:00:46.192586] Averaged stats: lr: 0.000045 loss: 0.3515 (0.3515) +[15:00:47.114211] val: [0/6] eta: 0:00:05 loss: 0.2042 (0.2042) time: 0.8983 data: 0.8850 max mem: 5539 +[15:00:47.154570] val: [5/6] eta: 0:00:00 loss: 0.2042 (0.6440) time: 0.1564 data: 0.1476 max mem: 5539 +[15:00:47.226899] val: Total time: 0:00:01 (0.1686 s / it) +[15:00:47.235627] val loss: 0.6440480550130209 +[15:00:47.235821] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8414, Kappa: 0.0000, Score: 0.4275 +[15:00:47.282291] Best epoch = 33, Best score = 0.4356 +[15:00:47.546162] log_dir: ./output_logs/retfound +[15:00:48.398157] Epoch: [56] [0/1] eta: 0:00:00 lr: 0.000042 loss: 0.3509 (0.3509) time: 0.8512 data: 0.8012 max mem: 5539 +[15:00:48.466517] Epoch: [56] Total time: 0:00:00 (0.9202 s / it) +[15:00:48.467364] Averaged stats: lr: 0.000042 loss: 0.3509 (0.3509) +[15:00:49.360159] val: [0/6] eta: 0:00:05 loss: 0.2059 (0.2059) time: 0.8696 data: 0.8564 max mem: 5539 +[15:00:49.416241] val: [5/6] eta: 0:00:00 loss: 0.2059 (0.6425) time: 0.1542 data: 0.1454 max mem: 5539 +[15:00:49.485148] val: Total time: 0:00:00 (0.1659 s / it) +[15:00:49.493935] val loss: 0.6425145467122396 +[15:00:49.494148] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8469, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8341, Kappa: 0.0000, Score: 0.4264 +[15:00:49.539307] Best epoch = 33, Best score = 0.4356 +[15:00:49.819021] log_dir: ./output_logs/retfound +[15:00:50.634877] Epoch: [57] [0/1] eta: 0:00:00 lr: 0.000039 loss: 0.5605 (0.5605) time: 0.8151 data: 0.7680 max mem: 5539 +[15:00:50.705600] Epoch: [57] Total time: 0:00:00 (0.8865 s / it) +[15:00:50.706433] Averaged stats: lr: 0.000039 loss: 0.5605 (0.5605) +[15:00:51.634944] val: [0/6] eta: 0:00:05 loss: 0.2078 (0.2078) time: 0.9050 data: 0.8926 max mem: 5539 +[15:00:51.675325] val: [5/6] eta: 0:00:00 loss: 0.2078 (0.6408) time: 0.1575 data: 0.1488 max mem: 5539 +[15:00:51.744131] val: Total time: 0:00:01 (0.1691 s / it) +[15:00:51.753029] val loss: 0.6408055623372396 +[15:00:51.753253] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8469, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8341, Kappa: 0.0000, Score: 0.4264 +[15:00:51.798713] Best epoch = 33, Best score = 0.4356 +[15:00:52.087537] log_dir: ./output_logs/retfound +[15:00:52.931536] Epoch: [58] [0/1] eta: 0:00:00 lr: 0.000036 loss: 0.6621 (0.6621) time: 0.8431 data: 0.7936 max mem: 5539 +[15:00:52.997880] Epoch: [58] Total time: 0:00:00 (0.9102 s / it) +[15:00:52.998708] Averaged stats: lr: 0.000036 loss: 0.6621 (0.6621) +[15:00:53.874174] val: [0/6] eta: 0:00:05 loss: 0.2103 (0.2103) time: 0.8640 data: 0.8503 max mem: 5539 +[15:00:53.914596] val: [5/6] eta: 0:00:00 loss: 0.2103 (0.6388) time: 0.1507 data: 0.1418 max mem: 5539 +[15:00:53.981895] val: Total time: 0:00:00 (0.1620 s / it) +[15:00:53.990478] val loss: 0.6387939453125 +[15:00:53.990640] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8469, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8343, Kappa: 0.0000, Score: 0.4264 +[15:00:54.035671] Best epoch = 33, Best score = 0.4356 +[15:00:54.340456] log_dir: ./output_logs/retfound +[15:00:55.140946] Epoch: [59] [0/1] eta: 0:00:00 lr: 0.000033 loss: 0.5901 (0.5901) time: 0.7997 data: 0.7518 max mem: 5539 +[15:00:55.209607] Epoch: [59] Total time: 0:00:00 (0.8690 s / it) +[15:00:55.210447] Averaged stats: lr: 0.000033 loss: 0.5901 (0.5901) +[15:00:56.169638] val: [0/6] eta: 0:00:05 loss: 0.2130 (0.2130) time: 0.9369 data: 0.9245 max mem: 5539 +[15:00:56.209434] val: [5/6] eta: 0:00:00 loss: 0.2130 (0.6368) time: 0.1627 data: 0.1542 max mem: 5539 +[15:00:56.278910] val: Total time: 0:00:01 (0.1745 s / it) +[15:00:56.287650] val loss: 0.6368230183919271 +[15:00:56.287814] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8445, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8218, Kappa: 0.0000, Score: 0.4257 +[15:00:56.345911] Best epoch = 33, Best score = 0.4356 +[15:00:56.601751] log_dir: ./output_logs/retfound +[15:00:57.430102] Epoch: [60] [0/1] eta: 0:00:00 lr: 0.000030 loss: 0.3851 (0.3851) time: 0.8276 data: 0.7802 max mem: 5539 +[15:00:57.503389] Epoch: [60] Total time: 0:00:00 (0.9015 s / it) +[15:00:57.504164] Averaged stats: lr: 0.000030 loss: 0.3851 (0.3851) +[15:00:58.422914] val: [0/6] eta: 0:00:05 loss: 0.2146 (0.2146) time: 0.9072 data: 0.8904 max mem: 5539 +[15:00:58.496598] val: [5/6] eta: 0:00:00 loss: 0.2146 (0.6356) time: 0.1634 data: 0.1539 max mem: 5539 +[15:00:58.570666] val: Total time: 0:00:01 (0.1759 s / it) +[15:00:58.579591] val loss: 0.6356124877929688 +[15:00:58.579775] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8438, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8218, Kappa: 0.0000, Score: 0.4254 +[15:00:58.673734] Best epoch = 33, Best score = 0.4356 +[15:00:58.970689] log_dir: ./output_logs/retfound +[15:00:59.848092] Epoch: [61] [0/1] eta: 0:00:00 lr: 0.000028 loss: 0.5033 (0.5033) time: 0.8765 data: 0.8278 max mem: 5539 +[15:00:59.915906] Epoch: [61] Total time: 0:00:00 (0.9450 s / it) +[15:00:59.916664] Averaged stats: lr: 0.000028 loss: 0.5033 (0.5033) +[15:01:00.838028] val: [0/6] eta: 0:00:05 loss: 0.2160 (0.2160) time: 0.8990 data: 0.8847 max mem: 5539 +[15:01:00.877738] val: [5/6] eta: 0:00:00 loss: 0.2160 (0.6343) time: 0.1564 data: 0.1475 max mem: 5539 +[15:01:00.948359] val: Total time: 0:00:01 (0.1683 s / it) +[15:01:00.957219] val loss: 0.6343485514322916 +[15:01:00.957400] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8214, Kappa: 0.0000, Score: 0.4251 +[15:01:01.003740] Best epoch = 33, Best score = 0.4356 +[15:01:01.371102] log_dir: ./output_logs/retfound +[15:01:02.168092] Epoch: [62] [0/1] eta: 0:00:00 lr: 0.000025 loss: 0.5230 (0.5230) time: 0.7962 data: 0.7455 max mem: 5539 +[15:01:02.278372] Epoch: [62] Total time: 0:00:00 (0.9071 s / it) +[15:01:02.279255] Averaged stats: lr: 0.000025 loss: 0.5230 (0.5230) +[15:01:03.120023] val: [0/6] eta: 0:00:04 loss: 0.2173 (0.2173) time: 0.8299 data: 0.8168 max mem: 5539 +[15:01:03.159516] val: [5/6] eta: 0:00:00 loss: 0.2173 (0.6332) time: 0.1448 data: 0.1362 max mem: 5539 +[15:01:03.228815] val: Total time: 0:00:00 (0.1565 s / it) +[15:01:03.237592] val loss: 0.6331812540690104 +[15:01:03.237764] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8422, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8171, Kappa: 0.0000, Score: 0.4249 +[15:01:03.328828] Best epoch = 33, Best score = 0.4356 +[15:01:03.612401] log_dir: ./output_logs/retfound +[15:01:04.407312] Epoch: [63] [0/1] eta: 0:00:00 lr: 0.000023 loss: 0.4280 (0.4280) time: 0.7941 data: 0.7443 max mem: 5539 +[15:01:04.473776] Epoch: [63] Total time: 0:00:00 (0.8612 s / it) +[15:01:04.474587] Averaged stats: lr: 0.000023 loss: 0.4280 (0.4280) +[15:01:05.363741] val: [0/6] eta: 0:00:05 loss: 0.2183 (0.2183) time: 0.8629 data: 0.8497 max mem: 5539 +[15:01:05.403982] val: [5/6] eta: 0:00:00 loss: 0.2183 (0.6325) time: 0.1504 data: 0.1417 max mem: 5539 +[15:01:05.494917] val: Total time: 0:00:00 (0.1658 s / it) +[15:01:05.503748] val loss: 0.6324615478515625 +[15:01:05.503931] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4246 +[15:01:05.564242] Best epoch = 33, Best score = 0.4356 +[15:01:05.778713] log_dir: ./output_logs/retfound +[15:01:06.535167] Epoch: [64] [0/1] eta: 0:00:00 lr: 0.000020 loss: 0.3963 (0.3963) time: 0.7557 data: 0.7080 max mem: 5539 +[15:01:06.602402] Epoch: [64] Total time: 0:00:00 (0.8235 s / it) +[15:01:06.603197] Averaged stats: lr: 0.000020 loss: 0.3963 (0.3963) +[15:01:07.550566] val: [0/6] eta: 0:00:05 loss: 0.2190 (0.2190) time: 0.9254 data: 0.9119 max mem: 5539 +[15:01:07.609996] val: [5/6] eta: 0:00:00 loss: 0.2190 (0.6318) time: 0.1640 data: 0.1553 max mem: 5539 +[15:01:07.685789] val: Total time: 0:00:01 (0.1769 s / it) +[15:01:07.694558] val loss: 0.6317698160807291 +[15:01:07.694736] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4246 +[15:01:07.751834] Best epoch = 33, Best score = 0.4356 +[15:01:07.978100] log_dir: ./output_logs/retfound +[15:01:08.808926] Epoch: [65] [0/1] eta: 0:00:00 lr: 0.000018 loss: 0.5812 (0.5812) time: 0.8300 data: 0.7802 max mem: 5539 +[15:01:08.873788] Epoch: [65] Total time: 0:00:00 (0.8956 s / it) +[15:01:08.874570] Averaged stats: lr: 0.000018 loss: 0.5812 (0.5812) +[15:01:09.772868] val: [0/6] eta: 0:00:05 loss: 0.2197 (0.2197) time: 0.8869 data: 0.8729 max mem: 5539 +[15:01:09.812225] val: [5/6] eta: 0:00:00 loss: 0.2197 (0.6313) time: 0.1543 data: 0.1456 max mem: 5539 +[15:01:09.886299] val: Total time: 0:00:01 (0.1668 s / it) +[15:01:09.895580] val loss: 0.6313095092773438 +[15:01:09.895756] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8406, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4244 +[15:01:09.941362] Best epoch = 33, Best score = 0.4356 +[15:01:10.165230] log_dir: ./output_logs/retfound +[15:01:11.000993] Epoch: [66] [0/1] eta: 0:00:00 lr: 0.000016 loss: 0.2145 (0.2145) time: 0.8349 data: 0.7826 max mem: 5539 +[15:01:11.068617] Epoch: [66] Total time: 0:00:00 (0.9032 s / it) +[15:01:11.069462] Averaged stats: lr: 0.000016 loss: 0.2145 (0.2145) +[15:01:12.013665] val: [0/6] eta: 0:00:05 loss: 0.2196 (0.2196) time: 0.9208 data: 0.9082 max mem: 5539 +[15:01:12.114638] val: [5/6] eta: 0:00:00 loss: 0.2196 (0.6311) time: 0.1702 data: 0.1616 max mem: 5539 +[15:01:12.185416] val: Total time: 0:00:01 (0.1822 s / it) +[15:01:12.194090] val loss: 0.631103515625 +[15:01:12.194261] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8406, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4244 +[15:01:12.238983] Best epoch = 33, Best score = 0.4356 +[15:01:12.624525] log_dir: ./output_logs/retfound +[15:01:13.387503] Epoch: [67] [0/1] eta: 0:00:00 lr: 0.000014 loss: 0.6783 (0.6783) time: 0.7622 data: 0.7140 max mem: 5539 +[15:01:13.450639] Epoch: [67] Total time: 0:00:00 (0.8260 s / it) +[15:01:13.451439] Averaged stats: lr: 0.000014 loss: 0.6783 (0.6783) +[15:01:14.304882] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.8421 data: 0.8287 max mem: 5539 +[15:01:14.358249] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6308) time: 0.1492 data: 0.1403 max mem: 5539 +[15:01:14.426451] val: Total time: 0:00:00 (0.1607 s / it) +[15:01:14.435136] val loss: 0.6308085123697916 +[15:01:14.435306] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4246 +[15:01:14.472954] Best epoch = 33, Best score = 0.4356 +[15:01:14.813907] log_dir: ./output_logs/retfound +[15:01:15.565441] Epoch: [68] [0/1] eta: 0:00:00 lr: 0.000012 loss: 0.5520 (0.5520) time: 0.7508 data: 0.7025 max mem: 5539 +[15:01:15.635516] Epoch: [68] Total time: 0:00:00 (0.8215 s / it) +[15:01:15.636306] Averaged stats: lr: 0.000012 loss: 0.5520 (0.5520) +[15:01:16.571015] val: [0/6] eta: 0:00:05 loss: 0.2200 (0.2200) time: 0.9160 data: 0.8999 max mem: 5539 +[15:01:16.610042] val: [5/6] eta: 0:00:00 loss: 0.2200 (0.6306) time: 0.1591 data: 0.1501 max mem: 5539 +[15:01:16.682443] val: Total time: 0:00:01 (0.1713 s / it) +[15:01:16.691138] val loss: 0.6305618286132812 +[15:01:16.691304] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8171, Kappa: 0.0000, Score: 0.4246 +[15:01:16.738901] Best epoch = 33, Best score = 0.4356 +[15:01:17.094284] log_dir: ./output_logs/retfound +[15:01:17.932344] Epoch: [69] [0/1] eta: 0:00:00 lr: 0.000010 loss: 0.4133 (0.4133) time: 0.8373 data: 0.7898 max mem: 5539 +[15:01:18.099660] Epoch: [69] Total time: 0:00:01 (1.0052 s / it) +[15:01:18.100461] Averaged stats: lr: 0.000010 loss: 0.4133 (0.4133) +[15:01:19.122507] val: [0/6] eta: 0:00:05 loss: 0.2200 (0.2200) time: 0.9878 data: 0.9722 max mem: 5539 +[15:01:19.161996] val: [5/6] eta: 0:00:00 loss: 0.2200 (0.6303) time: 0.1711 data: 0.1621 max mem: 5539 +[15:01:19.230041] val: Total time: 0:00:01 (0.1827 s / it) +[15:01:19.239034] val loss: 0.6303456624348959 +[15:01:19.239222] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8406, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8172, Kappa: 0.0000, Score: 0.4244 +[15:01:19.286135] Best epoch = 33, Best score = 0.4356 +[15:01:19.502892] log_dir: ./output_logs/retfound +[15:01:20.346390] Epoch: [70] [0/1] eta: 0:00:00 lr: 0.000009 loss: 0.5130 (0.5130) time: 0.8426 data: 0.7932 max mem: 5539 +[15:01:20.412215] Epoch: [70] Total time: 0:00:00 (0.9092 s / it) +[15:01:20.413063] Averaged stats: lr: 0.000009 loss: 0.5130 (0.5130) +[15:01:21.336177] val: [0/6] eta: 0:00:05 loss: 0.2200 (0.2200) time: 0.8998 data: 0.8868 max mem: 5539 +[15:01:21.375436] val: [5/6] eta: 0:00:00 loss: 0.2200 (0.6303) time: 0.1564 data: 0.1479 max mem: 5539 +[15:01:21.446252] val: Total time: 0:00:01 (0.1684 s / it) +[15:01:21.455119] val loss: 0.6302820841471354 +[15:01:21.455288] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8172, Kappa: 0.0000, Score: 0.4246 +[15:01:21.509106] Best epoch = 33, Best score = 0.4356 +[15:01:21.734318] log_dir: ./output_logs/retfound +[15:01:22.514956] Epoch: [71] [0/1] eta: 0:00:00 lr: 0.000007 loss: 0.5421 (0.5421) time: 0.7798 data: 0.7276 max mem: 5539 +[15:01:22.582810] Epoch: [71] Total time: 0:00:00 (0.8484 s / it) +[15:01:22.583617] Averaged stats: lr: 0.000007 loss: 0.5421 (0.5421) +[15:01:23.523911] val: [0/6] eta: 0:00:05 loss: 0.2201 (0.2201) time: 0.9175 data: 0.9032 max mem: 5539 +[15:01:23.598395] val: [5/6] eta: 0:00:00 loss: 0.2201 (0.6301) time: 0.1652 data: 0.1562 max mem: 5539 +[15:01:23.666994] val: Total time: 0:00:01 (0.1769 s / it) +[15:01:23.675763] val loss: 0.6301396687825521 +[15:01:23.675933] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8422, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8169, Kappa: 0.0000, Score: 0.4249 +[15:01:23.721950] Best epoch = 33, Best score = 0.4356 +[15:01:23.956570] log_dir: ./output_logs/retfound +[15:01:24.767493] Epoch: [72] [0/1] eta: 0:00:00 lr: 0.000006 loss: 0.5093 (0.5093) time: 0.8101 data: 0.7628 max mem: 5539 +[15:01:24.833857] Epoch: [72] Total time: 0:00:00 (0.8771 s / it) +[15:01:24.834586] Averaged stats: lr: 0.000006 loss: 0.5093 (0.5093) +[15:01:25.772631] val: [0/6] eta: 0:00:05 loss: 0.2200 (0.2200) time: 0.9244 data: 0.9090 max mem: 5539 +[15:01:25.811822] val: [5/6] eta: 0:00:00 loss: 0.2200 (0.6301) time: 0.1605 data: 0.1516 max mem: 5539 +[15:01:25.881378] val: Total time: 0:00:01 (0.1723 s / it) +[15:01:25.890215] val loss: 0.6301040649414062 +[15:01:25.890421] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8229, Kappa: 0.0000, Score: 0.4251 +[15:01:25.938854] Best epoch = 33, Best score = 0.4356 +[15:01:26.321668] log_dir: ./output_logs/retfound +[15:01:27.180953] Epoch: [73] [0/1] eta: 0:00:00 lr: 0.000005 loss: 0.2195 (0.2195) time: 0.8585 data: 0.8104 max mem: 5539 +[15:01:27.243578] Epoch: [73] Total time: 0:00:00 (0.9217 s / it) +[15:01:27.244418] Averaged stats: lr: 0.000005 loss: 0.2195 (0.2195) +[15:01:28.167225] val: [0/6] eta: 0:00:05 loss: 0.2199 (0.2199) time: 0.8995 data: 0.8858 max mem: 5539 +[15:01:28.207280] val: [5/6] eta: 0:00:00 loss: 0.2199 (0.6301) time: 0.1565 data: 0.1477 max mem: 5539 +[15:01:28.276206] val: Total time: 0:00:01 (0.1682 s / it) +[15:01:28.284887] val loss: 0.6300532023111979 +[15:01:28.285055] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8422, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8170, Kappa: 0.0000, Score: 0.4249 +[15:01:28.334020] Best epoch = 33, Best score = 0.4356 +[15:01:28.706604] log_dir: ./output_logs/retfound +[15:01:29.589588] Epoch: [74] [0/1] eta: 0:00:00 lr: 0.000004 loss: 0.6375 (0.6375) time: 0.8822 data: 0.8340 max mem: 5539 +[15:01:29.661606] Epoch: [74] Total time: 0:00:00 (0.9548 s / it) +[15:01:29.662410] Averaged stats: lr: 0.000004 loss: 0.6375 (0.6375) +[15:01:30.531566] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.8580 data: 0.8447 max mem: 5539 +[15:01:30.571570] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6300) time: 0.1496 data: 0.1409 max mem: 5539 +[15:01:30.704449] val: Total time: 0:00:01 (0.1719 s / it) +[15:01:30.713292] val loss: 0.6299972534179688 +[15:01:30.713477] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8232, Kappa: 0.0000, Score: 0.4251 +[15:01:30.759382] Best epoch = 33, Best score = 0.4356 +[15:01:31.002973] log_dir: ./output_logs/retfound +[15:01:31.767822] Epoch: [75] [0/1] eta: 0:00:00 lr: 0.000003 loss: 0.5551 (0.5551) time: 0.7641 data: 0.7156 max mem: 5539 +[15:01:31.837104] Epoch: [75] Total time: 0:00:00 (0.8340 s / it) +[15:01:31.837910] Averaged stats: lr: 0.000003 loss: 0.5551 (0.5551) +[15:01:32.746873] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.8859 data: 0.8719 max mem: 5539 +[15:01:32.786075] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6299) time: 0.1541 data: 0.1454 max mem: 5539 +[15:01:32.900062] val: Total time: 0:00:01 (0.1733 s / it) +[15:01:32.908772] val loss: 0.6299362182617188 +[15:01:32.909173] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8233, Kappa: 0.0000, Score: 0.4251 +[15:01:32.947486] Best epoch = 33, Best score = 0.4356 +[15:01:33.175918] log_dir: ./output_logs/retfound +[15:01:34.012083] Epoch: [76] [0/1] eta: 0:00:00 lr: 0.000002 loss: 0.5967 (0.5967) time: 0.8354 data: 0.7867 max mem: 5539 +[15:01:34.074714] Epoch: [76] Total time: 0:00:00 (0.8986 s / it) +[15:01:34.075494] Averaged stats: lr: 0.000002 loss: 0.5967 (0.5967) +[15:01:35.074147] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.9736 data: 0.9611 max mem: 5539 +[15:01:35.114541] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6300) time: 0.1689 data: 0.1603 max mem: 5539 +[15:01:35.185453] val: Total time: 0:00:01 (0.1809 s / it) +[15:01:35.194131] val loss: 0.6299692789713541 +[15:01:35.194290] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8233, Kappa: 0.0000, Score: 0.4251 +[15:01:35.240883] Best epoch = 33, Best score = 0.4356 +[15:01:35.667340] log_dir: ./output_logs/retfound +[15:01:36.519631] Epoch: [77] [0/1] eta: 0:00:00 lr: 0.000002 loss: 0.5454 (0.5454) time: 0.8515 data: 0.8029 max mem: 5539 +[15:01:36.734841] Epoch: [77] Total time: 0:00:01 (1.0674 s / it) +[15:01:36.735703] Averaged stats: lr: 0.000002 loss: 0.5454 (0.5454) +[15:01:37.656188] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.8977 data: 0.8823 max mem: 5539 +[15:01:37.696061] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6300) time: 0.1562 data: 0.1471 max mem: 5539 +[15:01:37.828285] val: Total time: 0:00:01 (0.1784 s / it) +[15:01:37.836946] val loss: 0.6299616495768229 +[15:01:37.837107] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8233, Kappa: 0.0000, Score: 0.4251 +[15:01:37.897671] Best epoch = 33, Best score = 0.4356 +[15:01:38.136599] log_dir: ./output_logs/retfound +[15:01:38.942534] Epoch: [78] [0/1] eta: 0:00:00 lr: 0.000001 loss: 0.5910 (0.5910) time: 0.8051 data: 0.7548 max mem: 5539 +[15:01:39.101513] Epoch: [78] Total time: 0:00:00 (0.9648 s / it) +[15:01:39.103174] Averaged stats: lr: 0.000001 loss: 0.5910 (0.5910) +[15:01:40.143692] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.9997 data: 0.9866 max mem: 5539 +[15:01:40.203682] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6300) time: 0.1765 data: 0.1678 max mem: 5539 +[15:01:40.274418] val: Total time: 0:00:01 (0.1885 s / it) +[15:01:40.284006] val loss: 0.6299540201822916 +[15:01:40.284198] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8233, Kappa: 0.0000, Score: 0.4251 +[15:01:40.350482] Best epoch = 33, Best score = 0.4356 +[15:01:40.834734] log_dir: ./output_logs/retfound +[15:01:41.658803] Epoch: [79] [0/1] eta: 0:00:00 lr: 0.000001 loss: 0.3429 (0.3429) time: 0.8232 data: 0.7694 max mem: 5539 +[15:01:41.831858] Epoch: [79] Total time: 0:00:00 (0.9969 s / it) +[15:01:41.832739] Averaged stats: lr: 0.000001 loss: 0.3429 (0.3429) +[15:01:42.780690] val: [0/6] eta: 0:00:05 loss: 0.2198 (0.2198) time: 0.9245 data: 0.9100 max mem: 5539 +[15:01:42.820322] val: [5/6] eta: 0:00:00 loss: 0.2198 (0.6299) time: 0.1606 data: 0.1518 max mem: 5539 +[15:01:43.020762] val: Total time: 0:00:01 (0.1942 s / it) +[15:01:43.034063] val loss: 0.6299387613932291 +[15:01:43.034335] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8430, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8233, Kappa: 0.0000, Score: 0.4251 +[15:01:43.287255] Best epoch = 33, Best score = 0.4356 +[15:01:46.066628] Test with the best model, epoch = 33: +[15:01:46.952098] test: [ 0/11] eta: 0:00:09 loss: 0.1628 (0.1628) time: 0.8635 data: 0.8517 max mem: 5539 +[15:01:47.318952] test: [10/11] eta: 0:00:00 loss: 0.1867 (0.5388) time: 0.1118 data: 0.0949 max mem: 5539 +[15:01:47.391319] test: Total time: 0:00:01 (0.1185 s / it) +[15:01:47.400837] val loss: 0.53875732421875 +[15:01:47.400952] Accuracy: 0.8095, F1 Score: 0.4474, ROC AUC: 0.6811, Hamming Loss: 0.1905, + Jaccard Score: 0.4048, Precision: 0.4048, Recall: 0.5000, + Average Precision: 0.6268, Kappa: 0.0000, Score: 0.3761 +[15:01:48.367133] Training time 0:03:44 +[rank0]:[W701 15:01:48.664167104 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/005/retfound acc=0.8095 auroc=0.6806066176470589 f1_macro=0.4474 qwk=0.0 diff --git a/results/downsample/papila/005/vit/confusion_matrix.png b/results/downsample/papila/005/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..b22430762fc1e42f6c0d532772ce0e04b9013df8 --- /dev/null +++ b/results/downsample/papila/005/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:990b3947aaa9a66a42f1942f19418620b575135e3f9acd191e07b3bde58862f0 +size 71508 diff --git a/results/downsample/papila/005/vit/log.csv b/results/downsample/papila/005/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..78c35eae77107cd84894e04f2fc5e4408a916e48 --- /dev/null +++ b/results/downsample/papila/005/vit/log.csv @@ -0,0 +1,27 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6040165424346924,0.5,0.55625,0.3775619783996073,0.0 +1,0.8981707096099854,0.5,0.55625,0.3775619783996073,7.39441178203743e-08 +2,0.6024819016456604,0.7380952380952381,0.5843750000000001,0.3777386680800219,1.478882356407486e-07 +3,1.2674440145492554,0.6190476190476191,0.603125,0.4044604700854701,2.2183235346112292e-07 +4,0.5617198348045349,0.6428571428571429,0.671875,0.5083668992825132,2.957764712814972e-07 +5,0.6496670246124268,0.6190476190476191,0.728125,0.5269720047651464,3.697205891018715e-07 +6,0.9378707408905029,0.6190476190476191,0.728125,0.5269720047651464,3.6955843521557525e-07 +7,0.47772127389907837,0.4523809523809524,0.765625,0.45825446582802004,3.6907225802970327e-07 +8,0.3764684200286865,0.2857142857142857,0.753125,0.3475725867269985,3.682629104642438e-07 +9,0.41049936413764954,0.30952380952380953,0.803125,0.37995642505828625,3.671318123898447e-07 +10,0.3508812189102173,0.40476190476190477,0.828125,0.44916048684431037,3.6568094813687817e-07 +11,0.508713960647583,0.5,0.828125,0.5091594720354097,3.6391286301425095e-07 +12,0.48519766330718994,0.5714285714285714,0.80625,0.5355510752688172,3.618306588440675e-07 +13,0.4946104884147644,0.5238095238095238,0.784375,0.4976716633681737,3.594379885199801e-07 +14,0.8590155839920044,0.7380952380952381,0.765625,0.627855362792418,3.567390495987718e-07 +15,0.5274534821510315,0.7857142857142857,0.759375,0.6515138564703855,3.537385769364163e-07 +16,0.8008548617362976,0.7619047619047619,0.784375,0.5753805532958217,3.504418343815308e-07 +17,0.5717078447341919,0.7142857142857143,0.790625,0.3736949233716475,3.4685460554079696e-07 +18,0.3504943549633026,0.7142857142857143,0.7875,0.3726532567049808,3.4298318363255025e-07 +19,0.5889942049980164,0.7380952380952381,0.79375,0.4475303347466884,3.3883436044633555e-07 +20,0.6673991680145264,0.8095238095238095,0.79375,0.625410272804774,3.344154144278013e-07 +21,0.32593899965286255,0.7380952380952381,0.7906249999999999,0.6361886961257514,3.297340979098317e-07 +22,0.3154484033584595,0.5,0.778125,0.48063750141373013,3.2479862351232145e-07 +23,0.5328191518783569,0.42857142857142855,0.778125,0.44746134288086337,3.1961764973444924e-07 +24,0.658480167388916,0.5238095238095238,0.796875,0.5018383300348402,3.1420026576472814e-07 +25,0.3611402213573456,0.7142857142857143,0.8125,0.6244747899159664,3.0855597553548053e-07 diff --git a/results/downsample/papila/005/vit/metrics.json b/results/downsample/papila/005/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..9f64a84edb70491ac49c68755030e6efbdc61e8e --- /dev/null +++ b/results/downsample/papila/005/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7142857142857143, + "balanced_accuracy": 0.65625, + "precision_macro": 0.6127320954907162, + "recall_macro": 0.65625, + "f1_macro": 0.6190476190476191, + "precision_weighted": 0.7777567260325879, + "recall_weighted": 0.7142857142857143, + "f1_weighted": 0.7369614512471655, + "cohen_kappa": 0.25222551928783377, + "quadratic_weighted_kappa": 0.25222551928783377, + "mcc": 0.2654384291727511, + "auroc": 0.703125, + "auprc": 0.41068571087808853, + "sensitivity": 0.5625, + "specificity": 0.75, + "precision_pos": 0.34615384615384615, + "f1_pos": 0.42857142857142855, + "per_class": { + "0": { + "precision": 0.8793103448275862, + "recall": 0.75, + "f1-score": 0.8095238095238095, + "support": 68.0 + }, + "1": { + "precision": 0.34615384615384615, + "recall": 0.5625, + "f1-score": 0.42857142857142855, + "support": 16.0 + }, + "accuracy": 0.7142857142857143, + "macro avg": { + "precision": 0.6127320954907162, + "recall": 0.65625, + "f1-score": 0.6190476190476191, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7777567260325879, + "recall": 0.7142857142857143, + "f1-score": 0.7369614512471655, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/005/vit/pr.png b/results/downsample/papila/005/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..38faecfea073ca4c0f5912e74fd726ca2c6f8adf --- /dev/null +++ b/results/downsample/papila/005/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:561e07640fcb650a4dcba11933767434ed4b3b6a25473ede010afe3ebb977abe +size 52104 diff --git a/results/downsample/papila/005/vit/roc.png b/results/downsample/papila/005/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..8cc08232fa90758b187628ea4c2adbad9b74f36d --- /dev/null +++ b/results/downsample/papila/005/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:868c3389af6dbeed6235917dbf8e168748c5da2e9ac73137248d7ea3d1a8f1ae +size 57648 diff --git a/results/downsample/papila/005/vit/test_pred.npz b/results/downsample/papila/005/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..f60b5001fd1480922c8d06fbbc1aa379031163b7 --- /dev/null +++ b/results/downsample/papila/005/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0f9dc2353989157ab95615fef1d5d28ce45ce298ff390e3ced8b58016f07222 +size 1854 diff --git a/results/downsample/papila/005/vit/train.log b/results/downsample/papila/005/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9a28aa180d9d6653db7f25c3d97697d6dd96f7a3 --- /dev/null +++ b/results/downsample/papila/005/vit/train.log @@ -0,0 +1,139 @@ +[vit] train=15 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.6040 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.8982 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6025 val_acc=0.7381 val_auc=0.5844 score=0.3777 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.2674 val_acc=0.6190 val_auc=0.6031 score=0.4045 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.5617 val_acc=0.6429 val_auc=0.6719 score=0.5084 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.6497 val_acc=0.6190 val_auc=0.7281 score=0.5270 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.9379 val_acc=0.6190 val_auc=0.7281 score=0.5270 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.4777 val_acc=0.4524 val_auc=0.7656 score=0.4583 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.3765 val_acc=0.2857 val_auc=0.7531 score=0.3476 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.4105 val_acc=0.3095 val_auc=0.8031 score=0.3800 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.3509 val_acc=0.4048 val_auc=0.8281 score=0.4492 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.5087 val_acc=0.5000 val_auc=0.8281 score=0.5092 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.4852 val_acc=0.5714 val_auc=0.8063 score=0.5356 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4946 val_acc=0.5238 val_auc=0.7844 score=0.4977 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.8590 val_acc=0.7381 val_auc=0.7656 score=0.6279 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.5275 val_acc=0.7857 val_auc=0.7594 score=0.6515 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.8009 val_acc=0.7619 val_auc=0.7844 score=0.5754 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.5717 val_acc=0.7143 val_auc=0.7906 score=0.3737 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.3505 val_acc=0.7143 val_auc=0.7875 score=0.3727 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.5890 val_acc=0.7381 val_auc=0.7937 score=0.4475 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.6674 val_acc=0.8095 val_auc=0.7937 score=0.6254 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.3259 val_acc=0.7381 val_auc=0.7906 score=0.6362 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3154 val_acc=0.5000 val_auc=0.7781 score=0.4806 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.5328 val_acc=0.4286 val_auc=0.7781 score=0.4475 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.6585 val_acc=0.5238 val_auc=0.7969 score=0.5018 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3611 val_acc=0.7143 val_auc=0.8125 score=0.6245 +[vit] early stop at ep25 (best ep15 score=0.6515) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=15 best_val_score=0.6515 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/005/vit acc=0.7143 auroc=0.703125 f1_macro=0.6190 qwk=0.25222551928783377 diff --git a/results/downsample/papila/010/resnet/confusion_matrix.png b/results/downsample/papila/010/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..0deadcac3adb95aa3dde61582ce117ef0d541c93 --- /dev/null +++ b/results/downsample/papila/010/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e147c3fc1c67ef380c002b3d210f8f424cc78e3c1e4efcc88b32285ecac6d17c +size 73580 diff --git a/results/downsample/papila/010/resnet/log.csv b/results/downsample/papila/010/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..19e41261604bc888599ff02cf55aaa288f7e87df --- /dev/null +++ b/results/downsample/papila/010/resnet/log.csv @@ -0,0 +1,25 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6950976848602295,0.2857142857142857,0.6500000000000001,0.3131975867269985,0.0 +1,0.6795051693916321,0.2619047619047619,0.5375,0.2596700182388088,0.00016666666666666666 +2,0.6728608012199402,0.30952380952380953,0.53125,0.28127114471794457,0.0003333333333333333 +3,0.6931819915771484,0.38095238095238093,0.49375,0.27166066237862574,0.0005 +4,0.6516298055648804,0.47619047619047616,0.50625,0.3096125482187429,0.0004997919498361457 +5,0.6672137975692749,0.6428571428571429,0.703125,0.5001057977472368,0.0004991681456235483 +6,0.6451507210731506,0.6190476190476191,0.76875,0.5405136714318131,0.000498129625622757 +7,0.702385663986206,0.6190476190476191,0.7875000000000001,0.5611150409530902,0.0004966781183478222 +8,0.6703510284423828,0.5714285714285714,0.696875,0.468754963694123,0.0004948160396893552 +9,0.6208155751228333,0.5952380952380952,0.7171875,0.5072955868076009,0.0004925464888935161 +10,0.6265327334403992,0.5238095238095238,0.671875,0.43056046785802926,0.0004898732434036243 +11,0.582933247089386,0.5714285714285714,0.74375,0.5147177419354838,0.00048680075257297753 +12,0.5731092095375061,0.5952380952380952,0.803125,0.5628007315353879,0.0004833341302593417 +13,0.541921079158783,0.6666666666666666,0.8187500000000001,0.6175381608339539,0.0004794791463134399 +14,0.5688995122909546,0.5952380952380952,0.80625,0.5369830868076009,0.00047524221697560476 +15,0.5205231308937073,0.5952380952380952,0.7250000000000001,0.475107485853515,0.0004706303941965803 +16,0.5258108973503113,0.5952380952380952,0.71875,0.47302415252018165,0.00046565135390024513 +17,0.5018647313117981,0.5952380952380952,0.65625,0.4310226818913128,0.0004603133832077953 +18,0.5410982370376587,0.6666666666666666,0.634375,0.4478802777349042,0.00045462536664464835 +19,0.49745121598243713,0.6190476190476191,0.5875,0.33749999999999997,0.0004485967713530281 +20,0.40457651019096375,0.6666666666666666,0.590625,0.4041666666666666,0.00044223763133484053 +21,0.45814451575279236,0.7142857142857143,0.546875,0.3901987425404944,0.0004355585307510675 +22,0.4777285158634186,0.7142857142857143,0.46875,0.36415707587382773,0.00042857058630547593 +23,0.29355186223983765,0.7380952380952381,0.5906250000000001,0.3798220014133551,0.00042128542874196107 diff --git a/results/downsample/papila/010/resnet/metrics.json b/results/downsample/papila/010/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..50e81ccc6d915b22c1c4fd8b1a1572cc2bed20de --- /dev/null +++ b/results/downsample/papila/010/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.4880952380952381, + "balanced_accuracy": 0.5643382352941176, + "precision_macro": 0.5408163265306122, + "recall_macro": 0.5643382352941176, + "f1_macro": 0.4604929051530993, + "precision_weighted": 0.7366375121477162, + "recall_weighted": 0.4880952380952381, + "f1_weighted": 0.536036132152637, + "cohen_kappa": 0.07194244604316535, + "quadratic_weighted_kappa": 0.07194244604316535, + "mcc": 0.10249000771134846, + "auroc": 0.609375, + "auprc": 0.2686309148198843, + "sensitivity": 0.6875, + "specificity": 0.4411764705882353, + "precision_pos": 0.22448979591836735, + "f1_pos": 0.3384615384615385, + "per_class": { + "0": { + "precision": 0.8571428571428571, + "recall": 0.4411764705882353, + "f1-score": 0.5825242718446602, + "support": 68.0 + }, + "1": { + "precision": 0.22448979591836735, + "recall": 0.6875, + "f1-score": 0.3384615384615385, + "support": 16.0 + }, + "accuracy": 0.4880952380952381, + "macro avg": { + "precision": 0.5408163265306122, + "recall": 0.5643382352941176, + "f1-score": 0.4604929051530993, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7366375121477162, + "recall": 0.4880952380952381, + "f1-score": 0.536036132152637, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/010/resnet/pr.png b/results/downsample/papila/010/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..9e5c30d3fb6eba8c7a4d1d3d475b6e309bf570e1 --- /dev/null +++ b/results/downsample/papila/010/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e922218c1a1e3ea6b8afa70b10b815888fce9c330667f01130275116a0314cb7 +size 50165 diff --git a/results/downsample/papila/010/resnet/roc.png b/results/downsample/papila/010/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..a434ddde37003e9f1ffc0dd871c294a51f91a0b4 --- /dev/null +++ b/results/downsample/papila/010/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6dc1f4ac2d206792f45df9b4705570674868ea97a42bcdc23fa5b52516d37fff +size 58148 diff --git a/results/downsample/papila/010/resnet/test_pred.npz b/results/downsample/papila/010/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..571e3e14ac78759dda2deb107fa5ce1341a7b028 --- /dev/null +++ b/results/downsample/papila/010/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84dd973371556ec7cb625c47910115b6741f45a58cdb02e41b0abd0ab6ecd4df +size 1854 diff --git a/results/downsample/papila/010/resnet/train.log b/results/downsample/papila/010/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..cd8915159b98d0a1d049173e2558bbe89c44bda1 --- /dev/null +++ b/results/downsample/papila/010/resnet/train.log @@ -0,0 +1,129 @@ +[resnet] train=29 val=42 test=84 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6951 val_acc=0.2857 val_auc=0.6500 score=0.3132 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6795 val_acc=0.2619 val_auc=0.5375 score=0.2597 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6729 val_acc=0.3095 val_auc=0.5312 score=0.2813 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6932 val_acc=0.3810 val_auc=0.4938 score=0.2717 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6516 val_acc=0.4762 val_auc=0.5062 score=0.3096 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6672 val_acc=0.6429 val_auc=0.7031 score=0.5001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6452 val_acc=0.6190 val_auc=0.7688 score=0.5405 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.7024 val_acc=0.6190 val_auc=0.7875 score=0.5611 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6704 val_acc=0.5714 val_auc=0.6969 score=0.4688 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6208 val_acc=0.5952 val_auc=0.7172 score=0.5073 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.6265 val_acc=0.5238 val_auc=0.6719 score=0.4306 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.5829 val_acc=0.5714 val_auc=0.7438 score=0.5147 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.5731 val_acc=0.5952 val_auc=0.8031 score=0.5628 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.5419 val_acc=0.6667 val_auc=0.8188 score=0.6175 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.5689 val_acc=0.5952 val_auc=0.8063 score=0.5370 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.5205 val_acc=0.5952 val_auc=0.7250 score=0.4751 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.5258 val_acc=0.5952 val_auc=0.7188 score=0.4730 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.5019 val_acc=0.5952 val_auc=0.6562 score=0.4310 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.5411 val_acc=0.6667 val_auc=0.6344 score=0.4479 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.4975 val_acc=0.6190 val_auc=0.5875 score=0.3375 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.4046 val_acc=0.6667 val_auc=0.5906 score=0.4042 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.4581 val_acc=0.7143 val_auc=0.5469 score=0.3902 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.4777 val_acc=0.7143 val_auc=0.4688 score=0.3642 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.2936 val_acc=0.7381 val_auc=0.5906 score=0.3798 +[resnet] early stop at ep23 (best ep13 score=0.6175) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=13 best_val_score=0.6175 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/010/resnet acc=0.4881 auroc=0.609375 f1_macro=0.4605 qwk=0.07194244604316535 diff --git a/results/downsample/papila/010/retfound/confusion_matrix.png b/results/downsample/papila/010/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7627f40aad23385b9ca50fc133405e269ee87449 --- /dev/null +++ b/results/downsample/papila/010/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7c220a1f8fb4a561324744530bb09bbf8cda1e520112fbfcbb7517b6f9ee185 +size 73143 diff --git a/results/downsample/papila/010/retfound/confusion_matrix_test.jpg b/results/downsample/papila/010/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a44063cdd273cf01b218e16dedbcc5eb44589de0 --- /dev/null +++ b/results/downsample/papila/010/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbf6a7f569061967235b05e90273681d0d6dd737bd7ced46cbededfec57d2cd +size 241221 diff --git a/results/downsample/papila/010/retfound/log.txt b/results/downsample/papila/010/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..bcc0f27efbbb336acebf1f41cf50098f8f92b5a1 --- /dev/null +++ b/results/downsample/papila/010/retfound/log.txt @@ -0,0 +1,80 @@ +{"train_lr": 0.0, "train_loss": 0.6928385496139526, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 2.9296875e-05, "train_loss": 0.6925130486488342, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 5.859375e-05, "train_loss": 0.6927408576011658, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 8.7890625e-05, "train_loss": 0.6929036378860474, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0001171875, "train_loss": 0.6711263060569763, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.000146484375, "train_loss": 0.65576171875, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00017578125, "train_loss": 0.6249023675918579, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.00020507812500000002, "train_loss": 0.5596354007720947, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000234375, "train_loss": 0.5167154669761658, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.00026367187499999996, "train_loss": 0.4763997495174408, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.00029296875, "train_loss": 0.5113850831985474, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.00029282175344836725, "train_loss": 0.5041259527206421, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.00029238105982496126, "train_loss": 0.2924031615257263, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.00029164755662809086, "train_loss": 0.5396362543106079, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.00029062272103557807, "train_loss": 0.8083943724632263, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0002893086169299184, "train_loss": 0.5720580816268921, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0002877078907418962, "train_loss": 0.7302327752113342, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.00028582376612102533, "train_loss": 0.662585437297821, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.00028366003744354786, "train_loss": 0.40493977069854736, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.00028122106217106567, "train_loss": 0.5518717169761658, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0002785117520751922, "train_loss": 0.549816906452179, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0002755375633458985, "train_loss": 0.3789835572242737, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00027230448560347243, "train_loss": 0.4811360538005829, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0002688190298362208, "train_loss": 0.3959309756755829, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0002650882152882052, "train_loss": 0.40986329317092896, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0002611195553234188, "train_loss": 0.37486979365348816, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0002569210422948709, "train_loss": 0.3924723267555237, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.00025250113144905187, "train_loss": 0.7294840216636658, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.0002478687238981918, "train_loss": 0.411376953125, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0002430331486946052, "train_loss": 0.5582438111305237, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00023800414404322103, "train_loss": 0.4812174439430237, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.000232791837690134, "train_loss": 0.566577136516571, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00022740672652667176, "train_loss": 0.49333494901657104, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022185965545005283, "train_loss": 0.6161864995956421, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00021616179552320664, "train_loss": 0.5009114742279053, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00021032462147773973, "train_loss": 0.46320801973342896, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00020435988860535324, "train_loss": 0.46202799677848816, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00019827960908424974, "train_loss": 0.5161946415901184, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.0001920960277882052, "train_loss": 0.5165120363235474, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00018582159762702254, "train_loss": 0.6889892816543579, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 0.0001794689544680288, "train_loss": 0.4135904908180237, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 0.0001730508916891197, "train_loss": 0.38878580927848816, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 0.00016658033441459928, "train_loss": 0.6777588129043579, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 0.00016007031348569967, "train_loss": 0.5455240607261658, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 0.00015353393921820096, "train_loss": 0.5132974982261658, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 0.00014698437499999998, "train_loss": 0.541308581829071, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 0.00014043481078179906, "train_loss": 0.5904866456985474, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 0.00013389843651430032, "train_loss": 0.6557210087776184, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 0.0001273884155854007, "train_loss": 0.4253987669944763, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 0.00012091785831088028, "train_loss": 0.4256184995174408, "epoch": 49, "n_parameters": 303303682} +{"train_lr": 0.00011449979553197115, "train_loss": 0.4089111387729645, "epoch": 50, "n_parameters": 303303682} +{"train_lr": 0.00010814715237297744, "train_loss": 0.40003255009651184, "epoch": 51, "n_parameters": 303303682} +{"train_lr": 0.00010187272221179478, "train_loss": 0.44401854276657104, "epoch": 52, "n_parameters": 303303682} +{"train_lr": 9.568914091575024e-05, "train_loss": 0.44519856572151184, "epoch": 53, "n_parameters": 303303682} +{"train_lr": 8.960886139464677e-05, "train_loss": 0.46158039569854736, "epoch": 54, "n_parameters": 303303682} +{"train_lr": 8.364412852226022e-05, "train_loss": 0.6363443732261658, "epoch": 55, "n_parameters": 303303682} +{"train_lr": 7.780695447679333e-05, "train_loss": 0.40217286348342896, "epoch": 56, "n_parameters": 303303682} +{"train_lr": 7.210909454994717e-05, "train_loss": 0.38379719853401184, "epoch": 57, "n_parameters": 303303682} +{"train_lr": 6.656202347332823e-05, "train_loss": 0.4485270082950592, "epoch": 58, "n_parameters": 303303682} +{"train_lr": 6.1176912309866e-05, "train_loss": 0.3871012330055237, "epoch": 59, "n_parameters": 303303682} +{"train_lr": 5.5964605956778944e-05, "train_loss": 0.5774332880973816, "epoch": 60, "n_parameters": 303303682} +{"train_lr": 5.0935601305394765e-05, "train_loss": 0.6259358525276184, "epoch": 61, "n_parameters": 303303682} +{"train_lr": 4.610002610180817e-05, "train_loss": 0.5669270753860474, "epoch": 62, "n_parameters": 303303682} +{"train_lr": 4.146761855094815e-05, "train_loss": 0.6846679449081421, "epoch": 63, "n_parameters": 303303682} +{"train_lr": 3.704770770512911e-05, "train_loss": 0.3912516236305237, "epoch": 64, "n_parameters": 303303682} +{"train_lr": 3.284919467658121e-05, "train_loss": 0.4475911557674408, "epoch": 65, "n_parameters": 303303682} +{"train_lr": 2.8880534711794795e-05, "train_loss": 0.38915202021598816, "epoch": 66, "n_parameters": 303303682} +{"train_lr": 2.5149720163779206e-05, "train_loss": 0.32262369990348816, "epoch": 67, "n_parameters": 303303682} +{"train_lr": 2.1664264396527578e-05, "train_loss": 0.6108642816543579, "epoch": 68, "n_parameters": 303303682} +{"train_lr": 1.843118665410151e-05, "train_loss": 0.5174316167831421, "epoch": 69, "n_parameters": 303303682} +{"train_lr": 1.5456997924807793e-05, "train_loss": 0.5917806029319763, "epoch": 70, "n_parameters": 303303682} +{"train_lr": 1.2747687828934333e-05, "train_loss": 0.4821370542049408, "epoch": 71, "n_parameters": 303303682} +{"train_lr": 1.0308712556452114e-05, "train_loss": 0.5819498896598816, "epoch": 72, "n_parameters": 303303682} +{"train_lr": 8.144983878974696e-06, "train_loss": 0.23827311396598816, "epoch": 73, "n_parameters": 303303682} +{"train_lr": 6.260859258103799e-06, "train_loss": 0.2804606258869171, "epoch": 74, "n_parameters": 303303682} +{"train_lr": 4.660133070081592e-06, "train_loss": 0.4970459043979645, "epoch": 75, "n_parameters": 303303682} +{"train_lr": 3.3460289644219218e-06, "train_loss": 0.5826497673988342, "epoch": 76, "n_parameters": 303303682} +{"train_lr": 2.3211933719091574e-06, "train_loss": 0.4211263060569763, "epoch": 77, "n_parameters": 303303682} +{"train_lr": 1.5876901750387716e-06, "train_loss": 0.4512532651424408, "epoch": 78, "n_parameters": 303303682} +{"train_lr": 1.146996551632778e-06, "train_loss": 0.40611571073532104, "epoch": 79, "n_parameters": 303303682} diff --git a/results/downsample/papila/010/retfound/metrics.json b/results/downsample/papila/010/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..f1708121338d6fb97b9aa29a5bee8ae2ae867bb6 --- /dev/null +++ b/results/downsample/papila/010/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8095238095238095, + "balanced_accuracy": 0.5, + "precision_macro": 0.40476190476190477, + "recall_macro": 0.5, + "f1_macro": 0.4473684210526316, + "precision_weighted": 0.655328798185941, + "recall_weighted": 0.8095238095238095, + "f1_weighted": 0.7243107769423559, + "cohen_kappa": 0.0, + "quadratic_weighted_kappa": 0.0, + "mcc": 0.0, + "auroc": 0.6387867647058825, + "auprc": 0.36409260533246335, + "sensitivity": 0.0, + "specificity": 1.0, + "precision_pos": null, + "f1_pos": 0.0, + "per_class": { + "0": { + "precision": 0.8095238095238095, + "recall": 1.0, + "f1-score": 0.8947368421052632, + "support": 68.0 + }, + "1": { + "precision": 0.0, + "recall": 0.0, + "f1-score": 0.0, + "support": 16.0 + }, + "accuracy": 0.8095238095238095, + "macro avg": { + "precision": 0.40476190476190477, + "recall": 0.5, + "f1-score": 0.4473684210526316, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.655328798185941, + "recall": 0.8095238095238095, + "f1-score": 0.7243107769423559, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/010/retfound/metrics_test.csv b/results/downsample/papila/010/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..9129a7c6cc1fcfd4ce78d2bb9e7983e64b5474fa --- /dev/null +++ b/results/downsample/papila/010/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5998874083161354,0.8095238095238095,0.4473684210526316,0.6401654411764706,0.19047619047619047,0.40476190476190477,0.40476190476190477,0.5,0.6181297948363298,0.0 diff --git a/results/downsample/papila/010/retfound/metrics_val.csv b/results/downsample/papila/010/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..6c4597d5448e2dd6d8a80f2f1009b07a4048570d --- /dev/null +++ b/results/downsample/papila/010/retfound/metrics_val.csv @@ -0,0 +1,81 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6928114096323649,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6928114096323649,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6928114096323649,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6804660360018412,0.7619047619047619,0.43243243243243246,0.596875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.5876969727887682,0.0 +0.6660102009773254,0.7619047619047619,0.43243243243243246,0.60234375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6107712174440005,0.0 +0.6497748494148254,0.7619047619047619,0.43243243243243246,0.6203125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6258483758712824,0.0 +0.6306532025337219,0.7619047619047619,0.43243243243243246,0.7,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6705877036996402,0.0 +0.6103013654549917,0.7619047619047619,0.43243243243243246,0.75,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6961419563209436,0.0 +0.5919175843397776,0.7619047619047619,0.43243243243243246,0.78671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7273689749890876,0.0 +0.5801133910814921,0.7619047619047619,0.43243243243243246,0.8046875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7568380086462012,0.0 +0.5818617045879364,0.7619047619047619,0.43243243243243246,0.8343750000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7982303699334035,0.0 +0.5954616963863373,0.7619047619047619,0.43243243243243246,0.846875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8111826073995092,0.0 +0.6188741028308868,0.7619047619047619,0.43243243243243246,0.859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8357683789664797,0.0 +0.6452745099862417,0.7619047619047619,0.43243243243243246,0.86640625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8545672625401795,0.0 +0.654808891316255,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8627962584382279,0.0 +0.6607157389322916,0.7619047619047619,0.43243243243243246,0.878125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8694846520296735,0.0 +0.6574266105890274,0.7619047619047619,0.43243243243243246,0.8851562499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8748913612557319,0.0 +0.6501010358333588,0.7619047619047619,0.43243243243243246,0.88515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8628610897912421,0.0 +0.644453247388204,0.7619047619047619,0.43243243243243246,0.875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8452214491098277,0.0 +0.6375264326731364,0.7619047619047619,0.43243243243243246,0.86875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8319864163014576,0.0 +0.6311808278163274,0.7619047619047619,0.43243243243243246,0.8578125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8195177009256494,0.0 +0.6267130672931671,0.7619047619047619,0.43243243243243246,0.83125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7893540472476694,0.0 +0.6217698852221171,0.7619047619047619,0.43243243243243246,0.8187500000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7642784431537599,0.0 +0.6187391479810079,0.7619047619047619,0.43243243243243246,0.796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7258067258051475,0.0 +0.6173502753178278,0.7619047619047619,0.43243243243243246,0.77578125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7110281113445545,0.0 +0.617540160814921,0.7619047619047619,0.43243243243243246,0.75234375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6911278878745847,0.0 +0.6180609663327535,0.7619047619047619,0.43243243243243246,0.73203125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6803412123835506,0.0 +0.615749771396319,0.7619047619047619,0.43243243243243246,0.70703125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6635559937757773,0.0 +0.6139675428469976,0.7619047619047619,0.43243243243243246,0.68828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6449481603966469,0.0 +0.6122273902098337,0.7619047619047619,0.43243243243243246,0.6734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6276994484604349,0.0 +0.6112338453531265,0.7619047619047619,0.43243243243243246,0.6585937499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6175950468754015,0.0 +0.6096604913473129,0.7619047619047619,0.43243243243243246,0.6484375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6091894932218631,0.0 +0.6091193159421285,0.7619047619047619,0.43243243243243246,0.64765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6109832500677277,0.0 +0.608538806438446,0.7619047619047619,0.43243243243243246,0.6492187500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.612551547436162,0.0 +0.608481173714002,0.7619047619047619,0.43243243243243246,0.64921875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.616515463577388,0.0 +0.6093654930591583,0.7619047619047619,0.43243243243243246,0.65625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6200854269476727,0.0 +0.6106994599103928,0.7619047619047619,0.43243243243243246,0.6570312500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6207560761275246,0.0 +0.6117736796538035,0.7619047619047619,0.43243243243243246,0.6546875000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6153992625600984,0.0 +0.6130581200122833,0.7619047619047619,0.43243243243243246,0.66328125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6197010400269664,0.0 +0.6130038946866989,0.7619047619047619,0.43243243243243246,0.6765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6294487091687206,0.0 +0.6141506632169088,0.7619047619047619,0.43243243243243246,0.67890625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6317306926925179,0.0 +0.616402859489123,0.7619047619047619,0.43243243243243246,0.69765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6437175590074609,0.0 +0.6167629559834799,0.7619047619047619,0.43243243243243246,0.7164062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6566397840105183,0.0 +0.616546630859375,0.7619047619047619,0.43243243243243246,0.72109375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6711947019189124,0.0 +0.6165595153967539,0.7619047619047619,0.43243243243243246,0.7257812499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6673835167247582,0.0 +0.6164781202872595,0.7619047619047619,0.43243243243243246,0.7296875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6741790844993978,0.0 +0.615570068359375,0.7619047619047619,0.43243243243243246,0.7242187499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6732083665939372,0.0 +0.6138929575681686,0.7619047619047619,0.43243243243243246,0.7289062500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6749904902155639,0.0 +0.6134209682544073,0.7619047619047619,0.43243243243243246,0.73515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6785038512439961,0.0 +0.6134155293305715,0.7619047619047619,0.43243243243243246,0.74375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6851369582897243,0.0 +0.6142103572686514,0.7619047619047619,0.43243243243243246,0.7531249999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6928732871523905,0.0 +0.6154663215080897,0.7619047619047619,0.43243243243243246,0.7609374999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7021999136813196,0.0 +0.6170640885829926,0.7619047619047619,0.43243243243243246,0.7671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7033017962536947,0.0 +0.6188185115655264,0.7619047619047619,0.43243243243243246,0.7765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7098175794502224,0.0 +0.6203626940647761,0.7619047619047619,0.43243243243243246,0.78125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7210924006435218,0.0 +0.6208177357912064,0.7619047619047619,0.43243243243243246,0.78125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7096951294497242,0.0 +0.6219645192225774,0.7619047619047619,0.43243243243243246,0.7828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7164210811559466,0.0 +0.6236707915862402,0.7619047619047619,0.43243243243243246,0.78671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7162335363213113,0.0 +0.6254719942808151,0.7619047619047619,0.43243243243243246,0.80078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7247393790916119,0.0 +0.6275594085454941,0.7619047619047619,0.43243243243243246,0.7992187500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7233942224958159,0.0 +0.6289021919171015,0.7619047619047619,0.43243243243243246,0.796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7251646118109623,0.0 +0.6295213351647059,0.7619047619047619,0.43243243243243246,0.79765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7252263707042428,0.0 +0.6297505845626196,0.7619047619047619,0.43243243243243246,0.79609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7244723041186545,0.0 +0.629614939292272,0.7619047619047619,0.43243243243243246,0.7984374999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7244814415455552,0.0 +0.6296352793773016,0.7619047619047619,0.43243243243243246,0.7984375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7247076741230463,0.0 +0.6297295341889063,0.7619047619047619,0.43243243243243246,0.796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7246504442152728,0.0 +0.6298523048559824,0.7619047619047619,0.43243243243243246,0.796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7228315587663565,0.0 +0.6304741750160853,0.7619047619047619,0.43243243243243246,0.79609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7230532623583403,0.0 +0.6307494988044103,0.7619047619047619,0.43243243243243246,0.79609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7247076741230463,0.0 +0.6309142907460531,0.7619047619047619,0.43243243243243246,0.796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7246645419596338,0.0 +0.6309583882490793,0.7619047619047619,0.43243243243243246,0.8,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7237455700506481,0.0 +0.6309468746185303,0.7619047619047619,0.43243243243243246,0.7992187500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7236331002438848,0.0 +0.6309041182200114,0.7619047619047619,0.43243243243243246,0.8015625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7236966447112816,0.0 +0.6309461941321691,0.7619047619047619,0.43243243243243246,0.7992187500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7229407925515772,0.0 +0.6311170806487402,0.7619047619047619,0.43243243243243246,0.7999999999999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7229631934360345,0.0 +0.6313354323307673,0.7619047619047619,0.43243243243243246,0.80078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.724450093707937,0.0 +0.6313374688227972,0.7619047619047619,0.43243243243243246,0.8,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7236331002438848,0.0 +0.6314317335685095,0.7619047619047619,0.43243243243243246,0.8,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7243991258849105,0.0 +0.6314988782008489,0.7619047619047619,0.43243243243243246,0.80078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7237052417030434,0.0 +0.6315443267424902,0.7619047619047619,0.43243243243243246,0.8023437499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7251635750363767,0.0 diff --git a/results/downsample/papila/010/retfound/pr.png b/results/downsample/papila/010/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..b27df6e61687ebd885c58ef028a0feed9403df7b --- /dev/null +++ b/results/downsample/papila/010/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:235dd9fbb1bb4a78cee89902d5b72669998134c0e20792a46befebe2bf46d4da +size 53031 diff --git a/results/downsample/papila/010/retfound/roc.png b/results/downsample/papila/010/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..3e28e44e2f554fb505d1bca30b41b0c1ad1243af --- /dev/null +++ b/results/downsample/papila/010/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7eaf1b03be670f1a2264f0b64a75247082f3a0bcedd2e694905e8d27349325b7 +size 61280 diff --git a/results/downsample/papila/010/retfound/test_pred.npz b/results/downsample/papila/010/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..4d85ff5c4a14203fac2aefe409e1cd8356771f00 --- /dev/null +++ b/results/downsample/papila/010/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:623c873b9f98ddee436bdff456d53947d9e04bb6086631293de5884874906566 +size 1518 diff --git a/results/downsample/papila/010/retfound/train.log b/results/downsample/papila/010/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1ac9c2078d3fdd4929c5d2c814c97a5f51fd059e --- /dev/null +++ b/results/downsample/papila/010/retfound/train.log @@ -0,0 +1,1043 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:56:57.812757929 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:56:58.297687] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:56:58.298024] Namespace(batch_size=15, +epochs=80, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/papila_10', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/010', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:57:01.227355] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:57:02.816410] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:57:02.957116] Sampler_train = +[14:57:03.000378] len of train_set: 15 +[14:57:03.306106] [Adaptation] Full fine-tuning: training all parameters. +[14:57:03.307144] number of trainable params (M): 303.30 +[14:57:03.307220] base lr: 5.00e-03 +[14:57:03.307272] actual lr: 2.93e-04 +[14:57:03.307313] accumulate grad iterations: 1 +[14:57:03.307356] effective batch size: 15 +[14:57:03.310225] criterion = CrossEntropyLoss() +[14:57:03.310296] Start training for 80 epochs +[14:57:03.312369] log_dir: ./output_logs/retfound +[14:57:04.932004] Epoch: [0] [0/1] eta: 0:00:01 lr: 0.000000 loss: 0.6928 (0.6928) time: 1.6187 data: 1.1638 max mem: 4514 +[14:57:04.990111] Epoch: [0] Total time: 0:00:01 (1.6776 s / it) +[14:57:04.991049] Averaged stats: lr: 0.000000 loss: 0.6928 (0.6928) +[14:57:06.390806] val: [0/3] eta: 0:00:04 loss: 0.6924 (0.6924) time: 1.3765 data: 1.3543 max mem: 4514 +[14:57:06.431529] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.4722 data: 0.4515 max mem: 4514 +[14:57:06.492843] val: Total time: 0:00:01 (0.4930 s / it) +[14:57:06.503832] val loss: 0.6928114096323649 +[14:57:06.504019] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:57:08.047635] Best epoch = 0, Best score = 0.2994 +[14:57:08.115111] log_dir: ./output_logs/retfound +[14:57:09.453545] Epoch: [1] [0/1] eta: 0:00:01 lr: 0.000029 loss: 0.6925 (0.6925) time: 1.3376 data: 1.2413 max mem: 4514 +[14:57:09.518145] Epoch: [1] Total time: 0:00:01 (1.4029 s / it) +[14:57:09.519060] Averaged stats: lr: 0.000029 loss: 0.6925 (0.6925) +[14:57:11.058364] val: [0/3] eta: 0:00:04 loss: 0.6924 (0.6924) time: 1.5169 data: 1.5031 max mem: 4514 +[14:57:11.076591] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.5115 data: 0.5011 max mem: 4514 +[14:57:11.142968] val: Total time: 0:00:01 (0.5340 s / it) +[14:57:11.152074] val loss: 0.6928114096323649 +[14:57:11.152243] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:57:11.183767] Best epoch = 0, Best score = 0.2994 +[14:57:11.451429] log_dir: ./output_logs/retfound +[14:57:12.645720] Epoch: [2] [0/1] eta: 0:00:01 lr: 0.000059 loss: 0.6927 (0.6927) time: 1.1935 data: 1.1525 max mem: 4514 +[14:57:12.712072] Epoch: [2] Total time: 0:00:01 (1.2605 s / it) +[14:57:12.712881] Averaged stats: lr: 0.000059 loss: 0.6927 (0.6927) +[14:57:14.127517] val: [0/3] eta: 0:00:04 loss: 0.6924 (0.6924) time: 1.3884 data: 1.3760 max mem: 4514 +[14:57:14.145567] val: [2/3] eta: 0:00:00 loss: 0.6924 (0.6928) time: 0.4686 data: 0.4588 max mem: 4514 +[14:57:14.212206] val: Total time: 0:00:01 (0.4912 s / it) +[14:57:14.221076] val loss: 0.6928114096323649 +[14:57:14.221250] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:57:14.251431] Best epoch = 0, Best score = 0.2994 +[14:57:14.523323] log_dir: ./output_logs/retfound +[14:57:15.709872] Epoch: [3] [0/1] eta: 0:00:01 lr: 0.000088 loss: 0.6929 (0.6929) time: 1.1858 data: 1.0949 max mem: 4752 +[14:57:15.772342] Epoch: [3] Total time: 0:00:01 (1.2489 s / it) +[14:57:15.773154] Averaged stats: lr: 0.000088 loss: 0.6929 (0.6929) +[14:57:17.276545] val: [0/3] eta: 0:00:04 loss: 0.6644 (0.6644) time: 1.4962 data: 1.4779 max mem: 4752 +[14:57:17.294650] val: [2/3] eta: 0:00:00 loss: 0.6644 (0.6805) time: 0.5046 data: 0.4927 max mem: 4752 +[14:57:17.363660] val: Total time: 0:00:01 (0.5280 s / it) +[14:57:17.372777] val loss: 0.6804660360018412 +[14:57:17.373055] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.5969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.5877, Kappa: 0.0000, Score: 0.3431 +[14:57:18.958536] Best epoch = 3, Best score = 0.3431 +[14:57:19.020950] log_dir: ./output_logs/retfound +[14:57:20.198243] Epoch: [4] [0/1] eta: 0:00:01 lr: 0.000117 loss: 0.6711 (0.6711) time: 1.1764 data: 1.1088 max mem: 6809 +[14:57:20.265670] Epoch: [4] Total time: 0:00:01 (1.2446 s / it) +[14:57:20.266552] Averaged stats: lr: 0.000117 loss: 0.6711 (0.6711) +[14:57:21.653972] val: [0/3] eta: 0:00:04 loss: 0.6301 (0.6301) time: 1.3648 data: 1.3517 max mem: 6809 +[14:57:21.718696] val: [2/3] eta: 0:00:00 loss: 0.6301 (0.6660) time: 0.4763 data: 0.4661 max mem: 6809 +[14:57:21.789255] val: Total time: 0:00:01 (0.5002 s / it) +[14:57:21.798576] val loss: 0.6660102009773254 +[14:57:21.798821] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6023, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6108, Kappa: 0.0000, Score: 0.3449 +[14:57:23.330333] Best epoch = 4, Best score = 0.3449 +[14:57:23.414869] log_dir: ./output_logs/retfound +[14:57:24.711091] Epoch: [5] [0/1] eta: 0:00:01 lr: 0.000146 loss: 0.6558 (0.6558) time: 1.2953 data: 1.2427 max mem: 6809 +[14:57:24.781394] Epoch: [5] Total time: 0:00:01 (1.3664 s / it) +[14:57:24.782221] Averaged stats: lr: 0.000146 loss: 0.6558 (0.6558) +[14:57:26.173613] val: [0/3] eta: 0:00:04 loss: 0.5895 (0.5895) time: 1.3805 data: 1.3663 max mem: 6809 +[14:57:26.192028] val: [2/3] eta: 0:00:00 loss: 0.5895 (0.6498) time: 0.4661 data: 0.4555 max mem: 6809 +[14:57:26.263535] val: Total time: 0:00:01 (0.4904 s / it) +[14:57:26.272812] val loss: 0.6497748494148254 +[14:57:26.273051] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6203, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6258, Kappa: 0.0000, Score: 0.3509 +[14:57:27.836750] Best epoch = 5, Best score = 0.3509 +[14:57:27.904999] log_dir: ./output_logs/retfound +[14:57:29.080331] Epoch: [6] [0/1] eta: 0:00:01 lr: 0.000176 loss: 0.6249 (0.6249) time: 1.1744 data: 1.1215 max mem: 6809 +[14:57:29.144561] Epoch: [6] Total time: 0:00:01 (1.2394 s / it) +[14:57:29.145378] Averaged stats: lr: 0.000176 loss: 0.6249 (0.6249) +[14:57:30.614910] val: [0/3] eta: 0:00:04 loss: 0.5377 (0.5377) time: 1.4583 data: 1.4433 max mem: 6809 +[14:57:30.633691] val: [2/3] eta: 0:00:00 loss: 0.5380 (0.6307) time: 0.4922 data: 0.4812 max mem: 6809 +[14:57:30.709534] val: Total time: 0:00:01 (0.5179 s / it) +[14:57:30.718283] val loss: 0.6306532025337219 +[14:57:30.718449] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7000, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6706, Kappa: 0.0000, Score: 0.3775 +[14:57:32.259480] Best epoch = 6, Best score = 0.3775 +[14:57:32.322782] log_dir: ./output_logs/retfound +[14:57:33.665714] Epoch: [7] [0/1] eta: 0:00:01 lr: 0.000205 loss: 0.5596 (0.5596) time: 1.3420 data: 1.2896 max mem: 6809 +[14:57:33.734078] Epoch: [7] Total time: 0:00:01 (1.4111 s / it) +[14:57:33.734963] Averaged stats: lr: 0.000205 loss: 0.5596 (0.5596) +[14:57:35.134908] val: [0/3] eta: 0:00:04 loss: 0.4751 (0.4751) time: 1.3886 data: 1.3757 max mem: 6809 +[14:57:35.214125] val: [2/3] eta: 0:00:00 loss: 0.4753 (0.6103) time: 0.4891 data: 0.4789 max mem: 6809 +[14:57:35.281809] val: Total time: 0:00:01 (0.5121 s / it) +[14:57:35.290645] val loss: 0.6103013654549917 +[14:57:35.290885] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6961, Kappa: 0.0000, Score: 0.3941 +[14:57:36.820367] Best epoch = 7, Best score = 0.3941 +[14:57:36.894626] log_dir: ./output_logs/retfound +[14:57:38.271551] Epoch: [8] [0/1] eta: 0:00:01 lr: 0.000234 loss: 0.5167 (0.5167) time: 1.3760 data: 1.3240 max mem: 6809 +[14:57:38.341761] Epoch: [8] Total time: 0:00:01 (1.4470 s / it) +[14:57:38.342572] Averaged stats: lr: 0.000234 loss: 0.5167 (0.5167) +[14:57:39.821852] val: [0/3] eta: 0:00:04 loss: 0.4036 (0.4036) time: 1.4558 data: 1.4429 max mem: 6809 +[14:57:39.840173] val: [2/3] eta: 0:00:00 loss: 0.4036 (0.5919) time: 0.4912 data: 0.4810 max mem: 6809 +[14:57:39.906388] val: Total time: 0:00:01 (0.5137 s / it) +[14:57:39.915135] val loss: 0.5919175843397776 +[14:57:39.915392] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7867, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7274, Kappa: 0.0000, Score: 0.4064 +[14:57:41.442406] Best epoch = 8, Best score = 0.4064 +[14:57:41.505983] log_dir: ./output_logs/retfound +[14:57:42.836818] Epoch: [9] [0/1] eta: 0:00:01 lr: 0.000264 loss: 0.4764 (0.4764) time: 1.3300 data: 1.2772 max mem: 6809 +[14:57:42.904110] Epoch: [9] Total time: 0:00:01 (1.3980 s / it) +[14:57:42.904955] Averaged stats: lr: 0.000264 loss: 0.4764 (0.4764) +[14:57:44.374832] val: [0/3] eta: 0:00:04 loss: 0.3249 (0.3249) time: 1.4581 data: 1.4450 max mem: 6809 +[14:57:44.393112] val: [2/3] eta: 0:00:00 loss: 0.3249 (0.5801) time: 0.4920 data: 0.4818 max mem: 6809 +[14:57:44.468791] val: Total time: 0:00:01 (0.5176 s / it) +[14:57:44.477923] val loss: 0.5801133910814921 +[14:57:44.478122] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8047, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7568, Kappa: 0.0000, Score: 0.4124 +[14:57:46.030139] Best epoch = 9, Best score = 0.4124 +[14:57:46.100817] log_dir: ./output_logs/retfound +[14:57:47.225954] Epoch: [10] [0/1] eta: 0:00:01 lr: 0.000293 loss: 0.5114 (0.5114) time: 1.1243 data: 1.0711 max mem: 6809 +[14:57:47.291387] Epoch: [10] Total time: 0:00:01 (1.1904 s / it) +[14:57:47.292140] Averaged stats: lr: 0.000293 loss: 0.5114 (0.5114) +[14:57:48.722333] val: [0/3] eta: 0:00:04 loss: 0.2561 (0.2561) time: 1.4194 data: 1.4067 max mem: 6809 +[14:57:48.740164] val: [2/3] eta: 0:00:00 loss: 0.2561 (0.5819) time: 0.4789 data: 0.4690 max mem: 6809 +[14:57:48.815508] val: Total time: 0:00:01 (0.5044 s / it) +[14:57:48.824206] val loss: 0.5818617045879364 +[14:57:48.824389] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8344, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7982, Kappa: 0.0000, Score: 0.4223 +[14:57:50.454103] Best epoch = 10, Best score = 0.4223 +[14:57:50.550021] log_dir: ./output_logs/retfound +[14:57:51.804670] Epoch: [11] [0/1] eta: 0:00:01 lr: 0.000293 loss: 0.5041 (0.5041) time: 1.2538 data: 1.2003 max mem: 6809 +[14:57:51.873136] Epoch: [11] Total time: 0:00:01 (1.3230 s / it) +[14:57:51.873986] Averaged stats: lr: 0.000293 loss: 0.5041 (0.5041) +[14:57:53.393050] val: [0/3] eta: 0:00:04 loss: 0.2042 (0.2042) time: 1.5077 data: 1.4917 max mem: 6809 +[14:57:53.452923] val: [2/3] eta: 0:00:00 loss: 0.2042 (0.5955) time: 0.5224 data: 0.5110 max mem: 6809 +[14:57:53.526284] val: Total time: 0:00:01 (0.5472 s / it) +[14:57:53.535156] val loss: 0.5954616963863373 +[14:57:53.535346] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8469, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8112, Kappa: 0.0000, Score: 0.4264 +[14:57:55.300464] Best epoch = 11, Best score = 0.4264 +[14:57:55.363715] log_dir: ./output_logs/retfound +[14:57:56.581720] Epoch: [12] [0/1] eta: 0:00:01 lr: 0.000292 loss: 0.2924 (0.2924) time: 1.2172 data: 1.1651 max mem: 6809 +[14:57:56.645840] Epoch: [12] Total time: 0:00:01 (1.2820 s / it) +[14:57:56.646642] Averaged stats: lr: 0.000292 loss: 0.2924 (0.2924) +[14:57:58.129611] val: [0/3] eta: 0:00:04 loss: 0.1624 (0.1624) time: 1.4599 data: 1.4468 max mem: 6809 +[14:57:58.147901] val: [2/3] eta: 0:00:00 loss: 0.1624 (0.6189) time: 0.4926 data: 0.4824 max mem: 6809 +[14:57:58.217527] val: Total time: 0:00:01 (0.5162 s / it) +[14:57:58.226654] val loss: 0.6188741028308868 +[14:57:58.226849] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8594, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8358, Kappa: 0.0000, Score: 0.4306 +[14:57:59.914467] Best epoch = 12, Best score = 0.4306 +[14:57:59.981878] log_dir: ./output_logs/retfound +[14:58:01.275271] Epoch: [13] [0/1] eta: 0:00:01 lr: 0.000292 loss: 0.5396 (0.5396) time: 1.2924 data: 1.2375 max mem: 6809 +[14:58:01.351944] Epoch: [13] Total time: 0:00:01 (1.3699 s / it) +[14:58:01.352718] Averaged stats: lr: 0.000292 loss: 0.5396 (0.5396) +[14:58:02.768033] val: [0/3] eta: 0:00:04 loss: 0.1335 (0.1335) time: 1.4040 data: 1.3900 max mem: 6809 +[14:58:02.786329] val: [2/3] eta: 0:00:00 loss: 0.1335 (0.6453) time: 0.4739 data: 0.4634 max mem: 6809 +[14:58:02.859622] val: Total time: 0:00:01 (0.4988 s / it) +[14:58:02.868869] val loss: 0.6452745099862417 +[14:58:02.869155] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8664, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8546, Kappa: 0.0000, Score: 0.4329 +[14:58:04.444625] Best epoch = 13, Best score = 0.4329 +[14:58:04.522943] log_dir: ./output_logs/retfound +[14:58:05.634803] Epoch: [14] [0/1] eta: 0:00:01 lr: 0.000291 loss: 0.8084 (0.8084) time: 1.1108 data: 1.0459 max mem: 6809 +[14:58:05.704091] Epoch: [14] Total time: 0:00:01 (1.1810 s / it) +[14:58:05.705094] Averaged stats: lr: 0.000291 loss: 0.8084 (0.8084) +[14:58:07.296210] val: [0/3] eta: 0:00:04 loss: 0.1259 (0.1259) time: 1.5788 data: 1.5652 max mem: 6809 +[14:58:07.314223] val: [2/3] eta: 0:00:00 loss: 0.1259 (0.6548) time: 0.5321 data: 0.5218 max mem: 6809 +[14:58:07.383751] val: Total time: 0:00:01 (0.5556 s / it) +[14:58:07.393218] val loss: 0.654808891316255 +[14:58:07.393420] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8719, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8628, Kappa: 0.0000, Score: 0.4348 +[14:58:09.034242] Best epoch = 14, Best score = 0.4348 +[14:58:09.096329] log_dir: ./output_logs/retfound +[14:58:10.374895] Epoch: [15] [0/1] eta: 0:00:01 lr: 0.000289 loss: 0.5721 (0.5721) time: 1.2777 data: 1.2248 max mem: 6809 +[14:58:10.439908] Epoch: [15] Total time: 0:00:01 (1.3434 s / it) +[14:58:10.441558] Averaged stats: lr: 0.000289 loss: 0.5721 (0.5721) +[14:58:11.847458] val: [0/3] eta: 0:00:04 loss: 0.1221 (0.1221) time: 1.3939 data: 1.3791 max mem: 6809 +[14:58:11.917431] val: [2/3] eta: 0:00:00 loss: 0.1221 (0.6607) time: 0.4877 data: 0.4758 max mem: 6809 +[14:58:12.008378] val: Total time: 0:00:01 (0.5185 s / it) +[14:58:12.017715] val loss: 0.6607157389322916 +[14:58:12.017982] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8781, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8695, Kappa: 0.0000, Score: 0.4369 +[14:58:13.688745] Best epoch = 15, Best score = 0.4369 +[14:58:13.753069] log_dir: ./output_logs/retfound +[14:58:14.983450] Epoch: [16] [0/1] eta: 0:00:01 lr: 0.000288 loss: 0.7302 (0.7302) time: 1.2294 data: 1.1761 max mem: 6809 +[14:58:15.053290] Epoch: [16] Total time: 0:00:01 (1.3001 s / it) +[14:58:15.054131] Averaged stats: lr: 0.000288 loss: 0.7302 (0.7302) +[14:58:16.471226] val: [0/3] eta: 0:00:04 loss: 0.1265 (0.1265) time: 1.4058 data: 1.3913 max mem: 6809 +[14:58:16.491534] val: [2/3] eta: 0:00:00 loss: 0.1265 (0.6574) time: 0.4752 data: 0.4643 max mem: 6809 +[14:58:16.560970] val: Total time: 0:00:01 (0.4989 s / it) +[14:58:16.570118] val loss: 0.6574266105890274 +[14:58:16.570410] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8852, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8749, Kappa: 0.0000, Score: 0.4392 +[14:58:18.161138] Best epoch = 16, Best score = 0.4392 +[14:58:18.238248] log_dir: ./output_logs/retfound +[14:58:19.497393] Epoch: [17] [0/1] eta: 0:00:01 lr: 0.000286 loss: 0.6626 (0.6626) time: 1.2581 data: 1.2031 max mem: 6809 +[14:58:19.571471] Epoch: [17] Total time: 0:00:01 (1.3331 s / it) +[14:58:19.572351] Averaged stats: lr: 0.000286 loss: 0.6626 (0.6626) +[14:58:21.148092] val: [0/3] eta: 0:00:04 loss: 0.1341 (0.1341) time: 1.5517 data: 1.5328 max mem: 6809 +[14:58:21.167676] val: [2/3] eta: 0:00:00 loss: 0.1341 (0.6501) time: 0.5235 data: 0.5110 max mem: 6809 +[14:58:21.270354] val: Total time: 0:00:01 (0.5582 s / it) +[14:58:21.279067] val loss: 0.6501010358333588 +[14:58:21.279303] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8852, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8629, Kappa: 0.0000, Score: 0.4392 +[14:58:22.978254] Best epoch = 17, Best score = 0.4392 +[14:58:23.040182] log_dir: ./output_logs/retfound +[14:58:24.298182] Epoch: [18] [0/1] eta: 0:00:01 lr: 0.000284 loss: 0.4049 (0.4049) time: 1.2571 data: 1.2037 max mem: 6809 +[14:58:24.370130] Epoch: [18] Total time: 0:00:01 (1.3298 s / it) +[14:58:24.370956] Averaged stats: lr: 0.000284 loss: 0.4049 (0.4049) +[14:58:25.788440] val: [0/3] eta: 0:00:04 loss: 0.1410 (0.1410) time: 1.4065 data: 1.3841 max mem: 6809 +[14:58:25.869933] val: [2/3] eta: 0:00:00 loss: 0.1410 (0.6445) time: 0.4958 data: 0.4818 max mem: 6809 +[14:58:25.945094] val: Total time: 0:00:01 (0.5214 s / it) +[14:58:25.954782] val loss: 0.644453247388204 +[14:58:25.955058] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8750, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8452, Kappa: 0.0000, Score: 0.4358 +[14:58:25.991001] Best epoch = 17, Best score = 0.4392 +[14:58:26.250557] log_dir: ./output_logs/retfound +[14:58:27.489890] Epoch: [19] [0/1] eta: 0:00:01 lr: 0.000281 loss: 0.5519 (0.5519) time: 1.2382 data: 1.1745 max mem: 6809 +[14:58:27.557940] Epoch: [19] Total time: 0:00:01 (1.3072 s / it) +[14:58:27.558905] Averaged stats: lr: 0.000281 loss: 0.5519 (0.5519) +[14:58:29.006219] val: [0/3] eta: 0:00:04 loss: 0.1492 (0.1492) time: 1.4343 data: 1.4211 max mem: 6809 +[14:58:29.052497] val: [2/3] eta: 0:00:00 loss: 0.1492 (0.6375) time: 0.4933 data: 0.4830 max mem: 6809 +[14:58:29.122729] val: Total time: 0:00:01 (0.5172 s / it) +[14:58:29.131415] val loss: 0.6375264326731364 +[14:58:29.131652] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8688, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8320, Kappa: 0.0000, Score: 0.4337 +[14:58:29.177615] Best epoch = 17, Best score = 0.4392 +[14:58:29.435661] log_dir: ./output_logs/retfound +[14:58:30.619906] Epoch: [20] [0/1] eta: 0:00:01 lr: 0.000279 loss: 0.5498 (0.5498) time: 1.1833 data: 1.1310 max mem: 6809 +[14:58:30.690454] Epoch: [20] Total time: 0:00:01 (1.2546 s / it) +[14:58:30.691213] Averaged stats: lr: 0.000279 loss: 0.5498 (0.5498) +[14:58:32.229315] val: [0/3] eta: 0:00:04 loss: 0.1583 (0.1583) time: 1.5263 data: 1.5128 max mem: 6809 +[14:58:32.247447] val: [2/3] eta: 0:00:00 loss: 0.1583 (0.6312) time: 0.5146 data: 0.5043 max mem: 6809 +[14:58:32.320945] val: Total time: 0:00:01 (0.5395 s / it) +[14:58:32.329533] val loss: 0.6311808278163274 +[14:58:32.329765] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8578, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8195, Kappa: 0.0000, Score: 0.4301 +[14:58:32.364526] Best epoch = 17, Best score = 0.4392 +[14:58:32.630733] log_dir: ./output_logs/retfound +[14:58:33.930370] Epoch: [21] [0/1] eta: 0:00:01 lr: 0.000276 loss: 0.3790 (0.3790) time: 1.2987 data: 1.2458 max mem: 6809 +[14:58:33.998388] Epoch: [21] Total time: 0:00:01 (1.3675 s / it) +[14:58:33.999208] Averaged stats: lr: 0.000276 loss: 0.3790 (0.3790) +[14:58:35.545751] val: [0/3] eta: 0:00:04 loss: 0.1660 (0.1660) time: 1.5355 data: 1.5210 max mem: 6809 +[14:58:35.565222] val: [2/3] eta: 0:00:00 loss: 0.1660 (0.6267) time: 0.5181 data: 0.5071 max mem: 6809 +[14:58:35.719459] val: Total time: 0:00:01 (0.5701 s / it) +[14:58:35.728654] val loss: 0.6267130672931671 +[14:58:35.728906] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8313, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7894, Kappa: 0.0000, Score: 0.4212 +[14:58:35.856600] Best epoch = 17, Best score = 0.4392 +[14:58:36.135547] log_dir: ./output_logs/retfound +[14:58:37.413605] Epoch: [22] [0/1] eta: 0:00:01 lr: 0.000272 loss: 0.4811 (0.4811) time: 1.2770 data: 1.2246 max mem: 6809 +[14:58:37.479129] Epoch: [22] Total time: 0:00:01 (1.3434 s / it) +[14:58:37.479973] Averaged stats: lr: 0.000272 loss: 0.4811 (0.4811) +[14:58:38.882319] val: [0/3] eta: 0:00:04 loss: 0.1747 (0.1747) time: 1.3910 data: 1.3775 max mem: 6809 +[14:58:38.901031] val: [2/3] eta: 0:00:00 loss: 0.1747 (0.6218) time: 0.4697 data: 0.4592 max mem: 6809 +[14:58:38.967452] val: Total time: 0:00:01 (0.4923 s / it) +[14:58:38.976449] val loss: 0.6217698852221171 +[14:58:38.976688] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8188, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7643, Kappa: 0.0000, Score: 0.4171 +[14:58:39.010323] Best epoch = 17, Best score = 0.4392 +[14:58:39.271265] log_dir: ./output_logs/retfound +[14:58:40.573909] Epoch: [23] [0/1] eta: 0:00:01 lr: 0.000269 loss: 0.3959 (0.3959) time: 1.3017 data: 1.2496 max mem: 6809 +[14:58:40.644947] Epoch: [23] Total time: 0:00:01 (1.3735 s / it) +[14:58:40.645729] Averaged stats: lr: 0.000269 loss: 0.3959 (0.3959) +[14:58:42.125595] val: [0/3] eta: 0:00:04 loss: 0.1810 (0.1810) time: 1.4687 data: 1.4542 max mem: 6809 +[14:58:42.143669] val: [2/3] eta: 0:00:00 loss: 0.1810 (0.6187) time: 0.4954 data: 0.4848 max mem: 6809 +[14:58:42.211472] val: Total time: 0:00:01 (0.5184 s / it) +[14:58:42.220275] val loss: 0.6187391479810079 +[14:58:42.220459] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7258, Kappa: 0.0000, Score: 0.4098 +[14:58:42.259472] Best epoch = 17, Best score = 0.4392 +[14:58:42.519477] log_dir: ./output_logs/retfound +[14:58:43.831631] Epoch: [24] [0/1] eta: 0:00:01 lr: 0.000265 loss: 0.4099 (0.4099) time: 1.3113 data: 1.2587 max mem: 6809 +[14:58:43.895219] Epoch: [24] Total time: 0:00:01 (1.3756 s / it) +[14:58:43.896059] Averaged stats: lr: 0.000265 loss: 0.4099 (0.4099) +[14:58:45.282586] val: [0/3] eta: 0:00:04 loss: 0.1848 (0.1848) time: 1.3727 data: 1.3577 max mem: 6809 +[14:58:45.300809] val: [2/3] eta: 0:00:00 loss: 0.1848 (0.6174) time: 0.4635 data: 0.4526 max mem: 6809 +[14:58:45.364092] val: Total time: 0:00:01 (0.4850 s / it) +[14:58:45.372709] val loss: 0.6173502753178278 +[14:58:45.372945] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7758, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7110, Kappa: 0.0000, Score: 0.4027 +[14:58:45.408668] Best epoch = 17, Best score = 0.4392 +[14:58:45.948829] log_dir: ./output_logs/retfound +[14:58:47.207020] Epoch: [25] [0/1] eta: 0:00:01 lr: 0.000261 loss: 0.3749 (0.3749) time: 1.2571 data: 1.1954 max mem: 6809 +[14:58:47.278749] Epoch: [25] Total time: 0:00:01 (1.3297 s / it) +[14:58:47.279590] Averaged stats: lr: 0.000261 loss: 0.3749 (0.3749) +[14:58:48.729086] val: [0/3] eta: 0:00:04 loss: 0.1862 (0.1862) time: 1.4376 data: 1.4235 max mem: 6809 +[14:58:48.747166] val: [2/3] eta: 0:00:00 loss: 0.1862 (0.6175) time: 0.4851 data: 0.4746 max mem: 6809 +[14:58:48.818652] val: Total time: 0:00:01 (0.5093 s / it) +[14:58:48.827300] val loss: 0.617540160814921 +[14:58:48.827537] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7523, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6911, Kappa: 0.0000, Score: 0.3949 +[14:58:48.860815] Best epoch = 17, Best score = 0.4392 +[14:58:49.127068] log_dir: ./output_logs/retfound +[14:58:50.314930] Epoch: [26] [0/1] eta: 0:00:01 lr: 0.000257 loss: 0.3925 (0.3925) time: 1.1869 data: 1.1326 max mem: 6809 +[14:58:50.383668] Epoch: [26] Total time: 0:00:01 (1.2564 s / it) +[14:58:50.384632] Averaged stats: lr: 0.000257 loss: 0.3925 (0.3925) +[14:58:51.834081] val: [0/3] eta: 0:00:04 loss: 0.1862 (0.1862) time: 1.4381 data: 1.4253 max mem: 6809 +[14:58:51.862842] val: [2/3] eta: 0:00:00 loss: 0.1862 (0.6181) time: 0.4888 data: 0.4787 max mem: 6809 +[14:58:51.931872] val: Total time: 0:00:01 (0.5122 s / it) +[14:58:51.940676] val loss: 0.6180609663327535 +[14:58:51.940900] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7320, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6803, Kappa: 0.0000, Score: 0.3882 +[14:58:51.976423] Best epoch = 17, Best score = 0.4392 +[14:58:52.265937] log_dir: ./output_logs/retfound +[14:58:53.586284] Epoch: [27] [0/1] eta: 0:00:01 lr: 0.000253 loss: 0.7295 (0.7295) time: 1.3195 data: 1.2606 max mem: 6809 +[14:58:53.653624] Epoch: [27] Total time: 0:00:01 (1.3875 s / it) +[14:58:53.654444] Averaged stats: lr: 0.000253 loss: 0.7295 (0.7295) +[14:58:55.094020] val: [0/3] eta: 0:00:04 loss: 0.1917 (0.1917) time: 1.4158 data: 1.4021 max mem: 6809 +[14:58:55.158182] val: [2/3] eta: 0:00:00 loss: 0.1917 (0.6157) time: 0.4931 data: 0.4825 max mem: 6809 +[14:58:55.227265] val: Total time: 0:00:01 (0.5166 s / it) +[14:58:55.236085] val loss: 0.615749771396319 +[14:58:55.236265] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7070, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6636, Kappa: 0.0000, Score: 0.3798 +[14:58:55.268956] Best epoch = 17, Best score = 0.4392 +[14:58:55.544266] log_dir: ./output_logs/retfound +[14:58:56.768015] Epoch: [28] [0/1] eta: 0:00:01 lr: 0.000248 loss: 0.4114 (0.4114) time: 1.2229 data: 1.1700 max mem: 6809 +[14:58:56.835137] Epoch: [28] Total time: 0:00:01 (1.2907 s / it) +[14:58:56.835892] Averaged stats: lr: 0.000248 loss: 0.4114 (0.4114) +[14:58:58.326800] val: [0/3] eta: 0:00:04 loss: 0.1973 (0.1973) time: 1.4671 data: 1.4541 max mem: 6809 +[14:58:58.344688] val: [2/3] eta: 0:00:00 loss: 0.1973 (0.6140) time: 0.4948 data: 0.4848 max mem: 6809 +[14:58:58.407446] val: Total time: 0:00:01 (0.5161 s / it) +[14:58:58.416096] val loss: 0.6139675428469976 +[14:58:58.416317] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6883, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6449, Kappa: 0.0000, Score: 0.3736 +[14:58:58.450641] Best epoch = 17, Best score = 0.4392 +[14:58:58.703663] log_dir: ./output_logs/retfound +[14:58:59.859485] Epoch: [29] [0/1] eta: 0:00:01 lr: 0.000243 loss: 0.5582 (0.5582) time: 1.1550 data: 1.1037 max mem: 6809 +[14:58:59.927914] Epoch: [29] Total time: 0:00:01 (1.2241 s / it) +[14:58:59.928633] Averaged stats: lr: 0.000243 loss: 0.5582 (0.5582) +[14:59:01.402747] val: [0/3] eta: 0:00:04 loss: 0.2027 (0.2027) time: 1.4553 data: 1.4415 max mem: 6809 +[14:59:01.420696] val: [2/3] eta: 0:00:00 loss: 0.2027 (0.6122) time: 0.4909 data: 0.4806 max mem: 6809 +[14:59:01.492332] val: Total time: 0:00:01 (0.5152 s / it) +[14:59:01.500882] val loss: 0.6122273902098337 +[14:59:01.501044] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6734, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6277, Kappa: 0.0000, Score: 0.3686 +[14:59:01.537633] Best epoch = 17, Best score = 0.4392 +[14:59:01.794811] log_dir: ./output_logs/retfound +[14:59:02.964140] Epoch: [30] [0/1] eta: 0:00:01 lr: 0.000238 loss: 0.4812 (0.4812) time: 1.1685 data: 1.1159 max mem: 6809 +[14:59:03.031270] Epoch: [30] Total time: 0:00:01 (1.2363 s / it) +[14:59:03.032216] Averaged stats: lr: 0.000238 loss: 0.4812 (0.4812) +[14:59:04.507503] val: [0/3] eta: 0:00:04 loss: 0.2052 (0.2052) time: 1.4520 data: 1.4387 max mem: 6809 +[14:59:04.552089] val: [2/3] eta: 0:00:00 loss: 0.2052 (0.6112) time: 0.4986 data: 0.4877 max mem: 6809 +[14:59:04.624673] val: Total time: 0:00:01 (0.5233 s / it) +[14:59:04.633345] val loss: 0.6112338453531265 +[14:59:04.633566] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6586, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6176, Kappa: 0.0000, Score: 0.3637 +[14:59:04.683027] Best epoch = 17, Best score = 0.4392 +[14:59:04.927688] log_dir: ./output_logs/retfound +[14:59:06.119827] Epoch: [31] [0/1] eta: 0:00:01 lr: 0.000233 loss: 0.5666 (0.5666) time: 1.1912 data: 1.1389 max mem: 6809 +[14:59:06.184814] Epoch: [31] Total time: 0:00:01 (1.2570 s / it) +[14:59:06.185595] Averaged stats: lr: 0.000233 loss: 0.5666 (0.5666) +[14:59:07.606959] val: [0/3] eta: 0:00:04 loss: 0.2096 (0.2096) time: 1.4105 data: 1.3978 max mem: 6809 +[14:59:07.637556] val: [2/3] eta: 0:00:00 loss: 0.2096 (0.6097) time: 0.4802 data: 0.4702 max mem: 6809 +[14:59:07.708919] val: Total time: 0:00:01 (0.5044 s / it) +[14:59:07.717585] val loss: 0.6096604913473129 +[14:59:07.717828] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6484, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6092, Kappa: 0.0000, Score: 0.3603 +[14:59:07.754158] Best epoch = 17, Best score = 0.4392 +[14:59:08.017940] log_dir: ./output_logs/retfound +[14:59:09.162389] Epoch: [32] [0/1] eta: 0:00:01 lr: 0.000227 loss: 0.4933 (0.4933) time: 1.1436 data: 1.0923 max mem: 6809 +[14:59:09.242090] Epoch: [32] Total time: 0:00:01 (1.2240 s / it) +[14:59:09.242734] Averaged stats: lr: 0.000227 loss: 0.4933 (0.4933) +[14:59:10.660294] val: [0/3] eta: 0:00:04 loss: 0.2115 (0.2115) time: 1.4059 data: 1.3928 max mem: 6809 +[14:59:10.731677] val: [2/3] eta: 0:00:00 loss: 0.2115 (0.6091) time: 0.4923 data: 0.4821 max mem: 6809 +[14:59:10.802550] val: Total time: 0:00:01 (0.5162 s / it) +[14:59:10.815847] val loss: 0.6091193159421285 +[14:59:10.816041] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6477, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6110, Kappa: 0.0000, Score: 0.3600 +[14:59:10.849832] Best epoch = 17, Best score = 0.4392 +[14:59:11.108115] log_dir: ./output_logs/retfound +[14:59:12.237082] Epoch: [33] [0/1] eta: 0:00:01 lr: 0.000222 loss: 0.6162 (0.6162) time: 1.1281 data: 1.0761 max mem: 6809 +[14:59:12.307800] Epoch: [33] Total time: 0:00:01 (1.1995 s / it) +[14:59:12.308622] Averaged stats: lr: 0.000222 loss: 0.6162 (0.6162) +[14:59:13.842293] val: [0/3] eta: 0:00:04 loss: 0.2133 (0.2133) time: 1.5227 data: 1.5094 max mem: 6809 +[14:59:13.861074] val: [2/3] eta: 0:00:00 loss: 0.2133 (0.6085) time: 0.5137 data: 0.5032 max mem: 6809 +[14:59:13.930261] val: Total time: 0:00:01 (0.5371 s / it) +[14:59:13.939229] val loss: 0.608538806438446 +[14:59:13.939410] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6492, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6126, Kappa: 0.0000, Score: 0.3606 +[14:59:13.988658] Best epoch = 17, Best score = 0.4392 +[14:59:14.262040] log_dir: ./output_logs/retfound +[14:59:15.368230] Epoch: [34] [0/1] eta: 0:00:01 lr: 0.000216 loss: 0.5009 (0.5009) time: 1.1054 data: 1.0536 max mem: 6809 +[14:59:15.450757] Epoch: [34] Total time: 0:00:01 (1.1886 s / it) +[14:59:15.451553] Averaged stats: lr: 0.000216 loss: 0.5009 (0.5009) +[14:59:16.898751] val: [0/3] eta: 0:00:04 loss: 0.2136 (0.2136) time: 1.4366 data: 1.4226 max mem: 6809 +[14:59:16.916868] val: [2/3] eta: 0:00:00 loss: 0.2136 (0.6085) time: 0.4847 data: 0.4743 max mem: 6809 +[14:59:16.984251] val: Total time: 0:00:01 (0.5076 s / it) +[14:59:16.992924] val loss: 0.608481173714002 +[14:59:16.993153] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6492, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6165, Kappa: 0.0000, Score: 0.3606 +[14:59:17.042588] Best epoch = 17, Best score = 0.4392 +[14:59:17.300411] log_dir: ./output_logs/retfound +[14:59:18.424239] Epoch: [35] [0/1] eta: 0:00:01 lr: 0.000210 loss: 0.4632 (0.4632) time: 1.1230 data: 1.0708 max mem: 6809 +[14:59:18.492482] Epoch: [35] Total time: 0:00:01 (1.1919 s / it) +[14:59:18.493299] Averaged stats: lr: 0.000210 loss: 0.4632 (0.4632) +[14:59:19.888203] val: [0/3] eta: 0:00:04 loss: 0.2112 (0.2112) time: 1.3843 data: 1.3704 max mem: 6809 +[14:59:19.906130] val: [2/3] eta: 0:00:00 loss: 0.2112 (0.6094) time: 0.4672 data: 0.4569 max mem: 6809 +[14:59:19.974916] val: Total time: 0:00:01 (0.4905 s / it) +[14:59:19.983665] val loss: 0.6093654930591583 +[14:59:19.983920] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6562, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6201, Kappa: 0.0000, Score: 0.3629 +[14:59:20.021708] Best epoch = 17, Best score = 0.4392 +[14:59:20.278112] log_dir: ./output_logs/retfound +[14:59:21.488909] Epoch: [36] [0/1] eta: 0:00:01 lr: 0.000204 loss: 0.4620 (0.4620) time: 1.2099 data: 1.1583 max mem: 6809 +[14:59:21.554814] Epoch: [36] Total time: 0:00:01 (1.2766 s / it) +[14:59:21.555578] Averaged stats: lr: 0.000204 loss: 0.4620 (0.4620) +[14:59:22.981855] val: [0/3] eta: 0:00:04 loss: 0.2079 (0.2079) time: 1.4151 data: 1.4024 max mem: 6809 +[14:59:22.999658] val: [2/3] eta: 0:00:00 loss: 0.2079 (0.6107) time: 0.4775 data: 0.4675 max mem: 6809 +[14:59:23.074899] val: Total time: 0:00:01 (0.5029 s / it) +[14:59:23.083567] val loss: 0.6106994599103928 +[14:59:23.083811] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6570, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6208, Kappa: 0.0000, Score: 0.3632 +[14:59:23.121214] Best epoch = 17, Best score = 0.4392 +[14:59:23.374440] log_dir: ./output_logs/retfound +[14:59:24.655579] Epoch: [37] [0/1] eta: 0:00:01 lr: 0.000198 loss: 0.5162 (0.5162) time: 1.2801 data: 1.2214 max mem: 6809 +[14:59:24.723245] Epoch: [37] Total time: 0:00:01 (1.3487 s / it) +[14:59:24.724185] Averaged stats: lr: 0.000198 loss: 0.5162 (0.5162) +[14:59:26.177308] val: [0/3] eta: 0:00:04 loss: 0.2048 (0.2048) time: 1.4408 data: 1.4278 max mem: 6809 +[14:59:26.195562] val: [2/3] eta: 0:00:00 loss: 0.2048 (0.6118) time: 0.4862 data: 0.4760 max mem: 6809 +[14:59:26.260880] val: Total time: 0:00:01 (0.5084 s / it) +[14:59:26.269664] val loss: 0.6117736796538035 +[14:59:26.269897] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6547, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6154, Kappa: 0.0000, Score: 0.3624 +[14:59:26.316486] Best epoch = 17, Best score = 0.4392 +[14:59:26.577603] log_dir: ./output_logs/retfound +[14:59:27.726593] Epoch: [38] [0/1] eta: 0:00:01 lr: 0.000192 loss: 0.5165 (0.5165) time: 1.1481 data: 1.0960 max mem: 6809 +[14:59:27.794400] Epoch: [38] Total time: 0:00:01 (1.2167 s / it) +[14:59:27.795241] Averaged stats: lr: 0.000192 loss: 0.5165 (0.5165) +[14:59:29.147983] val: [0/3] eta: 0:00:04 loss: 0.2016 (0.2016) time: 1.3416 data: 1.3277 max mem: 6809 +[14:59:29.165850] val: [2/3] eta: 0:00:00 loss: 0.2016 (0.6131) time: 0.4530 data: 0.4426 max mem: 6809 +[14:59:29.233410] val: Total time: 0:00:01 (0.4759 s / it) +[14:59:29.242065] val loss: 0.6130581200122833 +[14:59:29.242288] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6633, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6197, Kappa: 0.0000, Score: 0.3652 +[14:59:29.277350] Best epoch = 17, Best score = 0.4392 +[14:59:29.560396] log_dir: ./output_logs/retfound +[14:59:30.720388] Epoch: [39] [0/1] eta: 0:00:01 lr: 0.000186 loss: 0.6890 (0.6890) time: 1.1592 data: 1.1077 max mem: 6809 +[14:59:30.789316] Epoch: [39] Total time: 0:00:01 (1.2288 s / it) +[14:59:30.790096] Averaged stats: lr: 0.000186 loss: 0.6890 (0.6890) +[14:59:32.173888] val: [0/3] eta: 0:00:04 loss: 0.2011 (0.2011) time: 1.3609 data: 1.3474 max mem: 6809 +[14:59:32.191628] val: [2/3] eta: 0:00:00 loss: 0.2011 (0.6130) time: 0.4594 data: 0.4492 max mem: 6809 +[14:59:32.260847] val: Total time: 0:00:01 (0.4828 s / it) +[14:59:32.269523] val loss: 0.6130038946866989 +[14:59:32.269772] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6766, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6294, Kappa: 0.0000, Score: 0.3697 +[14:59:32.312726] Best epoch = 17, Best score = 0.4392 +[14:59:32.566896] log_dir: ./output_logs/retfound +[14:59:33.894623] Epoch: [40] [0/1] eta: 0:00:01 lr: 0.000179 loss: 0.4136 (0.4136) time: 1.3270 data: 1.2748 max mem: 6809 +[14:59:33.960246] Epoch: [40] Total time: 0:00:01 (1.3932 s / it) +[14:59:33.961044] Averaged stats: lr: 0.000179 loss: 0.4136 (0.4136) +[14:59:35.330956] val: [0/3] eta: 0:00:04 loss: 0.1981 (0.1981) time: 1.3593 data: 1.3465 max mem: 6809 +[14:59:35.348754] val: [2/3] eta: 0:00:00 loss: 0.1981 (0.6142) time: 0.4589 data: 0.4489 max mem: 6809 +[14:59:35.416568] val: Total time: 0:00:01 (0.4819 s / it) +[14:59:35.425546] val loss: 0.6141506632169088 +[14:59:35.425715] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6789, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6317, Kappa: 0.0000, Score: 0.3704 +[14:59:35.457196] Best epoch = 17, Best score = 0.4392 +[14:59:35.729121] log_dir: ./output_logs/retfound +[14:59:36.976584] Epoch: [41] [0/1] eta: 0:00:01 lr: 0.000173 loss: 0.3888 (0.3888) time: 1.2466 data: 1.1932 max mem: 6809 +[14:59:37.046457] Epoch: [41] Total time: 0:00:01 (1.3172 s / it) +[14:59:37.047463] Averaged stats: lr: 0.000173 loss: 0.3888 (0.3888) +[14:59:38.479439] val: [0/3] eta: 0:00:04 loss: 0.1926 (0.1926) time: 1.4095 data: 1.3964 max mem: 6809 +[14:59:38.497474] val: [2/3] eta: 0:00:00 loss: 0.1926 (0.6164) time: 0.4757 data: 0.4655 max mem: 6809 +[14:59:38.571035] val: Total time: 0:00:01 (0.5006 s / it) +[14:59:38.580199] val loss: 0.616402859489123 +[14:59:38.580426] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6437, Kappa: 0.0000, Score: 0.3767 +[14:59:38.615708] Best epoch = 17, Best score = 0.4392 +[14:59:38.879390] log_dir: ./output_logs/retfound +[14:59:40.173023] Epoch: [42] [0/1] eta: 0:00:01 lr: 0.000167 loss: 0.6778 (0.6778) time: 1.2928 data: 1.2402 max mem: 6809 +[14:59:40.241339] Epoch: [42] Total time: 0:00:01 (1.3618 s / it) +[14:59:40.242098] Averaged stats: lr: 0.000167 loss: 0.6778 (0.6778) +[14:59:41.696076] val: [0/3] eta: 0:00:04 loss: 0.1913 (0.1913) time: 1.4300 data: 1.4168 max mem: 6809 +[14:59:41.714595] val: [2/3] eta: 0:00:00 loss: 0.1913 (0.6168) time: 0.4826 data: 0.4723 max mem: 6809 +[14:59:41.785576] val: Total time: 0:00:01 (0.5067 s / it) +[14:59:41.794233] val loss: 0.6167629559834799 +[14:59:41.794431] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7164, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6566, Kappa: 0.0000, Score: 0.3829 +[14:59:41.828493] Best epoch = 17, Best score = 0.4392 +[14:59:42.097980] log_dir: ./output_logs/retfound +[14:59:43.220400] Epoch: [43] [0/1] eta: 0:00:01 lr: 0.000160 loss: 0.5455 (0.5455) time: 1.1216 data: 1.0695 max mem: 6809 +[14:59:43.292367] Epoch: [43] Total time: 0:00:01 (1.1942 s / it) +[14:59:43.293284] Averaged stats: lr: 0.000160 loss: 0.5455 (0.5455) +[14:59:44.797405] val: [0/3] eta: 0:00:04 loss: 0.1912 (0.1912) time: 1.4934 data: 1.4806 max mem: 6809 +[14:59:44.815237] val: [2/3] eta: 0:00:00 loss: 0.1912 (0.6165) time: 0.5036 data: 0.4936 max mem: 6809 +[14:59:44.883379] val: Total time: 0:00:01 (0.5267 s / it) +[14:59:44.894781] val loss: 0.616546630859375 +[14:59:44.894968] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7211, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6712, Kappa: 0.0000, Score: 0.3845 +[14:59:44.942164] Best epoch = 17, Best score = 0.4392 +[14:59:45.201959] log_dir: ./output_logs/retfound +[14:59:46.307420] Epoch: [44] [0/1] eta: 0:00:01 lr: 0.000154 loss: 0.5133 (0.5133) time: 1.1047 data: 1.0530 max mem: 6809 +[14:59:46.374286] Epoch: [44] Total time: 0:00:01 (1.1722 s / it) +[14:59:46.375078] Averaged stats: lr: 0.000154 loss: 0.5133 (0.5133) +[14:59:47.731649] val: [0/3] eta: 0:00:04 loss: 0.1904 (0.1904) time: 1.3450 data: 1.3324 max mem: 6809 +[14:59:47.749528] val: [2/3] eta: 0:00:00 loss: 0.1904 (0.6166) time: 0.4541 data: 0.4442 max mem: 6809 +[14:59:47.815892] val: Total time: 0:00:01 (0.4766 s / it) +[14:59:47.824663] val loss: 0.6165595153967539 +[14:59:47.824856] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7258, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6674, Kappa: 0.0000, Score: 0.3861 +[14:59:47.860006] Best epoch = 17, Best score = 0.4392 +[14:59:48.112762] log_dir: ./output_logs/retfound +[14:59:49.254636] Epoch: [45] [0/1] eta: 0:00:01 lr: 0.000147 loss: 0.5413 (0.5413) time: 1.1411 data: 1.0897 max mem: 6809 +[14:59:49.319287] Epoch: [45] Total time: 0:00:01 (1.2064 s / it) +[14:59:49.320020] Averaged stats: lr: 0.000147 loss: 0.5413 (0.5413) +[14:59:50.672141] val: [0/3] eta: 0:00:04 loss: 0.1900 (0.1900) time: 1.3414 data: 1.3282 max mem: 6809 +[14:59:50.715022] val: [2/3] eta: 0:00:00 loss: 0.1900 (0.6165) time: 0.4613 data: 0.4511 max mem: 6809 +[14:59:50.782976] val: Total time: 0:00:01 (0.4843 s / it) +[14:59:50.791465] val loss: 0.6164781202872595 +[14:59:50.791631] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7297, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6742, Kappa: 0.0000, Score: 0.3874 +[14:59:50.822539] Best epoch = 17, Best score = 0.4392 +[14:59:51.096435] log_dir: ./output_logs/retfound +[14:59:52.300844] Epoch: [46] [0/1] eta: 0:00:01 lr: 0.000140 loss: 0.5905 (0.5905) time: 1.2034 data: 1.1406 max mem: 6809 +[14:59:52.364724] Epoch: [46] Total time: 0:00:01 (1.2681 s / it) +[14:59:52.365704] Averaged stats: lr: 0.000140 loss: 0.5905 (0.5905) +[14:59:53.853937] val: [0/3] eta: 0:00:04 loss: 0.1910 (0.1910) time: 1.4755 data: 1.4612 max mem: 6809 +[14:59:53.871955] val: [2/3] eta: 0:00:00 loss: 0.1910 (0.6156) time: 0.4977 data: 0.4871 max mem: 6809 +[14:59:53.938860] val: Total time: 0:00:01 (0.5203 s / it) +[14:59:53.947586] val loss: 0.615570068359375 +[14:59:53.947767] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7242, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6732, Kappa: 0.0000, Score: 0.3856 +[14:59:53.991789] Best epoch = 17, Best score = 0.4392 +[14:59:54.251548] log_dir: ./output_logs/retfound +[14:59:55.507967] Epoch: [47] [0/1] eta: 0:00:01 lr: 0.000134 loss: 0.6557 (0.6557) time: 1.2556 data: 1.2042 max mem: 6809 +[14:59:55.573425] Epoch: [47] Total time: 0:00:01 (1.3218 s / it) +[14:59:55.574218] Averaged stats: lr: 0.000134 loss: 0.6557 (0.6557) +[14:59:56.895354] val: [0/3] eta: 0:00:03 loss: 0.1932 (0.1932) time: 1.3100 data: 1.2971 max mem: 6809 +[14:59:56.932564] val: [2/3] eta: 0:00:00 loss: 0.1932 (0.6139) time: 0.4489 data: 0.4387 max mem: 6809 +[14:59:57.000605] val: Total time: 0:00:01 (0.4720 s / it) +[14:59:57.009329] val loss: 0.6138929575681686 +[14:59:57.009505] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7289, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6750, Kappa: 0.0000, Score: 0.3871 +[14:59:57.053237] Best epoch = 17, Best score = 0.4392 +[14:59:57.310499] log_dir: ./output_logs/retfound +[14:59:58.474106] Epoch: [48] [0/1] eta: 0:00:01 lr: 0.000127 loss: 0.4254 (0.4254) time: 1.1628 data: 1.1116 max mem: 6809 +[14:59:58.539245] Epoch: [48] Total time: 0:00:01 (1.2286 s / it) +[14:59:58.539941] Averaged stats: lr: 0.000127 loss: 0.4254 (0.4254) +[14:59:59.908933] val: [0/3] eta: 0:00:04 loss: 0.1930 (0.1930) time: 1.3581 data: 1.3453 max mem: 6809 +[14:59:59.961714] val: [2/3] eta: 0:00:00 loss: 0.1930 (0.6134) time: 0.4701 data: 0.4601 max mem: 6809 +[15:00:00.028104] val: Total time: 0:00:01 (0.4926 s / it) +[15:00:00.036669] val loss: 0.6134209682544073 +[15:00:00.036840] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7352, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6785, Kappa: 0.0000, Score: 0.3892 +[15:00:00.071045] Best epoch = 17, Best score = 0.4392 +[15:00:00.328958] log_dir: ./output_logs/retfound +[15:00:01.442003] Epoch: [49] [0/1] eta: 0:00:01 lr: 0.000121 loss: 0.4256 (0.4256) time: 1.1122 data: 1.0599 max mem: 6809 +[15:00:01.513585] Epoch: [49] Total time: 0:00:01 (1.1845 s / it) +[15:00:01.514845] Averaged stats: lr: 0.000121 loss: 0.4256 (0.4256) +[15:00:02.984137] val: [0/3] eta: 0:00:04 loss: 0.1921 (0.1921) time: 1.4563 data: 1.4419 max mem: 6809 +[15:00:03.002570] val: [2/3] eta: 0:00:00 loss: 0.1921 (0.6134) time: 0.4914 data: 0.4807 max mem: 6809 +[15:00:03.073678] val: Total time: 0:00:01 (0.5155 s / it) +[15:00:03.082562] val loss: 0.6134155293305715 +[15:00:03.082747] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7438, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6851, Kappa: 0.0000, Score: 0.3921 +[15:00:03.125435] Best epoch = 17, Best score = 0.4392 +[15:00:03.395540] log_dir: ./output_logs/retfound +[15:00:04.629825] Epoch: [50] [0/1] eta: 0:00:01 lr: 0.000114 loss: 0.4089 (0.4089) time: 1.2335 data: 1.1815 max mem: 6809 +[15:00:04.693220] Epoch: [50] Total time: 0:00:01 (1.2975 s / it) +[15:00:04.693936] Averaged stats: lr: 0.000114 loss: 0.4089 (0.4089) +[15:00:06.060594] val: [0/3] eta: 0:00:04 loss: 0.1895 (0.1895) time: 1.3565 data: 1.3434 max mem: 6809 +[15:00:06.078658] val: [2/3] eta: 0:00:00 loss: 0.1895 (0.6142) time: 0.4580 data: 0.4479 max mem: 6809 +[15:00:06.143781] val: Total time: 0:00:01 (0.4801 s / it) +[15:00:06.152500] val loss: 0.6142103572686514 +[15:00:06.152678] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7531, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6929, Kappa: 0.0000, Score: 0.3952 +[15:00:06.186925] Best epoch = 17, Best score = 0.4392 +[15:00:06.467166] log_dir: ./output_logs/retfound +[15:00:07.733594] Epoch: [51] [0/1] eta: 0:00:01 lr: 0.000108 loss: 0.4000 (0.4000) time: 1.2657 data: 1.2144 max mem: 6809 +[15:00:07.800445] Epoch: [51] Total time: 0:00:01 (1.3331 s / it) +[15:00:07.801088] Averaged stats: lr: 0.000108 loss: 0.4000 (0.4000) +[15:00:09.215751] val: [0/3] eta: 0:00:04 loss: 0.1863 (0.1863) time: 1.3924 data: 1.3798 max mem: 6809 +[15:00:09.233574] val: [2/3] eta: 0:00:00 loss: 0.1863 (0.6155) time: 0.4699 data: 0.4600 max mem: 6809 +[15:00:09.302412] val: Total time: 0:00:01 (0.4932 s / it) +[15:00:09.312288] val loss: 0.6154663215080897 +[15:00:09.312472] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7609, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7022, Kappa: 0.0000, Score: 0.3978 +[15:00:09.358352] Best epoch = 17, Best score = 0.4392 +[15:00:09.643204] log_dir: ./output_logs/retfound +[15:00:10.971804] Epoch: [52] [0/1] eta: 0:00:01 lr: 0.000102 loss: 0.4440 (0.4440) time: 1.3277 data: 1.2751 max mem: 6809 +[15:00:11.036391] Epoch: [52] Total time: 0:00:01 (1.3930 s / it) +[15:00:11.037154] Averaged stats: lr: 0.000102 loss: 0.4440 (0.4440) +[15:00:12.426612] val: [0/3] eta: 0:00:04 loss: 0.1824 (0.1824) time: 1.3787 data: 1.3656 max mem: 6809 +[15:00:12.444462] val: [2/3] eta: 0:00:00 loss: 0.1824 (0.6171) time: 0.4654 data: 0.4553 max mem: 6809 +[15:00:12.513417] val: Total time: 0:00:01 (0.4887 s / it) +[15:00:12.522046] val loss: 0.6170640885829926 +[15:00:12.522214] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7672, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7033, Kappa: 0.0000, Score: 0.3999 +[15:00:12.562754] Best epoch = 17, Best score = 0.4392 +[15:00:12.813680] log_dir: ./output_logs/retfound +[15:00:14.049459] Epoch: [53] [0/1] eta: 0:00:01 lr: 0.000096 loss: 0.4452 (0.4452) time: 1.2349 data: 1.1823 max mem: 6809 +[15:00:14.117995] Epoch: [53] Total time: 0:00:01 (1.3042 s / it) +[15:00:14.118833] Averaged stats: lr: 0.000096 loss: 0.4452 (0.4452) +[15:00:15.534071] val: [0/3] eta: 0:00:04 loss: 0.1789 (0.1789) time: 1.4038 data: 1.3907 max mem: 6809 +[15:00:15.552151] val: [2/3] eta: 0:00:00 loss: 0.1789 (0.6188) time: 0.4738 data: 0.4636 max mem: 6809 +[15:00:15.619876] val: Total time: 0:00:01 (0.4967 s / it) +[15:00:15.628624] val loss: 0.6188185115655264 +[15:00:15.628826] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7766, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7098, Kappa: 0.0000, Score: 0.4030 +[15:00:15.661706] Best epoch = 17, Best score = 0.4392 +[15:00:15.935708] log_dir: ./output_logs/retfound +[15:00:17.056941] Epoch: [54] [0/1] eta: 0:00:01 lr: 0.000090 loss: 0.4616 (0.4616) time: 1.1204 data: 1.0690 max mem: 6809 +[15:00:17.126643] Epoch: [54] Total time: 0:00:01 (1.1908 s / it) +[15:00:17.127424] Averaged stats: lr: 0.000090 loss: 0.4616 (0.4616) +[15:00:18.462136] val: [0/3] eta: 0:00:03 loss: 0.1758 (0.1758) time: 1.3228 data: 1.3091 max mem: 6809 +[15:00:18.480080] val: [2/3] eta: 0:00:00 loss: 0.1758 (0.6204) time: 0.4468 data: 0.4364 max mem: 6809 +[15:00:18.548317] val: Total time: 0:00:01 (0.4699 s / it) +[15:00:18.556999] val loss: 0.6203626940647761 +[15:00:18.557167] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7812, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7211, Kappa: 0.0000, Score: 0.4046 +[15:00:18.592742] Best epoch = 17, Best score = 0.4392 +[15:00:18.851139] log_dir: ./output_logs/retfound +[15:00:20.091175] Epoch: [55] [0/1] eta: 0:00:01 lr: 0.000084 loss: 0.6363 (0.6363) time: 1.2392 data: 1.1873 max mem: 6809 +[15:00:20.158992] Epoch: [55] Total time: 0:00:01 (1.3077 s / it) +[15:00:20.159674] Averaged stats: lr: 0.000084 loss: 0.6363 (0.6363) +[15:00:21.530848] val: [0/3] eta: 0:00:04 loss: 0.1743 (0.1743) time: 1.3543 data: 1.3408 max mem: 6809 +[15:00:21.548789] val: [2/3] eta: 0:00:00 loss: 0.1743 (0.6208) time: 0.4572 data: 0.4470 max mem: 6809 +[15:00:21.618399] val: Total time: 0:00:01 (0.4808 s / it) +[15:00:21.627187] val loss: 0.6208177357912064 +[15:00:21.627362] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7812, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7097, Kappa: 0.0000, Score: 0.4046 +[15:00:21.666378] Best epoch = 17, Best score = 0.4392 +[15:00:21.917134] log_dir: ./output_logs/retfound +[15:00:23.080424] Epoch: [56] [0/1] eta: 0:00:01 lr: 0.000078 loss: 0.4022 (0.4022) time: 1.1625 data: 1.1102 max mem: 6809 +[15:00:23.150955] Epoch: [56] Total time: 0:00:01 (1.2337 s / it) +[15:00:23.151753] Averaged stats: lr: 0.000078 loss: 0.4022 (0.4022) +[15:00:24.621006] val: [0/3] eta: 0:00:04 loss: 0.1720 (0.1720) time: 1.4577 data: 1.4448 max mem: 6809 +[15:00:24.638912] val: [2/3] eta: 0:00:00 loss: 0.1720 (0.6220) time: 0.4917 data: 0.4817 max mem: 6809 +[15:00:24.708082] val: Total time: 0:00:01 (0.5151 s / it) +[15:00:24.716770] val loss: 0.6219645192225774 +[15:00:24.716971] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7828, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7164, Kappa: 0.0000, Score: 0.4051 +[15:00:24.759835] Best epoch = 17, Best score = 0.4392 +[15:00:25.020302] log_dir: ./output_logs/retfound +[15:00:26.206328] Epoch: [57] [0/1] eta: 0:00:01 lr: 0.000072 loss: 0.3838 (0.3838) time: 1.1852 data: 1.1325 max mem: 6809 +[15:00:26.269980] Epoch: [57] Total time: 0:00:01 (1.2495 s / it) +[15:00:26.271011] Averaged stats: lr: 0.000072 loss: 0.3838 (0.3838) +[15:00:27.629734] val: [0/3] eta: 0:00:04 loss: 0.1691 (0.1691) time: 1.3476 data: 1.3350 max mem: 6809 +[15:00:27.647607] val: [2/3] eta: 0:00:00 loss: 0.1691 (0.6237) time: 0.4550 data: 0.4451 max mem: 6809 +[15:00:27.714180] val: Total time: 0:00:01 (0.4775 s / it) +[15:00:27.722784] val loss: 0.6236707915862402 +[15:00:27.722955] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7867, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7162, Kappa: 0.0000, Score: 0.4064 +[15:00:27.755907] Best epoch = 17, Best score = 0.4392 +[15:00:28.108060] log_dir: ./output_logs/retfound +[15:00:29.310630] Epoch: [58] [0/1] eta: 0:00:01 lr: 0.000067 loss: 0.4485 (0.4485) time: 1.2018 data: 1.1498 max mem: 6809 +[15:00:29.377322] Epoch: [58] Total time: 0:00:01 (1.2691 s / it) +[15:00:29.378017] Averaged stats: lr: 0.000067 loss: 0.4485 (0.4485) +[15:00:30.910374] val: [0/3] eta: 0:00:04 loss: 0.1661 (0.1661) time: 1.5167 data: 1.5031 max mem: 6809 +[15:00:30.928202] val: [2/3] eta: 0:00:00 loss: 0.1661 (0.6255) time: 0.5113 data: 0.5011 max mem: 6809 +[15:00:31.058559] val: Total time: 0:00:01 (0.5552 s / it) +[15:00:31.067739] val loss: 0.6254719942808151 +[15:00:31.067996] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8008, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7247, Kappa: 0.0000, Score: 0.4111 +[15:00:31.104226] Best epoch = 17, Best score = 0.4392 +[15:00:31.402648] log_dir: ./output_logs/retfound +[15:00:32.615865] Epoch: [59] [0/1] eta: 0:00:01 lr: 0.000061 loss: 0.3871 (0.3871) time: 1.2124 data: 1.1604 max mem: 6809 +[15:00:32.678787] Epoch: [59] Total time: 0:00:01 (1.2760 s / it) +[15:00:32.679577] Averaged stats: lr: 0.000061 loss: 0.3871 (0.3871) +[15:00:34.123543] val: [0/3] eta: 0:00:04 loss: 0.1629 (0.1629) time: 1.4329 data: 1.4192 max mem: 6809 +[15:00:34.141475] val: [2/3] eta: 0:00:00 loss: 0.1629 (0.6276) time: 0.4834 data: 0.4732 max mem: 6809 +[15:00:34.209559] val: Total time: 0:00:01 (0.5065 s / it) +[15:00:34.218249] val loss: 0.6275594085454941 +[15:00:34.218428] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7992, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7234, Kappa: 0.0000, Score: 0.4106 +[15:00:34.254553] Best epoch = 17, Best score = 0.4392 +[15:00:34.508863] log_dir: ./output_logs/retfound +[15:00:35.763095] Epoch: [60] [0/1] eta: 0:00:01 lr: 0.000056 loss: 0.5774 (0.5774) time: 1.2534 data: 1.2014 max mem: 6809 +[15:00:35.830374] Epoch: [60] Total time: 0:00:01 (1.3214 s / it) +[15:00:35.831197] Averaged stats: lr: 0.000056 loss: 0.5774 (0.5774) +[15:00:37.290602] val: [0/3] eta: 0:00:04 loss: 0.1607 (0.1607) time: 1.4484 data: 1.4355 max mem: 6809 +[15:00:37.308480] val: [2/3] eta: 0:00:00 loss: 0.1607 (0.6289) time: 0.4886 data: 0.4786 max mem: 6809 +[15:00:37.375990] val: Total time: 0:00:01 (0.5115 s / it) +[15:00:37.384648] val loss: 0.6289021919171015 +[15:00:37.384821] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7252, Kappa: 0.0000, Score: 0.4098 +[15:00:37.420303] Best epoch = 17, Best score = 0.4392 +[15:00:37.671876] log_dir: ./output_logs/retfound +[15:00:38.953382] Epoch: [61] [0/1] eta: 0:00:01 lr: 0.000051 loss: 0.6259 (0.6259) time: 1.2807 data: 1.2288 max mem: 6809 +[15:00:39.017101] Epoch: [61] Total time: 0:00:01 (1.3451 s / it) +[15:00:39.017845] Averaged stats: lr: 0.000051 loss: 0.6259 (0.6259) +[15:00:40.428583] val: [0/3] eta: 0:00:04 loss: 0.1597 (0.1597) time: 1.3999 data: 1.3873 max mem: 6809 +[15:00:40.446388] val: [2/3] eta: 0:00:00 loss: 0.1597 (0.6295) time: 0.4724 data: 0.4625 max mem: 6809 +[15:00:40.518160] val: Total time: 0:00:01 (0.4967 s / it) +[15:00:40.526866] val loss: 0.6295213351647059 +[15:00:40.527051] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7252, Kappa: 0.0000, Score: 0.4100 +[15:00:40.558706] Best epoch = 17, Best score = 0.4392 +[15:00:40.828879] log_dir: ./output_logs/retfound +[15:00:42.029857] Epoch: [62] [0/1] eta: 0:00:01 lr: 0.000046 loss: 0.5669 (0.5669) time: 1.2002 data: 1.1473 max mem: 6809 +[15:00:42.100882] Epoch: [62] Total time: 0:00:01 (1.2719 s / it) +[15:00:42.101761] Averaged stats: lr: 0.000046 loss: 0.5669 (0.5669) +[15:00:43.460694] val: [0/3] eta: 0:00:04 loss: 0.1593 (0.1593) time: 1.3382 data: 1.3206 max mem: 6809 +[15:00:43.486171] val: [2/3] eta: 0:00:00 loss: 0.1593 (0.6298) time: 0.4543 data: 0.4403 max mem: 6809 +[15:00:43.557075] val: Total time: 0:00:01 (0.4784 s / it) +[15:00:43.571840] val loss: 0.6297505845626196 +[15:00:43.572068] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7961, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7245, Kappa: 0.0000, Score: 0.4095 +[15:00:43.628386] Best epoch = 17, Best score = 0.4392 +[15:00:43.904964] log_dir: ./output_logs/retfound +[15:00:45.062866] Epoch: [63] [0/1] eta: 0:00:01 lr: 0.000041 loss: 0.6847 (0.6847) time: 1.1571 data: 1.1052 max mem: 6809 +[15:00:45.128157] Epoch: [63] Total time: 0:00:01 (1.2231 s / it) +[15:00:45.128895] Averaged stats: lr: 0.000041 loss: 0.6847 (0.6847) +[15:00:46.544783] val: [0/3] eta: 0:00:04 loss: 0.1594 (0.1594) time: 1.3949 data: 1.3797 max mem: 6809 +[15:00:46.562876] val: [2/3] eta: 0:00:00 loss: 0.1594 (0.6296) time: 0.4708 data: 0.4600 max mem: 6809 +[15:00:46.629690] val: Total time: 0:00:01 (0.4935 s / it) +[15:00:46.638379] val loss: 0.629614939292272 +[15:00:46.638561] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7984, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7245, Kappa: 0.0000, Score: 0.4103 +[15:00:46.671756] Best epoch = 17, Best score = 0.4392 +[15:00:46.963570] log_dir: ./output_logs/retfound +[15:00:48.157928] Epoch: [64] [0/1] eta: 0:00:01 lr: 0.000037 loss: 0.3913 (0.3913) time: 1.1936 data: 1.1423 max mem: 6809 +[15:00:48.226165] Epoch: [64] Total time: 0:00:01 (1.2624 s / it) +[15:00:48.226979] Averaged stats: lr: 0.000037 loss: 0.3913 (0.3913) +[15:00:49.614462] val: [0/3] eta: 0:00:04 loss: 0.1592 (0.1592) time: 1.3761 data: 1.3628 max mem: 6809 +[15:00:49.632254] val: [2/3] eta: 0:00:00 loss: 0.1592 (0.6296) time: 0.4645 data: 0.4543 max mem: 6809 +[15:00:49.699081] val: Total time: 0:00:01 (0.4871 s / it) +[15:00:49.707861] val loss: 0.6296352793773016 +[15:00:49.708023] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7984, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7247, Kappa: 0.0000, Score: 0.4103 +[15:00:49.767002] Best epoch = 17, Best score = 0.4392 +[15:00:50.018889] log_dir: ./output_logs/retfound +[15:00:51.173189] Epoch: [65] [0/1] eta: 0:00:01 lr: 0.000033 loss: 0.4476 (0.4476) time: 1.1535 data: 1.0994 max mem: 6809 +[15:00:51.238646] Epoch: [65] Total time: 0:00:01 (1.2196 s / it) +[15:00:51.239384] Averaged stats: lr: 0.000033 loss: 0.4476 (0.4476) +[15:00:52.504852] val: [0/3] eta: 0:00:03 loss: 0.1590 (0.1590) time: 1.2444 data: 1.2317 max mem: 6809 +[15:00:52.544000] val: [2/3] eta: 0:00:00 loss: 0.1590 (0.6297) time: 0.4277 data: 0.4176 max mem: 6809 +[15:00:52.609458] val: Total time: 0:00:01 (0.4499 s / it) +[15:00:52.618109] val loss: 0.6297295341889063 +[15:00:52.618340] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7247, Kappa: 0.0000, Score: 0.4098 +[15:00:52.650183] Best epoch = 17, Best score = 0.4392 +[15:00:52.955940] log_dir: ./output_logs/retfound +[15:00:54.218871] Epoch: [66] [0/1] eta: 0:00:01 lr: 0.000029 loss: 0.3892 (0.3892) time: 1.2621 data: 1.2103 max mem: 6809 +[15:00:54.282828] Epoch: [66] Total time: 0:00:01 (1.3267 s / it) +[15:00:54.283552] Averaged stats: lr: 0.000029 loss: 0.3892 (0.3892) +[15:00:55.642206] val: [0/3] eta: 0:00:04 loss: 0.1587 (0.1587) time: 1.3473 data: 1.3334 max mem: 6809 +[15:00:55.660242] val: [2/3] eta: 0:00:00 loss: 0.1587 (0.6299) time: 0.4549 data: 0.4445 max mem: 6809 +[15:00:55.726333] val: Total time: 0:00:01 (0.4773 s / it) +[15:00:55.734991] val loss: 0.6298523048559824 +[15:00:55.735180] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7228, Kappa: 0.0000, Score: 0.4098 +[15:00:55.792851] Best epoch = 17, Best score = 0.4392 +[15:00:56.064157] log_dir: ./output_logs/retfound +[15:00:57.296495] Epoch: [67] [0/1] eta: 0:00:01 lr: 0.000025 loss: 0.3226 (0.3226) time: 1.2314 data: 1.1797 max mem: 6809 +[15:00:57.361970] Epoch: [67] Total time: 0:00:01 (1.2976 s / it) +[15:00:57.362762] Averaged stats: lr: 0.000025 loss: 0.3226 (0.3226) +[15:00:58.889354] val: [0/3] eta: 0:00:04 loss: 0.1578 (0.1578) time: 1.5014 data: 1.4890 max mem: 6809 +[15:00:58.907197] val: [2/3] eta: 0:00:00 loss: 0.1578 (0.6305) time: 0.5062 data: 0.4964 max mem: 6809 +[15:00:58.975063] val: Total time: 0:00:01 (0.5292 s / it) +[15:00:58.983957] val loss: 0.6304741750160853 +[15:00:58.984137] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7961, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7231, Kappa: 0.0000, Score: 0.4095 +[15:00:59.021110] Best epoch = 17, Best score = 0.4392 +[15:00:59.346683] log_dir: ./output_logs/retfound +[15:01:00.511139] Epoch: [68] [0/1] eta: 0:00:01 lr: 0.000022 loss: 0.6109 (0.6109) time: 1.1636 data: 1.1114 max mem: 6809 +[15:01:00.577624] Epoch: [68] Total time: 0:00:01 (1.2308 s / it) +[15:01:00.578422] Averaged stats: lr: 0.000022 loss: 0.6109 (0.6109) +[15:01:02.023560] val: [0/3] eta: 0:00:04 loss: 0.1573 (0.1573) time: 1.4339 data: 1.4215 max mem: 6809 +[15:01:02.041391] val: [2/3] eta: 0:00:00 loss: 0.1573 (0.6307) time: 0.4837 data: 0.4739 max mem: 6809 +[15:01:02.111837] val: Total time: 0:00:01 (0.5076 s / it) +[15:01:02.120678] val loss: 0.6307494988044103 +[15:01:02.120848] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7961, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7247, Kappa: 0.0000, Score: 0.4095 +[15:01:02.166748] Best epoch = 17, Best score = 0.4392 +[15:01:02.490550] log_dir: ./output_logs/retfound +[15:01:03.695022] Epoch: [69] [0/1] eta: 0:00:01 lr: 0.000018 loss: 0.5174 (0.5174) time: 1.2036 data: 1.1518 max mem: 6809 +[15:01:03.760100] Epoch: [69] Total time: 0:00:01 (1.2694 s / it) +[15:01:03.760881] Averaged stats: lr: 0.000018 loss: 0.5174 (0.5174) +[15:01:05.157443] val: [0/3] eta: 0:00:04 loss: 0.1570 (0.1570) time: 1.3855 data: 1.3716 max mem: 6809 +[15:01:05.175446] val: [2/3] eta: 0:00:00 loss: 0.1570 (0.6309) time: 0.4677 data: 0.4573 max mem: 6809 +[15:01:05.241130] val: Total time: 0:00:01 (0.4899 s / it) +[15:01:05.249789] val loss: 0.6309142907460531 +[15:01:05.249971] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7969, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7247, Kappa: 0.0000, Score: 0.4098 +[15:01:05.283822] Best epoch = 17, Best score = 0.4392 +[15:01:05.613741] log_dir: ./output_logs/retfound +[15:01:06.790842] Epoch: [70] [0/1] eta: 0:00:01 lr: 0.000015 loss: 0.5918 (0.5918) time: 1.1763 data: 1.1246 max mem: 6809 +[15:01:06.854684] Epoch: [70] Total time: 0:00:01 (1.2408 s / it) +[15:01:06.855486] Averaged stats: lr: 0.000015 loss: 0.5918 (0.5918) +[15:01:08.336254] val: [0/3] eta: 0:00:04 loss: 0.1569 (0.1569) time: 1.4695 data: 1.4557 max mem: 6809 +[15:01:08.354181] val: [2/3] eta: 0:00:00 loss: 0.1569 (0.6310) time: 0.4956 data: 0.4853 max mem: 6809 +[15:01:08.422480] val: Total time: 0:00:01 (0.5188 s / it) +[15:01:08.431195] val loss: 0.6309583882490793 +[15:01:08.431414] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8000, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7237, Kappa: 0.0000, Score: 0.4108 +[15:01:08.463875] Best epoch = 17, Best score = 0.4392 +[15:01:08.820291] log_dir: ./output_logs/retfound +[15:01:10.030422] Epoch: [71] [0/1] eta: 0:00:01 lr: 0.000013 loss: 0.4821 (0.4821) time: 1.2093 data: 1.1569 max mem: 6809 +[15:01:10.093767] Epoch: [71] Total time: 0:00:01 (1.2733 s / it) +[15:01:10.094437] Averaged stats: lr: 0.000013 loss: 0.4821 (0.4821) +[15:01:11.519895] val: [0/3] eta: 0:00:04 loss: 0.1569 (0.1569) time: 1.4138 data: 1.3950 max mem: 6809 +[15:01:11.548173] val: [2/3] eta: 0:00:00 loss: 0.1569 (0.6309) time: 0.4804 data: 0.4651 max mem: 6809 +[15:01:11.616328] val: Total time: 0:00:01 (0.5037 s / it) +[15:01:11.628447] val loss: 0.6309468746185303 +[15:01:11.628633] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7992, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7236, Kappa: 0.0000, Score: 0.4106 +[15:01:11.664599] Best epoch = 17, Best score = 0.4392 +[15:01:11.907321] log_dir: ./output_logs/retfound +[15:01:13.147702] Epoch: [72] [0/1] eta: 0:00:01 lr: 0.000010 loss: 0.5819 (0.5819) time: 1.2395 data: 1.1876 max mem: 6809 +[15:01:13.222541] Epoch: [72] Total time: 0:00:01 (1.3150 s / it) +[15:01:13.223369] Averaged stats: lr: 0.000010 loss: 0.5819 (0.5819) +[15:01:14.632555] val: [0/3] eta: 0:00:04 loss: 0.1569 (0.1569) time: 1.3861 data: 1.3730 max mem: 6809 +[15:01:14.650540] val: [2/3] eta: 0:00:00 loss: 0.1569 (0.6309) time: 0.4679 data: 0.4577 max mem: 6809 +[15:01:14.812668] val: Total time: 0:00:01 (0.5223 s / it) +[15:01:14.822065] val loss: 0.6309041182200114 +[15:01:14.822242] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8016, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7237, Kappa: 0.0000, Score: 0.4113 +[15:01:14.875146] Best epoch = 17, Best score = 0.4392 +[15:01:15.107583] log_dir: ./output_logs/retfound +[15:01:16.303823] Epoch: [73] [0/1] eta: 0:00:01 lr: 0.000008 loss: 0.2383 (0.2383) time: 1.1954 data: 1.1432 max mem: 6809 +[15:01:16.458012] Epoch: [73] Total time: 0:00:01 (1.3503 s / it) +[15:01:16.458855] Averaged stats: lr: 0.000008 loss: 0.2383 (0.2383) +[15:01:17.882330] val: [0/3] eta: 0:00:04 loss: 0.1568 (0.1568) time: 1.4123 data: 1.3988 max mem: 6809 +[15:01:17.901997] val: [2/3] eta: 0:00:00 loss: 0.1568 (0.6309) time: 0.4771 data: 0.4663 max mem: 6809 +[15:01:18.097428] val: Total time: 0:00:01 (0.5427 s / it) +[15:01:18.106777] val loss: 0.6309461941321691 +[15:01:18.106963] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7992, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7229, Kappa: 0.0000, Score: 0.4106 +[15:01:18.160113] Best epoch = 17, Best score = 0.4392 +[15:01:18.397256] log_dir: ./output_logs/retfound +[15:01:19.615147] Epoch: [74] [0/1] eta: 0:00:01 lr: 0.000006 loss: 0.2805 (0.2805) time: 1.2170 data: 1.1647 max mem: 6809 +[15:01:19.744780] Epoch: [74] Total time: 0:00:01 (1.3474 s / it) +[15:01:19.745620] Averaged stats: lr: 0.000006 loss: 0.2805 (0.2805) +[15:01:21.152249] val: [0/3] eta: 0:00:04 loss: 0.1565 (0.1565) time: 1.3957 data: 1.3827 max mem: 6809 +[15:01:21.202315] val: [2/3] eta: 0:00:00 loss: 0.1565 (0.6311) time: 0.4817 data: 0.4715 max mem: 6809 +[15:01:21.386686] val: Total time: 0:00:01 (0.5436 s / it) +[15:01:21.395388] val loss: 0.6311170806487402 +[15:01:21.395571] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8000, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7230, Kappa: 0.0000, Score: 0.4108 +[15:01:21.429352] Best epoch = 17, Best score = 0.4392 +[15:01:21.695229] log_dir: ./output_logs/retfound +[15:01:22.865426] Epoch: [75] [0/1] eta: 0:00:01 lr: 0.000005 loss: 0.4970 (0.4970) time: 1.1694 data: 1.1165 max mem: 6809 +[15:01:23.034079] Epoch: [75] Total time: 0:00:01 (1.3387 s / it) +[15:01:23.035081] Averaged stats: lr: 0.000005 loss: 0.4970 (0.4970) +[15:01:24.501765] val: [0/3] eta: 0:00:04 loss: 0.1564 (0.1564) time: 1.4547 data: 1.4411 max mem: 6809 +[15:01:24.519612] val: [2/3] eta: 0:00:00 loss: 0.1564 (0.6313) time: 0.4907 data: 0.4804 max mem: 6809 +[15:01:24.676670] val: Total time: 0:00:01 (0.5434 s / it) +[15:01:24.685353] val loss: 0.6313354323307673 +[15:01:24.685563] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8008, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7245, Kappa: 0.0000, Score: 0.4111 +[15:01:24.729512] Best epoch = 17, Best score = 0.4392 +[15:01:25.006474] log_dir: ./output_logs/retfound +[15:01:26.198833] Epoch: [76] [0/1] eta: 0:00:01 lr: 0.000003 loss: 0.5826 (0.5826) time: 1.1916 data: 1.1404 max mem: 6809 +[15:01:26.324086] Epoch: [76] Total time: 0:00:01 (1.3175 s / it) +[15:01:26.324831] Averaged stats: lr: 0.000003 loss: 0.5826 (0.5826) +[15:01:27.793580] val: [0/3] eta: 0:00:04 loss: 0.1563 (0.1563) time: 1.4447 data: 1.4308 max mem: 6809 +[15:01:27.811537] val: [2/3] eta: 0:00:00 loss: 0.1563 (0.6313) time: 0.4874 data: 0.4770 max mem: 6809 +[15:01:27.962218] val: Total time: 0:00:01 (0.5380 s / it) +[15:01:27.970944] val loss: 0.6313374688227972 +[15:01:27.971127] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8000, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7236, Kappa: 0.0000, Score: 0.4108 +[15:01:28.005387] Best epoch = 17, Best score = 0.4392 +[15:01:28.266739] log_dir: ./output_logs/retfound +[15:01:29.427340] Epoch: [77] [0/1] eta: 0:00:01 lr: 0.000002 loss: 0.4211 (0.4211) time: 1.1598 data: 1.1080 max mem: 6809 +[15:01:29.609316] Epoch: [77] Total time: 0:00:01 (1.3424 s / it) +[15:01:29.610188] Averaged stats: lr: 0.000002 loss: 0.4211 (0.4211) +[15:01:31.027293] val: [0/3] eta: 0:00:04 loss: 0.1562 (0.1562) time: 1.3942 data: 1.3813 max mem: 6809 +[15:01:31.045261] val: [2/3] eta: 0:00:00 loss: 0.1562 (0.6314) time: 0.4706 data: 0.4605 max mem: 6809 +[15:01:31.253482] val: Total time: 0:00:01 (0.5403 s / it) +[15:01:31.262182] val loss: 0.6314317335685095 +[15:01:31.262357] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8000, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7244, Kappa: 0.0000, Score: 0.4108 +[15:01:31.296985] Best epoch = 17, Best score = 0.4392 +[15:01:31.555749] log_dir: ./output_logs/retfound +[15:01:32.686446] Epoch: [78] [0/1] eta: 0:00:01 lr: 0.000002 loss: 0.4513 (0.4513) time: 1.1299 data: 1.0780 max mem: 6809 +[15:01:32.899347] Epoch: [78] Total time: 0:00:01 (1.3435 s / it) +[15:01:32.900153] Averaged stats: lr: 0.000002 loss: 0.4513 (0.4513) +[15:01:34.315252] val: [0/3] eta: 0:00:04 loss: 0.1562 (0.1562) time: 1.3918 data: 1.3760 max mem: 6809 +[15:01:34.334856] val: [2/3] eta: 0:00:00 loss: 0.1562 (0.6315) time: 0.4703 data: 0.4587 max mem: 6809 +[15:01:34.539868] val: Total time: 0:00:01 (0.5390 s / it) +[15:01:34.548804] val loss: 0.6314988782008489 +[15:01:34.548977] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8008, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7237, Kappa: 0.0000, Score: 0.4111 +[15:01:34.580180] Best epoch = 17, Best score = 0.4392 +[15:01:34.838222] log_dir: ./output_logs/retfound +[15:01:36.080522] Epoch: [79] [0/1] eta: 0:00:01 lr: 0.000001 loss: 0.4061 (0.4061) time: 1.2415 data: 1.1896 max mem: 6809 +[15:01:36.187667] Epoch: [79] Total time: 0:00:01 (1.3493 s / it) +[15:01:36.188437] Averaged stats: lr: 0.000001 loss: 0.4061 (0.4061) +[15:01:37.607279] val: [0/3] eta: 0:00:04 loss: 0.1561 (0.1561) time: 1.4074 data: 1.3946 max mem: 6809 +[15:01:37.625164] val: [2/3] eta: 0:00:00 loss: 0.1561 (0.6315) time: 0.4749 data: 0.4649 max mem: 6809 +[15:01:37.827162] val: Total time: 0:00:01 (0.5426 s / it) +[15:01:37.835871] val loss: 0.6315443267424902 +[15:01:37.836062] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8023, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7252, Kappa: 0.0000, Score: 0.4116 +[15:01:37.893802] Best epoch = 17, Best score = 0.4392 +[15:01:40.755232] Test with the best model, epoch = 17: +[15:01:42.066447] test: [0/6] eta: 0:00:07 loss: 0.1329 (0.1329) time: 1.3010 data: 1.2871 max mem: 6809 +[15:01:42.212949] test: [5/6] eta: 0:00:00 loss: 0.1342 (0.5999) time: 0.2411 data: 0.2196 max mem: 6809 +[15:01:42.485967] test: Total time: 0:00:01 (0.2868 s / it) +[15:01:42.495209] val loss: 0.5998874083161354 +[15:01:42.495321] Accuracy: 0.8095, F1 Score: 0.4474, ROC AUC: 0.6402, Hamming Loss: 0.1905, + Jaccard Score: 0.4048, Precision: 0.4048, Recall: 0.5000, + Average Precision: 0.6181, Kappa: 0.0000, Score: 0.3625 +[15:01:43.636574] Training time 0:04:40 +[rank0]:[W701 15:01:44.995314793 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/010/retfound acc=0.8095 auroc=0.6387867647058825 f1_macro=0.4474 qwk=0.0 diff --git a/results/downsample/papila/010/vit/confusion_matrix.png b/results/downsample/papila/010/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..8030cad4600d6ed1febcf08c0d0107a126dc23b3 --- /dev/null +++ b/results/downsample/papila/010/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:681a4721e63744f34ca2d8dd541c61261e03d7654801e3ef04da62d3dde5271f +size 70939 diff --git a/results/downsample/papila/010/vit/log.csv b/results/downsample/papila/010/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..0c4a145b84b14eb86bb00e72f53d1481df25689e --- /dev/null +++ b/results/downsample/papila/010/vit/log.csv @@ -0,0 +1,29 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6503244042396545,0.5,0.55625,0.3775619783996073,0.0 +1,0.8696306943893433,0.5,0.55625,0.3775619783996073,7.39441178203743e-08 +2,0.9053922891616821,0.2619047619047619,0.515625,0.2523783515721421,1.478882356407486e-07 +3,1.3667328357696533,0.2619047619047619,0.515625,0.2523783515721421,2.2183235346112292e-07 +4,1.0063855648040771,0.23809523809523808,0.534375,0.2422275641025641,2.957764712814972e-07 +5,0.9050484299659729,0.2857142857142857,0.625,0.30486425339366513,3.697205891018715e-07 +6,0.7792922258377075,0.6666666666666666,0.6937500000000001,0.4676719444015709,3.6955843521557525e-07 +7,0.6777358055114746,0.7142857142857143,0.73125,0.4516570758738278,3.6907225802970327e-07 +8,0.8703470826148987,0.6904761904761905,0.74375,0.5270709408275939,3.682629104642438e-07 +9,0.5216395258903503,0.5,0.765625,0.4883261387020763,3.671318123898447e-07 +10,0.5001521706581116,0.3333333333333333,0.778125,0.38708790398678933,3.6568094813687817e-07 +11,0.45891866087913513,0.30952380952380953,0.7765625,0.37110225839161964,3.6391286301425095e-07 +12,0.5629663467407227,0.35714285714285715,0.7718750000000001,0.4002711232431439,3.618306588440675e-07 +13,0.4027242958545685,0.5238095238095238,0.775,0.5066498316498317,3.594379885199801e-07 +14,0.6634818315505981,0.6904761904761905,0.775,0.5937096460448448,3.567390495987718e-07 +15,0.6491262912750244,0.7142857142857143,0.78125,0.5796474358974358,3.537385769364163e-07 +16,0.4568294286727905,0.7857142857142857,0.7875,0.6232588800230597,3.504418343815308e-07 +17,0.5585417747497559,0.8095238095238095,0.784375,0.6459294426366574,3.4685460554079696e-07 +18,0.3401409387588501,0.8095238095238095,0.775,0.6428044426366574,3.4298318363255025e-07 +19,0.37649503350257874,0.7142857142857143,0.778125,0.5572604761537274,3.3883436044633555e-07 +20,0.4255063533782959,0.6666666666666666,0.78125,0.521741452991453,3.344154144278013e-07 +21,0.4428254961967468,0.6904761904761905,0.778125,0.5785639136031607,3.297340979098317e-07 +22,0.364983469247818,0.6666666666666666,0.75625,0.5698553689348262,3.2479862351232145e-07 +23,0.5103057622909546,0.6428571428571429,0.759375,0.4974187853107345,3.1961764973444924e-07 +24,0.3181127905845642,0.7619047619047619,0.7562500000000001,0.5906250000000001,3.1420026576472814e-07 +25,0.43093088269233704,0.8095238095238095,0.75,0.6344711093033241,3.0855597553548053e-07 +26,0.41581475734710693,0.7857142857142857,0.746875,0.5561500205086136,3.026946810497112e-07 +27,0.4480593204498291,0.7619047619047619,0.746875,0.4971773059617548,2.9662666500962944e-07 diff --git a/results/downsample/papila/010/vit/metrics.json b/results/downsample/papila/010/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..43b1ce9bf99b1956be290781aad6758ee5783eee --- /dev/null +++ b/results/downsample/papila/010/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7976190476190477, + "balanced_accuracy": 0.6838235294117647, + "precision_macro": 0.6755926251097454, + "recall_macro": 0.6838235294117647, + "f1_macro": 0.6794612794612795, + "precision_weighted": 0.8025001045194197, + "recall_weighted": 0.7976190476190477, + "f1_weighted": 0.7999358666025332, + "cohen_kappa": 0.3590664272890485, + "quadratic_weighted_kappa": 0.3590664272890485, + "mcc": 0.35932189516560364, + "auroc": 0.6378676470588236, + "auprc": 0.3906037442385183, + "sensitivity": 0.5, + "specificity": 0.8676470588235294, + "precision_pos": 0.47058823529411764, + "f1_pos": 0.48484848484848486, + "per_class": { + "0": { + "precision": 0.8805970149253731, + "recall": 0.8676470588235294, + "f1-score": 0.8740740740740741, + "support": 68.0 + }, + "1": { + "precision": 0.47058823529411764, + "recall": 0.5, + "f1-score": 0.48484848484848486, + "support": 16.0 + }, + "accuracy": 0.7976190476190477, + "macro avg": { + "precision": 0.6755926251097454, + "recall": 0.6838235294117647, + "f1-score": 0.6794612794612795, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8025001045194197, + "recall": 0.7976190476190477, + "f1-score": 0.7999358666025332, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/010/vit/pr.png b/results/downsample/papila/010/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..4cd3d892e9868377479c5ed182d51cf92cb7afc0 --- /dev/null +++ b/results/downsample/papila/010/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:faaaf7f115e08196ef7dc927c7ef0a80253b819afc5dfb197b1b8b5f4f999edb +size 52174 diff --git a/results/downsample/papila/010/vit/roc.png b/results/downsample/papila/010/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..a84c1b949914c37ce005b705571199b09054cc1a --- /dev/null +++ b/results/downsample/papila/010/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b6e12089ffab7daa942df55761ed0433e6f56f70a3c95e441718c99ac0944ac +size 58017 diff --git a/results/downsample/papila/010/vit/test_pred.npz b/results/downsample/papila/010/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..62cc4a638f82c51dc0607dfc6483b27c99176cc7 --- /dev/null +++ b/results/downsample/papila/010/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:abc73845b326be904b218a9e1da76b1539fc4029c54589fcf50a5b784c7bccbd +size 1854 diff --git a/results/downsample/papila/010/vit/train.log b/results/downsample/papila/010/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..a98f247452ffa881ad1cade7b931a2d400b4e6bb --- /dev/null +++ b/results/downsample/papila/010/vit/train.log @@ -0,0 +1,149 @@ +[vit] train=29 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.6503 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.8696 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.9054 val_acc=0.2619 val_auc=0.5156 score=0.2524 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.3667 val_acc=0.2619 val_auc=0.5156 score=0.2524 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=1.0064 val_acc=0.2381 val_auc=0.5344 score=0.2422 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.9050 val_acc=0.2857 val_auc=0.6250 score=0.3049 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.7793 val_acc=0.6667 val_auc=0.6938 score=0.4677 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.6777 val_acc=0.7143 val_auc=0.7312 score=0.4517 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.8703 val_acc=0.6905 val_auc=0.7438 score=0.5271 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.5216 val_acc=0.5000 val_auc=0.7656 score=0.4883 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.5002 val_acc=0.3333 val_auc=0.7781 score=0.3871 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4589 val_acc=0.3095 val_auc=0.7766 score=0.3711 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.5630 val_acc=0.3571 val_auc=0.7719 score=0.4003 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4027 val_acc=0.5238 val_auc=0.7750 score=0.5066 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.6635 val_acc=0.6905 val_auc=0.7750 score=0.5937 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.6491 val_acc=0.7143 val_auc=0.7812 score=0.5796 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4568 val_acc=0.7857 val_auc=0.7875 score=0.6233 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.5585 val_acc=0.8095 val_auc=0.7844 score=0.6459 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.3401 val_acc=0.8095 val_auc=0.7750 score=0.6428 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.3765 val_acc=0.7143 val_auc=0.7781 score=0.5573 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.4255 val_acc=0.6667 val_auc=0.7812 score=0.5217 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4428 val_acc=0.6905 val_auc=0.7781 score=0.5786 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.3650 val_acc=0.6667 val_auc=0.7562 score=0.5699 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.5103 val_acc=0.6429 val_auc=0.7594 score=0.4974 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3181 val_acc=0.7619 val_auc=0.7563 score=0.5906 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.4309 val_acc=0.8095 val_auc=0.7500 score=0.6345 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4158 val_acc=0.7857 val_auc=0.7469 score=0.5562 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.4481 val_acc=0.7619 val_auc=0.7469 score=0.4972 +[vit] early stop at ep27 (best ep17 score=0.6459) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=17 best_val_score=0.6459 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/010/vit acc=0.7976 auroc=0.6378676470588236 f1_macro=0.6795 qwk=0.3590664272890485 diff --git a/results/downsample/papila/025/papila_025pct/confusion_matrix_test.jpg b/results/downsample/papila/025/papila_025pct/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4cd3251670565565b2aba16cb2b7c25a1e6ad676 --- /dev/null +++ b/results/downsample/papila/025/papila_025pct/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f8b514f58159f3872ec1655ef1c6c84c126b8d069b9404ae87be6d9ca1160f6 +size 252679 diff --git a/results/downsample/papila/025/papila_025pct/log.txt b/results/downsample/papila/025/papila_025pct/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..12ce2982d7b971eec24a41a266f4aa0d27c6be4a --- /dev/null +++ b/results/downsample/papila/025/papila_025pct/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 1.5625e-05, "train_loss": 0.6927871704101562, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 7.8125e-05, "train_loss": 0.69268798828125, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014062500000000002, "train_loss": 0.6571502685546875, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00020312500000000002, "train_loss": 0.6088829040527344, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.000265625, "train_loss": 0.5208396911621094, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00032812499999999997, "train_loss": 0.5095539093017578, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.000390625, "train_loss": 0.5811576843261719, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.000453125, "train_loss": 0.5780129432678223, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000515625, "train_loss": 0.5630178451538086, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000578125, "train_loss": 0.502619743347168, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006248797296535528, "train_loss": 0.5125598907470703, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006234377833473366, "train_loss": 0.5149345397949219, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006200818848757455, "train_loss": 0.5596179962158203, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006148327244688065, "train_loss": 0.4870719909667969, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006077226649459515, "train_loss": 0.5089321136474609, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005987955421884734, "train_loss": 0.5342826843261719, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005881063948767417, "train_loss": 0.5201177597045898, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005757211251584354, "train_loss": 0.5254964828491211, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.000561716092339888, "train_loss": 0.4832305908203125, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005461776421055717, "train_loss": 0.5030412673950195, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005292015741682404, "train_loss": 0.5327243804931641, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005108925516318428, "train_loss": 0.5036191940307617, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004913634557086838, "train_loss": 0.47374820709228516, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00047073468976921304, "train_loss": 0.4982309341430664, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004491334370152081, "train_loss": 0.48442840576171875, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.000426692876353043, "train_loss": 0.5116918087005615, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004035513613014434, "train_loss": 0.4750664234161377, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003798515669960373, "train_loss": 0.48610401153564453, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035573961054969873, "train_loss": 0.3859574794769287, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003313641501919586, "train_loss": 0.5084846019744873, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030687546874158533, "train_loss": 0.36304616928100586, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002824245470630336, "train_loss": 0.35725951194763184, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00025816213321920806, "train_loss": 0.4331321716308594, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00023423781305952714, "train_loss": 0.43616461753845215, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00021079908797341933, "train_loss": 0.4032329320907593, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018799046549520726, "train_loss": 0.41208159923553467, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.0001659525683671057, "train_loss": 0.47698259353637695, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014482126755325402, "train_loss": 0.4409008026123047, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012472684455004014, "train_loss": 0.4225883483886719, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010579318815735249, "train_loss": 0.4559565782546997, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.813703066293163e-05, "train_loss": 0.38010454177856445, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 7.186722814899894e-05, "train_loss": 0.3405710458755493, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.708408935831119e-05, "train_loss": 0.39614832401275635, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.387875725740529e-05, "train_loss": 0.392925500869751, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.233264710990193e-05, "train_loss": 0.40308690071105957, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.251694452433174e-05, "train_loss": 0.5007079839706421, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.449216657118709e-05, "train_loss": 0.4213298559188843, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 8.30778867505851e-06, "train_loss": 0.3859215974807739, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 4.001939582190023e-06, "train_loss": 0.4069554805755615, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.601166284079265e-06, "train_loss": 0.44500911235809326, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/025/papila_025pct/metrics_test.csv b/results/downsample/papila/025/papila_025pct/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..355a60f6b34d0375da6c972656bc56c3d225cadc --- /dev/null +++ b/results/downsample/papila/025/papila_025pct/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5708357890446981,0.75,0.636063544460491,0.7424172794117647,0.25,0.49977313974591653,0.626984126984127,0.6544117647058824,0.6491555898315275,0.27586206896551724 diff --git a/results/downsample/papila/025/papila_025pct/metrics_val.csv b/results/downsample/papila/025/papila_025pct/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..8476897c4f3857c363bb4a3bc12c80466a49100d --- /dev/null +++ b/results/downsample/papila/025/papila_025pct/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6931304931640625,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6932540833950043,0.7619047619047619,0.43243243243243246,0.58828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6089108172812516,0.0 +0.6989273130893707,0.7619047619047619,0.43243243243243246,0.7203125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6753175390396292,0.0 +0.7273658514022827,0.7619047619047619,0.43243243243243246,0.78828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7398871570771733,0.0 +0.8092338442802429,0.7619047619047619,0.43243243243243246,0.83671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.794398513120576,0.0 +0.9765872955322266,0.7619047619047619,0.43243243243243246,0.840625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8080592074201579,0.0 +1.1360862255096436,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8160704268171326,0.0 +1.1362947225570679,0.7619047619047619,0.43243243243243246,0.83515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8047602452800908,0.0 +1.048771858215332,0.7619047619047619,0.43243243243243246,0.84609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.821423338897352,0.0 +0.959237277507782,0.7619047619047619,0.43243243243243246,0.8382812500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8237510817462563,0.0 +0.8907642364501953,0.7619047619047619,0.43243243243243246,0.8304687500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8238541475496797,0.0 +0.870038628578186,0.7619047619047619,0.43243243243243246,0.83671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8253195472416519,0.0 +0.8961948156356812,0.7619047619047619,0.43243243243243246,0.85625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8396165894215959,0.0 +0.9349063634872437,0.7619047619047619,0.43243243243243246,0.8625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8293608118389826,0.0 +0.9833486676216125,0.7619047619047619,0.43243243243243246,0.86328125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8261446791739725,0.0 +0.9895151257514954,0.7619047619047619,0.43243243243243246,0.8703125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.818424611578652,0.0 +0.949261486530304,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8167529996830518,0.0 +0.9138736724853516,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8079969934955282,0.0 +0.8665733337402344,0.7619047619047619,0.43243243243243246,0.87578125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8033945313692888,0.0 +0.821393609046936,0.7619047619047619,0.43243243243243246,0.8796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8057996011558242,0.0 +0.7693771123886108,0.7619047619047619,0.43243243243243246,0.875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7805092411093666,0.0 +0.7240272760391235,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7787167627543883,0.0 +0.7130428552627563,0.7619047619047619,0.43243243243243246,0.86875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7768020149473707,0.0 +0.7460571527481079,0.7619047619047619,0.513888888888889,0.8625,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.7721991110801352,0.09482758620689657 +0.7539762258529663,0.7380952380952381,0.49945828819068255,0.859375,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.7936246206992423,0.049382716049382824 +0.8885501623153687,0.7619047619047619,0.513888888888889,0.8648437500000001,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.799238479879278,0.09482758620689657 +1.0140803456306458,0.7619047619047619,0.513888888888889,0.8625,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.7979932943050483,0.09482758620689657 +0.9593358635902405,0.7380952380952381,0.49945828819068255,0.86875,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.8017961407641447,0.049382716049382824 +0.9262599945068359,0.7619047619047619,0.5714285714285714,0.8625,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7698219032525597,0.17322834645669294 +0.7272065877914429,0.7857142857142857,0.6347826086956522,0.878125,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7819408264797454,0.2867924528301887 +0.6347376108169556,0.8095238095238095,0.6911764705882353,0.88125,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7843458962662806,0.3913043478260869 +0.6057861745357513,0.8095238095238095,0.6911764705882353,0.8820312499999999,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7843458962662806,0.3913043478260869 +0.5361908078193665,0.8333333333333334,0.7418788410886743,0.9,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8206289297541346,0.4878048780487805 +0.49860239028930664,0.8333333333333334,0.7619433198380567,0.9031250000000001,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.823045868751956,0.5242718446601942 +0.5100588202476501,0.8333333333333334,0.7619433198380567,0.9031250000000001,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.823045868751956,0.5242718446601942 +0.5540881752967834,0.8095238095238095,0.6911764705882353,0.9031250000000001,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.823045868751956,0.3913043478260869 +0.6192449331283569,0.8095238095238095,0.6911764705882353,0.909375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8300352543804135,0.3913043478260869 +0.6298291683197021,0.8095238095238095,0.6911764705882353,0.9125,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8314255250499589,0.3913043478260869 +0.6357086896896362,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8378880083498789,0.3913043478260869 +0.6205177307128906,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.6146293878555298,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.6064767241477966,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5936670303344727,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5689648687839508,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836717240457713,0.3913043478260869 +0.5564109683036804,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8366992806875981,0.3913043478260869 +0.5510595440864563,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5462074279785156,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5437152981758118,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5424394011497498,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5420194268226624,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8366992806875981,0.3913043478260869 diff --git a/results/downsample/papila/025/papila_025pct/test_pred.npz b/results/downsample/papila/025/papila_025pct/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e4db737d28e62daf002120c8cfc2a8d2960ebbe9 --- /dev/null +++ b/results/downsample/papila/025/papila_025pct/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46dbd7f2a4fa1b8b00100bff82950d7240ddd917e7aaa789231b88a83aa7064d +size 1518 diff --git a/results/downsample/papila/025/resnet/confusion_matrix.png b/results/downsample/papila/025/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..edd5aef6d67c0ea894982ccd8905a9f4795e07dc --- /dev/null +++ b/results/downsample/papila/025/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:db2e39fadee49d581b035931a0e5e0c3eb1be8b9627c173072e8382068c89108 +size 72023 diff --git a/results/downsample/papila/025/resnet/log.csv b/results/downsample/papila/025/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..0355844e9e28fbf18eed18c2289c1130c42459be --- /dev/null +++ b/results/downsample/papila/025/resnet/log.csv @@ -0,0 +1,23 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.701056718826294,0.2857142857142857,0.6656249999999999,0.31840592006033175,0.0 +1,0.7000800967216492,0.2619047619047619,0.596875,0.27946168490547546,0.00016666666666666666 +2,0.6960475444793701,0.23809523809523808,0.46249999999999997,0.21268853695324283,0.0003333333333333333 +3,0.6956083178520203,0.23809523809523808,0.35,0.1673861759849021,0.0005 +4,0.6833346486091614,0.23809523809523808,0.34375,0.14314290161892904,0.0004994417196557883 +5,0.6811696290969849,0.2857142857142857,0.309375,0.10334901433691761,0.000497769372038695 +6,0.6825703978538513,0.30952380952380953,0.36875,0.17450139084118288,0.0004949904262591467 +7,0.6762101054191589,0.35714285714285715,0.40625,0.2294208351509864,0.0004911172937635942 +8,0.6682584285736084,0.4523809523809524,0.503125,0.31466113507258575,0.0004861672729019797 +9,0.6578769683837891,0.5238095238095238,0.565625,0.3334996498599439,0.0004801624716691072 +10,0.6617310047149658,0.5952380952380952,0.625,0.39631073446327686,0.0004731297089649703 +11,0.6480985879898071,0.7142857142857143,0.628125,0.4527483306216529,0.00046510039481503486 +12,0.6489280462265015,0.6904761904761905,0.625,0.35575836295688634,0.0004561103900854401 +13,0.6214346885681152,0.7619047619047619,0.60625,0.4049888250319285,0.00044619984631966527 +14,0.612542450428009,0.7619047619047619,0.634375,0.41436382503192853,0.00043541302641198946 +15,0.6121289730072021,0.7619047619047619,0.6125,0.4070721583652619,0.00042379810691866064 +16,0.5796793699264526,0.7857142857142857,0.5875,0.4205205451277168,0.0004114069628897006 +17,0.5902805924415588,0.7619047619047619,0.509375,0.37269715836526185,0.0003982949361823388 +18,0.5874336957931519,0.7619047619047619,0.48750000000000004,0.3654054916985952,0.0003845205882908432 +19,0.5729873776435852,0.7142857142857143,0.42812500000000003,0.25286159003831415,0.00037014543879667093 +20,0.5772780776023865,0.6666666666666666,0.465625,0.23604822834645675,0.0003552336906070838 +21,0.5382381081581116,0.6428571428571429,0.503125,0.28414908509818204,0.0003398519432093782 diff --git a/results/downsample/papila/025/resnet/metrics.json b/results/downsample/papila/025/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..c1714f3c3d9f7b49182bb31b0f2922d186dd372a --- /dev/null +++ b/results/downsample/papila/025/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7023809523809523, + "balanced_accuracy": 0.5294117647058824, + "precision_macro": 0.5280948200175593, + "recall_macro": 0.5294117647058824, + "f1_macro": 0.5286195286195285, + "precision_weighted": 0.7093523976754882, + "recall_weighted": 0.7023809523809523, + "f1_weighted": 0.7057880391213723, + "cohen_kappa": 0.05745062836624781, + "quadratic_weighted_kappa": 0.05745062836624781, + "mcc": 0.05749150322649658, + "auroc": 0.6240808823529411, + "auprc": 0.3282503325752311, + "sensitivity": 0.25, + "specificity": 0.8088235294117647, + "precision_pos": 0.23529411764705882, + "f1_pos": 0.24242424242424243, + "per_class": { + "0": { + "precision": 0.8208955223880597, + "recall": 0.8088235294117647, + "f1-score": 0.8148148148148148, + "support": 68.0 + }, + "1": { + "precision": 0.23529411764705882, + "recall": 0.25, + "f1-score": 0.24242424242424243, + "support": 16.0 + }, + "accuracy": 0.7023809523809523, + "macro avg": { + "precision": 0.5280948200175593, + "recall": 0.5294117647058824, + "f1-score": 0.5286195286195285, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7093523976754882, + "recall": 0.7023809523809523, + "f1-score": 0.7057880391213723, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/025/resnet/pr.png b/results/downsample/papila/025/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..84cf55c877ab54f6c3dd36142cbaeac3a5237f79 --- /dev/null +++ b/results/downsample/papila/025/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3142da5dd31b2294be64ff7c870807be67db66a447d9feed0172c3ba6cccb5ed +size 51866 diff --git a/results/downsample/papila/025/resnet/roc.png b/results/downsample/papila/025/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..60bc98e496a3f2adf7a4f6ac47d1e01a093caf02 --- /dev/null +++ b/results/downsample/papila/025/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb7e953aa7cb11502d24f5d7e5b2ece3f81ec445de29bbca0c882fdc9aa50814 +size 57771 diff --git a/results/downsample/papila/025/resnet/test_pred.npz b/results/downsample/papila/025/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..84e343d7ef1924e405b380cda798cf22d28c1553 --- /dev/null +++ b/results/downsample/papila/025/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7150295986cc9e7cb233a22b62c2d0cebb4ce20cb49aea2b1750ff91de9d7256 +size 1854 diff --git a/results/downsample/papila/025/resnet/train.log b/results/downsample/papila/025/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..2a1376663915a7dd896edc9914c7afec3705df4e --- /dev/null +++ b/results/downsample/papila/025/resnet/train.log @@ -0,0 +1,119 @@ +[resnet] train=73 val=42 test=84 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.7011 val_acc=0.2857 val_auc=0.6656 score=0.3184 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.7001 val_acc=0.2619 val_auc=0.5969 score=0.2795 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6960 val_acc=0.2381 val_auc=0.4625 score=0.2127 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6956 val_acc=0.2381 val_auc=0.3500 score=0.1674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6833 val_acc=0.2381 val_auc=0.3438 score=0.1431 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6812 val_acc=0.2857 val_auc=0.3094 score=0.1033 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6826 val_acc=0.3095 val_auc=0.3688 score=0.1745 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6762 val_acc=0.3571 val_auc=0.4062 score=0.2294 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6683 val_acc=0.4524 val_auc=0.5031 score=0.3147 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.6579 val_acc=0.5238 val_auc=0.5656 score=0.3335 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.6617 val_acc=0.5952 val_auc=0.6250 score=0.3963 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.6481 val_acc=0.7143 val_auc=0.6281 score=0.4527 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.6489 val_acc=0.6905 val_auc=0.6250 score=0.3558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.6214 val_acc=0.7619 val_auc=0.6062 score=0.4050 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.6125 val_acc=0.7619 val_auc=0.6344 score=0.4144 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.6121 val_acc=0.7619 val_auc=0.6125 score=0.4071 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.5797 val_acc=0.7857 val_auc=0.5875 score=0.4205 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.5903 val_acc=0.7619 val_auc=0.5094 score=0.3727 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.5874 val_acc=0.7619 val_auc=0.4875 score=0.3654 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.5730 val_acc=0.7143 val_auc=0.4281 score=0.2529 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.5773 val_acc=0.6667 val_auc=0.4656 score=0.2360 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.5382 val_acc=0.6429 val_auc=0.5031 score=0.2841 +[resnet] early stop at ep21 (best ep11 score=0.4527) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=11 best_val_score=0.4527 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/025/resnet acc=0.7024 auroc=0.6240808823529411 f1_macro=0.5286 qwk=0.05745062836624781 diff --git a/results/downsample/papila/025/retfound/confusion_matrix.png b/results/downsample/papila/025/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..1e2ad5d38f0576c5efdd7908d4f54737c8b1d66e --- /dev/null +++ b/results/downsample/papila/025/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5009a391ba785a0b67fdf2261db7050e3bac1b4a97ef502604e0eb0d35f38532 +size 70637 diff --git a/results/downsample/papila/025/retfound/confusion_matrix_test.jpg b/results/downsample/papila/025/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4cd3251670565565b2aba16cb2b7c25a1e6ad676 --- /dev/null +++ b/results/downsample/papila/025/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f8b514f58159f3872ec1655ef1c6c84c126b8d069b9404ae87be6d9ca1160f6 +size 252679 diff --git a/results/downsample/papila/025/retfound/log.txt b/results/downsample/papila/025/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..12ce2982d7b971eec24a41a266f4aa0d27c6be4a --- /dev/null +++ b/results/downsample/papila/025/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 1.5625e-05, "train_loss": 0.6927871704101562, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 7.8125e-05, "train_loss": 0.69268798828125, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014062500000000002, "train_loss": 0.6571502685546875, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00020312500000000002, "train_loss": 0.6088829040527344, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.000265625, "train_loss": 0.5208396911621094, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.00032812499999999997, "train_loss": 0.5095539093017578, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.000390625, "train_loss": 0.5811576843261719, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.000453125, "train_loss": 0.5780129432678223, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.000515625, "train_loss": 0.5630178451538086, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.000578125, "train_loss": 0.502619743347168, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006248797296535528, "train_loss": 0.5125598907470703, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006234377833473366, "train_loss": 0.5149345397949219, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006200818848757455, "train_loss": 0.5596179962158203, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006148327244688065, "train_loss": 0.4870719909667969, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006077226649459515, "train_loss": 0.5089321136474609, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005987955421884734, "train_loss": 0.5342826843261719, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005881063948767417, "train_loss": 0.5201177597045898, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005757211251584354, "train_loss": 0.5254964828491211, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.000561716092339888, "train_loss": 0.4832305908203125, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005461776421055717, "train_loss": 0.5030412673950195, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005292015741682404, "train_loss": 0.5327243804931641, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005108925516318428, "train_loss": 0.5036191940307617, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004913634557086838, "train_loss": 0.47374820709228516, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00047073468976921304, "train_loss": 0.4982309341430664, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004491334370152081, "train_loss": 0.48442840576171875, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.000426692876353043, "train_loss": 0.5116918087005615, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004035513613014434, "train_loss": 0.4750664234161377, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003798515669960373, "train_loss": 0.48610401153564453, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035573961054969873, "train_loss": 0.3859574794769287, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003313641501919586, "train_loss": 0.5084846019744873, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030687546874158533, "train_loss": 0.36304616928100586, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002824245470630336, "train_loss": 0.35725951194763184, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.00025816213321920806, "train_loss": 0.4331321716308594, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00023423781305952714, "train_loss": 0.43616461753845215, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00021079908797341933, "train_loss": 0.4032329320907593, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018799046549520726, "train_loss": 0.41208159923553467, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.0001659525683671057, "train_loss": 0.47698259353637695, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014482126755325402, "train_loss": 0.4409008026123047, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012472684455004014, "train_loss": 0.4225883483886719, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010579318815735249, "train_loss": 0.4559565782546997, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.813703066293163e-05, "train_loss": 0.38010454177856445, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 7.186722814899894e-05, "train_loss": 0.3405710458755493, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.708408935831119e-05, "train_loss": 0.39614832401275635, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.387875725740529e-05, "train_loss": 0.392925500869751, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.233264710990193e-05, "train_loss": 0.40308690071105957, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.251694452433174e-05, "train_loss": 0.5007079839706421, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.449216657118709e-05, "train_loss": 0.4213298559188843, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 8.30778867505851e-06, "train_loss": 0.3859215974807739, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 4.001939582190023e-06, "train_loss": 0.4069554805755615, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.601166284079265e-06, "train_loss": 0.44500911235809326, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/025/retfound/metrics.json b/results/downsample/papila/025/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..f6084f9c5c1127532aefa35bfd6c8dd6a9c81f6c --- /dev/null +++ b/results/downsample/papila/025/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.75, + "balanced_accuracy": 0.6544117647058824, + "precision_macro": 0.626984126984127, + "recall_macro": 0.6544117647058824, + "f1_macro": 0.636063544460491, + "precision_weighted": 0.7792894935752079, + "recall_weighted": 0.75, + "f1_weighted": 0.7621208995254796, + "cohen_kappa": 0.27586206896551724, + "quadratic_weighted_kappa": 0.27586206896551724, + "mcc": 0.28005601680560194, + "auroc": 0.7421875, + "auprc": 0.38651604740348866, + "sensitivity": 0.5, + "specificity": 0.8088235294117647, + "precision_pos": 0.38095238095238093, + "f1_pos": 0.43243243243243246, + "per_class": { + "0": { + "precision": 0.873015873015873, + "recall": 0.8088235294117647, + "f1-score": 0.8396946564885496, + "support": 68.0 + }, + "1": { + "precision": 0.38095238095238093, + "recall": 0.5, + "f1-score": 0.43243243243243246, + "support": 16.0 + }, + "accuracy": 0.75, + "macro avg": { + "precision": 0.626984126984127, + "recall": 0.6544117647058824, + "f1-score": 0.636063544460491, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7792894935752079, + "recall": 0.75, + "f1-score": 0.7621208995254796, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/025/retfound/metrics_test.csv b/results/downsample/papila/025/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..355a60f6b34d0375da6c972656bc56c3d225cadc --- /dev/null +++ b/results/downsample/papila/025/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5708357890446981,0.75,0.636063544460491,0.7424172794117647,0.25,0.49977313974591653,0.626984126984127,0.6544117647058824,0.6491555898315275,0.27586206896551724 diff --git a/results/downsample/papila/025/retfound/metrics_val.csv b/results/downsample/papila/025/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..8476897c4f3857c363bb4a3bc12c80466a49100d --- /dev/null +++ b/results/downsample/papila/025/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6931304931640625,0.7619047619047619,0.43243243243243246,0.465625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.48897243107769417,0.0 +0.6932540833950043,0.7619047619047619,0.43243243243243246,0.58828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6089108172812516,0.0 +0.6989273130893707,0.7619047619047619,0.43243243243243246,0.7203125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6753175390396292,0.0 +0.7273658514022827,0.7619047619047619,0.43243243243243246,0.78828125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7398871570771733,0.0 +0.8092338442802429,0.7619047619047619,0.43243243243243246,0.83671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.794398513120576,0.0 +0.9765872955322266,0.7619047619047619,0.43243243243243246,0.840625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8080592074201579,0.0 +1.1360862255096436,0.7619047619047619,0.43243243243243246,0.8414062499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8160704268171326,0.0 +1.1362947225570679,0.7619047619047619,0.43243243243243246,0.83515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8047602452800908,0.0 +1.048771858215332,0.7619047619047619,0.43243243243243246,0.84609375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.821423338897352,0.0 +0.959237277507782,0.7619047619047619,0.43243243243243246,0.8382812500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8237510817462563,0.0 +0.8907642364501953,0.7619047619047619,0.43243243243243246,0.8304687500000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8238541475496797,0.0 +0.870038628578186,0.7619047619047619,0.43243243243243246,0.83671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8253195472416519,0.0 +0.8961948156356812,0.7619047619047619,0.43243243243243246,0.85625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8396165894215959,0.0 +0.9349063634872437,0.7619047619047619,0.43243243243243246,0.8625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8293608118389826,0.0 +0.9833486676216125,0.7619047619047619,0.43243243243243246,0.86328125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8261446791739725,0.0 +0.9895151257514954,0.7619047619047619,0.43243243243243246,0.8703125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.818424611578652,0.0 +0.949261486530304,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8167529996830518,0.0 +0.9138736724853516,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8079969934955282,0.0 +0.8665733337402344,0.7619047619047619,0.43243243243243246,0.87578125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8033945313692888,0.0 +0.821393609046936,0.7619047619047619,0.43243243243243246,0.8796875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8057996011558242,0.0 +0.7693771123886108,0.7619047619047619,0.43243243243243246,0.875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7805092411093666,0.0 +0.7240272760391235,0.7619047619047619,0.43243243243243246,0.871875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7787167627543883,0.0 +0.7130428552627563,0.7619047619047619,0.43243243243243246,0.86875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7768020149473707,0.0 +0.7460571527481079,0.7619047619047619,0.513888888888889,0.8625,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.7721991110801352,0.09482758620689657 +0.7539762258529663,0.7380952380952381,0.49945828819068255,0.859375,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.7936246206992423,0.049382716049382824 +0.8885501623153687,0.7619047619047619,0.513888888888889,0.8648437500000001,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.799238479879278,0.09482758620689657 +1.0140803456306458,0.7619047619047619,0.513888888888889,0.8625,0.23809523809523808,0.42350332594235035,0.6375,0.534375,0.7979932943050483,0.09482758620689657 +0.9593358635902405,0.7380952380952381,0.49945828819068255,0.86875,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.8017961407641447,0.049382716049382824 +0.9262599945068359,0.7619047619047619,0.5714285714285714,0.8625,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7698219032525597,0.17322834645669294 +0.7272065877914429,0.7857142857142857,0.6347826086956522,0.878125,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7819408264797454,0.2867924528301887 +0.6347376108169556,0.8095238095238095,0.6911764705882353,0.88125,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7843458962662806,0.3913043478260869 +0.6057861745357513,0.8095238095238095,0.6911764705882353,0.8820312499999999,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7843458962662806,0.3913043478260869 +0.5361908078193665,0.8333333333333334,0.7418788410886743,0.9,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8206289297541346,0.4878048780487805 +0.49860239028930664,0.8333333333333334,0.7619433198380567,0.9031250000000001,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.823045868751956,0.5242718446601942 +0.5100588202476501,0.8333333333333334,0.7619433198380567,0.9031250000000001,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.823045868751956,0.5242718446601942 +0.5540881752967834,0.8095238095238095,0.6911764705882353,0.9031250000000001,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.823045868751956,0.3913043478260869 +0.6192449331283569,0.8095238095238095,0.6911764705882353,0.909375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8300352543804135,0.3913043478260869 +0.6298291683197021,0.8095238095238095,0.6911764705882353,0.9125,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8314255250499589,0.3913043478260869 +0.6357086896896362,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8378880083498789,0.3913043478260869 +0.6205177307128906,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.6146293878555298,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.6064767241477966,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5936670303344727,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5689648687839508,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836717240457713,0.3913043478260869 +0.5564109683036804,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8366992806875981,0.3913043478260869 +0.5510595440864563,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5462074279785156,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5437152981758118,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5424394011497498,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.836699280687598,0.3913043478260869 +0.5420194268226624,0.8095238095238095,0.6911764705882353,0.91875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8366992806875981,0.3913043478260869 diff --git a/results/downsample/papila/025/retfound/pr.png b/results/downsample/papila/025/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..c4a62ea7385015be27e0db56c1cc94964ebec4d8 --- /dev/null +++ b/results/downsample/papila/025/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3afd628d74ec1866148596c15f33aefa4b86ab3b89d585f8304df3fbc94e017a +size 59659 diff --git a/results/downsample/papila/025/retfound/roc.png b/results/downsample/papila/025/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..e9a9acd027d1db1966b492198d7ec4f93cbd0954 --- /dev/null +++ b/results/downsample/papila/025/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58e1c054e8bf39bd5d4d213dcbda934f339d35f1fe1b9719a275e32f1ff54703 +size 58027 diff --git a/results/downsample/papila/025/retfound/test_pred.npz b/results/downsample/papila/025/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e4db737d28e62daf002120c8cfc2a8d2960ebbe9 --- /dev/null +++ b/results/downsample/papila/025/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46dbd7f2a4fa1b8b00100bff82950d7240ddd917e7aaa789231b88a83aa7064d +size 1518 diff --git a/results/downsample/papila/025/retfound/train.log b/results/downsample/papila/025/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..6817627ca4d9a42f96945ea47595837cfae5427d --- /dev/null +++ b/results/downsample/papila/025/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:54:43.016824411 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:54:43.864725] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:54:43.865173] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/papila_25', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/025', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:54:46.846044] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:54:48.419685] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:54:48.785504] Sampler_train = +[14:54:48.829807] len of train_set: 64 +[14:54:49.368213] [Adaptation] Full fine-tuning: training all parameters. +[14:54:49.369244] number of trainable params (M): 303.30 +[14:54:49.369340] base lr: 5.00e-03 +[14:54:49.369406] actual lr: 6.25e-04 +[14:54:49.369468] accumulate grad iterations: 1 +[14:54:49.369527] effective batch size: 32 +[14:54:49.372623] criterion = CrossEntropyLoss() +[14:54:49.372759] Start training for 50 epochs +[14:54:49.375039] log_dir: ./output_logs/retfound +[14:54:52.267557] Epoch: [0] [0/2] eta: 0:00:05 lr: 0.000000 loss: 0.6927 (0.6927) time: 2.8911 data: 2.3433 max mem: 7340 +[14:54:52.360848] Epoch: [0] [1/2] eta: 0:00:01 lr: 0.000031 loss: 0.6927 (0.6928) time: 1.4918 data: 1.1717 max mem: 7340 +[14:54:52.430008] Epoch: [0] Total time: 0:00:03 (1.5274 s / it) +[14:54:52.431464] Averaged stats: lr: 0.000031 loss: 0.6927 (0.6928) +[14:54:55.318508] val: [0/2] eta: 0:00:05 loss: 0.6924 (0.6924) time: 2.8637 data: 2.8175 max mem: 7340 +[14:54:55.420731] val: [1/2] eta: 0:00:01 loss: 0.6924 (0.6931) time: 1.4827 data: 1.4088 max mem: 7340 +[14:54:55.491799] val: Total time: 0:00:03 (1.5188 s / it) +[14:54:55.506729] val loss: 0.6931304931640625 +[14:54:55.506929] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.4656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4890, Kappa: 0.0000, Score: 0.2994 +[14:54:57.259590] Best epoch = 0, Best score = 0.2994 +[14:54:57.333088] log_dir: ./output_logs/retfound +[14:54:59.774937] Epoch: [1] [0/2] eta: 0:00:04 lr: 0.000063 loss: 0.6925 (0.6925) time: 2.4409 data: 2.3533 max mem: 7340 +[14:54:59.875961] Epoch: [1] [1/2] eta: 0:00:01 lr: 0.000094 loss: 0.6925 (0.6927) time: 1.2707 data: 1.1767 max mem: 7340 +[14:54:59.950692] Epoch: [1] Total time: 0:00:02 (1.3087 s / it) +[14:54:59.951544] Averaged stats: lr: 0.000094 loss: 0.6925 (0.6927) +[14:55:02.659736] val: [0/2] eta: 0:00:05 loss: 0.6665 (0.6665) time: 2.6955 data: 2.6601 max mem: 7340 +[14:55:02.679646] val: [1/2] eta: 0:00:01 loss: 0.6665 (0.6933) time: 1.3574 data: 1.3302 max mem: 7340 +[14:55:02.751520] val: Total time: 0:00:02 (1.3946 s / it) +[14:55:02.764395] val loss: 0.6932540833950043 +[14:55:02.764613] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.5883, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6089, Kappa: 0.0000, Score: 0.3402 +[14:55:04.354173] Best epoch = 1, Best score = 0.3402 +[14:55:04.416504] log_dir: ./output_logs/retfound +[14:55:06.779665] Epoch: [2] [0/2] eta: 0:00:04 lr: 0.000125 loss: 0.6755 (0.6755) time: 2.3623 data: 2.2795 max mem: 9671 +[14:55:06.844970] Epoch: [2] [1/2] eta: 0:00:01 lr: 0.000156 loss: 0.6388 (0.6572) time: 1.2134 data: 1.1398 max mem: 9671 +[14:55:06.923400] Epoch: [2] Total time: 0:00:02 (1.2534 s / it) +[14:55:06.924328] Averaged stats: lr: 0.000156 loss: 0.6388 (0.6572) +[14:55:09.685207] val: [0/2] eta: 0:00:05 loss: 0.5744 (0.5744) time: 2.7329 data: 2.6978 max mem: 9671 +[14:55:09.701724] val: [1/2] eta: 0:00:01 loss: 0.5744 (0.6989) time: 1.3744 data: 1.3490 max mem: 9671 +[14:55:09.777486] val: Total time: 0:00:02 (1.4130 s / it) +[14:55:09.786518] val loss: 0.6989273130893707 +[14:55:09.786748] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7203, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6753, Kappa: 0.0000, Score: 0.3842 +[14:55:11.540734] Best epoch = 2, Best score = 0.3842 +[14:55:11.614336] log_dir: ./output_logs/retfound +[14:55:13.795154] Epoch: [3] [0/2] eta: 0:00:04 lr: 0.000188 loss: 0.6125 (0.6125) time: 2.1796 data: 2.1036 max mem: 9671 +[14:55:13.861992] Epoch: [3] [1/2] eta: 0:00:01 lr: 0.000219 loss: 0.6052 (0.6089) time: 1.1229 data: 1.0519 max mem: 9671 +[14:55:13.945474] Epoch: [3] Total time: 0:00:02 (1.1654 s / it) +[14:55:13.946469] Averaged stats: lr: 0.000219 loss: 0.6052 (0.6089) +[14:55:16.855180] val: [0/2] eta: 0:00:05 loss: 0.4396 (0.4396) time: 2.8857 data: 2.8499 max mem: 9671 +[14:55:16.873937] val: [1/2] eta: 0:00:01 loss: 0.4396 (0.7274) time: 1.4517 data: 1.4250 max mem: 9671 +[14:55:16.956586] val: Total time: 0:00:02 (1.4941 s / it) +[14:55:16.971451] val loss: 0.7273658514022827 +[14:55:16.971724] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7883, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7399, Kappa: 0.0000, Score: 0.4069 +[14:55:18.688203] Best epoch = 3, Best score = 0.4069 +[14:55:18.759306] log_dir: ./output_logs/retfound +[14:55:21.087458] Epoch: [4] [0/2] eta: 0:00:04 lr: 0.000250 loss: 0.6041 (0.6041) time: 2.3269 data: 2.2564 max mem: 9671 +[14:55:21.153380] Epoch: [4] [1/2] eta: 0:00:01 lr: 0.000281 loss: 0.4376 (0.5208) time: 1.1960 data: 1.1282 max mem: 9671 +[14:55:21.225961] Epoch: [4] Total time: 0:00:02 (1.2333 s / it) +[14:55:21.226791] Averaged stats: lr: 0.000281 loss: 0.4376 (0.5208) +[14:55:24.059465] val: [0/2] eta: 0:00:05 loss: 0.2898 (0.2898) time: 2.8212 data: 2.7887 max mem: 9671 +[14:55:24.076293] val: [1/2] eta: 0:00:01 loss: 0.2898 (0.8092) time: 1.4187 data: 1.3944 max mem: 9671 +[14:55:24.146945] val: Total time: 0:00:02 (1.4547 s / it) +[14:55:24.155956] val loss: 0.8092338442802429 +[14:55:24.156141] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8367, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7944, Kappa: 0.0000, Score: 0.4231 +[14:55:25.821195] Best epoch = 4, Best score = 0.4231 +[14:55:25.907104] log_dir: ./output_logs/retfound +[14:55:28.064317] Epoch: [5] [0/2] eta: 0:00:04 lr: 0.000313 loss: 0.4972 (0.4972) time: 2.1562 data: 2.0195 max mem: 9671 +[14:55:28.208185] Epoch: [5] [1/2] eta: 0:00:01 lr: 0.000344 loss: 0.4972 (0.5096) time: 1.1496 data: 1.0098 max mem: 9671 +[14:55:28.278941] Epoch: [5] Total time: 0:00:02 (1.1858 s / it) +[14:55:28.288447] Averaged stats: lr: 0.000344 loss: 0.4972 (0.5096) +[14:55:30.980441] val: [0/2] eta: 0:00:05 loss: 0.1636 (0.1636) time: 2.6810 data: 2.6643 max mem: 9671 +[14:55:30.989943] val: [1/2] eta: 0:00:01 loss: 0.1636 (0.9766) time: 1.3450 data: 1.3322 max mem: 9671 +[14:55:31.061962] val: Total time: 0:00:02 (1.3816 s / it) +[14:55:31.070832] val loss: 0.9765872955322266 +[14:55:31.071011] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8406, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8081, Kappa: 0.0000, Score: 0.4244 +[14:55:32.806232] Best epoch = 5, Best score = 0.4244 +[14:55:32.875114] log_dir: ./output_logs/retfound +[14:55:35.297352] Epoch: [6] [0/2] eta: 0:00:04 lr: 0.000375 loss: 0.4703 (0.4703) time: 2.4211 data: 2.2731 max mem: 9671 +[14:55:35.430853] Epoch: [6] [1/2] eta: 0:00:01 lr: 0.000406 loss: 0.4703 (0.5812) time: 1.2769 data: 1.1366 max mem: 9671 +[14:55:35.504424] Epoch: [6] Total time: 0:00:02 (1.3145 s / it) +[14:55:35.512953] Averaged stats: lr: 0.000406 loss: 0.4703 (0.5812) +[14:55:38.316448] val: [0/2] eta: 0:00:05 loss: 0.1038 (0.1038) time: 2.7913 data: 2.7742 max mem: 9671 +[14:55:38.326232] val: [1/2] eta: 0:00:01 loss: 0.1038 (1.1361) time: 1.4002 data: 1.3872 max mem: 9671 +[14:55:38.416359] val: Total time: 0:00:02 (1.4459 s / it) +[14:55:38.426147] val loss: 1.1360862255096436 +[14:55:38.426359] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8414, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8161, Kappa: 0.0000, Score: 0.4246 +[14:55:40.168789] Best epoch = 6, Best score = 0.4246 +[14:55:40.235269] log_dir: ./output_logs/retfound +[14:55:42.556872] Epoch: [7] [0/2] eta: 0:00:04 lr: 0.000438 loss: 0.5844 (0.5844) time: 2.3205 data: 2.1702 max mem: 9671 +[14:55:42.696869] Epoch: [7] [1/2] eta: 0:00:01 lr: 0.000469 loss: 0.5716 (0.5780) time: 1.2293 data: 1.0851 max mem: 9671 +[14:55:42.771956] Epoch: [7] Total time: 0:00:02 (1.2683 s / it) +[14:55:42.780588] Averaged stats: lr: 0.000469 loss: 0.5716 (0.5780) +[14:55:45.498705] val: [0/2] eta: 0:00:05 loss: 0.1034 (0.1034) time: 2.6911 data: 2.6739 max mem: 9671 +[14:55:45.508611] val: [1/2] eta: 0:00:01 loss: 0.1034 (1.1363) time: 1.3502 data: 1.3370 max mem: 9671 +[14:55:45.576306] val: Total time: 0:00:02 (1.3847 s / it) +[14:55:45.585831] val loss: 1.1362947225570679 +[14:55:45.586074] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8352, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8048, Kappa: 0.0000, Score: 0.4225 +[14:55:45.632422] Best epoch = 6, Best score = 0.4246 +[14:55:45.882826] log_dir: ./output_logs/retfound +[14:55:48.104892] Epoch: [8] [0/2] eta: 0:00:04 lr: 0.000500 loss: 0.6924 (0.6924) time: 2.2210 data: 2.0845 max mem: 9671 +[14:55:48.248943] Epoch: [8] [1/2] eta: 0:00:01 lr: 0.000531 loss: 0.4336 (0.5630) time: 1.1821 data: 1.0423 max mem: 9671 +[14:55:48.325070] Epoch: [8] Total time: 0:00:02 (1.2210 s / it) +[14:55:48.333553] Averaged stats: lr: 0.000531 loss: 0.4336 (0.5630) +[14:55:51.048567] val: [0/2] eta: 0:00:05 loss: 0.1293 (0.1293) time: 2.6918 data: 2.6555 max mem: 9671 +[14:55:51.064825] val: [1/2] eta: 0:00:01 loss: 0.1293 (1.0488) time: 1.3537 data: 1.3278 max mem: 9671 +[14:55:51.136343] val: Total time: 0:00:02 (1.3902 s / it) +[14:55:51.146928] val loss: 1.048771858215332 +[14:55:51.147129] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8461, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8214, Kappa: 0.0000, Score: 0.4262 +[14:55:52.894297] Best epoch = 8, Best score = 0.4262 +[14:55:52.970062] log_dir: ./output_logs/retfound +[14:55:55.138420] Epoch: [9] [0/2] eta: 0:00:04 lr: 0.000562 loss: 0.6117 (0.6117) time: 2.1665 data: 2.0892 max mem: 9671 +[14:55:55.216252] Epoch: [9] [1/2] eta: 0:00:01 lr: 0.000594 loss: 0.3935 (0.5026) time: 1.1216 data: 1.0488 max mem: 9671 +[14:55:55.290326] Epoch: [9] Total time: 0:00:02 (1.1600 s / it) +[14:55:55.291479] Averaged stats: lr: 0.000594 loss: 0.3935 (0.5026) +[14:55:58.054423] val: [0/2] eta: 0:00:05 loss: 0.1664 (0.1664) time: 2.7392 data: 2.7050 max mem: 9671 +[14:55:58.070858] val: [1/2] eta: 0:00:01 loss: 0.1664 (0.9592) time: 1.3775 data: 1.3526 max mem: 9671 +[14:55:58.140828] val: Total time: 0:00:02 (1.4132 s / it) +[14:55:58.149910] val loss: 0.959237277507782 +[14:55:58.150114] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8383, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8238, Kappa: 0.0000, Score: 0.4236 +[14:55:58.203234] Best epoch = 8, Best score = 0.4262 +[14:55:58.470988] log_dir: ./output_logs/retfound +[14:56:00.708597] Epoch: [10] [0/2] eta: 0:00:04 lr: 0.000625 loss: 0.6419 (0.6419) time: 2.2366 data: 2.1637 max mem: 9671 +[14:56:00.774374] Epoch: [10] [1/2] eta: 0:00:01 lr: 0.000625 loss: 0.3832 (0.5126) time: 1.1508 data: 1.0819 max mem: 9671 +[14:56:00.843318] Epoch: [10] Total time: 0:00:02 (1.1861 s / it) +[14:56:00.844159] Averaged stats: lr: 0.000625 loss: 0.3832 (0.5126) +[14:56:03.606734] val: [0/2] eta: 0:00:05 loss: 0.2054 (0.2054) time: 2.7400 data: 2.7132 max mem: 9671 +[14:56:03.624046] val: [1/2] eta: 0:00:01 loss: 0.2054 (0.8908) time: 1.3784 data: 1.3567 max mem: 9671 +[14:56:03.695674] val: Total time: 0:00:02 (1.4148 s / it) +[14:56:03.710323] val loss: 0.8907642364501953 +[14:56:03.710531] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8305, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8239, Kappa: 0.0000, Score: 0.4210 +[14:56:03.745894] Best epoch = 8, Best score = 0.4262 +[14:56:04.030497] log_dir: ./output_logs/retfound +[14:56:06.274788] Epoch: [11] [0/2] eta: 0:00:04 lr: 0.000624 loss: 0.5653 (0.5653) time: 2.2431 data: 2.0963 max mem: 9671 +[14:56:06.413909] Epoch: [11] [1/2] eta: 0:00:01 lr: 0.000623 loss: 0.4645 (0.5149) time: 1.1907 data: 1.0482 max mem: 9671 +[14:56:06.490826] Epoch: [11] Total time: 0:00:02 (1.2301 s / it) +[14:56:06.498861] Averaged stats: lr: 0.000623 loss: 0.4645 (0.5149) +[14:56:09.250444] val: [0/2] eta: 0:00:05 loss: 0.2178 (0.2178) time: 2.7288 data: 2.7115 max mem: 9671 +[14:56:09.260602] val: [1/2] eta: 0:00:01 loss: 0.2178 (0.8700) time: 1.3692 data: 1.3558 max mem: 9671 +[14:56:09.328976] val: Total time: 0:00:02 (1.4040 s / it) +[14:56:09.337830] val loss: 0.870038628578186 +[14:56:09.338009] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8367, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8253, Kappa: 0.0000, Score: 0.4231 +[14:56:09.383499] Best epoch = 8, Best score = 0.4262 +[14:56:09.659158] log_dir: ./output_logs/retfound +[14:56:11.873752] Epoch: [12] [0/2] eta: 0:00:04 lr: 0.000621 loss: 0.4471 (0.4471) time: 2.2136 data: 2.0660 max mem: 9671 +[14:56:12.017625] Epoch: [12] [1/2] eta: 0:00:01 lr: 0.000619 loss: 0.4471 (0.5596) time: 1.1784 data: 1.0331 max mem: 9671 +[14:56:12.087693] Epoch: [12] Total time: 0:00:02 (1.2142 s / it) +[14:56:12.095639] Averaged stats: lr: 0.000619 loss: 0.4471 (0.5596) +[14:56:14.965418] val: [0/2] eta: 0:00:05 loss: 0.1975 (0.1975) time: 2.8620 data: 2.8272 max mem: 9671 +[14:56:14.981578] val: [1/2] eta: 0:00:01 loss: 0.1975 (0.8962) time: 1.4388 data: 1.4137 max mem: 9671 +[14:56:15.050150] val: Total time: 0:00:02 (1.4737 s / it) +[14:56:15.059142] val loss: 0.8961948156356812 +[14:56:15.059372] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8562, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8396, Kappa: 0.0000, Score: 0.4296 +[14:56:16.772173] Best epoch = 12, Best score = 0.4296 +[14:56:16.844998] log_dir: ./output_logs/retfound +[14:56:19.246081] Epoch: [13] [0/2] eta: 0:00:04 lr: 0.000616 loss: 0.4263 (0.4263) time: 2.4001 data: 2.2634 max mem: 9671 +[14:56:19.389215] Epoch: [13] [1/2] eta: 0:00:01 lr: 0.000613 loss: 0.4263 (0.4871) time: 1.2712 data: 1.1317 max mem: 9671 +[14:56:19.459532] Epoch: [13] Total time: 0:00:02 (1.3072 s / it) +[14:56:19.467550] Averaged stats: lr: 0.000613 loss: 0.4263 (0.4871) +[14:56:22.178044] val: [0/2] eta: 0:00:05 loss: 0.1724 (0.1724) time: 2.6987 data: 2.6632 max mem: 9671 +[14:56:22.194715] val: [1/2] eta: 0:00:01 loss: 0.1724 (0.9349) time: 1.3574 data: 1.3317 max mem: 9671 +[14:56:22.264228] val: Total time: 0:00:02 (1.3928 s / it) +[14:56:22.273810] val loss: 0.9349063634872437 +[14:56:22.273986] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8625, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8294, Kappa: 0.0000, Score: 0.4316 +[14:56:24.003587] Best epoch = 13, Best score = 0.4316 +[14:56:24.067160] log_dir: ./output_logs/retfound +[14:56:26.216549] Epoch: [14] [0/2] eta: 0:00:04 lr: 0.000610 loss: 0.5052 (0.5052) time: 2.1485 data: 2.0778 max mem: 9671 +[14:56:26.282275] Epoch: [14] [1/2] eta: 0:00:01 lr: 0.000606 loss: 0.5052 (0.5089) time: 1.1067 data: 1.0390 max mem: 9671 +[14:56:26.354179] Epoch: [14] Total time: 0:00:02 (1.1434 s / it) +[14:56:26.355007] Averaged stats: lr: 0.000606 loss: 0.5052 (0.5089) +[14:56:29.120872] val: [0/2] eta: 0:00:05 loss: 0.1465 (0.1465) time: 2.7420 data: 2.7062 max mem: 9671 +[14:56:29.141006] val: [1/2] eta: 0:00:01 loss: 0.1465 (0.9833) time: 1.3806 data: 1.3533 max mem: 9671 +[14:56:29.217884] val: Total time: 0:00:02 (1.4203 s / it) +[14:56:29.226978] val loss: 0.9833486676216125 +[14:56:29.227220] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8633, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8261, Kappa: 0.0000, Score: 0.4319 +[14:56:30.945668] Best epoch = 14, Best score = 0.4319 +[14:56:31.025532] log_dir: ./output_logs/retfound +[14:56:33.331691] Epoch: [15] [0/2] eta: 0:00:04 lr: 0.000601 loss: 0.5157 (0.5157) time: 2.3052 data: 2.2356 max mem: 9671 +[14:56:33.397240] Epoch: [15] [1/2] eta: 0:00:01 lr: 0.000596 loss: 0.5157 (0.5343) time: 1.1851 data: 1.1178 max mem: 9671 +[14:56:33.468564] Epoch: [15] Total time: 0:00:02 (1.2214 s / it) +[14:56:33.469408] Averaged stats: lr: 0.000596 loss: 0.5157 (0.5343) +[14:56:36.212699] val: [0/2] eta: 0:00:05 loss: 0.1412 (0.1412) time: 2.7205 data: 2.6860 max mem: 9671 +[14:56:36.229138] val: [1/2] eta: 0:00:01 loss: 0.1412 (0.9895) time: 1.3682 data: 1.3431 max mem: 9671 +[14:56:36.306742] val: Total time: 0:00:02 (1.4076 s / it) +[14:56:36.315923] val loss: 0.9895151257514954 +[14:56:36.316127] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8703, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8184, Kappa: 0.0000, Score: 0.4342 +[14:56:38.014967] Best epoch = 15, Best score = 0.4342 +[14:56:38.075743] log_dir: ./output_logs/retfound +[14:56:40.437588] Epoch: [16] [0/2] eta: 0:00:04 lr: 0.000591 loss: 0.5783 (0.5783) time: 2.3610 data: 2.2909 max mem: 9671 +[14:56:40.502995] Epoch: [16] [1/2] eta: 0:00:01 lr: 0.000585 loss: 0.4620 (0.5201) time: 1.2129 data: 1.1455 max mem: 9671 +[14:56:40.572711] Epoch: [16] Total time: 0:00:02 (1.2484 s / it) +[14:56:40.573504] Averaged stats: lr: 0.000585 loss: 0.4620 (0.5201) +[14:56:43.367079] val: [0/2] eta: 0:00:05 loss: 0.1574 (0.1574) time: 2.7822 data: 2.7482 max mem: 9671 +[14:56:43.383678] val: [1/2] eta: 0:00:01 loss: 0.1574 (0.9493) time: 1.3991 data: 1.3742 max mem: 9671 +[14:56:43.452137] val: Total time: 0:00:02 (1.4340 s / it) +[14:56:43.461031] val loss: 0.949261486530304 +[14:56:43.461203] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8719, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8168, Kappa: 0.0000, Score: 0.4348 +[14:56:45.157082] Best epoch = 16, Best score = 0.4348 +[14:56:45.243721] log_dir: ./output_logs/retfound +[14:56:47.501277] Epoch: [17] [0/2] eta: 0:00:04 lr: 0.000579 loss: 0.4577 (0.4577) time: 2.2566 data: 2.1861 max mem: 9671 +[14:56:47.567553] Epoch: [17] [1/2] eta: 0:00:01 lr: 0.000572 loss: 0.4577 (0.5255) time: 1.1611 data: 1.0931 max mem: 9671 +[14:56:47.648569] Epoch: [17] Total time: 0:00:02 (1.2023 s / it) +[14:56:47.649457] Averaged stats: lr: 0.000572 loss: 0.4577 (0.5255) +[14:56:50.493885] val: [0/2] eta: 0:00:05 loss: 0.1705 (0.1705) time: 2.8205 data: 2.7882 max mem: 9671 +[14:56:50.511398] val: [1/2] eta: 0:00:01 loss: 0.1705 (0.9139) time: 1.4187 data: 1.3941 max mem: 9671 +[14:56:50.592722] val: Total time: 0:00:02 (1.4602 s / it) +[14:56:50.602159] val loss: 0.9138736724853516 +[14:56:50.602330] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8719, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8080, Kappa: 0.0000, Score: 0.4348 +[14:56:50.650049] Best epoch = 16, Best score = 0.4348 +[14:56:50.900694] log_dir: ./output_logs/retfound +[14:56:53.325514] Epoch: [18] [0/2] eta: 0:00:04 lr: 0.000565 loss: 0.6856 (0.6856) time: 2.4238 data: 2.2888 max mem: 9671 +[14:56:53.468902] Epoch: [18] [1/2] eta: 0:00:01 lr: 0.000558 loss: 0.2809 (0.4832) time: 1.2833 data: 1.1445 max mem: 9671 +[14:56:53.538646] Epoch: [18] Total time: 0:00:02 (1.3189 s / it) +[14:56:53.548666] Averaged stats: lr: 0.000558 loss: 0.2809 (0.4832) +[14:56:56.318928] val: [0/2] eta: 0:00:05 loss: 0.1921 (0.1921) time: 2.7482 data: 2.7312 max mem: 9671 +[14:56:56.328782] val: [1/2] eta: 0:00:01 loss: 0.1921 (0.8666) time: 1.3788 data: 1.3657 max mem: 9671 +[14:56:56.398971] val: Total time: 0:00:02 (1.4145 s / it) +[14:56:56.407744] val loss: 0.8665733337402344 +[14:56:56.407941] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8758, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8034, Kappa: 0.0000, Score: 0.4361 +[14:56:58.119123] Best epoch = 18, Best score = 0.4361 +[14:56:58.192088] log_dir: ./output_logs/retfound +[14:57:00.483639] Epoch: [19] [0/2] eta: 0:00:04 lr: 0.000550 loss: 0.5303 (0.5303) time: 2.2906 data: 2.1501 max mem: 9671 +[14:57:00.620039] Epoch: [19] [1/2] eta: 0:00:01 lr: 0.000542 loss: 0.4757 (0.5030) time: 1.2130 data: 1.0751 max mem: 9671 +[14:57:00.689587] Epoch: [19] Total time: 0:00:02 (1.2487 s / it) +[14:57:00.698135] Averaged stats: lr: 0.000542 loss: 0.4757 (0.5030) +[14:57:03.526538] val: [0/2] eta: 0:00:05 loss: 0.2113 (0.2113) time: 2.8082 data: 2.7910 max mem: 9671 +[14:57:03.537945] val: [1/2] eta: 0:00:01 loss: 0.2113 (0.8214) time: 1.4093 data: 1.3958 max mem: 9671 +[14:57:03.609106] val: Total time: 0:00:02 (1.4458 s / it) +[14:57:03.622905] val loss: 0.821393609046936 +[14:57:03.623169] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8797, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8058, Kappa: 0.0000, Score: 0.4374 +[14:57:05.391698] Best epoch = 19, Best score = 0.4374 +[14:57:05.446978] log_dir: ./output_logs/retfound +[14:57:07.751400] Epoch: [20] [0/2] eta: 0:00:04 lr: 0.000534 loss: 0.5622 (0.5622) time: 2.3033 data: 2.1627 max mem: 9671 +[14:57:07.896998] Epoch: [20] [1/2] eta: 0:00:01 lr: 0.000525 loss: 0.5033 (0.5327) time: 1.2241 data: 1.0814 max mem: 9671 +[14:57:07.970578] Epoch: [20] Total time: 0:00:02 (1.2617 s / it) +[14:57:07.978891] Averaged stats: lr: 0.000525 loss: 0.5033 (0.5327) +[14:57:10.749730] val: [0/2] eta: 0:00:05 loss: 0.2366 (0.2366) time: 2.7511 data: 2.7158 max mem: 9671 +[14:57:10.766216] val: [1/2] eta: 0:00:01 loss: 0.2366 (0.7694) time: 1.3835 data: 1.3579 max mem: 9671 +[14:57:10.840254] val: Total time: 0:00:02 (1.4212 s / it) +[14:57:10.849771] val loss: 0.7693771123886108 +[14:57:10.849952] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8750, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7805, Kappa: 0.0000, Score: 0.4358 +[14:57:10.888492] Best epoch = 19, Best score = 0.4374 +[14:57:11.131123] log_dir: ./output_logs/retfound +[14:57:13.338589] Epoch: [21] [0/2] eta: 0:00:04 lr: 0.000516 loss: 0.5044 (0.5044) time: 2.2065 data: 2.1370 max mem: 9671 +[14:57:13.404169] Epoch: [21] [1/2] eta: 0:00:01 lr: 0.000506 loss: 0.5028 (0.5036) time: 1.1357 data: 1.0686 max mem: 9671 +[14:57:13.471906] Epoch: [21] Total time: 0:00:02 (1.1703 s / it) +[14:57:13.472688] Averaged stats: lr: 0.000506 loss: 0.5028 (0.5036) +[14:57:16.148355] val: [0/2] eta: 0:00:05 loss: 0.2534 (0.2534) time: 2.6635 data: 2.6292 max mem: 9671 +[14:57:16.164684] val: [1/2] eta: 0:00:01 loss: 0.2534 (0.7240) time: 1.3397 data: 1.3147 max mem: 9671 +[14:57:16.236285] val: Total time: 0:00:02 (1.3761 s / it) +[14:57:16.245144] val loss: 0.7240272760391235 +[14:57:16.245310] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8719, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7787, Kappa: 0.0000, Score: 0.4348 +[14:57:16.286275] Best epoch = 19, Best score = 0.4374 +[14:57:16.543864] log_dir: ./output_logs/retfound +[14:57:18.794895] Epoch: [22] [0/2] eta: 0:00:04 lr: 0.000496 loss: 0.5097 (0.5097) time: 2.2501 data: 2.1804 max mem: 9671 +[14:57:18.861449] Epoch: [22] [1/2] eta: 0:00:01 lr: 0.000486 loss: 0.4378 (0.4737) time: 1.1580 data: 1.0903 max mem: 9671 +[14:57:18.935603] Epoch: [22] Total time: 0:00:02 (1.1958 s / it) +[14:57:18.936408] Averaged stats: lr: 0.000486 loss: 0.4378 (0.4737) +[14:57:21.587155] val: [0/2] eta: 0:00:05 loss: 0.2322 (0.2322) time: 2.6379 data: 2.6018 max mem: 9671 +[14:57:21.603728] val: [1/2] eta: 0:00:01 loss: 0.2322 (0.7130) time: 1.3269 data: 1.3010 max mem: 9671 +[14:57:21.675527] val: Total time: 0:00:02 (1.3635 s / it) +[14:57:21.684416] val loss: 0.7130428552627563 +[14:57:21.684605] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8688, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7768, Kappa: 0.0000, Score: 0.4337 +[14:57:21.736276] Best epoch = 19, Best score = 0.4374 +[14:57:21.971985] log_dir: ./output_logs/retfound +[14:57:24.290320] Epoch: [23] [0/2] eta: 0:00:04 lr: 0.000476 loss: 0.3182 (0.3182) time: 2.3173 data: 2.1804 max mem: 9671 +[14:57:24.434220] Epoch: [23] [1/2] eta: 0:00:01 lr: 0.000465 loss: 0.3182 (0.4982) time: 1.2302 data: 1.0902 max mem: 9671 +[14:57:24.502213] Epoch: [23] Total time: 0:00:02 (1.2650 s / it) +[14:57:24.510175] Averaged stats: lr: 0.000465 loss: 0.3182 (0.4982) +[14:57:27.271337] val: [0/2] eta: 0:00:05 loss: 0.1810 (0.1810) time: 2.7379 data: 2.7211 max mem: 9671 +[14:57:27.280684] val: [1/2] eta: 0:00:01 loss: 0.1810 (0.7461) time: 1.3734 data: 1.3606 max mem: 9671 +[14:57:27.353472] val: Total time: 0:00:02 (1.4103 s / it) +[14:57:27.362281] val loss: 0.7460571527481079 +[14:57:27.362505] Accuracy: 0.7619, F1 Score: 0.5139, ROC AUC: 0.8625, Hamming Loss: 0.2381, + Jaccard Score: 0.4235, Precision: 0.6375, Recall: 0.5344, + Average Precision: 0.7722, Kappa: 0.0948, Score: 0.4904 +[14:57:29.074486] Best epoch = 23, Best score = 0.4904 +[14:57:29.161754] log_dir: ./output_logs/retfound +[14:57:31.544877] Epoch: [24] [0/2] eta: 0:00:04 lr: 0.000455 loss: 0.5256 (0.5256) time: 2.3821 data: 2.2343 max mem: 9671 +[14:57:31.678607] Epoch: [24] [1/2] eta: 0:00:01 lr: 0.000444 loss: 0.4432 (0.4844) time: 1.2575 data: 1.1172 max mem: 9671 +[14:57:31.754554] Epoch: [24] Total time: 0:00:02 (1.2963 s / it) +[14:57:31.763588] Averaged stats: lr: 0.000444 loss: 0.4432 (0.4844) +[14:57:34.517858] val: [0/2] eta: 0:00:05 loss: 0.1567 (0.1567) time: 2.7318 data: 2.7148 max mem: 9671 +[14:57:34.527243] val: [1/2] eta: 0:00:01 loss: 0.1567 (0.7540) time: 1.3704 data: 1.3575 max mem: 9671 +[14:57:34.595716] val: Total time: 0:00:02 (1.4052 s / it) +[14:57:34.605280] val loss: 0.7539762258529663 +[14:57:34.605515] Accuracy: 0.7381, F1 Score: 0.4995, ROC AUC: 0.8594, Hamming Loss: 0.2619, + Jaccard Score: 0.4075, Precision: 0.5513, Recall: 0.5188, + Average Precision: 0.7936, Kappa: 0.0494, Score: 0.4694 +[14:57:34.639187] Best epoch = 23, Best score = 0.4904 +[14:57:34.905460] log_dir: ./output_logs/retfound +[14:57:37.137909] Epoch: [25] [0/2] eta: 0:00:04 lr: 0.000432 loss: 0.6646 (0.6646) time: 2.2314 data: 2.1617 max mem: 9671 +[14:57:37.203757] Epoch: [25] [1/2] eta: 0:00:01 lr: 0.000421 loss: 0.3588 (0.5117) time: 1.1482 data: 1.0809 max mem: 9671 +[14:57:37.271538] Epoch: [25] Total time: 0:00:02 (1.1830 s / it) +[14:57:37.279844] Averaged stats: lr: 0.000421 loss: 0.3588 (0.5117) +[14:57:39.990436] val: [0/2] eta: 0:00:05 loss: 0.1034 (0.1034) time: 2.6996 data: 2.6642 max mem: 9671 +[14:57:40.006940] val: [1/2] eta: 0:00:01 loss: 0.1034 (0.8886) time: 1.3578 data: 1.3322 max mem: 9671 +[14:57:40.077709] val: Total time: 0:00:02 (1.3938 s / it) +[14:57:40.086430] val loss: 0.8885501623153687 +[14:57:40.086605] Accuracy: 0.7619, F1 Score: 0.5139, ROC AUC: 0.8648, Hamming Loss: 0.2381, + Jaccard Score: 0.4235, Precision: 0.6375, Recall: 0.5344, + Average Precision: 0.7992, Kappa: 0.0948, Score: 0.4912 +[14:57:41.779276] Best epoch = 25, Best score = 0.4912 +[14:57:41.839004] log_dir: ./output_logs/retfound +[14:57:44.082376] Epoch: [26] [0/2] eta: 0:00:04 lr: 0.000409 loss: 0.5470 (0.5470) time: 2.2424 data: 2.1730 max mem: 9671 +[14:57:44.147652] Epoch: [26] [1/2] eta: 0:00:01 lr: 0.000398 loss: 0.4031 (0.4751) time: 1.1535 data: 1.0865 max mem: 9671 +[14:57:44.218040] Epoch: [26] Total time: 0:00:02 (1.1894 s / it) +[14:57:44.218837] Averaged stats: lr: 0.000398 loss: 0.4031 (0.4751) +[14:57:46.815061] val: [0/2] eta: 0:00:05 loss: 0.0833 (0.0833) time: 2.5887 data: 2.5532 max mem: 9671 +[14:57:46.831619] val: [1/2] eta: 0:00:01 loss: 0.0833 (1.0141) time: 1.3024 data: 1.2767 max mem: 9671 +[14:57:46.903345] val: Total time: 0:00:02 (1.3389 s / it) +[14:57:46.911980] val loss: 1.0140803456306458 +[14:57:46.912213] Accuracy: 0.7619, F1 Score: 0.5139, ROC AUC: 0.8625, Hamming Loss: 0.2381, + Jaccard Score: 0.4235, Precision: 0.6375, Recall: 0.5344, + Average Precision: 0.7980, Kappa: 0.0948, Score: 0.4904 +[14:57:46.959603] Best epoch = 25, Best score = 0.4912 +[14:57:47.220675] log_dir: ./output_logs/retfound +[14:57:49.513500] Epoch: [27] [0/2] eta: 0:00:04 lr: 0.000386 loss: 0.4127 (0.4127) time: 2.2918 data: 2.1544 max mem: 9671 +[14:57:49.654351] Epoch: [27] [1/2] eta: 0:00:01 lr: 0.000374 loss: 0.4127 (0.4861) time: 1.2159 data: 1.0773 max mem: 9671 +[14:57:49.730509] Epoch: [27] Total time: 0:00:02 (1.2548 s / it) +[14:57:49.739707] Averaged stats: lr: 0.000374 loss: 0.4127 (0.4861) +[14:57:52.503509] val: [0/2] eta: 0:00:05 loss: 0.1088 (0.1088) time: 2.7560 data: 2.7364 max mem: 9671 +[14:57:52.514401] val: [1/2] eta: 0:00:01 loss: 0.1088 (0.9593) time: 1.3831 data: 1.3683 max mem: 9671 +[14:57:52.585643] val: Total time: 0:00:02 (1.4195 s / it) +[14:57:52.599092] val loss: 0.9593358635902405 +[14:57:52.599383] Accuracy: 0.7381, F1 Score: 0.4995, ROC AUC: 0.8688, Hamming Loss: 0.2619, + Jaccard Score: 0.4075, Precision: 0.5513, Recall: 0.5188, + Average Precision: 0.8018, Kappa: 0.0494, Score: 0.4725 +[14:57:52.630579] Best epoch = 25, Best score = 0.4912 +[14:57:52.897889] log_dir: ./output_logs/retfound +[14:57:55.416275] Epoch: [28] [0/2] eta: 0:00:05 lr: 0.000362 loss: 0.3604 (0.3604) time: 2.5173 data: 2.3807 max mem: 9671 +[14:57:55.559302] Epoch: [28] [1/2] eta: 0:00:01 lr: 0.000350 loss: 0.3604 (0.3860) time: 1.3298 data: 1.1904 max mem: 9671 +[14:57:55.633992] Epoch: [28] Total time: 0:00:02 (1.3680 s / it) +[14:57:55.643182] Averaged stats: lr: 0.000350 loss: 0.3604 (0.3860) +[14:57:58.336361] val: [0/2] eta: 0:00:05 loss: 0.1265 (0.1265) time: 2.6703 data: 2.6534 max mem: 9671 +[14:57:58.346125] val: [1/2] eta: 0:00:01 loss: 0.1265 (0.9263) time: 1.3398 data: 1.3267 max mem: 9671 +[14:57:58.423808] val: Total time: 0:00:02 (1.3792 s / it) +[14:57:58.432546] val loss: 0.9262599945068359 +[14:57:58.432769] Accuracy: 0.7619, F1 Score: 0.5714, ROC AUC: 0.8625, Hamming Loss: 0.2381, + Jaccard Score: 0.4583, Precision: 0.6447, Recall: 0.5687, + Average Precision: 0.7698, Kappa: 0.1732, Score: 0.5357 +[14:58:00.174493] Best epoch = 28, Best score = 0.5357 +[14:58:00.235491] log_dir: ./output_logs/retfound +[14:58:02.722099] Epoch: [29] [0/2] eta: 0:00:04 lr: 0.000337 loss: 0.5876 (0.5876) time: 2.4855 data: 2.3485 max mem: 9671 +[14:58:02.866008] Epoch: [29] [1/2] eta: 0:00:01 lr: 0.000325 loss: 0.4294 (0.5085) time: 1.3142 data: 1.1743 max mem: 9671 +[14:58:02.940757] Epoch: [29] Total time: 0:00:02 (1.3526 s / it) +[14:58:02.949931] Averaged stats: lr: 0.000325 loss: 0.4294 (0.5085) +[14:58:05.624100] val: [0/2] eta: 0:00:05 loss: 0.1795 (0.1795) time: 2.6643 data: 2.6474 max mem: 9671 +[14:58:05.633513] val: [1/2] eta: 0:00:01 loss: 0.1795 (0.7272) time: 1.3366 data: 1.3238 max mem: 9671 +[14:58:05.707994] val: Total time: 0:00:02 (1.3744 s / it) +[14:58:05.717543] val loss: 0.7272065877914429 +[14:58:05.717713] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8781, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.7819, Kappa: 0.2868, Score: 0.5999 +[14:58:07.423116] Best epoch = 29, Best score = 0.5999 +[14:58:07.489175] log_dir: ./output_logs/retfound +[14:58:09.875003] Epoch: [30] [0/2] eta: 0:00:04 lr: 0.000313 loss: 0.3972 (0.3972) time: 2.3849 data: 2.2393 max mem: 9671 +[14:58:10.009030] Epoch: [30] [1/2] eta: 0:00:01 lr: 0.000301 loss: 0.3289 (0.3630) time: 1.2591 data: 1.1197 max mem: 9671 +[14:58:10.084641] Epoch: [30] Total time: 0:00:02 (1.2977 s / it) +[14:58:10.093407] Averaged stats: lr: 0.000301 loss: 0.3289 (0.3630) +[14:58:12.774752] val: [0/2] eta: 0:00:05 loss: 0.2110 (0.2110) time: 2.6697 data: 2.6355 max mem: 9671 +[14:58:12.791462] val: [1/2] eta: 0:00:01 loss: 0.2110 (0.6347) time: 1.3429 data: 1.3178 max mem: 9671 +[14:58:12.872346] val: Total time: 0:00:02 (1.3840 s / it) +[14:58:12.881987] val loss: 0.6347376108169556 +[14:58:12.882179] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8812, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7843, Kappa: 0.3913, Score: 0.6546 +[14:58:14.679315] Best epoch = 30, Best score = 0.6546 +[14:58:14.742595] log_dir: ./output_logs/retfound +[14:58:17.068550] Epoch: [31] [0/2] eta: 0:00:04 lr: 0.000289 loss: 0.2376 (0.2376) time: 2.3249 data: 2.1785 max mem: 9671 +[14:58:17.202448] Epoch: [31] [1/2] eta: 0:00:01 lr: 0.000276 loss: 0.2376 (0.3573) time: 1.2290 data: 1.0893 max mem: 9671 +[14:58:17.278896] Epoch: [31] Total time: 0:00:02 (1.2681 s / it) +[14:58:17.288737] Averaged stats: lr: 0.000276 loss: 0.2376 (0.3573) +[14:58:20.001155] val: [0/2] eta: 0:00:05 loss: 0.2192 (0.2192) time: 2.7000 data: 2.6647 max mem: 9671 +[14:58:20.017502] val: [1/2] eta: 0:00:01 loss: 0.2192 (0.6058) time: 1.3579 data: 1.3324 max mem: 9671 +[14:58:20.100100] val: Total time: 0:00:02 (1.3999 s / it) +[14:58:20.109878] val loss: 0.6057861745357513 +[14:58:20.110053] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8820, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7843, Kappa: 0.3913, Score: 0.6548 +[14:58:22.053666] Best epoch = 31, Best score = 0.6548 +[14:58:22.185155] log_dir: ./output_logs/retfound +[14:58:24.443649] Epoch: [32] [0/2] eta: 0:00:04 lr: 0.000264 loss: 0.4484 (0.4484) time: 2.2575 data: 2.1172 max mem: 9671 +[14:58:24.577900] Epoch: [32] [1/2] eta: 0:00:01 lr: 0.000252 loss: 0.4178 (0.4331) time: 1.1955 data: 1.0586 max mem: 9671 +[14:58:24.672311] Epoch: [32] Total time: 0:00:02 (1.2435 s / it) +[14:58:24.680935] Averaged stats: lr: 0.000252 loss: 0.4178 (0.4331) +[14:58:27.371295] val: [0/2] eta: 0:00:05 loss: 0.2556 (0.2556) time: 2.6777 data: 2.6427 max mem: 9671 +[14:58:27.387852] val: [1/2] eta: 0:00:01 loss: 0.2556 (0.5362) time: 1.3469 data: 1.3214 max mem: 9671 +[14:58:27.455477] val: Total time: 0:00:02 (1.3813 s / it) +[14:58:27.465029] val loss: 0.5361908078193665 +[14:58:27.465215] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.9000, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8206, Kappa: 0.4878, Score: 0.7099 +[14:58:29.204066] Best epoch = 32, Best score = 0.7099 +[14:58:29.266697] log_dir: ./output_logs/retfound +[14:58:31.516555] Epoch: [33] [0/2] eta: 0:00:04 lr: 0.000240 loss: 0.4264 (0.4264) time: 2.2488 data: 2.1086 max mem: 9671 +[14:58:31.689975] Epoch: [33] [1/2] eta: 0:00:01 lr: 0.000228 loss: 0.4264 (0.4362) time: 1.2108 data: 1.0710 max mem: 9671 +[14:58:31.767611] Epoch: [33] Total time: 0:00:02 (1.2504 s / it) +[14:58:31.776651] Averaged stats: lr: 0.000228 loss: 0.4264 (0.4362) +[14:58:34.428867] val: [0/2] eta: 0:00:05 loss: 0.2724 (0.2724) time: 2.6293 data: 2.6046 max mem: 9671 +[14:58:34.445412] val: [1/2] eta: 0:00:01 loss: 0.2724 (0.4986) time: 1.3227 data: 1.3024 max mem: 9671 +[14:58:34.513098] val: Total time: 0:00:02 (1.3572 s / it) +[14:58:34.521927] val loss: 0.49860239028930664 +[14:58:34.522105] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.9031, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.8230, Kappa: 0.5243, Score: 0.7298 +[14:58:36.273487] Best epoch = 33, Best score = 0.7298 +[14:58:36.341919] log_dir: ./output_logs/retfound +[14:58:38.703632] Epoch: [34] [0/2] eta: 0:00:04 lr: 0.000217 loss: 0.4632 (0.4632) time: 2.3608 data: 2.2249 max mem: 9671 +[14:58:38.846902] Epoch: [34] [1/2] eta: 0:00:01 lr: 0.000205 loss: 0.3432 (0.4032) time: 1.2517 data: 1.1125 max mem: 9671 +[14:58:38.920817] Epoch: [34] Total time: 0:00:02 (1.2894 s / it) +[14:58:38.929483] Averaged stats: lr: 0.000205 loss: 0.3432 (0.4032) +[14:58:41.619438] val: [0/2] eta: 0:00:05 loss: 0.2555 (0.2555) time: 2.6667 data: 2.6317 max mem: 9671 +[14:58:41.635724] val: [1/2] eta: 0:00:01 loss: 0.2555 (0.5101) time: 1.3412 data: 1.3159 max mem: 9671 +[14:58:41.709329] val: Total time: 0:00:02 (1.3787 s / it) +[14:58:41.718134] val loss: 0.5100588202476501 +[14:58:41.718310] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.9031, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.8230, Kappa: 0.5243, Score: 0.7298 +[14:58:41.756082] Best epoch = 33, Best score = 0.7298 +[14:58:42.005454] log_dir: ./output_logs/retfound +[14:58:44.126711] Epoch: [35] [0/2] eta: 0:00:04 lr: 0.000194 loss: 0.5314 (0.5314) time: 2.1205 data: 2.0515 max mem: 9671 +[14:58:44.373845] Epoch: [35] [1/2] eta: 0:00:01 lr: 0.000182 loss: 0.2928 (0.4121) time: 1.1835 data: 1.1165 max mem: 9671 +[14:58:44.453774] Epoch: [35] Total time: 0:00:02 (1.2241 s / it) +[14:58:44.454591] Averaged stats: lr: 0.000182 loss: 0.2928 (0.4121) +[14:58:47.106772] val: [0/2] eta: 0:00:05 loss: 0.2247 (0.2247) time: 2.6286 data: 2.5924 max mem: 9671 +[14:58:47.123373] val: [1/2] eta: 0:00:01 loss: 0.2247 (0.5541) time: 1.3223 data: 1.2963 max mem: 9671 +[14:58:47.195350] val: Total time: 0:00:02 (1.3589 s / it) +[14:58:47.204406] val loss: 0.5540881752967834 +[14:58:47.204594] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9031, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8230, Kappa: 0.3913, Score: 0.6619 +[14:58:47.244065] Best epoch = 33, Best score = 0.7298 +[14:58:47.492602] log_dir: ./output_logs/retfound +[14:58:49.820241] Epoch: [36] [0/2] eta: 0:00:04 lr: 0.000171 loss: 0.5297 (0.5297) time: 2.3267 data: 2.1892 max mem: 9671 +[14:58:49.963574] Epoch: [36] [1/2] eta: 0:00:01 lr: 0.000161 loss: 0.4242 (0.4770) time: 1.2347 data: 1.0946 max mem: 9671 +[14:58:50.034745] Epoch: [36] Total time: 0:00:02 (1.2710 s / it) +[14:58:50.044281] Averaged stats: lr: 0.000161 loss: 0.4242 (0.4770) +[14:58:52.708335] val: [0/2] eta: 0:00:05 loss: 0.1878 (0.1878) time: 2.6397 data: 2.6055 max mem: 9671 +[14:58:52.724818] val: [1/2] eta: 0:00:01 loss: 0.1878 (0.6192) time: 1.3278 data: 1.3028 max mem: 9671 +[14:58:52.796648] val: Total time: 0:00:02 (1.3644 s / it) +[14:58:52.805390] val loss: 0.6192449331283569 +[14:58:52.805566] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9094, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8300, Kappa: 0.3913, Score: 0.6640 +[14:58:52.851518] Best epoch = 33, Best score = 0.7298 +[14:58:53.107755] log_dir: ./output_logs/retfound +[14:58:55.539657] Epoch: [37] [0/2] eta: 0:00:04 lr: 0.000150 loss: 0.5286 (0.5286) time: 2.4311 data: 2.2833 max mem: 9671 +[14:58:55.685054] Epoch: [37] [1/2] eta: 0:00:01 lr: 0.000140 loss: 0.3532 (0.4409) time: 1.2878 data: 1.1417 max mem: 9671 +[14:58:55.752256] Epoch: [37] Total time: 0:00:02 (1.3222 s / it) +[14:58:55.760338] Averaged stats: lr: 0.000140 loss: 0.3532 (0.4409) +[14:58:58.167366] val: [0/2] eta: 0:00:04 loss: 0.1770 (0.1770) time: 2.3837 data: 2.3668 max mem: 9671 +[14:58:58.176683] val: [1/2] eta: 0:00:01 loss: 0.1770 (0.6298) time: 1.1963 data: 1.1834 max mem: 9671 +[14:58:58.245288] val: Total time: 0:00:02 (1.2311 s / it) +[14:58:58.254149] val loss: 0.6298291683197021 +[14:58:58.254337] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9125, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8314, Kappa: 0.3913, Score: 0.6650 +[14:58:58.287678] Best epoch = 33, Best score = 0.7298 +[14:58:58.542471] log_dir: ./output_logs/retfound +[14:59:00.589367] Epoch: [38] [0/2] eta: 0:00:04 lr: 0.000130 loss: 0.5269 (0.5269) time: 2.0460 data: 1.9079 max mem: 9671 +[14:59:00.799030] Epoch: [38] [1/2] eta: 0:00:01 lr: 0.000120 loss: 0.3183 (0.4226) time: 1.1275 data: 0.9915 max mem: 9671 +[14:59:00.867774] Epoch: [38] Total time: 0:00:02 (1.1626 s / it) +[14:59:00.875600] Averaged stats: lr: 0.000120 loss: 0.3183 (0.4226) +[14:59:03.302060] val: [0/2] eta: 0:00:04 loss: 0.1686 (0.1686) time: 2.4150 data: 2.3837 max mem: 9671 +[14:59:03.319394] val: [1/2] eta: 0:00:01 loss: 0.1686 (0.6357) time: 1.2159 data: 1.1919 max mem: 9671 +[14:59:03.394142] val: Total time: 0:00:02 (1.2539 s / it) +[14:59:03.402841] val loss: 0.6357086896896362 +[14:59:03.403020] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8379, Kappa: 0.3913, Score: 0.6671 +[14:59:03.441115] Best epoch = 33, Best score = 0.7298 +[14:59:03.695717] log_dir: ./output_logs/retfound +[14:59:05.862297] Epoch: [39] [0/2] eta: 0:00:04 lr: 0.000110 loss: 0.4934 (0.4934) time: 2.1658 data: 2.0959 max mem: 9671 +[14:59:05.927365] Epoch: [39] [1/2] eta: 0:00:01 lr: 0.000101 loss: 0.4185 (0.4560) time: 1.1151 data: 1.0480 max mem: 9671 +[14:59:05.996380] Epoch: [39] Total time: 0:00:02 (1.1503 s / it) +[14:59:05.997126] Averaged stats: lr: 0.000101 loss: 0.4185 (0.4560) +[14:59:08.521940] val: [0/2] eta: 0:00:05 loss: 0.1713 (0.1713) time: 2.5064 data: 2.4700 max mem: 9671 +[14:59:08.538379] val: [1/2] eta: 0:00:01 loss: 0.1713 (0.6205) time: 1.2612 data: 1.2351 max mem: 9671 +[14:59:08.606625] val: Total time: 0:00:02 (1.2959 s / it) +[14:59:08.615346] val loss: 0.6205177307128906 +[14:59:08.615516] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:08.656138] Best epoch = 33, Best score = 0.7298 +[14:59:08.906591] log_dir: ./output_logs/retfound +[14:59:11.106142] Epoch: [40] [0/2] eta: 0:00:04 lr: 0.000092 loss: 0.3863 (0.3863) time: 2.1987 data: 2.0631 max mem: 9671 +[14:59:11.249096] Epoch: [40] [1/2] eta: 0:00:01 lr: 0.000084 loss: 0.3739 (0.3801) time: 1.1705 data: 1.0316 max mem: 9671 +[14:59:11.314163] Epoch: [40] Total time: 0:00:02 (1.2037 s / it) +[14:59:11.323213] Averaged stats: lr: 0.000084 loss: 0.3739 (0.3801) +[14:59:13.924461] val: [0/2] eta: 0:00:05 loss: 0.1719 (0.1719) time: 2.5781 data: 2.5611 max mem: 9671 +[14:59:13.933984] val: [1/2] eta: 0:00:01 loss: 0.1719 (0.6146) time: 1.2936 data: 1.2806 max mem: 9671 +[14:59:14.002190] val: Total time: 0:00:02 (1.3283 s / it) +[14:59:14.010970] val loss: 0.6146293878555298 +[14:59:14.011135] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:14.045874] Best epoch = 33, Best score = 0.7298 +[14:59:14.296019] log_dir: ./output_logs/retfound +[14:59:16.329872] Epoch: [41] [0/2] eta: 0:00:04 lr: 0.000076 loss: 0.2726 (0.2726) time: 2.0330 data: 1.8870 max mem: 9671 +[14:59:16.529958] Epoch: [41] [1/2] eta: 0:00:01 lr: 0.000068 loss: 0.2726 (0.3406) time: 1.1162 data: 0.9728 max mem: 9671 +[14:59:16.600935] Epoch: [41] Total time: 0:00:02 (1.1524 s / it) +[14:59:16.610395] Averaged stats: lr: 0.000068 loss: 0.2726 (0.3406) +[14:59:19.244091] val: [0/2] eta: 0:00:05 loss: 0.1740 (0.1740) time: 2.6225 data: 2.5866 max mem: 9671 +[14:59:19.260602] val: [1/2] eta: 0:00:01 loss: 0.1740 (0.6065) time: 1.3193 data: 1.2933 max mem: 9671 +[14:59:19.329831] val: Total time: 0:00:02 (1.3545 s / it) +[14:59:19.338600] val loss: 0.6064767241477966 +[14:59:19.338783] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:19.383811] Best epoch = 33, Best score = 0.7298 +[14:59:19.625113] log_dir: ./output_logs/retfound +[14:59:21.828769] Epoch: [42] [0/2] eta: 0:00:04 lr: 0.000061 loss: 0.5306 (0.5306) time: 2.2028 data: 2.1339 max mem: 9671 +[14:59:21.893881] Epoch: [42] [1/2] eta: 0:00:01 lr: 0.000054 loss: 0.2617 (0.3961) time: 1.1336 data: 1.0670 max mem: 9671 +[14:59:21.978594] Epoch: [42] Total time: 0:00:02 (1.1767 s / it) +[14:59:21.979360] Averaged stats: lr: 0.000054 loss: 0.2617 (0.3961) +[14:59:24.452064] val: [0/2] eta: 0:00:04 loss: 0.1783 (0.1783) time: 2.4496 data: 2.4151 max mem: 9671 +[14:59:24.468656] val: [1/2] eta: 0:00:01 loss: 0.1783 (0.5937) time: 1.2328 data: 1.2076 max mem: 9671 +[14:59:24.536673] val: Total time: 0:00:02 (1.2675 s / it) +[14:59:24.545414] val loss: 0.5936670303344727 +[14:59:24.545579] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:24.582810] Best epoch = 33, Best score = 0.7298 +[14:59:24.826490] log_dir: ./output_logs/retfound +[14:59:26.903822] Epoch: [43] [0/2] eta: 0:00:04 lr: 0.000047 loss: 0.3380 (0.3380) time: 2.0764 data: 1.9407 max mem: 9671 +[14:59:27.040506] Epoch: [43] [1/2] eta: 0:00:01 lr: 0.000041 loss: 0.3380 (0.3929) time: 1.1062 data: 0.9704 max mem: 9671 +[14:59:27.111029] Epoch: [43] Total time: 0:00:02 (1.1422 s / it) +[14:59:27.118922] Averaged stats: lr: 0.000041 loss: 0.3380 (0.3929) +[14:59:29.586308] val: [0/2] eta: 0:00:04 loss: 0.1884 (0.1884) time: 2.4557 data: 2.4376 max mem: 9671 +[14:59:29.595984] val: [1/2] eta: 0:00:01 loss: 0.1884 (0.5690) time: 1.2324 data: 1.2189 max mem: 9671 +[14:59:29.667731] val: Total time: 0:00:02 (1.2689 s / it) +[14:59:29.676542] val loss: 0.5689648687839508 +[14:59:29.676773] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:29.721815] Best epoch = 33, Best score = 0.7298 +[14:59:29.966667] log_dir: ./output_logs/retfound +[14:59:32.188286] Epoch: [44] [0/2] eta: 0:00:04 lr: 0.000035 loss: 0.4207 (0.4207) time: 2.2207 data: 2.0783 max mem: 9671 +[14:59:32.332986] Epoch: [44] [1/2] eta: 0:00:01 lr: 0.000030 loss: 0.3855 (0.4031) time: 1.1824 data: 1.0392 max mem: 9671 +[14:59:32.408138] Epoch: [44] Total time: 0:00:02 (1.2207 s / it) +[14:59:32.416889] Averaged stats: lr: 0.000030 loss: 0.3855 (0.4031) +[14:59:34.898401] val: [0/2] eta: 0:00:04 loss: 0.1937 (0.1937) time: 2.4699 data: 2.4530 max mem: 9671 +[14:59:34.907754] val: [1/2] eta: 0:00:01 loss: 0.1937 (0.5564) time: 1.2394 data: 1.2266 max mem: 9671 +[14:59:34.974811] val: Total time: 0:00:02 (1.2735 s / it) +[14:59:34.983525] val loss: 0.5564109683036804 +[14:59:34.983759] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:35.033579] Best epoch = 33, Best score = 0.7298 +[14:59:35.261892] log_dir: ./output_logs/retfound +[14:59:37.320726] Epoch: [45] [0/2] eta: 0:00:04 lr: 0.000025 loss: 0.3846 (0.3846) time: 2.0580 data: 1.9127 max mem: 9671 +[14:59:37.455229] Epoch: [45] [1/2] eta: 0:00:01 lr: 0.000020 loss: 0.3846 (0.5007) time: 1.0959 data: 0.9564 max mem: 9671 +[14:59:37.529460] Epoch: [45] Total time: 0:00:02 (1.1337 s / it) +[14:59:37.538131] Averaged stats: lr: 0.000020 loss: 0.3846 (0.5007) +[14:59:39.976404] val: [0/2] eta: 0:00:04 loss: 0.1963 (0.1963) time: 2.4269 data: 2.3904 max mem: 9671 +[14:59:39.992782] val: [1/2] eta: 0:00:01 loss: 0.1963 (0.5511) time: 1.2214 data: 1.1952 max mem: 9671 +[14:59:40.062628] val: Total time: 0:00:02 (1.2569 s / it) +[14:59:40.071421] val loss: 0.5510595440864563 +[14:59:40.071671] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:40.109391] Best epoch = 33, Best score = 0.7298 +[14:59:40.359800] log_dir: ./output_logs/retfound +[14:59:42.434579] Epoch: [46] [0/2] eta: 0:00:04 lr: 0.000016 loss: 0.3950 (0.3950) time: 2.0739 data: 2.0044 max mem: 9671 +[14:59:42.500036] Epoch: [46] [1/2] eta: 0:00:01 lr: 0.000013 loss: 0.3950 (0.4213) time: 1.0694 data: 1.0022 max mem: 9671 +[14:59:42.571775] Epoch: [46] Total time: 0:00:02 (1.1059 s / it) +[14:59:42.572548] Averaged stats: lr: 0.000013 loss: 0.3950 (0.4213) +[14:59:45.048632] val: [0/2] eta: 0:00:04 loss: 0.1986 (0.1986) time: 2.4646 data: 2.4303 max mem: 9671 +[14:59:45.065135] val: [1/2] eta: 0:00:01 loss: 0.1986 (0.5462) time: 1.2403 data: 1.2152 max mem: 9671 +[14:59:45.135479] val: Total time: 0:00:02 (1.2761 s / it) +[14:59:45.144094] val loss: 0.5462074279785156 +[14:59:45.144318] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:45.182279] Best epoch = 33, Best score = 0.7298 +[14:59:45.422848] log_dir: ./output_logs/retfound +[14:59:47.609242] Epoch: [47] [0/2] eta: 0:00:04 lr: 0.000010 loss: 0.2231 (0.2231) time: 2.1855 data: 2.0479 max mem: 9671 +[14:59:47.752575] Epoch: [47] [1/2] eta: 0:00:01 lr: 0.000007 loss: 0.2231 (0.3859) time: 1.1641 data: 1.0240 max mem: 9671 +[14:59:47.825870] Epoch: [47] Total time: 0:00:02 (1.2014 s / it) +[14:59:47.835096] Averaged stats: lr: 0.000007 loss: 0.2231 (0.3859) +[14:59:50.385247] val: [0/2] eta: 0:00:05 loss: 0.1997 (0.1997) time: 2.5387 data: 2.5216 max mem: 9671 +[14:59:50.394728] val: [1/2] eta: 0:00:01 loss: 0.1997 (0.5437) time: 1.2738 data: 1.2608 max mem: 9671 +[14:59:50.464396] val: Total time: 0:00:02 (1.3093 s / it) +[14:59:50.473072] val loss: 0.5437152981758118 +[14:59:50.473275] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:50.521422] Best epoch = 33, Best score = 0.7298 +[14:59:50.764276] log_dir: ./output_logs/retfound +[14:59:53.091822] Epoch: [48] [0/2] eta: 0:00:04 lr: 0.000005 loss: 0.4238 (0.4238) time: 2.3266 data: 2.1798 max mem: 9671 +[14:59:53.224649] Epoch: [48] [1/2] eta: 0:00:01 lr: 0.000003 loss: 0.3902 (0.4070) time: 1.2294 data: 1.0899 max mem: 9671 +[14:59:53.294323] Epoch: [48] Total time: 0:00:02 (1.2650 s / it) +[14:59:53.302999] Averaged stats: lr: 0.000003 loss: 0.3902 (0.4070) +[14:59:55.746140] val: [0/2] eta: 0:00:04 loss: 0.2003 (0.2003) time: 2.4316 data: 2.4081 max mem: 9671 +[14:59:55.762810] val: [1/2] eta: 0:00:01 loss: 0.2003 (0.5424) time: 1.2239 data: 1.2041 max mem: 9671 +[14:59:55.836565] val: Total time: 0:00:02 (1.2613 s / it) +[14:59:55.845325] val loss: 0.5424394011497498 +[14:59:55.845548] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[14:59:55.881476] Best epoch = 33, Best score = 0.7298 +[14:59:56.112792] log_dir: ./output_logs/retfound +[14:59:58.259886] Epoch: [49] [0/2] eta: 0:00:04 lr: 0.000002 loss: 0.3958 (0.3958) time: 2.1462 data: 2.0748 max mem: 9671 +[14:59:58.325074] Epoch: [49] [1/2] eta: 0:00:01 lr: 0.000001 loss: 0.3958 (0.4450) time: 1.1054 data: 1.0374 max mem: 9671 +[14:59:58.400428] Epoch: [49] Total time: 0:00:02 (1.1437 s / it) +[14:59:58.401179] Averaged stats: lr: 0.000001 loss: 0.3958 (0.4450) +[15:00:00.958377] val: [0/2] eta: 0:00:05 loss: 0.2005 (0.2005) time: 2.5461 data: 2.5112 max mem: 9671 +[15:00:00.974840] val: [1/2] eta: 0:00:01 loss: 0.2005 (0.5420) time: 1.2810 data: 1.2557 max mem: 9671 +[15:00:01.046055] val: Total time: 0:00:02 (1.3172 s / it) +[15:00:01.054765] val loss: 0.5420194268226624 +[15:00:01.055014] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.9187, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8367, Kappa: 0.3913, Score: 0.6671 +[15:00:01.091510] Best epoch = 33, Best score = 0.7298 +[15:00:04.356208] Test with the best model, epoch = 33: +[15:00:06.758998] test: [0/3] eta: 0:00:07 loss: 0.4429 (0.4429) time: 2.3802 data: 2.3635 max mem: 9671 +[15:00:10.062449] test: [2/3] eta: 0:00:01 loss: 0.4429 (0.5708) time: 1.8943 data: 0.7879 max mem: 9671 +[15:00:10.135447] test: Total time: 0:00:05 (1.9191 s / it) +[15:00:10.144488] val loss: 0.5708357890446981 +[15:00:10.144595] Accuracy: 0.7500, F1 Score: 0.6361, ROC AUC: 0.7424, Hamming Loss: 0.2500, + Jaccard Score: 0.4998, Precision: 0.6270, Recall: 0.6544, + Average Precision: 0.6492, Kappa: 0.2759, Score: 0.5514 +[15:00:10.817155] Training time 0:05:21 +[rank0]:[W701 15:00:11.157221587 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/025/retfound acc=0.7500 auroc=0.7421875 f1_macro=0.6361 qwk=0.27586206896551724 diff --git a/results/downsample/papila/025/vit/confusion_matrix.png b/results/downsample/papila/025/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..0611930805a0287c64c3db6a97cefd874be90a14 --- /dev/null +++ b/results/downsample/papila/025/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60ae560985b4b0f79c26d90d702899331a43f9d6bc7f9f4a75360ac80dc25817 +size 74616 diff --git a/results/downsample/papila/025/vit/log.csv b/results/downsample/papila/025/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..dc4cb3bfdfaa95faf5ae4f191fb6bb4debe6c740 --- /dev/null +++ b/results/downsample/papila/025/vit/log.csv @@ -0,0 +1,20 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8867765069007874,0.5,0.55625,0.3775619783996073,0.0 +1,0.9031227231025696,0.5,0.55625,0.3775619783996073,7.39441178203743e-08 +2,0.7297886610031128,0.5,0.55625,0.3775619783996073,1.478882356407486e-07 +3,0.8079441785812378,0.2619047619047619,0.70625,0.31592001823880883,2.2183235346112292e-07 +4,0.6448674201965332,0.5952380952380952,0.775,0.5408636650536502,2.957764712814972e-07 +5,0.9438716769218445,0.30952380952380953,0.825,0.38724809172495295,3.697205891018715e-07 +6,0.6608027815818787,0.23809523809523808,0.840625,0.34431089743589743,3.6927027871539135e-07 +7,0.7632479071617126,0.2857142857142857,0.828125,0.37257258672699844,3.679215414228887e-07 +8,0.6710209846496582,0.7142857142857143,0.809375,0.6234331232492997,3.6568094813687817e-07 +9,0.6062873601913452,0.6666666666666666,0.834375,0.5958970356014929,3.625594148026254e-07 +10,0.6242647171020508,0.47619047619047616,0.821875,0.4920004730836581,3.585721492167595e-07 +11,0.545183002948761,0.42857142857142855,0.81875,0.46100300954753015,3.537385769364163e-07 +12,0.7380874752998352,0.5476190476190477,0.8343750000000001,0.5418069457235336,3.480822466398767e-07 +13,0.4875471591949463,0.6190476190476191,0.834375,0.57674004095309,3.4163071539977294e-07 +14,0.585501492023468,0.6666666666666666,0.840625,0.5979803689348263,3.344154144278013e-07 +15,0.5873349905014038,0.6190476190476191,0.83125,0.5613470047651464,3.2647149594501757e-07 +16,0.4456626772880554,0.6190476190476191,0.815625,0.5704900409530901,3.178376619237501e-07 +17,0.5656861066818237,0.5714285714285714,0.8125,0.5376344086021505,3.0855597553548053e-07 +18,0.5642045140266418,0.6428571428571429,0.8,0.567423122627018,2.986716562233006e-07 diff --git a/results/downsample/papila/025/vit/metrics.json b/results/downsample/papila/025/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..b5b03f4a862c34627e85ea859a47d6b251d23b5a --- /dev/null +++ b/results/downsample/papila/025/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.6071428571428571, + "balanced_accuracy": 0.5900735294117647, + "precision_macro": 0.5571428571428572, + "recall_macro": 0.5900735294117647, + "f1_macro": 0.5354449472096531, + "precision_weighted": 0.7428571428571429, + "recall_weighted": 0.6071428571428571, + "f1_weighted": 0.6484234719528837, + "cohen_kappa": 0.12389380530973448, + "quadratic_weighted_kappa": 0.12389380530973448, + "mcc": 0.14348601079588785, + "auroc": 0.6443014705882353, + "auprc": 0.4116801309868954, + "sensitivity": 0.5625, + "specificity": 0.6176470588235294, + "precision_pos": 0.2571428571428571, + "f1_pos": 0.35294117647058826, + "per_class": { + "0": { + "precision": 0.8571428571428571, + "recall": 0.6176470588235294, + "f1-score": 0.717948717948718, + "support": 68.0 + }, + "1": { + "precision": 0.2571428571428571, + "recall": 0.5625, + "f1-score": 0.35294117647058826, + "support": 16.0 + }, + "accuracy": 0.6071428571428571, + "macro avg": { + "precision": 0.5571428571428572, + "recall": 0.5900735294117647, + "f1-score": 0.5354449472096531, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7428571428571429, + "recall": 0.6071428571428571, + "f1-score": 0.6484234719528837, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/025/vit/pr.png b/results/downsample/papila/025/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..efb5c8583ee96ec44a1d91c60c4e4ff7175945f0 --- /dev/null +++ b/results/downsample/papila/025/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92838490fa468e27e2b29bce6a9264580a978ea7023a29c448cb4ac8d3aa5e4a +size 51702 diff --git a/results/downsample/papila/025/vit/roc.png b/results/downsample/papila/025/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..5ee6e874f85251eb2255c3e2719cb4170ba9d57c --- /dev/null +++ b/results/downsample/papila/025/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3dea02b8f23077d2413ebff540153b684512d7390fca16a92425bc0e638d3985 +size 57588 diff --git a/results/downsample/papila/025/vit/test_pred.npz b/results/downsample/papila/025/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..c85d064f6a1abc5a3b610ef0c9c80921102c7af7 --- /dev/null +++ b/results/downsample/papila/025/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c49948301ac69e294c7c557f2fcc066b65de861ecb7f783e78bf0432396a5cf6 +size 1854 diff --git a/results/downsample/papila/025/vit/train.log b/results/downsample/papila/025/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..170e82bffd9f30bdaf2b2d4f524cac323d50d159 --- /dev/null +++ b/results/downsample/papila/025/vit/train.log @@ -0,0 +1,104 @@ +[vit] train=73 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8868 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.9031 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.7298 val_acc=0.5000 val_auc=0.5563 score=0.3776 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.8079 val_acc=0.2619 val_auc=0.7063 score=0.3159 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6449 val_acc=0.5952 val_auc=0.7750 score=0.5409 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.9439 val_acc=0.3095 val_auc=0.8250 score=0.3872 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.6608 val_acc=0.2381 val_auc=0.8406 score=0.3443 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.7632 val_acc=0.2857 val_auc=0.8281 score=0.3726 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.6710 val_acc=0.7143 val_auc=0.8094 score=0.6234 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.6063 val_acc=0.6667 val_auc=0.8344 score=0.5959 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.6243 val_acc=0.4762 val_auc=0.8219 score=0.4920 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.5452 val_acc=0.4286 val_auc=0.8187 score=0.4610 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.7381 val_acc=0.5476 val_auc=0.8344 score=0.5418 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.4875 val_acc=0.6190 val_auc=0.8344 score=0.5767 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.5855 val_acc=0.6667 val_auc=0.8406 score=0.5980 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.5873 val_acc=0.6190 val_auc=0.8313 score=0.5613 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4457 val_acc=0.6190 val_auc=0.8156 score=0.5705 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.5657 val_acc=0.5714 val_auc=0.8125 score=0.5376 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.5642 val_acc=0.6429 val_auc=0.8000 score=0.5674 +[vit] early stop at ep18 (best ep8 score=0.6234) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=8 best_val_score=0.6234 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/025/vit acc=0.6071 auroc=0.6443014705882353 f1_macro=0.5354 qwk=0.12389380530973448 diff --git a/results/downsample/papila/050/papila_050pct/confusion_matrix_test.jpg b/results/downsample/papila/050/papila_050pct/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b2afd6b6cfbfd6473c7d9fb89e2efd8b27a0e1ed --- /dev/null +++ b/results/downsample/papila/050/papila_050pct/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e3cd94cffac2aca509eaeb3e83f69cf66fc1cf96ccd9bcfc0c853064f4dcb31 +size 258766 diff --git a/results/downsample/papila/050/papila_050pct/log.txt b/results/downsample/papila/050/papila_050pct/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..e07a3aa7430bbd39893ab8cb24289f9684723fa5 --- /dev/null +++ b/results/downsample/papila/050/papila_050pct/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.34375e-05, "train_loss": 0.6926918029785156, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.593750000000001e-05, "train_loss": 0.6574020385742188, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014843750000000002, "train_loss": 0.5783367156982422, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021093750000000002, "train_loss": 0.4873313903808594, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002734375, "train_loss": 0.540806770324707, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003359375, "train_loss": 0.5708413124084473, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00039843750000000003, "train_loss": 0.48935413360595703, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004609375, "train_loss": 0.5359764099121094, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005234375, "train_loss": 0.493929386138916, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005859375, "train_loss": 0.47930383682250977, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247895471787854, "train_loss": 0.4681119918823242, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006231077089086987, "train_loss": 0.49184608459472656, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006195139534919123, "train_loss": 0.5033931732177734, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006140304376256159, "train_loss": 0.4772026538848877, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006066909690085789, "train_loss": 0.47235870361328125, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005975407979054466, "train_loss": 0.40575075149536133, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005866363381638103, "train_loss": 0.4771219491958618, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005740448194040781, "train_loss": 0.484890878200531, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005598438725265087, "train_loss": 0.47937941551208496, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005441210510908961, "train_loss": 0.4379233121871948, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005269732915197597, "train_loss": 0.43051815032958984, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005085063154530642, "train_loss": 0.49863606691360474, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004888339779391536, "train_loss": 0.45597130060195923, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00046807756548051713, "train_loss": 0.42668789625167847, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00044636504826216976, "train_loss": 0.4274568557739258, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004238302911729072, "train_loss": 0.4756513237953186, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004006122284837521, "train_loss": 0.4869039058685303, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003768540072719723, "train_loss": 0.4019651412963867, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035270210487174864, "train_loss": 0.44496965408325195, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.000328305425792701, "train_loss": 0.42784976959228516, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030381438367407094, "train_loss": 0.4405871033668518, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002793799739346176, "train_loss": 0.38128840923309326, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002551528428356505, "train_loss": 0.41892170906066895, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.0002312823586967332, "train_loss": 0.463517427444458, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.0002079156909903239, "train_loss": 0.3807825446128845, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018519690299303795, "train_loss": 0.48010849952697754, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016326606358764225, "train_loss": 0.45887988805770874, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014225838369181528, "train_loss": 0.4060739278793335, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012230338263787712, "train_loss": 0.4463691711425781, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010352408964303667, "train_loss": 0.379574179649353, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.60362852933556e-05, "train_loss": 0.38536226749420166, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.994778771793487e-05, "train_loss": 0.39244067668914795, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.535778785429517e-05, "train_loss": 0.32479822635650635, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.2356237903263196e-05, "train_loss": 0.4007827639579773, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.102329674374109e-05, "train_loss": 0.3887771964073181, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.1428835726562727e-05, "train_loss": 0.3725169897079468, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3632007894380988e-05, "train_loss": 0.41518163681030273, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.680883283489666e-06, "train_loss": 0.42567282915115356, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.6121525560646343e-06, "train_loss": 0.48031461238861084, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.4509007900307425e-06, "train_loss": 0.36185771226882935, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/050/papila_050pct/metrics_test.csv b/results/downsample/papila/050/papila_050pct/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..e41d3d419b0f975409037d8781ddfc4257c003f4 --- /dev/null +++ b/results/downsample/papila/050/papila_050pct/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5038713614145914,0.7976190476190477,0.7053847740870642,0.8090533088235294,0.20238095238095238,0.5703203203203203,0.6904761904761905,0.7316176470588236,0.7303672406024357,0.4137931034482759 diff --git a/results/downsample/papila/050/papila_050pct/metrics_val.csv b/results/downsample/papila/050/papila_050pct/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..9df723443182fe31f6ba21b26813b41a34d9d2e9 --- /dev/null +++ b/results/downsample/papila/050/papila_050pct/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.693167120218277,0.7619047619047619,0.43243243243243246,0.60078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.5976375859912446,0.0 +0.6988800168037415,0.7619047619047619,0.43243243243243246,0.69765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.657925853525252,0.0 +0.7387275695800781,0.7619047619047619,0.43243243243243246,0.7945312499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7325213107087549,0.0 +0.8862209320068359,0.7619047619047619,0.43243243243243246,0.83125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8012618299074655,0.0 +1.0715103149414062,0.7619047619047619,0.43243243243243246,0.8140625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7734005402206623,0.0 +1.0457447171211243,0.7619047619047619,0.43243243243243246,0.584375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6046350599584542,0.0 +0.9472068548202515,0.7619047619047619,0.43243243243243246,0.24062500000000003,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.4075970098449705,0.0 +0.996890664100647,0.7619047619047619,0.43243243243243246,0.6687500000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.621096383637135,0.0 +0.9891384243965149,0.7619047619047619,0.43243243243243246,0.765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7126468739365517,0.0 +0.8918132781982422,0.7619047619047619,0.43243243243243246,0.790625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7573208232810256,0.0 +0.9426374435424805,0.7619047619047619,0.43243243243243246,0.79765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7513803184006123,0.0 +0.8659271001815796,0.7619047619047619,0.43243243243243246,0.79375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7585529316028172,0.0 +0.7070209383964539,0.7857142857142857,0.5292652552926526,0.825,0.21428571428571427,0.4402439024390244,0.8902439024390244,0.55,0.7739663759773165,0.1447963800904979 +0.7854663729667664,0.7619047619047619,0.43243243243243246,0.82890625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7719302823496224,0.0 +0.9815189242362976,0.7619047619047619,0.43243243243243246,0.83515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7996472698626722,0.0 +1.1473761796951294,0.7857142857142857,0.5292652552926526,0.83203125,0.21428571428571427,0.4402439024390244,0.8902439024390244,0.55,0.7981918244517225,0.1447963800904979 +0.9500231146812439,0.7380952380952381,0.49945828819068255,0.83125,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.7721278252736342,0.049382716049382824 +0.7822439074516296,0.7619047619047619,0.5714285714285714,0.825,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7678374768122866,0.17322834645669294 +0.6437538862228394,0.7857142857142857,0.6681299385425812,0.8187500000000001,0.21428571428571427,0.535425101214575,0.7,0.653125,0.7608902471433694,0.3414634146341464 +0.7901833653450012,0.7619047619047619,0.5714285714285714,0.821875,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7656390153066914,0.17322834645669294 +0.9415563344955444,0.7857142857142857,0.6347826086956522,0.82265625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7638525568009973,0.2867924528301887 +1.097848653793335,0.8095238095238095,0.6571428571428571,0.8296875,0.19047619047619047,0.5337995337995338,0.7828947368421053,0.634375,0.795656832214718,0.3385826771653543 +0.665777862071991,0.8333333333333334,0.7619433198380567,0.83203125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945817508209687,0.5242718446601942 +0.7156535387039185,0.7857142857142857,0.6681299385425812,0.83984375,0.21428571428571427,0.535425101214575,0.7,0.653125,0.8123745843529664,0.3414634146341464 +1.3100290298461914,0.8095238095238095,0.611111111111111,0.84453125,0.19047619047619047,0.5,0.9,0.6,0.8156212562312035,0.27586206896551724 +1.18218994140625,0.8095238095238095,0.611111111111111,0.8453125,0.19047619047619047,0.5,0.9,0.6,0.8153927517746767,0.27586206896551724 +0.7669225931167603,0.7857142857142857,0.6347826086956522,0.8390625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7945931867563617,0.2867924528301887 +0.8134706616401672,0.7857142857142857,0.6347826086956522,0.84375,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7988651121489767,0.2867924528301887 +0.9359757900238037,0.7857142857142857,0.6347826086956522,0.84921875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8094472119237253,0.2867924528301887 +0.987418532371521,0.7857142857142857,0.6347826086956522,0.846875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8067595915730614,0.2867924528301887 +0.9265840649604797,0.7857142857142857,0.6347826086956522,0.85234375,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8217465815096481,0.2867924528301887 +0.8141542077064514,0.7857142857142857,0.6347826086956522,0.84921875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.812493975809674,0.2867924528301887 +0.7478845715522766,0.8095238095238095,0.7171717171717171,0.84609375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8106461476357767,0.43624161073825496 +0.7601932287216187,0.7857142857142857,0.6681299385425812,0.84375,0.21428571428571427,0.535425101214575,0.7,0.653125,0.8189430087452315,0.3414634146341464 +0.8489300012588501,0.7857142857142857,0.6347826086956522,0.84453125,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8198383928159252,0.2867924528301887 +0.9358091950416565,0.7857142857142857,0.6347826086956522,0.8390625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8108210178658877,0.2867924528301887 +0.8780083656311035,0.7857142857142857,0.6347826086956522,0.840625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8152734613808312,0.2867924528301887 +0.765245795249939,0.8095238095238095,0.6911764705882353,0.83984375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8164486866931533,0.3913043478260869 +0.6599004864692688,0.8095238095238095,0.7171717171717171,0.8335937499999999,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8119813970106735,0.43624161073825496 +0.6236768960952759,0.8095238095238095,0.7171717171717171,0.83125,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8112890893183657,0.43624161073825496 +0.6044119298458099,0.8333333333333334,0.7777777777777777,0.828125,0.16666666666666666,0.65,0.7697947214076246,0.7875,0.8095983052461162,0.5558912386706949 +0.6177876591682434,0.7857142857142857,0.6939271255060728,0.828125,0.21428571428571427,0.556949806949807,0.702020202020202,0.6875,0.809672710008021,0.38834951456310673 +0.644670307636261,0.8095238095238095,0.7171717171717171,0.834375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8181718028269667,0.43624161073825496 +0.6628798842430115,0.8095238095238095,0.7171717171717171,0.8375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8198625868992162,0.43624161073825496 +0.678103506565094,0.8333333333333334,0.7418788410886743,0.8382812500000001,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8198625868992163,0.4878048780487805 +0.6932103633880615,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7041279077529907,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7076523303985596,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7084973454475403,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7082802057266235,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 diff --git a/results/downsample/papila/050/papila_050pct/test_pred.npz b/results/downsample/papila/050/papila_050pct/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..f73fa88d5a957325c49463807e7b54e237b314bb --- /dev/null +++ b/results/downsample/papila/050/papila_050pct/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66b5fbea2218dec79001fdc668e5ab5bed1644c25466a0d9a09f003af23aa2de +size 1518 diff --git a/results/downsample/papila/050/resnet/confusion_matrix.png b/results/downsample/papila/050/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..11a43ed4be817521440f6ed250a510a34c40d212 --- /dev/null +++ b/results/downsample/papila/050/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5e6ff5b81a7043f7d4b2ea567d5a14cb30ae889a3a883ff6b707694813ce8d9 +size 71212 diff --git a/results/downsample/papila/050/resnet/log.csv b/results/downsample/papila/050/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..1735a9e727c894f7e18a36351bc0bd46e35e2931 --- /dev/null +++ b/results/downsample/papila/050/resnet/log.csv @@ -0,0 +1,19 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6965464651584625,0.2857142857142857,0.590625,0.2934059200603318,8.333333333333333e-05 +1,0.6973766088485718,0.3333333333333333,0.484375,0.2891712373201227,0.00025 +2,0.6939314007759094,0.30952380952380953,0.45625,0.24623449807273343,0.0004166666666666667 +3,0.6797415912151337,0.47619047619047616,0.475,0.2767370083261556,0.0004998603909325636 +4,0.6749887764453888,0.5,0.49374999999999997,0.2960886589955185,0.0004987444537721261 +5,0.65753173828125,0.5476190476190477,0.49687499999999996,0.2972860169491525,0.0004965175634975071 +6,0.6528246402740479,0.6428571428571429,0.603125,0.44533545197740115,0.0004931896659412593 +7,0.6279943883419037,0.6190476190476191,0.709375,0.5043882787826586,0.0004887756243017281 +8,0.6017423272132874,0.47619047619047616,0.7000000000000001,0.42650132970499105,0.00048329515276040064 +9,0.5807418823242188,0.5238095238095238,0.603125,0.3692874318116566,0.0004767727284335852 +10,0.544448733329773,0.6428571428571429,0.6,0.44429378531073455,0.0004692374820516679 +11,0.5186592638492584,0.6666666666666666,0.571875,0.3979166666666667,0.00046072306785419927 +12,0.4823641777038574,0.6904761904761905,0.640625,0.40334587375175773,0.0004512675132818908 +13,0.42710140347480774,0.7380952380952381,0.6796875,0.45387085839941665,0.00044091304913683303 +14,0.4169328808784485,0.7142857142857143,0.6562500000000001,0.4266570758738278,0.0004297059209694824 +15,0.3663795739412308,0.6666666666666666,0.621875,0.38011610794748374,0.0004176961825348059 +16,0.3312646597623825,0.7142857142857143,0.59375,0.44128999728831947,0.0004049374722400612 +17,0.2988763749599457,0.6904761904761905,0.6124999999999999,0.45834272446810215,0.0003914867735826488 diff --git a/results/downsample/papila/050/resnet/metrics.json b/results/downsample/papila/050/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..d02419ae864e9c3c4d342da3ee6be57477f2ffde --- /dev/null +++ b/results/downsample/papila/050/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7142857142857143, + "balanced_accuracy": 0.7040441176470589, + "precision_macro": 0.6370370370370371, + "recall_macro": 0.7040441176470589, + "f1_macro": 0.6407697790449038, + "precision_weighted": 0.8044091710758378, + "recall_weighted": 0.7142857142857143, + "f1_weighted": 0.7413705325323287, + "cohen_kappa": 0.30578512396694213, + "quadratic_weighted_kappa": 0.30578512396694213, + "mcc": 0.33443445580376163, + "auroc": 0.7141544117647058, + "auprc": 0.3752473287062459, + "sensitivity": 0.6875, + "specificity": 0.7205882352941176, + "precision_pos": 0.36666666666666664, + "f1_pos": 0.4782608695652174, + "per_class": { + "0": { + "precision": 0.9074074074074074, + "recall": 0.7205882352941176, + "f1-score": 0.8032786885245902, + "support": 68.0 + }, + "1": { + "precision": 0.36666666666666664, + "recall": 0.6875, + "f1-score": 0.4782608695652174, + "support": 16.0 + }, + "accuracy": 0.7142857142857143, + "macro avg": { + "precision": 0.6370370370370371, + "recall": 0.7040441176470589, + "f1-score": 0.6407697790449038, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8044091710758378, + "recall": 0.7142857142857143, + "f1-score": 0.7413705325323287, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/050/resnet/pr.png b/results/downsample/papila/050/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..cefd0ff3a46b5d4fce7d7a89304a7bb2a5d770e9 --- /dev/null +++ b/results/downsample/papila/050/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7644e5dc8dc16651ccec6deeb536c8c6b317f572781bdcafff3bf3cdc6e6b9c4 +size 51940 diff --git a/results/downsample/papila/050/resnet/roc.png b/results/downsample/papila/050/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..51101cd7f6601eadae9e64331badb4c94055e932 --- /dev/null +++ b/results/downsample/papila/050/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:11729f89ce3174854f6314b9c661ca5e12625b4e124ce0b036c770156a906369 +size 57157 diff --git a/results/downsample/papila/050/resnet/test_pred.npz b/results/downsample/papila/050/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..4a04191dbb970ab1899e76fa9aa2c89b0d4cf65b --- /dev/null +++ b/results/downsample/papila/050/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bbd8aec31d19132de5aec3756edb683e4785a590e492977b5b763bf2905676df +size 1854 diff --git a/results/downsample/papila/050/resnet/train.log b/results/downsample/papila/050/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..feab464987591560392c5bd8a5ef692a0cbcc037 --- /dev/null +++ b/results/downsample/papila/050/resnet/train.log @@ -0,0 +1,99 @@ +[resnet] train=146 val=42 test=84 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6965 val_acc=0.2857 val_auc=0.5906 score=0.2934 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6974 val_acc=0.3333 val_auc=0.4844 score=0.2892 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6939 val_acc=0.3095 val_auc=0.4562 score=0.2462 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6797 val_acc=0.4762 val_auc=0.4750 score=0.2767 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6750 val_acc=0.5000 val_auc=0.4937 score=0.2961 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6575 val_acc=0.5476 val_auc=0.4969 score=0.2973 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.6528 val_acc=0.6429 val_auc=0.6031 score=0.4453 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.6280 val_acc=0.6190 val_auc=0.7094 score=0.5044 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.6017 val_acc=0.4762 val_auc=0.7000 score=0.4265 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.5807 val_acc=0.5238 val_auc=0.6031 score=0.3693 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.5444 val_acc=0.6429 val_auc=0.6000 score=0.4443 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.5187 val_acc=0.6667 val_auc=0.5719 score=0.3979 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.4824 val_acc=0.6905 val_auc=0.6406 score=0.4033 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.4271 val_acc=0.7381 val_auc=0.6797 score=0.4539 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.4169 val_acc=0.7143 val_auc=0.6563 score=0.4267 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.3664 val_acc=0.6667 val_auc=0.6219 score=0.3801 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.3313 val_acc=0.7143 val_auc=0.5938 score=0.4413 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.2989 val_acc=0.6905 val_auc=0.6125 score=0.4583 +[resnet] early stop at ep17 (best ep7 score=0.5044) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=7 best_val_score=0.5044 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/050/resnet acc=0.7143 auroc=0.7141544117647058 f1_macro=0.6408 qwk=0.30578512396694213 diff --git a/results/downsample/papila/050/retfound/confusion_matrix.png b/results/downsample/papila/050/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..18d2896b08393ed44d7fafd34f5c8f8ecbd23daa --- /dev/null +++ b/results/downsample/papila/050/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6096a2cd8467bf5996dac3f74cb5ab1df410f54dfb61365cbf36315196d047bb +size 72547 diff --git a/results/downsample/papila/050/retfound/confusion_matrix_test.jpg b/results/downsample/papila/050/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b2afd6b6cfbfd6473c7d9fb89e2efd8b27a0e1ed --- /dev/null +++ b/results/downsample/papila/050/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e3cd94cffac2aca509eaeb3e83f69cf66fc1cf96ccd9bcfc0c853064f4dcb31 +size 258766 diff --git a/results/downsample/papila/050/retfound/log.txt b/results/downsample/papila/050/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..e07a3aa7430bbd39893ab8cb24289f9684723fa5 --- /dev/null +++ b/results/downsample/papila/050/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.34375e-05, "train_loss": 0.6926918029785156, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 8.593750000000001e-05, "train_loss": 0.6574020385742188, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00014843750000000002, "train_loss": 0.5783367156982422, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021093750000000002, "train_loss": 0.4873313903808594, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002734375, "train_loss": 0.540806770324707, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003359375, "train_loss": 0.5708413124084473, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.00039843750000000003, "train_loss": 0.48935413360595703, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004609375, "train_loss": 0.5359764099121094, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005234375, "train_loss": 0.493929386138916, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005859375, "train_loss": 0.47930383682250977, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247895471787854, "train_loss": 0.4681119918823242, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006231077089086987, "train_loss": 0.49184608459472656, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006195139534919123, "train_loss": 0.5033931732177734, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006140304376256159, "train_loss": 0.4772026538848877, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006066909690085789, "train_loss": 0.47235870361328125, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005975407979054466, "train_loss": 0.40575075149536133, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005866363381638103, "train_loss": 0.4771219491958618, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005740448194040781, "train_loss": 0.484890878200531, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005598438725265087, "train_loss": 0.47937941551208496, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005441210510908961, "train_loss": 0.4379233121871948, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005269732915197597, "train_loss": 0.43051815032958984, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005085063154530642, "train_loss": 0.49863606691360474, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.0004888339779391536, "train_loss": 0.45597130060195923, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.00046807756548051713, "train_loss": 0.42668789625167847, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.00044636504826216976, "train_loss": 0.4274568557739258, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004238302911729072, "train_loss": 0.4756513237953186, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.0004006122284837521, "train_loss": 0.4869039058685303, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003768540072719723, "train_loss": 0.4019651412963867, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035270210487174864, "train_loss": 0.44496965408325195, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.000328305425792701, "train_loss": 0.42784976959228516, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.00030381438367407094, "train_loss": 0.4405871033668518, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002793799739346176, "train_loss": 0.38128840923309326, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002551528428356505, "train_loss": 0.41892170906066895, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.0002312823586967332, "train_loss": 0.463517427444458, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.0002079156909903239, "train_loss": 0.3807825446128845, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018519690299303795, "train_loss": 0.48010849952697754, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016326606358764225, "train_loss": 0.45887988805770874, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014225838369181528, "train_loss": 0.4060739278793335, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012230338263787712, "train_loss": 0.4463691711425781, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010352408964303667, "train_loss": 0.379574179649353, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.60362852933556e-05, "train_loss": 0.38536226749420166, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.994778771793487e-05, "train_loss": 0.39244067668914795, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.535778785429517e-05, "train_loss": 0.32479822635650635, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.2356237903263196e-05, "train_loss": 0.4007827639579773, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.102329674374109e-05, "train_loss": 0.3887771964073181, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.1428835726562727e-05, "train_loss": 0.3725169897079468, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3632007894380988e-05, "train_loss": 0.41518163681030273, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.680883283489666e-06, "train_loss": 0.42567282915115356, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.6121525560646343e-06, "train_loss": 0.48031461238861084, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.4509007900307425e-06, "train_loss": 0.36185771226882935, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/050/retfound/metrics.json b/results/downsample/papila/050/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..6f1b9286daa017305f28ed64c5fec6c6be39f89b --- /dev/null +++ b/results/downsample/papila/050/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.7976190476190477, + "balanced_accuracy": 0.7316176470588236, + "precision_macro": 0.6904761904761905, + "recall_macro": 0.7316176470588236, + "f1_macro": 0.7053847740870642, + "precision_weighted": 0.8231292517006802, + "recall_weighted": 0.7976190476190477, + "f1_weighted": 0.8074312043777693, + "cohen_kappa": 0.4137931034482759, + "quadratic_weighted_kappa": 0.4137931034482759, + "mcc": 0.42008402520840293, + "auroc": 0.8088235294117647, + "auprc": 0.5191885241446983, + "sensitivity": 0.625, + "specificity": 0.8382352941176471, + "precision_pos": 0.47619047619047616, + "f1_pos": 0.5405405405405406, + "per_class": { + "0": { + "precision": 0.9047619047619048, + "recall": 0.8382352941176471, + "f1-score": 0.8702290076335878, + "support": 68.0 + }, + "1": { + "precision": 0.47619047619047616, + "recall": 0.625, + "f1-score": 0.5405405405405406, + "support": 16.0 + }, + "accuracy": 0.7976190476190477, + "macro avg": { + "precision": 0.6904761904761905, + "recall": 0.7316176470588236, + "f1-score": 0.7053847740870642, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8231292517006802, + "recall": 0.7976190476190477, + "f1-score": 0.8074312043777693, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/050/retfound/metrics_test.csv b/results/downsample/papila/050/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..e41d3d419b0f975409037d8781ddfc4257c003f4 --- /dev/null +++ b/results/downsample/papila/050/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.5038713614145914,0.7976190476190477,0.7053847740870642,0.8090533088235294,0.20238095238095238,0.5703203203203203,0.6904761904761905,0.7316176470588236,0.7303672406024357,0.4137931034482759 diff --git a/results/downsample/papila/050/retfound/metrics_val.csv b/results/downsample/papila/050/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..9df723443182fe31f6ba21b26813b41a34d9d2e9 --- /dev/null +++ b/results/downsample/papila/050/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.693167120218277,0.7619047619047619,0.43243243243243246,0.60078125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.5976375859912446,0.0 +0.6988800168037415,0.7619047619047619,0.43243243243243246,0.69765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.657925853525252,0.0 +0.7387275695800781,0.7619047619047619,0.43243243243243246,0.7945312499999999,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7325213107087549,0.0 +0.8862209320068359,0.7619047619047619,0.43243243243243246,0.83125,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.8012618299074655,0.0 +1.0715103149414062,0.7619047619047619,0.43243243243243246,0.8140625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7734005402206623,0.0 +1.0457447171211243,0.7619047619047619,0.43243243243243246,0.584375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6046350599584542,0.0 +0.9472068548202515,0.7619047619047619,0.43243243243243246,0.24062500000000003,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.4075970098449705,0.0 +0.996890664100647,0.7619047619047619,0.43243243243243246,0.6687500000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.621096383637135,0.0 +0.9891384243965149,0.7619047619047619,0.43243243243243246,0.765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7126468739365517,0.0 +0.8918132781982422,0.7619047619047619,0.43243243243243246,0.790625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7573208232810256,0.0 +0.9426374435424805,0.7619047619047619,0.43243243243243246,0.79765625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7513803184006123,0.0 +0.8659271001815796,0.7619047619047619,0.43243243243243246,0.79375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7585529316028172,0.0 +0.7070209383964539,0.7857142857142857,0.5292652552926526,0.825,0.21428571428571427,0.4402439024390244,0.8902439024390244,0.55,0.7739663759773165,0.1447963800904979 +0.7854663729667664,0.7619047619047619,0.43243243243243246,0.82890625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7719302823496224,0.0 +0.9815189242362976,0.7619047619047619,0.43243243243243246,0.83515625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7996472698626722,0.0 +1.1473761796951294,0.7857142857142857,0.5292652552926526,0.83203125,0.21428571428571427,0.4402439024390244,0.8902439024390244,0.55,0.7981918244517225,0.1447963800904979 +0.9500231146812439,0.7380952380952381,0.49945828819068255,0.83125,0.2619047619047619,0.40752032520325204,0.5512820512820513,0.51875,0.7721278252736342,0.049382716049382824 +0.7822439074516296,0.7619047619047619,0.5714285714285714,0.825,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7678374768122866,0.17322834645669294 +0.6437538862228394,0.7857142857142857,0.6681299385425812,0.8187500000000001,0.21428571428571427,0.535425101214575,0.7,0.653125,0.7608902471433694,0.3414634146341464 +0.7901833653450012,0.7619047619047619,0.5714285714285714,0.821875,0.23809523809523808,0.4583333333333333,0.6447368421052632,0.56875,0.7656390153066914,0.17322834645669294 +0.9415563344955444,0.7857142857142857,0.6347826086956522,0.82265625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7638525568009973,0.2867924528301887 +1.097848653793335,0.8095238095238095,0.6571428571428571,0.8296875,0.19047619047619047,0.5337995337995338,0.7828947368421053,0.634375,0.795656832214718,0.3385826771653543 +0.665777862071991,0.8333333333333334,0.7619433198380567,0.83203125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945817508209687,0.5242718446601942 +0.7156535387039185,0.7857142857142857,0.6681299385425812,0.83984375,0.21428571428571427,0.535425101214575,0.7,0.653125,0.8123745843529664,0.3414634146341464 +1.3100290298461914,0.8095238095238095,0.611111111111111,0.84453125,0.19047619047619047,0.5,0.9,0.6,0.8156212562312035,0.27586206896551724 +1.18218994140625,0.8095238095238095,0.611111111111111,0.8453125,0.19047619047619047,0.5,0.9,0.6,0.8153927517746767,0.27586206896551724 +0.7669225931167603,0.7857142857142857,0.6347826086956522,0.8390625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7945931867563617,0.2867924528301887 +0.8134706616401672,0.7857142857142857,0.6347826086956522,0.84375,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.7988651121489767,0.2867924528301887 +0.9359757900238037,0.7857142857142857,0.6347826086956522,0.84921875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8094472119237253,0.2867924528301887 +0.987418532371521,0.7857142857142857,0.6347826086956522,0.846875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8067595915730614,0.2867924528301887 +0.9265840649604797,0.7857142857142857,0.6347826086956522,0.85234375,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8217465815096481,0.2867924528301887 +0.8141542077064514,0.7857142857142857,0.6347826086956522,0.84921875,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.812493975809674,0.2867924528301887 +0.7478845715522766,0.8095238095238095,0.7171717171717171,0.84609375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8106461476357767,0.43624161073825496 +0.7601932287216187,0.7857142857142857,0.6681299385425812,0.84375,0.21428571428571427,0.535425101214575,0.7,0.653125,0.8189430087452315,0.3414634146341464 +0.8489300012588501,0.7857142857142857,0.6347826086956522,0.84453125,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8198383928159252,0.2867924528301887 +0.9358091950416565,0.7857142857142857,0.6347826086956522,0.8390625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8108210178658877,0.2867924528301887 +0.8780083656311035,0.7857142857142857,0.6347826086956522,0.840625,0.21428571428571427,0.5096153846153846,0.7054054054054054,0.61875,0.8152734613808312,0.2867924528301887 +0.765245795249939,0.8095238095238095,0.6911764705882353,0.83984375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8164486866931533,0.3913043478260869 +0.6599004864692688,0.8095238095238095,0.7171717171717171,0.8335937499999999,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8119813970106735,0.43624161073825496 +0.6236768960952759,0.8095238095238095,0.7171717171717171,0.83125,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8112890893183657,0.43624161073825496 +0.6044119298458099,0.8333333333333334,0.7777777777777777,0.828125,0.16666666666666666,0.65,0.7697947214076246,0.7875,0.8095983052461162,0.5558912386706949 +0.6177876591682434,0.7857142857142857,0.6939271255060728,0.828125,0.21428571428571427,0.556949806949807,0.702020202020202,0.6875,0.809672710008021,0.38834951456310673 +0.644670307636261,0.8095238095238095,0.7171717171717171,0.834375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8181718028269667,0.43624161073825496 +0.6628798842430115,0.8095238095238095,0.7171717171717171,0.8375,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.8198625868992162,0.43624161073825496 +0.678103506565094,0.8333333333333334,0.7418788410886743,0.8382812500000001,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8198625868992163,0.4878048780487805 +0.6932103633880615,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7041279077529907,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7076523303985596,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7084973454475403,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 +0.7082802057266235,0.8095238095238095,0.6911764705882353,0.840625,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8213718880555763,0.3913043478260869 diff --git a/results/downsample/papila/050/retfound/pr.png b/results/downsample/papila/050/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..5edffd9f2762342ee6d32335da29232157d95f3a --- /dev/null +++ b/results/downsample/papila/050/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8f73de607ff4764736680f1792692871748fb3d6e98c0af6ec200534f3e52c1 +size 52683 diff --git a/results/downsample/papila/050/retfound/roc.png b/results/downsample/papila/050/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..cf8887999d8a197dc7241a0d8510106348c69412 --- /dev/null +++ b/results/downsample/papila/050/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e5748bc5ae317c32ae1027d77ec255de437f0b8a58a074fce5923c177cf29bda +size 57732 diff --git a/results/downsample/papila/050/retfound/test_pred.npz b/results/downsample/papila/050/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..f73fa88d5a957325c49463807e7b54e237b314bb --- /dev/null +++ b/results/downsample/papila/050/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66b5fbea2218dec79001fdc668e5ab5bed1644c25466a0d9a09f003af23aa2de +size 1518 diff --git a/results/downsample/papila/050/retfound/train.log b/results/downsample/papila/050/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..2ab6d7b56c07cdefd111dd9d493acf31af6dea5c --- /dev/null +++ b/results/downsample/papila/050/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:53:42.387098626 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:53:42.827837] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:53:42.828117] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/papila_50', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/050', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:53:45.822981] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:53:47.312365] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:53:47.841544] Sampler_train = +[14:53:47.885815] len of train_set: 128 +[14:53:48.098275] [Adaptation] Full fine-tuning: training all parameters. +[14:53:48.099271] number of trainable params (M): 303.30 +[14:53:48.099348] base lr: 5.00e-03 +[14:53:48.099407] actual lr: 6.25e-04 +[14:53:48.099468] accumulate grad iterations: 1 +[14:53:48.099527] effective batch size: 32 +[14:53:48.102420] criterion = CrossEntropyLoss() +[14:53:48.102509] Start training for 50 epochs +[14:53:48.104521] log_dir: ./output_logs/retfound +[14:53:50.820120] Epoch: [0] [0/4] eta: 0:00:10 lr: 0.000000 loss: 0.6928 (0.6928) time: 2.7147 data: 2.1865 max mem: 7340 +[14:53:51.043030] Epoch: [0] [3/4] eta: 0:00:00 lr: 0.000047 loss: 0.6927 (0.6927) time: 0.7342 data: 0.5467 max mem: 7340 +[14:53:51.108359] Epoch: [0] Total time: 0:00:03 (0.7509 s / it) +[14:53:51.109349] Averaged stats: lr: 0.000047 loss: 0.6927 (0.6927) +[14:53:53.801759] val: [0/2] eta: 0:00:05 loss: 0.6772 (0.6772) time: 2.6692 data: 2.6456 max mem: 7340 +[14:53:53.832840] val: [1/2] eta: 0:00:01 loss: 0.6772 (0.6932) time: 1.3499 data: 1.3228 max mem: 7340 +[14:53:53.903073] val: Total time: 0:00:02 (1.3855 s / it) +[14:53:53.915622] val loss: 0.693167120218277 +[14:53:53.915866] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6008, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.5976, Kappa: 0.0000, Score: 0.3444 +[14:53:55.887062] Best epoch = 0, Best score = 0.3444 +[14:53:55.966544] log_dir: ./output_logs/retfound +[14:53:58.325543] Epoch: [1] [0/4] eta: 0:00:09 lr: 0.000063 loss: 0.6795 (0.6795) time: 2.3580 data: 2.2425 max mem: 9657 +[14:53:58.521657] Epoch: [1] [3/4] eta: 0:00:00 lr: 0.000109 loss: 0.6510 (0.6574) time: 0.6383 data: 0.5607 max mem: 9657 +[14:53:58.596085] Epoch: [1] Total time: 0:00:02 (0.6573 s / it) +[14:53:58.597030] Averaged stats: lr: 0.000109 loss: 0.6510 (0.6574) +[14:54:01.483247] val: [0/2] eta: 0:00:05 loss: 0.5744 (0.5744) time: 2.8703 data: 2.8500 max mem: 9657 +[14:54:01.499840] val: [1/2] eta: 0:00:01 loss: 0.5744 (0.6989) time: 1.4430 data: 1.4251 max mem: 9657 +[14:54:01.624768] val: Total time: 0:00:03 (1.5064 s / it) +[14:54:01.641032] val loss: 0.6988800168037415 +[14:54:01.641242] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6579, Kappa: 0.0000, Score: 0.3767 +[14:54:03.539205] Best epoch = 1, Best score = 0.3767 +[14:54:03.673710] log_dir: ./output_logs/retfound +[14:54:06.020227] Epoch: [2] [0/4] eta: 0:00:09 lr: 0.000125 loss: 0.5967 (0.5967) time: 2.3455 data: 2.2760 max mem: 9657 +[14:54:06.216499] Epoch: [2] [3/4] eta: 0:00:00 lr: 0.000172 loss: 0.5897 (0.5783) time: 0.6353 data: 0.5691 max mem: 9657 +[14:54:06.287693] Epoch: [2] Total time: 0:00:02 (0.6534 s / it) +[14:54:06.288588] Averaged stats: lr: 0.000172 loss: 0.5897 (0.5783) +[14:54:09.116198] val: [0/2] eta: 0:00:05 loss: 0.4057 (0.4057) time: 2.8042 data: 2.7873 max mem: 9657 +[14:54:09.125315] val: [1/2] eta: 0:00:01 loss: 0.4057 (0.7387) time: 1.4064 data: 1.3937 max mem: 9657 +[14:54:09.193406] val: Total time: 0:00:02 (1.4410 s / it) +[14:54:09.202361] val loss: 0.7387275695800781 +[14:54:09.202525] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7945, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7325, Kappa: 0.0000, Score: 0.4090 +[14:54:10.923827] Best epoch = 2, Best score = 0.4090 +[14:54:10.991278] log_dir: ./output_logs/retfound +[14:54:13.284394] Epoch: [3] [0/4] eta: 0:00:09 lr: 0.000188 loss: 0.5060 (0.5060) time: 2.2921 data: 2.2222 max mem: 9657 +[14:54:13.479791] Epoch: [3] [3/4] eta: 0:00:00 lr: 0.000234 loss: 0.4378 (0.4873) time: 0.6217 data: 0.5556 max mem: 9657 +[14:54:13.557646] Epoch: [3] Total time: 0:00:02 (0.6416 s / it) +[14:54:13.558494] Averaged stats: lr: 0.000234 loss: 0.4378 (0.4873) +[14:54:16.088647] val: [0/2] eta: 0:00:05 loss: 0.2114 (0.2114) time: 2.5198 data: 2.5023 max mem: 9657 +[14:54:16.098596] val: [1/2] eta: 0:00:01 loss: 0.2114 (0.8862) time: 1.2646 data: 1.2512 max mem: 9657 +[14:54:16.168467] val: Total time: 0:00:02 (1.3002 s / it) +[14:54:16.178486] val loss: 0.8862209320068359 +[14:54:16.178703] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8313, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.8013, Kappa: 0.0000, Score: 0.4212 +[14:54:17.907789] Best epoch = 3, Best score = 0.4212 +[14:54:17.979579] log_dir: ./output_logs/retfound +[14:54:20.084316] Epoch: [4] [0/4] eta: 0:00:08 lr: 0.000250 loss: 0.6236 (0.6236) time: 2.1038 data: 2.0332 max mem: 9657 +[14:54:20.280860] Epoch: [4] [3/4] eta: 0:00:00 lr: 0.000297 loss: 0.4668 (0.5408) time: 0.5749 data: 0.5084 max mem: 9657 +[14:54:20.349451] Epoch: [4] Total time: 0:00:02 (0.5924 s / it) +[14:54:20.350306] Averaged stats: lr: 0.000297 loss: 0.4668 (0.5408) +[14:54:23.175315] val: [0/2] eta: 0:00:05 loss: 0.1240 (0.1240) time: 2.8130 data: 2.7911 max mem: 9657 +[14:54:23.190951] val: [1/2] eta: 0:00:01 loss: 0.1240 (1.0715) time: 1.4140 data: 1.3956 max mem: 9657 +[14:54:23.258860] val: Total time: 0:00:02 (1.4488 s / it) +[14:54:23.268319] val loss: 1.0715103149414062 +[14:54:23.268530] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8141, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7734, Kappa: 0.0000, Score: 0.4155 +[14:54:23.301328] Best epoch = 3, Best score = 0.4212 +[14:54:23.575893] log_dir: ./output_logs/retfound +[14:54:25.755338] Epoch: [5] [0/4] eta: 0:00:08 lr: 0.000313 loss: 0.5933 (0.5933) time: 2.1785 data: 2.1089 max mem: 9657 +[14:54:25.973210] Epoch: [5] [3/4] eta: 0:00:00 lr: 0.000359 loss: 0.5304 (0.5708) time: 0.5989 data: 0.5326 max mem: 9657 +[14:54:26.054028] Epoch: [5] Total time: 0:00:02 (0.6195 s / it) +[14:54:26.055262] Averaged stats: lr: 0.000359 loss: 0.5304 (0.5708) +[14:54:28.923111] val: [0/2] eta: 0:00:05 loss: 0.1536 (0.1536) time: 2.8410 data: 2.8241 max mem: 9657 +[14:54:28.932045] val: [1/2] eta: 0:00:01 loss: 0.1536 (1.0457) time: 1.4247 data: 1.4121 max mem: 9657 +[14:54:29.002925] val: Total time: 0:00:02 (1.4608 s / it) +[14:54:29.014931] val loss: 1.0457447171211243 +[14:54:29.015180] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.5844, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6046, Kappa: 0.0000, Score: 0.3389 +[14:54:29.061953] Best epoch = 3, Best score = 0.4212 +[14:54:29.298603] log_dir: ./output_logs/retfound +[14:54:31.675233] Epoch: [6] [0/4] eta: 0:00:09 lr: 0.000375 loss: 0.5678 (0.5678) time: 2.3755 data: 2.3063 max mem: 9657 +[14:54:31.870845] Epoch: [6] [3/4] eta: 0:00:00 lr: 0.000422 loss: 0.4436 (0.4894) time: 0.6426 data: 0.5766 max mem: 9657 +[14:54:31.938881] Epoch: [6] Total time: 0:00:02 (0.6600 s / it) +[14:54:31.939644] Averaged stats: lr: 0.000422 loss: 0.4436 (0.4894) +[14:54:34.463387] val: [0/2] eta: 0:00:05 loss: 0.2311 (0.2311) time: 2.5038 data: 2.4868 max mem: 9657 +[14:54:34.472677] val: [1/2] eta: 0:00:01 loss: 0.2311 (0.9472) time: 1.2563 data: 1.2435 max mem: 9657 +[14:54:34.536093] val: Total time: 0:00:02 (1.2887 s / it) +[14:54:34.545031] val loss: 0.9472068548202515 +[14:54:34.545280] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.2406, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.4076, Kappa: 0.0000, Score: 0.2244 +[14:54:34.579149] Best epoch = 3, Best score = 0.4212 +[14:54:34.876610] log_dir: ./output_logs/retfound +[14:54:36.988479] Epoch: [7] [0/4] eta: 0:00:08 lr: 0.000438 loss: 0.3803 (0.3803) time: 2.1110 data: 2.0401 max mem: 9657 +[14:54:37.184476] Epoch: [7] [3/4] eta: 0:00:00 lr: 0.000484 loss: 0.3887 (0.5360) time: 0.5766 data: 0.5101 max mem: 9657 +[14:54:37.254601] Epoch: [7] Total time: 0:00:02 (0.5945 s / it) +[14:54:37.255412] Averaged stats: lr: 0.000484 loss: 0.3887 (0.5360) +[14:54:39.924392] val: [0/2] eta: 0:00:05 loss: 0.1683 (0.1683) time: 2.6575 data: 2.6404 max mem: 9657 +[14:54:39.933809] val: [1/2] eta: 0:00:01 loss: 0.1683 (0.9969) time: 1.3332 data: 1.3203 max mem: 9657 +[14:54:39.995924] val: Total time: 0:00:02 (1.3649 s / it) +[14:54:40.005021] val loss: 0.996890664100647 +[14:54:40.005243] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6688, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6211, Kappa: 0.0000, Score: 0.3671 +[14:54:40.042096] Best epoch = 3, Best score = 0.4212 +[14:54:40.324424] log_dir: ./output_logs/retfound +[14:54:42.569342] Epoch: [8] [0/4] eta: 0:00:08 lr: 0.000500 loss: 0.4144 (0.4144) time: 2.2440 data: 2.1756 max mem: 9657 +[14:54:42.773128] Epoch: [8] [3/4] eta: 0:00:00 lr: 0.000547 loss: 0.4144 (0.4939) time: 0.6118 data: 0.5461 max mem: 9657 +[14:54:42.845549] Epoch: [8] Total time: 0:00:02 (0.6302 s / it) +[14:54:42.846225] Averaged stats: lr: 0.000547 loss: 0.4144 (0.4939) +[14:54:45.528283] val: [0/2] eta: 0:00:05 loss: 0.1522 (0.1522) time: 2.6757 data: 2.6581 max mem: 9657 +[14:54:45.539304] val: [1/2] eta: 0:00:01 loss: 0.1522 (0.9891) time: 1.3431 data: 1.3292 max mem: 9657 +[14:54:45.624962] val: Total time: 0:00:02 (1.3866 s / it) +[14:54:45.635260] val loss: 0.9891384243965149 +[14:54:45.635447] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7656, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7126, Kappa: 0.0000, Score: 0.3994 +[14:54:45.665347] Best epoch = 3, Best score = 0.4212 +[14:54:45.944989] log_dir: ./output_logs/retfound +[14:54:48.265149] Epoch: [9] [0/4] eta: 0:00:09 lr: 0.000562 loss: 0.4039 (0.4039) time: 2.3191 data: 2.2490 max mem: 9657 +[14:54:48.462081] Epoch: [9] [3/4] eta: 0:00:00 lr: 0.000609 loss: 0.4379 (0.4793) time: 0.6288 data: 0.5623 max mem: 9657 +[14:54:48.550675] Epoch: [9] Total time: 0:00:02 (0.6514 s / it) +[14:54:48.551683] Averaged stats: lr: 0.000609 loss: 0.4379 (0.4793) +[14:54:51.099613] val: [0/2] eta: 0:00:05 loss: 0.1830 (0.1830) time: 2.5369 data: 2.5199 max mem: 9657 +[14:54:51.108790] val: [1/2] eta: 0:00:01 loss: 0.1830 (0.8918) time: 1.2728 data: 1.2600 max mem: 9657 +[14:54:51.176209] val: Total time: 0:00:02 (1.3071 s / it) +[14:54:51.184964] val loss: 0.8918132781982422 +[14:54:51.185184] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7906, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7573, Kappa: 0.0000, Score: 0.4077 +[14:54:51.218960] Best epoch = 3, Best score = 0.4212 +[14:54:51.499431] log_dir: ./output_logs/retfound +[14:54:53.629537] Epoch: [10] [0/4] eta: 0:00:08 lr: 0.000625 loss: 0.4874 (0.4874) time: 2.1292 data: 2.0583 max mem: 9657 +[14:54:53.825272] Epoch: [10] [3/4] eta: 0:00:00 lr: 0.000624 loss: 0.4874 (0.4681) time: 0.5811 data: 0.5146 max mem: 9657 +[14:54:53.892066] Epoch: [10] Total time: 0:00:02 (0.5981 s / it) +[14:54:53.892811] Averaged stats: lr: 0.000624 loss: 0.4874 (0.4681) +[14:54:56.685443] val: [0/2] eta: 0:00:05 loss: 0.1485 (0.1485) time: 2.7812 data: 2.7609 max mem: 9657 +[14:54:56.698080] val: [1/2] eta: 0:00:01 loss: 0.1485 (0.9426) time: 1.3966 data: 1.3805 max mem: 9657 +[14:54:56.772239] val: Total time: 0:00:02 (1.4344 s / it) +[14:54:56.783660] val loss: 0.9426374435424805 +[14:54:56.783874] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7977, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7514, Kappa: 0.0000, Score: 0.4100 +[14:54:56.827081] Best epoch = 3, Best score = 0.4212 +[14:54:57.134660] log_dir: ./output_logs/retfound +[14:54:59.385562] Epoch: [11] [0/4] eta: 0:00:09 lr: 0.000624 loss: 0.5614 (0.5614) time: 2.2501 data: 2.1808 max mem: 9657 +[14:54:59.582779] Epoch: [11] [3/4] eta: 0:00:00 lr: 0.000622 loss: 0.4553 (0.4918) time: 0.6116 data: 0.5453 max mem: 9657 +[14:54:59.654997] Epoch: [11] Total time: 0:00:02 (0.6300 s / it) +[14:54:59.655751] Averaged stats: lr: 0.000622 loss: 0.4553 (0.4918) +[14:55:02.399894] val: [0/2] eta: 0:00:05 loss: 0.1672 (0.1672) time: 2.7215 data: 2.7044 max mem: 9657 +[14:55:02.411491] val: [1/2] eta: 0:00:01 loss: 0.1672 (0.8659) time: 1.3662 data: 1.3523 max mem: 9657 +[14:55:02.480809] val: Total time: 0:00:02 (1.4015 s / it) +[14:55:02.489723] val loss: 0.8659271001815796 +[14:55:02.489914] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7937, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7586, Kappa: 0.0000, Score: 0.4087 +[14:55:02.533256] Best epoch = 3, Best score = 0.4212 +[14:55:02.812732] log_dir: ./output_logs/retfound +[14:55:04.923832] Epoch: [12] [0/4] eta: 0:00:08 lr: 0.000621 loss: 0.8006 (0.8006) time: 2.1102 data: 2.0390 max mem: 9657 +[14:55:05.143313] Epoch: [12] [3/4] eta: 0:00:00 lr: 0.000618 loss: 0.4040 (0.5034) time: 0.5823 data: 0.5157 max mem: 9657 +[14:55:05.209985] Epoch: [12] Total time: 0:00:02 (0.5993 s / it) +[14:55:05.210750] Averaged stats: lr: 0.000618 loss: 0.4040 (0.5034) +[14:55:08.096107] val: [0/2] eta: 0:00:05 loss: 0.2665 (0.2665) time: 2.8612 data: 2.8441 max mem: 9657 +[14:55:08.105488] val: [1/2] eta: 0:00:01 loss: 0.2665 (0.7070) time: 1.4350 data: 1.4221 max mem: 9657 +[14:55:08.172545] val: Total time: 0:00:02 (1.4692 s / it) +[14:55:08.181306] val loss: 0.7070209383964539 +[14:55:08.181477] Accuracy: 0.7857, F1 Score: 0.5293, ROC AUC: 0.8250, Hamming Loss: 0.2143, + Jaccard Score: 0.4402, Precision: 0.8902, Recall: 0.5500, + Average Precision: 0.7740, Kappa: 0.1448, Score: 0.4997 +[14:55:09.895040] Best epoch = 12, Best score = 0.4997 +[14:55:09.958337] log_dir: ./output_logs/retfound +[14:55:12.361668] Epoch: [13] [0/4] eta: 0:00:09 lr: 0.000616 loss: 0.4536 (0.4536) time: 2.4024 data: 2.3326 max mem: 9657 +[14:55:12.559396] Epoch: [13] [3/4] eta: 0:00:00 lr: 0.000612 loss: 0.4536 (0.4772) time: 0.6499 data: 0.5833 max mem: 9657 +[14:55:12.632764] Epoch: [13] Total time: 0:00:02 (0.6686 s / it) +[14:55:12.633604] Averaged stats: lr: 0.000612 loss: 0.4536 (0.4772) +[14:55:15.493266] val: [0/2] eta: 0:00:05 loss: 0.1724 (0.1724) time: 2.8352 data: 2.8182 max mem: 9657 +[14:55:15.502567] val: [1/2] eta: 0:00:01 loss: 0.1724 (0.7855) time: 1.4220 data: 1.4092 max mem: 9657 +[14:55:15.567831] val: Total time: 0:00:02 (1.4552 s / it) +[14:55:15.578175] val loss: 0.7854663729667664 +[14:55:15.578446] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8289, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7719, Kappa: 0.0000, Score: 0.4204 +[14:55:15.625049] Best epoch = 12, Best score = 0.4997 +[14:55:15.864920] log_dir: ./output_logs/retfound +[14:55:17.914384] Epoch: [14] [0/4] eta: 0:00:08 lr: 0.000610 loss: 0.5041 (0.5041) time: 2.0485 data: 1.9779 max mem: 9657 +[14:55:18.130628] Epoch: [14] [3/4] eta: 0:00:00 lr: 0.000604 loss: 0.4286 (0.4724) time: 0.5659 data: 0.4996 max mem: 9657 +[14:55:18.200610] Epoch: [14] Total time: 0:00:02 (0.5839 s / it) +[14:55:18.201381] Averaged stats: lr: 0.000604 loss: 0.4286 (0.4724) +[14:55:21.034495] val: [0/2] eta: 0:00:05 loss: 0.0872 (0.0872) time: 2.8214 data: 2.8045 max mem: 9657 +[14:55:21.044101] val: [1/2] eta: 0:00:01 loss: 0.0872 (0.9815) time: 1.4152 data: 1.4023 max mem: 9657 +[14:55:21.111985] val: Total time: 0:00:02 (1.4498 s / it) +[14:55:21.120910] val loss: 0.9815189242362976 +[14:55:21.121113] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8352, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7996, Kappa: 0.0000, Score: 0.4225 +[14:55:21.166056] Best epoch = 12, Best score = 0.4997 +[14:55:21.419884] log_dir: ./output_logs/retfound +[14:55:23.672221] Epoch: [15] [0/4] eta: 0:00:09 lr: 0.000601 loss: 0.4289 (0.4289) time: 2.2514 data: 2.1823 max mem: 9657 +[14:55:23.871599] Epoch: [15] [3/4] eta: 0:00:00 lr: 0.000594 loss: 0.4158 (0.4058) time: 0.6125 data: 0.5463 max mem: 9657 +[14:55:23.940223] Epoch: [15] Total time: 0:00:02 (0.6300 s / it) +[14:55:23.941021] Averaged stats: lr: 0.000594 loss: 0.4158 (0.4058) +[14:55:26.688501] val: [0/2] eta: 0:00:05 loss: 0.0577 (0.0577) time: 2.7245 data: 2.7069 max mem: 9657 +[14:55:26.698138] val: [1/2] eta: 0:00:01 loss: 0.0577 (1.1474) time: 1.3668 data: 1.3536 max mem: 9657 +[14:55:26.767751] val: Total time: 0:00:02 (1.4023 s / it) +[14:55:26.776995] val loss: 1.1473761796951294 +[14:55:26.777284] Accuracy: 0.7857, F1 Score: 0.5293, ROC AUC: 0.8320, Hamming Loss: 0.2143, + Jaccard Score: 0.4402, Precision: 0.8902, Recall: 0.5500, + Average Precision: 0.7982, Kappa: 0.1448, Score: 0.5020 +[14:55:28.507131] Best epoch = 15, Best score = 0.5020 +[14:55:28.569102] log_dir: ./output_logs/retfound +[14:55:30.821569] Epoch: [16] [0/4] eta: 0:00:09 lr: 0.000591 loss: 0.5790 (0.5790) time: 2.2516 data: 2.1823 max mem: 9657 +[14:55:31.017579] Epoch: [16] [3/4] eta: 0:00:00 lr: 0.000582 loss: 0.4673 (0.4771) time: 0.6117 data: 0.5456 max mem: 9657 +[14:55:31.085493] Epoch: [16] Total time: 0:00:02 (0.6291 s / it) +[14:55:31.086149] Averaged stats: lr: 0.000582 loss: 0.4673 (0.4771) +[14:55:33.660725] val: [0/2] eta: 0:00:05 loss: 0.1054 (0.1054) time: 2.5647 data: 2.5477 max mem: 9657 +[14:55:33.670439] val: [1/2] eta: 0:00:01 loss: 0.1054 (0.9500) time: 1.2869 data: 1.2739 max mem: 9657 +[14:55:33.738360] val: Total time: 0:00:02 (1.3215 s / it) +[14:55:33.747271] val loss: 0.9500231146812439 +[14:55:33.747461] Accuracy: 0.7381, F1 Score: 0.4995, ROC AUC: 0.8313, Hamming Loss: 0.2619, + Jaccard Score: 0.4075, Precision: 0.5513, Recall: 0.5188, + Average Precision: 0.7721, Kappa: 0.0494, Score: 0.4600 +[14:55:33.793047] Best epoch = 15, Best score = 0.5020 +[14:55:34.065481] log_dir: ./output_logs/retfound +[14:55:36.337283] Epoch: [17] [0/4] eta: 0:00:09 lr: 0.000579 loss: 0.5105 (0.5105) time: 2.2709 data: 2.2018 max mem: 9657 +[14:55:36.533172] Epoch: [17] [3/4] eta: 0:00:00 lr: 0.000569 loss: 0.4574 (0.4849) time: 0.6165 data: 0.5505 max mem: 9657 +[14:55:36.602585] Epoch: [17] Total time: 0:00:02 (0.6342 s / it) +[14:55:36.603361] Averaged stats: lr: 0.000569 loss: 0.4574 (0.4849) +[14:55:39.421942] val: [0/2] eta: 0:00:05 loss: 0.1489 (0.1489) time: 2.7967 data: 2.7797 max mem: 9657 +[14:55:39.431238] val: [1/2] eta: 0:00:01 loss: 0.1489 (0.7822) time: 1.4027 data: 1.3899 max mem: 9657 +[14:55:39.498199] val: Total time: 0:00:02 (1.4368 s / it) +[14:55:39.507192] val loss: 0.7822439074516296 +[14:55:39.507519] Accuracy: 0.7619, F1 Score: 0.5714, ROC AUC: 0.8250, Hamming Loss: 0.2381, + Jaccard Score: 0.4583, Precision: 0.6447, Recall: 0.5687, + Average Precision: 0.7678, Kappa: 0.1732, Score: 0.5232 +[14:55:41.301070] Best epoch = 17, Best score = 0.5232 +[14:55:41.372707] log_dir: ./output_logs/retfound +[14:55:43.600384] Epoch: [18] [0/4] eta: 0:00:08 lr: 0.000565 loss: 0.3333 (0.3333) time: 2.2268 data: 2.1566 max mem: 9657 +[14:55:43.842832] Epoch: [18] [3/4] eta: 0:00:00 lr: 0.000554 loss: 0.3696 (0.4794) time: 0.6171 data: 0.5508 max mem: 9657 +[14:55:43.914953] Epoch: [18] Total time: 0:00:02 (0.6355 s / it) +[14:55:43.915723] Averaged stats: lr: 0.000554 loss: 0.3696 (0.4794) +[14:55:46.605774] val: [0/2] eta: 0:00:05 loss: 0.2231 (0.2231) time: 2.6789 data: 2.6623 max mem: 9657 +[14:55:46.614792] val: [1/2] eta: 0:00:01 loss: 0.2231 (0.6438) time: 1.3437 data: 1.3312 max mem: 9657 +[14:55:46.681003] val: Total time: 0:00:02 (1.3774 s / it) +[14:55:46.691162] val loss: 0.6437538862228394 +[14:55:46.691519] Accuracy: 0.7857, F1 Score: 0.6681, ROC AUC: 0.8188, Hamming Loss: 0.2143, + Jaccard Score: 0.5354, Precision: 0.7000, Recall: 0.6531, + Average Precision: 0.7609, Kappa: 0.3415, Score: 0.6094 +[14:55:48.376651] Best epoch = 18, Best score = 0.6094 +[14:55:48.442328] log_dir: ./output_logs/retfound +[14:55:50.840578] Epoch: [19] [0/4] eta: 0:00:09 lr: 0.000550 loss: 0.4532 (0.4532) time: 2.3974 data: 2.3273 max mem: 9657 +[14:55:51.036511] Epoch: [19] [3/4] eta: 0:00:00 lr: 0.000538 loss: 0.4371 (0.4379) time: 0.6482 data: 0.5819 max mem: 9657 +[14:55:51.107298] Epoch: [19] Total time: 0:00:02 (0.6662 s / it) +[14:55:51.108311] Averaged stats: lr: 0.000538 loss: 0.4371 (0.4379) +[14:55:53.739095] val: [0/2] eta: 0:00:05 loss: 0.1399 (0.1399) time: 2.6075 data: 2.5896 max mem: 9657 +[14:55:53.748355] val: [1/2] eta: 0:00:01 loss: 0.1399 (0.7902) time: 1.3081 data: 1.2948 max mem: 9657 +[14:55:53.816797] val: Total time: 0:00:02 (1.3430 s / it) +[14:55:53.826291] val loss: 0.7901833653450012 +[14:55:53.826506] Accuracy: 0.7619, F1 Score: 0.5714, ROC AUC: 0.8219, Hamming Loss: 0.2381, + Jaccard Score: 0.4583, Precision: 0.6447, Recall: 0.5687, + Average Precision: 0.7656, Kappa: 0.1732, Score: 0.5222 +[14:55:53.860507] Best epoch = 18, Best score = 0.6094 +[14:55:54.152136] log_dir: ./output_logs/retfound +[14:55:56.525279] Epoch: [20] [0/4] eta: 0:00:09 lr: 0.000534 loss: 0.4661 (0.4661) time: 2.3722 data: 2.3026 max mem: 9657 +[14:55:56.721057] Epoch: [20] [3/4] eta: 0:00:00 lr: 0.000520 loss: 0.3735 (0.4305) time: 0.6418 data: 0.5757 max mem: 9657 +[14:55:56.785767] Epoch: [20] Total time: 0:00:02 (0.6584 s / it) +[14:55:56.786527] Averaged stats: lr: 0.000520 loss: 0.3735 (0.4305) +[14:55:59.535782] val: [0/2] eta: 0:00:05 loss: 0.1011 (0.1011) time: 2.7265 data: 2.7095 max mem: 9657 +[14:55:59.545128] val: [1/2] eta: 0:00:01 loss: 0.1011 (0.9416) time: 1.3676 data: 1.3548 max mem: 9657 +[14:55:59.617357] val: Total time: 0:00:02 (1.4044 s / it) +[14:55:59.626314] val loss: 0.9415563344955444 +[14:55:59.626645] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8227, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.7639, Kappa: 0.2868, Score: 0.5814 +[14:55:59.671497] Best epoch = 18, Best score = 0.6094 +[14:55:59.940017] log_dir: ./output_logs/retfound +[14:56:02.128102] Epoch: [21] [0/4] eta: 0:00:08 lr: 0.000516 loss: 0.5421 (0.5421) time: 2.1872 data: 2.1180 max mem: 9657 +[14:56:02.347038] Epoch: [21] [3/4] eta: 0:00:00 lr: 0.000501 loss: 0.4351 (0.4986) time: 0.6013 data: 0.5351 max mem: 9657 +[14:56:02.418201] Epoch: [21] Total time: 0:00:02 (0.6195 s / it) +[14:56:02.419011] Averaged stats: lr: 0.000501 loss: 0.4351 (0.4986) +[14:56:05.045996] val: [0/2] eta: 0:00:05 loss: 0.0794 (0.0794) time: 2.6136 data: 2.5966 max mem: 9657 +[14:56:05.055165] val: [1/2] eta: 0:00:01 loss: 0.0794 (1.0978) time: 1.3111 data: 1.2984 max mem: 9657 +[14:56:05.123954] val: Total time: 0:00:02 (1.3461 s / it) +[14:56:05.132779] val loss: 1.097848653793335 +[14:56:05.133028] Accuracy: 0.8095, F1 Score: 0.6571, ROC AUC: 0.8297, Hamming Loss: 0.1905, + Jaccard Score: 0.5338, Precision: 0.7829, Recall: 0.6344, + Average Precision: 0.7957, Kappa: 0.3386, Score: 0.6085 +[14:56:05.165819] Best epoch = 18, Best score = 0.6094 +[14:56:05.460541] log_dir: ./output_logs/retfound +[14:56:07.698162] Epoch: [22] [0/4] eta: 0:00:08 lr: 0.000496 loss: 0.3067 (0.3067) time: 2.2366 data: 2.1604 max mem: 9657 +[14:56:07.894242] Epoch: [22] [3/4] eta: 0:00:00 lr: 0.000481 loss: 0.3426 (0.4560) time: 0.6080 data: 0.5401 max mem: 9657 +[14:56:07.963237] Epoch: [22] Total time: 0:00:02 (0.6256 s / it) +[14:56:07.964015] Averaged stats: lr: 0.000481 loss: 0.3426 (0.4560) +[14:56:10.916348] val: [0/2] eta: 0:00:05 loss: 0.2472 (0.2472) time: 2.9447 data: 2.9280 max mem: 9657 +[14:56:10.925663] val: [1/2] eta: 0:00:01 loss: 0.2472 (0.6658) time: 1.4767 data: 1.4641 max mem: 9657 +[14:56:10.994666] val: Total time: 0:00:03 (1.5119 s / it) +[14:56:11.003500] val loss: 0.665777862071991 +[14:56:11.003764] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8320, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.7946, Kappa: 0.5243, Score: 0.7061 +[14:56:12.709688] Best epoch = 22, Best score = 0.7061 +[14:56:12.772055] log_dir: ./output_logs/retfound +[14:56:15.000495] Epoch: [23] [0/4] eta: 0:00:08 lr: 0.000476 loss: 0.4409 (0.4409) time: 2.2274 data: 2.1576 max mem: 9657 +[14:56:15.195970] Epoch: [23] [3/4] eta: 0:00:00 lr: 0.000460 loss: 0.4032 (0.4267) time: 0.6055 data: 0.5395 max mem: 9657 +[14:56:15.262669] Epoch: [23] Total time: 0:00:02 (0.6226 s / it) +[14:56:15.263447] Averaged stats: lr: 0.000460 loss: 0.4032 (0.4267) +[14:56:17.930664] val: [0/2] eta: 0:00:05 loss: 0.1838 (0.1838) time: 2.6554 data: 2.6383 max mem: 9657 +[14:56:17.939963] val: [1/2] eta: 0:00:01 loss: 0.1838 (0.7157) time: 1.3320 data: 1.3192 max mem: 9657 +[14:56:18.009351] val: Total time: 0:00:02 (1.3674 s / it) +[14:56:18.018451] val loss: 0.7156535387039185 +[14:56:18.018626] Accuracy: 0.7857, F1 Score: 0.6681, ROC AUC: 0.8398, Hamming Loss: 0.2143, + Jaccard Score: 0.5354, Precision: 0.7000, Recall: 0.6531, + Average Precision: 0.8124, Kappa: 0.3415, Score: 0.6165 +[14:56:18.052150] Best epoch = 22, Best score = 0.7061 +[14:56:18.346484] log_dir: ./output_logs/retfound +[14:56:20.667437] Epoch: [24] [0/4] eta: 0:00:09 lr: 0.000455 loss: 0.4179 (0.4179) time: 2.3199 data: 2.2496 max mem: 9657 +[14:56:20.864981] Epoch: [24] [3/4] eta: 0:00:00 lr: 0.000438 loss: 0.4179 (0.4275) time: 0.6292 data: 0.5624 max mem: 9657 +[14:56:20.935644] Epoch: [24] Total time: 0:00:02 (0.6472 s / it) +[14:56:20.936412] Averaged stats: lr: 0.000438 loss: 0.4179 (0.4275) +[14:56:23.598185] val: [0/2] eta: 0:00:05 loss: 0.0337 (0.0337) time: 2.6396 data: 2.6226 max mem: 9657 +[14:56:23.607352] val: [1/2] eta: 0:00:01 loss: 0.0337 (1.3100) time: 1.3241 data: 1.3113 max mem: 9657 +[14:56:23.678367] val: Total time: 0:00:02 (1.3602 s / it) +[14:56:23.687201] val loss: 1.3100290298461914 +[14:56:23.687373] Accuracy: 0.8095, F1 Score: 0.6111, ROC AUC: 0.8445, Hamming Loss: 0.1905, + Jaccard Score: 0.5000, Precision: 0.9000, Recall: 0.6000, + Average Precision: 0.8156, Kappa: 0.2759, Score: 0.5772 +[14:56:23.733316] Best epoch = 22, Best score = 0.7061 +[14:56:24.010003] log_dir: ./output_logs/retfound +[14:56:26.289987] Epoch: [25] [0/4] eta: 0:00:09 lr: 0.000432 loss: 0.2786 (0.2786) time: 2.2790 data: 2.2100 max mem: 9657 +[14:56:26.486932] Epoch: [25] [3/4] eta: 0:00:00 lr: 0.000415 loss: 0.3872 (0.4757) time: 0.6188 data: 0.5526 max mem: 9657 +[14:56:26.555814] Epoch: [25] Total time: 0:00:02 (0.6364 s / it) +[14:56:26.556577] Averaged stats: lr: 0.000415 loss: 0.3872 (0.4757) +[14:56:29.099121] val: [0/2] eta: 0:00:05 loss: 0.0575 (0.0575) time: 2.5313 data: 2.5140 max mem: 9657 +[14:56:29.108318] val: [1/2] eta: 0:00:01 loss: 0.0575 (1.1822) time: 1.2700 data: 1.2571 max mem: 9657 +[14:56:29.175328] val: Total time: 0:00:02 (1.3041 s / it) +[14:56:29.184246] val loss: 1.18218994140625 +[14:56:29.184441] Accuracy: 0.8095, F1 Score: 0.6111, ROC AUC: 0.8453, Hamming Loss: 0.1905, + Jaccard Score: 0.5000, Precision: 0.9000, Recall: 0.6000, + Average Precision: 0.8154, Kappa: 0.2759, Score: 0.5774 +[14:56:29.217998] Best epoch = 22, Best score = 0.7061 +[14:56:29.505576] log_dir: ./output_logs/retfound +[14:56:31.493913] Epoch: [26] [0/4] eta: 0:00:07 lr: 0.000409 loss: 0.5900 (0.5900) time: 1.9874 data: 1.9184 max mem: 9657 +[14:56:31.696703] Epoch: [26] [3/4] eta: 0:00:00 lr: 0.000392 loss: 0.4443 (0.4869) time: 0.5474 data: 0.4816 max mem: 9657 +[14:56:31.778641] Epoch: [26] Total time: 0:00:02 (0.5682 s / it) +[14:56:31.779458] Averaged stats: lr: 0.000392 loss: 0.4443 (0.4869) +[14:56:34.393807] val: [0/2] eta: 0:00:05 loss: 0.1692 (0.1692) time: 2.5987 data: 2.5819 max mem: 9657 +[14:56:34.402836] val: [1/2] eta: 0:00:01 loss: 0.1692 (0.7669) time: 1.3036 data: 1.2910 max mem: 9657 +[14:56:34.471955] val: Total time: 0:00:02 (1.3388 s / it) +[14:56:34.480831] val loss: 0.7669225931167603 +[14:56:34.481077] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8391, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.7946, Kappa: 0.2868, Score: 0.5869 +[14:56:34.531766] Best epoch = 22, Best score = 0.7061 +[14:56:34.791463] log_dir: ./output_logs/retfound +[14:56:37.102120] Epoch: [27] [0/4] eta: 0:00:09 lr: 0.000386 loss: 0.3972 (0.3972) time: 2.3097 data: 2.2393 max mem: 9657 +[14:56:37.297163] Epoch: [27] [3/4] eta: 0:00:00 lr: 0.000368 loss: 0.3972 (0.4020) time: 0.6260 data: 0.5599 max mem: 9657 +[14:56:37.368697] Epoch: [27] Total time: 0:00:02 (0.6443 s / it) +[14:56:37.369442] Averaged stats: lr: 0.000368 loss: 0.3972 (0.4020) +[14:56:39.984626] val: [0/2] eta: 0:00:05 loss: 0.1408 (0.1408) time: 2.5920 data: 2.5750 max mem: 9657 +[14:56:39.994282] val: [1/2] eta: 0:00:01 loss: 0.1408 (0.8135) time: 1.3005 data: 1.2876 max mem: 9657 +[14:56:40.062051] val: Total time: 0:00:02 (1.3350 s / it) +[14:56:40.070825] val loss: 0.8134706616401672 +[14:56:40.071071] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8438, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.7989, Kappa: 0.2868, Score: 0.5884 +[14:56:40.119006] Best epoch = 22, Best score = 0.7061 +[14:56:40.370017] log_dir: ./output_logs/retfound +[14:56:42.575641] Epoch: [28] [0/4] eta: 0:00:08 lr: 0.000362 loss: 0.4524 (0.4524) time: 2.2047 data: 2.1356 max mem: 9657 +[14:56:42.770593] Epoch: [28] [3/4] eta: 0:00:00 lr: 0.000344 loss: 0.4424 (0.4450) time: 0.5997 data: 0.5340 max mem: 9657 +[14:56:42.839581] Epoch: [28] Total time: 0:00:02 (0.6173 s / it) +[14:56:42.840363] Averaged stats: lr: 0.000344 loss: 0.4424 (0.4450) +[14:56:45.464365] val: [0/2] eta: 0:00:05 loss: 0.1032 (0.1032) time: 2.6023 data: 2.5848 max mem: 9657 +[14:56:45.473714] val: [1/2] eta: 0:00:01 loss: 0.1032 (0.9360) time: 1.3056 data: 1.2925 max mem: 9657 +[14:56:45.541980] val: Total time: 0:00:02 (1.3404 s / it) +[14:56:45.550879] val loss: 0.9359757900238037 +[14:56:45.551119] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8492, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8094, Kappa: 0.2868, Score: 0.5903 +[14:56:45.585308] Best epoch = 22, Best score = 0.7061 +[14:56:45.848981] log_dir: ./output_logs/retfound +[14:56:48.063266] Epoch: [29] [0/4] eta: 0:00:08 lr: 0.000337 loss: 0.3955 (0.3955) time: 2.2132 data: 2.1419 max mem: 9657 +[14:56:48.280124] Epoch: [29] [3/4] eta: 0:00:00 lr: 0.000319 loss: 0.3759 (0.4278) time: 0.6073 data: 0.5407 max mem: 9657 +[14:56:48.358784] Epoch: [29] Total time: 0:00:02 (0.6274 s / it) +[14:56:48.359557] Averaged stats: lr: 0.000319 loss: 0.3759 (0.4278) +[14:56:51.011422] val: [0/2] eta: 0:00:05 loss: 0.0963 (0.0963) time: 2.6405 data: 2.6231 max mem: 9657 +[14:56:51.020890] val: [1/2] eta: 0:00:01 loss: 0.0963 (0.9874) time: 1.3247 data: 1.3116 max mem: 9657 +[14:56:51.098178] val: Total time: 0:00:02 (1.3640 s / it) +[14:56:51.106945] val loss: 0.987418532371521 +[14:56:51.107195] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8469, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8068, Kappa: 0.2868, Score: 0.5895 +[14:56:51.139450] Best epoch = 22, Best score = 0.7061 +[14:56:51.402975] log_dir: ./output_logs/retfound +[14:56:53.704155] Epoch: [30] [0/4] eta: 0:00:09 lr: 0.000313 loss: 0.3595 (0.3595) time: 2.3001 data: 2.2313 max mem: 9657 +[14:56:53.899735] Epoch: [30] [3/4] eta: 0:00:00 lr: 0.000295 loss: 0.3994 (0.4406) time: 0.6238 data: 0.5579 max mem: 9657 +[14:56:53.968734] Epoch: [30] Total time: 0:00:02 (0.6414 s / it) +[14:56:53.969533] Averaged stats: lr: 0.000295 loss: 0.3994 (0.4406) +[14:56:56.504984] val: [0/2] eta: 0:00:05 loss: 0.1127 (0.1127) time: 2.5240 data: 2.5071 max mem: 9657 +[14:56:56.514492] val: [1/2] eta: 0:00:01 loss: 0.1127 (0.9266) time: 1.2665 data: 1.2536 max mem: 9657 +[14:56:56.581536] val: Total time: 0:00:02 (1.3007 s / it) +[14:56:56.590359] val loss: 0.9265840649604797 +[14:56:56.590555] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8523, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8217, Kappa: 0.2868, Score: 0.5913 +[14:56:56.625453] Best epoch = 22, Best score = 0.7061 +[14:56:56.889688] log_dir: ./output_logs/retfound +[14:56:59.266157] Epoch: [31] [0/4] eta: 0:00:09 lr: 0.000289 loss: 0.3638 (0.3638) time: 2.3756 data: 2.3065 max mem: 9657 +[14:56:59.461540] Epoch: [31] [3/4] eta: 0:00:00 lr: 0.000270 loss: 0.3606 (0.3813) time: 0.6426 data: 0.5767 max mem: 9657 +[14:56:59.531869] Epoch: [31] Total time: 0:00:02 (0.6605 s / it) +[14:56:59.532644] Averaged stats: lr: 0.000270 loss: 0.3606 (0.3813) +[14:57:02.408019] val: [0/2] eta: 0:00:05 loss: 0.1679 (0.1679) time: 2.8518 data: 2.8346 max mem: 9657 +[14:57:02.418866] val: [1/2] eta: 0:00:01 loss: 0.1679 (0.8142) time: 1.4310 data: 1.4174 max mem: 9657 +[14:57:02.485916] val: Total time: 0:00:02 (1.4652 s / it) +[14:57:02.497769] val loss: 0.8141542077064514 +[14:57:02.497976] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8492, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8125, Kappa: 0.2868, Score: 0.5903 +[14:57:02.550681] Best epoch = 22, Best score = 0.7061 +[14:57:02.802620] log_dir: ./output_logs/retfound +[14:57:05.055194] Epoch: [32] [0/4] eta: 0:00:09 lr: 0.000264 loss: 0.3088 (0.3088) time: 2.2516 data: 2.1822 max mem: 9657 +[14:57:05.249683] Epoch: [32] [3/4] eta: 0:00:00 lr: 0.000246 loss: 0.3903 (0.4189) time: 0.6114 data: 0.5456 max mem: 9657 +[14:57:05.318159] Epoch: [32] Total time: 0:00:02 (0.6288 s / it) +[14:57:05.318909] Averaged stats: lr: 0.000246 loss: 0.3903 (0.4189) +[14:57:07.882696] val: [0/2] eta: 0:00:05 loss: 0.2242 (0.2242) time: 2.5526 data: 2.5300 max mem: 9657 +[14:57:07.895432] val: [1/2] eta: 0:00:01 loss: 0.2242 (0.7479) time: 1.2824 data: 1.2651 max mem: 9657 +[14:57:07.968528] val: Total time: 0:00:02 (1.3197 s / it) +[14:57:07.982156] val loss: 0.7478845715522766 +[14:57:07.982391] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8461, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.8106, Kappa: 0.4362, Score: 0.6665 +[14:57:08.029811] Best epoch = 22, Best score = 0.7061 +[14:57:08.273911] log_dir: ./output_logs/retfound +[14:57:10.485227] Epoch: [33] [0/4] eta: 0:00:08 lr: 0.000240 loss: 0.5746 (0.5746) time: 2.2103 data: 2.1390 max mem: 9657 +[14:57:10.792987] Epoch: [33] [3/4] eta: 0:00:00 lr: 0.000222 loss: 0.4433 (0.4635) time: 0.6294 data: 0.5629 max mem: 9657 +[14:57:10.861689] Epoch: [33] Total time: 0:00:02 (0.6469 s / it) +[14:57:10.862441] Averaged stats: lr: 0.000222 loss: 0.4433 (0.4635) +[14:57:13.580636] val: [0/2] eta: 0:00:05 loss: 0.2122 (0.2122) time: 2.7075 data: 2.6903 max mem: 9657 +[14:57:13.590014] val: [1/2] eta: 0:00:01 loss: 0.2122 (0.7602) time: 1.3581 data: 1.3452 max mem: 9657 +[14:57:13.655305] val: Total time: 0:00:02 (1.3914 s / it) +[14:57:13.664083] val loss: 0.7601932287216187 +[14:57:13.664316] Accuracy: 0.7857, F1 Score: 0.6681, ROC AUC: 0.8438, Hamming Loss: 0.2143, + Jaccard Score: 0.5354, Precision: 0.7000, Recall: 0.6531, + Average Precision: 0.8189, Kappa: 0.3415, Score: 0.6178 +[14:57:13.696222] Best epoch = 22, Best score = 0.7061 +[14:57:13.983857] log_dir: ./output_logs/retfound +[14:57:16.287480] Epoch: [34] [0/4] eta: 0:00:09 lr: 0.000217 loss: 0.4239 (0.4239) time: 2.3027 data: 2.2333 max mem: 9657 +[14:57:16.483202] Epoch: [34] [3/4] eta: 0:00:00 lr: 0.000199 loss: 0.3663 (0.3808) time: 0.6245 data: 0.5584 max mem: 9657 +[14:57:16.558292] Epoch: [34] Total time: 0:00:02 (0.6436 s / it) +[14:57:16.559055] Averaged stats: lr: 0.000199 loss: 0.3663 (0.3808) +[14:57:19.174011] val: [0/2] eta: 0:00:05 loss: 0.1693 (0.1693) time: 2.6046 data: 2.5877 max mem: 9657 +[14:57:19.182931] val: [1/2] eta: 0:00:01 loss: 0.1693 (0.8489) time: 1.3065 data: 1.2939 max mem: 9657 +[14:57:19.254744] val: Total time: 0:00:02 (1.3430 s / it) +[14:57:19.263912] val loss: 0.8489300012588501 +[14:57:19.264091] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8445, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8198, Kappa: 0.2868, Score: 0.5887 +[14:57:19.296309] Best epoch = 22, Best score = 0.7061 +[14:57:19.561356] log_dir: ./output_logs/retfound +[14:57:21.764348] Epoch: [35] [0/4] eta: 0:00:08 lr: 0.000194 loss: 0.6013 (0.6013) time: 2.2019 data: 2.1207 max mem: 9657 +[14:57:22.021686] Epoch: [35] [3/4] eta: 0:00:00 lr: 0.000177 loss: 0.3931 (0.4801) time: 0.6146 data: 0.5438 max mem: 9657 +[14:57:22.100236] Epoch: [35] Total time: 0:00:02 (0.6347 s / it) +[14:57:22.101021] Averaged stats: lr: 0.000177 loss: 0.3931 (0.4801) +[14:57:24.632091] val: [0/2] eta: 0:00:05 loss: 0.1303 (0.1303) time: 2.5191 data: 2.5020 max mem: 9657 +[14:57:24.641258] val: [1/2] eta: 0:00:01 loss: 0.1303 (0.9358) time: 1.2639 data: 1.2511 max mem: 9657 +[14:57:24.713465] val: Total time: 0:00:02 (1.3006 s / it) +[14:57:24.722163] val loss: 0.9358091950416565 +[14:57:24.722361] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8391, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8108, Kappa: 0.2868, Score: 0.5869 +[14:57:24.754739] Best epoch = 22, Best score = 0.7061 +[14:57:25.029923] log_dir: ./output_logs/retfound +[14:57:27.312536] Epoch: [36] [0/4] eta: 0:00:09 lr: 0.000171 loss: 0.3651 (0.3651) time: 2.2818 data: 2.2126 max mem: 9657 +[14:57:27.506637] Epoch: [36] [3/4] eta: 0:00:00 lr: 0.000155 loss: 0.3651 (0.4589) time: 0.6188 data: 0.5532 max mem: 9657 +[14:57:27.575438] Epoch: [36] Total time: 0:00:02 (0.6363 s / it) +[14:57:27.576134] Averaged stats: lr: 0.000155 loss: 0.3651 (0.4589) +[14:57:30.112264] val: [0/2] eta: 0:00:05 loss: 0.1335 (0.1335) time: 2.5257 data: 2.5087 max mem: 9657 +[14:57:30.121423] val: [1/2] eta: 0:00:01 loss: 0.1335 (0.8780) time: 1.2671 data: 1.2544 max mem: 9657 +[14:57:30.203374] val: Total time: 0:00:02 (1.3087 s / it) +[14:57:30.212154] val loss: 0.8780083656311035 +[14:57:30.212327] Accuracy: 0.7857, F1 Score: 0.6348, ROC AUC: 0.8406, Hamming Loss: 0.2143, + Jaccard Score: 0.5096, Precision: 0.7054, Recall: 0.6188, + Average Precision: 0.8153, Kappa: 0.2868, Score: 0.5874 +[14:57:30.260711] Best epoch = 22, Best score = 0.7061 +[14:57:30.528074] log_dir: ./output_logs/retfound +[14:57:32.883654] Epoch: [37] [0/4] eta: 0:00:09 lr: 0.000150 loss: 0.2904 (0.2904) time: 2.3545 data: 2.2856 max mem: 9657 +[14:57:33.078665] Epoch: [37] [3/4] eta: 0:00:00 lr: 0.000135 loss: 0.2904 (0.4061) time: 0.6372 data: 0.5714 max mem: 9657 +[14:57:33.147783] Epoch: [37] Total time: 0:00:02 (0.6549 s / it) +[14:57:33.148513] Averaged stats: lr: 0.000135 loss: 0.2904 (0.4061) +[14:57:35.716668] val: [0/2] eta: 0:00:05 loss: 0.1639 (0.1639) time: 2.5455 data: 2.5285 max mem: 9657 +[14:57:35.725714] val: [1/2] eta: 0:00:01 loss: 0.1639 (0.7652) time: 1.2770 data: 1.2643 max mem: 9657 +[14:57:35.793095] val: Total time: 0:00:02 (1.3113 s / it) +[14:57:35.802226] val loss: 0.765245795249939 +[14:57:35.802458] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8398, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8164, Kappa: 0.3913, Score: 0.6408 +[14:57:35.839542] Best epoch = 22, Best score = 0.7061 +[14:57:36.107407] log_dir: ./output_logs/retfound +[14:57:38.218251] Epoch: [38] [0/4] eta: 0:00:08 lr: 0.000130 loss: 0.3784 (0.3784) time: 2.1099 data: 2.0399 max mem: 9657 +[14:57:38.502115] Epoch: [38] [3/4] eta: 0:00:00 lr: 0.000115 loss: 0.3784 (0.4464) time: 0.5983 data: 0.5320 max mem: 9657 +[14:57:38.576059] Epoch: [38] Total time: 0:00:02 (0.6171 s / it) +[14:57:38.576845] Averaged stats: lr: 0.000115 loss: 0.3784 (0.4464) +[14:57:41.326644] val: [0/2] eta: 0:00:05 loss: 0.2152 (0.2152) time: 2.7377 data: 2.7209 max mem: 9657 +[14:57:41.335964] val: [1/2] eta: 0:00:01 loss: 0.2152 (0.6599) time: 1.3733 data: 1.3605 max mem: 9657 +[14:57:41.406409] val: Total time: 0:00:02 (1.4091 s / it) +[14:57:41.415178] val loss: 0.6599004864692688 +[14:57:41.415408] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8336, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.8120, Kappa: 0.4362, Score: 0.6623 +[14:57:41.460255] Best epoch = 22, Best score = 0.7061 +[14:57:41.727057] log_dir: ./output_logs/retfound +[14:57:44.098582] Epoch: [39] [0/4] eta: 0:00:09 lr: 0.000110 loss: 0.3822 (0.3822) time: 2.3706 data: 2.3014 max mem: 9657 +[14:57:44.293884] Epoch: [39] [3/4] eta: 0:00:00 lr: 0.000097 loss: 0.3822 (0.3796) time: 0.6413 data: 0.5754 max mem: 9657 +[14:57:44.368729] Epoch: [39] Total time: 0:00:02 (0.6604 s / it) +[14:57:44.369506] Averaged stats: lr: 0.000097 loss: 0.3822 (0.3796) +[14:57:47.029673] val: [0/2] eta: 0:00:05 loss: 0.2428 (0.2428) time: 2.6482 data: 2.6313 max mem: 9657 +[14:57:47.038995] val: [1/2] eta: 0:00:01 loss: 0.2428 (0.6237) time: 1.3285 data: 1.3157 max mem: 9657 +[14:57:47.109216] val: Total time: 0:00:02 (1.3643 s / it) +[14:57:47.118029] val loss: 0.6236768960952759 +[14:57:47.118203] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8313, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.8113, Kappa: 0.4362, Score: 0.6616 +[14:57:47.149500] Best epoch = 22, Best score = 0.7061 +[14:57:47.405803] log_dir: ./output_logs/retfound +[14:57:49.552286] Epoch: [40] [0/4] eta: 0:00:08 lr: 0.000092 loss: 0.5698 (0.5698) time: 2.1456 data: 2.0697 max mem: 9657 +[14:57:49.748449] Epoch: [40] [3/4] eta: 0:00:00 lr: 0.000080 loss: 0.3245 (0.3854) time: 0.5853 data: 0.5175 max mem: 9657 +[14:57:49.818945] Epoch: [40] Total time: 0:00:02 (0.6032 s / it) +[14:57:49.819719] Averaged stats: lr: 0.000080 loss: 0.3245 (0.3854) +[14:57:52.543865] val: [0/2] eta: 0:00:05 loss: 0.2649 (0.2649) time: 2.7135 data: 2.6965 max mem: 9657 +[14:57:52.553465] val: [1/2] eta: 0:00:01 loss: 0.2649 (0.6044) time: 1.3613 data: 1.3483 max mem: 9657 +[14:57:52.627848] val: Total time: 0:00:02 (1.3991 s / it) +[14:57:52.636504] val loss: 0.6044119298458099 +[14:57:52.636680] Accuracy: 0.8333, F1 Score: 0.7778, ROC AUC: 0.8281, Hamming Loss: 0.1667, + Jaccard Score: 0.6500, Precision: 0.7698, Recall: 0.7875, + Average Precision: 0.8096, Kappa: 0.5559, Score: 0.7206 +[14:57:54.539457] Best epoch = 40, Best score = 0.7206 +[14:57:54.613755] log_dir: ./output_logs/retfound +[14:57:56.815590] Epoch: [41] [0/4] eta: 0:00:08 lr: 0.000076 loss: 0.3139 (0.3139) time: 2.2009 data: 2.1265 max mem: 9657 +[14:57:57.069045] Epoch: [41] [3/4] eta: 0:00:00 lr: 0.000064 loss: 0.3262 (0.3924) time: 0.6134 data: 0.5461 max mem: 9657 +[14:57:57.143706] Epoch: [41] Total time: 0:00:02 (0.6324 s / it) +[14:57:57.144519] Averaged stats: lr: 0.000064 loss: 0.3262 (0.3924) +[14:57:59.812344] val: [0/2] eta: 0:00:05 loss: 0.2478 (0.2478) time: 2.6564 data: 2.6387 max mem: 9657 +[14:57:59.821217] val: [1/2] eta: 0:00:01 loss: 0.2478 (0.6178) time: 1.3324 data: 1.3194 max mem: 9657 +[14:57:59.888956] val: Total time: 0:00:02 (1.3668 s / it) +[14:57:59.898481] val loss: 0.6177876591682434 +[14:57:59.898766] Accuracy: 0.7857, F1 Score: 0.6939, ROC AUC: 0.8281, Hamming Loss: 0.2143, + Jaccard Score: 0.5569, Precision: 0.7020, Recall: 0.6875, + Average Precision: 0.8097, Kappa: 0.3883, Score: 0.6368 +[14:57:59.946621] Best epoch = 40, Best score = 0.7206 +[14:58:00.232857] log_dir: ./output_logs/retfound +[14:58:02.710913] Epoch: [42] [0/4] eta: 0:00:09 lr: 0.000061 loss: 0.4462 (0.4462) time: 2.4770 data: 2.4073 max mem: 9657 +[14:58:02.907116] Epoch: [42] [3/4] eta: 0:00:00 lr: 0.000050 loss: 0.2929 (0.3248) time: 0.6681 data: 0.6019 max mem: 9657 +[14:58:02.976892] Epoch: [42] Total time: 0:00:02 (0.6860 s / it) +[14:58:02.977604] Averaged stats: lr: 0.000050 loss: 0.2929 (0.3248) +[14:58:05.623437] val: [0/2] eta: 0:00:05 loss: 0.2199 (0.2199) time: 2.6360 data: 2.6190 max mem: 9657 +[14:58:05.632477] val: [1/2] eta: 0:00:01 loss: 0.2199 (0.6447) time: 1.3222 data: 1.3096 max mem: 9657 +[14:58:05.704889] val: Total time: 0:00:02 (1.3590 s / it) +[14:58:05.713710] val loss: 0.644670307636261 +[14:58:05.713919] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8344, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.8182, Kappa: 0.4362, Score: 0.6626 +[14:58:05.748713] Best epoch = 40, Best score = 0.7206 +[14:58:06.038368] log_dir: ./output_logs/retfound +[14:58:08.378014] Epoch: [43] [0/4] eta: 0:00:09 lr: 0.000047 loss: 0.3963 (0.3963) time: 2.3386 data: 2.2686 max mem: 9657 +[14:58:08.574944] Epoch: [43] [3/4] eta: 0:00:00 lr: 0.000038 loss: 0.3621 (0.4008) time: 0.6337 data: 0.5675 max mem: 9657 +[14:58:08.642736] Epoch: [43] Total time: 0:00:02 (0.6511 s / it) +[14:58:08.643472] Averaged stats: lr: 0.000038 loss: 0.3621 (0.4008) +[14:58:11.323280] val: [0/2] eta: 0:00:05 loss: 0.2050 (0.2050) time: 2.6602 data: 2.6432 max mem: 9657 +[14:58:11.333037] val: [1/2] eta: 0:00:01 loss: 0.2050 (0.6629) time: 1.3343 data: 1.3217 max mem: 9657 +[14:58:11.400100] val: Total time: 0:00:02 (1.3688 s / it) +[14:58:11.409003] val loss: 0.6628798842430115 +[14:58:11.409171] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8375, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.8199, Kappa: 0.4362, Score: 0.6636 +[14:58:11.442141] Best epoch = 40, Best score = 0.7206 +[14:58:11.702003] log_dir: ./output_logs/retfound +[14:58:14.007807] Epoch: [44] [0/4] eta: 0:00:09 lr: 0.000035 loss: 0.3702 (0.3702) time: 2.3049 data: 2.2357 max mem: 9657 +[14:58:14.229769] Epoch: [44] [3/4] eta: 0:00:00 lr: 0.000027 loss: 0.3732 (0.3888) time: 0.6316 data: 0.5655 max mem: 9657 +[14:58:14.302022] Epoch: [44] Total time: 0:00:02 (0.6500 s / it) +[14:58:14.302759] Averaged stats: lr: 0.000027 loss: 0.3732 (0.3888) +[14:58:16.889539] val: [0/2] eta: 0:00:05 loss: 0.1938 (0.1938) time: 2.5753 data: 2.5583 max mem: 9657 +[14:58:16.898550] val: [1/2] eta: 0:00:01 loss: 0.1938 (0.6781) time: 1.2919 data: 1.2792 max mem: 9657 +[14:58:16.970462] val: Total time: 0:00:02 (1.3284 s / it) +[14:58:16.979284] val loss: 0.678103506565094 +[14:58:16.979465] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8383, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8199, Kappa: 0.4878, Score: 0.6893 +[14:58:17.019525] Best epoch = 40, Best score = 0.7206 +[14:58:17.282337] log_dir: ./output_logs/retfound +[14:58:19.565514] Epoch: [45] [0/4] eta: 0:00:09 lr: 0.000025 loss: 0.5618 (0.5618) time: 2.2823 data: 2.2128 max mem: 9657 +[14:58:19.808502] Epoch: [45] [3/4] eta: 0:00:00 lr: 0.000018 loss: 0.2835 (0.3725) time: 0.6312 data: 0.5652 max mem: 9657 +[14:58:19.877301] Epoch: [45] Total time: 0:00:02 (0.6487 s / it) +[14:58:19.878095] Averaged stats: lr: 0.000018 loss: 0.2835 (0.3725) +[14:58:22.603853] val: [0/2] eta: 0:00:05 loss: 0.1825 (0.1825) time: 2.7145 data: 2.6975 max mem: 9657 +[14:58:22.612985] val: [1/2] eta: 0:00:01 loss: 0.1825 (0.6932) time: 1.3616 data: 1.3488 max mem: 9657 +[14:58:22.681419] val: Total time: 0:00:02 (1.3964 s / it) +[14:58:22.694161] val loss: 0.6932103633880615 +[14:58:22.694357] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8406, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8214, Kappa: 0.3913, Score: 0.6410 +[14:58:22.729483] Best epoch = 40, Best score = 0.7206 +[14:58:22.998964] log_dir: ./output_logs/retfound +[14:58:25.292103] Epoch: [46] [0/4] eta: 0:00:09 lr: 0.000016 loss: 0.3904 (0.3904) time: 2.2917 data: 2.2185 max mem: 9657 +[14:58:25.490636] Epoch: [46] [3/4] eta: 0:00:00 lr: 0.000011 loss: 0.3904 (0.4152) time: 0.6224 data: 0.5547 max mem: 9657 +[14:58:25.557464] Epoch: [46] Total time: 0:00:02 (0.6396 s / it) +[14:58:25.558218] Averaged stats: lr: 0.000011 loss: 0.3904 (0.4152) +[14:58:28.172681] val: [0/2] eta: 0:00:05 loss: 0.1747 (0.1747) time: 2.6035 data: 2.5865 max mem: 9657 +[14:58:28.181847] val: [1/2] eta: 0:00:01 loss: 0.1747 (0.7041) time: 1.3061 data: 1.2933 max mem: 9657 +[14:58:28.249595] val: Total time: 0:00:02 (1.3406 s / it) +[14:58:28.258355] val loss: 0.7041279077529907 +[14:58:28.258645] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8406, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8214, Kappa: 0.3913, Score: 0.6410 +[14:58:28.294468] Best epoch = 40, Best score = 0.7206 +[14:58:28.548025] log_dir: ./output_logs/retfound +[14:58:30.800497] Epoch: [47] [0/4] eta: 0:00:09 lr: 0.000010 loss: 0.3822 (0.3822) time: 2.2516 data: 2.1826 max mem: 9657 +[14:58:30.994761] Epoch: [47] [3/4] eta: 0:00:00 lr: 0.000006 loss: 0.3822 (0.4257) time: 0.6113 data: 0.5457 max mem: 9657 +[14:58:31.064346] Epoch: [47] Total time: 0:00:02 (0.6290 s / it) +[14:58:31.065092] Averaged stats: lr: 0.000006 loss: 0.3822 (0.4257) +[14:58:33.642558] val: [0/2] eta: 0:00:05 loss: 0.1724 (0.1724) time: 2.5656 data: 2.5488 max mem: 9657 +[14:58:33.651373] val: [1/2] eta: 0:00:01 loss: 0.1724 (0.7077) time: 1.2870 data: 1.2745 max mem: 9657 +[14:58:33.718739] val: Total time: 0:00:02 (1.3213 s / it) +[14:58:33.727458] val loss: 0.7076523303985596 +[14:58:33.727624] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8406, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8214, Kappa: 0.3913, Score: 0.6410 +[14:58:33.763673] Best epoch = 40, Best score = 0.7206 +[14:58:34.018234] log_dir: ./output_logs/retfound +[14:58:36.321223] Epoch: [48] [0/4] eta: 0:00:09 lr: 0.000005 loss: 0.4920 (0.4920) time: 2.3022 data: 2.2335 max mem: 9657 +[14:58:36.515517] Epoch: [48] [3/4] eta: 0:00:00 lr: 0.000003 loss: 0.4193 (0.4803) time: 0.6240 data: 0.5584 max mem: 9657 +[14:58:36.589281] Epoch: [48] Total time: 0:00:02 (0.6427 s / it) +[14:58:36.590079] Averaged stats: lr: 0.000003 loss: 0.4193 (0.4803) +[14:58:39.104254] val: [0/2] eta: 0:00:05 loss: 0.1719 (0.1719) time: 2.5028 data: 2.4858 max mem: 9657 +[14:58:39.113421] val: [1/2] eta: 0:00:01 loss: 0.1719 (0.7085) time: 1.2557 data: 1.2430 max mem: 9657 +[14:58:39.180024] val: Total time: 0:00:02 (1.2896 s / it) +[14:58:39.188722] val loss: 0.7084973454475403 +[14:58:39.188899] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8406, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8214, Kappa: 0.3913, Score: 0.6410 +[14:58:39.223874] Best epoch = 40, Best score = 0.7206 +[14:58:39.476501] log_dir: ./output_logs/retfound +[14:58:41.724037] Epoch: [49] [0/4] eta: 0:00:08 lr: 0.000002 loss: 0.4247 (0.4247) time: 2.2466 data: 2.1769 max mem: 9657 +[14:58:41.919229] Epoch: [49] [3/4] eta: 0:00:00 lr: 0.000001 loss: 0.3583 (0.3619) time: 0.6103 data: 0.5443 max mem: 9657 +[14:58:41.987867] Epoch: [49] Total time: 0:00:02 (0.6278 s / it) +[14:58:41.988606] Averaged stats: lr: 0.000001 loss: 0.3583 (0.3619) +[14:58:44.563826] val: [0/2] eta: 0:00:05 loss: 0.1720 (0.1720) time: 2.5646 data: 2.5475 max mem: 9657 +[14:58:44.572772] val: [1/2] eta: 0:00:01 loss: 0.1720 (0.7083) time: 1.2865 data: 1.2738 max mem: 9657 +[14:58:44.638658] val: Total time: 0:00:02 (1.3200 s / it) +[14:58:44.647242] val loss: 0.7082802057266235 +[14:58:44.647472] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8406, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8214, Kappa: 0.3913, Score: 0.6410 +[14:58:44.686352] Best epoch = 40, Best score = 0.7206 +[14:58:47.874918] Test with the best model, epoch = 40: +[14:58:50.477556] test: [0/3] eta: 0:00:07 loss: 0.3731 (0.3731) time: 2.5917 data: 2.5744 max mem: 9657 +[14:58:50.540925] test: [2/3] eta: 0:00:00 loss: 0.3731 (0.5039) time: 0.8849 data: 0.8582 max mem: 9657 +[14:58:50.603927] test: Total time: 0:00:02 (0.9063 s / it) +[14:58:50.614026] val loss: 0.5038713614145914 +[14:58:50.614133] Accuracy: 0.7976, F1 Score: 0.7054, ROC AUC: 0.8091, Hamming Loss: 0.2024, + Jaccard Score: 0.5703, Precision: 0.6905, Recall: 0.7316, + Average Precision: 0.7304, Kappa: 0.4138, Score: 0.6427 +[14:58:51.293957] Training time 0:05:03 +[rank0]:[W701 14:58:51.656895535 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/050/retfound acc=0.7976 auroc=0.8088235294117647 f1_macro=0.7054 qwk=0.4137931034482759 diff --git a/results/downsample/papila/050/vit/confusion_matrix.png b/results/downsample/papila/050/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..b19b326d424d12ccf9ba8da69edc634ab65f328d --- /dev/null +++ b/results/downsample/papila/050/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29d5ce94c38b847688b3bc323ea9bb9ab25537c0ca0d7b82a2b6af8262253e4f +size 70957 diff --git a/results/downsample/papila/050/vit/log.csv b/results/downsample/papila/050/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..a5babfdd7862b733f023aa3c7e83eef4ea1a5c34 --- /dev/null +++ b/results/downsample/papila/050/vit/log.csv @@ -0,0 +1,31 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.7593023777008057,0.42857142857142855,0.565625,0.3531881313131313,3.697205891018715e-08 +1,0.7366253435611725,0.5714285714285714,0.6156250000000001,0.4578169210722263,1.1091617673056146e-07 +2,0.7318039834499359,0.35714285714285715,0.715625,0.37217665912518855,1.8486029455093576e-07 +3,0.7052681148052216,0.6904761904761905,0.765625,0.5558185028248588,2.5880441237131e-07 +4,0.7180840075016022,0.42857142857142855,0.79375,0.4526696762141967,3.327485301916844e-07 +5,0.6304374933242798,0.5714285714285714,0.796875,0.5324260752688171,3.6960797720518913e-07 +6,0.6011975109577179,0.6666666666666666,0.728125,0.5040331196581197,3.687079050660697e-07 +7,0.5684238374233246,0.5238095238095238,0.675,0.4473715651135006,3.6691214584881775e-07 +8,0.5141042768955231,0.5,0.671875,0.3783507050111585,3.6422944831183056e-07 +9,0.5242394506931305,0.5714285714285714,0.70625,0.432545518207283,3.6067288228784465e-07 +10,0.4669893980026245,0.5952380952380952,0.7625,0.5366969983869836,3.5625977500902404e-07 +11,0.49912726879119873,0.7380952380952381,0.834375,0.5957611514730096,3.5101162669042713e-07 +12,0.5155495405197144,0.7380952380952381,0.86875,0.6072194848063428,3.449540057831202e-07 +13,0.5049877762794495,0.7619047619047619,0.865625,0.705997882572495,3.381164244072535e-07 +14,0.48200932145118713,0.7619047619047619,0.865625,0.705997882572495,3.3053219457197917e-07 +15,0.4502372443675995,0.7619047619047619,0.8625,0.6646367521367521,3.222382658826906e-07 +16,0.46321888267993927,0.7380952380952381,0.8625,0.6262715653156564,3.13275045526259e-07 +17,0.44813093543052673,0.7142857142857143,0.875,0.671531086907193,3.036862014112813e-07 +18,0.4423871636390686,0.7142857142857143,0.8875,0.6756977535738597,2.9351844942241844e-07 +19,0.4433220624923706,0.8095238095238095,0.884375,0.7527748599439775,2.8282132582530113e-07 +20,0.40044471621513367,0.8095238095238095,0.884375,0.7527748599439775,2.716469459308236e-07 +21,0.42482155561447144,0.7380952380952381,0.88125,0.6920825121487021,2.60049750194587e-07 +22,0.4015875309705734,0.7380952380952381,0.8875,0.6941658454820354,2.4808623898847216e-07 +23,0.43150418996810913,0.7857142857142857,0.88125,0.7310618062361747,2.3581469733650727e-07 +24,0.3805621862411499,0.7142857142857143,0.88125,0.6129807692307692,2.232949109560898e-07 +25,0.37962038815021515,0.7380952380952381,0.884375,0.6931241788153688,2.1058787498798055e-07 +26,0.41069748997688293,0.7380952380952381,0.875,0.6899991788153687,1.9775549683410935e-07 +27,0.3684448003768921,0.7380952380952381,0.871875,0.688957512148702,1.8486029455093576e-07 +28,0.41774076223373413,0.7380952380952381,0.86875,0.6879158454820354,1.7196509226776213e-07 +29,0.41247278451919556,0.7380952380952381,0.871875,0.6632720294590847,1.5913271411389099e-07 diff --git a/results/downsample/papila/050/vit/metrics.json b/results/downsample/papila/050/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..8f09fcceb8d53e39dbe136ee82cbd66ee3e59c79 --- /dev/null +++ b/results/downsample/papila/050/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.6904761904761905, + "balanced_accuracy": 0.6893382352941176, + "precision_macro": 0.6237980769230769, + "recall_macro": 0.6893382352941176, + "f1_macro": 0.6208333333333333, + "precision_weighted": 0.7971611721611721, + "recall_weighted": 0.6904761904761905, + "f1_weighted": 0.7214285714285714, + "cohen_kappa": 0.27393617021276595, + "quadratic_weighted_kappa": 0.27393617021276595, + "mcc": 0.30620064936195557, + "auroc": 0.7297794117647058, + "auprc": 0.4713836824237678, + "sensitivity": 0.6875, + "specificity": 0.6911764705882353, + "precision_pos": 0.34375, + "f1_pos": 0.4583333333333333, + "per_class": { + "0": { + "precision": 0.9038461538461539, + "recall": 0.6911764705882353, + "f1-score": 0.7833333333333333, + "support": 68.0 + }, + "1": { + "precision": 0.34375, + "recall": 0.6875, + "f1-score": 0.4583333333333333, + "support": 16.0 + }, + "accuracy": 0.6904761904761905, + "macro avg": { + "precision": 0.6237980769230769, + "recall": 0.6893382352941176, + "f1-score": 0.6208333333333333, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.7971611721611721, + "recall": 0.6904761904761905, + "f1-score": 0.7214285714285714, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/050/vit/pr.png b/results/downsample/papila/050/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..c6fe07c6fa1bca014df16a394bffef1cd9a4fbd6 --- /dev/null +++ b/results/downsample/papila/050/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:26ced839496f964c40a2329120ffbb94df569a5927e390ccf25641a1b9062dd2 +size 52083 diff --git a/results/downsample/papila/050/vit/roc.png b/results/downsample/papila/050/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..e85ca2fa5f10a62d65e357c4a2e66661f9060d8a --- /dev/null +++ b/results/downsample/papila/050/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:973bdcec526c9e65fb3568bacfc768bc079deb849fcb0e8421487724b861e533 +size 57359 diff --git a/results/downsample/papila/050/vit/test_pred.npz b/results/downsample/papila/050/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..578611f06c8e59f7f0896c8c12eb9535131eb79a --- /dev/null +++ b/results/downsample/papila/050/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2cbafcacd2d77eb0e2940a1a9e5948f1ca4849eda44525126d7baa8d3d3fabd8 +size 1854 diff --git a/results/downsample/papila/050/vit/train.log b/results/downsample/papila/050/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..c36536b8c3347c1b7ff4ef84e80d572a62819edf --- /dev/null +++ b/results/downsample/papila/050/vit/train.log @@ -0,0 +1,159 @@ +[vit] train=146 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.7593 val_acc=0.4286 val_auc=0.5656 score=0.3532 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7366 val_acc=0.5714 val_auc=0.6156 score=0.4578 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.7318 val_acc=0.3571 val_auc=0.7156 score=0.3722 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.7053 val_acc=0.6905 val_auc=0.7656 score=0.5558 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.7181 val_acc=0.4286 val_auc=0.7937 score=0.4527 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.6304 val_acc=0.5714 val_auc=0.7969 score=0.5324 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.6012 val_acc=0.6667 val_auc=0.7281 score=0.5040 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5684 val_acc=0.5238 val_auc=0.6750 score=0.4474 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5141 val_acc=0.5000 val_auc=0.6719 score=0.3784 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.5242 val_acc=0.5714 val_auc=0.7063 score=0.4325 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.4670 val_acc=0.5952 val_auc=0.7625 score=0.5367 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.4991 val_acc=0.7381 val_auc=0.8344 score=0.5958 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.5155 val_acc=0.7381 val_auc=0.8688 score=0.6072 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.5050 val_acc=0.7619 val_auc=0.8656 score=0.7060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4820 val_acc=0.7619 val_auc=0.8656 score=0.7060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4502 val_acc=0.7619 val_auc=0.8625 score=0.6646 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4632 val_acc=0.7381 val_auc=0.8625 score=0.6263 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.4481 val_acc=0.7143 val_auc=0.8750 score=0.6715 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4424 val_acc=0.7143 val_auc=0.8875 score=0.6757 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4433 val_acc=0.8095 val_auc=0.8844 score=0.7528 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.4004 val_acc=0.8095 val_auc=0.8844 score=0.7528 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4248 val_acc=0.7381 val_auc=0.8812 score=0.6921 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.4016 val_acc=0.7381 val_auc=0.8875 score=0.6942 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4315 val_acc=0.7857 val_auc=0.8812 score=0.7311 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.3806 val_acc=0.7143 val_auc=0.8812 score=0.6130 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3796 val_acc=0.7381 val_auc=0.8844 score=0.6931 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4107 val_acc=0.7381 val_auc=0.8750 score=0.6900 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3684 val_acc=0.7381 val_auc=0.8719 score=0.6890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.4177 val_acc=0.7381 val_auc=0.8688 score=0.6879 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.4125 val_acc=0.7381 val_auc=0.8719 score=0.6633 +[vit] early stop at ep29 (best ep19 score=0.7528) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=19 best_val_score=0.7528 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/050/vit acc=0.6905 auroc=0.7297794117647058 f1_macro=0.6208 qwk=0.27393617021276595 diff --git a/results/downsample/papila/100/papila_100pct/confusion_matrix_test.jpg b/results/downsample/papila/100/papila_100pct/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..bb78e1f3cd4bbca66196dee1305d5bd217471669 --- /dev/null +++ b/results/downsample/papila/100/papila_100pct/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f47a6c410ec7a3629543c9c4cb1093b2b9635932c2ab9c8ee089a837df3e300 +size 258603 diff --git a/results/downsample/papila/100/papila_100pct/log.txt b/results/downsample/papila/100/papila_100pct/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..5177c358e09ebc9ed7d60591cc8f079a13f8fa59 --- /dev/null +++ b/results/downsample/papila/100/papila_100pct/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.777777777777778e-05, "train_loss": 0.6827528211805556, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.027777777777779e-05, "train_loss": 0.5992067125108507, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015277777777777777, "train_loss": 0.5172894795735677, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021527777777777778, "train_loss": 0.5202977922227647, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002777777777777778, "train_loss": 0.4997452629937066, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003402777777777778, "train_loss": 0.5041298336452908, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.0004027777777777778, "train_loss": 0.4900501039293077, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004652777777777778, "train_loss": 0.5020377900865343, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005277777777777777, "train_loss": 0.4713812934027778, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005902777777777778, "train_loss": 0.44863936636183, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247307916747557, "train_loss": 0.45878227551778156, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006229157334469918, "train_loss": 0.4269802040523953, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006191899416650879, "train_loss": 0.4496183395385742, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006135763870743313, "train_loss": 0.4470519489712185, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006061096791054699, "train_loss": 0.45556622081332737, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005968358524960642, "train_loss": 0.3853343062930637, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005858120834710217, "train_loss": 0.4626990424262153, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005731063372321567, "train_loss": 0.4492563141716851, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005587969489301213, "train_loss": 0.42685553762647843, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005429721407021519, "train_loss": 0.39957908789316815, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005257294777532569, "train_loss": 0.38024022844102645, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005071752668342834, "train_loss": 0.3989546100298564, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00048742390082544794, "train_loss": 0.3949376742045085, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0004665971534661909, "train_loss": 0.3674582905239529, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004448234285795783, "train_loss": 0.3995681140157912, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004222369684200342, "train_loss": 0.3284359508090549, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.00039897702602520257, "train_loss": 0.34738753901587593, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003751870066746631, "train_loss": 0.39103036456637913, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035101358374869636, "train_loss": 0.41759975088967216, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003266057944381195, "train_loss": 0.3715149561564128, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0003021141208804459, "train_loss": 0.3839879764450921, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002776895623874664, "train_loss": 0.34202857149971855, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002534827044842803, "train_loss": 0.3478359712494744, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022964279049945762, "train_loss": 0.39175425635443795, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00020631680143029074, "train_loss": 0.34169335497750175, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018364854975606965, "train_loss": 0.37303222550286186, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016177779278632443, "train_loss": 0.3513450225194295, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014083937101053823, "train_loss": 0.3313828508059184, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012096237676169061, "train_loss": 0.3487516575389438, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010226935831909702, "train_loss": 0.32215451531940037, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.48755643575155e-05, "train_loss": 0.35140036212073433, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.888823340074255e-05, "train_loss": 0.319819880856408, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.4405932660453715e-05, "train_loss": 0.3057285149892171, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.1517950336566605e-05, "train_loss": 0.3327277766333686, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.0303745125797026e-05, "train_loss": 0.34663278195593095, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.0832456332370918e-05, "train_loss": 0.37178613079918754, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3162477601222926e-05, "train_loss": 0.3343882891866896, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.341096901757339e-06, "train_loss": 0.36498680379655624, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.404204981791873e-06, "train_loss": 0.3455526265833113, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.3760740891625011e-06, "train_loss": 0.32421498828464085, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/100/papila_100pct/metrics_test.csv b/results/downsample/papila/100/papila_100pct/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..3b498a846f59c6dbc4ece55587c9a674e54ca0fb --- /dev/null +++ b/results/downsample/papila/100/papila_100pct/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.47513073682785034,0.8333333333333334,0.7418788410886743,0.8370863970588236,0.16666666666666666,0.6137387387387387,0.7323232323232323,0.7536764705882353,0.7821000271484148,0.4842105263157894 diff --git a/results/downsample/papila/100/papila_100pct/metrics_val.csv b/results/downsample/papila/100/papila_100pct/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..b96c44224ca570b2dd37800ded75d415ef0ae917 --- /dev/null +++ b/results/downsample/papila/100/papila_100pct/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6945861876010895,0.7619047619047619,0.43243243243243246,0.68671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6626308049825587,0.0 +0.7484893798828125,0.7619047619047619,0.43243243243243246,0.77265625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7154166661741241,0.0 +0.9287281036376953,0.7619047619047619,0.43243243243243246,0.80859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7687674774801041,0.0 +0.8795062899589539,0.7619047619047619,0.43243243243243246,0.79375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.750414290631294,0.0 +0.8250083923339844,0.7619047619047619,0.43243243243243246,0.75546875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6976555889940705,0.0 +0.8058345913887024,0.7619047619047619,0.43243243243243246,0.7093750000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6559102729397122,0.0 +0.7665609121322632,0.7619047619047619,0.43243243243243246,0.734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6563890461269495,0.0 +0.8525406122207642,0.7619047619047619,0.43243243243243246,0.75,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6717796242170622,0.0 +0.6245647370815277,0.7142857142857143,0.5367647058823529,0.75,0.2857142857142857,0.4214285714285714,0.5555555555555556,0.5375,0.6939900141268103,0.08695652173913038 +0.7787399291992188,0.7619047619047619,0.6139705882352942,0.79375,0.23809523809523808,0.4871794871794872,0.6527777777777778,0.603125,0.730369188055292,0.23913043478260865 +0.5484199523925781,0.7619047619047619,0.671875,0.815625,0.23809523809523808,0.5315315315315315,0.671875,0.671875,0.7518565475904775,0.34375 +1.2771737575531006,0.7857142857142857,0.590465872156013,0.85,0.21428571428571427,0.4784090909090909,0.7307692307692307,0.584375,0.80314870255797,0.22222222222222232 +0.9774452447891235,0.8333333333333334,0.7159420289855072,0.84453125,0.16666666666666666,0.5897129186602871,0.818918918918919,0.684375,0.8015047226759677,0.44528301886792454 +0.8752020001411438,0.8095238095238095,0.6911764705882353,0.86796875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8321278862142879,0.3913043478260869 +0.7527960538864136,0.8095238095238095,0.7375,0.8531249999999999,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8224014286197165,0.475 +0.9505177736282349,0.8333333333333334,0.7418788410886743,0.859375,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8066661561331343,0.4878048780487805 +1.0000401735305786,0.8095238095238095,0.6911764705882353,0.84375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7957254531473945,0.3913043478260869 +0.8677769899368286,0.8095238095238095,0.7171717171717171,0.815625,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.75446910044971,0.43624161073825496 +1.133644163608551,0.7857142857142857,0.6681299385425812,0.8070312499999999,0.21428571428571427,0.535425101214575,0.7,0.653125,0.7543627853796022,0.3414634146341464 +0.6373273730278015,0.7857142857142857,0.7142857142857142,0.825,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.785276188218651,0.42900302114803623 +0.6148214936256409,0.7857142857142857,0.7142857142857142,0.8531249999999999,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8130521054771913,0.42900302114803623 +0.6647389531135559,0.8571428571428571,0.7878787878787878,0.859375,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.830052265583969,0.5771812080536913 +0.7311892509460449,0.8333333333333334,0.7418788410886743,0.840625,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8170925996103228,0.4878048780487805 +0.8558176159858704,0.8095238095238095,0.7375,0.828125,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8103413120078523,0.475 +0.7548640370368958,0.8095238095238095,0.7375,0.821875,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8070984081914674,0.475 +0.7577589750289917,0.8571428571428571,0.7878787878787878,0.815625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7927907195509698,0.5771812080536913 +0.659015029668808,0.7380952380952381,0.6707056307911619,0.82890625,0.2619047619047619,0.5236928104575164,0.6618037135278514,0.690625,0.7990907690725269,0.34560906515580736 +1.1874847412109375,0.8095238095238095,0.6911764705882353,0.8179687499999999,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.79050604102645,0.3913043478260869 +0.6474664807319641,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.8007156258962138,0.5242718446601942 +0.5518224239349365,0.7857142857142857,0.7142857142857142,0.84375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8078735127959484,0.42900302114803623 +0.6501250863075256,0.7857142857142857,0.7142857142857142,0.84453125,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8077895392728555,0.42900302114803623 +0.648048460483551,0.7857142857142857,0.7142857142857142,0.84609375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8087281862237307,0.42900302114803623 +0.8128635883331299,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8008065294158258,0.5771812080536913 +0.7290278077125549,0.8571428571428571,0.7878787878787878,0.83671875,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8033269098506084,0.5771812080536913 +0.9214009940624237,0.8333333333333334,0.7418788410886743,0.828125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7963100577502364,0.4878048780487805 +0.9986748993396759,0.8333333333333334,0.7418788410886743,0.83125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7981890016514915,0.4878048780487805 +0.887900710105896,0.8333333333333334,0.7418788410886743,0.825,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7945901607993612,0.4878048780487805 +0.7909270226955414,0.8571428571428571,0.7878787878787878,0.82890625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929611826223955,0.5771812080536913 +0.7527517080307007,0.8571428571428571,0.7878787878787878,0.8343750000000001,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7963039314782308,0.5771812080536913 +0.7334095239639282,0.8333333333333334,0.7619433198380567,0.83203125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945614241682891,0.5242718446601942 +0.7272949814796448,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945614241682891,0.5242718446601942 +0.7320790588855743,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7945614241682891,0.5771812080536913 +0.7541973888874054,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7751908898353577,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7738757431507111,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7761802673339844,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929513475101896,0.5771812080536913 +0.7720681726932526,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7686194181442261,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7694710791110992,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7701238989830017,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 diff --git a/results/downsample/papila/100/papila_100pct/test_pred.npz b/results/downsample/papila/100/papila_100pct/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..3d702b56e1fa0fa773506bb7a507fe8432118b5c --- /dev/null +++ b/results/downsample/papila/100/papila_100pct/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:288713ad75144f9b05b057382287103c78c0443830c6b60555e9c81300d12da3 +size 1518 diff --git a/results/downsample/papila/100/resnet/confusion_matrix.png b/results/downsample/papila/100/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..315f6d5b49bbef41ad7f01c4fe8d475867ac0006 --- /dev/null +++ b/results/downsample/papila/100/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5d540f973ca1fe879d63a1019f85ee2107bc683576651d53411181cf83ec3c36 +size 72526 diff --git a/results/downsample/papila/100/resnet/log.csv b/results/downsample/papila/100/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..6d3d98999e4217fbd70f1ec790c21f7d68559d4d --- /dev/null +++ b/results/downsample/papila/100/resnet/log.csv @@ -0,0 +1,30 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6922027319669724,0.3333333333333333,0.41250000000000003,0.24582295558322956,0.000125 +1,0.6898889243602753,0.5952380952380952,0.4,0.2221741807842983,0.0002916666666666667 +2,0.6811905652284622,0.6666666666666666,0.590625,0.36969944128081705,0.0004583333333333333 +3,0.6693146228790283,0.7619047619047619,0.6499999999999999,0.464885639295088,0.0004996859161456965 +4,0.6437167376279831,0.7857142857142857,0.734375,0.46947887846105013,0.0004982915790812436 +5,0.609171137213707,0.7857142857142857,0.590625,0.46777103145941173,0.0004957883115509159 +6,0.5673289448022842,0.7857142857142857,0.684375,0.49902103145941173,0.0004921872937551814 +7,0.5232248678803444,0.7380952380952381,0.6906249999999999,0.45751669173274995,0.000487504608713676 +8,0.4348279908299446,0.7857142857142857,0.715625,0.545733353841947,0.00048176117043453436 +9,0.39112991839647293,0.7142857142857143,0.7312500000000001,0.5166666666666666,0.000474982630507352 +10,0.3208153024315834,0.6428571428571429,0.728125,0.48700211864406784,0.00046719926353695914 +11,0.25122571736574173,0.5714285714285714,0.715625,0.47500496369412293,0.00045844583192968674 +12,0.1940660960972309,0.4523809523809524,0.6312500000000001,0.35736946840591904,0.00044876143063602076 +13,0.1818530410528183,0.6904761904761905,0.6187499999999999,0.4604260578014354,0.00043818931254306284 +14,0.13051196932792664,0.7857142857142857,0.7625000000000001,0.5906977843922425,0.00042677669529663686 +15,0.07865838054567575,0.7857142857142857,0.69375,0.5920088800230597,0.0004145745504158204 +16,0.08366507478058338,0.8809523809523809,0.7000000000000001,0.7002916780603409,0.0004016373756417668 +17,0.09364914707839489,0.8571428571428571,0.80625,0.7060368712702472,0.00038802295153756415 +18,0.06360051967203617,0.8809523809523809,0.78125,0.7436746950327273,0.0003737920834262134 +19,0.07134038070216775,0.8809523809523809,0.7359375000000001,0.7122708447270076,0.0003590083298192957 +20,0.04667547903954983,0.8809523809523809,0.653125,0.6846666780603409,0.0003437377185492303 +21,0.02803577808663249,0.8571428571428571,0.659375,0.6353897169103863,0.0003280484518729466 +22,0.03978420910425484,0.8809523809523809,0.621875,0.6742500113936742,0.00031201060186404833 +23,0.016235881950706244,0.8809523809523809,0.6406249999999999,0.6805000113936742,0.0002956957974539226 +24,0.018899165326729417,0.8571428571428571,0.646875,0.6312230502437196,0.0002791769045195441 +25,0.019129932625219226,0.8333333333333334,0.6437499999999999,0.6244779063791516,0.0002625277004467798 +26,0.0318123537581414,0.8333333333333334,0.678125,0.6547800548327503,0.00024582254462267474 +27,0.02134297462180257,0.7619047619047619,0.715625,0.5978906745725506,0.00022913604632837759 +28,0.017061496968381107,0.7380952380952381,0.715625,0.577313231982323,0.00021254273151597963 diff --git a/results/downsample/papila/100/resnet/metrics.json b/results/downsample/papila/100/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..068ed22788be147812d7c7de349b507039a4cad6 --- /dev/null +++ b/results/downsample/papila/100/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8809523809523809, + "balanced_accuracy": 0.7352941176470589, + "precision_macro": 0.845945945945946, + "recall_macro": 0.7352941176470589, + "f1_macro": 0.772481040086674, + "precision_weighted": 0.8743886743886743, + "recall_weighted": 0.8809523809523809, + "f1_weighted": 0.8697312077593767, + "cohen_kappa": 0.5493562231759657, + "quadratic_weighted_kappa": 0.5493562231759657, + "mcc": 0.5706103612971936, + "auroc": 0.7720588235294117, + "auprc": 0.618210816095377, + "sensitivity": 0.5, + "specificity": 0.9705882352941176, + "precision_pos": 0.8, + "f1_pos": 0.6153846153846154, + "per_class": { + "0": { + "precision": 0.8918918918918919, + "recall": 0.9705882352941176, + "f1-score": 0.9295774647887324, + "support": 68.0 + }, + "1": { + "precision": 0.8, + "recall": 0.5, + "f1-score": 0.6153846153846154, + "support": 16.0 + }, + "accuracy": 0.8809523809523809, + "macro avg": { + "precision": 0.845945945945946, + "recall": 0.7352941176470589, + "f1-score": 0.772481040086674, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8743886743886743, + "recall": 0.8809523809523809, + "f1-score": 0.8697312077593767, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/100/resnet/pr.png b/results/downsample/papila/100/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..4d33808425dcee88b7e54acf364f99fd86739543 --- /dev/null +++ b/results/downsample/papila/100/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19d54c41b4f4c33e46425238e5953a0b2e4dc7bfdebb3ace023d623a91d6204e +size 48559 diff --git a/results/downsample/papila/100/resnet/roc.png b/results/downsample/papila/100/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..d2f9cae67c46e762aa9bbdf40ef744dd7c2100a5 --- /dev/null +++ b/results/downsample/papila/100/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c10b463ff813bcba3701326ab53273ce95f230ed7d90ca1500214870728a814 +size 57089 diff --git a/results/downsample/papila/100/resnet/test_pred.npz b/results/downsample/papila/100/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..d182273697b0b15176d4dd20ca7255b0c3d6bb62 --- /dev/null +++ b/results/downsample/papila/100/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d96262876ecc1e941617b83d3a1f857b3f455f953cc811a1cd9bf089b43df00 +size 1854 diff --git a/results/downsample/papila/100/resnet/train.log b/results/downsample/papila/100/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b8fbdc8054638065b27b9fe0b20362aa44304b42 --- /dev/null +++ b/results/downsample/papila/100/resnet/train.log @@ -0,0 +1,154 @@ +[resnet] train=294 val=42 test=84 classes=['0', '1'] +[resnet] optim groups=2 layer_decay=1.0 drop_path=0.0 ls=0.0 lr=0.0005 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6922 val_acc=0.3333 val_auc=0.4125 score=0.2458 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6899 val_acc=0.5952 val_auc=0.4000 score=0.2222 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6812 val_acc=0.6667 val_auc=0.5906 score=0.3697 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6693 val_acc=0.7619 val_auc=0.6500 score=0.4649 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6437 val_acc=0.7857 val_auc=0.7344 score=0.4695 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6092 val_acc=0.7857 val_auc=0.5906 score=0.4678 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.5673 val_acc=0.7857 val_auc=0.6844 score=0.4990 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.5232 val_acc=0.7381 val_auc=0.6906 score=0.4575 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.4348 val_acc=0.7857 val_auc=0.7156 score=0.5457 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.3911 val_acc=0.7143 val_auc=0.7313 score=0.5167 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.3208 val_acc=0.6429 val_auc=0.7281 score=0.4870 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.2512 val_acc=0.5714 val_auc=0.7156 score=0.4750 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.1941 val_acc=0.4524 val_auc=0.6313 score=0.3574 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.1819 val_acc=0.6905 val_auc=0.6187 score=0.4604 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.1305 val_acc=0.7857 val_auc=0.7625 score=0.5907 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.0787 val_acc=0.7857 val_auc=0.6937 score=0.5920 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0837 val_acc=0.8810 val_auc=0.7000 score=0.7003 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0936 val_acc=0.8571 val_auc=0.8063 score=0.7060 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0636 val_acc=0.8810 val_auc=0.7812 score=0.7437 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0713 val_acc=0.8810 val_auc=0.7359 score=0.7123 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0467 val_acc=0.8810 val_auc=0.6531 score=0.6847 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0280 val_acc=0.8571 val_auc=0.6594 score=0.6354 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0398 val_acc=0.8810 val_auc=0.6219 score=0.6743 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0162 val_acc=0.8810 val_auc=0.6406 score=0.6805 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0189 val_acc=0.8571 val_auc=0.6469 score=0.6312 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0191 val_acc=0.8333 val_auc=0.6437 score=0.6245 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0318 val_acc=0.8333 val_auc=0.6781 score=0.6548 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0213 val_acc=0.7619 val_auc=0.7156 score=0.5979 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0171 val_acc=0.7381 val_auc=0.7156 score=0.5773 +[resnet] early stop at ep28 (best ep18 score=0.7437) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=18 best_val_score=0.7437 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/100/resnet acc=0.8810 auroc=0.7720588235294117 f1_macro=0.7725 qwk=0.5493562231759657 diff --git a/results/downsample/papila/100/retfound/confusion_matrix.png b/results/downsample/papila/100/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..ac0fad06654ce85a88b195d14352067b13a9b547 --- /dev/null +++ b/results/downsample/papila/100/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfb2b0d9931914bfb7b6864539b77eb6062e8faacbf80f642185893ca80e37ac +size 74572 diff --git a/results/downsample/papila/100/retfound/confusion_matrix_test.jpg b/results/downsample/papila/100/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..bb78e1f3cd4bbca66196dee1305d5bd217471669 --- /dev/null +++ b/results/downsample/papila/100/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f47a6c410ec7a3629543c9c4cb1093b2b9635932c2ab9c8ee089a837df3e300 +size 258603 diff --git a/results/downsample/papila/100/retfound/log.txt b/results/downsample/papila/100/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..5177c358e09ebc9ed7d60591cc8f079a13f8fa59 --- /dev/null +++ b/results/downsample/papila/100/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.777777777777778e-05, "train_loss": 0.6827528211805556, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.027777777777779e-05, "train_loss": 0.5992067125108507, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015277777777777777, "train_loss": 0.5172894795735677, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021527777777777778, "train_loss": 0.5202977922227647, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002777777777777778, "train_loss": 0.4997452629937066, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003402777777777778, "train_loss": 0.5041298336452908, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.0004027777777777778, "train_loss": 0.4900501039293077, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004652777777777778, "train_loss": 0.5020377900865343, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005277777777777777, "train_loss": 0.4713812934027778, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005902777777777778, "train_loss": 0.44863936636183, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247307916747557, "train_loss": 0.45878227551778156, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006229157334469918, "train_loss": 0.4269802040523953, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006191899416650879, "train_loss": 0.4496183395385742, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006135763870743313, "train_loss": 0.4470519489712185, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006061096791054699, "train_loss": 0.45556622081332737, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005968358524960642, "train_loss": 0.3853343062930637, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005858120834710217, "train_loss": 0.4626990424262153, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005731063372321567, "train_loss": 0.4492563141716851, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005587969489301213, "train_loss": 0.42685553762647843, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005429721407021519, "train_loss": 0.39957908789316815, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005257294777532569, "train_loss": 0.38024022844102645, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005071752668342834, "train_loss": 0.3989546100298564, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00048742390082544794, "train_loss": 0.3949376742045085, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0004665971534661909, "train_loss": 0.3674582905239529, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004448234285795783, "train_loss": 0.3995681140157912, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004222369684200342, "train_loss": 0.3284359508090549, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.00039897702602520257, "train_loss": 0.34738753901587593, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003751870066746631, "train_loss": 0.39103036456637913, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035101358374869636, "train_loss": 0.41759975088967216, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003266057944381195, "train_loss": 0.3715149561564128, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0003021141208804459, "train_loss": 0.3839879764450921, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002776895623874664, "train_loss": 0.34202857149971855, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002534827044842803, "train_loss": 0.3478359712494744, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022964279049945762, "train_loss": 0.39175425635443795, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00020631680143029074, "train_loss": 0.34169335497750175, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018364854975606965, "train_loss": 0.37303222550286186, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016177779278632443, "train_loss": 0.3513450225194295, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014083937101053823, "train_loss": 0.3313828508059184, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012096237676169061, "train_loss": 0.3487516575389438, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010226935831909702, "train_loss": 0.32215451531940037, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.48755643575155e-05, "train_loss": 0.35140036212073433, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.888823340074255e-05, "train_loss": 0.319819880856408, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.4405932660453715e-05, "train_loss": 0.3057285149892171, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.1517950336566605e-05, "train_loss": 0.3327277766333686, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.0303745125797026e-05, "train_loss": 0.34663278195593095, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.0832456332370918e-05, "train_loss": 0.37178613079918754, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3162477601222926e-05, "train_loss": 0.3343882891866896, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.341096901757339e-06, "train_loss": 0.36498680379655624, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.404204981791873e-06, "train_loss": 0.3455526265833113, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.3760740891625011e-06, "train_loss": 0.32421498828464085, "epoch": 49, "n_parameters": 303303682} diff --git a/results/downsample/papila/100/retfound/metrics.json b/results/downsample/papila/100/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..381827cdf3d3dc0b25d94392a9ba5ef26d6496e5 --- /dev/null +++ b/results/downsample/papila/100/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8333333333333334, + "balanced_accuracy": 0.7536764705882353, + "precision_macro": 0.7323232323232323, + "recall_macro": 0.7536764705882353, + "f1_macro": 0.7418788410886743, + "precision_weighted": 0.8417508417508417, + "recall_weighted": 0.8333333333333334, + "f1_weighted": 0.8369915130231197, + "cohen_kappa": 0.4842105263157894, + "quadratic_weighted_kappa": 0.4842105263157894, + "mcc": 0.48553038055886144, + "auroc": 0.8373161764705882, + "auprc": 0.6334668074576429, + "sensitivity": 0.625, + "specificity": 0.8823529411764706, + "precision_pos": 0.5555555555555556, + "f1_pos": 0.5882352941176471, + "per_class": { + "0": { + "precision": 0.9090909090909091, + "recall": 0.8823529411764706, + "f1-score": 0.8955223880597015, + "support": 68.0 + }, + "1": { + "precision": 0.5555555555555556, + "recall": 0.625, + "f1-score": 0.5882352941176471, + "support": 16.0 + }, + "accuracy": 0.8333333333333334, + "macro avg": { + "precision": 0.7323232323232323, + "recall": 0.7536764705882353, + "f1-score": 0.7418788410886743, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8417508417508417, + "recall": 0.8333333333333334, + "f1-score": 0.8369915130231197, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/100/retfound/metrics_test.csv b/results/downsample/papila/100/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..3b498a846f59c6dbc4ece55587c9a674e54ca0fb --- /dev/null +++ b/results/downsample/papila/100/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.47513073682785034,0.8333333333333334,0.7418788410886743,0.8370863970588236,0.16666666666666666,0.6137387387387387,0.7323232323232323,0.7536764705882353,0.7821000271484148,0.4842105263157894 diff --git a/results/downsample/papila/100/retfound/metrics_val.csv b/results/downsample/papila/100/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..b96c44224ca570b2dd37800ded75d415ef0ae917 --- /dev/null +++ b/results/downsample/papila/100/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6945861876010895,0.7619047619047619,0.43243243243243246,0.68671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6626308049825587,0.0 +0.7484893798828125,0.7619047619047619,0.43243243243243246,0.77265625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7154166661741241,0.0 +0.9287281036376953,0.7619047619047619,0.43243243243243246,0.80859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7687674774801041,0.0 +0.8795062899589539,0.7619047619047619,0.43243243243243246,0.79375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.750414290631294,0.0 +0.8250083923339844,0.7619047619047619,0.43243243243243246,0.75546875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6976555889940705,0.0 +0.8058345913887024,0.7619047619047619,0.43243243243243246,0.7093750000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6559102729397122,0.0 +0.7665609121322632,0.7619047619047619,0.43243243243243246,0.734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6563890461269495,0.0 +0.8525406122207642,0.7619047619047619,0.43243243243243246,0.75,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6717796242170622,0.0 +0.6245647370815277,0.7142857142857143,0.5367647058823529,0.75,0.2857142857142857,0.4214285714285714,0.5555555555555556,0.5375,0.6939900141268103,0.08695652173913038 +0.7787399291992188,0.7619047619047619,0.6139705882352942,0.79375,0.23809523809523808,0.4871794871794872,0.6527777777777778,0.603125,0.730369188055292,0.23913043478260865 +0.5484199523925781,0.7619047619047619,0.671875,0.815625,0.23809523809523808,0.5315315315315315,0.671875,0.671875,0.7518565475904775,0.34375 +1.2771737575531006,0.7857142857142857,0.590465872156013,0.85,0.21428571428571427,0.4784090909090909,0.7307692307692307,0.584375,0.80314870255797,0.22222222222222232 +0.9774452447891235,0.8333333333333334,0.7159420289855072,0.84453125,0.16666666666666666,0.5897129186602871,0.818918918918919,0.684375,0.8015047226759677,0.44528301886792454 +0.8752020001411438,0.8095238095238095,0.6911764705882353,0.86796875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8321278862142879,0.3913043478260869 +0.7527960538864136,0.8095238095238095,0.7375,0.8531249999999999,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8224014286197165,0.475 +0.9505177736282349,0.8333333333333334,0.7418788410886743,0.859375,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8066661561331343,0.4878048780487805 +1.0000401735305786,0.8095238095238095,0.6911764705882353,0.84375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7957254531473945,0.3913043478260869 +0.8677769899368286,0.8095238095238095,0.7171717171717171,0.815625,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.75446910044971,0.43624161073825496 +1.133644163608551,0.7857142857142857,0.6681299385425812,0.8070312499999999,0.21428571428571427,0.535425101214575,0.7,0.653125,0.7543627853796022,0.3414634146341464 +0.6373273730278015,0.7857142857142857,0.7142857142857142,0.825,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.785276188218651,0.42900302114803623 +0.6148214936256409,0.7857142857142857,0.7142857142857142,0.8531249999999999,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8130521054771913,0.42900302114803623 +0.6647389531135559,0.8571428571428571,0.7878787878787878,0.859375,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.830052265583969,0.5771812080536913 +0.7311892509460449,0.8333333333333334,0.7418788410886743,0.840625,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8170925996103228,0.4878048780487805 +0.8558176159858704,0.8095238095238095,0.7375,0.828125,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8103413120078523,0.475 +0.7548640370368958,0.8095238095238095,0.7375,0.821875,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8070984081914674,0.475 +0.7577589750289917,0.8571428571428571,0.7878787878787878,0.815625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7927907195509698,0.5771812080536913 +0.659015029668808,0.7380952380952381,0.6707056307911619,0.82890625,0.2619047619047619,0.5236928104575164,0.6618037135278514,0.690625,0.7990907690725269,0.34560906515580736 +1.1874847412109375,0.8095238095238095,0.6911764705882353,0.8179687499999999,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.79050604102645,0.3913043478260869 +0.6474664807319641,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.8007156258962138,0.5242718446601942 +0.5518224239349365,0.7857142857142857,0.7142857142857142,0.84375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8078735127959484,0.42900302114803623 +0.6501250863075256,0.7857142857142857,0.7142857142857142,0.84453125,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8077895392728555,0.42900302114803623 +0.648048460483551,0.7857142857142857,0.7142857142857142,0.84609375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8087281862237307,0.42900302114803623 +0.8128635883331299,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8008065294158258,0.5771812080536913 +0.7290278077125549,0.8571428571428571,0.7878787878787878,0.83671875,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8033269098506084,0.5771812080536913 +0.9214009940624237,0.8333333333333334,0.7418788410886743,0.828125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7963100577502364,0.4878048780487805 +0.9986748993396759,0.8333333333333334,0.7418788410886743,0.83125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7981890016514915,0.4878048780487805 +0.887900710105896,0.8333333333333334,0.7418788410886743,0.825,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7945901607993612,0.4878048780487805 +0.7909270226955414,0.8571428571428571,0.7878787878787878,0.82890625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929611826223955,0.5771812080536913 +0.7527517080307007,0.8571428571428571,0.7878787878787878,0.8343750000000001,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7963039314782308,0.5771812080536913 +0.7334095239639282,0.8333333333333334,0.7619433198380567,0.83203125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945614241682891,0.5242718446601942 +0.7272949814796448,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945614241682891,0.5242718446601942 +0.7320790588855743,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7945614241682891,0.5771812080536913 +0.7541973888874054,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7751908898353577,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7738757431507111,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7761802673339844,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929513475101896,0.5771812080536913 +0.7720681726932526,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7686194181442261,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7694710791110992,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7701238989830017,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 diff --git a/results/downsample/papila/100/retfound/pr.png b/results/downsample/papila/100/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..567bef62299e8292c76bc9859243b2c14a03e919 --- /dev/null +++ b/results/downsample/papila/100/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da1e69338a15119bd42de1548ecd850a11f9fa7010485387bbacc23cc853bd79 +size 52261 diff --git a/results/downsample/papila/100/retfound/roc.png b/results/downsample/papila/100/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..6c1790997e7d1da62528b52c94485f129cb23717 --- /dev/null +++ b/results/downsample/papila/100/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:703e193e0dfa2492530843c855420951e3e1b0af10a25f4df2ae2fa4e96aeaf5 +size 57625 diff --git a/results/downsample/papila/100/retfound/test_pred.npz b/results/downsample/papila/100/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..3d702b56e1fa0fa773506bb7a507fe8432118b5c --- /dev/null +++ b/results/downsample/papila/100/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:288713ad75144f9b05b057382287103c78c0443830c6b60555e9c81300d12da3 +size 1518 diff --git a/results/downsample/papila/100/retfound/train.log b/results/downsample/papila/100/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..15c178bea5a399d4868c7a6fae023020cae404f6 --- /dev/null +++ b/results/downsample/papila/100/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W701 14:52:56.682172985 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[14:52:57.303553] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[14:52:57.303762] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/tmp/downsample_exp/papila_100', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/100', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[14:53:00.048289] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:53:01.578248] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[14:53:02.982910] Sampler_train = +[14:53:03.030595] len of train_set: 288 +[14:53:03.238856] [Adaptation] Full fine-tuning: training all parameters. +[14:53:03.239864] number of trainable params (M): 303.30 +[14:53:03.239960] base lr: 5.00e-03 +[14:53:03.240030] actual lr: 6.25e-04 +[14:53:03.240095] accumulate grad iterations: 1 +[14:53:03.240154] effective batch size: 32 +[14:53:03.242866] criterion = CrossEntropyLoss() +[14:53:03.242990] Start training for 50 epochs +[14:53:03.244762] log_dir: ./output_logs/retfound +[14:53:06.175540] Epoch: [0] [0/9] eta: 0:00:26 lr: 0.000000 loss: 0.6927 (0.6927) time: 2.9300 data: 2.4401 max mem: 7340 +[14:53:06.728081] Epoch: [0] [8/9] eta: 0:00:00 lr: 0.000056 loss: 0.6903 (0.6828) time: 0.3869 data: 0.2712 max mem: 9671 +[14:53:06.787826] Epoch: [0] Total time: 0:00:03 (0.3937 s / it) +[14:53:06.788705] Averaged stats: lr: 0.000056 loss: 0.6903 (0.6828) +[14:53:09.769357] val: [0/2] eta: 0:00:05 loss: 0.6260 (0.6260) time: 2.9740 data: 2.9511 max mem: 9671 +[14:53:10.038203] val: [1/2] eta: 0:00:01 loss: 0.6260 (0.6946) time: 1.6211 data: 1.4756 max mem: 9671 +[14:53:10.113473] val: Total time: 0:00:03 (1.6593 s / it) +[14:53:10.124059] val loss: 0.6945861876010895 +[14:53:10.124251] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6867, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6626, Kappa: 0.0000, Score: 0.3731 +[14:53:11.899629] Best epoch = 0, Best score = 0.3731 +[14:53:12.001018] log_dir: ./output_logs/retfound +[14:53:14.346364] Epoch: [1] [0/9] eta: 0:00:21 lr: 0.000063 loss: 0.6313 (0.6313) time: 2.3444 data: 2.2416 max mem: 9671 +[14:53:14.871459] Epoch: [1] [8/9] eta: 0:00:00 lr: 0.000118 loss: 0.5999 (0.5992) time: 0.3188 data: 0.2491 max mem: 9671 +[14:53:14.948577] Epoch: [1] Total time: 0:00:02 (0.3275 s / it) +[14:53:14.949332] Averaged stats: lr: 0.000118 loss: 0.5999 (0.5992) +[14:53:17.658061] val: [0/2] eta: 0:00:05 loss: 0.3876 (0.3876) time: 2.6945 data: 2.6777 max mem: 9671 +[14:53:17.667600] val: [1/2] eta: 0:00:01 loss: 0.3876 (0.7485) time: 1.3518 data: 1.3389 max mem: 9671 +[14:53:17.742841] val: Total time: 0:00:02 (1.3900 s / it) +[14:53:17.751871] val loss: 0.7484893798828125 +[14:53:17.752072] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7727, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7154, Kappa: 0.0000, Score: 0.4017 +[14:53:19.464112] Best epoch = 1, Best score = 0.4017 +[14:53:19.528970] log_dir: ./output_logs/retfound +[14:53:21.840547] Epoch: [2] [0/9] eta: 0:00:20 lr: 0.000125 loss: 0.5245 (0.5245) time: 2.3106 data: 2.2381 max mem: 9671 +[14:53:22.361742] Epoch: [2] [8/9] eta: 0:00:00 lr: 0.000181 loss: 0.5136 (0.5173) time: 0.3146 data: 0.2487 max mem: 9671 +[14:53:22.434232] Epoch: [2] Total time: 0:00:02 (0.3228 s / it) +[14:53:22.435037] Averaged stats: lr: 0.000181 loss: 0.5136 (0.5173) +[14:53:25.117125] val: [0/2] eta: 0:00:05 loss: 0.1880 (0.1880) time: 2.6671 data: 2.6486 max mem: 9671 +[14:53:25.126364] val: [1/2] eta: 0:00:01 loss: 0.1880 (0.9287) time: 1.3379 data: 1.3243 max mem: 9671 +[14:53:25.210147] val: Total time: 0:00:02 (1.3805 s / it) +[14:53:25.219006] val loss: 0.9287281036376953 +[14:53:25.219171] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8086, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7688, Kappa: 0.0000, Score: 0.4137 +[14:53:26.931753] Best epoch = 2, Best score = 0.4137 +[14:53:26.997384] log_dir: ./output_logs/retfound +[14:53:29.349431] Epoch: [3] [0/9] eta: 0:00:21 lr: 0.000188 loss: 0.4505 (0.4505) time: 2.3509 data: 2.2654 max mem: 9671 +[14:53:29.880013] Epoch: [3] [8/9] eta: 0:00:00 lr: 0.000243 loss: 0.5323 (0.5203) time: 0.3201 data: 0.2518 max mem: 9671 +[14:53:29.951250] Epoch: [3] Total time: 0:00:02 (0.3282 s / it) +[14:53:29.952071] Averaged stats: lr: 0.000243 loss: 0.5323 (0.5203) +[14:53:32.801070] val: [0/2] eta: 0:00:05 loss: 0.2116 (0.2116) time: 2.8337 data: 2.8167 max mem: 9671 +[14:53:32.810467] val: [1/2] eta: 0:00:01 loss: 0.2116 (0.8795) time: 1.4213 data: 1.4084 max mem: 9671 +[14:53:32.880546] val: Total time: 0:00:02 (1.4570 s / it) +[14:53:32.889349] val loss: 0.8795062899589539 +[14:53:32.889552] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7937, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7504, Kappa: 0.0000, Score: 0.4087 +[14:53:32.926596] Best epoch = 2, Best score = 0.4137 +[14:53:33.177665] log_dir: ./output_logs/retfound +[14:53:35.443048] Epoch: [4] [0/9] eta: 0:00:20 lr: 0.000250 loss: 0.5317 (0.5317) time: 2.2644 data: 2.1956 max mem: 9671 +[14:53:36.029428] Epoch: [4] [8/9] eta: 0:00:00 lr: 0.000306 loss: 0.5317 (0.4997) time: 0.3167 data: 0.2514 max mem: 9671 +[14:53:36.099048] Epoch: [4] Total time: 0:00:02 (0.3246 s / it) +[14:53:36.099846] Averaged stats: lr: 0.000306 loss: 0.5317 (0.4997) +[14:53:38.870687] val: [0/2] eta: 0:00:05 loss: 0.2462 (0.2462) time: 2.7556 data: 2.7386 max mem: 9671 +[14:53:38.879866] val: [1/2] eta: 0:00:01 loss: 0.2462 (0.8250) time: 1.3821 data: 1.3694 max mem: 9671 +[14:53:38.948489] val: Total time: 0:00:02 (1.4171 s / it) +[14:53:38.957245] val loss: 0.8250083923339844 +[14:53:38.957428] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7555, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6977, Kappa: 0.0000, Score: 0.3960 +[14:53:38.995237] Best epoch = 2, Best score = 0.4137 +[14:53:39.267265] log_dir: ./output_logs/retfound +[14:53:41.448977] Epoch: [5] [0/9] eta: 0:00:19 lr: 0.000313 loss: 0.4567 (0.4567) time: 2.1807 data: 2.1110 max mem: 9671 +[14:53:42.067450] Epoch: [5] [8/9] eta: 0:00:00 lr: 0.000368 loss: 0.4888 (0.5041) time: 0.3109 data: 0.2457 max mem: 9671 +[14:53:42.139201] Epoch: [5] Total time: 0:00:02 (0.3191 s / it) +[14:53:42.140088] Averaged stats: lr: 0.000368 loss: 0.4888 (0.5041) +[14:53:45.024323] val: [0/2] eta: 0:00:05 loss: 0.2470 (0.2470) time: 2.8686 data: 2.8512 max mem: 9671 +[14:53:45.034608] val: [1/2] eta: 0:00:01 loss: 0.2470 (0.8058) time: 1.4392 data: 1.4256 max mem: 9671 +[14:53:45.114442] val: Total time: 0:00:02 (1.4797 s / it) +[14:53:45.125539] val loss: 0.8058345913887024 +[14:53:45.125780] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7094, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6559, Kappa: 0.0000, Score: 0.3806 +[14:53:45.171047] Best epoch = 2, Best score = 0.4137 +[14:53:45.429515] log_dir: ./output_logs/retfound +[14:53:47.941210] Epoch: [6] [0/9] eta: 0:00:22 lr: 0.000375 loss: 0.5451 (0.5451) time: 2.5109 data: 2.4410 max mem: 9671 +[14:53:48.469873] Epoch: [6] [8/9] eta: 0:00:00 lr: 0.000431 loss: 0.5040 (0.4901) time: 0.3376 data: 0.2713 max mem: 9671 +[14:53:48.544286] Epoch: [6] Total time: 0:00:03 (0.3461 s / it) +[14:53:48.545102] Averaged stats: lr: 0.000431 loss: 0.5040 (0.4901) +[14:53:51.338715] val: [0/2] eta: 0:00:05 loss: 0.2326 (0.2326) time: 2.7784 data: 2.7616 max mem: 9671 +[14:53:51.348718] val: [1/2] eta: 0:00:01 loss: 0.2326 (0.7666) time: 1.3939 data: 1.3808 max mem: 9671 +[14:53:51.419763] val: Total time: 0:00:02 (1.4301 s / it) +[14:53:51.429149] val loss: 0.7665609121322632 +[14:53:51.429327] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7344, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6564, Kappa: 0.0000, Score: 0.3889 +[14:53:51.462621] Best epoch = 2, Best score = 0.4137 +[14:53:51.717479] log_dir: ./output_logs/retfound +[14:53:53.920813] Epoch: [7] [0/9] eta: 0:00:19 lr: 0.000438 loss: 0.4717 (0.4717) time: 2.2024 data: 2.1256 max mem: 9671 +[14:53:54.437651] Epoch: [7] [8/9] eta: 0:00:00 lr: 0.000493 loss: 0.4926 (0.5020) time: 0.3021 data: 0.2362 max mem: 9671 +[14:53:54.580566] Epoch: [7] Total time: 0:00:02 (0.3181 s / it) +[14:53:54.581303] Averaged stats: lr: 0.000493 loss: 0.4926 (0.5020) +[14:53:57.278527] val: [0/2] eta: 0:00:05 loss: 0.1709 (0.1709) time: 2.6863 data: 2.6690 max mem: 9671 +[14:53:57.289435] val: [1/2] eta: 0:00:01 loss: 0.1709 (0.8525) time: 1.3483 data: 1.3346 max mem: 9671 +[14:53:57.360585] val: Total time: 0:00:02 (1.3845 s / it) +[14:53:57.370394] val loss: 0.8525406122207642 +[14:53:57.370543] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6718, Kappa: 0.0000, Score: 0.3941 +[14:53:57.405590] Best epoch = 2, Best score = 0.4137 +[14:53:57.675730] log_dir: ./output_logs/retfound +[14:53:59.953899] Epoch: [8] [0/9] eta: 0:00:20 lr: 0.000500 loss: 0.4438 (0.4438) time: 2.2772 data: 2.2087 max mem: 9671 +[14:54:00.556980] Epoch: [8] [8/9] eta: 0:00:00 lr: 0.000556 loss: 0.4622 (0.4714) time: 0.3200 data: 0.2543 max mem: 9671 +[14:54:00.627186] Epoch: [8] Total time: 0:00:02 (0.3279 s / it) +[14:54:00.627923] Averaged stats: lr: 0.000556 loss: 0.4622 (0.4714) +[14:54:03.349499] val: [0/2] eta: 0:00:05 loss: 0.4000 (0.4000) time: 2.7063 data: 2.6895 max mem: 9671 +[14:54:03.359137] val: [1/2] eta: 0:00:01 loss: 0.4000 (0.6246) time: 1.3577 data: 1.3448 max mem: 9671 +[14:54:03.427681] val: Total time: 0:00:02 (1.3926 s / it) +[14:54:03.436506] val loss: 0.6245647370815277 +[14:54:03.436681] Accuracy: 0.7143, F1 Score: 0.5368, ROC AUC: 0.7500, Hamming Loss: 0.2857, + Jaccard Score: 0.4214, Precision: 0.5556, Recall: 0.5375, + Average Precision: 0.6940, Kappa: 0.0870, Score: 0.4579 +[14:54:05.180118] Best epoch = 8, Best score = 0.4579 +[14:54:05.248400] log_dir: ./output_logs/retfound +[14:54:07.467256] Epoch: [9] [0/9] eta: 0:00:19 lr: 0.000562 loss: 0.5492 (0.5492) time: 2.2180 data: 2.1377 max mem: 9671 +[14:54:07.986982] Epoch: [9] [8/9] eta: 0:00:00 lr: 0.000618 loss: 0.4048 (0.4486) time: 0.3041 data: 0.2376 max mem: 9671 +[14:54:08.055233] Epoch: [9] Total time: 0:00:02 (0.3118 s / it) +[14:54:08.056123] Averaged stats: lr: 0.000618 loss: 0.4048 (0.4486) +[14:54:10.804630] val: [0/2] eta: 0:00:05 loss: 0.1913 (0.1913) time: 2.7257 data: 2.7088 max mem: 9671 +[14:54:10.814009] val: [1/2] eta: 0:00:01 loss: 0.1913 (0.7787) time: 1.3673 data: 1.3545 max mem: 9671 +[14:54:10.884982] val: Total time: 0:00:02 (1.4034 s / it) +[14:54:10.893971] val loss: 0.7787399291992188 +[14:54:10.894157] Accuracy: 0.7619, F1 Score: 0.6140, ROC AUC: 0.7937, Hamming Loss: 0.2381, + Jaccard Score: 0.4872, Precision: 0.6528, Recall: 0.6031, + Average Precision: 0.7304, Kappa: 0.2391, Score: 0.5490 +[14:54:12.594971] Best epoch = 9, Best score = 0.5490 +[14:54:12.679785] log_dir: ./output_logs/retfound +[14:54:14.962054] Epoch: [10] [0/9] eta: 0:00:20 lr: 0.000625 loss: 0.4722 (0.4722) time: 2.2814 data: 2.2116 max mem: 9671 +[14:54:15.479872] Epoch: [10] [8/9] eta: 0:00:00 lr: 0.000624 loss: 0.4722 (0.4588) time: 0.3110 data: 0.2458 max mem: 9671 +[14:54:15.552198] Epoch: [10] Total time: 0:00:02 (0.3191 s / it) +[14:54:15.553013] Averaged stats: lr: 0.000624 loss: 0.4722 (0.4588) +[14:54:18.166371] val: [0/2] eta: 0:00:05 loss: 0.4142 (0.4142) time: 2.5982 data: 2.5812 max mem: 9671 +[14:54:18.175996] val: [1/2] eta: 0:00:01 loss: 0.4142 (0.5484) time: 1.3036 data: 1.2907 max mem: 9671 +[14:54:18.246882] val: Total time: 0:00:02 (1.3397 s / it) +[14:54:18.255950] val loss: 0.5484199523925781 +[14:54:18.256130] Accuracy: 0.7619, F1 Score: 0.6719, ROC AUC: 0.8156, Hamming Loss: 0.2381, + Jaccard Score: 0.5315, Precision: 0.6719, Recall: 0.6719, + Average Precision: 0.7519, Kappa: 0.3438, Score: 0.6104 +[14:54:20.039207] Best epoch = 10, Best score = 0.6104 +[14:54:20.126080] log_dir: ./output_logs/retfound +[14:54:22.506855] Epoch: [11] [0/9] eta: 0:00:21 lr: 0.000624 loss: 0.3691 (0.3691) time: 2.3798 data: 2.3102 max mem: 9671 +[14:54:23.027359] Epoch: [11] [8/9] eta: 0:00:00 lr: 0.000622 loss: 0.4582 (0.4270) time: 0.3222 data: 0.2569 max mem: 9671 +[14:54:23.099785] Epoch: [11] Total time: 0:00:02 (0.3304 s / it) +[14:54:23.100639] Averaged stats: lr: 0.000622 loss: 0.4582 (0.4270) +[14:54:26.005444] val: [0/2] eta: 0:00:05 loss: 0.0844 (0.0844) time: 2.8901 data: 2.8728 max mem: 9671 +[14:54:26.016542] val: [1/2] eta: 0:00:01 loss: 0.0844 (1.2772) time: 1.4503 data: 1.4365 max mem: 9671 +[14:54:26.109905] val: Total time: 0:00:02 (1.4977 s / it) +[14:54:26.124177] val loss: 1.2771737575531006 +[14:54:26.124499] Accuracy: 0.7857, F1 Score: 0.5905, ROC AUC: 0.8500, Hamming Loss: 0.2143, + Jaccard Score: 0.4784, Precision: 0.7308, Recall: 0.5844, + Average Precision: 0.8031, Kappa: 0.2222, Score: 0.5542 +[14:54:26.159828] Best epoch = 10, Best score = 0.6104 +[14:54:26.411089] log_dir: ./output_logs/retfound +[14:54:28.537435] Epoch: [12] [0/9] eta: 0:00:19 lr: 0.000621 loss: 0.5461 (0.5461) time: 2.1253 data: 2.0517 max mem: 9671 +[14:54:29.105127] Epoch: [12] [8/9] eta: 0:00:00 lr: 0.000617 loss: 0.4572 (0.4496) time: 0.2991 data: 0.2334 max mem: 9671 +[14:54:29.176035] Epoch: [12] Total time: 0:00:02 (0.3072 s / it) +[14:54:29.176854] Averaged stats: lr: 0.000617 loss: 0.4572 (0.4496) +[14:54:32.002024] val: [0/2] eta: 0:00:05 loss: 0.1181 (0.1181) time: 2.8143 data: 2.7975 max mem: 9671 +[14:54:32.013088] val: [1/2] eta: 0:00:01 loss: 0.1181 (0.9774) time: 1.4124 data: 1.3988 max mem: 9671 +[14:54:32.082659] val: Total time: 0:00:02 (1.4478 s / it) +[14:54:32.091582] val loss: 0.9774452447891235 +[14:54:32.091766] Accuracy: 0.8333, F1 Score: 0.7159, ROC AUC: 0.8445, Hamming Loss: 0.1667, + Jaccard Score: 0.5897, Precision: 0.8189, Recall: 0.6844, + Average Precision: 0.8015, Kappa: 0.4453, Score: 0.6686 +[14:54:33.798539] Best epoch = 12, Best score = 0.6686 +[14:54:33.874552] log_dir: ./output_logs/retfound +[14:54:36.253247] Epoch: [13] [0/9] eta: 0:00:21 lr: 0.000616 loss: 0.6730 (0.6730) time: 2.3779 data: 2.3082 max mem: 9671 +[14:54:36.772657] Epoch: [13] [8/9] eta: 0:00:00 lr: 0.000611 loss: 0.4303 (0.4471) time: 0.3218 data: 0.2565 max mem: 9671 +[14:54:36.843667] Epoch: [13] Total time: 0:00:02 (0.3299 s / it) +[14:54:36.844598] Averaged stats: lr: 0.000611 loss: 0.4303 (0.4471) +[14:54:39.745214] val: [0/2] eta: 0:00:05 loss: 0.1100 (0.1100) time: 2.8852 data: 2.8683 max mem: 9671 +[14:54:39.754688] val: [1/2] eta: 0:00:01 loss: 0.1100 (0.8752) time: 1.4471 data: 1.4342 max mem: 9671 +[14:54:39.826990] val: Total time: 0:00:02 (1.4838 s / it) +[14:54:39.836558] val loss: 0.8752020001411438 +[14:54:39.836727] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8680, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8321, Kappa: 0.3913, Score: 0.6501 +[14:54:39.876689] Best epoch = 12, Best score = 0.6686 +[14:54:40.127215] log_dir: ./output_logs/retfound +[14:54:42.437543] Epoch: [14] [0/9] eta: 0:00:20 lr: 0.000610 loss: 0.3318 (0.3318) time: 2.3095 data: 2.2380 max mem: 9671 +[14:54:42.956194] Epoch: [14] [8/9] eta: 0:00:00 lr: 0.000602 loss: 0.4897 (0.4556) time: 0.3142 data: 0.2487 max mem: 9671 +[14:54:43.025775] Epoch: [14] Total time: 0:00:02 (0.3220 s / it) +[14:54:43.026561] Averaged stats: lr: 0.000602 loss: 0.4897 (0.4556) +[14:54:45.704351] val: [0/2] eta: 0:00:05 loss: 0.3347 (0.3347) time: 2.6669 data: 2.6493 max mem: 9671 +[14:54:45.714559] val: [1/2] eta: 0:00:01 loss: 0.3347 (0.7528) time: 1.3383 data: 1.3247 max mem: 9671 +[14:54:45.794739] val: Total time: 0:00:02 (1.3790 s / it) +[14:54:45.803750] val loss: 0.7527960538864136 +[14:54:45.803964] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8531, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8224, Kappa: 0.4750, Score: 0.6885 +[14:54:47.553233] Best epoch = 14, Best score = 0.6885 +[14:54:47.626683] log_dir: ./output_logs/retfound +[14:54:50.002829] Epoch: [15] [0/9] eta: 0:00:21 lr: 0.000601 loss: 0.4373 (0.4373) time: 2.3753 data: 2.3044 max mem: 9671 +[14:54:50.521700] Epoch: [15] [8/9] eta: 0:00:00 lr: 0.000592 loss: 0.3949 (0.3853) time: 0.3215 data: 0.2561 max mem: 9671 +[14:54:50.591789] Epoch: [15] Total time: 0:00:02 (0.3294 s / it) +[14:54:50.592613] Averaged stats: lr: 0.000592 loss: 0.3949 (0.3853) +[14:54:53.276239] val: [0/2] eta: 0:00:05 loss: 0.2078 (0.2078) time: 2.6689 data: 2.6520 max mem: 9671 +[14:54:53.285714] val: [1/2] eta: 0:00:01 loss: 0.2078 (0.9505) time: 1.3389 data: 1.3260 max mem: 9671 +[14:54:53.356148] val: Total time: 0:00:02 (1.3748 s / it) +[14:54:53.366295] val loss: 0.9505177736282349 +[14:54:53.366511] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8594, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8067, Kappa: 0.4878, Score: 0.6964 +[14:54:55.142930] Best epoch = 15, Best score = 0.6964 +[14:54:55.232196] log_dir: ./output_logs/retfound +[14:54:57.371429] Epoch: [16] [0/9] eta: 0:00:19 lr: 0.000591 loss: 0.6279 (0.6279) time: 2.1384 data: 2.0648 max mem: 9671 +[14:54:57.891392] Epoch: [16] [8/9] eta: 0:00:00 lr: 0.000580 loss: 0.4289 (0.4627) time: 0.2953 data: 0.2295 max mem: 9671 +[14:54:57.960558] Epoch: [16] Total time: 0:00:02 (0.3031 s / it) +[14:54:57.961428] Averaged stats: lr: 0.000580 loss: 0.4289 (0.4627) +[14:55:00.688913] val: [0/2] eta: 0:00:05 loss: 0.1866 (0.1866) time: 2.7089 data: 2.6920 max mem: 9671 +[14:55:00.698206] val: [1/2] eta: 0:00:01 loss: 0.1866 (1.0000) time: 1.3588 data: 1.3461 max mem: 9671 +[14:55:00.769106] val: Total time: 0:00:02 (1.3949 s / it) +[14:55:00.778799] val loss: 1.0000401735305786 +[14:55:00.778986] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8438, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7957, Kappa: 0.3913, Score: 0.6421 +[14:55:00.825354] Best epoch = 15, Best score = 0.6964 +[14:55:01.082035] log_dir: ./output_logs/retfound +[14:55:03.417055] Epoch: [17] [0/9] eta: 0:00:21 lr: 0.000579 loss: 0.4267 (0.4267) time: 2.3337 data: 2.2532 max mem: 9671 +[14:55:03.957695] Epoch: [17] [8/9] eta: 0:00:00 lr: 0.000567 loss: 0.4068 (0.4493) time: 0.3193 data: 0.2504 max mem: 9671 +[14:55:04.026415] Epoch: [17] Total time: 0:00:02 (0.3271 s / it) +[14:55:04.027187] Averaged stats: lr: 0.000567 loss: 0.4068 (0.4493) +[14:55:06.892133] val: [0/2] eta: 0:00:05 loss: 0.2248 (0.2248) time: 2.8419 data: 2.8252 max mem: 9671 +[14:55:06.901417] val: [1/2] eta: 0:00:01 loss: 0.2248 (0.8678) time: 1.4254 data: 1.4126 max mem: 9671 +[14:55:06.969273] val: Total time: 0:00:02 (1.4599 s / it) +[14:55:06.978074] val loss: 0.8677769899368286 +[14:55:06.978250] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8156, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.7545, Kappa: 0.4362, Score: 0.6563 +[14:55:07.022726] Best epoch = 15, Best score = 0.6964 +[14:55:07.254766] log_dir: ./output_logs/retfound +[14:55:09.647672] Epoch: [18] [0/9] eta: 0:00:21 lr: 0.000565 loss: 0.4027 (0.4027) time: 2.3919 data: 2.3238 max mem: 9671 +[14:55:10.168094] Epoch: [18] [8/9] eta: 0:00:00 lr: 0.000552 loss: 0.4138 (0.4269) time: 0.3235 data: 0.2583 max mem: 9671 +[14:55:10.246638] Epoch: [18] Total time: 0:00:02 (0.3324 s / it) +[14:55:10.247409] Averaged stats: lr: 0.000552 loss: 0.4138 (0.4269) +[14:55:12.866011] val: [0/2] eta: 0:00:05 loss: 0.2075 (0.2075) time: 2.6076 data: 2.5903 max mem: 9671 +[14:55:12.875840] val: [1/2] eta: 0:00:01 loss: 0.2075 (1.1336) time: 1.3084 data: 1.2952 max mem: 9671 +[14:55:12.951232] val: Total time: 0:00:02 (1.3468 s / it) +[14:55:12.960119] val loss: 1.133644163608551 +[14:55:12.960297] Accuracy: 0.7857, F1 Score: 0.6681, ROC AUC: 0.8070, Hamming Loss: 0.2143, + Jaccard Score: 0.5354, Precision: 0.7000, Recall: 0.6531, + Average Precision: 0.7544, Kappa: 0.3415, Score: 0.6055 +[14:55:12.994671] Best epoch = 15, Best score = 0.6964 +[14:55:13.249835] log_dir: ./output_logs/retfound +[14:55:15.401158] Epoch: [19] [0/9] eta: 0:00:19 lr: 0.000550 loss: 0.5522 (0.5522) time: 2.1504 data: 2.0809 max mem: 9671 +[14:55:15.921881] Epoch: [19] [8/9] eta: 0:00:00 lr: 0.000536 loss: 0.4279 (0.3996) time: 0.2967 data: 0.2313 max mem: 9671 +[14:55:15.995295] Epoch: [19] Total time: 0:00:02 (0.3050 s / it) +[14:55:15.996024] Averaged stats: lr: 0.000536 loss: 0.4279 (0.3996) +[14:55:18.631468] val: [0/2] eta: 0:00:05 loss: 0.4255 (0.4255) time: 2.6236 data: 2.6064 max mem: 9671 +[14:55:18.640669] val: [1/2] eta: 0:00:01 loss: 0.4255 (0.6373) time: 1.3161 data: 1.3033 max mem: 9671 +[14:55:18.710446] val: Total time: 0:00:02 (1.3516 s / it) +[14:55:18.719396] val loss: 0.6373273730278015 +[14:55:18.719594] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8250, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.7853, Kappa: 0.4290, Score: 0.6561 +[14:55:18.753277] Best epoch = 15, Best score = 0.6964 +[14:55:19.020877] log_dir: ./output_logs/retfound +[14:55:21.351278] Epoch: [20] [0/9] eta: 0:00:20 lr: 0.000534 loss: 0.3223 (0.3223) time: 2.3295 data: 2.2591 max mem: 9671 +[14:55:21.871982] Epoch: [20] [8/9] eta: 0:00:00 lr: 0.000518 loss: 0.3818 (0.3802) time: 0.3166 data: 0.2511 max mem: 9671 +[14:55:21.944085] Epoch: [20] Total time: 0:00:02 (0.3248 s / it) +[14:55:21.944910] Averaged stats: lr: 0.000518 loss: 0.3818 (0.3802) +[14:55:24.569968] val: [0/2] eta: 0:00:05 loss: 0.3854 (0.3854) time: 2.6141 data: 2.5962 max mem: 9671 +[14:55:24.580619] val: [1/2] eta: 0:00:01 loss: 0.3854 (0.6148) time: 1.3121 data: 1.2982 max mem: 9671 +[14:55:24.649921] val: Total time: 0:00:02 (1.3475 s / it) +[14:55:24.658764] val loss: 0.6148214936256409 +[14:55:24.658953] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8531, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8131, Kappa: 0.4290, Score: 0.6655 +[14:55:24.692239] Best epoch = 15, Best score = 0.6964 +[14:55:24.935867] log_dir: ./output_logs/retfound +[14:55:27.105613] Epoch: [21] [0/9] eta: 0:00:19 lr: 0.000516 loss: 0.4815 (0.4815) time: 2.1689 data: 2.0936 max mem: 9671 +[14:55:27.631455] Epoch: [21] [8/9] eta: 0:00:00 lr: 0.000499 loss: 0.3667 (0.3990) time: 0.2993 data: 0.2333 max mem: 9671 +[14:55:27.703318] Epoch: [21] Total time: 0:00:02 (0.3075 s / it) +[14:55:27.704075] Averaged stats: lr: 0.000499 loss: 0.3667 (0.3990) +[14:55:30.252923] val: [0/2] eta: 0:00:05 loss: 0.3414 (0.3414) time: 2.5261 data: 2.5089 max mem: 9671 +[14:55:30.262481] val: [1/2] eta: 0:00:01 loss: 0.3414 (0.6647) time: 1.2676 data: 1.2545 max mem: 9671 +[14:55:30.334072] val: Total time: 0:00:02 (1.3040 s / it) +[14:55:30.342816] val loss: 0.6647389531135559 +[14:55:30.343012] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8594, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8301, Kappa: 0.5772, Score: 0.7415 +[14:55:32.041143] Best epoch = 21, Best score = 0.7415 +[14:55:32.124179] log_dir: ./output_logs/retfound +[14:55:34.416523] Epoch: [22] [0/9] eta: 0:00:20 lr: 0.000496 loss: 0.3751 (0.3751) time: 2.2915 data: 2.2220 max mem: 9671 +[14:55:34.937042] Epoch: [22] [8/9] eta: 0:00:00 lr: 0.000478 loss: 0.4090 (0.3949) time: 0.3124 data: 0.2469 max mem: 9671 +[14:55:35.005817] Epoch: [22] Total time: 0:00:02 (0.3202 s / it) +[14:55:35.006583] Averaged stats: lr: 0.000478 loss: 0.4090 (0.3949) +[14:55:37.670474] val: [0/2] eta: 0:00:05 loss: 0.2182 (0.2182) time: 2.6414 data: 2.6245 max mem: 9671 +[14:55:37.679841] val: [1/2] eta: 0:00:01 loss: 0.2182 (0.7312) time: 1.3251 data: 1.3123 max mem: 9671 +[14:55:37.752944] val: Total time: 0:00:02 (1.3623 s / it) +[14:55:37.761669] val loss: 0.7311892509460449 +[14:55:37.761846] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8406, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8171, Kappa: 0.4878, Score: 0.6901 +[14:55:37.808870] Best epoch = 21, Best score = 0.7415 +[14:55:38.058605] log_dir: ./output_logs/retfound +[14:55:40.483782] Epoch: [23] [0/9] eta: 0:00:21 lr: 0.000476 loss: 0.4109 (0.4109) time: 2.4243 data: 2.3556 max mem: 9671 +[14:55:41.005175] Epoch: [23] [8/9] eta: 0:00:00 lr: 0.000457 loss: 0.3560 (0.3675) time: 0.3272 data: 0.2618 max mem: 9671 +[14:55:41.075838] Epoch: [23] Total time: 0:00:03 (0.3352 s / it) +[14:55:41.076616] Averaged stats: lr: 0.000457 loss: 0.3560 (0.3675) +[14:55:43.791290] val: [0/2] eta: 0:00:05 loss: 0.4150 (0.4150) time: 2.6914 data: 2.6742 max mem: 9671 +[14:55:43.800439] val: [1/2] eta: 0:00:01 loss: 0.4150 (0.8558) time: 1.3500 data: 1.3371 max mem: 9671 +[14:55:43.870687] val: Total time: 0:00:02 (1.3857 s / it) +[14:55:43.879418] val loss: 0.8558176159858704 +[14:55:43.879586] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8281, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8103, Kappa: 0.4750, Score: 0.6802 +[14:55:43.925118] Best epoch = 21, Best score = 0.7415 +[14:55:44.153454] log_dir: ./output_logs/retfound +[14:55:46.374703] Epoch: [24] [0/9] eta: 0:00:19 lr: 0.000455 loss: 0.3960 (0.3960) time: 2.2203 data: 2.1517 max mem: 9671 +[14:55:46.900967] Epoch: [24] [8/9] eta: 0:00:00 lr: 0.000435 loss: 0.3916 (0.3996) time: 0.3051 data: 0.2399 max mem: 9671 +[14:55:46.968588] Epoch: [24] Total time: 0:00:02 (0.3128 s / it) +[14:55:46.969339] Averaged stats: lr: 0.000435 loss: 0.3916 (0.3996) +[14:55:49.676416] val: [0/2] eta: 0:00:05 loss: 0.3662 (0.3662) time: 2.6962 data: 2.6794 max mem: 9671 +[14:55:49.685890] val: [1/2] eta: 0:00:01 loss: 0.3662 (0.7549) time: 1.3525 data: 1.3398 max mem: 9671 +[14:55:49.755177] val: Total time: 0:00:02 (1.3878 s / it) +[14:55:49.763990] val loss: 0.7548640370368958 +[14:55:49.764167] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8219, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8071, Kappa: 0.4750, Score: 0.6781 +[14:55:49.811130] Best epoch = 21, Best score = 0.7415 +[14:55:50.058767] log_dir: ./output_logs/retfound +[14:55:52.211124] Epoch: [25] [0/9] eta: 0:00:19 lr: 0.000432 loss: 0.4206 (0.4206) time: 2.1515 data: 2.0820 max mem: 9671 +[14:55:52.773328] Epoch: [25] [8/9] eta: 0:00:00 lr: 0.000412 loss: 0.3320 (0.3284) time: 0.3015 data: 0.2356 max mem: 9671 +[14:55:52.845280] Epoch: [25] Total time: 0:00:02 (0.3096 s / it) +[14:55:52.846025] Averaged stats: lr: 0.000412 loss: 0.3320 (0.3284) +[14:55:55.774555] val: [0/2] eta: 0:00:05 loss: 0.2804 (0.2804) time: 2.9064 data: 2.8895 max mem: 9671 +[14:55:55.783991] val: [1/2] eta: 0:00:01 loss: 0.2804 (0.7578) time: 1.4577 data: 1.4448 max mem: 9671 +[14:55:55.852288] val: Total time: 0:00:02 (1.4925 s / it) +[14:55:55.861670] val loss: 0.7577589750289917 +[14:55:55.861908] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8156, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7928, Kappa: 0.5772, Score: 0.7269 +[14:55:55.895596] Best epoch = 21, Best score = 0.7415 +[14:55:56.155681] log_dir: ./output_logs/retfound +[14:55:58.218956] Epoch: [26] [0/9] eta: 0:00:18 lr: 0.000409 loss: 0.2213 (0.2213) time: 2.0623 data: 1.9934 max mem: 9671 +[14:55:58.760718] Epoch: [26] [8/9] eta: 0:00:00 lr: 0.000388 loss: 0.3055 (0.3474) time: 0.2893 data: 0.2241 max mem: 9671 +[14:55:58.831316] Epoch: [26] Total time: 0:00:02 (0.2973 s / it) +[14:55:58.832087] Averaged stats: lr: 0.000388 loss: 0.3055 (0.3474) +[14:56:01.523523] val: [0/2] eta: 0:00:05 loss: 0.5462 (0.5462) time: 2.6807 data: 2.6638 max mem: 9671 +[14:56:01.533032] val: [1/2] eta: 0:00:01 loss: 0.5462 (0.6590) time: 1.3448 data: 1.3320 max mem: 9671 +[14:56:01.601172] val: Total time: 0:00:02 (1.3795 s / it) +[14:56:01.609974] val loss: 0.659015029668808 +[14:56:01.610160] Accuracy: 0.7381, F1 Score: 0.6707, ROC AUC: 0.8289, Hamming Loss: 0.2619, + Jaccard Score: 0.5237, Precision: 0.6618, Recall: 0.6906, + Average Precision: 0.7991, Kappa: 0.3456, Score: 0.6151 +[14:56:01.645217] Best epoch = 21, Best score = 0.7415 +[14:56:01.929843] log_dir: ./output_logs/retfound +[14:56:04.189675] Epoch: [27] [0/9] eta: 0:00:20 lr: 0.000386 loss: 0.2824 (0.2824) time: 2.2590 data: 2.1910 max mem: 9671 +[14:56:04.708381] Epoch: [27] [8/9] eta: 0:00:00 lr: 0.000364 loss: 0.3541 (0.3910) time: 0.3086 data: 0.2435 max mem: 9671 +[14:56:04.786777] Epoch: [27] Total time: 0:00:02 (0.3174 s / it) +[14:56:04.787539] Averaged stats: lr: 0.000364 loss: 0.3541 (0.3910) +[14:56:07.523425] val: [0/2] eta: 0:00:05 loss: 0.1278 (0.1278) time: 2.7132 data: 2.6961 max mem: 9671 +[14:56:07.532912] val: [1/2] eta: 0:00:01 loss: 0.1278 (1.1875) time: 1.3611 data: 1.3481 max mem: 9671 +[14:56:07.603546] val: Total time: 0:00:02 (1.3970 s / it) +[14:56:07.612400] val loss: 1.1874847412109375 +[14:56:07.612662] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8180, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7905, Kappa: 0.3913, Score: 0.6335 +[14:56:07.659422] Best epoch = 21, Best score = 0.7415 +[14:56:07.899344] log_dir: ./output_logs/retfound +[14:56:10.255072] Epoch: [28] [0/9] eta: 0:00:21 lr: 0.000362 loss: 0.3359 (0.3359) time: 2.3548 data: 2.2834 max mem: 9671 +[14:56:10.775142] Epoch: [28] [8/9] eta: 0:00:00 lr: 0.000340 loss: 0.3616 (0.4176) time: 0.3193 data: 0.2538 max mem: 9671 +[14:56:10.848879] Epoch: [28] Total time: 0:00:02 (0.3277 s / it) +[14:56:10.849641] Averaged stats: lr: 0.000340 loss: 0.3616 (0.4176) +[14:56:13.536478] val: [0/2] eta: 0:00:05 loss: 0.2967 (0.2967) time: 2.6757 data: 2.6567 max mem: 9671 +[14:56:13.546173] val: [1/2] eta: 0:00:01 loss: 0.2967 (0.6475) time: 1.3424 data: 1.3284 max mem: 9671 +[14:56:13.630281] val: Total time: 0:00:02 (1.3851 s / it) +[14:56:13.640005] val loss: 0.6474664807319641 +[14:56:13.640259] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.8007, Kappa: 0.5243, Score: 0.7058 +[14:56:13.687865] Best epoch = 21, Best score = 0.7415 +[14:56:13.937611] log_dir: ./output_logs/retfound +[14:56:16.202699] Epoch: [29] [0/9] eta: 0:00:20 lr: 0.000337 loss: 0.3300 (0.3300) time: 2.2641 data: 2.1946 max mem: 9671 +[14:56:16.722289] Epoch: [29] [8/9] eta: 0:00:00 lr: 0.000316 loss: 0.3625 (0.3715) time: 0.3092 data: 0.2439 max mem: 9671 +[14:56:16.794517] Epoch: [29] Total time: 0:00:02 (0.3174 s / it) +[14:56:16.795262] Averaged stats: lr: 0.000316 loss: 0.3625 (0.3715) +[14:56:19.345922] val: [0/2] eta: 0:00:05 loss: 0.4714 (0.4714) time: 2.5287 data: 2.5119 max mem: 9671 +[14:56:19.355103] val: [1/2] eta: 0:00:01 loss: 0.4714 (0.5518) time: 1.2687 data: 1.2560 max mem: 9671 +[14:56:19.425367] val: Total time: 0:00:02 (1.3044 s / it) +[14:56:19.434695] val loss: 0.5518224239349365 +[14:56:19.434888] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8438, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8079, Kappa: 0.4290, Score: 0.6623 +[14:56:19.479358] Best epoch = 21, Best score = 0.7415 +[14:56:19.729696] log_dir: ./output_logs/retfound +[14:56:21.925046] Epoch: [30] [0/9] eta: 0:00:19 lr: 0.000313 loss: 0.3485 (0.3485) time: 2.1943 data: 2.1225 max mem: 9671 +[14:56:22.445451] Epoch: [30] [8/9] eta: 0:00:00 lr: 0.000291 loss: 0.4160 (0.3840) time: 0.3016 data: 0.2359 max mem: 9671 +[14:56:22.516717] Epoch: [30] Total time: 0:00:02 (0.3097 s / it) +[14:56:22.517468] Averaged stats: lr: 0.000291 loss: 0.4160 (0.3840) +[14:56:25.218750] val: [0/2] eta: 0:00:05 loss: 0.4225 (0.4225) time: 2.6899 data: 2.6731 max mem: 9671 +[14:56:25.228297] val: [1/2] eta: 0:00:01 loss: 0.4225 (0.6501) time: 1.3495 data: 1.3366 max mem: 9671 +[14:56:25.297058] val: Total time: 0:00:02 (1.3845 s / it) +[14:56:25.305878] val loss: 0.6501250863075256 +[14:56:25.306069] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8445, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8078, Kappa: 0.4290, Score: 0.6626 +[14:56:25.353014] Best epoch = 21, Best score = 0.7415 +[14:56:25.601077] log_dir: ./output_logs/retfound +[14:56:27.900346] Epoch: [31] [0/9] eta: 0:00:20 lr: 0.000289 loss: 0.4108 (0.4108) time: 2.2983 data: 2.2294 max mem: 9671 +[14:56:28.420317] Epoch: [31] [8/9] eta: 0:00:00 lr: 0.000267 loss: 0.3838 (0.3420) time: 0.3131 data: 0.2478 max mem: 9671 +[14:56:28.490837] Epoch: [31] Total time: 0:00:02 (0.3211 s / it) +[14:56:28.491610] Averaged stats: lr: 0.000267 loss: 0.3838 (0.3420) +[14:56:31.208012] val: [0/2] eta: 0:00:05 loss: 0.3892 (0.3892) time: 2.6939 data: 2.6747 max mem: 9671 +[14:56:31.218032] val: [1/2] eta: 0:00:01 loss: 0.3892 (0.6480) time: 1.3517 data: 1.3374 max mem: 9671 +[14:56:31.291948] val: Total time: 0:00:02 (1.3893 s / it) +[14:56:31.300723] val loss: 0.648048460483551 +[14:56:31.300919] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8461, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8087, Kappa: 0.4290, Score: 0.6631 +[14:56:31.330948] Best epoch = 21, Best score = 0.7415 +[14:56:31.607148] log_dir: ./output_logs/retfound +[14:56:33.925751] Epoch: [32] [0/9] eta: 0:00:20 lr: 0.000264 loss: 0.3498 (0.3498) time: 2.3178 data: 2.2390 max mem: 9671 +[14:56:34.444600] Epoch: [32] [8/9] eta: 0:00:00 lr: 0.000243 loss: 0.3191 (0.3478) time: 0.3151 data: 0.2488 max mem: 9671 +[14:56:34.517256] Epoch: [32] Total time: 0:00:02 (0.3233 s / it) +[14:56:34.517945] Averaged stats: lr: 0.000243 loss: 0.3191 (0.3478) +[14:56:37.300084] val: [0/2] eta: 0:00:05 loss: 0.3005 (0.3005) time: 2.7755 data: 2.7588 max mem: 9671 +[14:56:37.309437] val: [1/2] eta: 0:00:01 loss: 0.3005 (0.8129) time: 1.3922 data: 1.3794 max mem: 9671 +[14:56:37.379090] val: Total time: 0:00:02 (1.4276 s / it) +[14:56:37.387760] val loss: 0.8128635883331299 +[14:56:37.388012] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8313, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8008, Kappa: 0.5772, Score: 0.7321 +[14:56:37.417430] Best epoch = 21, Best score = 0.7415 +[14:56:37.684153] log_dir: ./output_logs/retfound +[14:56:40.007441] Epoch: [33] [0/9] eta: 0:00:20 lr: 0.000240 loss: 0.5475 (0.5475) time: 2.3221 data: 2.2522 max mem: 9671 +[14:56:40.641158] Epoch: [33] [8/9] eta: 0:00:00 lr: 0.000219 loss: 0.4062 (0.3918) time: 0.3283 data: 0.2630 max mem: 9671 +[14:56:40.706965] Epoch: [33] Total time: 0:00:03 (0.3358 s / it) +[14:56:40.707736] Averaged stats: lr: 0.000219 loss: 0.4062 (0.3918) +[14:56:43.338411] val: [0/2] eta: 0:00:05 loss: 0.3385 (0.3385) time: 2.6197 data: 2.6028 max mem: 9671 +[14:56:43.347610] val: [1/2] eta: 0:00:01 loss: 0.3385 (0.7290) time: 1.3142 data: 1.3014 max mem: 9671 +[14:56:43.414825] val: Total time: 0:00:02 (1.3484 s / it) +[14:56:43.423565] val loss: 0.7290278077125549 +[14:56:43.423817] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8367, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8033, Kappa: 0.5772, Score: 0.7339 +[14:56:43.468167] Best epoch = 21, Best score = 0.7415 +[14:56:43.727323] log_dir: ./output_logs/retfound +[14:56:46.000957] Epoch: [34] [0/9] eta: 0:00:20 lr: 0.000217 loss: 0.4328 (0.4328) time: 2.2727 data: 2.2025 max mem: 9671 +[14:56:46.576286] Epoch: [34] [8/9] eta: 0:00:00 lr: 0.000196 loss: 0.3536 (0.3417) time: 0.3164 data: 0.2509 max mem: 9671 +[14:56:46.651145] Epoch: [34] Total time: 0:00:02 (0.3248 s / it) +[14:56:46.651939] Averaged stats: lr: 0.000196 loss: 0.3536 (0.3417) +[14:56:49.436744] val: [0/2] eta: 0:00:05 loss: 0.2532 (0.2532) time: 2.7621 data: 2.7452 max mem: 9671 +[14:56:49.445884] val: [1/2] eta: 0:00:01 loss: 0.2532 (0.9214) time: 1.3854 data: 1.3727 max mem: 9671 +[14:56:49.516701] val: Total time: 0:00:02 (1.4213 s / it) +[14:56:49.525485] val loss: 0.9214009940624237 +[14:56:49.525655] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8281, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7963, Kappa: 0.4878, Score: 0.6859 +[14:56:49.573249] Best epoch = 21, Best score = 0.7415 +[14:56:49.825945] log_dir: ./output_logs/retfound +[14:56:52.102133] Epoch: [35] [0/9] eta: 0:00:20 lr: 0.000194 loss: 0.4965 (0.4965) time: 2.2751 data: 2.1958 max mem: 9671 +[14:56:52.707823] Epoch: [35] [8/9] eta: 0:00:00 lr: 0.000174 loss: 0.3498 (0.3730) time: 0.3200 data: 0.2529 max mem: 9671 +[14:56:52.778645] Epoch: [35] Total time: 0:00:02 (0.3281 s / it) +[14:56:52.779402] Averaged stats: lr: 0.000174 loss: 0.3498 (0.3730) +[14:56:55.350790] val: [0/2] eta: 0:00:05 loss: 0.2596 (0.2596) time: 2.5643 data: 2.5473 max mem: 9671 +[14:56:55.360367] val: [1/2] eta: 0:00:01 loss: 0.2596 (0.9987) time: 1.2866 data: 1.2737 max mem: 9671 +[14:56:55.430323] val: Total time: 0:00:02 (1.3222 s / it) +[14:56:55.439262] val loss: 0.9986748993396759 +[14:56:55.439456] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7982, Kappa: 0.4878, Score: 0.6870 +[14:56:55.486893] Best epoch = 21, Best score = 0.7415 +[14:56:55.736907] log_dir: ./output_logs/retfound +[14:56:57.956849] Epoch: [36] [0/9] eta: 0:00:19 lr: 0.000171 loss: 0.4374 (0.4374) time: 2.2191 data: 2.1503 max mem: 9671 +[14:56:58.476025] Epoch: [36] [8/9] eta: 0:00:00 lr: 0.000152 loss: 0.3730 (0.3513) time: 0.3042 data: 0.2390 max mem: 9671 +[14:56:58.546095] Epoch: [36] Total time: 0:00:02 (0.3121 s / it) +[14:56:58.546897] Averaged stats: lr: 0.000152 loss: 0.3730 (0.3513) +[14:57:01.124862] val: [0/2] eta: 0:00:05 loss: 0.2994 (0.2994) time: 2.5667 data: 2.5496 max mem: 9671 +[14:57:01.136139] val: [1/2] eta: 0:00:01 loss: 0.2994 (0.8879) time: 1.2887 data: 1.2749 max mem: 9671 +[14:57:01.208352] val: Total time: 0:00:02 (1.3255 s / it) +[14:57:01.222289] val loss: 0.887900710105896 +[14:57:01.222495] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8250, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7946, Kappa: 0.4878, Score: 0.6849 +[14:57:01.256554] Best epoch = 21, Best score = 0.7415 +[14:57:01.522359] log_dir: ./output_logs/retfound +[14:57:03.791737] Epoch: [37] [0/9] eta: 0:00:20 lr: 0.000150 loss: 0.3205 (0.3205) time: 2.2685 data: 2.1991 max mem: 9671 +[14:57:04.311580] Epoch: [37] [8/9] eta: 0:00:00 lr: 0.000132 loss: 0.3006 (0.3314) time: 0.3097 data: 0.2444 max mem: 9671 +[14:57:04.384181] Epoch: [37] Total time: 0:00:02 (0.3180 s / it) +[14:57:04.384952] Averaged stats: lr: 0.000132 loss: 0.3006 (0.3314) +[14:57:06.964870] val: [0/2] eta: 0:00:05 loss: 0.3475 (0.3475) time: 2.5692 data: 2.5522 max mem: 9671 +[14:57:06.974073] val: [1/2] eta: 0:00:01 loss: 0.3475 (0.7909) time: 1.2889 data: 1.2761 max mem: 9671 +[14:57:07.038232] val: Total time: 0:00:02 (1.3216 s / it) +[14:57:07.047040] val loss: 0.7909270226955414 +[14:57:07.047223] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8289, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7930, Kappa: 0.5772, Score: 0.7313 +[14:57:07.083542] Best epoch = 21, Best score = 0.7415 +[14:57:07.333767] log_dir: ./output_logs/retfound +[14:57:09.460803] Epoch: [38] [0/9] eta: 0:00:19 lr: 0.000130 loss: 0.3957 (0.3957) time: 2.1262 data: 2.0560 max mem: 9671 +[14:57:10.122409] Epoch: [38] [8/9] eta: 0:00:00 lr: 0.000112 loss: 0.3663 (0.3488) time: 0.3097 data: 0.2439 max mem: 9671 +[14:57:10.193840] Epoch: [38] Total time: 0:00:02 (0.3178 s / it) +[14:57:10.194656] Averaged stats: lr: 0.000112 loss: 0.3663 (0.3488) +[14:57:12.873542] val: [0/2] eta: 0:00:05 loss: 0.3682 (0.3682) time: 2.6626 data: 2.6458 max mem: 9671 +[14:57:12.882720] val: [1/2] eta: 0:00:01 loss: 0.3682 (0.7528) time: 1.3356 data: 1.3230 max mem: 9671 +[14:57:12.947970] val: Total time: 0:00:02 (1.3688 s / it) +[14:57:12.956756] val loss: 0.7527517080307007 +[14:57:12.956948] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8344, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7963, Kappa: 0.5772, Score: 0.7331 +[14:57:12.995826] Best epoch = 21, Best score = 0.7415 +[14:57:13.241546] log_dir: ./output_logs/retfound +[14:57:15.391758] Epoch: [39] [0/9] eta: 0:00:19 lr: 0.000110 loss: 0.2798 (0.2798) time: 2.1493 data: 2.0795 max mem: 9671 +[14:57:15.928478] Epoch: [39] [8/9] eta: 0:00:00 lr: 0.000094 loss: 0.2936 (0.3222) time: 0.2984 data: 0.2330 max mem: 9671 +[14:57:16.004135] Epoch: [39] Total time: 0:00:02 (0.3069 s / it) +[14:57:16.004902] Averaged stats: lr: 0.000094 loss: 0.2936 (0.3222) +[14:57:18.651371] val: [0/2] eta: 0:00:05 loss: 0.3959 (0.3959) time: 2.6353 data: 2.6184 max mem: 9671 +[14:57:18.660520] val: [1/2] eta: 0:00:01 loss: 0.3959 (0.7334) time: 1.3220 data: 1.3092 max mem: 9671 +[14:57:18.733420] val: Total time: 0:00:02 (1.3590 s / it) +[14:57:18.742174] val loss: 0.7334095239639282 +[14:57:18.742399] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8320, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.7946, Kappa: 0.5243, Score: 0.7061 +[14:57:18.793369] Best epoch = 21, Best score = 0.7415 +[14:57:19.032411] log_dir: ./output_logs/retfound +[14:57:21.352589] Epoch: [40] [0/9] eta: 0:00:20 lr: 0.000092 loss: 0.2801 (0.2801) time: 2.3191 data: 2.2500 max mem: 9671 +[14:57:21.887846] Epoch: [40] [8/9] eta: 0:00:00 lr: 0.000078 loss: 0.3197 (0.3514) time: 0.3171 data: 0.2501 max mem: 9671 +[14:57:21.958667] Epoch: [40] Total time: 0:00:02 (0.3251 s / it) +[14:57:21.959380] Averaged stats: lr: 0.000078 loss: 0.3197 (0.3514) +[14:57:24.566431] val: [0/2] eta: 0:00:05 loss: 0.4082 (0.4082) time: 2.5961 data: 2.5793 max mem: 9671 +[14:57:24.576418] val: [1/2] eta: 0:00:01 loss: 0.4082 (0.7273) time: 1.3027 data: 1.2899 max mem: 9671 +[14:57:24.647354] val: Total time: 0:00:02 (1.3389 s / it) +[14:57:24.655941] val loss: 0.7272949814796448 +[14:57:24.656111] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.7946, Kappa: 0.5243, Score: 0.7058 +[14:57:24.690126] Best epoch = 21, Best score = 0.7415 +[14:57:24.952360] log_dir: ./output_logs/retfound +[14:57:27.221545] Epoch: [41] [0/9] eta: 0:00:20 lr: 0.000076 loss: 0.4471 (0.4471) time: 2.2684 data: 2.1960 max mem: 9671 +[14:57:27.738868] Epoch: [41] [8/9] eta: 0:00:00 lr: 0.000062 loss: 0.3231 (0.3198) time: 0.3095 data: 0.2440 max mem: 9671 +[14:57:27.815290] Epoch: [41] Total time: 0:00:02 (0.3181 s / it) +[14:57:27.816053] Averaged stats: lr: 0.000062 loss: 0.3231 (0.3198) +[14:57:30.354637] val: [0/2] eta: 0:00:05 loss: 0.4002 (0.4002) time: 2.5274 data: 2.5106 max mem: 9671 +[14:57:30.363803] val: [1/2] eta: 0:00:01 loss: 0.4002 (0.7321) time: 1.2680 data: 1.2553 max mem: 9671 +[14:57:30.430941] val: Total time: 0:00:02 (1.3022 s / it) +[14:57:30.439610] val loss: 0.7320790588855743 +[14:57:30.439786] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8313, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7946, Kappa: 0.5772, Score: 0.7321 +[14:57:30.475880] Best epoch = 21, Best score = 0.7415 +[14:57:30.735133] log_dir: ./output_logs/retfound +[14:57:33.076013] Epoch: [42] [0/9] eta: 0:00:21 lr: 0.000061 loss: 0.2169 (0.2169) time: 2.3400 data: 2.2688 max mem: 9671 +[14:57:33.596910] Epoch: [42] [8/9] eta: 0:00:00 lr: 0.000048 loss: 0.2901 (0.3057) time: 0.3178 data: 0.2522 max mem: 9671 +[14:57:33.664727] Epoch: [42] Total time: 0:00:02 (0.3255 s / it) +[14:57:33.665749] Averaged stats: lr: 0.000048 loss: 0.2901 (0.3057) +[14:57:36.308336] val: [0/2] eta: 0:00:05 loss: 0.3896 (0.3896) time: 2.6318 data: 2.6150 max mem: 9671 +[14:57:36.317723] val: [1/2] eta: 0:00:01 loss: 0.3896 (0.7542) time: 1.3203 data: 1.3075 max mem: 9671 +[14:57:36.384581] val: Total time: 0:00:02 (1.3544 s / it) +[14:57:36.393217] val loss: 0.7541973888874054 +[14:57:36.393420] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:57:36.446217] Best epoch = 21, Best score = 0.7415 +[14:57:36.679714] log_dir: ./output_logs/retfound +[14:57:38.967905] Epoch: [43] [0/9] eta: 0:00:20 lr: 0.000047 loss: 0.2695 (0.2695) time: 2.2873 data: 2.2171 max mem: 9671 +[14:57:39.498752] Epoch: [43] [8/9] eta: 0:00:00 lr: 0.000036 loss: 0.3200 (0.3327) time: 0.3130 data: 0.2476 max mem: 9671 +[14:57:39.567551] Epoch: [43] Total time: 0:00:02 (0.3209 s / it) +[14:57:39.568347] Averaged stats: lr: 0.000036 loss: 0.3200 (0.3327) +[14:57:42.200056] val: [0/2] eta: 0:00:05 loss: 0.3908 (0.3908) time: 2.6095 data: 2.5925 max mem: 9671 +[14:57:42.209489] val: [1/2] eta: 0:00:01 loss: 0.3908 (0.7752) time: 1.3092 data: 1.2963 max mem: 9671 +[14:57:42.276502] val: Total time: 0:00:02 (1.3433 s / it) +[14:57:42.285237] val loss: 0.7751908898353577 +[14:57:42.285410] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:57:42.319571] Best epoch = 21, Best score = 0.7415 +[14:57:42.594247] log_dir: ./output_logs/retfound +[14:57:44.775425] Epoch: [44] [0/9] eta: 0:00:19 lr: 0.000035 loss: 0.3314 (0.3314) time: 2.1803 data: 2.1069 max mem: 9671 +[14:57:45.312941] Epoch: [44] [8/9] eta: 0:00:00 lr: 0.000026 loss: 0.3314 (0.3466) time: 0.3019 data: 0.2361 max mem: 9671 +[14:57:45.381497] Epoch: [44] Total time: 0:00:02 (0.3097 s / it) +[14:57:45.382334] Averaged stats: lr: 0.000026 loss: 0.3314 (0.3466) +[14:57:48.122472] val: [0/2] eta: 0:00:05 loss: 0.4027 (0.4027) time: 2.7175 data: 2.7005 max mem: 9671 +[14:57:48.132868] val: [1/2] eta: 0:00:01 loss: 0.4027 (0.7739) time: 1.3636 data: 1.3503 max mem: 9671 +[14:57:48.208404] val: Total time: 0:00:02 (1.4021 s / it) +[14:57:48.217323] val loss: 0.7738757431507111 +[14:57:48.217542] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:57:48.270034] Best epoch = 21, Best score = 0.7415 +[14:57:48.516895] log_dir: ./output_logs/retfound +[14:57:50.813397] Epoch: [45] [0/9] eta: 0:00:20 lr: 0.000025 loss: 0.3766 (0.3766) time: 2.2954 data: 2.2244 max mem: 9671 +[14:57:51.408895] Epoch: [45] [8/9] eta: 0:00:00 lr: 0.000017 loss: 0.3525 (0.3718) time: 0.3211 data: 0.2556 max mem: 9671 +[14:57:51.484139] Epoch: [45] Total time: 0:00:02 (0.3297 s / it) +[14:57:51.484949] Averaged stats: lr: 0.000017 loss: 0.3525 (0.3718) +[14:57:54.126529] val: [0/2] eta: 0:00:05 loss: 0.4050 (0.4050) time: 2.6188 data: 2.5983 max mem: 9671 +[14:57:54.145625] val: [1/2] eta: 0:00:01 loss: 0.4050 (0.7762) time: 1.3185 data: 1.2992 max mem: 9671 +[14:57:54.302505] val: Total time: 0:00:02 (1.3979 s / it) +[14:57:54.315918] val loss: 0.7761802673339844 +[14:57:54.316161] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7930, Kappa: 0.5772, Score: 0.7311 +[14:57:54.451158] Best epoch = 21, Best score = 0.7415 +[14:57:54.760231] log_dir: ./output_logs/retfound +[14:57:57.078960] Epoch: [46] [0/9] eta: 0:00:20 lr: 0.000016 loss: 0.3260 (0.3260) time: 2.3179 data: 2.2465 max mem: 9671 +[14:57:57.601393] Epoch: [46] [8/9] eta: 0:00:00 lr: 0.000010 loss: 0.3260 (0.3344) time: 0.3155 data: 0.2497 max mem: 9671 +[14:57:57.674095] Epoch: [46] Total time: 0:00:02 (0.3237 s / it) +[14:57:57.674872] Averaged stats: lr: 0.000010 loss: 0.3260 (0.3344) +[14:58:00.395809] val: [0/2] eta: 0:00:05 loss: 0.4028 (0.4028) time: 2.6980 data: 2.6811 max mem: 9671 +[14:58:00.404973] val: [1/2] eta: 0:00:01 loss: 0.4028 (0.7721) time: 1.3533 data: 1.3406 max mem: 9671 +[14:58:00.479539] val: Total time: 0:00:02 (1.3912 s / it) +[14:58:00.488157] val loss: 0.7720681726932526 +[14:58:00.488324] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:58:00.521852] Best epoch = 21, Best score = 0.7415 +[14:58:00.770225] log_dir: ./output_logs/retfound +[14:58:03.006390] Epoch: [47] [0/9] eta: 0:00:20 lr: 0.000010 loss: 0.3762 (0.3762) time: 2.2353 data: 2.1658 max mem: 9671 +[14:58:03.526097] Epoch: [47] [8/9] eta: 0:00:00 lr: 0.000005 loss: 0.3166 (0.3650) time: 0.3060 data: 0.2407 max mem: 9671 +[14:58:03.594697] Epoch: [47] Total time: 0:00:02 (0.3138 s / it) +[14:58:03.595462] Averaged stats: lr: 0.000005 loss: 0.3166 (0.3650) +[14:58:06.270271] val: [0/2] eta: 0:00:05 loss: 0.4020 (0.4020) time: 2.6511 data: 2.6342 max mem: 9671 +[14:58:06.279337] val: [1/2] eta: 0:00:01 loss: 0.4020 (0.7686) time: 1.3298 data: 1.3172 max mem: 9671 +[14:58:06.348937] val: Total time: 0:00:02 (1.3652 s / it) +[14:58:06.357564] val loss: 0.7686194181442261 +[14:58:06.357739] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:58:06.405900] Best epoch = 21, Best score = 0.7415 +[14:58:06.646286] log_dir: ./output_logs/retfound +[14:58:09.158395] Epoch: [48] [0/9] eta: 0:00:22 lr: 0.000005 loss: 0.3463 (0.3463) time: 2.5113 data: 2.4422 max mem: 9671 +[14:58:09.677464] Epoch: [48] [8/9] eta: 0:00:00 lr: 0.000002 loss: 0.3568 (0.3456) time: 0.3366 data: 0.2714 max mem: 9671 +[14:58:09.748279] Epoch: [48] Total time: 0:00:03 (0.3446 s / it) +[14:58:09.749032] Averaged stats: lr: 0.000002 loss: 0.3568 (0.3456) +[14:58:12.455163] val: [0/2] eta: 0:00:05 loss: 0.3997 (0.3997) time: 2.6951 data: 2.6783 max mem: 9671 +[14:58:12.464387] val: [1/2] eta: 0:00:01 loss: 0.3997 (0.7695) time: 1.3519 data: 1.3392 max mem: 9671 +[14:58:12.539649] val: Total time: 0:00:02 (1.3901 s / it) +[14:58:12.548369] val loss: 0.7694710791110992 +[14:58:12.548555] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:58:12.583459] Best epoch = 21, Best score = 0.7415 +[14:58:12.839203] log_dir: ./output_logs/retfound +[14:58:15.238783] Epoch: [49] [0/9] eta: 0:00:21 lr: 0.000002 loss: 0.3057 (0.3057) time: 2.3987 data: 2.3292 max mem: 9671 +[14:58:15.812939] Epoch: [49] [8/9] eta: 0:00:00 lr: 0.000001 loss: 0.3238 (0.3242) time: 0.3302 data: 0.2649 max mem: 9671 +[14:58:15.888979] Epoch: [49] Total time: 0:00:03 (0.3388 s / it) +[14:58:15.889806] Averaged stats: lr: 0.000001 loss: 0.3238 (0.3242) +[14:58:18.525371] val: [0/2] eta: 0:00:05 loss: 0.3984 (0.3984) time: 2.6242 data: 2.6073 max mem: 9671 +[14:58:18.534742] val: [1/2] eta: 0:00:01 loss: 0.3984 (0.7701) time: 1.3165 data: 1.3037 max mem: 9671 +[14:58:18.606261] val: Total time: 0:00:02 (1.3529 s / it) +[14:58:18.615012] val loss: 0.7701238989830017 +[14:58:18.615190] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:58:18.650709] Best epoch = 21, Best score = 0.7415 +[14:58:21.508200] Test with the best model, epoch = 21: +[14:58:24.158066] test: [0/3] eta: 0:00:07 loss: 0.3573 (0.3573) time: 2.6395 data: 2.6222 max mem: 9671 +[14:58:24.243004] test: [2/3] eta: 0:00:00 loss: 0.3573 (0.4751) time: 0.9079 data: 0.8805 max mem: 9671 +[14:58:24.314325] test: Total time: 0:00:02 (0.9322 s / it) +[14:58:24.323443] val loss: 0.47513073682785034 +[14:58:24.323566] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8371, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7323, Recall: 0.7537, + Average Precision: 0.7821, Kappa: 0.4842, Score: 0.6877 +[14:58:24.989534] Training time 0:05:21 +[rank0]:[W701 14:58:25.317431248 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/100/retfound acc=0.8333 auroc=0.8373161764705882 f1_macro=0.7419 qwk=0.4842105263157894 diff --git a/results/downsample/papila/100/vit/confusion_matrix.png b/results/downsample/papila/100/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..f40a1551c4b96a2f0b620479cc865f469b1bd4e1 --- /dev/null +++ b/results/downsample/papila/100/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27f4ffb4f4a539debbfc987395c1999c389c1adbc49f6f2fc9065c069227fe56 +size 73078 diff --git a/results/downsample/papila/100/vit/log.csv b/results/downsample/papila/100/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..ce68bd8cef0bffa9e5989839d31502d27b226910 --- /dev/null +++ b/results/downsample/papila/100/vit/log.csv @@ -0,0 +1,30 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8081949055194855,0.38095238095238093,0.553125,0.32084904394735353,5.545808836528073e-08 +1,0.7463895827531815,0.6904761904761905,0.69375,0.5318601694915255,1.29402206185655e-07 +2,0.684089183807373,0.2857142857142857,0.78125,0.35694758672699844,2.0334632400602932e-07 +3,0.6925588548183441,0.7380952380952381,0.721875,0.5793965653156564,2.7729044182640365e-07 +4,0.646022841334343,0.5238095238095238,0.7124999999999999,0.4441021345246959,3.512345596467779e-07 +5,0.6103057861328125,0.5714285714285714,0.753125,0.5036502544055598,3.694672444929302e-07 +6,0.5391916632652283,0.5476190476190477,0.765625,0.49254615505996213,3.683426684987483e-07 +7,0.5554562211036682,0.5952380952380952,0.784375,0.5296914201409342,3.663241848222764e-07 +8,0.5353901088237762,0.6428571428571429,0.78125,0.5261474644139035,3.6342162731308893e-07 +9,0.5195781961083412,0.6904761904761905,0.784375,0.5620685028248588,3.5964913693960667e-07 +10,0.5137002021074295,0.6428571428571429,0.8125,0.5859409589478831,3.550250928957199e-07 +11,0.5435433834791183,0.6904761904761905,0.80625,0.6041263127115114,3.495720230592029e-07 +12,0.5050823017954826,0.6666666666666666,0.815625,0.5331997863247863,3.4331649423815874e-07 +13,0.5033866092562675,0.7380952380952381,0.859375,0.6728023928071764,3.362889827402e-07 +14,0.4588450863957405,0.7619047619047619,0.88125,0.6862313201211271,3.285237258949416e-07 +15,0.44873566925525665,0.7380952380952381,0.859375,0.6433848870056497,3.2005855525317277e-07 +16,0.42817943543195724,0.7380952380952381,0.859375,0.6433848870056497,3.1093471227534435e-07 +17,0.46573343873023987,0.8095238095238095,0.865625,0.7235844017094016,3.0119664740731875e-07 +18,0.4151291847229004,0.8095238095238095,0.88125,0.728792735042735,2.908918035222662e-07 +19,0.4494623392820358,0.7857142857142857,0.86875,0.6879721898037188,2.800703847837553e-07 +20,0.4139930456876755,0.7619047619047619,0.840625,0.6573450854700854,2.687851120561149e-07 +21,0.40749695897102356,0.7380952380952381,0.815625,0.6106465653156564,2.570909660536824e-07 +22,0.40808261930942535,0.7619047619047619,0.821875,0.6664396534544604,2.450449194802903e-07 +23,0.4032032936811447,0.7380952380952381,0.815625,0.6106465653156564,2.3270565946397963e-07 +24,0.40627045184373856,0.7380952380952381,0.815625,0.6106465653156564,2.2013330163921197e-07 +25,0.3788785859942436,0.7380952380952381,0.790625,0.6023132319823231,2.0738909726954043e-07 +26,0.4226858466863632,0.7619047619047619,0.7718750000000001,0.5712138866291551,1.9453513483761127e-07 +27,0.3840523436665535,0.7142857142857143,0.78125,0.558302142820394,1.8163403755631687e-07 +28,0.3793400377035141,0.7380952380952381,0.78125,0.5531960385467029,1.6874865827479806e-07 diff --git a/results/downsample/papila/100/vit/metrics.json b/results/downsample/papila/100/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..0cc8ef3739d1d1749dce7989beef7cd71e6acc91 --- /dev/null +++ b/results/downsample/papila/100/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.75, + "balanced_accuracy": 0.7022058823529411, + "precision_macro": 0.6491525423728814, + "recall_macro": 0.7022058823529411, + "f1_macro": 0.6612252736700595, + "precision_weighted": 0.8033898305084746, + "recall_weighted": 0.75, + "f1_weighted": 0.7685807566737085, + "cohen_kappa": 0.33282904689863835, + "quadratic_weighted_kappa": 0.33282904689863835, + "mcc": 0.3473299378728699, + "auroc": 0.7849264705882353, + "auprc": 0.5894183392973115, + "sensitivity": 0.625, + "specificity": 0.7794117647058824, + "precision_pos": 0.4, + "f1_pos": 0.4878048780487805, + "per_class": { + "0": { + "precision": 0.8983050847457628, + "recall": 0.7794117647058824, + "f1-score": 0.8346456692913385, + "support": 68.0 + }, + "1": { + "precision": 0.4, + "recall": 0.625, + "f1-score": 0.4878048780487805, + "support": 16.0 + }, + "accuracy": 0.75, + "macro avg": { + "precision": 0.6491525423728814, + "recall": 0.7022058823529411, + "f1-score": 0.6612252736700595, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8033898305084746, + "recall": 0.75, + "f1-score": 0.7685807566737085, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/downsample/papila/100/vit/pr.png b/results/downsample/papila/100/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a103b55caac4fc2eb8c4d001a8654bda6534474b --- /dev/null +++ b/results/downsample/papila/100/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fe3680bdd3b1a11889ce1112c7916bb49503f03877f55757061e7ece3d1a591 +size 48629 diff --git a/results/downsample/papila/100/vit/roc.png b/results/downsample/papila/100/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..afae9df8b9cddf511ed3a568506a50daa8e8e961 --- /dev/null +++ b/results/downsample/papila/100/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b8bee8ac5f985f7ba6a106dbbc229eaf5b1647dfc2870d38fa996b35b97178a +size 57359 diff --git a/results/downsample/papila/100/vit/test_pred.npz b/results/downsample/papila/100/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..959718674514d2e52f123d6e6b6a8afd86d1df1e --- /dev/null +++ b/results/downsample/papila/100/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:793dde3491b2fe8d25ef8123d6a8ae58d8dfb2a380dfce74b22024499db7532c +size 1854 diff --git a/results/downsample/papila/100/vit/train.log b/results/downsample/papila/100/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e295a528ad8dac3f1aff81705c151e377dad1d89 --- /dev/null +++ b/results/downsample/papila/100/vit/train.log @@ -0,0 +1,154 @@ +[vit] train=294 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8082 val_acc=0.3810 val_auc=0.5531 score=0.3208 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7464 val_acc=0.6905 val_auc=0.6937 score=0.5319 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6841 val_acc=0.2857 val_auc=0.7812 score=0.3569 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6926 val_acc=0.7381 val_auc=0.7219 score=0.5794 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6460 val_acc=0.5238 val_auc=0.7125 score=0.4441 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.6103 val_acc=0.5714 val_auc=0.7531 score=0.5037 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5392 val_acc=0.5476 val_auc=0.7656 score=0.4925 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5555 val_acc=0.5952 val_auc=0.7844 score=0.5297 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5354 val_acc=0.6429 val_auc=0.7812 score=0.5261 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.5196 val_acc=0.6905 val_auc=0.7844 score=0.5621 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.5137 val_acc=0.6429 val_auc=0.8125 score=0.5859 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.5435 val_acc=0.6905 val_auc=0.8063 score=0.6041 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.5051 val_acc=0.6667 val_auc=0.8156 score=0.5332 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.5034 val_acc=0.7381 val_auc=0.8594 score=0.6728 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4588 val_acc=0.7619 val_auc=0.8812 score=0.6862 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4487 val_acc=0.7381 val_auc=0.8594 score=0.6434 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4282 val_acc=0.7381 val_auc=0.8594 score=0.6434 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.4657 val_acc=0.8095 val_auc=0.8656 score=0.7236 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4151 val_acc=0.8095 val_auc=0.8812 score=0.7288 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4495 val_acc=0.7857 val_auc=0.8688 score=0.6880 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.4140 val_acc=0.7619 val_auc=0.8406 score=0.6573 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4075 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.4081 val_acc=0.7619 val_auc=0.8219 score=0.6664 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4032 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.4063 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3789 val_acc=0.7381 val_auc=0.7906 score=0.6023 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4227 val_acc=0.7619 val_auc=0.7719 score=0.5712 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3841 val_acc=0.7143 val_auc=0.7812 score=0.5583 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3793 val_acc=0.7381 val_auc=0.7812 score=0.5532 +[vit] early stop at ep28 (best ep18 score=0.7288) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=18 best_val_score=0.7288 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/downsample/papila/100/vit acc=0.7500 auroc=0.7849264705882353 f1_macro=0.6612 qwk=0.33282904689863835 diff --git a/results/downsample/summary.csv b/results/downsample/summary.csv new file mode 100644 index 0000000000000000000000000000000000000000..c8c4bba9e1a8f8d08baa15c0a94d17289f4cdf3f --- /dev/null +++ b/results/downsample/summary.csv @@ -0,0 +1,46 @@ +dataset,frac,n_train,model,acc,auroc,f1,kappa,mcc,n_test +adam,100,280,retfound,0.925,0.9516129032258065,0.89247311827957,0.7849462365591398,0.7849462365591398,80 +adam,100,280,resnet,0.7375,0.7795698924731183,0.6836753906985502,0.38144329896907225,0.4031935495202789,80 +adam,100,280,vit,0.9125,0.9318996415770608,0.8769501208525599,0.7539543057996485,0.7544204852635336,80 +adam,50,140,retfound,0.925,0.9444444444444444,0.89247311827957,0.7849462365591398,0.7849462365591398,80 +adam,50,140,resnet,0.875,0.9247311827956989,0.7948717948717949,0.5934959349593496,0.6119778033738925,80 +adam,50,140,vit,0.8625,0.9005376344086021,0.8066359041968798,0.6133567662565905,0.613736012487117,80 +adam,25,70,retfound,0.875,0.921146953405018,0.8333333333333333,0.6677740863787376,0.6737497456694623,80 +adam,25,70,resnet,0.825,0.8494623655913978,0.7491039426523298,0.4982078853046594,0.4982078853046595,80 +adam,25,70,vit,0.8125,0.8351254480286737,0.7540479606476738,0.5106035889070146,0.5175363875271632,80 +adam,10,28,retfound,0.775,0.8409498207885304,0.43661971830985913,0.0,0.0,80 +adam,10,28,resnet,0.775,0.7706093189964158,0.6893874029335634,0.3793103448275862,0.38021562140115595,80 +adam,10,28,vit,0.8,0.8221326164874552,0.7012138188608776,0.4029850746268657,0.4041119295602435,80 +adam,5,14,retfound,0.775,0.8006272401433692,0.43661971830985913,0.0,0.0,80 +adam,5,14,resnet,0.825,0.7840501792114696,0.6785304247990815,0.375,0.41907903806247476,80 +adam,5,14,vit,0.7125,0.7661290322580645,0.6342675412442854,0.2755905511811023,0.28254302982294827,80 +airogs,100,5000,retfound,0.908,0.970758,0.9079764419691441,0.8160000000000001,0.8164181131383057,1000 +airogs,100,5000,resnet,0.898,0.963978,0.8979983679738875,0.796,0.7960254732227212,1000 +airogs,100,5000,vit,0.873,0.945154,0.872984631140368,0.746,0.7461805975595589,1000 +airogs,50,2500,retfound,0.898,0.9645199999999999,0.897999591998368,0.796,0.796006368076417,1000 +airogs,50,2500,resnet,0.879,0.9442860000000001,0.878999878999879,0.758,0.758001516004548,1000 +airogs,50,2500,vit,0.863,0.939084,0.862999862999863,0.726,0.7260014520043561,1000 +airogs,25,1250,retfound,0.881,0.9543219999999999,0.8809903602191778,0.762,0.7621234740049935,1000 +airogs,25,1250,resnet,0.855,0.939054,0.8549986949882549,0.71,0.7100127803450703,1000 +airogs,25,1250,vit,0.832,0.9150260000000001,0.8317031243112851,0.6639999999999999,0.6663550623441318,1000 +airogs,10,500,retfound,0.851,0.9329420000000002,0.850981968818227,0.702,0.7021699456927752,1000 +airogs,10,500,resnet,0.819,0.891534,0.8188026761142884,0.638,0.6393941202477768,1000 +airogs,10,500,vit,0.8,0.890196,0.8,0.6,0.6,1000 +airogs,5,250,retfound,0.834,0.9092500000000001,0.8339574931182383,0.6679999999999999,0.6683422788926326,1000 +airogs,5,250,resnet,0.774,0.8552180000000001,0.7738471206535618,0.548,0.5487424019308351,1000 +airogs,5,250,vit,0.766,0.8468760000000001,0.7659541270088938,0.532,0.5322086667040412,1000 +papila,100,294,retfound,0.8333333333333334,0.8373161764705882,0.7418788410886743,0.4842105263157894,0.48553038055886144,84 +papila,100,294,resnet,0.8809523809523809,0.7720588235294117,0.772481040086674,0.5493562231759657,0.5706103612971936,84 +papila,100,294,vit,0.75,0.7849264705882353,0.6612252736700595,0.33282904689863835,0.3473299378728699,84 +papila,50,146,retfound,0.7976190476190477,0.8088235294117647,0.7053847740870642,0.4137931034482759,0.42008402520840293,84 +papila,50,146,resnet,0.7142857142857143,0.7141544117647058,0.6407697790449038,0.30578512396694213,0.33443445580376163,84 +papila,50,146,vit,0.6904761904761905,0.7297794117647058,0.6208333333333333,0.27393617021276595,0.30620064936195557,84 +papila,25,73,retfound,0.75,0.7421875,0.636063544460491,0.27586206896551724,0.28005601680560194,84 +papila,25,73,resnet,0.7023809523809523,0.6240808823529411,0.5286195286195285,0.05745062836624781,0.05749150322649658,84 +papila,25,73,vit,0.6071428571428571,0.6443014705882353,0.5354449472096531,0.12389380530973448,0.14348601079588785,84 +papila,10,29,retfound,0.8095238095238095,0.6387867647058825,0.4473684210526316,0.0,0.0,84 +papila,10,29,resnet,0.4880952380952381,0.609375,0.4604929051530993,0.07194244604316535,0.10249000771134846,84 +papila,10,29,vit,0.7976190476190477,0.6378676470588236,0.6794612794612795,0.3590664272890485,0.35932189516560364,84 +papila,5,15,retfound,0.8095238095238095,0.6806066176470589,0.4473684210526316,0.0,0.0,84 +papila,5,15,resnet,0.8214285714285714,0.5647977941176471,0.6221889055472264,0.2622950819672131,0.29250896965085227,84 +papila,5,15,vit,0.7142857142857143,0.703125,0.6190476190476191,0.25222551928783377,0.2654384291727511,84 diff --git a/results/idrid/resnet/confusion_matrix.png b/results/idrid/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..7e2a2d4d2aab46d6dc41530954ea8db949a2b798 --- /dev/null +++ b/results/idrid/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05e03433a6c2d50d44d37634d7d6b35dae5f5086c981dc00d6233dd5d80ccdc5 +size 89653 diff --git a/results/idrid/resnet/log.csv b/results/idrid/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..b2975b66bfa17aeef94356f8c00c37b1c193b29f --- /dev/null +++ b/results/idrid/resnet/log.csv @@ -0,0 +1,32 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.610917568206787,0.24444444444444444,0.5564363573986669,0.2880297479958223,0.000125 +1,1.6061263382434845,0.26666666666666666,0.5978772118457363,0.3089813620970104,0.0002916666666666667 +2,1.5950437784194946,0.17777777777777778,0.6373878192392627,0.268972231539734,0.0004583333333333333 +3,1.5709975957870483,0.28888888888888886,0.741618501868101,0.38811929440831205,0.0004996859161456965 +4,1.5270738303661346,0.37777777777777777,0.8134109168541807,0.48202965984408114,0.0004982915790812436 +5,1.4991064071655273,0.37777777777777777,0.8025730556755015,0.482792881290798,0.0004957883115509159 +6,1.4465978145599365,0.4,0.8333796398952776,0.49914158654809365,0.0004921872937551814 +7,1.381913661956787,0.4222222222222222,0.8624843815501393,0.5362160622617173,0.000487504608713676 +8,1.2907358705997467,0.4,0.8444602539279765,0.5026224788523757,0.00048176117043453436 +9,1.173436462879181,0.4888888888888889,0.8316577700549234,0.5544458799900291,0.000474982630507352 +10,1.0984624326229095,0.4,0.8374054198081863,0.5063069805599729,0.00046719926353695914 +11,0.9870758503675461,0.35555555555555557,0.8143335782269222,0.4668401369391739,0.00045844583192968674 +12,0.9702458530664444,0.5111111111111111,0.8440861127458963,0.5687596692983202,0.00044876143063602076 +13,0.8469409793615341,0.6222222222222222,0.8601033934434096,0.6508990896882053,0.00043818931254306284 +14,0.809129387140274,0.5555555555555556,0.8289437558433148,0.5868376928001268,0.00042677669529663686 +15,0.6975472569465637,0.5555555555555556,0.8528418852045556,0.5843890365090614,0.0004145745504158204 +16,0.6515198796987534,0.6444444444444445,0.8735213102344217,0.6638122759510522,0.0004016373756417668 +17,0.5912734866142273,0.5333333333333333,0.8448454194895654,0.590425770842543,0.00038802295153756415 +18,0.564646415412426,0.6,0.8778549044387056,0.6398007486620158,0.0003737920834262134 +19,0.48417840898036957,0.6222222222222222,0.8513811039773349,0.6179886690079833,0.0003590083298192957 +20,0.4074878543615341,0.7555555555555555,0.8705619184819264,0.7573485834158161,0.0003437377185492303 +21,0.38012996315956116,0.5333333333333333,0.8457767919977222,0.5406446560516328,0.0003280484518729466 +22,0.33787114173173904,0.6666666666666666,0.8267917406951086,0.6688508561967549,0.00031201060186404833 +23,0.3281714469194412,0.6888888888888889,0.8697811511313918,0.7124347908161536,0.0002956957974539226 +24,0.30072716996073723,0.6888888888888889,0.8664853120346304,0.7127239298778325,0.0002791769045195441 +25,0.2503240071237087,0.6222222222222222,0.8636887226502303,0.6593118319221184,0.0002625277004467798 +26,0.254607368260622,0.6666666666666666,0.8593973098383684,0.6828244022072298,0.00024582254462267474 +27,0.2855723649263382,0.6888888888888889,0.8891374700065559,0.7097529965622164,0.00022913604632837759 +28,0.2312307320535183,0.6888888888888889,0.8687372157951548,0.6993186294694024,0.00021254273151597963 +29,0.22885946929454803,0.6444444444444445,0.8613922976242543,0.6673661883882719,0.00019611670995752162 +30,0.20332366228103638,0.6222222222222222,0.8615436923877182,0.6446214908505451,0.00017993134425276094 diff --git a/results/idrid/resnet/metrics.json b/results/idrid/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..0658a8ccf7246123e965c22c6669383da1f4af50 --- /dev/null +++ b/results/idrid/resnet/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 92, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.6195652173913043, + "balanced_accuracy": 0.5597343453510437, + "precision_macro": 0.5642299729256252, + "recall_macro": 0.5597343453510437, + "f1_macro": 0.551470880540648, + "precision_weighted": 0.6321425109232293, + "recall_weighted": 0.6195652173913043, + "f1_weighted": 0.6165810021345915, + "cohen_kappa": 0.49968924798011194, + "quadratic_weighted_kappa": 0.839103628171686, + "mcc": 0.5055826890609914, + "auroc_macro_ovr": 0.8880725893746568, + "auroc_weighted_ovr": 0.8858624931153432, + "auprc_macro": 0.6335777447015164, + "auroc_per_class": { + "0": 0.9871794871794872, + "1": 0.8988505747126436, + "2": 0.8233738762559492, + "3": 0.827843137254902, + "4": 0.9031158714703019 + }, + "per_class": { + "0": { + "precision": 0.8846153846153846, + "recall": 0.8846153846153846, + "f1-score": 0.8846153846153846, + "support": 26.0 + }, + "1": { + "precision": 0.3333333333333333, + "recall": 0.2, + "f1-score": 0.25, + "support": 5.0 + }, + "2": { + "precision": 0.6086956521739131, + "recall": 0.45161290322580644, + "f1-score": 0.5185185185185185, + "support": 31.0 + }, + "3": { + "precision": 0.4230769230769231, + "recall": 0.6470588235294118, + "f1-score": 0.5116279069767442, + "support": 17.0 + }, + "4": { + "precision": 0.5714285714285714, + "recall": 0.6153846153846154, + "f1-score": 0.5925925925925926, + "support": 13.0 + }, + "accuracy": 0.6195652173913043, + "macro avg": { + "precision": 0.5642299729256252, + "recall": 0.5597343453510437, + "f1-score": 0.551470880540648, + "support": 92.0 + }, + "weighted avg": { + "precision": 0.6321425109232293, + "recall": 0.6195652173913043, + "f1-score": 0.6165810021345915, + "support": 92.0 + } + } +} \ No newline at end of file diff --git a/results/idrid/resnet/pr.png b/results/idrid/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..ee5de70199cc6f7a6a296ffbb74d13ca3f99e3c6 --- /dev/null +++ b/results/idrid/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:249abe4b8271f3c02bdb925c27fa9fb35c431766d5fce6668182b8708f8e9049 +size 91504 diff --git a/results/idrid/resnet/roc.png b/results/idrid/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..34387f865577d39d42b1262148ee0cdda15795d1 --- /dev/null +++ b/results/idrid/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5fb09850b830041347df9d8c8f17277f85d59f6429b8ab69f52b45e9920fc5f0 +size 83163 diff --git a/results/idrid/resnet/test_pred.npz b/results/idrid/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..e4fc0aedde377b3d01b420a004c862fd2b5218f0 --- /dev/null +++ b/results/idrid/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f07355ef0e8daa7d69118882c1d2bb01997c8cca645272bce83c306ad3084a6 +size 3086 diff --git a/results/idrid/resnet/train.log b/results/idrid/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..bab1728c571f7c59523531a0adfff158451658d3 --- /dev/null +++ b/results/idrid/resnet/train.log @@ -0,0 +1,163 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=318 val=45 test=92 classes=['0', '1', '2', '3', '4'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=1.6109 val_acc=0.2444 val_auc=0.5564 score=0.2880 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=1.6061 val_acc=0.2667 val_auc=0.5979 score=0.3090 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=1.5950 val_acc=0.1778 val_auc=0.6374 score=0.2690 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=1.5710 val_acc=0.2889 val_auc=0.7416 score=0.3881 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=1.5271 val_acc=0.3778 val_auc=0.8134 score=0.4820 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=1.4991 val_acc=0.3778 val_auc=0.8026 score=0.4828 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=1.4466 val_acc=0.4000 val_auc=0.8334 score=0.4991 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=1.3819 val_acc=0.4222 val_auc=0.8625 score=0.5362 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=1.2907 val_acc=0.4000 val_auc=0.8445 score=0.5026 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=1.1734 val_acc=0.4889 val_auc=0.8317 score=0.5544 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=1.0985 val_acc=0.4000 val_auc=0.8374 score=0.5063 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.9871 val_acc=0.3556 val_auc=0.8143 score=0.4668 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.9702 val_acc=0.5111 val_auc=0.8441 score=0.5688 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.8469 val_acc=0.6222 val_auc=0.8601 score=0.6509 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.8091 val_acc=0.5556 val_auc=0.8289 score=0.5868 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.6975 val_acc=0.5556 val_auc=0.8528 score=0.5844 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.6515 val_acc=0.6444 val_auc=0.8735 score=0.6638 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.5913 val_acc=0.5333 val_auc=0.8448 score=0.5904 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.5646 val_acc=0.6000 val_auc=0.8779 score=0.6398 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.4842 val_acc=0.6222 val_auc=0.8514 score=0.6180 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.4075 val_acc=0.7556 val_auc=0.8706 score=0.7573 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.3801 val_acc=0.5333 val_auc=0.8458 score=0.5406 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.3379 val_acc=0.6667 val_auc=0.8268 score=0.6689 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.3282 val_acc=0.6889 val_auc=0.8698 score=0.7124 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.3007 val_acc=0.6889 val_auc=0.8665 score=0.7127 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.2503 val_acc=0.6222 val_auc=0.8637 score=0.6593 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.2546 val_acc=0.6667 val_auc=0.8594 score=0.6828 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.2856 val_acc=0.6889 val_auc=0.8891 score=0.7098 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.2312 val_acc=0.6889 val_auc=0.8687 score=0.6993 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.2289 val_acc=0.6444 val_auc=0.8614 score=0.6674 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep30 loss=0.2033 val_acc=0.6222 val_auc=0.8615 score=0.6446 +[resnet] early stop at ep30 (best ep20 score=0.7573) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=20 best_val_score=0.7573 -> saved test_pred.npz (92 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/idrid/resnet acc=0.6196 auroc_macro_ovr=0.8880725893746568 f1_macro=0.5515 qwk=0.839103628171686 diff --git a/results/idrid/retfound/confusion_matrix.png b/results/idrid/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..c5dc21f3fae0641d3e01ba680506fc66dcfeb49f --- /dev/null +++ b/results/idrid/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3da449f44b7e60cd28ec7093203708f2bb814b7e577fb6542e441829b449641e +size 93333 diff --git a/results/idrid/retfound/confusion_matrix_test.jpg b/results/idrid/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..1108b88d6debd8455b2d447c6a426205f5d6b71d --- /dev/null +++ b/results/idrid/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:332f54e5377cc1e72f9e9adc03b7bdf5950169ba08cbb3a3197706696726e159 +size 336484 diff --git a/results/idrid/retfound/log.txt b/results/idrid/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..22991e7477682915e7a5e50c13b109d49bdb0fca --- /dev/null +++ b/results/idrid/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.777777777777778e-05, "train_loss": 1.6020541720920138, "epoch": 0, "n_parameters": 303306757} +{"train_lr": 9.027777777777779e-05, "train_loss": 1.5425075954861112, "epoch": 1, "n_parameters": 303306757} +{"train_lr": 0.00015277777777777777, "train_loss": 1.469145033094618, "epoch": 2, "n_parameters": 303306757} +{"train_lr": 0.00021527777777777778, "train_loss": 1.4359130859375, "epoch": 3, "n_parameters": 303306757} +{"train_lr": 0.0002777777777777778, "train_loss": 1.4466467963324652, "epoch": 4, "n_parameters": 303306757} +{"train_lr": 0.0003402777777777778, "train_loss": 1.3976016574435763, "epoch": 5, "n_parameters": 303306757} +{"train_lr": 0.0004027777777777778, "train_loss": 1.31461673312717, "epoch": 6, "n_parameters": 303306757} +{"train_lr": 0.0004652777777777778, "train_loss": 1.3565527598063152, "epoch": 7, "n_parameters": 303306757} +{"train_lr": 0.0005277777777777777, "train_loss": 1.2889660729302301, "epoch": 8, "n_parameters": 303306757} +{"train_lr": 0.0005902777777777778, "train_loss": 1.3096612294514973, "epoch": 9, "n_parameters": 303306757} +{"train_lr": 0.0006247307916747557, "train_loss": 1.269482930501302, "epoch": 10, "n_parameters": 303306757} +{"train_lr": 0.0006229157334469918, "train_loss": 1.2491995493570964, "epoch": 11, "n_parameters": 303306757} +{"train_lr": 0.0006191899416650879, "train_loss": 1.2184555265638564, "epoch": 12, "n_parameters": 303306757} +{"train_lr": 0.0006135763870743313, "train_loss": 1.2288381788465712, "epoch": 13, "n_parameters": 303306757} +{"train_lr": 0.0006061096791054699, "train_loss": 1.1407895618014865, "epoch": 14, "n_parameters": 303306757} +{"train_lr": 0.0005968358524960642, "train_loss": 1.148459964328342, "epoch": 15, "n_parameters": 303306757} +{"train_lr": 0.0005858120834710217, "train_loss": 1.1219838460286458, "epoch": 16, "n_parameters": 303306757} +{"train_lr": 0.0005731063372321567, "train_loss": 1.1042601267496746, "epoch": 17, "n_parameters": 303306757} +{"train_lr": 0.0005587969489301213, "train_loss": 1.0724815792507596, "epoch": 18, "n_parameters": 303306757} +{"train_lr": 0.0005429721407021519, "train_loss": 1.1397572623358831, "epoch": 19, "n_parameters": 303306757} +{"train_lr": 0.0005257294777532569, "train_loss": 1.0966654883490667, "epoch": 20, "n_parameters": 303306757} +{"train_lr": 0.0005071752668342834, "train_loss": 1.1608192655775282, "epoch": 21, "n_parameters": 303306757} +{"train_lr": 0.00048742390082544794, "train_loss": 1.103729936811659, "epoch": 22, "n_parameters": 303306757} +{"train_lr": 0.0004665971534661909, "train_loss": 1.0378839174906414, "epoch": 23, "n_parameters": 303306757} +{"train_lr": 0.0004448234285795783, "train_loss": 1.0426340632968478, "epoch": 24, "n_parameters": 303306757} +{"train_lr": 0.0004222369684200342, "train_loss": 1.0116879145304363, "epoch": 25, "n_parameters": 303306757} +{"train_lr": 0.00039897702602520257, "train_loss": 1.020840088526408, "epoch": 26, "n_parameters": 303306757} +{"train_lr": 0.0003751870066746631, "train_loss": 1.0474363035625882, "epoch": 27, "n_parameters": 303306757} +{"train_lr": 0.00035101358374869636, "train_loss": 1.0168669488694932, "epoch": 28, "n_parameters": 303306757} +{"train_lr": 0.0003266057944381195, "train_loss": 1.0094067255655925, "epoch": 29, "n_parameters": 303306757} +{"train_lr": 0.0003021141208804459, "train_loss": 1.0227819681167603, "epoch": 30, "n_parameters": 303306757} +{"train_lr": 0.0002776895623874664, "train_loss": 1.0181847943200006, "epoch": 31, "n_parameters": 303306757} +{"train_lr": 0.0002534827044842803, "train_loss": 1.0049101776546903, "epoch": 32, "n_parameters": 303306757} +{"train_lr": 0.00022964279049945762, "train_loss": 0.9673961533440484, "epoch": 33, "n_parameters": 303306757} +{"train_lr": 0.00020631680143029074, "train_loss": 1.0027337074279785, "epoch": 34, "n_parameters": 303306757} +{"train_lr": 0.00018364854975606965, "train_loss": 0.9875240590837266, "epoch": 35, "n_parameters": 303306757} +{"train_lr": 0.00016177779278632443, "train_loss": 1.0082972049713135, "epoch": 36, "n_parameters": 303306757} +{"train_lr": 0.00014083937101053823, "train_loss": 1.0269517368740506, "epoch": 37, "n_parameters": 303306757} +{"train_lr": 0.00012096237676169061, "train_loss": 0.9670819309022691, "epoch": 38, "n_parameters": 303306757} +{"train_lr": 0.00010226935831909702, "train_loss": 0.9489992989434136, "epoch": 39, "n_parameters": 303306757} +{"train_lr": 8.48755643575155e-05, "train_loss": 0.963928673002455, "epoch": 40, "n_parameters": 303306757} +{"train_lr": 6.888823340074255e-05, "train_loss": 0.982821438047621, "epoch": 41, "n_parameters": 303306757} +{"train_lr": 5.4405932660453715e-05, "train_loss": 0.9591609372033013, "epoch": 42, "n_parameters": 303306757} +{"train_lr": 4.1517950336566605e-05, "train_loss": 0.9107767475975884, "epoch": 43, "n_parameters": 303306757} +{"train_lr": 3.0303745125797026e-05, "train_loss": 0.9499408668941922, "epoch": 44, "n_parameters": 303306757} +{"train_lr": 2.0832456332370918e-05, "train_loss": 0.9128559960259331, "epoch": 45, "n_parameters": 303306757} +{"train_lr": 1.3162477601222926e-05, "train_loss": 0.9679558277130127, "epoch": 46, "n_parameters": 303306757} +{"train_lr": 7.341096901757339e-06, "train_loss": 0.9742760393354628, "epoch": 47, "n_parameters": 303306757} +{"train_lr": 3.404204981791873e-06, "train_loss": 0.9302464061313205, "epoch": 48, "n_parameters": 303306757} +{"train_lr": 1.3760740891625011e-06, "train_loss": 0.9742009374830458, "epoch": 49, "n_parameters": 303306757} diff --git a/results/idrid/retfound/metrics.json b/results/idrid/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..c686fec40cd9737201492b39069a41ff8d66a9d3 --- /dev/null +++ b/results/idrid/retfound/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 92, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.6847826086956522, + "balanced_accuracy": 0.5656838417749233, + "precision_macro": 0.54, + "recall_macro": 0.5656838417749233, + "f1_macro": 0.5510162668523325, + "precision_weighted": 0.6492753623188405, + "recall_weighted": 0.6847826086956522, + "f1_weighted": 0.6649176612719022, + "cohen_kappa": 0.5736657078938958, + "quadratic_weighted_kappa": 0.8781603660004815, + "mcc": 0.5761521015650642, + "auroc_macro_ovr": 0.9069453534022897, + "auroc_weighted_ovr": 0.9040685958780689, + "auprc_macro": 0.6864908963225116, + "auroc_per_class": { + "0": 0.9889277389277389, + "1": 0.9287356321839081, + "2": 0.8595980962453729, + "3": 0.8309803921568627, + "4": 0.9264849074975657 + }, + "per_class": { + "0": { + "precision": 0.8333333333333334, + "recall": 0.9615384615384616, + "f1-score": 0.8928571428571429, + "support": 26.0 + }, + "1": { + "precision": 0.0, + "recall": 0.0, + "f1-score": 0.0, + "support": 5.0 + }, + "2": { + "precision": 0.6666666666666666, + "recall": 0.6451612903225806, + "f1-score": 0.6557377049180327, + "support": 31.0 + }, + "3": { + "precision": 0.45, + "recall": 0.5294117647058824, + "f1-score": 0.4864864864864865, + "support": 17.0 + }, + "4": { + "precision": 0.75, + "recall": 0.6923076923076923, + "f1-score": 0.72, + "support": 13.0 + }, + "accuracy": 0.6847826086956522, + "macro avg": { + "precision": 0.54, + "recall": 0.5656838417749233, + "f1-score": 0.5510162668523325, + "support": 92.0 + }, + "weighted avg": { + "precision": 0.6492753623188405, + "recall": 0.6847826086956522, + "f1-score": 0.6649176612719022, + "support": 92.0 + } + } +} \ No newline at end of file diff --git a/results/idrid/retfound/metrics_test.csv b/results/idrid/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..2da1c00271b4f1ae66307f18d78b844c75ed9f50 --- /dev/null +++ b/results/idrid/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.7299975752830505,0.6847826086956522,0.5510162668523325,0.9069453534022897,0.12608695652173912,0.4356370124761155,0.54,0.5656838417749233,0.6864908963225116,0.5736657078938958 diff --git a/results/idrid/retfound/metrics_val.csv b/results/idrid/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..faa3cf1106e7d2bd9539eded8ee32e48fc40ad1a --- /dev/null +++ b/results/idrid/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +1.5962324738502502,0.4222222222222222,0.19818181818181818,0.523056266548147,0.2311111111111111,0.1338709677419355,0.1695852534562212,0.275,0.30564967092659157,0.16309012875536466 +1.5760169625282288,0.4222222222222222,0.19603174603174603,0.5414418970278032,0.2311111111111111,0.13031358885017422,0.23500000000000001,0.24903846153846154,0.38832919319319553,0.11363636363636365 +1.5724135041236877,0.35555555555555557,0.10491803278688525,0.5951560437125794,0.2577777777777778,0.07111111111111111,0.07111111111111111,0.2,0.43567289732665965,0.0 +1.5160839557647705,0.5111111111111111,0.25476190476190474,0.6697717256458235,0.19555555555555557,0.19096774193548388,0.21093117408906883,0.32211538461538464,0.44177869970108147,0.27312775330396477 +1.381072998046875,0.4888888888888889,0.24893435635123615,0.6593600941636227,0.20444444444444446,0.18234234234234234,0.24000000000000005,0.2980769230769231,0.4566871574041945,0.2247191011235954 +1.2845553159713745,0.5777777777777777,0.29316770186335406,0.7089456101434851,0.1688888888888889,0.2375,0.2533333333333333,0.3596153846153846,0.49321136343047733,0.3666666666666667 +1.2352371215820312,0.4222222222222222,0.2858937885253675,0.7532491666763921,0.2311111111111111,0.18834586466165412,0.3079365079365079,0.3560897435897436,0.450938870728893,0.22155688622754488 +1.0971596240997314,0.5777777777777777,0.4200330966109367,0.852719664177162,0.1688888888888889,0.30023735810113517,0.5784461152882205,0.46794871794871795,0.5524616139502413,0.437869822485207 +1.2994247078895569,0.4666666666666667,0.23496503496503496,0.7961682315681915,0.21333333333333335,0.16701754385964912,0.23333333333333334,0.2826923076923077,0.5186548526980037,0.18918918918918926 +1.1164644956588745,0.5333333333333333,0.3665656565656566,0.8026583423686471,0.18666666666666668,0.2670708601743085,0.35277777777777775,0.4272435897435898,0.5277705864818487,0.34917355371900827 +0.9869260191917419,0.6222222222222222,0.46764891464850483,0.8678345539111376,0.1511111111111111,0.34689075630252103,0.4893939393939394,0.47628205128205137,0.5871291833833923,0.47994561522773616 +1.019225001335144,0.5333333333333333,0.36756598240469207,0.8496055664108993,0.18666666666666668,0.2596491228070176,0.4066666666666666,0.380448717948718,0.588907489849227,0.3359100491918482 +1.0121880173683167,0.6,0.3944061302681992,0.8833098095628168,0.16,0.30190374331550807,0.3488888888888889,0.455448717948718,0.6597358831634901,0.43866943866943864 +1.0137562155723572,0.5777777777777777,0.45044065804935374,0.8584718362184279,0.1688888888888889,0.326007326007326,0.4597619047619048,0.4509615384615385,0.5651624602871923,0.41398217957505135 +1.0793166756629944,0.6444444444444445,0.4378804657874425,0.8816808631313442,0.14222222222222222,0.3343295019157088,0.6084656084656085,0.442948717948718,0.6211477152283297,0.4786386676321507 +0.8901780843734741,0.6444444444444445,0.46669459572685373,0.8832715692529245,0.14222222222222222,0.3535714285714286,0.5790476190476191,0.492948717948718,0.6205487339614885,0.5108695652173912 +1.0445513129234314,0.6,0.3590131763386599,0.8968956600021155,0.16,0.27403361344537813,0.354,0.39711538461538465,0.7050772463954769,0.4151624548736461 +0.9862746000289917,0.6222222222222222,0.4489004329004329,0.8823809702769205,0.1511111111111111,0.3440202517788725,0.5033333333333333,0.4567307692307693,0.601745947595339,0.45628997867803833 +1.1765361428260803,0.5555555555555556,0.3135688684075781,0.8747199400066282,0.17777777777777778,0.23131578947368422,0.4256410256410256,0.3596153846153846,0.6156634370388001,0.3416239941477689 +0.8740264773368835,0.6444444444444445,0.5011965811965812,0.8901614093274077,0.14222222222222222,0.38263736263736264,0.48571428571428565,0.5221153846153846,0.7066553809394366,0.5190380761523046 +1.1071023344993591,0.5777777777777777,0.325374677002584,0.8884788345343679,0.1688888888888889,0.24500000000000002,0.4374727668845316,0.3721153846153846,0.6420901804428536,0.37316715542521994 +0.8539901971817017,0.6666666666666666,0.509683908045977,0.9037370967753887,0.13333333333333333,0.3853323147440795,0.5017857142857143,0.5221153846153846,0.6600408609235564,0.540503744043567 +0.9264477491378784,0.6666666666666666,0.47434077079107506,0.8961589948007351,0.13333333333333333,0.3621721415839063,0.49444444444444446,0.492948717948718,0.6930423243294316,0.5328719723183392 +1.0080241560935974,0.6444444444444445,0.45599343185550084,0.9042704280525051,0.14222222222222222,0.3411764705882353,0.5700000000000001,0.46794871794871795,0.6959003746561051,0.4954449894884373 +0.9226003885269165,0.6666666666666666,0.4872072072072072,0.9079454759336475,0.13333333333333333,0.37547619047619046,0.5142857142857143,0.4887820512820514,0.7014067033019461,0.528960223307746 +0.8886241912841797,0.7111111111111111,0.5431989063568011,0.9174946401740709,0.11555555555555555,0.42326007326007326,0.6102673796791444,0.5512820512820513,0.7449263750179045,0.5987654320987654 +1.0057958364486694,0.6666666666666666,0.4792063492063491,0.9141340811850034,0.13333333333333333,0.36962962962962964,0.613095238095238,0.48044871794871796,0.7427160117988958,0.5195729537366548 +0.8995458483695984,0.7111111111111111,0.5376470588235295,0.9186248555785443,0.11555555555555555,0.41818181818181815,0.6079532163742691,0.5387820512820514,0.7440768835767567,0.5943134535367545 +0.9491428732872009,0.6444444444444445,0.45008547008547006,0.9144781455627486,0.14222222222222222,0.3448677248677249,0.49976190476190474,0.455448717948718,0.7355457187074947,0.488272921108742 +0.8871878385543823,0.6888888888888889,0.5123076923076924,0.9142366940532538,0.12444444444444444,0.39484848484848484,0.5325563909774436,0.5221153846153846,0.7072094658813407,0.5661157024793388 +0.8569319844245911,0.7333333333333333,0.566793990323402,0.9102930157811875,0.10666666666666667,0.4488095238095238,0.5673015873015873,0.5721153846153847,0.7155749219232082,0.6291208791208791 +1.0129761695861816,0.6888888888888889,0.4987468671679197,0.9037745243238426,0.12444444444444444,0.38733333333333336,0.6210389610389611,0.5054487179487179,0.7156892674444169,0.5550847457627119 +0.9665532112121582,0.6666666666666666,0.48946725788831047,0.9035023024998967,0.13333333333333333,0.37114285714285716,0.5315151515151515,0.4887820512820514,0.7053677186523897,0.5246478873239437 +0.9573783278465271,0.6666666666666666,0.46831831831831827,0.9047806697265397,0.13333333333333333,0.36357142857142855,0.49523809523809526,0.48044871794871796,0.7154918866138613,0.5283018867924527 +0.9578981995582581,0.6888888888888889,0.5111111111111111,0.9079535838657732,0.12444444444444444,0.3968530020703934,0.5347619047619048,0.5137820512820513,0.7304216729223241,0.5628036086051353 +1.079646646976471,0.6666666666666666,0.47492063492063485,0.9038468647229674,0.13333333333333333,0.3660606060606061,0.5987577639751553,0.48044871794871796,0.6637248129496021,0.5222929936305734 +1.0088809728622437,0.6888888888888889,0.49310823428470485,0.9060105502918254,0.12444444444444444,0.38500000000000006,0.6063492063492063,0.5054487179487179,0.6714624110444467,0.5591322603219034 +0.9201102256774902,0.6888888888888889,0.5260317460317461,0.9157642293919359,0.12444444444444444,0.40793756967670014,0.5977443609022556,0.5262820512820514,0.7482638847738905,0.5646164478230822 +1.0064192414283752,0.6444444444444445,0.45405759552101016,0.9102644857707005,0.14222222222222222,0.3461904761904762,0.5954285714285714,0.455448717948718,0.7143004273951188,0.4846098783106657 +1.072479784488678,0.6222222222222222,0.4253968253968254,0.9179605203562702,0.1511111111111111,0.32298850574712645,0.5714285714285714,0.430448717948718,0.7404309477569665,0.44924406047516197 +1.0099626779556274,0.6666666666666666,0.5000854700854701,0.9140506492551401,0.13333333333333333,0.38296296296296295,0.5997619047619047,0.4887820512820514,0.7271873112872271,0.5202558635394456 +1.0447663068771362,0.6444444444444445,0.45405759552101016,0.911636447918525,0.14222222222222222,0.3461904761904762,0.5954285714285714,0.455448717948718,0.7170068789028473,0.4846098783106657 +1.0410215854644775,0.6444444444444445,0.45087912087912085,0.9120966063528213,0.14222222222222222,0.3396296296296296,0.5883333333333334,0.455448717948718,0.7182568789028474,0.48571428571428565 +1.0116748213768005,0.6,0.3854151268785415,0.9101193140738048,0.16,0.2978571428571429,0.40876190476190477,0.40544871794871795,0.7023687591129057,0.4218415417558886 +0.992044985294342,0.6444444444444445,0.4561583445793972,0.9114706654251563,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7076572206513673,0.4954449894884373 +0.9809765815734863,0.6444444444444445,0.4561583445793972,0.9114209306771457,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7134112733689455,0.4954449894884373 +0.9713383913040161,0.6444444444444445,0.4561583445793972,0.9120966063528213,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7147294727134607,0.4954449894884373 +0.9727200269699097,0.6444444444444445,0.4561583445793972,0.9116655718700628,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7140046597619643,0.4954449894884373 +0.9715694189071655,0.6444444444444445,0.4561583445793972,0.9116655718700628,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7140046597619643,0.4954449894884373 +0.9706709980964661,0.6444444444444445,0.4561583445793972,0.9116655718700628,0.14222222222222222,0.3503636363636364,0.4696103896103896,0.4637820512820513,0.7140046597619643,0.4954449894884373 diff --git a/results/idrid/retfound/pr.png b/results/idrid/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..eb4064e173e350a03c25974a7e1d2b1dd27afbc1 --- /dev/null +++ b/results/idrid/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6b3e02f43752d6b05eeba1578242def3642ced3cdf20e1950f035f588306d32 +size 90720 diff --git a/results/idrid/retfound/roc.png b/results/idrid/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..cf9780adf09b496c19648dbd9080d94823194eff --- /dev/null +++ b/results/idrid/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46286dcbfa30913c302bdbf49b92e4e047c89b732c8c89720a27842cf8f577aa +size 82799 diff --git a/results/idrid/retfound/test_pred.npz b/results/idrid/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..abfe0c2177da3c977f7f2be9399eba287474ba93 --- /dev/null +++ b/results/idrid/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:863f2fe0492c5217f7d12f253ec63b083d9937d7a2ef28ee4d8b6ccc026625e7 +size 2166 diff --git a/results/idrid/retfound/train.log b/results/idrid/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..946188026d45ee8f3bfbf97d9dafac2e3d665e8f --- /dev/null +++ b/results/idrid/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:57:00.453645923 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:57:01.603350] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:57:01.603571] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/DR/idrid-dataset', +nb_classes=5, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/idrid', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:57:04.579076] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:57:06.334251] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:57:06.645401] Sampler_train = +[13:57:06.689046] len of train_set: 288 +[13:57:07.382010] [Adaptation] Full fine-tuning: training all parameters. +[13:57:07.383064] number of trainable params (M): 303.31 +[13:57:07.383146] base lr: 5.00e-03 +[13:57:07.383208] actual lr: 6.25e-04 +[13:57:07.383268] accumulate grad iterations: 1 +[13:57:07.383330] effective batch size: 32 +[13:57:07.385983] criterion = CrossEntropyLoss() +[13:57:07.386082] Start training for 50 epochs +[13:57:07.387775] log_dir: ./output_logs/retfound +[13:57:11.931751] Epoch: [0] [0/9] eta: 0:00:40 lr: 0.000000 loss: 1.6095 (1.6095) time: 4.5431 data: 3.7168 max mem: 7340 +[13:57:13.093303] Epoch: [0] [8/9] eta: 0:00:00 lr: 0.000056 loss: 1.6056 (1.6021) time: 0.6338 data: 0.4131 max mem: 9672 +[13:57:13.174878] Epoch: [0] Total time: 0:00:05 (0.6430 s / it) +[13:57:13.183982] Averaged stats: lr: 0.000056 loss: 1.6056 (1.6021) +[13:57:18.192457] val: [0/2] eta: 0:00:09 loss: 1.5551 (1.5551) time: 4.9947 data: 4.9520 max mem: 9672 +[13:57:18.287243] val: [1/2] eta: 0:00:02 loss: 1.5551 (1.5962) time: 2.5444 data: 2.4761 max mem: 9672 +[13:57:18.367653] val: Total time: 0:00:05 (2.5852 s / it) +[13:57:18.380355] val loss: 1.5962324738502502 +[13:57:18.380678] Accuracy: 0.4222, F1 Score: 0.1982, ROC AUC: 0.5231, Hamming Loss: 0.2311, + Jaccard Score: 0.1339, Precision: 0.1696, Recall: 0.2750, + Average Precision: 0.3056, Kappa: 0.1631, Score: 0.2948 +[13:57:20.219294] Best epoch = 0, Best score = 0.2948 +[13:57:20.297552] log_dir: ./output_logs/retfound +[13:57:24.329539] Epoch: [1] [0/9] eta: 0:00:36 lr: 0.000063 loss: 1.5881 (1.5881) time: 4.0311 data: 3.8962 max mem: 9672 +[13:57:25.469443] Epoch: [1] [8/9] eta: 0:00:00 lr: 0.000118 loss: 1.5361 (1.5425) time: 0.5745 data: 0.4330 max mem: 9672 +[13:57:25.556231] Epoch: [1] Total time: 0:00:05 (0.5843 s / it) +[13:57:25.564325] Averaged stats: lr: 0.000118 loss: 1.5361 (1.5425) +[13:57:30.718116] val: [0/2] eta: 0:00:10 loss: 1.3929 (1.3929) time: 5.1418 data: 5.1053 max mem: 9672 +[13:57:30.735686] val: [1/2] eta: 0:00:02 loss: 1.3929 (1.5760) time: 2.5794 data: 2.5527 max mem: 9672 +[13:57:30.815623] val: Total time: 0:00:05 (2.6199 s / it) +[13:57:30.826812] val loss: 1.5760169625282288 +[13:57:30.827005] Accuracy: 0.4222, F1 Score: 0.1960, ROC AUC: 0.5414, Hamming Loss: 0.2311, + Jaccard Score: 0.1303, Precision: 0.2350, Recall: 0.2490, + Average Precision: 0.3883, Kappa: 0.1136, Score: 0.2837 +[13:57:30.865306] Best epoch = 0, Best score = 0.2948 +[13:57:31.132456] log_dir: ./output_logs/retfound +[13:57:34.848730] Epoch: [2] [0/9] eta: 0:00:33 lr: 0.000125 loss: 1.5021 (1.5021) time: 3.7152 data: 3.5698 max mem: 9672 +[13:57:35.982220] Epoch: [2] [8/9] eta: 0:00:00 lr: 0.000181 loss: 1.4673 (1.4691) time: 0.5387 data: 0.3967 max mem: 9672 +[13:57:36.066745] Epoch: [2] Total time: 0:00:04 (0.5482 s / it) +[13:57:36.075210] Averaged stats: lr: 0.000181 loss: 1.4673 (1.4691) +[13:57:41.076936] val: [0/2] eta: 0:00:09 loss: 1.2487 (1.2487) time: 4.9862 data: 4.9510 max mem: 9672 +[13:57:41.094263] val: [1/2] eta: 0:00:02 loss: 1.2487 (1.5724) time: 2.5014 data: 2.4755 max mem: 9672 +[13:57:41.169577] val: Total time: 0:00:05 (2.5397 s / it) +[13:57:41.180095] val loss: 1.5724135041236877 +[13:57:41.180338] Accuracy: 0.3556, F1 Score: 0.1049, ROC AUC: 0.5952, Hamming Loss: 0.2578, + Jaccard Score: 0.0711, Precision: 0.0711, Recall: 0.2000, + Average Precision: 0.4357, Kappa: 0.0000, Score: 0.2334 +[13:57:41.221351] Best epoch = 0, Best score = 0.2948 +[13:57:41.491021] log_dir: ./output_logs/retfound +[13:57:45.003522] Epoch: [3] [0/9] eta: 0:00:31 lr: 0.000188 loss: 1.4260 (1.4260) time: 3.5114 data: 3.3662 max mem: 9672 +[13:57:46.132697] Epoch: [3] [8/9] eta: 0:00:00 lr: 0.000243 loss: 1.4174 (1.4359) time: 0.5155 data: 0.3741 max mem: 9672 +[13:57:46.217643] Epoch: [3] Total time: 0:00:04 (0.5252 s / it) +[13:57:46.225548] Averaged stats: lr: 0.000243 loss: 1.4174 (1.4359) +[13:57:51.130222] val: [0/2] eta: 0:00:09 loss: 1.2569 (1.2569) time: 4.8928 data: 4.8587 max mem: 9672 +[13:57:51.147711] val: [1/2] eta: 0:00:02 loss: 1.2569 (1.5161) time: 2.4549 data: 2.4294 max mem: 9672 +[13:57:51.229530] val: Total time: 0:00:04 (2.4964 s / it) +[13:57:51.240271] val loss: 1.5160839557647705 +[13:57:51.240539] Accuracy: 0.5111, F1 Score: 0.2548, ROC AUC: 0.6698, Hamming Loss: 0.1956, + Jaccard Score: 0.1910, Precision: 0.2109, Recall: 0.3221, + Average Precision: 0.4418, Kappa: 0.2731, Score: 0.3992 +[13:57:53.040232] Best epoch = 3, Best score = 0.3992 +[13:57:53.103470] log_dir: ./output_logs/retfound +[13:57:56.866905] Epoch: [4] [0/9] eta: 0:00:33 lr: 0.000250 loss: 1.4665 (1.4665) time: 3.7620 data: 3.6166 max mem: 9672 +[13:57:58.008585] Epoch: [4] [8/9] eta: 0:00:00 lr: 0.000306 loss: 1.4532 (1.4466) time: 0.5448 data: 0.4019 max mem: 9672 +[13:57:58.103258] Epoch: [4] Total time: 0:00:04 (0.5555 s / it) +[13:57:58.111913] Averaged stats: lr: 0.000306 loss: 1.4532 (1.4466) +[13:58:03.222744] val: [0/2] eta: 0:00:10 loss: 1.3120 (1.3120) time: 5.0988 data: 5.0640 max mem: 9672 +[13:58:03.240247] val: [1/2] eta: 0:00:02 loss: 1.3120 (1.3811) time: 2.5578 data: 2.5321 max mem: 9672 +[13:58:03.313836] val: Total time: 0:00:05 (2.5952 s / it) +[13:58:03.324394] val loss: 1.381072998046875 +[13:58:03.324591] Accuracy: 0.4889, F1 Score: 0.2489, ROC AUC: 0.6594, Hamming Loss: 0.2044, + Jaccard Score: 0.1823, Precision: 0.2400, Recall: 0.2981, + Average Precision: 0.4567, Kappa: 0.2247, Score: 0.3777 +[13:58:03.362606] Best epoch = 3, Best score = 0.3992 +[13:58:03.607830] log_dir: ./output_logs/retfound +[13:58:07.195283] Epoch: [5] [0/9] eta: 0:00:32 lr: 0.000313 loss: 1.5345 (1.5345) time: 3.5863 data: 3.4407 max mem: 9672 +[13:58:08.480139] Epoch: [5] [8/9] eta: 0:00:00 lr: 0.000368 loss: 1.4093 (1.3976) time: 0.5411 data: 0.3976 max mem: 9672 +[13:58:08.559560] Epoch: [5] Total time: 0:00:04 (0.5502 s / it) +[13:58:08.568662] Averaged stats: lr: 0.000368 loss: 1.4093 (1.3976) +[13:58:13.559434] val: [0/2] eta: 0:00:09 loss: 1.0430 (1.0430) time: 4.9751 data: 4.9400 max mem: 9672 +[13:58:13.577043] val: [1/2] eta: 0:00:02 loss: 1.0430 (1.2846) time: 2.4961 data: 2.4701 max mem: 9672 +[13:58:13.662147] val: Total time: 0:00:05 (2.5392 s / it) +[13:58:13.673004] val loss: 1.2845553159713745 +[13:58:13.673206] Accuracy: 0.5778, F1 Score: 0.2932, ROC AUC: 0.7089, Hamming Loss: 0.1689, + Jaccard Score: 0.2375, Precision: 0.2533, Recall: 0.3596, + Average Precision: 0.4932, Kappa: 0.3667, Score: 0.4563 +[13:58:15.409457] Best epoch = 5, Best score = 0.4563 +[13:58:15.473850] log_dir: ./output_logs/retfound +[13:58:19.201203] Epoch: [6] [0/9] eta: 0:00:33 lr: 0.000375 loss: 1.3983 (1.3983) time: 3.7263 data: 3.5786 max mem: 9672 +[13:58:20.342283] Epoch: [6] [8/9] eta: 0:00:00 lr: 0.000431 loss: 1.3523 (1.3146) time: 0.5407 data: 0.3977 max mem: 9672 +[13:58:20.423851] Epoch: [6] Total time: 0:00:04 (0.5500 s / it) +[13:58:20.433060] Averaged stats: lr: 0.000431 loss: 1.3523 (1.3146) +[13:58:25.416027] val: [0/2] eta: 0:00:09 loss: 1.3030 (1.3030) time: 4.9694 data: 4.9331 max mem: 9672 +[13:58:25.433501] val: [1/2] eta: 0:00:02 loss: 1.1675 (1.2352) time: 2.4931 data: 2.4666 max mem: 9672 +[13:58:25.516544] val: Total time: 0:00:05 (2.5352 s / it) +[13:58:25.527262] val loss: 1.2352371215820312 +[13:58:25.527486] Accuracy: 0.4222, F1 Score: 0.2859, ROC AUC: 0.7532, Hamming Loss: 0.2311, + Jaccard Score: 0.1883, Precision: 0.3079, Recall: 0.3561, + Average Precision: 0.4509, Kappa: 0.2216, Score: 0.4202 +[13:58:25.570914] Best epoch = 5, Best score = 0.4563 +[13:58:25.812659] log_dir: ./output_logs/retfound +[13:58:29.466946] Epoch: [7] [0/9] eta: 0:00:32 lr: 0.000438 loss: 1.4041 (1.4041) time: 3.6526 data: 3.5032 max mem: 9672 +[13:58:30.610373] Epoch: [7] [8/9] eta: 0:00:00 lr: 0.000493 loss: 1.3267 (1.3566) time: 0.5328 data: 0.3894 max mem: 9672 +[13:58:30.694906] Epoch: [7] Total time: 0:00:04 (0.5425 s / it) +[13:58:30.703243] Averaged stats: lr: 0.000493 loss: 1.3267 (1.3566) +[13:58:35.597418] val: [0/2] eta: 0:00:09 loss: 1.0661 (1.0661) time: 4.8868 data: 4.8505 max mem: 9672 +[13:58:35.615007] val: [1/2] eta: 0:00:02 loss: 1.0661 (1.0972) time: 2.4519 data: 2.4253 max mem: 9672 +[13:58:35.690927] val: Total time: 0:00:04 (2.4904 s / it) +[13:58:35.701721] val loss: 1.0971596240997314 +[13:58:35.701909] Accuracy: 0.5778, F1 Score: 0.4200, ROC AUC: 0.8527, Hamming Loss: 0.1689, + Jaccard Score: 0.3002, Precision: 0.5784, Recall: 0.4679, + Average Precision: 0.5525, Kappa: 0.4379, Score: 0.5702 +[13:58:37.731876] Best epoch = 7, Best score = 0.5702 +[13:58:37.798784] log_dir: ./output_logs/retfound +[13:58:41.431383] Epoch: [8] [0/9] eta: 0:00:32 lr: 0.000500 loss: 1.2979 (1.2979) time: 3.6312 data: 3.4855 max mem: 9672 +[13:58:42.570910] Epoch: [8] [8/9] eta: 0:00:00 lr: 0.000556 loss: 1.3193 (1.2890) time: 0.5300 data: 0.3874 max mem: 9672 +[13:58:42.653481] Epoch: [8] Total time: 0:00:04 (0.5394 s / it) +[13:58:42.661544] Averaged stats: lr: 0.000556 loss: 1.3193 (1.2890) +[13:58:47.578714] val: [0/2] eta: 0:00:09 loss: 0.9970 (0.9970) time: 4.9044 data: 4.8713 max mem: 9672 +[13:58:47.596509] val: [1/2] eta: 0:00:02 loss: 0.9970 (1.2994) time: 2.4608 data: 2.4357 max mem: 9672 +[13:58:47.675551] val: Total time: 0:00:05 (2.5009 s / it) +[13:58:47.686062] val loss: 1.2994247078895569 +[13:58:47.686248] Accuracy: 0.4667, F1 Score: 0.2350, ROC AUC: 0.7962, Hamming Loss: 0.2133, + Jaccard Score: 0.1670, Precision: 0.2333, Recall: 0.2827, + Average Precision: 0.5187, Kappa: 0.1892, Score: 0.4068 +[13:58:47.727124] Best epoch = 7, Best score = 0.5702 +[13:58:47.990347] log_dir: ./output_logs/retfound +[13:58:51.650234] Epoch: [9] [0/9] eta: 0:00:32 lr: 0.000562 loss: 1.7064 (1.7064) time: 3.6588 data: 3.5127 max mem: 9672 +[13:58:52.835256] Epoch: [9] [8/9] eta: 0:00:00 lr: 0.000618 loss: 1.2605 (1.3097) time: 0.5381 data: 0.3947 max mem: 9672 +[13:58:52.915645] Epoch: [9] Total time: 0:00:04 (0.5472 s / it) +[13:58:52.925329] Averaged stats: lr: 0.000618 loss: 1.2605 (1.3097) +[13:58:57.885946] val: [0/2] eta: 0:00:09 loss: 1.1316 (1.1316) time: 4.9444 data: 4.9061 max mem: 9672 +[13:58:57.903247] val: [1/2] eta: 0:00:02 loss: 1.1013 (1.1165) time: 2.4805 data: 2.4531 max mem: 9672 +[13:58:57.980202] val: Total time: 0:00:05 (2.5196 s / it) +[13:58:57.990787] val loss: 1.1164644956588745 +[13:58:57.990968] Accuracy: 0.5333, F1 Score: 0.3666, ROC AUC: 0.8027, Hamming Loss: 0.1867, + Jaccard Score: 0.2671, Precision: 0.3528, Recall: 0.4272, + Average Precision: 0.5278, Kappa: 0.3492, Score: 0.5061 +[13:58:58.027340] Best epoch = 7, Best score = 0.5702 +[13:58:58.288454] log_dir: ./output_logs/retfound +[13:59:02.020539] Epoch: [10] [0/9] eta: 0:00:33 lr: 0.000625 loss: 1.2122 (1.2122) time: 3.7311 data: 3.5855 max mem: 9672 +[13:59:03.159109] Epoch: [10] [8/9] eta: 0:00:00 lr: 0.000624 loss: 1.2122 (1.2695) time: 0.5409 data: 0.3984 max mem: 9672 +[13:59:03.240999] Epoch: [10] Total time: 0:00:04 (0.5503 s / it) +[13:59:03.250031] Averaged stats: lr: 0.000624 loss: 1.2122 (1.2695) +[13:59:08.029094] val: [0/2] eta: 0:00:09 loss: 0.8646 (0.8646) time: 4.7574 data: 4.7211 max mem: 9672 +[13:59:08.046668] val: [1/2] eta: 0:00:02 loss: 0.8646 (0.9869) time: 2.3872 data: 2.3606 max mem: 9672 +[13:59:08.127828] val: Total time: 0:00:04 (2.4284 s / it) +[13:59:08.138437] val loss: 0.9869260191917419 +[13:59:08.138671] Accuracy: 0.6222, F1 Score: 0.4676, ROC AUC: 0.8678, Hamming Loss: 0.1511, + Jaccard Score: 0.3469, Precision: 0.4894, Recall: 0.4763, + Average Precision: 0.5871, Kappa: 0.4799, Score: 0.6051 +[13:59:09.952593] Best epoch = 10, Best score = 0.6051 +[13:59:10.029114] log_dir: ./output_logs/retfound +[13:59:13.851650] Epoch: [11] [0/9] eta: 0:00:34 lr: 0.000624 loss: 1.2263 (1.2263) time: 3.8215 data: 3.6771 max mem: 9672 +[13:59:14.995844] Epoch: [11] [8/9] eta: 0:00:00 lr: 0.000622 loss: 1.2839 (1.2492) time: 0.5517 data: 0.4086 max mem: 9672 +[13:59:15.076819] Epoch: [11] Total time: 0:00:05 (0.5608 s / it) +[13:59:15.085669] Averaged stats: lr: 0.000622 loss: 1.2839 (1.2492) +[13:59:20.162608] val: [0/2] eta: 0:00:10 loss: 0.8131 (0.8131) time: 5.0645 data: 5.0289 max mem: 9672 +[13:59:20.180141] val: [1/2] eta: 0:00:02 loss: 0.8131 (1.0192) time: 2.5407 data: 2.5145 max mem: 9672 +[13:59:20.261634] val: Total time: 0:00:05 (2.5821 s / it) +[13:59:20.273436] val loss: 1.019225001335144 +[13:59:20.273629] Accuracy: 0.5333, F1 Score: 0.3676, ROC AUC: 0.8496, Hamming Loss: 0.1867, + Jaccard Score: 0.2596, Precision: 0.4067, Recall: 0.3804, + Average Precision: 0.5889, Kappa: 0.3359, Score: 0.5177 +[13:59:20.307081] Best epoch = 10, Best score = 0.6051 +[13:59:20.562467] log_dir: ./output_logs/retfound +[13:59:24.037396] Epoch: [12] [0/9] eta: 0:00:31 lr: 0.000621 loss: 1.1053 (1.1053) time: 3.4737 data: 3.3279 max mem: 9672 +[13:59:25.176469] Epoch: [12] [8/9] eta: 0:00:00 lr: 0.000617 loss: 1.2010 (1.2185) time: 0.5124 data: 0.3699 max mem: 9672 +[13:59:25.257995] Epoch: [12] Total time: 0:00:04 (0.5217 s / it) +[13:59:25.267247] Averaged stats: lr: 0.000617 loss: 1.2010 (1.2185) +[13:59:30.257026] val: [0/2] eta: 0:00:09 loss: 0.7289 (0.7289) time: 4.9777 data: 4.9449 max mem: 9672 +[13:59:30.271637] val: [1/2] eta: 0:00:02 loss: 0.7289 (1.0122) time: 2.4958 data: 2.4725 max mem: 9672 +[13:59:30.349226] val: Total time: 0:00:05 (2.5352 s / it) +[13:59:30.360223] val loss: 1.0121880173683167 +[13:59:30.360441] Accuracy: 0.6000, F1 Score: 0.3944, ROC AUC: 0.8833, Hamming Loss: 0.1600, + Jaccard Score: 0.3019, Precision: 0.3489, Recall: 0.4554, + Average Precision: 0.6597, Kappa: 0.4387, Score: 0.5721 +[13:59:30.400487] Best epoch = 10, Best score = 0.6051 +[13:59:30.650826] log_dir: ./output_logs/retfound +[13:59:34.498686] Epoch: [13] [0/9] eta: 0:00:34 lr: 0.000616 loss: 1.2531 (1.2531) time: 3.8463 data: 3.7010 max mem: 9672 +[13:59:35.642076] Epoch: [13] [8/9] eta: 0:00:00 lr: 0.000611 loss: 1.2531 (1.2288) time: 0.5543 data: 0.4113 max mem: 9672 +[13:59:35.723165] Epoch: [13] Total time: 0:00:05 (0.5636 s / it) +[13:59:35.732603] Averaged stats: lr: 0.000611 loss: 1.2531 (1.2288) +[13:59:41.099314] val: [0/2] eta: 0:00:10 loss: 0.9487 (0.9487) time: 5.3539 data: 5.3174 max mem: 9672 +[13:59:41.119299] val: [1/2] eta: 0:00:02 loss: 0.9487 (1.0138) time: 2.6866 data: 2.6588 max mem: 9672 +[13:59:41.220336] val: Total time: 0:00:05 (2.7377 s / it) +[13:59:41.231287] val loss: 1.0137562155723572 +[13:59:41.231499] Accuracy: 0.5778, F1 Score: 0.4504, ROC AUC: 0.8585, Hamming Loss: 0.1689, + Jaccard Score: 0.3260, Precision: 0.4598, Recall: 0.4510, + Average Precision: 0.5652, Kappa: 0.4140, Score: 0.5743 +[13:59:41.283379] Best epoch = 10, Best score = 0.6051 +[13:59:41.533308] log_dir: ./output_logs/retfound +[13:59:45.129195] Epoch: [14] [0/9] eta: 0:00:32 lr: 0.000610 loss: 1.3122 (1.3122) time: 3.5945 data: 3.4485 max mem: 9672 +[13:59:46.340843] Epoch: [14] [8/9] eta: 0:00:00 lr: 0.000602 loss: 1.0947 (1.1408) time: 0.5339 data: 0.3914 max mem: 9672 +[13:59:46.421782] Epoch: [14] Total time: 0:00:04 (0.5431 s / it) +[13:59:46.431682] Averaged stats: lr: 0.000602 loss: 1.0947 (1.1408) +[13:59:51.435320] val: [0/2] eta: 0:00:09 loss: 0.6678 (0.6678) time: 4.9926 data: 4.9586 max mem: 9672 +[13:59:51.455038] val: [1/2] eta: 0:00:02 loss: 0.6678 (1.0793) time: 2.5058 data: 2.4794 max mem: 9672 +[13:59:51.539482] val: Total time: 0:00:05 (2.5487 s / it) +[13:59:51.550194] val loss: 1.0793166756629944 +[13:59:51.550430] Accuracy: 0.6444, F1 Score: 0.4379, ROC AUC: 0.8817, Hamming Loss: 0.1422, + Jaccard Score: 0.3343, Precision: 0.6085, Recall: 0.4429, + Average Precision: 0.6211, Kappa: 0.4786, Score: 0.5994 +[13:59:51.592511] Best epoch = 10, Best score = 0.6051 +[13:59:51.872036] log_dir: ./output_logs/retfound +[13:59:55.743790] Epoch: [15] [0/9] eta: 0:00:34 lr: 0.000601 loss: 0.9294 (0.9294) time: 3.8706 data: 3.7271 max mem: 9672 +[13:59:56.884480] Epoch: [15] [8/9] eta: 0:00:00 lr: 0.000592 loss: 1.1506 (1.1485) time: 0.5566 data: 0.4142 max mem: 9672 +[13:59:56.962833] Epoch: [15] Total time: 0:00:05 (0.5656 s / it) +[13:59:56.971885] Averaged stats: lr: 0.000592 loss: 1.1506 (1.1485) +[14:00:01.654966] val: [0/2] eta: 0:00:09 loss: 0.7557 (0.7557) time: 4.6656 data: 4.6305 max mem: 9672 +[14:00:01.674939] val: [1/2] eta: 0:00:02 loss: 0.7557 (0.8902) time: 2.3425 data: 2.3153 max mem: 9672 +[14:00:01.752532] val: Total time: 0:00:04 (2.3819 s / it) +[14:00:01.764998] val loss: 0.8901780843734741 +[14:00:01.765269] Accuracy: 0.6444, F1 Score: 0.4667, ROC AUC: 0.8833, Hamming Loss: 0.1422, + Jaccard Score: 0.3536, Precision: 0.5790, Recall: 0.4929, + Average Precision: 0.6205, Kappa: 0.5109, Score: 0.6203 +[14:00:03.708226] Best epoch = 15, Best score = 0.6203 +[14:00:03.769711] log_dir: ./output_logs/retfound +[14:00:07.204808] Epoch: [16] [0/9] eta: 0:00:30 lr: 0.000591 loss: 1.1552 (1.1552) time: 3.4339 data: 3.2852 max mem: 9672 +[14:00:08.343580] Epoch: [16] [8/9] eta: 0:00:00 lr: 0.000580 loss: 1.1178 (1.1220) time: 0.5080 data: 0.3652 max mem: 9672 +[14:00:08.424276] Epoch: [16] Total time: 0:00:04 (0.5172 s / it) +[14:00:08.432177] Averaged stats: lr: 0.000580 loss: 1.1178 (1.1220) +[14:00:13.430544] val: [0/2] eta: 0:00:09 loss: 0.6182 (0.6182) time: 4.9867 data: 4.9514 max mem: 9672 +[14:00:13.447915] val: [1/2] eta: 0:00:02 loss: 0.6182 (1.0446) time: 2.5018 data: 2.4758 max mem: 9672 +[14:00:13.524505] val: Total time: 0:00:05 (2.5406 s / it) +[14:00:13.536115] val loss: 1.0445513129234314 +[14:00:13.536324] Accuracy: 0.6000, F1 Score: 0.3590, ROC AUC: 0.8969, Hamming Loss: 0.1600, + Jaccard Score: 0.2740, Precision: 0.3540, Recall: 0.3971, + Average Precision: 0.7051, Kappa: 0.4152, Score: 0.5570 +[14:00:13.579811] Best epoch = 15, Best score = 0.6203 +[14:00:13.822343] log_dir: ./output_logs/retfound +[14:00:17.649958] Epoch: [17] [0/9] eta: 0:00:34 lr: 0.000579 loss: 0.9844 (0.9844) time: 3.8265 data: 3.6813 max mem: 9672 +[14:00:18.790265] Epoch: [17] [8/9] eta: 0:00:00 lr: 0.000567 loss: 1.1257 (1.1043) time: 0.5518 data: 0.4091 max mem: 9672 +[14:00:18.872911] Epoch: [17] Total time: 0:00:05 (0.5612 s / it) +[14:00:18.882784] Averaged stats: lr: 0.000567 loss: 1.1257 (1.1043) +[14:00:23.844814] val: [0/2] eta: 0:00:09 loss: 0.7076 (0.7076) time: 4.9474 data: 4.9124 max mem: 9672 +[14:00:23.864578] val: [1/2] eta: 0:00:02 loss: 0.7076 (0.9863) time: 2.4833 data: 2.4563 max mem: 9672 +[14:00:23.949283] val: Total time: 0:00:05 (2.5262 s / it) +[14:00:23.960003] val loss: 0.9862746000289917 +[14:00:23.960236] Accuracy: 0.6222, F1 Score: 0.4489, ROC AUC: 0.8824, Hamming Loss: 0.1511, + Jaccard Score: 0.3440, Precision: 0.5033, Recall: 0.4567, + Average Precision: 0.6017, Kappa: 0.4563, Score: 0.5959 +[14:00:24.002820] Best epoch = 15, Best score = 0.6203 +[14:00:24.256522] log_dir: ./output_logs/retfound +[14:00:27.943736] Epoch: [18] [0/9] eta: 0:00:33 lr: 0.000565 loss: 1.2906 (1.2906) time: 3.6863 data: 3.5388 max mem: 9672 +[14:00:29.179934] Epoch: [18] [8/9] eta: 0:00:00 lr: 0.000552 loss: 1.0842 (1.0725) time: 0.5469 data: 0.4054 max mem: 9672 +[14:00:29.261459] Epoch: [18] Total time: 0:00:05 (0.5561 s / it) +[14:00:29.270763] Averaged stats: lr: 0.000552 loss: 1.0842 (1.0725) +[14:00:34.407514] val: [0/2] eta: 0:00:10 loss: 0.6664 (0.6664) time: 5.1240 data: 5.0885 max mem: 9672 +[14:00:34.425076] val: [1/2] eta: 0:00:02 loss: 0.6664 (1.1765) time: 2.5705 data: 2.5443 max mem: 9672 +[14:00:34.505435] val: Total time: 0:00:05 (2.6113 s / it) +[14:00:34.515997] val loss: 1.1765361428260803 +[14:00:34.516182] Accuracy: 0.5556, F1 Score: 0.3136, ROC AUC: 0.8747, Hamming Loss: 0.1778, + Jaccard Score: 0.2313, Precision: 0.4256, Recall: 0.3596, + Average Precision: 0.6157, Kappa: 0.3416, Score: 0.5100 +[14:00:34.553766] Best epoch = 15, Best score = 0.6203 +[14:00:34.825341] log_dir: ./output_logs/retfound +[14:00:38.753948] Epoch: [19] [0/9] eta: 0:00:35 lr: 0.000550 loss: 1.2732 (1.2732) time: 3.9275 data: 3.7878 max mem: 9672 +[14:00:39.897568] Epoch: [19] [8/9] eta: 0:00:00 lr: 0.000536 loss: 1.1827 (1.1398) time: 0.5633 data: 0.4209 max mem: 9672 +[14:00:39.978223] Epoch: [19] Total time: 0:00:05 (0.5725 s / it) +[14:00:39.987716] Averaged stats: lr: 0.000536 loss: 1.1827 (1.1398) +[14:00:45.076354] val: [0/2] eta: 0:00:10 loss: 0.8435 (0.8435) time: 5.0724 data: 5.0371 max mem: 9672 +[14:00:45.096094] val: [1/2] eta: 0:00:02 loss: 0.8435 (0.8740) time: 2.5457 data: 2.5186 max mem: 9672 +[14:00:45.176689] val: Total time: 0:00:05 (2.5867 s / it) +[14:00:45.187591] val loss: 0.8740264773368835 +[14:00:45.187813] Accuracy: 0.6444, F1 Score: 0.5012, ROC AUC: 0.8902, Hamming Loss: 0.1422, + Jaccard Score: 0.3826, Precision: 0.4857, Recall: 0.5221, + Average Precision: 0.7067, Kappa: 0.5190, Score: 0.6368 +[14:00:47.013991] Best epoch = 19, Best score = 0.6368 +[14:00:47.098065] log_dir: ./output_logs/retfound +[14:00:50.648580] Epoch: [20] [0/9] eta: 0:00:31 lr: 0.000534 loss: 0.9797 (0.9797) time: 3.5495 data: 3.4046 max mem: 9672 +[14:00:52.024713] Epoch: [20] [8/9] eta: 0:00:00 lr: 0.000518 loss: 1.1551 (1.0967) time: 0.5472 data: 0.4039 max mem: 9672 +[14:00:52.097545] Epoch: [20] Total time: 0:00:04 (0.5555 s / it) +[14:00:52.105313] Averaged stats: lr: 0.000518 loss: 1.1551 (1.0967) +[14:00:57.232767] val: [0/2] eta: 0:00:10 loss: 0.6351 (0.6351) time: 5.1106 data: 5.0758 max mem: 9672 +[14:00:57.250319] val: [1/2] eta: 0:00:02 loss: 0.6351 (1.1071) time: 2.5637 data: 2.5380 max mem: 9672 +[14:00:57.326452] val: Total time: 0:00:05 (2.6024 s / it) +[14:00:57.337940] val loss: 1.1071023344993591 +[14:00:57.338262] Accuracy: 0.5778, F1 Score: 0.3254, ROC AUC: 0.8885, Hamming Loss: 0.1689, + Jaccard Score: 0.2450, Precision: 0.4375, Recall: 0.3721, + Average Precision: 0.6421, Kappa: 0.3732, Score: 0.5290 +[14:00:57.379429] Best epoch = 19, Best score = 0.6368 +[14:00:57.647759] log_dir: ./output_logs/retfound +[14:01:01.364435] Epoch: [21] [0/9] eta: 0:00:33 lr: 0.000516 loss: 1.1534 (1.1534) time: 3.7157 data: 3.5663 max mem: 9672 +[14:01:02.512261] Epoch: [21] [8/9] eta: 0:00:00 lr: 0.000499 loss: 1.1946 (1.1608) time: 0.5403 data: 0.3963 max mem: 9672 +[14:01:02.594705] Epoch: [21] Total time: 0:00:04 (0.5496 s / it) +[14:01:02.602944] Averaged stats: lr: 0.000499 loss: 1.1946 (1.1608) +[14:01:07.249171] val: [0/2] eta: 0:00:09 loss: 0.6836 (0.6836) time: 4.6338 data: 4.5970 max mem: 9672 +[14:01:07.266283] val: [1/2] eta: 0:00:02 loss: 0.6836 (0.8540) time: 2.3252 data: 2.2985 max mem: 9672 +[14:01:07.352362] val: Total time: 0:00:04 (2.3688 s / it) +[14:01:07.363037] val loss: 0.8539901971817017 +[14:01:07.363228] Accuracy: 0.6667, F1 Score: 0.5097, ROC AUC: 0.9037, Hamming Loss: 0.1333, + Jaccard Score: 0.3853, Precision: 0.5018, Recall: 0.5221, + Average Precision: 0.6600, Kappa: 0.5405, Score: 0.6513 +[14:01:09.204207] Best epoch = 21, Best score = 0.6513 +[14:01:09.269771] log_dir: ./output_logs/retfound +[14:01:12.990791] Epoch: [22] [0/9] eta: 0:00:33 lr: 0.000496 loss: 1.2249 (1.2249) time: 3.7193 data: 3.5769 max mem: 9672 +[14:01:14.133353] Epoch: [22] [8/9] eta: 0:00:00 lr: 0.000478 loss: 1.1064 (1.1037) time: 0.5401 data: 0.3975 max mem: 9672 +[14:01:14.209980] Epoch: [22] Total time: 0:00:04 (0.5488 s / it) +[14:01:14.217859] Averaged stats: lr: 0.000478 loss: 1.1064 (1.1037) +[14:01:18.901978] val: [0/2] eta: 0:00:09 loss: 0.6482 (0.6482) time: 4.6673 data: 4.6309 max mem: 9672 +[14:01:18.919745] val: [1/2] eta: 0:00:02 loss: 0.6482 (0.9264) time: 2.3422 data: 2.3155 max mem: 9672 +[14:01:18.998092] val: Total time: 0:00:04 (2.3820 s / it) +[14:01:19.009611] val loss: 0.9264477491378784 +[14:01:19.009821] Accuracy: 0.6667, F1 Score: 0.4743, ROC AUC: 0.8962, Hamming Loss: 0.1333, + Jaccard Score: 0.3622, Precision: 0.4944, Recall: 0.4929, + Average Precision: 0.6930, Kappa: 0.5329, Score: 0.6345 +[14:01:19.046669] Best epoch = 21, Best score = 0.6513 +[14:01:19.310108] log_dir: ./output_logs/retfound +[14:01:22.987500] Epoch: [23] [0/9] eta: 0:00:33 lr: 0.000476 loss: 1.0456 (1.0456) time: 3.6764 data: 3.5284 max mem: 9672 +[14:01:24.130372] Epoch: [23] [8/9] eta: 0:00:00 lr: 0.000457 loss: 1.0456 (1.0379) time: 0.5354 data: 0.3921 max mem: 9672 +[14:01:24.208240] Epoch: [23] Total time: 0:00:04 (0.5442 s / it) +[14:01:24.216813] Averaged stats: lr: 0.000457 loss: 1.0456 (1.0379) +[14:01:29.185549] val: [0/2] eta: 0:00:09 loss: 0.6080 (0.6080) time: 4.9517 data: 4.9170 max mem: 9672 +[14:01:29.205482] val: [1/2] eta: 0:00:02 loss: 0.6080 (1.0080) time: 2.4854 data: 2.4586 max mem: 9672 +[14:01:29.285861] val: Total time: 0:00:05 (2.5263 s / it) +[14:01:29.296718] val loss: 1.0080241560935974 +[14:01:29.296929] Accuracy: 0.6444, F1 Score: 0.4560, ROC AUC: 0.9043, Hamming Loss: 0.1422, + Jaccard Score: 0.3412, Precision: 0.5700, Recall: 0.4679, + Average Precision: 0.6959, Kappa: 0.4954, Score: 0.6186 +[14:01:29.339603] Best epoch = 21, Best score = 0.6513 +[14:01:29.594904] log_dir: ./output_logs/retfound +[14:01:33.516735] Epoch: [24] [0/9] eta: 0:00:35 lr: 0.000455 loss: 1.0622 (1.0622) time: 3.9204 data: 3.7682 max mem: 9672 +[14:01:34.656000] Epoch: [24] [8/9] eta: 0:00:00 lr: 0.000435 loss: 1.0482 (1.0426) time: 0.5621 data: 0.4188 max mem: 9672 +[14:01:34.739472] Epoch: [24] Total time: 0:00:05 (0.5716 s / it) +[14:01:34.748764] Averaged stats: lr: 0.000435 loss: 1.0482 (1.0426) +[14:01:39.610430] val: [0/2] eta: 0:00:09 loss: 0.6110 (0.6110) time: 4.8455 data: 4.8106 max mem: 9672 +[14:01:39.630011] val: [1/2] eta: 0:00:02 loss: 0.6110 (0.9226) time: 2.4322 data: 2.4054 max mem: 9672 +[14:01:39.703010] val: Total time: 0:00:04 (2.4694 s / it) +[14:01:39.717463] val loss: 0.9226003885269165 +[14:01:39.717713] Accuracy: 0.6667, F1 Score: 0.4872, ROC AUC: 0.9079, Hamming Loss: 0.1333, + Jaccard Score: 0.3755, Precision: 0.5143, Recall: 0.4888, + Average Precision: 0.7014, Kappa: 0.5290, Score: 0.6414 +[14:01:39.762557] Best epoch = 21, Best score = 0.6513 +[14:01:40.009624] log_dir: ./output_logs/retfound +[14:01:43.558062] Epoch: [25] [0/9] eta: 0:00:31 lr: 0.000432 loss: 0.8416 (0.8416) time: 3.5471 data: 3.4014 max mem: 9672 +[14:01:44.701746] Epoch: [25] [8/9] eta: 0:00:00 lr: 0.000412 loss: 1.0055 (1.0117) time: 0.5211 data: 0.3780 max mem: 9672 +[14:01:44.777524] Epoch: [25] Total time: 0:00:04 (0.5297 s / it) +[14:01:44.786193] Averaged stats: lr: 0.000412 loss: 1.0055 (1.0117) +[14:01:49.811811] val: [0/2] eta: 0:00:10 loss: 0.6360 (0.6360) time: 5.0121 data: 4.9765 max mem: 9672 +[14:01:49.829465] val: [1/2] eta: 0:00:02 loss: 0.6360 (0.8886) time: 2.5145 data: 2.4883 max mem: 9672 +[14:01:49.907660] val: Total time: 0:00:05 (2.5543 s / it) +[14:01:49.920998] val loss: 0.8886241912841797 +[14:01:49.921175] Accuracy: 0.7111, F1 Score: 0.5432, ROC AUC: 0.9175, Hamming Loss: 0.1156, + Jaccard Score: 0.4233, Precision: 0.6103, Recall: 0.5513, + Average Precision: 0.7449, Kappa: 0.5988, Score: 0.6865 +[14:01:51.691055] Best epoch = 25, Best score = 0.6865 +[14:01:51.752679] log_dir: ./output_logs/retfound +[14:01:55.295368] Epoch: [26] [0/9] eta: 0:00:31 lr: 0.000409 loss: 1.2479 (1.2479) time: 3.5416 data: 3.3951 max mem: 9672 +[14:01:56.712496] Epoch: [26] [8/9] eta: 0:00:00 lr: 0.000388 loss: 1.0371 (1.0208) time: 0.5509 data: 0.4080 max mem: 9672 +[14:01:56.801634] Epoch: [26] Total time: 0:00:05 (0.5610 s / it) +[14:01:56.811554] Averaged stats: lr: 0.000388 loss: 1.0371 (1.0208) +[14:02:01.799835] val: [0/2] eta: 0:00:09 loss: 0.5489 (0.5489) time: 4.9708 data: 4.9352 max mem: 9672 +[14:02:01.819921] val: [1/2] eta: 0:00:02 loss: 0.5489 (1.0058) time: 2.4950 data: 2.4677 max mem: 9672 +[14:02:01.897983] val: Total time: 0:00:05 (2.5348 s / it) +[14:02:01.914206] val loss: 1.0057958364486694 +[14:02:01.914485] Accuracy: 0.6667, F1 Score: 0.4792, ROC AUC: 0.9141, Hamming Loss: 0.1333, + Jaccard Score: 0.3696, Precision: 0.6131, Recall: 0.4804, + Average Precision: 0.7427, Kappa: 0.5196, Score: 0.6376 +[14:02:01.959586] Best epoch = 25, Best score = 0.6865 +[14:02:02.210072] log_dir: ./output_logs/retfound +[14:02:05.785702] Epoch: [27] [0/9] eta: 0:00:32 lr: 0.000386 loss: 1.3024 (1.3024) time: 3.5746 data: 3.4263 max mem: 9672 +[14:02:07.036071] Epoch: [27] [8/9] eta: 0:00:00 lr: 0.000364 loss: 1.0271 (1.0474) time: 0.5360 data: 0.3934 max mem: 9672 +[14:02:07.123315] Epoch: [27] Total time: 0:00:04 (0.5459 s / it) +[14:02:07.132914] Averaged stats: lr: 0.000364 loss: 1.0271 (1.0474) +[14:02:12.140698] val: [0/2] eta: 0:00:09 loss: 0.6023 (0.6023) time: 4.9904 data: 4.9545 max mem: 9672 +[14:02:12.158161] val: [1/2] eta: 0:00:02 loss: 0.6023 (0.8995) time: 2.5036 data: 2.4773 max mem: 9672 +[14:02:12.232341] val: Total time: 0:00:05 (2.5413 s / it) +[14:02:12.242909] val loss: 0.8995458483695984 +[14:02:12.243147] Accuracy: 0.7111, F1 Score: 0.5376, ROC AUC: 0.9186, Hamming Loss: 0.1156, + Jaccard Score: 0.4182, Precision: 0.6080, Recall: 0.5388, + Average Precision: 0.7441, Kappa: 0.5943, Score: 0.6835 +[14:02:12.286431] Best epoch = 25, Best score = 0.6865 +[14:02:12.559573] log_dir: ./output_logs/retfound +[14:02:16.399382] Epoch: [28] [0/9] eta: 0:00:34 lr: 0.000362 loss: 1.0796 (1.0796) time: 3.8388 data: 3.6923 max mem: 9672 +[14:02:17.553662] Epoch: [28] [8/9] eta: 0:00:00 lr: 0.000340 loss: 1.0172 (1.0169) time: 0.5547 data: 0.4124 max mem: 9672 +[14:02:17.633618] Epoch: [28] Total time: 0:00:05 (0.5638 s / it) +[14:02:17.642152] Averaged stats: lr: 0.000340 loss: 1.0172 (1.0169) +[14:02:22.776826] val: [0/2] eta: 0:00:10 loss: 0.5620 (0.5620) time: 5.1167 data: 5.0813 max mem: 9672 +[14:02:22.794371] val: [1/2] eta: 0:00:02 loss: 0.5620 (0.9491) time: 2.5668 data: 2.5407 max mem: 9672 +[14:02:22.871735] val: Total time: 0:00:05 (2.6061 s / it) +[14:02:22.882433] val loss: 0.9491428732872009 +[14:02:22.882631] Accuracy: 0.6444, F1 Score: 0.4501, ROC AUC: 0.9145, Hamming Loss: 0.1422, + Jaccard Score: 0.3449, Precision: 0.4998, Recall: 0.4554, + Average Precision: 0.7355, Kappa: 0.4883, Score: 0.6176 +[14:02:22.923865] Best epoch = 25, Best score = 0.6865 +[14:02:23.191383] log_dir: ./output_logs/retfound +[14:02:27.213879] Epoch: [29] [0/9] eta: 0:00:36 lr: 0.000337 loss: 1.0353 (1.0353) time: 4.0213 data: 3.8769 max mem: 9672 +[14:02:28.357255] Epoch: [29] [8/9] eta: 0:00:00 lr: 0.000316 loss: 1.0353 (1.0094) time: 0.5737 data: 0.4309 max mem: 9672 +[14:02:28.438376] Epoch: [29] Total time: 0:00:05 (0.5830 s / it) +[14:02:28.448032] Averaged stats: lr: 0.000316 loss: 1.0353 (1.0094) +[14:02:33.703756] val: [0/2] eta: 0:00:10 loss: 0.6169 (0.6169) time: 5.2380 data: 5.2023 max mem: 9672 +[14:02:33.721403] val: [1/2] eta: 0:00:02 loss: 0.6169 (0.8872) time: 2.6275 data: 2.6012 max mem: 9672 +[14:02:33.828184] val: Total time: 0:00:05 (2.6815 s / it) +[14:02:33.843868] val loss: 0.8871878385543823 +[14:02:33.844105] Accuracy: 0.6889, F1 Score: 0.5123, ROC AUC: 0.9142, Hamming Loss: 0.1244, + Jaccard Score: 0.3948, Precision: 0.5326, Recall: 0.5221, + Average Precision: 0.7072, Kappa: 0.5661, Score: 0.6642 +[14:02:33.880528] Best epoch = 25, Best score = 0.6865 +[14:02:34.154697] log_dir: ./output_logs/retfound +[14:02:38.020654] Epoch: [30] [0/9] eta: 0:00:34 lr: 0.000313 loss: 1.2064 (1.2064) time: 3.8649 data: 3.7202 max mem: 9672 +[14:02:39.164034] Epoch: [30] [8/9] eta: 0:00:00 lr: 0.000291 loss: 0.9632 (1.0228) time: 0.5564 data: 0.4135 max mem: 9672 +[14:02:39.240449] Epoch: [30] Total time: 0:00:05 (0.5651 s / it) +[14:02:39.249102] Averaged stats: lr: 0.000291 loss: 0.9632 (1.0228) +[14:02:44.183385] val: [0/2] eta: 0:00:09 loss: 0.6312 (0.6312) time: 4.9177 data: 4.8811 max mem: 9672 +[14:02:44.200917] val: [1/2] eta: 0:00:02 loss: 0.6312 (0.8569) time: 2.4673 data: 2.4406 max mem: 9672 +[14:02:44.300073] val: Total time: 0:00:05 (2.5175 s / it) +[14:02:44.311013] val loss: 0.8569319844245911 +[14:02:44.311214] Accuracy: 0.7333, F1 Score: 0.5668, ROC AUC: 0.9103, Hamming Loss: 0.1067, + Jaccard Score: 0.4488, Precision: 0.5673, Recall: 0.5721, + Average Precision: 0.7156, Kappa: 0.6291, Score: 0.7021 +[14:02:46.215874] Best epoch = 30, Best score = 0.7021 +[14:02:46.286734] log_dir: ./output_logs/retfound +[14:02:49.837868] Epoch: [31] [0/9] eta: 0:00:31 lr: 0.000289 loss: 0.9862 (0.9862) time: 3.5501 data: 3.4054 max mem: 9672 +[14:02:50.819887] Epoch: [31] [8/9] eta: 0:00:00 lr: 0.000267 loss: 0.9842 (1.0182) time: 0.5035 data: 0.3784 max mem: 9672 +[14:02:50.901394] Epoch: [31] Total time: 0:00:04 (0.5127 s / it) +[14:02:50.902192] Averaged stats: lr: 0.000267 loss: 0.9842 (1.0182) +[14:02:56.117346] val: [0/2] eta: 0:00:10 loss: 0.5608 (0.5608) time: 5.1972 data: 5.1600 max mem: 9672 +[14:02:56.135934] val: [1/2] eta: 0:00:02 loss: 0.5608 (1.0130) time: 2.6075 data: 2.5801 max mem: 9672 +[14:02:56.214424] val: Total time: 0:00:05 (2.6476 s / it) +[14:02:56.225465] val loss: 1.0129761695861816 +[14:02:56.225728] Accuracy: 0.6889, F1 Score: 0.4987, ROC AUC: 0.9038, Hamming Loss: 0.1244, + Jaccard Score: 0.3873, Precision: 0.6210, Recall: 0.5054, + Average Precision: 0.7157, Kappa: 0.5551, Score: 0.6525 +[14:02:56.268139] Best epoch = 30, Best score = 0.7021 +[14:02:56.533846] log_dir: ./output_logs/retfound +[14:03:00.272988] Epoch: [32] [0/9] eta: 0:00:33 lr: 0.000264 loss: 0.7977 (0.7977) time: 3.7379 data: 3.5930 max mem: 9672 +[14:03:01.414536] Epoch: [32] [8/9] eta: 0:00:00 lr: 0.000243 loss: 0.9603 (1.0049) time: 0.5420 data: 0.3993 max mem: 9672 +[14:03:01.489235] Epoch: [32] Total time: 0:00:04 (0.5506 s / it) +[14:03:01.498693] Averaged stats: lr: 0.000243 loss: 0.9603 (1.0049) +[14:03:06.379031] val: [0/2] eta: 0:00:09 loss: 0.5829 (0.5829) time: 4.8632 data: 4.8261 max mem: 9672 +[14:03:06.400423] val: [1/2] eta: 0:00:02 loss: 0.5829 (0.9666) time: 2.4419 data: 2.4131 max mem: 9672 +[14:03:06.481850] val: Total time: 0:00:04 (2.4834 s / it) +[14:03:06.492577] val loss: 0.9665532112121582 +[14:03:06.492814] Accuracy: 0.6667, F1 Score: 0.4895, ROC AUC: 0.9035, Hamming Loss: 0.1333, + Jaccard Score: 0.3711, Precision: 0.5315, Recall: 0.4888, + Average Precision: 0.7054, Kappa: 0.5246, Score: 0.6392 +[14:03:06.532010] Best epoch = 30, Best score = 0.7021 +[14:03:06.824314] log_dir: ./output_logs/retfound +[14:03:10.541508] Epoch: [33] [0/9] eta: 0:00:33 lr: 0.000240 loss: 0.8847 (0.8847) time: 3.7121 data: 3.5715 max mem: 9672 +[14:03:11.689111] Epoch: [33] [8/9] eta: 0:00:00 lr: 0.000219 loss: 0.9105 (0.9674) time: 0.5399 data: 0.3969 max mem: 9672 +[14:03:11.771336] Epoch: [33] Total time: 0:00:04 (0.5496 s / it) +[14:03:11.779476] Averaged stats: lr: 0.000219 loss: 0.9105 (0.9674) +[14:03:16.525094] val: [0/2] eta: 0:00:09 loss: 0.5811 (0.5811) time: 4.7318 data: 4.6970 max mem: 9672 +[14:03:16.544886] val: [1/2] eta: 0:00:02 loss: 0.5811 (0.9574) time: 2.3755 data: 2.3486 max mem: 9672 +[14:03:16.623114] val: Total time: 0:00:04 (2.4152 s / it) +[14:03:16.633860] val loss: 0.9573783278465271 +[14:03:16.634119] Accuracy: 0.6667, F1 Score: 0.4683, ROC AUC: 0.9048, Hamming Loss: 0.1333, + Jaccard Score: 0.3636, Precision: 0.4952, Recall: 0.4804, + Average Precision: 0.7155, Kappa: 0.5283, Score: 0.6338 +[14:03:16.676600] Best epoch = 30, Best score = 0.7021 +[14:03:16.962328] log_dir: ./output_logs/retfound +[14:03:20.686724] Epoch: [34] [0/9] eta: 0:00:33 lr: 0.000217 loss: 0.9055 (0.9055) time: 3.7232 data: 3.5791 max mem: 9672 +[14:03:21.819910] Epoch: [34] [8/9] eta: 0:00:00 lr: 0.000196 loss: 1.0164 (1.0027) time: 0.5395 data: 0.3978 max mem: 9672 +[14:03:21.908952] Epoch: [34] Total time: 0:00:04 (0.5496 s / it) +[14:03:21.918789] Averaged stats: lr: 0.000196 loss: 1.0164 (1.0027) +[14:03:26.980808] val: [0/2] eta: 0:00:10 loss: 0.5906 (0.5906) time: 5.0438 data: 5.0094 max mem: 9672 +[14:03:26.998274] val: [1/2] eta: 0:00:02 loss: 0.5906 (0.9579) time: 2.5303 data: 2.5048 max mem: 9672 +[14:03:27.083119] val: Total time: 0:00:05 (2.5733 s / it) +[14:03:27.093979] val loss: 0.9578981995582581 +[14:03:27.094227] Accuracy: 0.6889, F1 Score: 0.5111, ROC AUC: 0.9080, Hamming Loss: 0.1244, + Jaccard Score: 0.3969, Precision: 0.5348, Recall: 0.5138, + Average Precision: 0.7304, Kappa: 0.5628, Score: 0.6606 +[14:03:27.135363] Best epoch = 30, Best score = 0.7021 +[14:03:27.403881] log_dir: ./output_logs/retfound +[14:03:31.212142] Epoch: [35] [0/9] eta: 0:00:34 lr: 0.000194 loss: 1.0706 (1.0706) time: 3.8070 data: 3.6583 max mem: 9672 +[14:03:32.347596] Epoch: [35] [8/9] eta: 0:00:00 lr: 0.000174 loss: 1.0706 (0.9875) time: 0.5491 data: 0.4066 max mem: 9672 +[14:03:32.433281] Epoch: [35] Total time: 0:00:05 (0.5588 s / it) +[14:03:32.442512] Averaged stats: lr: 0.000174 loss: 1.0706 (0.9875) +[14:03:37.449384] val: [0/2] eta: 0:00:09 loss: 0.5963 (0.5963) time: 4.9895 data: 4.9534 max mem: 9672 +[14:03:37.466891] val: [1/2] eta: 0:00:02 loss: 0.5963 (1.0796) time: 2.5032 data: 2.4768 max mem: 9672 +[14:03:37.545031] val: Total time: 0:00:05 (2.5429 s / it) +[14:03:37.555691] val loss: 1.079646646976471 +[14:03:37.555881] Accuracy: 0.6667, F1 Score: 0.4749, ROC AUC: 0.9038, Hamming Loss: 0.1333, + Jaccard Score: 0.3661, Precision: 0.5988, Recall: 0.4804, + Average Precision: 0.6637, Kappa: 0.5223, Score: 0.6337 +[14:03:37.611043] Best epoch = 30, Best score = 0.7021 +[14:03:37.881351] log_dir: ./output_logs/retfound +[14:03:41.672597] Epoch: [36] [0/9] eta: 0:00:34 lr: 0.000171 loss: 0.9289 (0.9289) time: 3.7901 data: 3.6449 max mem: 9672 +[14:03:42.809952] Epoch: [36] [8/9] eta: 0:00:00 lr: 0.000152 loss: 1.0378 (1.0083) time: 0.5474 data: 0.4051 max mem: 9672 +[14:03:42.894708] Epoch: [36] Total time: 0:00:05 (0.5570 s / it) +[14:03:42.902652] Averaged stats: lr: 0.000152 loss: 1.0378 (1.0083) +[14:03:47.995907] val: [0/2] eta: 0:00:10 loss: 0.6240 (0.6240) time: 5.0700 data: 5.0341 max mem: 9672 +[14:03:48.013411] val: [1/2] eta: 0:00:02 loss: 0.6240 (1.0089) time: 2.5434 data: 2.5171 max mem: 9672 +[14:03:48.092090] val: Total time: 0:00:05 (2.5834 s / it) +[14:03:48.102685] val loss: 1.0088809728622437 +[14:03:48.102973] Accuracy: 0.6889, F1 Score: 0.4931, ROC AUC: 0.9060, Hamming Loss: 0.1244, + Jaccard Score: 0.3850, Precision: 0.6063, Recall: 0.5054, + Average Precision: 0.6715, Kappa: 0.5591, Score: 0.6528 +[14:03:48.147549] Best epoch = 30, Best score = 0.7021 +[14:03:48.391366] log_dir: ./output_logs/retfound +[14:03:52.245511] Epoch: [37] [0/9] eta: 0:00:34 lr: 0.000150 loss: 1.1337 (1.1337) time: 3.8528 data: 3.7054 max mem: 9672 +[14:03:53.391467] Epoch: [37] [8/9] eta: 0:00:00 lr: 0.000132 loss: 1.0010 (1.0270) time: 0.5553 data: 0.4118 max mem: 9672 +[14:03:53.479440] Epoch: [37] Total time: 0:00:05 (0.5653 s / it) +[14:03:53.488495] Averaged stats: lr: 0.000132 loss: 1.0010 (1.0270) +[14:03:58.290976] val: [0/2] eta: 0:00:09 loss: 0.6292 (0.6292) time: 4.7838 data: 4.7485 max mem: 9672 +[14:03:58.308582] val: [1/2] eta: 0:00:02 loss: 0.6292 (0.9201) time: 2.4004 data: 2.3743 max mem: 9672 +[14:03:58.386241] val: Total time: 0:00:04 (2.4398 s / it) +[14:03:58.396960] val loss: 0.9201102256774902 +[14:03:58.397146] Accuracy: 0.6889, F1 Score: 0.5260, ROC AUC: 0.9158, Hamming Loss: 0.1244, + Jaccard Score: 0.4079, Precision: 0.5977, Recall: 0.5263, + Average Precision: 0.7483, Kappa: 0.5646, Score: 0.6688 +[14:03:58.429575] Best epoch = 30, Best score = 0.7021 +[14:03:58.696115] log_dir: ./output_logs/retfound +[14:04:02.735311] Epoch: [38] [0/9] eta: 0:00:36 lr: 0.000130 loss: 1.0621 (1.0621) time: 4.0382 data: 3.8918 max mem: 9672 +[14:04:03.875649] Epoch: [38] [8/9] eta: 0:00:00 lr: 0.000112 loss: 0.9292 (0.9671) time: 0.5753 data: 0.4325 max mem: 9672 +[14:04:03.958407] Epoch: [38] Total time: 0:00:05 (0.5847 s / it) +[14:04:03.966367] Averaged stats: lr: 0.000112 loss: 0.9292 (0.9671) +[14:04:09.220902] val: [0/2] eta: 0:00:10 loss: 0.5620 (0.5620) time: 5.2376 data: 5.2021 max mem: 9672 +[14:04:09.238413] val: [1/2] eta: 0:00:02 loss: 0.5620 (1.0064) time: 2.6272 data: 2.6011 max mem: 9672 +[14:04:09.315343] val: Total time: 0:00:05 (2.6663 s / it) +[14:04:09.325914] val loss: 1.0064192414283752 +[14:04:09.326114] Accuracy: 0.6444, F1 Score: 0.4541, ROC AUC: 0.9103, Hamming Loss: 0.1422, + Jaccard Score: 0.3462, Precision: 0.5954, Recall: 0.4554, + Average Precision: 0.7143, Kappa: 0.4846, Score: 0.6163 +[14:04:09.365435] Best epoch = 30, Best score = 0.7021 +[14:04:09.623470] log_dir: ./output_logs/retfound +[14:04:12.978994] Epoch: [39] [0/9] eta: 0:00:30 lr: 0.000110 loss: 1.1013 (1.1013) time: 3.3545 data: 3.2058 max mem: 9672 +[14:04:14.373883] Epoch: [39] [8/9] eta: 0:00:00 lr: 0.000094 loss: 0.9623 (0.9490) time: 0.5276 data: 0.3843 max mem: 9672 +[14:04:14.448841] Epoch: [39] Total time: 0:00:04 (0.5361 s / it) +[14:04:14.458359] Averaged stats: lr: 0.000094 loss: 0.9623 (0.9490) +[14:04:19.361019] val: [0/2] eta: 0:00:09 loss: 0.5394 (0.5394) time: 4.8905 data: 4.8551 max mem: 9672 +[14:04:19.380726] val: [1/2] eta: 0:00:02 loss: 0.5394 (1.0725) time: 2.4548 data: 2.4276 max mem: 9672 +[14:04:19.464616] val: Total time: 0:00:04 (2.4974 s / it) +[14:04:19.475527] val loss: 1.072479784488678 +[14:04:19.475733] Accuracy: 0.6222, F1 Score: 0.4254, ROC AUC: 0.9180, Hamming Loss: 0.1511, + Jaccard Score: 0.3230, Precision: 0.5714, Recall: 0.4304, + Average Precision: 0.7404, Kappa: 0.4492, Score: 0.5975 +[14:04:19.515415] Best epoch = 30, Best score = 0.7021 +[14:04:19.774854] log_dir: ./output_logs/retfound +[14:04:23.831027] Epoch: [40] [0/9] eta: 0:00:36 lr: 0.000092 loss: 1.2488 (1.2488) time: 4.0550 data: 3.9097 max mem: 9672 +[14:04:24.968154] Epoch: [40] [8/9] eta: 0:00:00 lr: 0.000078 loss: 0.9583 (0.9639) time: 0.5768 data: 0.4345 max mem: 9672 +[14:04:25.055692] Epoch: [40] Total time: 0:00:05 (0.5867 s / it) +[14:04:25.065238] Averaged stats: lr: 0.000078 loss: 0.9583 (0.9639) +[14:04:30.106013] val: [0/2] eta: 0:00:10 loss: 0.5461 (0.5461) time: 5.0239 data: 4.9884 max mem: 9672 +[14:04:30.125771] val: [1/2] eta: 0:00:02 loss: 0.5461 (1.0100) time: 2.5215 data: 2.4943 max mem: 9672 +[14:04:30.208979] val: Total time: 0:00:05 (2.5638 s / it) +[14:04:30.219986] val loss: 1.0099626779556274 +[14:04:30.220191] Accuracy: 0.6667, F1 Score: 0.5001, ROC AUC: 0.9141, Hamming Loss: 0.1333, + Jaccard Score: 0.3830, Precision: 0.5998, Recall: 0.4888, + Average Precision: 0.7272, Kappa: 0.5203, Score: 0.6448 +[14:04:30.261015] Best epoch = 30, Best score = 0.7021 +[14:04:30.514445] log_dir: ./output_logs/retfound +[14:04:34.495292] Epoch: [41] [0/9] eta: 0:00:35 lr: 0.000076 loss: 1.1611 (1.1611) time: 3.9798 data: 3.8338 max mem: 9672 +[14:04:35.634738] Epoch: [41] [8/9] eta: 0:00:00 lr: 0.000062 loss: 0.9897 (0.9828) time: 0.5687 data: 0.4261 max mem: 9672 +[14:04:35.711560] Epoch: [41] Total time: 0:00:05 (0.5774 s / it) +[14:04:35.721335] Averaged stats: lr: 0.000062 loss: 0.9897 (0.9828) +[14:04:40.691112] val: [0/2] eta: 0:00:09 loss: 0.5341 (0.5341) time: 4.9577 data: 4.9212 max mem: 9672 +[14:04:40.710902] val: [1/2] eta: 0:00:02 loss: 0.5341 (1.0448) time: 2.4884 data: 2.4607 max mem: 9672 +[14:04:40.787584] val: Total time: 0:00:05 (2.5274 s / it) +[14:04:40.798242] val loss: 1.0447663068771362 +[14:04:40.798519] Accuracy: 0.6444, F1 Score: 0.4541, ROC AUC: 0.9116, Hamming Loss: 0.1422, + Jaccard Score: 0.3462, Precision: 0.5954, Recall: 0.4554, + Average Precision: 0.7170, Kappa: 0.4846, Score: 0.6168 +[14:04:40.841025] Best epoch = 30, Best score = 0.7021 +[14:04:41.099765] log_dir: ./output_logs/retfound +[14:04:45.149129] Epoch: [42] [0/9] eta: 0:00:36 lr: 0.000061 loss: 1.1133 (1.1133) time: 4.0482 data: 3.9033 max mem: 9672 +[14:04:46.286381] Epoch: [42] [8/9] eta: 0:00:00 lr: 0.000048 loss: 0.9555 (0.9592) time: 0.5761 data: 0.4338 max mem: 9672 +[14:04:46.368702] Epoch: [42] Total time: 0:00:05 (0.5854 s / it) +[14:04:46.377058] Averaged stats: lr: 0.000048 loss: 0.9555 (0.9592) +[14:04:51.409432] val: [0/2] eta: 0:00:10 loss: 0.5345 (0.5345) time: 5.0197 data: 4.9820 max mem: 9672 +[14:04:51.429097] val: [1/2] eta: 0:00:02 loss: 0.5345 (1.0410) time: 2.5193 data: 2.4911 max mem: 9672 +[14:04:51.508151] val: Total time: 0:00:05 (2.5595 s / it) +[14:04:51.525570] val loss: 1.0410215854644775 +[14:04:51.525801] Accuracy: 0.6444, F1 Score: 0.4509, ROC AUC: 0.9121, Hamming Loss: 0.1422, + Jaccard Score: 0.3396, Precision: 0.5883, Recall: 0.4554, + Average Precision: 0.7183, Kappa: 0.4857, Score: 0.6162 +[14:04:51.569054] Best epoch = 30, Best score = 0.7021 +[14:04:51.852760] log_dir: ./output_logs/retfound +[14:04:55.826232] Epoch: [43] [0/9] eta: 0:00:35 lr: 0.000047 loss: 0.9757 (0.9757) time: 3.9722 data: 3.8256 max mem: 9672 +[14:04:56.966770] Epoch: [43] [8/9] eta: 0:00:00 lr: 0.000036 loss: 0.9020 (0.9108) time: 0.5680 data: 0.4251 max mem: 9672 +[14:04:57.059327] Epoch: [43] Total time: 0:00:05 (0.5785 s / it) +[14:04:57.067956] Averaged stats: lr: 0.000036 loss: 0.9020 (0.9108) +[14:05:02.334787] val: [0/2] eta: 0:00:10 loss: 0.5411 (0.5411) time: 5.2501 data: 5.2154 max mem: 9672 +[14:05:02.354577] val: [1/2] eta: 0:00:02 loss: 0.5411 (1.0117) time: 2.6346 data: 2.6078 max mem: 9672 +[14:05:02.433308] val: Total time: 0:00:05 (2.6746 s / it) +[14:05:02.443948] val loss: 1.0116748213768005 +[14:05:02.444154] Accuracy: 0.6000, F1 Score: 0.3854, ROC AUC: 0.9101, Hamming Loss: 0.1600, + Jaccard Score: 0.2979, Precision: 0.4088, Recall: 0.4054, + Average Precision: 0.7024, Kappa: 0.4218, Score: 0.5725 +[14:05:02.489321] Best epoch = 30, Best score = 0.7021 +[14:05:02.744517] log_dir: ./output_logs/retfound +[14:05:06.624547] Epoch: [44] [0/9] eta: 0:00:34 lr: 0.000035 loss: 1.0773 (1.0773) time: 3.8789 data: 3.7326 max mem: 9672 +[14:05:07.867418] Epoch: [44] [8/9] eta: 0:00:00 lr: 0.000026 loss: 0.9590 (0.9499) time: 0.5690 data: 0.4270 max mem: 9672 +[14:05:07.948679] Epoch: [44] Total time: 0:00:05 (0.5782 s / it) +[14:05:07.958077] Averaged stats: lr: 0.000026 loss: 0.9590 (0.9499) +[14:05:12.893991] val: [0/2] eta: 0:00:09 loss: 0.5476 (0.5476) time: 4.9185 data: 4.8829 max mem: 9672 +[14:05:12.911547] val: [1/2] eta: 0:00:02 loss: 0.5476 (0.9920) time: 2.4677 data: 2.4415 max mem: 9672 +[14:05:12.987586] val: Total time: 0:00:05 (2.5063 s / it) +[14:05:12.998574] val loss: 0.992044985294342 +[14:05:12.998767] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9115, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7077, Kappa: 0.4954, Score: 0.6210 +[14:05:13.035930] Best epoch = 30, Best score = 0.7021 +[14:05:13.297383] log_dir: ./output_logs/retfound +[14:05:17.228010] Epoch: [45] [0/9] eta: 0:00:35 lr: 0.000025 loss: 0.8672 (0.8672) time: 3.9293 data: 3.7906 max mem: 9672 +[14:05:18.373582] Epoch: [45] [8/9] eta: 0:00:00 lr: 0.000017 loss: 0.8581 (0.9129) time: 0.5638 data: 0.4213 max mem: 9672 +[14:05:18.461110] Epoch: [45] Total time: 0:00:05 (0.5737 s / it) +[14:05:18.462045] Averaged stats: lr: 0.000017 loss: 0.8581 (0.9129) +[14:05:23.532992] val: [0/2] eta: 0:00:10 loss: 0.5529 (0.5529) time: 5.0592 data: 5.0222 max mem: 9672 +[14:05:23.550678] val: [1/2] eta: 0:00:02 loss: 0.5529 (0.9810) time: 2.5381 data: 2.5112 max mem: 9672 +[14:05:23.630515] val: Total time: 0:00:05 (2.5787 s / it) +[14:05:23.641342] val loss: 0.9809765815734863 +[14:05:23.641548] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9114, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7134, Kappa: 0.4954, Score: 0.6210 +[14:05:23.682085] Best epoch = 30, Best score = 0.7021 +[14:05:23.935295] log_dir: ./output_logs/retfound +[14:05:27.914624] Epoch: [46] [0/9] eta: 0:00:35 lr: 0.000016 loss: 0.9652 (0.9652) time: 3.9782 data: 3.8318 max mem: 9672 +[14:05:29.128479] Epoch: [46] [8/9] eta: 0:00:00 lr: 0.000010 loss: 0.9464 (0.9680) time: 0.5768 data: 0.4343 max mem: 9672 +[14:05:29.213186] Epoch: [46] Total time: 0:00:05 (0.5864 s / it) +[14:05:29.221173] Averaged stats: lr: 0.000010 loss: 0.9464 (0.9680) +[14:05:34.421688] val: [0/2] eta: 0:00:10 loss: 0.5572 (0.5572) time: 5.1828 data: 5.1473 max mem: 9672 +[14:05:34.439291] val: [1/2] eta: 0:00:02 loss: 0.5572 (0.9713) time: 2.5999 data: 2.5737 max mem: 9672 +[14:05:34.518041] val: Total time: 0:00:05 (2.6399 s / it) +[14:05:34.529026] val loss: 0.9713383913040161 +[14:05:34.529272] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9121, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7147, Kappa: 0.4954, Score: 0.6212 +[14:05:34.573384] Best epoch = 30, Best score = 0.7021 +[14:05:34.834306] log_dir: ./output_logs/retfound +[14:05:38.870059] Epoch: [47] [0/9] eta: 0:00:36 lr: 0.000010 loss: 0.9560 (0.9560) time: 4.0345 data: 3.8851 max mem: 9672 +[14:05:40.013227] Epoch: [47] [8/9] eta: 0:00:00 lr: 0.000005 loss: 0.9749 (0.9743) time: 0.5752 data: 0.4318 max mem: 9672 +[14:05:40.105604] Epoch: [47] Total time: 0:00:05 (0.5857 s / it) +[14:05:40.114133] Averaged stats: lr: 0.000005 loss: 0.9749 (0.9743) +[14:05:45.212772] val: [0/2] eta: 0:00:10 loss: 0.5569 (0.5569) time: 5.0849 data: 5.0493 max mem: 9672 +[14:05:45.232627] val: [1/2] eta: 0:00:02 loss: 0.5569 (0.9727) time: 2.5520 data: 2.5247 max mem: 9672 +[14:05:45.309522] val: Total time: 0:00:05 (2.5911 s / it) +[14:05:45.320378] val loss: 0.9727200269699097 +[14:05:45.320624] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9117, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7140, Kappa: 0.4954, Score: 0.6211 +[14:05:45.359718] Best epoch = 30, Best score = 0.7021 +[14:05:45.613821] log_dir: ./output_logs/retfound +[14:05:49.358068] Epoch: [48] [0/9] eta: 0:00:33 lr: 0.000005 loss: 1.0635 (1.0635) time: 3.7430 data: 3.6066 max mem: 9672 +[14:05:50.669944] Epoch: [48] [8/9] eta: 0:00:00 lr: 0.000002 loss: 0.9388 (0.9302) time: 0.5615 data: 0.4193 max mem: 9672 +[14:05:50.763005] Epoch: [48] Total time: 0:00:05 (0.5721 s / it) +[14:05:50.772598] Averaged stats: lr: 0.000002 loss: 0.9388 (0.9302) +[14:05:55.776041] val: [0/2] eta: 0:00:09 loss: 0.5577 (0.5577) time: 4.9904 data: 4.9546 max mem: 9672 +[14:05:55.793707] val: [1/2] eta: 0:00:02 loss: 0.5577 (0.9716) time: 2.5037 data: 2.4774 max mem: 9672 +[14:05:55.869239] val: Total time: 0:00:05 (2.5421 s / it) +[14:05:55.884383] val loss: 0.9715694189071655 +[14:05:55.884589] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9117, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7140, Kappa: 0.4954, Score: 0.6211 +[14:05:55.933648] Best epoch = 30, Best score = 0.7021 +[14:05:56.175083] log_dir: ./output_logs/retfound +[14:05:59.784316] Epoch: [49] [0/9] eta: 0:00:32 lr: 0.000002 loss: 0.8373 (0.8373) time: 3.6079 data: 3.4598 max mem: 9672 +[14:06:00.922965] Epoch: [49] [8/9] eta: 0:00:00 lr: 0.000001 loss: 0.9645 (0.9742) time: 0.5273 data: 0.3845 max mem: 9672 +[14:06:01.007695] Epoch: [49] Total time: 0:00:04 (0.5369 s / it) +[14:06:01.016897] Averaged stats: lr: 0.000001 loss: 0.9645 (0.9742) +[14:06:06.186001] val: [0/2] eta: 0:00:10 loss: 0.5578 (0.5578) time: 5.1522 data: 5.1178 max mem: 9672 +[14:06:06.203428] val: [1/2] eta: 0:00:02 loss: 0.5578 (0.9707) time: 2.5845 data: 2.5589 max mem: 9672 +[14:06:06.280819] val: Total time: 0:00:05 (2.6238 s / it) +[14:06:06.291570] val loss: 0.9706709980964661 +[14:06:06.291866] Accuracy: 0.6444, F1 Score: 0.4562, ROC AUC: 0.9117, Hamming Loss: 0.1422, + Jaccard Score: 0.3504, Precision: 0.4696, Recall: 0.4638, + Average Precision: 0.7140, Kappa: 0.4954, Score: 0.6211 +[14:06:06.334523] Best epoch = 30, Best score = 0.7021 +[14:06:09.946591] Test with the best model, epoch = 30: +[14:06:15.032845] test: [0/3] eta: 0:00:15 loss: 0.4703 (0.4703) time: 5.0705 data: 5.0335 max mem: 9672 +[14:06:15.204375] test: [2/3] eta: 0:00:01 loss: 0.8226 (0.7300) time: 1.7471 data: 1.6779 max mem: 9672 +[14:06:15.282506] test: Total time: 0:00:05 (1.7736 s / it) +[14:06:15.297498] val loss: 0.7299975752830505 +[14:06:15.297714] Accuracy: 0.6848, F1 Score: 0.5510, ROC AUC: 0.9069, Hamming Loss: 0.1261, + Jaccard Score: 0.4356, Precision: 0.5400, Recall: 0.5657, + Average Precision: 0.6865, Kappa: 0.5737, Score: 0.6772 +[14:06:16.098246] Training time 0:09:08 +[rank0]:[W615 14:06:16.552285193 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/idrid/retfound acc=0.6848 auroc_macro_ovr=0.9069453534022897 f1_macro=0.5510 qwk=0.8781603660004815 diff --git a/results/idrid/vit/confusion_matrix.png b/results/idrid/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..3991319ae2786e4d29e5761312dc683c66a7ee9f --- /dev/null +++ b/results/idrid/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45af255539dae75a47a8b70f36ac2c1691829f3dfa94a980b27c3d38f0e561f0 +size 90971 diff --git a/results/idrid/vit/log.csv b/results/idrid/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..db4fe595bac58637ecb7e7a031a32e6f44e4b19a --- /dev/null +++ b/results/idrid/vit/log.csv @@ -0,0 +1,36 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,2.255853235721588,0.26666666666666666,0.46668995077675907,0.22004619826945424,5.545808836528073e-08 +1,1.9137949347496033,0.06666666666666667,0.5680067539780851,0.19341874138557025,1.29402206185655e-07 +2,1.7849106788635254,0.24444444444444444,0.6802171654567404,0.3101538644960425,2.0334632400602932e-07 +3,1.5728134214878082,0.3111111111111111,0.7359477669020571,0.38371775739369723,2.7729044182640365e-07 +4,1.5866973102092743,0.24444444444444444,0.7781716510014426,0.3923571314680458,3.512345596467779e-07 +5,1.5119765102863312,0.4,0.8068141004337874,0.4622396074352408,3.694672444929302e-07 +6,1.396763265132904,0.4666666666666667,0.8295699665336794,0.5114927704128608,3.683426684987483e-07 +7,1.4076372683048248,0.4444444444444444,0.8502568627167666,0.5483655053887678,3.663241848222764e-07 +8,1.3394615054130554,0.5555555555555556,0.8438629101620281,0.5719263565786763,3.6342162731308893e-07 +9,1.3054597973823547,0.4444444444444444,0.8441879499016627,0.5196877673614705,3.5964913693960667e-07 +10,1.2223477959632874,0.5111111111111111,0.8545222729515032,0.5825804920361982,3.550250928957199e-07 +11,1.2130805253982544,0.5555555555555556,0.861831783899747,0.6001595305395843,3.495720230592029e-07 +12,1.2484467923641205,0.5777777777777777,0.8549069082499315,0.6127067095118969,3.4331649423815874e-07 +13,1.224834680557251,0.6888888888888889,0.8295543295593415,0.6692596451604279,3.362889827402e-07 +14,1.2271062433719635,0.6888888888888889,0.83638391414053,0.662957434911584,3.285237258949416e-07 +15,1.177821397781372,0.6444444444444445,0.8450220075420557,0.6519326235043468,3.2005855525317277e-07 +16,1.1487193703651428,0.6,0.8452808477467659,0.6311603517679623,3.1093471227534435e-07 +17,1.130027323961258,0.6,0.844115889685256,0.6323499862276211,3.0119664740731875e-07 +18,1.1175923645496368,0.6444444444444445,0.8614340228514246,0.6516922426521572,2.908918035222662e-07 +19,1.0861523300409317,0.6222222222222222,0.8610047736651584,0.6512571718600839,2.800703847837553e-07 +20,1.0249593257904053,0.6,0.8385106167483872,0.6122106363514556,2.687851120561149e-07 +21,1.067935287952423,0.6666666666666666,0.8390688970229869,0.6703261784767737,2.570909660536824e-07 +22,0.9945769309997559,0.5777777777777777,0.8514019521186722,0.6063686006712975,2.450449194802903e-07 +23,1.0193547159433365,0.6666666666666666,0.8473950648569735,0.6662403235298863,2.3270565946397963e-07 +24,1.014438956975937,0.7111111111111111,0.8445100442494187,0.7007142075842365,2.2013330163921197e-07 +25,0.986349031329155,0.6444444444444445,0.844012508051201,0.6506386760599451,2.0738909726954043e-07 +26,0.9955818355083466,0.6444444444444445,0.8415140215009902,0.6419115279402781,1.9453513483761127e-07 +27,0.9631227105855942,0.6666666666666666,0.8468715237660707,0.6570629576944733,1.8163403755631687e-07 +28,0.9266996681690216,0.6444444444444445,0.84437657539101,0.6526651462657025,1.6874865827479806e-07 +29,0.9320847243070602,0.6666666666666666,0.8406898814025918,0.6706148217225106,1.5594177326568057e-07 +30,0.9506816565990448,0.6888888888888889,0.8463673129671525,0.6801149940148078,1.4327577638538533e-07 +31,0.9696063846349716,0.6444444444444445,0.8548697365708191,0.642522909725347,1.3081237509753143e-07 +32,0.8741843700408936,0.6444444444444445,0.845540925618111,0.6522636418727036,1.1861228984037665e-07 +33,0.8824443817138672,0.6666666666666666,0.838892467586333,0.6630774932432079,1.0673495820294563e-07 +34,0.9259753376245499,0.6888888888888889,0.8402879708713711,0.6779205807923397,9.523824535107058e-08 diff --git a/results/idrid/vit/metrics.json b/results/idrid/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..028d64a291070c89ee6526ab46db5e4dd11d7e53 --- /dev/null +++ b/results/idrid/vit/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 92, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.6521739130434783, + "balanced_accuracy": 0.6544095752444898, + "precision_macro": 0.6084321042941733, + "recall_macro": 0.6544095752444898, + "f1_macro": 0.619198400469587, + "precision_weighted": 0.6513323965722766, + "recall_weighted": 0.6521739130434783, + "f1_weighted": 0.6467931831739865, + "cohen_kappa": 0.5399281137677763, + "quadratic_weighted_kappa": 0.8356333169048792, + "mcc": 0.5417135610864549, + "auroc_macro_ovr": 0.8645763865877646, + "auroc_weighted_ovr": 0.8689293204988744, + "auprc_macro": 0.6594039846666954, + "auroc_per_class": { + "0": 0.9854312354312355, + "1": 0.8528735632183908, + "2": 0.8117398202009519, + "3": 0.7862745098039216, + "4": 0.8865628042843233 + }, + "per_class": { + "0": { + "precision": 0.8275862068965517, + "recall": 0.9230769230769231, + "f1-score": 0.8727272727272727, + "support": 26.0 + }, + "1": { + "precision": 0.4444444444444444, + "recall": 0.8, + "f1-score": 0.5714285714285714, + "support": 5.0 + }, + "2": { + "precision": 0.6428571428571429, + "recall": 0.5806451612903226, + "f1-score": 0.6101694915254238, + "support": 31.0 + }, + "3": { + "precision": 0.4, + "recall": 0.35294117647058826, + "f1-score": 0.375, + "support": 17.0 + }, + "4": { + "precision": 0.7272727272727273, + "recall": 0.6153846153846154, + "f1-score": 0.6666666666666666, + "support": 13.0 + }, + "accuracy": 0.6521739130434783, + "macro avg": { + "precision": 0.6084321042941733, + "recall": 0.6544095752444898, + "f1-score": 0.619198400469587, + "support": 92.0 + }, + "weighted avg": { + "precision": 0.6513323965722766, + "recall": 0.6521739130434783, + "f1-score": 0.6467931831739865, + "support": 92.0 + } + } +} \ No newline at end of file diff --git a/results/idrid/vit/pr.png b/results/idrid/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..8ad92b5e2f8ed24d35429d7cc5196d6b07ae433c --- /dev/null +++ b/results/idrid/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:111baed1d92ec2ddc9c67bf77b193a98838b91ec4f0f0bcbeea320330360e41f +size 100029 diff --git a/results/idrid/vit/roc.png b/results/idrid/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..216e73c2b3c2202eb0948066890dd4c5353b5a04 --- /dev/null +++ b/results/idrid/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c50935c08dfa983229f51d063637b353ffcd69697fc1223517997991b6963816 +size 82008 diff --git a/results/idrid/vit/test_pred.npz b/results/idrid/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..6edb0ce52345a8c02186baaa432a98a173b0a11f --- /dev/null +++ b/results/idrid/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5fb0f6570e56d0ab942e1809f834f03a257ed5a1faaf34f7daaf11effcb114d0 +size 3086 diff --git a/results/idrid/vit/train.log b/results/idrid/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0b648ed40f4b03d4b5a25e425b8541c4d5b9bee5 --- /dev/null +++ b/results/idrid/vit/train.log @@ -0,0 +1,184 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=318 val=45 test=92 classes=['0', '1', '2', '3', '4'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=2.2559 val_acc=0.2667 val_auc=0.4667 score=0.2200 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=1.9138 val_acc=0.0667 val_auc=0.5680 score=0.1934 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=1.7849 val_acc=0.2444 val_auc=0.6802 score=0.3102 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.5728 val_acc=0.3111 val_auc=0.7359 score=0.3837 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=1.5867 val_acc=0.2444 val_auc=0.7782 score=0.3924 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=1.5120 val_acc=0.4000 val_auc=0.8068 score=0.4622 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=1.3968 val_acc=0.4667 val_auc=0.8296 score=0.5115 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=1.4076 val_acc=0.4444 val_auc=0.8503 score=0.5484 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=1.3395 val_acc=0.5556 val_auc=0.8439 score=0.5719 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=1.3055 val_acc=0.4444 val_auc=0.8442 score=0.5197 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=1.2223 val_acc=0.5111 val_auc=0.8545 score=0.5826 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=1.2131 val_acc=0.5556 val_auc=0.8618 score=0.6002 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=1.2484 val_acc=0.5778 val_auc=0.8549 score=0.6127 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=1.2248 val_acc=0.6889 val_auc=0.8296 score=0.6693 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=1.2271 val_acc=0.6889 val_auc=0.8364 score=0.6630 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=1.1778 val_acc=0.6444 val_auc=0.8450 score=0.6519 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=1.1487 val_acc=0.6000 val_auc=0.8453 score=0.6312 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=1.1300 val_acc=0.6000 val_auc=0.8441 score=0.6323 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=1.1176 val_acc=0.6444 val_auc=0.8614 score=0.6517 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=1.0862 val_acc=0.6222 val_auc=0.8610 score=0.6513 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=1.0250 val_acc=0.6000 val_auc=0.8385 score=0.6122 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=1.0679 val_acc=0.6667 val_auc=0.8391 score=0.6703 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.9946 val_acc=0.5778 val_auc=0.8514 score=0.6064 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=1.0194 val_acc=0.6667 val_auc=0.8474 score=0.6662 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=1.0144 val_acc=0.7111 val_auc=0.8445 score=0.7007 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.9863 val_acc=0.6444 val_auc=0.8440 score=0.6506 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.9956 val_acc=0.6444 val_auc=0.8415 score=0.6419 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.9631 val_acc=0.6667 val_auc=0.8469 score=0.6571 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.9267 val_acc=0.6444 val_auc=0.8444 score=0.6527 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.9321 val_acc=0.6667 val_auc=0.8407 score=0.6706 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.9507 val_acc=0.6889 val_auc=0.8464 score=0.6801 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.9696 val_acc=0.6444 val_auc=0.8549 score=0.6425 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep32 loss=0.8742 val_acc=0.6444 val_auc=0.8455 score=0.6523 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep33 loss=0.8824 val_acc=0.6667 val_auc=0.8389 score=0.6631 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep34 loss=0.9260 val_acc=0.6889 val_auc=0.8403 score=0.6779 +[vit] early stop at ep34 (best ep24 score=0.7007) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=24 best_val_score=0.7007 -> saved test_pred.npz (92 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/idrid/vit acc=0.6522 auroc_macro_ovr=0.8645763865877646 f1_macro=0.6192 qwk=0.8356333169048792 diff --git a/results/mmac/resnet/confusion_matrix.png b/results/mmac/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..f2ac9b9fe96303df9637e27d626fd3ff6e9d53cf --- /dev/null +++ b/results/mmac/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4935390efc43933ad187b6860b20e47ddc94d0c7024b1aad46977cf51b1e8934 +size 100210 diff --git a/results/mmac/resnet/log.csv b/results/mmac/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..11bf2850fb6f9a3367be1a3a61c036c825fffe9a --- /dev/null +++ b/results/mmac/resnet/log.csv @@ -0,0 +1,26 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.6097462495168051,0.3237410071942446,0.6196130887808357,0.34083220221933747,0.00015555555555555556 +1,1.5640982151031495,0.5899280575539568,0.8303039981365105,0.5364692851877073,0.0003222222222222222 +2,1.4231563409169514,0.48201438848920863,0.8688891903475777,0.5031619767073843,0.000488888888888889 +3,1.1615390698115031,0.5107913669064749,0.8799484713569494,0.5401026505874335,0.0004995136524488283 +4,0.8302007595698039,0.7482014388489209,0.950335122487054,0.7353582823313505,0.0004979153985213637 +5,0.6137118697166443,0.6690647482014388,0.9184720972139949,0.653735422284507,0.000495209894241724 +6,0.48933639625708264,0.7985611510791367,0.9510667577604316,0.7971432413606158,0.0004914092230487934 +7,0.49812817176183066,0.6330935251798561,0.9151805469902436,0.6477480003863855,0.0004865303596627461 +8,0.31694190700848895,0.7985611510791367,0.9560042086079161,0.7958257320656855,0.00048059509427182646 +9,0.23524039884408315,0.762589928057554,0.9520729698095527,0.7659711529674591,0.00047362993521207807 +10,0.19624598721663158,0.7697841726618705,0.9470207340696062,0.7781491735421461,0.0004656659905746797 +11,0.22037608275810877,0.7913669064748201,0.958197839236211,0.7497251147796407,0.00045673882926965906 +12,0.22500026176373164,0.8129496402877698,0.9579475798882073,0.8081389147822872,0.00044688832216650565 +13,0.18665400644143423,0.8273381294964028,0.9659032379339614,0.8168479059080701,0.00043615846402118917 +14,0.1090852717558543,0.8705035971223022,0.9686229572144791,0.8748890184269115,0.0004245971769848993 +15,0.1074158363044262,0.8633093525179856,0.9683591633066614,0.8608044063167334,0.0004122560965720853 +16,0.09458980138103167,0.841726618705036,0.9677807673989941,0.8449769675812275,0.00039919034104371357 +17,0.09716446039577326,0.8273381294964028,0.9303546509665255,0.7898215294878669,0.00038545826523573424 +18,0.0929977028320233,0.8201438848920863,0.96453581585355,0.8267258295334518,0.00037112119993221794 +19,0.07291377472380797,0.8489208633093526,0.971068650546872,0.8434924732928609,0.00035624317794718564 +20,0.045595853279034294,0.8489208633093526,0.947987527481305,0.815778050574948,0.0003408906481385174 +21,0.05221998356282711,0.8561151079136691,0.9663516668708528,0.859533309251724,0.0003251321786312219 +22,0.042320853533844155,0.8489208633093526,0.9625669808830297,0.8460866781083919,0.00030903815057554555 +23,0.05039242648830016,0.8561151079136691,0.9526640879933594,0.8475725561684578,0.0002926804438076697 +24,0.03699336002270381,0.8273381294964028,0.9589967456275467,0.8249251378005608,0.0002761321158169134 diff --git a/results/mmac/resnet/metrics.json b/results/mmac/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..e92d9c91471f10e67dd42e181911ef65785ea748 --- /dev/null +++ b/results/mmac/resnet/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 279, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.8243727598566308, + "balanced_accuracy": 0.7258992648541173, + "precision_macro": 0.7483974831358766, + "recall_macro": 0.7258992648541173, + "f1_macro": 0.7359690118310807, + "precision_weighted": 0.8243097752689594, + "recall_weighted": 0.8243727598566308, + "f1_weighted": 0.8235575517740886, + "cohen_kappa": 0.7509836065573771, + "quadratic_weighted_kappa": 0.8789472745751832, + "mcc": 0.7515394812272348, + "auroc_macro_ovr": 0.9444763426613452, + "auroc_weighted_ovr": 0.9578794455001811, + "auprc_macro": 0.7623651868121825, + "auroc_per_class": { + "0": 0.9805143310297949, + "1": 0.9476829405795467, + "2": 0.9590419722265565, + "3": 0.8636363636363635, + "4": 0.9715061058344641 + }, + "per_class": { + "0": { + "precision": 0.9130434782608695, + "recall": 0.865979381443299, + "f1-score": 0.8888888888888888, + "support": 97.0 + }, + "1": { + "precision": 0.8, + "recall": 0.8571428571428571, + "f1-score": 0.8275862068965517, + "support": 98.0 + }, + "2": { + "precision": 0.8135593220338984, + "recall": 0.8275862068965517, + "f1-score": 0.8205128205128205, + "support": 58.0 + }, + "3": { + "precision": 0.6153846153846154, + "recall": 0.5333333333333333, + "f1-score": 0.5714285714285714, + "support": 15.0 + }, + "4": { + "precision": 0.6, + "recall": 0.5454545454545454, + "f1-score": 0.5714285714285714, + "support": 11.0 + }, + "accuracy": 0.8243727598566308, + "macro avg": { + "precision": 0.7483974831358766, + "recall": 0.7258992648541173, + "f1-score": 0.7359690118310807, + "support": 279.0 + }, + "weighted avg": { + "precision": 0.8243097752689594, + "recall": 0.8243727598566308, + "f1-score": 0.8235575517740886, + "support": 279.0 + } + } +} \ No newline at end of file diff --git a/results/mmac/resnet/pr.png b/results/mmac/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..e093abc6a07e2b6f9e2897551ff48634d241419e --- /dev/null +++ b/results/mmac/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b7bca94536450361a7ef727c2d944833f0faa409d699a1fe7a7d9610b46b131 +size 93364 diff --git a/results/mmac/resnet/roc.png b/results/mmac/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..f84dd76800b573037f362de375b35e574842874d --- /dev/null +++ b/results/mmac/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86cbe85a5400c0fc30a969c3da2e730443a5d5eb5c1d8bb35a2176d44361c5d3 +size 81716 diff --git a/results/mmac/resnet/test_pred.npz b/results/mmac/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..0201cc6fec782c02a1d2fcf8155509a4a8555762 --- /dev/null +++ b/results/mmac/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:105084f2f8d80bbdf4226f5d3b1b6f7c999b9f1b528322f7a77fcdd8cc345e44 +size 8322 diff --git a/results/mmac/resnet/train.log b/results/mmac/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..afbe2cffbac3483af6411a704eb2a8491c10d1ea --- /dev/null +++ b/results/mmac/resnet/train.log @@ -0,0 +1,133 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=973 val=139 test=279 classes=['0', '1', '2', '3', '4'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=1.6097 val_acc=0.3237 val_auc=0.6196 score=0.3408 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=1.5641 val_acc=0.5899 val_auc=0.8303 score=0.5365 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=1.4232 val_acc=0.4820 val_auc=0.8689 score=0.5032 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=1.1615 val_acc=0.5108 val_auc=0.8799 score=0.5401 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.8302 val_acc=0.7482 val_auc=0.9503 score=0.7354 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6137 val_acc=0.6691 val_auc=0.9185 score=0.6537 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.4893 val_acc=0.7986 val_auc=0.9511 score=0.7971 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.4981 val_acc=0.6331 val_auc=0.9152 score=0.6477 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.3169 val_acc=0.7986 val_auc=0.9560 score=0.7958 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.2352 val_acc=0.7626 val_auc=0.9521 score=0.7660 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.1962 val_acc=0.7698 val_auc=0.9470 score=0.7781 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.2204 val_acc=0.7914 val_auc=0.9582 score=0.7497 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.2250 val_acc=0.8129 val_auc=0.9579 score=0.8081 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.1867 val_acc=0.8273 val_auc=0.9659 score=0.8168 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.1091 val_acc=0.8705 val_auc=0.9686 score=0.8749 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.1074 val_acc=0.8633 val_auc=0.9684 score=0.8608 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.0946 val_acc=0.8417 val_auc=0.9678 score=0.8450 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.0972 val_acc=0.8273 val_auc=0.9304 score=0.7898 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0930 val_acc=0.8201 val_auc=0.9645 score=0.8267 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0729 val_acc=0.8489 val_auc=0.9711 score=0.8435 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0456 val_acc=0.8489 val_auc=0.9480 score=0.8158 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0522 val_acc=0.8561 val_auc=0.9664 score=0.8595 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0423 val_acc=0.8489 val_auc=0.9626 score=0.8461 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0504 val_acc=0.8561 val_auc=0.9527 score=0.8476 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0370 val_acc=0.8273 val_auc=0.9590 score=0.8249 +[resnet] early stop at ep24 (best ep14 score=0.8749) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=14 best_val_score=0.8749 -> saved test_pred.npz (279 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/mmac/resnet acc=0.8244 auroc_macro_ovr=0.9444763426613452 f1_macro=0.7360 qwk=0.8789472745751832 diff --git a/results/mmac/retfound/confusion_matrix.png b/results/mmac/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..07b7a8a96c1f3295dd7658e4b3fc683aba81f3e4 --- /dev/null +++ b/results/mmac/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:193b4f3c7e9d47a8aafce810ebcb7e2914f1e31cf74df024bd323319b8bca71b +size 95984 diff --git a/results/mmac/retfound/confusion_matrix_test.jpg b/results/mmac/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..07ff27ed8dd74a07b8c9459989814f15f8b12959 --- /dev/null +++ b/results/mmac/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b21d94ec46c098c8672ea84d7661fdc4cb3cfbbb0a13df561f38881396a7b267 +size 339418 diff --git a/results/mmac/retfound/log.txt b/results/mmac/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..e2d2a8db26208086f53685a9d9b98240ce09738e --- /dev/null +++ b/results/mmac/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 3.0208333333333334e-05, "train_loss": 1.5535105387369792, "epoch": 0, "n_parameters": 303306757} +{"train_lr": 9.270833333333336e-05, "train_loss": 1.3521804809570312, "epoch": 1, "n_parameters": 303306757} +{"train_lr": 0.00015520833333333332, "train_loss": 1.1724380493164062, "epoch": 2, "n_parameters": 303306757} +{"train_lr": 0.00021770833333333334, "train_loss": 0.9836680094401041, "epoch": 3, "n_parameters": 303306757} +{"train_lr": 0.00028020833333333337, "train_loss": 0.8754009246826172, "epoch": 4, "n_parameters": 303306757} +{"train_lr": 0.0003427083333333334, "train_loss": 0.8619959513346355, "epoch": 5, "n_parameters": 303306757} +{"train_lr": 0.0004052083333333333, "train_loss": 0.808201805750529, "epoch": 6, "n_parameters": 303306757} +{"train_lr": 0.00046770833333333327, "train_loss": 0.7841961065928141, "epoch": 7, "n_parameters": 303306757} +{"train_lr": 0.0005302083333333333, "train_loss": 0.7557985941569011, "epoch": 8, "n_parameters": 303306757} +{"train_lr": 0.0005927083333333335, "train_loss": 0.7590953787167867, "epoch": 9, "n_parameters": 303306757} +{"train_lr": 0.0006246951886200701, "train_loss": 0.7279052098592123, "epoch": 10, "n_parameters": 303306757} +{"train_lr": 0.0006228055438622347, "train_loss": 0.73025541305542, "epoch": 11, "n_parameters": 303306757} +{"train_lr": 0.0006190058449056913, "train_loss": 0.7005573570728302, "epoch": 12, "n_parameters": 303306757} +{"train_lr": 0.0006133195181580349, "train_loss": 0.713701456785202, "epoch": 13, "n_parameters": 303306757} +{"train_lr": 0.0006057816217145604, "train_loss": 0.6676824152469635, "epoch": 14, "n_parameters": 303306757} +{"train_lr": 0.0005964386292134469, "train_loss": 0.6528079469998678, "epoch": 15, "n_parameters": 303306757} +{"train_lr": 0.0005853481433103239, "train_loss": 0.6654120067755381, "epoch": 16, "n_parameters": 303306757} +{"train_lr": 0.0005725785405387451, "train_loss": 0.664145161708196, "epoch": 17, "n_parameters": 303306757} +{"train_lr": 0.0005582085497461208, "train_loss": 0.6687527408202489, "epoch": 18, "n_parameters": 303306757} +{"train_lr": 0.0005423267667041947, "train_loss": 0.6771090696255366, "epoch": 19, "n_parameters": 303306757} +{"train_lr": 0.0005250311078866487, "train_loss": 0.6486600339412689, "epoch": 20, "n_parameters": 303306757} +{"train_lr": 0.0005064282067814835, "train_loss": 0.6116965899864832, "epoch": 21, "n_parameters": 303306757} +{"train_lr": 0.00048663275646010454, "train_loss": 0.659013032913208, "epoch": 22, "n_parameters": 303306757} +{"train_lr": 0.0004657668024563984, "train_loss": 0.5751704454421998, "epoch": 23, "n_parameters": 303306757} +{"train_lr": 0.00044395899031543187, "train_loss": 0.6024237036705017, "epoch": 24, "n_parameters": 303306757} +{"train_lr": 0.0004213437724508874, "train_loss": 0.570877734820048, "epoch": 25, "n_parameters": 303306757} +{"train_lr": 0.00039806057920122295, "train_loss": 0.5844285567601522, "epoch": 26, "n_parameters": 303306757} +{"train_lr": 0.00037425295919526997, "train_loss": 0.5988928576310476, "epoch": 27, "n_parameters": 303306757} +{"train_lr": 0.00035006769432720225, "train_loss": 0.6173951029777527, "epoch": 28, "n_parameters": 303306757} +{"train_lr": 0.0003256538947973482, "train_loss": 0.5680068333943685, "epoch": 29, "n_parameters": 303306757} +{"train_lr": 0.0003011620797982215, "train_loss": 0.5746919999519984, "epoch": 30, "n_parameters": 303306757} +{"train_lr": 0.00027674324951364664, "train_loss": 0.5695249656836192, "epoch": 31, "n_parameters": 303306757} +{"train_lr": 0.0002525479541524131, "train_loss": 0.5344164133071899, "epoch": 32, "n_parameters": 303306757} +{"train_lr": 0.00022872536575617525, "train_loss": 0.5515508552392324, "epoch": 33, "n_parameters": 303306757} +{"train_lr": 0.00020542235850421015, "train_loss": 0.5662560880184173, "epoch": 34, "n_parameters": 303306757} +{"train_lr": 0.00018278260318526244, "train_loss": 0.5562593698501587, "epoch": 35, "n_parameters": 303306757} +{"train_lr": 0.00016094568141935615, "train_loss": 0.4990649511416753, "epoch": 36, "n_parameters": 303306757} +{"train_lr": 0.00014004622509069135, "train_loss": 0.5092185089985529, "epoch": 37, "n_parameters": 303306757} +{"train_lr": 0.00012021308629730876, "train_loss": 0.5353374898433685, "epoch": 38, "n_parameters": 303306757} +{"train_lr": 0.00010156854293505751, "train_loss": 0.5241690337657928, "epoch": 39, "n_parameters": 303306757} +{"train_lr": 8.422754481370566e-05, "train_loss": 0.5310941249132156, "epoch": 40, "n_parameters": 303306757} +{"train_lr": 6.829700495313831e-05, "train_loss": 0.5278316070636113, "epoch": 41, "n_parameters": 303306757} +{"train_lr": 5.387514042903803e-05, "train_loss": 0.5098071197668711, "epoch": 42, "n_parameters": 303306757} +{"train_lr": 4.105086683195245e-05, "train_loss": 0.5050264914830526, "epoch": 43, "n_parameters": 303306757} +{"train_lr": 2.9903250073111336e-05, "train_loss": 0.5105938464403152, "epoch": 44, "n_parameters": 303306757} +{"train_lr": 2.0501018916791655e-05, "train_loss": 0.4953600257635117, "epoch": 45, "n_parameters": 303306757} +{"train_lr": 1.2902141244631288e-05, "train_loss": 0.4830067833264669, "epoch": 46, "n_parameters": 303306757} +{"train_lr": 7.153466664362297e-06, "train_loss": 0.49064323405424753, "epoch": 47, "n_parameters": 303306757} +{"train_lr": 3.2904376664003598e-06, "train_loss": 0.5370309183994929, "epoch": 48, "n_parameters": 303306757} +{"train_lr": 1.3368711091053904e-06, "train_loss": 0.4606425344944, "epoch": 49, "n_parameters": 303306757} diff --git a/results/mmac/retfound/metrics.json b/results/mmac/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..9c308d5af29f49391585709bf4209a5d38f0b11b --- /dev/null +++ b/results/mmac/retfound/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 279, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.8566308243727598, + "balanced_accuracy": 0.7423382427333471, + "precision_macro": 0.8010902460042265, + "recall_macro": 0.7423382427333471, + "f1_macro": 0.7602307525273241, + "precision_weighted": 0.8588280529947351, + "recall_weighted": 0.8566308243727598, + "f1_weighted": 0.8551887713696437, + "cohen_kappa": 0.7966583459359, + "quadratic_weighted_kappa": 0.9263793516080929, + "mcc": 0.7971033534126032, + "auroc_macro_ovr": 0.9672777830979756, + "auroc_weighted_ovr": 0.9724144852750178, + "auprc_macro": 0.8270549335220301, + "auroc_per_class": { + "0": 0.9904270986745213, + "1": 0.9613823429924455, + "2": 0.9701591511936339, + "3": 0.9222222222222223, + "4": 0.9921981004070556 + }, + "per_class": { + "0": { + "precision": 0.9468085106382979, + "recall": 0.9175257731958762, + "f1-score": 0.9319371727748691, + "support": 97.0 + }, + "1": { + "precision": 0.8431372549019608, + "recall": 0.8775510204081632, + "f1-score": 0.86, + "support": 98.0 + }, + "2": { + "precision": 0.819672131147541, + "recall": 0.8620689655172413, + "f1-score": 0.8403361344537815, + "support": 58.0 + }, + "3": { + "precision": 0.5625, + "recall": 0.6, + "f1-score": 0.5806451612903226, + "support": 15.0 + }, + "4": { + "precision": 0.8333333333333334, + "recall": 0.45454545454545453, + "f1-score": 0.5882352941176471, + "support": 11.0 + }, + "accuracy": 0.8566308243727598, + "macro avg": { + "precision": 0.8010902460042265, + "recall": 0.7423382427333471, + "f1-score": 0.7602307525273241, + "support": 279.0 + }, + "weighted avg": { + "precision": 0.8588280529947351, + "recall": 0.8566308243727598, + "f1-score": 0.8551887713696437, + "support": 279.0 + } + } +} \ No newline at end of file diff --git a/results/mmac/retfound/metrics_test.csv b/results/mmac/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..e48a39fee2c82dc084268c8d1533a5d6f1c2c11d --- /dev/null +++ b/results/mmac/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.47984142270353103,0.8566308243727598,0.7602307525273241,0.9672777830979756,0.05734767025089606,0.635466048287424,0.8010902460042265,0.7423382427333471,0.8270549335220301,0.7966583459359 diff --git a/results/mmac/retfound/metrics_val.csv b/results/mmac/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..88b6d1da941aab622e0b388d91ffd7fd3d297781 --- /dev/null +++ b/results/mmac/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +1.5148043632507324,0.35251798561151076,0.10425531914893618,0.704577763398414,0.2589928057553957,0.07050359712230216,0.07050359712230216,0.2,0.35401667565547723,0.0 +1.384571123123169,0.4028776978417266,0.15919638372677045,0.8771350662704046,0.23884892086330936,0.10340909090909092,0.27424242424242423,0.22916666666666666,0.6176892632203947,0.07829352081169616 +0.9155297040939331,0.7266187050359713,0.45339360022958813,0.9342915755347697,0.1093525179856115,0.3678047319556753,0.4350544804033176,0.4831192821956368,0.6967415243667883,0.6062322946175638 +0.8349391222000122,0.7122302158273381,0.43067152348816523,0.9437171183035845,0.11510791366906475,0.34647557653198263,0.4239372822299652,0.44953671123621863,0.6994042370908049,0.5759286095644879 +0.6916433572769165,0.7266187050359713,0.5122988156696022,0.9467051614303363,0.1093525179856115,0.3992775041050903,0.5164132104454684,0.5216045038705137,0.6860736876518463,0.6059091248228008 +0.6218647718429565,0.7482014388489209,0.5577318475396873,0.9564659677026801,0.10071942446043165,0.44261254674802464,0.5527777777777778,0.5717335209946047,0.7192328580779352,0.6424634379363563 +0.6278911352157592,0.7553956834532374,0.6479856157752267,0.9543172152797508,0.09784172661870504,0.506020202020202,0.7723344760368354,0.6886113065916022,0.7110891040853018,0.6565656565656566 +0.5456351161003112,0.8201438848920863,0.6266691228393356,0.9663830061635998,0.07194244604316546,0.527699417491113,0.623249299719888,0.6953964344358433,0.7731618155447997,0.7430303926643496 +0.5968305110931397,0.7697841726618705,0.6776889617635673,0.9678357501542619,0.0920863309352518,0.5296650175221603,0.7440224089635854,0.655116701853155,0.8180289459530993,0.6697111457637186 +0.5600282669067382,0.7841726618705036,0.6251181498240321,0.9737644479652516,0.08633093525179857,0.5153153988868275,0.6158338412814116,0.6797677691766362,0.8589765948758933,0.6906528189910979 +0.648695421218872,0.8201438848920863,0.6208046996185228,0.9677171444348007,0.07194244604316546,0.5219618182260497,0.6080380264413248,0.6505248651184612,0.810356189193344,0.7425544525114831 +0.5922170877456665,0.7985611510791367,0.6338480617440656,0.9658824007712392,0.08057553956834532,0.5291125541125542,0.6328376886271624,0.6442675346000468,0.8024524232228949,0.7124916894437467 +0.4726423978805542,0.8489208633093526,0.8017755060473506,0.9697135879129656,0.060431654676258995,0.6811981566820277,0.8683385431467272,0.7935520760028149,0.8304874173857929,0.7836335334667557 +0.4674602270126343,0.841726618705036,0.7415776588065547,0.9727211944855192,0.06330935251798561,0.6127239709443099,0.7525975144579795,0.7343947923997185,0.8298006416581915,0.7761674718196457 +0.4194536328315735,0.8561151079136691,0.8073877551020407,0.9746700401908465,0.05755395683453238,0.6897471320698385,0.8650762527233116,0.8032635467980296,0.8743684620193557,0.7951363301400147 +0.5885109603404999,0.7985611510791367,0.6104007820136853,0.9643507816615153,0.08057553956834532,0.5070198412698412,0.6380846293961049,0.5979562514661036,0.7878594480649821,0.7089658266656697 +0.5172239661216735,0.8129496402877698,0.7300935047065078,0.973784578714576,0.07482014388489208,0.5953750380749315,0.8323426573426573,0.7425140745953553,0.8945599679669792,0.7323755924170616 +0.47414124608039854,0.8201438848920863,0.7632215270110005,0.974778426136995,0.07194244604316546,0.6260537557985801,0.8093772893772894,0.7371569317382124,0.8956860730243148,0.7414242131110946 +0.5710845947265625,0.8273381294964028,0.6816275624686841,0.9697630377141655,0.06906474820143885,0.6001063452676356,0.663780472401162,0.7049378372038471,0.8249152589975817,0.7506540100156962 +0.4621611088514328,0.8345323741007195,0.6638918730380189,0.9755202867595919,0.06618705035971223,0.5710774410774412,0.6265064102564103,0.7232641332395027,0.8798802165813303,0.7654782863849765 +0.4679856836795807,0.8489208633093526,0.7549860309182342,0.9747847251127002,0.060431654676258995,0.6360588235294118,0.7942691029900333,0.7507301196340606,0.8666770625588937,0.7850515463917526 +0.5458193361759186,0.7913669064748201,0.7277483888454137,0.971173022020183,0.08345323741007195,0.5862689878818912,0.76257058287796,0.7138253577292986,0.83886770087881,0.7043854502786742 +0.4953368067741394,0.8345323741007195,0.7589478023664799,0.9728896031948923,0.06618705035971223,0.6264164900933051,0.8228005226480836,0.7329580107905231,0.8428998967644631,0.7651682091964154 +0.4528108716011047,0.8489208633093526,0.7535431235431236,0.9772770747890659,0.060431654676258995,0.6344563331405436,0.7888141321044546,0.7509852216748768,0.889434527980326,0.7854465270121279 +0.4765301883220673,0.8345323741007195,0.6983917822325806,0.9754376382986703,0.06618705035971223,0.5870625479144749,0.7292853246044736,0.6862655406990382,0.8506442135890036,0.7638499039739991 +0.5353950619697571,0.8273381294964028,0.6714608959342843,0.9740363015401906,0.06906474820143885,0.551533019987656,0.7981095615714157,0.6678131597466572,0.8368601173271315,0.7542903439640569 +0.5428592979907989,0.8057553956834532,0.6398163148420879,0.9730429435024975,0.0776978417266187,0.5177921218139689,0.7676081070726163,0.6339784189537884,0.834085884939838,0.7244088706124248 +0.45543997287750243,0.8273381294964028,0.7163488436755763,0.9752106635766216,0.06906474820143885,0.5847743395468641,0.8039461222771755,0.6997947454844007,0.8581969904182245,0.7556938850238007 +0.3940145254135132,0.8489208633093526,0.8143590769407723,0.9777931618118292,0.060431654676258995,0.6924794763804051,0.8482730035921525,0.7952029087497068,0.9013704247378568,0.7866695900021925 +0.46051245033740995,0.8345323741007195,0.7806774668630337,0.974053220099694,0.06618705035971223,0.6530552797087057,0.8153829273644941,0.7597349284541403,0.8436124695613201,0.7657017222425797 +0.38845452964305877,0.841726618705036,0.7870551241911073,0.9784326419914311,0.06330935251798561,0.6617193872926673,0.8179194847020934,0.7797765657987333,0.8968123533522732,0.7759378663540446 +0.4829346999526024,0.8345323741007195,0.6726890756302522,0.9748927941674894,0.06618705035971223,0.556959706959707,0.7960522273425499,0.6747947454844005,0.837666227493927,0.7656330181071769 +0.5373015165328979,0.8201438848920863,0.6643668441134858,0.9745594206444608,0.07194244604316546,0.5435892490665876,0.793888168697606,0.6634764250527797,0.8337826021545173,0.7443913203383596 +0.4446850061416626,0.841726618705036,0.7880364289798252,0.9773944348510073,0.06330935251798561,0.6629151332664213,0.8284707737084647,0.7662063101102509,0.8878548443268123,0.7757406864183045 +0.5135580390691757,0.841726618705036,0.7586261813967131,0.9738011874335442,0.06330935251798561,0.629722952355056,0.815993265993266,0.7371246774571898,0.8327130879097666,0.7754607533592774 +0.43940497636795045,0.841726618705036,0.7878407014979905,0.9757373505416025,0.06330935251798561,0.6627375017803732,0.8242038241733919,0.7663763781374618,0.845565068816953,0.7758064516129033 +0.4637205421924591,0.8345323741007195,0.7926036392985546,0.975890683726437,0.06618705035971223,0.6648312499366364,0.8232539682539682,0.7837995543044805,0.8636391532661272,0.7666423357664234 +0.5187703669071198,0.841726618705036,0.7641882483987745,0.9745161060671821,0.06330935251798561,0.6338416422287391,0.8321226795803067,0.7342247243725076,0.8352177434336951,0.7740672330993721 +0.4630659490823746,0.8561151079136691,0.7959311043566363,0.9756143905067939,0.05755395683453238,0.6756951201533244,0.831160009585882,0.774709711470795,0.8425605382577273,0.7962175634071251 +0.4450608491897583,0.8345323741007195,0.7762668516190558,0.976315617376027,0.06618705035971223,0.6497504887933602,0.8086277873070327,0.759479826413324,0.860819237626656,0.765993265993266 +0.46497651040554044,0.8489208633093526,0.7978191194076769,0.9762065490468395,0.060431654676258995,0.6738630275756601,0.8482651072124756,0.764913206661975,0.8466292645288129,0.784081662844885 +0.49512178897857667,0.8561151079136691,0.803750320827562,0.9751993924031778,0.05755395683453238,0.6825549770250231,0.8565505226480836,0.7716396903589021,0.8408682011701833,0.7946065755448837 +0.4630673497915268,0.8561151079136691,0.7959311043566363,0.9756077128011273,0.05755395683453238,0.6756951201533244,0.831160009585882,0.774709711470795,0.8394698387867839,0.7962175634071251 +0.4626917392015457,0.8705035971223022,0.8062802201774224,0.9759157381091527,0.051798561151079135,0.6915174575223,0.8468498833290121,0.7827029087497066,0.8405028427915371,0.8160023532872481 +0.46102696359157563,0.8561151079136691,0.7905639347765613,0.9765996970073972,0.05755395683453238,0.6714471243042672,0.828619488165325,0.7716396903589021,0.8614335977168197,0.7958734121447977 +0.45225479304790495,0.8705035971223022,0.8062802201774224,0.9766908360271078,0.051798561151079135,0.6915174575223,0.8468498833290121,0.7827029087497066,0.8617903823971119,0.8160023532872481 +0.4584751546382904,0.8633093525179856,0.8021757228674782,0.976537437154497,0.05467625899280575,0.6847804576376004,0.8438904241259161,0.77853624208304,0.8614141120495462,0.8057659777892182 +0.4597661018371582,0.8705035971223022,0.8062802201774224,0.9763209869380468,0.051798561151079135,0.6915174575223,0.8468498833290121,0.7827029087497066,0.8600586355583241,0.8160023532872481 +0.46119038164615633,0.8633093525179856,0.8012375571697605,0.9760591852476737,0.05467625899280575,0.6836712990280858,0.8404761904761905,0.7786212760966454,0.8589442333691306,0.8060655015420767 +0.461053204536438,0.8633093525179856,0.8012375571697605,0.9760591852476738,0.05467625899280575,0.6836712990280858,0.8404761904761905,0.7786212760966454,0.8589052335667295,0.8060655015420767 diff --git a/results/mmac/retfound/pr.png b/results/mmac/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..76b70419864675efaa5339d869402cebec2bb1f6 --- /dev/null +++ b/results/mmac/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:090f41a8c62150e90e8c312ebe81f2a4eabc574791e9891beada2495a867c442 +size 84470 diff --git a/results/mmac/retfound/roc.png b/results/mmac/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..41a3f15402a7953094347b9e0850740da5ce93d2 --- /dev/null +++ b/results/mmac/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73ace34b884c9cd21d303bd7e85edcad5a39e2c4d96ccc617bc5bc2733b93f4d +size 81704 diff --git a/results/mmac/retfound/test_pred.npz b/results/mmac/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..26ae3b98ce0e1fd16acdab6992fbd45d7015ab3e --- /dev/null +++ b/results/mmac/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:108cd9f96ed7a9ecc3bc8da1f53ac60df7d453c9857d25b73ede5c41ce1a5e07 +size 5532 diff --git a/results/mmac/retfound/train.log b/results/mmac/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..a9413c1284f25eb6fae3e73635758ed25e333379 --- /dev/null +++ b/results/mmac/retfound/train.log @@ -0,0 +1,783 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:52:23.591203868 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:52:24.998443] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:52:24.998785] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/Myopia/Classification_of_Myopic_Maculopathy', +nb_classes=5, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/mmac', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:52:31.551192] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:36.888407] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:52:38.165620] Sampler_train = +[13:52:38.226587] len of train_set: 960 +[13:52:38.934427] [Adaptation] Full fine-tuning: training all parameters. +[13:52:38.935562] number of trainable params (M): 303.31 +[13:52:38.935636] base lr: 5.00e-03 +[13:52:38.935682] actual lr: 6.25e-04 +[13:52:38.935733] accumulate grad iterations: 1 +[13:52:38.935778] effective batch size: 32 +[13:52:38.939561] criterion = CrossEntropyLoss() +[13:52:38.939687] Start training for 50 epochs +[13:52:38.942577] log_dir: ./output_logs/retfound +[13:52:42.093563] Epoch: [0] [ 0/30] eta: 0:01:34 lr: 0.000000 loss: 1.6090 (1.6090) time: 3.1496 data: 1.7667 max mem: 7340 +[13:52:45.044713] Epoch: [0] [20/30] eta: 0:00:02 lr: 0.000042 loss: 1.5865 (1.5809) time: 0.1475 data: 0.0002 max mem: 9672 +[13:52:46.334662] Epoch: [0] [29/30] eta: 0:00:00 lr: 0.000060 loss: 1.5183 (1.5535) time: 0.1434 data: 0.0002 max mem: 9672 +[13:52:46.462627] Epoch: [0] Total time: 0:00:07 (0.2507 s / it) +[13:52:46.471601] Averaged stats: lr: 0.000060 loss: 1.5183 (1.5535) +[13:52:48.119762] val: [0/5] eta: 0:00:08 loss: 1.3714 (1.3714) time: 1.6393 data: 1.5927 max mem: 9672 +[13:52:48.319436] val: [4/5] eta: 0:00:00 loss: 1.3714 (1.5148) time: 0.3677 data: 0.3187 max mem: 9672 +[13:52:48.450461] val: Total time: 0:00:01 (0.3942 s / it) +[13:52:48.467866] val loss: 1.5148043632507324 +[13:52:48.468063] Accuracy: 0.3525, F1 Score: 0.1043, ROC AUC: 0.7046, Hamming Loss: 0.2590, + Jaccard Score: 0.0705, Precision: 0.0705, Recall: 0.2000, + Average Precision: 0.3540, Kappa: 0.0000, Score: 0.2696 +[13:52:50.416570] Best epoch = 0, Best score = 0.2696 +[13:52:50.494032] log_dir: ./output_logs/retfound +[13:52:52.378734] Epoch: [1] [ 0/30] eta: 0:00:56 lr: 0.000063 loss: 1.4557 (1.4557) time: 1.8831 data: 1.6320 max mem: 9672 +[13:52:55.245576] Epoch: [1] [20/30] eta: 0:00:02 lr: 0.000104 loss: 1.3954 (1.3765) time: 0.1433 data: 0.0002 max mem: 9672 +[13:52:56.534711] Epoch: [1] [29/30] eta: 0:00:00 lr: 0.000123 loss: 1.3466 (1.3522) time: 0.1434 data: 0.0001 max mem: 9672 +[13:52:56.675572] Epoch: [1] Total time: 0:00:06 (0.2060 s / it) +[13:52:56.684131] Averaged stats: lr: 0.000123 loss: 1.3466 (1.3522) +[13:52:58.147248] val: [0/5] eta: 0:00:07 loss: 1.0444 (1.0444) time: 1.4559 data: 1.4200 max mem: 9672 +[13:52:58.318933] val: [4/5] eta: 0:00:00 loss: 1.0444 (1.3846) time: 0.3254 data: 0.2938 max mem: 9672 +[13:52:58.475481] val: Total time: 0:00:01 (0.3570 s / it) +[13:52:58.490583] val loss: 1.384571123123169 +[13:52:58.490812] Accuracy: 0.4029, F1 Score: 0.1592, ROC AUC: 0.8771, Hamming Loss: 0.2388, + Jaccard Score: 0.1034, Precision: 0.2742, Recall: 0.2292, + Average Precision: 0.6177, Kappa: 0.0783, Score: 0.3715 +[13:53:00.287895] Best epoch = 1, Best score = 0.3715 +[13:53:00.348634] log_dir: ./output_logs/retfound +[13:53:02.029105] Epoch: [2] [ 0/30] eta: 0:00:50 lr: 0.000125 loss: 1.4745 (1.4745) time: 1.6793 data: 1.5300 max mem: 9672 +[13:53:04.895558] Epoch: [2] [20/30] eta: 0:00:02 lr: 0.000167 loss: 1.1794 (1.2117) time: 0.1433 data: 0.0002 max mem: 9672 +[13:53:06.189001] Epoch: [2] [29/30] eta: 0:00:00 lr: 0.000185 loss: 1.1133 (1.1724) time: 0.1434 data: 0.0001 max mem: 9672 +[13:53:06.331752] Epoch: [2] Total time: 0:00:05 (0.1994 s / it) +[13:53:06.341302] Averaged stats: lr: 0.000185 loss: 1.1133 (1.1724) +[13:53:07.782683] val: [0/5] eta: 0:00:07 loss: 0.5806 (0.5806) time: 1.4250 data: 1.3885 max mem: 9672 +[13:53:08.014701] val: [4/5] eta: 0:00:00 loss: 0.6848 (0.9155) time: 0.3313 data: 0.3008 max mem: 9672 +[13:53:08.147994] val: Total time: 0:00:01 (0.3582 s / it) +[13:53:08.161593] val loss: 0.9155297040939331 +[13:53:08.161783] Accuracy: 0.7266, F1 Score: 0.4534, ROC AUC: 0.9343, Hamming Loss: 0.1094, + Jaccard Score: 0.3678, Precision: 0.4351, Recall: 0.4831, + Average Precision: 0.6967, Kappa: 0.6062, Score: 0.6646 +[13:53:10.033012] Best epoch = 2, Best score = 0.6646 +[13:53:10.101926] log_dir: ./output_logs/retfound +[13:53:11.793335] Epoch: [3] [ 0/30] eta: 0:00:50 lr: 0.000188 loss: 0.9751 (0.9751) time: 1.6904 data: 1.5456 max mem: 9672 +[13:53:14.660077] Epoch: [3] [20/30] eta: 0:00:02 lr: 0.000229 loss: 0.9944 (1.0196) time: 0.1433 data: 0.0002 max mem: 9672 +[13:53:15.948419] Epoch: [3] [29/30] eta: 0:00:00 lr: 0.000248 loss: 0.9610 (0.9837) time: 0.1436 data: 0.0001 max mem: 9672 +[13:53:16.090681] Epoch: [3] Total time: 0:00:05 (0.1996 s / it) +[13:53:16.100604] Averaged stats: lr: 0.000248 loss: 0.9610 (0.9837) +[13:53:17.691702] val: [0/5] eta: 0:00:07 loss: 0.4705 (0.4705) time: 1.5803 data: 1.5441 max mem: 9672 +[13:53:17.813254] val: [4/5] eta: 0:00:00 loss: 0.4705 (0.8349) time: 0.3403 data: 0.3089 max mem: 9672 +[13:53:17.933227] val: Total time: 0:00:01 (0.3645 s / it) +[13:53:17.946295] val loss: 0.8349391222000122 +[13:53:17.946497] Accuracy: 0.7122, F1 Score: 0.4307, ROC AUC: 0.9437, Hamming Loss: 0.1151, + Jaccard Score: 0.3465, Precision: 0.4239, Recall: 0.4495, + Average Precision: 0.6994, Kappa: 0.5759, Score: 0.6501 +[13:53:17.986548] Best epoch = 2, Best score = 0.6646 +[13:53:18.264318] log_dir: ./output_logs/retfound +[13:53:19.877112] Epoch: [4] [ 0/30] eta: 0:00:48 lr: 0.000250 loss: 1.0670 (1.0670) time: 1.6117 data: 1.4644 max mem: 9672 +[13:53:22.749886] Epoch: [4] [20/30] eta: 0:00:02 lr: 0.000292 loss: 0.8891 (0.8935) time: 0.1436 data: 0.0002 max mem: 9672 +[13:53:24.036327] Epoch: [4] [29/30] eta: 0:00:00 lr: 0.000310 loss: 0.8416 (0.8754) time: 0.1430 data: 0.0001 max mem: 9672 +[13:53:24.161368] Epoch: [4] Total time: 0:00:05 (0.1966 s / it) +[13:53:24.170997] Averaged stats: lr: 0.000310 loss: 0.8416 (0.8754) +[13:53:25.722132] val: [0/5] eta: 0:00:07 loss: 0.0815 (0.0815) time: 1.5397 data: 1.5040 max mem: 9672 +[13:53:25.843271] val: [4/5] eta: 0:00:00 loss: 0.7376 (0.6916) time: 0.3321 data: 0.3009 max mem: 9672 +[13:53:25.961156] val: Total time: 0:00:01 (0.3559 s / it) +[13:53:25.973523] val loss: 0.6916433572769165 +[13:53:25.973728] Accuracy: 0.7266, F1 Score: 0.5123, ROC AUC: 0.9467, Hamming Loss: 0.1094, + Jaccard Score: 0.3993, Precision: 0.5164, Recall: 0.5216, + Average Precision: 0.6861, Kappa: 0.6059, Score: 0.6883 +[13:53:27.769389] Best epoch = 4, Best score = 0.6883 +[13:53:27.838434] log_dir: ./output_logs/retfound +[13:53:29.477712] Epoch: [5] [ 0/30] eta: 0:00:49 lr: 0.000313 loss: 0.7009 (0.7009) time: 1.6382 data: 1.4899 max mem: 9672 +[13:53:32.343619] Epoch: [5] [20/30] eta: 0:00:02 lr: 0.000354 loss: 0.8337 (0.8434) time: 0.1432 data: 0.0001 max mem: 9672 +[13:53:33.633695] Epoch: [5] [29/30] eta: 0:00:00 lr: 0.000373 loss: 0.8548 (0.8620) time: 0.1435 data: 0.0001 max mem: 9672 +[13:53:33.767948] Epoch: [5] Total time: 0:00:05 (0.1976 s / it) +[13:53:33.777701] Averaged stats: lr: 0.000373 loss: 0.8548 (0.8620) +[13:53:35.282414] val: [0/5] eta: 0:00:07 loss: 0.2918 (0.2918) time: 1.4889 data: 1.4513 max mem: 9672 +[13:53:35.440451] val: [4/5] eta: 0:00:00 loss: 0.4999 (0.6219) time: 0.3293 data: 0.2972 max mem: 9672 +[13:53:35.599126] val: Total time: 0:00:01 (0.3613 s / it) +[13:53:35.613524] val loss: 0.6218647718429565 +[13:53:35.613748] Accuracy: 0.7482, F1 Score: 0.5577, ROC AUC: 0.9565, Hamming Loss: 0.1007, + Jaccard Score: 0.4426, Precision: 0.5528, Recall: 0.5717, + Average Precision: 0.7192, Kappa: 0.6425, Score: 0.7189 +[13:53:37.329372] Best epoch = 5, Best score = 0.7189 +[13:53:37.393477] log_dir: ./output_logs/retfound +[13:53:39.093702] Epoch: [6] [ 0/30] eta: 0:00:50 lr: 0.000375 loss: 0.8696 (0.8696) time: 1.6991 data: 1.5524 max mem: 9672 +[13:53:41.514415] Epoch: [6] [20/30] eta: 0:00:01 lr: 0.000417 loss: 0.7520 (0.8091) time: 0.1210 data: 0.0002 max mem: 9672 +[13:53:42.799659] Epoch: [6] [29/30] eta: 0:00:00 lr: 0.000435 loss: 0.7895 (0.8082) time: 0.1211 data: 0.0001 max mem: 9672 +[13:53:42.942987] Epoch: [6] Total time: 0:00:05 (0.1850 s / it) +[13:53:42.951998] Averaged stats: lr: 0.000435 loss: 0.7895 (0.8082) +[13:53:44.562029] val: [0/5] eta: 0:00:07 loss: 0.4487 (0.4487) time: 1.5993 data: 1.5625 max mem: 9672 +[13:53:44.693018] val: [4/5] eta: 0:00:00 loss: 0.5246 (0.6279) time: 0.3459 data: 0.3134 max mem: 9672 +[13:53:44.818406] val: Total time: 0:00:01 (0.3713 s / it) +[13:53:44.832136] val loss: 0.6278911352157592 +[13:53:44.832332] Accuracy: 0.7554, F1 Score: 0.6480, ROC AUC: 0.9543, Hamming Loss: 0.0978, + Jaccard Score: 0.5060, Precision: 0.7723, Recall: 0.6886, + Average Precision: 0.7111, Kappa: 0.6566, Score: 0.7530 +[13:53:46.493479] Best epoch = 6, Best score = 0.7530 +[13:53:46.559288] log_dir: ./output_logs/retfound +[13:53:48.236270] Epoch: [7] [ 0/30] eta: 0:00:50 lr: 0.000438 loss: 0.6355 (0.6355) time: 1.6757 data: 1.5265 max mem: 9672 +[13:53:51.099999] Epoch: [7] [20/30] eta: 0:00:02 lr: 0.000479 loss: 0.7854 (0.7973) time: 0.1431 data: 0.0002 max mem: 9672 +[13:53:52.385791] Epoch: [7] [29/30] eta: 0:00:00 lr: 0.000498 loss: 0.7602 (0.7842) time: 0.1429 data: 0.0001 max mem: 9672 +[13:53:52.532112] Epoch: [7] Total time: 0:00:05 (0.1991 s / it) +[13:53:52.541465] Averaged stats: lr: 0.000498 loss: 0.7602 (0.7842) +[13:53:54.203380] val: [0/5] eta: 0:00:08 loss: 0.1889 (0.1889) time: 1.6457 data: 1.6110 max mem: 9672 +[13:53:54.354770] val: [4/5] eta: 0:00:00 loss: 0.2889 (0.5456) time: 0.3593 data: 0.3281 max mem: 9672 +[13:53:54.482442] val: Total time: 0:00:01 (0.3851 s / it) +[13:53:54.499814] val loss: 0.5456351161003112 +[13:53:54.500004] Accuracy: 0.8201, F1 Score: 0.6267, ROC AUC: 0.9664, Hamming Loss: 0.0719, + Jaccard Score: 0.5277, Precision: 0.6232, Recall: 0.6954, + Average Precision: 0.7732, Kappa: 0.7430, Score: 0.7787 +[13:53:56.159043] Best epoch = 7, Best score = 0.7787 +[13:53:56.225705] log_dir: ./output_logs/retfound +[13:53:57.982043] Epoch: [8] [ 0/30] eta: 0:00:52 lr: 0.000500 loss: 0.5424 (0.5424) time: 1.7551 data: 1.6074 max mem: 9672 +[13:54:00.857172] Epoch: [8] [20/30] eta: 0:00:02 lr: 0.000542 loss: 0.7756 (0.7736) time: 0.1437 data: 0.0001 max mem: 9672 +[13:54:02.151732] Epoch: [8] [29/30] eta: 0:00:00 lr: 0.000560 loss: 0.7182 (0.7558) time: 0.1438 data: 0.0001 max mem: 9672 +[13:54:02.295264] Epoch: [8] Total time: 0:00:06 (0.2023 s / it) +[13:54:02.304568] Averaged stats: lr: 0.000560 loss: 0.7182 (0.7558) +[13:54:03.936091] val: [0/5] eta: 0:00:08 loss: 0.5778 (0.5778) time: 1.6156 data: 1.5789 max mem: 9672 +[13:54:04.057530] val: [4/5] eta: 0:00:00 loss: 0.5778 (0.5968) time: 0.3473 data: 0.3159 max mem: 9672 +[13:54:04.180376] val: Total time: 0:00:01 (0.3721 s / it) +[13:54:04.192514] val loss: 0.5968305110931397 +[13:54:04.192701] Accuracy: 0.7698, F1 Score: 0.6777, ROC AUC: 0.9678, Hamming Loss: 0.0921, + Jaccard Score: 0.5297, Precision: 0.7440, Recall: 0.6551, + Average Precision: 0.8180, Kappa: 0.6697, Score: 0.7717 +[13:54:04.235163] Best epoch = 7, Best score = 0.7787 +[13:54:04.481487] log_dir: ./output_logs/retfound +[13:54:06.235585] Epoch: [9] [ 0/30] eta: 0:00:52 lr: 0.000562 loss: 0.8186 (0.8186) time: 1.7527 data: 1.6067 max mem: 9672 +[13:54:09.117315] Epoch: [9] [20/30] eta: 0:00:02 lr: 0.000604 loss: 0.7689 (0.7893) time: 0.1440 data: 0.0002 max mem: 9672 +[13:54:10.410063] Epoch: [9] [29/30] eta: 0:00:00 lr: 0.000623 loss: 0.6713 (0.7591) time: 0.1441 data: 0.0001 max mem: 9672 +[13:54:10.559630] Epoch: [9] Total time: 0:00:06 (0.2026 s / it) +[13:54:10.567981] Averaged stats: lr: 0.000623 loss: 0.6713 (0.7591) +[13:54:12.168246] val: [0/5] eta: 0:00:07 loss: 0.4485 (0.4485) time: 1.5848 data: 1.5481 max mem: 9672 +[13:54:12.304298] val: [4/5] eta: 0:00:00 loss: 0.4550 (0.5600) time: 0.3441 data: 0.3126 max mem: 9672 +[13:54:12.433565] val: Total time: 0:00:01 (0.3702 s / it) +[13:54:12.446012] val loss: 0.5600282669067382 +[13:54:12.446217] Accuracy: 0.7842, F1 Score: 0.6251, ROC AUC: 0.9738, Hamming Loss: 0.0863, + Jaccard Score: 0.5153, Precision: 0.6158, Recall: 0.6798, + Average Precision: 0.8590, Kappa: 0.6907, Score: 0.7632 +[13:54:12.487949] Best epoch = 7, Best score = 0.7787 +[13:54:12.770835] log_dir: ./output_logs/retfound +[13:54:14.437682] Epoch: [10] [ 0/30] eta: 0:00:49 lr: 0.000625 loss: 0.7567 (0.7567) time: 1.6658 data: 1.5230 max mem: 9672 +[13:54:17.308702] Epoch: [10] [20/30] eta: 0:00:02 lr: 0.000625 loss: 0.6977 (0.7326) time: 0.1435 data: 0.0002 max mem: 9672 +[13:54:18.603539] Epoch: [10] [29/30] eta: 0:00:00 lr: 0.000624 loss: 0.6977 (0.7279) time: 0.1439 data: 0.0001 max mem: 9672 +[13:54:18.750524] Epoch: [10] Total time: 0:00:05 (0.1993 s / it) +[13:54:18.758498] Averaged stats: lr: 0.000624 loss: 0.6977 (0.7279) +[13:54:20.444717] val: [0/5] eta: 0:00:08 loss: 0.1891 (0.1891) time: 1.6705 data: 1.6358 max mem: 9672 +[13:54:20.566095] val: [4/5] eta: 0:00:00 loss: 0.3332 (0.6487) time: 0.3582 data: 0.3273 max mem: 9672 +[13:54:20.694983] val: Total time: 0:00:01 (0.3843 s / it) +[13:54:20.707189] val loss: 0.648695421218872 +[13:54:20.707361] Accuracy: 0.8201, F1 Score: 0.6208, ROC AUC: 0.9677, Hamming Loss: 0.0719, + Jaccard Score: 0.5220, Precision: 0.6080, Recall: 0.6505, + Average Precision: 0.8104, Kappa: 0.7426, Score: 0.7770 +[13:54:20.748824] Best epoch = 7, Best score = 0.7787 +[13:54:20.996182] log_dir: ./output_logs/retfound +[13:54:22.692171] Epoch: [11] [ 0/30] eta: 0:00:50 lr: 0.000624 loss: 0.3874 (0.3874) time: 1.6950 data: 1.5496 max mem: 9672 +[13:54:25.561904] Epoch: [11] [20/30] eta: 0:00:02 lr: 0.000622 loss: 0.7793 (0.7278) time: 0.1434 data: 0.0001 max mem: 9672 +[13:54:26.853436] Epoch: [11] [29/30] eta: 0:00:00 lr: 0.000621 loss: 0.7617 (0.7303) time: 0.1433 data: 0.0001 max mem: 9672 +[13:54:27.006385] Epoch: [11] Total time: 0:00:06 (0.2003 s / it) +[13:54:27.008118] Averaged stats: lr: 0.000621 loss: 0.7617 (0.7303) +[13:54:28.561409] val: [0/5] eta: 0:00:07 loss: 0.0880 (0.0880) time: 1.5375 data: 1.5003 max mem: 9672 +[13:54:28.695726] val: [4/5] eta: 0:00:00 loss: 0.5029 (0.5922) time: 0.3342 data: 0.3024 max mem: 9672 +[13:54:28.826291] val: Total time: 0:00:01 (0.3607 s / it) +[13:54:28.838450] val loss: 0.5922170877456665 +[13:54:28.838647] Accuracy: 0.7986, F1 Score: 0.6338, ROC AUC: 0.9659, Hamming Loss: 0.0806, + Jaccard Score: 0.5291, Precision: 0.6328, Recall: 0.6443, + Average Precision: 0.8025, Kappa: 0.7125, Score: 0.7707 +[13:54:28.881759] Best epoch = 7, Best score = 0.7787 +[13:54:29.160978] log_dir: ./output_logs/retfound +[13:54:30.970376] Epoch: [12] [ 0/30] eta: 0:00:54 lr: 0.000621 loss: 0.8659 (0.8659) time: 1.8083 data: 1.6614 max mem: 9672 +[13:54:33.829071] Epoch: [12] [20/30] eta: 0:00:02 lr: 0.000618 loss: 0.6847 (0.7020) time: 0.1429 data: 0.0002 max mem: 9672 +[13:54:35.118821] Epoch: [12] [29/30] eta: 0:00:00 lr: 0.000617 loss: 0.6365 (0.7006) time: 0.1431 data: 0.0001 max mem: 9672 +[13:54:35.252592] Epoch: [12] Total time: 0:00:06 (0.2030 s / it) +[13:54:35.262240] Averaged stats: lr: 0.000617 loss: 0.6365 (0.7006) +[13:54:36.929296] val: [0/5] eta: 0:00:08 loss: 0.0861 (0.0861) time: 1.6546 data: 1.6168 max mem: 9672 +[13:54:37.050494] val: [4/5] eta: 0:00:00 loss: 0.3604 (0.4726) time: 0.3550 data: 0.3235 max mem: 9672 +[13:54:37.181012] val: Total time: 0:00:01 (0.3814 s / it) +[13:54:37.193352] val loss: 0.4726423978805542 +[13:54:37.193531] Accuracy: 0.8489, F1 Score: 0.8018, ROC AUC: 0.9697, Hamming Loss: 0.0604, + Jaccard Score: 0.6812, Precision: 0.8683, Recall: 0.7936, + Average Precision: 0.8305, Kappa: 0.7836, Score: 0.8517 +[13:54:38.902413] Best epoch = 12, Best score = 0.8517 +[13:54:38.973679] log_dir: ./output_logs/retfound +[13:54:40.770741] Epoch: [13] [ 0/30] eta: 0:00:53 lr: 0.000616 loss: 0.6513 (0.6513) time: 1.7961 data: 1.6532 max mem: 9672 +[13:54:43.304960] Epoch: [13] [20/30] eta: 0:00:02 lr: 0.000612 loss: 0.7097 (0.7256) time: 0.1267 data: 0.0001 max mem: 9672 +[13:54:44.598988] Epoch: [13] [29/30] eta: 0:00:00 lr: 0.000610 loss: 0.6510 (0.7137) time: 0.1434 data: 0.0001 max mem: 9672 +[13:54:44.747892] Epoch: [13] Total time: 0:00:05 (0.1925 s / it) +[13:54:44.756634] Averaged stats: lr: 0.000610 loss: 0.6510 (0.7137) +[13:54:46.255077] val: [0/5] eta: 0:00:07 loss: 0.1827 (0.1827) time: 1.4873 data: 1.4517 max mem: 9672 +[13:54:46.440042] val: [4/5] eta: 0:00:00 loss: 0.3526 (0.4675) time: 0.3343 data: 0.3030 max mem: 9672 +[13:54:46.570921] val: Total time: 0:00:01 (0.3607 s / it) +[13:54:46.583550] val loss: 0.4674602270126343 +[13:54:46.583717] Accuracy: 0.8417, F1 Score: 0.7416, ROC AUC: 0.9727, Hamming Loss: 0.0633, + Jaccard Score: 0.6127, Precision: 0.7526, Recall: 0.7344, + Average Precision: 0.8298, Kappa: 0.7762, Score: 0.8302 +[13:54:46.624936] Best epoch = 12, Best score = 0.8517 +[13:54:46.882562] log_dir: ./output_logs/retfound +[13:54:48.663117] Epoch: [14] [ 0/30] eta: 0:00:53 lr: 0.000610 loss: 0.8709 (0.8709) time: 1.7796 data: 1.6337 max mem: 9672 +[13:54:51.535123] Epoch: [14] [20/30] eta: 0:00:02 lr: 0.000604 loss: 0.6729 (0.6818) time: 0.1436 data: 0.0002 max mem: 9672 +[13:54:52.828017] Epoch: [14] [29/30] eta: 0:00:00 lr: 0.000602 loss: 0.6481 (0.6677) time: 0.1429 data: 0.0001 max mem: 9672 +[13:54:52.969984] Epoch: [14] Total time: 0:00:06 (0.2029 s / it) +[13:54:52.979772] Averaged stats: lr: 0.000602 loss: 0.6481 (0.6677) +[13:54:54.648640] val: [0/5] eta: 0:00:08 loss: 0.0811 (0.0811) time: 1.6614 data: 1.6242 max mem: 9672 +[13:54:54.769599] val: [4/5] eta: 0:00:00 loss: 0.4197 (0.4195) time: 0.3564 data: 0.3250 max mem: 9672 +[13:54:54.894878] val: Total time: 0:00:01 (0.3817 s / it) +[13:54:54.906937] val loss: 0.4194536328315735 +[13:54:54.907225] Accuracy: 0.8561, F1 Score: 0.8074, ROC AUC: 0.9747, Hamming Loss: 0.0576, + Jaccard Score: 0.6897, Precision: 0.8651, Recall: 0.8033, + Average Precision: 0.8744, Kappa: 0.7951, Score: 0.8591 +[13:54:56.701878] Best epoch = 14, Best score = 0.8591 +[13:54:56.764044] log_dir: ./output_logs/retfound +[13:54:58.582426] Epoch: [15] [ 0/30] eta: 0:00:54 lr: 0.000601 loss: 0.7167 (0.7167) time: 1.8173 data: 1.6699 max mem: 9672 +[13:55:01.450540] Epoch: [15] [20/30] eta: 0:00:02 lr: 0.000595 loss: 0.6171 (0.6704) time: 0.1434 data: 0.0002 max mem: 9672 +[13:55:02.740712] Epoch: [15] [29/30] eta: 0:00:00 lr: 0.000591 loss: 0.6345 (0.6528) time: 0.1433 data: 0.0001 max mem: 9672 +[13:55:02.892095] Epoch: [15] Total time: 0:00:06 (0.2043 s / it) +[13:55:02.900789] Averaged stats: lr: 0.000591 loss: 0.6345 (0.6528) +[13:55:04.515293] val: [0/5] eta: 0:00:07 loss: 0.1534 (0.1534) time: 1.5987 data: 1.5632 max mem: 9672 +[13:55:04.635950] val: [4/5] eta: 0:00:00 loss: 0.2817 (0.5885) time: 0.3438 data: 0.3128 max mem: 9672 +[13:55:04.757712] val: Total time: 0:00:01 (0.3684 s / it) +[13:55:04.770760] val loss: 0.5885109603404999 +[13:55:04.771034] Accuracy: 0.7986, F1 Score: 0.6104, ROC AUC: 0.9644, Hamming Loss: 0.0806, + Jaccard Score: 0.5070, Precision: 0.6381, Recall: 0.5980, + Average Precision: 0.7879, Kappa: 0.7090, Score: 0.7612 +[13:55:04.804901] Best epoch = 14, Best score = 0.8591 +[13:55:05.084091] log_dir: ./output_logs/retfound +[13:55:06.892525] Epoch: [16] [ 0/30] eta: 0:00:54 lr: 0.000591 loss: 0.5479 (0.5479) time: 1.8073 data: 1.6625 max mem: 9672 +[13:55:09.757596] Epoch: [16] [20/30] eta: 0:00:02 lr: 0.000583 loss: 0.6558 (0.6492) time: 0.1432 data: 0.0001 max mem: 9672 +[13:55:11.054807] Epoch: [16] [29/30] eta: 0:00:00 lr: 0.000579 loss: 0.5857 (0.6654) time: 0.1434 data: 0.0001 max mem: 9672 +[13:55:11.191949] Epoch: [16] Total time: 0:00:06 (0.2036 s / it) +[13:55:11.199982] Averaged stats: lr: 0.000579 loss: 0.5857 (0.6654) +[13:55:12.935193] val: [0/5] eta: 0:00:08 loss: 0.3198 (0.3198) time: 1.7191 data: 1.6843 max mem: 9672 +[13:55:13.056946] val: [4/5] eta: 0:00:00 loss: 0.3819 (0.5172) time: 0.3681 data: 0.3370 max mem: 9672 +[13:55:13.184012] val: Total time: 0:00:01 (0.3937 s / it) +[13:55:13.199552] val loss: 0.5172239661216735 +[13:55:13.199735] Accuracy: 0.8129, F1 Score: 0.7301, ROC AUC: 0.9738, Hamming Loss: 0.0748, + Jaccard Score: 0.5954, Precision: 0.8323, Recall: 0.7425, + Average Precision: 0.8946, Kappa: 0.7324, Score: 0.8121 +[13:55:13.241691] Best epoch = 14, Best score = 0.8591 +[13:55:13.489613] log_dir: ./output_logs/retfound +[13:55:15.307550] Epoch: [17] [ 0/30] eta: 0:00:54 lr: 0.000579 loss: 0.4765 (0.4765) time: 1.8169 data: 1.6714 max mem: 9672 +[13:55:18.179372] Epoch: [17] [20/30] eta: 0:00:02 lr: 0.000570 loss: 0.6074 (0.6307) time: 0.1435 data: 0.0002 max mem: 9672 +[13:55:19.478249] Epoch: [17] [29/30] eta: 0:00:00 lr: 0.000566 loss: 0.6851 (0.6641) time: 0.1438 data: 0.0002 max mem: 9672 +[13:55:19.638006] Epoch: [17] Total time: 0:00:06 (0.2049 s / it) +[13:55:19.647307] Averaged stats: lr: 0.000566 loss: 0.6851 (0.6641) +[13:55:21.206841] val: [0/5] eta: 0:00:07 loss: 0.1685 (0.1685) time: 1.5491 data: 1.5143 max mem: 9672 +[13:55:21.466522] val: [4/5] eta: 0:00:00 loss: 0.2938 (0.4741) time: 0.3616 data: 0.3305 max mem: 9672 +[13:55:21.597277] val: Total time: 0:00:01 (0.3880 s / it) +[13:55:21.615762] val loss: 0.47414124608039854 +[13:55:21.616014] Accuracy: 0.8201, F1 Score: 0.7632, ROC AUC: 0.9748, Hamming Loss: 0.0719, + Jaccard Score: 0.6261, Precision: 0.8094, Recall: 0.7372, + Average Precision: 0.8957, Kappa: 0.7414, Score: 0.8265 +[13:55:21.660848] Best epoch = 14, Best score = 0.8591 +[13:55:21.940329] log_dir: ./output_logs/retfound +[13:55:23.655227] Epoch: [18] [ 0/30] eta: 0:00:51 lr: 0.000565 loss: 0.6698 (0.6698) time: 1.7139 data: 1.5695 max mem: 9672 +[13:55:26.515183] Epoch: [18] [20/30] eta: 0:00:02 lr: 0.000555 loss: 0.6744 (0.6748) time: 0.1430 data: 0.0002 max mem: 9672 +[13:55:27.796675] Epoch: [18] [29/30] eta: 0:00:00 lr: 0.000551 loss: 0.6717 (0.6688) time: 0.1427 data: 0.0001 max mem: 9672 +[13:55:27.925926] Epoch: [18] Total time: 0:00:05 (0.1995 s / it) +[13:55:27.935558] Averaged stats: lr: 0.000551 loss: 0.6717 (0.6688) +[13:55:29.672166] val: [0/5] eta: 0:00:08 loss: 0.1137 (0.1137) time: 1.7246 data: 1.6885 max mem: 9672 +[13:55:29.793897] val: [4/5] eta: 0:00:00 loss: 0.3431 (0.5711) time: 0.3692 data: 0.3378 max mem: 9672 +[13:55:29.919701] val: Total time: 0:00:01 (0.3945 s / it) +[13:55:29.931675] val loss: 0.5710845947265625 +[13:55:29.931900] Accuracy: 0.8273, F1 Score: 0.6816, ROC AUC: 0.9698, Hamming Loss: 0.0691, + Jaccard Score: 0.6001, Precision: 0.6638, Recall: 0.7049, + Average Precision: 0.8249, Kappa: 0.7507, Score: 0.8007 +[13:55:29.984949] Best epoch = 14, Best score = 0.8591 +[13:55:30.228132] log_dir: ./output_logs/retfound +[13:55:31.928804] Epoch: [19] [ 0/30] eta: 0:00:50 lr: 0.000550 loss: 0.9024 (0.9024) time: 1.6997 data: 1.5538 max mem: 9672 +[13:55:34.800922] Epoch: [19] [20/30] eta: 0:00:02 lr: 0.000539 loss: 0.6465 (0.6793) time: 0.1435 data: 0.0001 max mem: 9672 +[13:55:36.086470] Epoch: [19] [29/30] eta: 0:00:00 lr: 0.000534 loss: 0.6717 (0.6771) time: 0.1430 data: 0.0001 max mem: 9672 +[13:55:36.238386] Epoch: [19] Total time: 0:00:06 (0.2003 s / it) +[13:55:36.246408] Averaged stats: lr: 0.000534 loss: 0.6717 (0.6771) +[13:55:37.906342] val: [0/5] eta: 0:00:08 loss: 0.0875 (0.0875) time: 1.6497 data: 1.6133 max mem: 9672 +[13:55:38.110695] val: [4/5] eta: 0:00:00 loss: 0.4873 (0.4622) time: 0.3707 data: 0.3389 max mem: 9672 +[13:55:38.239377] val: Total time: 0:00:01 (0.3967 s / it) +[13:55:38.251612] val loss: 0.4621611088514328 +[13:55:38.251822] Accuracy: 0.8345, F1 Score: 0.6639, ROC AUC: 0.9755, Hamming Loss: 0.0662, + Jaccard Score: 0.5711, Precision: 0.6265, Recall: 0.7233, + Average Precision: 0.8799, Kappa: 0.7655, Score: 0.8016 +[13:55:38.295245] Best epoch = 14, Best score = 0.8591 +[13:55:38.528971] log_dir: ./output_logs/retfound +[13:55:40.324429] Epoch: [20] [ 0/30] eta: 0:00:53 lr: 0.000534 loss: 0.9199 (0.9199) time: 1.7945 data: 1.6465 max mem: 9672 +[13:55:43.196093] Epoch: [20] [20/30] eta: 0:00:02 lr: 0.000522 loss: 0.5783 (0.6463) time: 0.1435 data: 0.0001 max mem: 9672 +[13:55:44.485340] Epoch: [20] [29/30] eta: 0:00:00 lr: 0.000516 loss: 0.6143 (0.6487) time: 0.1434 data: 0.0001 max mem: 9672 +[13:55:44.823206] Epoch: [20] Total time: 0:00:06 (0.2098 s / it) +[13:55:44.832035] Averaged stats: lr: 0.000516 loss: 0.6143 (0.6487) +[13:55:46.467848] val: [0/5] eta: 0:00:08 loss: 0.2576 (0.2576) time: 1.6202 data: 1.5824 max mem: 9672 +[13:55:46.590043] val: [4/5] eta: 0:00:00 loss: 0.3210 (0.4680) time: 0.3483 data: 0.3166 max mem: 9672 +[13:55:46.719252] val: Total time: 0:00:01 (0.3745 s / it) +[13:55:46.731665] val loss: 0.4679856836795807 +[13:55:46.731834] Accuracy: 0.8489, F1 Score: 0.7550, ROC AUC: 0.9748, Hamming Loss: 0.0604, + Jaccard Score: 0.6361, Precision: 0.7943, Recall: 0.7507, + Average Precision: 0.8667, Kappa: 0.7851, Score: 0.8383 +[13:55:46.776732] Best epoch = 14, Best score = 0.8591 +[13:55:46.999542] log_dir: ./output_logs/retfound +[13:55:48.708325] Epoch: [21] [ 0/30] eta: 0:00:51 lr: 0.000516 loss: 0.5436 (0.5436) time: 1.7078 data: 1.5579 max mem: 9672 +[13:55:51.578544] Epoch: [21] [20/30] eta: 0:00:02 lr: 0.000503 loss: 0.5815 (0.6056) time: 0.1435 data: 0.0001 max mem: 9672 +[13:55:52.857476] Epoch: [21] [29/30] eta: 0:00:00 lr: 0.000497 loss: 0.5773 (0.6117) time: 0.1427 data: 0.0001 max mem: 9672 +[13:55:53.013322] Epoch: [21] Total time: 0:00:06 (0.2005 s / it) +[13:55:53.021938] Averaged stats: lr: 0.000497 loss: 0.5773 (0.6117) +[13:55:54.738222] val: [0/5] eta: 0:00:08 loss: 0.0253 (0.0253) time: 1.6985 data: 1.6623 max mem: 9672 +[13:55:54.859950] val: [4/5] eta: 0:00:00 loss: 0.5873 (0.5458) time: 0.3639 data: 0.3326 max mem: 9672 +[13:55:54.988153] val: Total time: 0:00:01 (0.3898 s / it) +[13:55:55.002590] val loss: 0.5458193361759186 +[13:55:55.002920] Accuracy: 0.7914, F1 Score: 0.7277, ROC AUC: 0.9712, Hamming Loss: 0.0835, + Jaccard Score: 0.5863, Precision: 0.7626, Recall: 0.7138, + Average Precision: 0.8389, Kappa: 0.7044, Score: 0.8011 +[13:55:55.046581] Best epoch = 14, Best score = 0.8591 +[13:55:55.282406] log_dir: ./output_logs/retfound +[13:55:57.094462] Epoch: [22] [ 0/30] eta: 0:00:54 lr: 0.000496 loss: 0.5792 (0.5792) time: 1.8111 data: 1.6662 max mem: 9672 +[13:55:59.965067] Epoch: [22] [20/30] eta: 0:00:02 lr: 0.000483 loss: 0.7030 (0.6881) time: 0.1435 data: 0.0001 max mem: 9672 +[13:56:01.263790] Epoch: [22] [29/30] eta: 0:00:00 lr: 0.000477 loss: 0.6064 (0.6590) time: 0.1435 data: 0.0001 max mem: 9672 +[13:56:01.405586] Epoch: [22] Total time: 0:00:06 (0.2041 s / it) +[13:56:01.415363] Averaged stats: lr: 0.000477 loss: 0.6064 (0.6590) +[13:56:02.907293] val: [0/5] eta: 0:00:07 loss: 0.2673 (0.2673) time: 1.4763 data: 1.4394 max mem: 9672 +[13:56:03.140283] val: [4/5] eta: 0:00:00 loss: 0.4009 (0.4953) time: 0.3417 data: 0.3099 max mem: 9672 +[13:56:03.270753] val: Total time: 0:00:01 (0.3682 s / it) +[13:56:03.284249] val loss: 0.4953368067741394 +[13:56:03.284438] Accuracy: 0.8345, F1 Score: 0.7589, ROC AUC: 0.9729, Hamming Loss: 0.0662, + Jaccard Score: 0.6264, Precision: 0.8228, Recall: 0.7330, + Average Precision: 0.8429, Kappa: 0.7652, Score: 0.8323 +[13:56:03.324072] Best epoch = 14, Best score = 0.8591 +[13:56:03.564730] log_dir: ./output_logs/retfound +[13:56:05.338862] Epoch: [23] [ 0/30] eta: 0:00:53 lr: 0.000476 loss: 0.6561 (0.6561) time: 1.7732 data: 1.6260 max mem: 9672 +[13:56:08.209717] Epoch: [23] [20/30] eta: 0:00:02 lr: 0.000462 loss: 0.5443 (0.5608) time: 0.1435 data: 0.0002 max mem: 9672 +[13:56:09.505121] Epoch: [23] [29/30] eta: 0:00:00 lr: 0.000455 loss: 0.5481 (0.5752) time: 0.1440 data: 0.0001 max mem: 9672 +[13:56:09.656300] Epoch: [23] Total time: 0:00:06 (0.2030 s / it) +[13:56:09.664370] Averaged stats: lr: 0.000455 loss: 0.5481 (0.5752) +[13:56:11.363521] val: [0/5] eta: 0:00:08 loss: 0.0842 (0.0842) time: 1.6864 data: 1.6512 max mem: 9672 +[13:56:11.486292] val: [4/5] eta: 0:00:00 loss: 0.3760 (0.4528) time: 0.3617 data: 0.3304 max mem: 9672 +[13:56:11.612775] val: Total time: 0:00:01 (0.3872 s / it) +[13:56:11.625154] val loss: 0.4528108716011047 +[13:56:11.625349] Accuracy: 0.8489, F1 Score: 0.7535, ROC AUC: 0.9773, Hamming Loss: 0.0604, + Jaccard Score: 0.6345, Precision: 0.7888, Recall: 0.7510, + Average Precision: 0.8894, Kappa: 0.7854, Score: 0.8388 +[13:56:11.675606] Best epoch = 14, Best score = 0.8591 +[13:56:11.907987] log_dir: ./output_logs/retfound +[13:56:13.657545] Epoch: [24] [ 0/30] eta: 0:00:52 lr: 0.000455 loss: 0.6420 (0.6420) time: 1.7484 data: 1.5992 max mem: 9672 +[13:56:16.524697] Epoch: [24] [20/30] eta: 0:00:02 lr: 0.000440 loss: 0.5988 (0.6247) time: 0.1433 data: 0.0002 max mem: 9672 +[13:56:17.815535] Epoch: [24] [29/30] eta: 0:00:00 lr: 0.000433 loss: 0.5581 (0.6024) time: 0.1436 data: 0.0001 max mem: 9672 +[13:56:17.957182] Epoch: [24] Total time: 0:00:06 (0.2016 s / it) +[13:56:17.965315] Averaged stats: lr: 0.000433 loss: 0.5581 (0.6024) +[13:56:19.619011] val: [0/5] eta: 0:00:08 loss: 0.1161 (0.1161) time: 1.6395 data: 1.6028 max mem: 9672 +[13:56:19.741257] val: [4/5] eta: 0:00:00 loss: 0.4242 (0.4765) time: 0.3522 data: 0.3207 max mem: 9672 +[13:56:19.867429] val: Total time: 0:00:01 (0.3777 s / it) +[13:56:19.879744] val loss: 0.4765301883220673 +[13:56:19.879950] Accuracy: 0.8345, F1 Score: 0.6984, ROC AUC: 0.9754, Hamming Loss: 0.0662, + Jaccard Score: 0.5871, Precision: 0.7293, Recall: 0.6863, + Average Precision: 0.8506, Kappa: 0.7638, Score: 0.8126 +[13:56:19.918364] Best epoch = 14, Best score = 0.8591 +[13:56:20.157709] log_dir: ./output_logs/retfound +[13:56:21.887250] Epoch: [25] [ 0/30] eta: 0:00:51 lr: 0.000432 loss: 0.5540 (0.5540) time: 1.7284 data: 1.5840 max mem: 9672 +[13:56:24.758213] Epoch: [25] [20/30] eta: 0:00:02 lr: 0.000417 loss: 0.5538 (0.5667) time: 0.1435 data: 0.0002 max mem: 9672 +[13:56:26.048374] Epoch: [25] [29/30] eta: 0:00:00 lr: 0.000410 loss: 0.5538 (0.5709) time: 0.1433 data: 0.0001 max mem: 9672 +[13:56:26.192593] Epoch: [25] Total time: 0:00:06 (0.2012 s / it) +[13:56:26.202223] Averaged stats: lr: 0.000410 loss: 0.5538 (0.5709) +[13:56:28.043649] val: [0/5] eta: 0:00:09 loss: 0.1237 (0.1237) time: 1.8271 data: 1.7916 max mem: 9672 +[13:56:28.165062] val: [4/5] eta: 0:00:00 loss: 0.4356 (0.5354) time: 0.3896 data: 0.3584 max mem: 9672 +[13:56:28.292777] val: Total time: 0:00:02 (0.4154 s / it) +[13:56:28.309749] val loss: 0.5353950619697571 +[13:56:28.309954] Accuracy: 0.8273, F1 Score: 0.6715, ROC AUC: 0.9740, Hamming Loss: 0.0691, + Jaccard Score: 0.5515, Precision: 0.7981, Recall: 0.6678, + Average Precision: 0.8369, Kappa: 0.7543, Score: 0.7999 +[13:56:28.346468] Best epoch = 14, Best score = 0.8591 +[13:56:28.588427] log_dir: ./output_logs/retfound +[13:56:30.356312] Epoch: [26] [ 0/30] eta: 0:00:53 lr: 0.000409 loss: 1.0587 (1.0587) time: 1.7668 data: 1.6215 max mem: 9672 +[13:56:33.211762] Epoch: [26] [20/30] eta: 0:00:02 lr: 0.000394 loss: 0.5588 (0.5982) time: 0.1427 data: 0.0001 max mem: 9672 +[13:56:34.506962] Epoch: [26] [29/30] eta: 0:00:00 lr: 0.000387 loss: 0.6390 (0.5844) time: 0.1429 data: 0.0001 max mem: 9672 +[13:56:34.660923] Epoch: [26] Total time: 0:00:06 (0.2024 s / it) +[13:56:34.662588] Averaged stats: lr: 0.000387 loss: 0.6390 (0.5844) +[13:56:36.424284] val: [0/5] eta: 0:00:08 loss: 0.1098 (0.1098) time: 1.7463 data: 1.7097 max mem: 9672 +[13:56:36.545646] val: [4/5] eta: 0:00:00 loss: 0.4014 (0.5429) time: 0.3734 data: 0.3420 max mem: 9672 +[13:56:36.682531] val: Total time: 0:00:02 (0.4010 s / it) +[13:56:36.694532] val loss: 0.5428592979907989 +[13:56:36.694751] Accuracy: 0.8058, F1 Score: 0.6398, ROC AUC: 0.9730, Hamming Loss: 0.0777, + Jaccard Score: 0.5178, Precision: 0.7676, Recall: 0.6340, + Average Precision: 0.8341, Kappa: 0.7244, Score: 0.7791 +[13:56:36.729108] Best epoch = 14, Best score = 0.8591 +[13:56:36.994845] log_dir: ./output_logs/retfound +[13:56:38.817617] Epoch: [27] [ 0/30] eta: 0:00:54 lr: 0.000386 loss: 0.6745 (0.6745) time: 1.8215 data: 1.6729 max mem: 9672 +[13:56:41.691389] Epoch: [27] [20/30] eta: 0:00:02 lr: 0.000370 loss: 0.5801 (0.6123) time: 0.1436 data: 0.0002 max mem: 9672 +[13:56:42.982003] Epoch: [27] [29/30] eta: 0:00:00 lr: 0.000363 loss: 0.5324 (0.5989) time: 0.1436 data: 0.0001 max mem: 9672 +[13:56:43.125471] Epoch: [27] Total time: 0:00:06 (0.2043 s / it) +[13:56:43.133878] Averaged stats: lr: 0.000363 loss: 0.5324 (0.5989) +[13:56:44.785044] val: [0/5] eta: 0:00:08 loss: 0.1944 (0.1944) time: 1.6381 data: 1.6024 max mem: 9672 +[13:56:44.908694] val: [4/5] eta: 0:00:00 loss: 0.4350 (0.4554) time: 0.3522 data: 0.3206 max mem: 9672 +[13:56:45.044082] val: Total time: 0:00:01 (0.3796 s / it) +[13:56:45.062766] val loss: 0.45543997287750243 +[13:56:45.062994] Accuracy: 0.8273, F1 Score: 0.7163, ROC AUC: 0.9752, Hamming Loss: 0.0691, + Jaccard Score: 0.5848, Precision: 0.8039, Recall: 0.6998, + Average Precision: 0.8582, Kappa: 0.7557, Score: 0.8158 +[13:56:45.098733] Best epoch = 14, Best score = 0.8591 +[13:56:45.332428] log_dir: ./output_logs/retfound +[13:56:47.248235] Epoch: [28] [ 0/30] eta: 0:00:57 lr: 0.000362 loss: 0.3080 (0.3080) time: 1.9148 data: 1.7675 max mem: 9672 +[13:56:50.123002] Epoch: [28] [20/30] eta: 0:00:02 lr: 0.000346 loss: 0.6184 (0.6403) time: 0.1437 data: 0.0002 max mem: 9672 +[13:56:51.416075] Epoch: [28] [29/30] eta: 0:00:00 lr: 0.000338 loss: 0.6131 (0.6174) time: 0.1439 data: 0.0001 max mem: 9672 +[13:56:51.557755] Epoch: [28] Total time: 0:00:06 (0.2075 s / it) +[13:56:51.567501] Averaged stats: lr: 0.000338 loss: 0.6131 (0.6174) +[13:56:53.227653] val: [0/5] eta: 0:00:08 loss: 0.1431 (0.1431) time: 1.6449 data: 1.6080 max mem: 9672 +[13:56:53.349153] val: [4/5] eta: 0:00:00 loss: 0.4351 (0.3940) time: 0.3531 data: 0.3217 max mem: 9672 +[13:56:53.474751] val: Total time: 0:00:01 (0.3785 s / it) +[13:56:53.487235] val loss: 0.3940145254135132 +[13:56:53.487428] Accuracy: 0.8489, F1 Score: 0.8144, ROC AUC: 0.9778, Hamming Loss: 0.0604, + Jaccard Score: 0.6925, Precision: 0.8483, Recall: 0.7952, + Average Precision: 0.9014, Kappa: 0.7867, Score: 0.8596 +[13:56:55.285854] Best epoch = 28, Best score = 0.8596 +[13:56:55.349529] log_dir: ./output_logs/retfound +[13:56:57.102753] Epoch: [29] [ 0/30] eta: 0:00:52 lr: 0.000337 loss: 0.5759 (0.5759) time: 1.7523 data: 1.6065 max mem: 9672 +[13:56:59.979866] Epoch: [29] [20/30] eta: 0:00:02 lr: 0.000321 loss: 0.5183 (0.5784) time: 0.1438 data: 0.0002 max mem: 9672 +[13:57:01.268062] Epoch: [29] [29/30] eta: 0:00:00 lr: 0.000314 loss: 0.5264 (0.5680) time: 0.1434 data: 0.0002 max mem: 9672 +[13:57:01.391333] Epoch: [29] Total time: 0:00:06 (0.2014 s / it) +[13:57:01.400596] Averaged stats: lr: 0.000314 loss: 0.5264 (0.5680) +[13:57:03.043985] val: [0/5] eta: 0:00:08 loss: 0.1238 (0.1238) time: 1.6323 data: 1.5953 max mem: 9672 +[13:57:03.165772] val: [4/5] eta: 0:00:00 loss: 0.4417 (0.4605) time: 0.3507 data: 0.3191 max mem: 9672 +[13:57:03.289372] val: Total time: 0:00:01 (0.3756 s / it) +[13:57:03.302480] val loss: 0.46051245033740995 +[13:57:03.302674] Accuracy: 0.8345, F1 Score: 0.7807, ROC AUC: 0.9741, Hamming Loss: 0.0662, + Jaccard Score: 0.6531, Precision: 0.8154, Recall: 0.7597, + Average Precision: 0.8436, Kappa: 0.7657, Score: 0.8401 +[13:57:03.343107] Best epoch = 28, Best score = 0.8596 +[13:57:03.577188] log_dir: ./output_logs/retfound +[13:57:05.348296] Epoch: [30] [ 0/30] eta: 0:00:53 lr: 0.000313 loss: 0.6743 (0.6743) time: 1.7701 data: 1.6241 max mem: 9672 +[13:57:08.219951] Epoch: [30] [20/30] eta: 0:00:02 lr: 0.000297 loss: 0.5646 (0.5654) time: 0.1435 data: 0.0001 max mem: 9672 +[13:57:09.518032] Epoch: [30] [29/30] eta: 0:00:00 lr: 0.000289 loss: 0.5418 (0.5747) time: 0.1439 data: 0.0001 max mem: 9672 +[13:57:09.663499] Epoch: [30] Total time: 0:00:06 (0.2029 s / it) +[13:57:09.665525] Averaged stats: lr: 0.000289 loss: 0.5418 (0.5747) +[13:57:11.384491] val: [0/5] eta: 0:00:08 loss: 0.1538 (0.1538) time: 1.7050 data: 1.6697 max mem: 9672 +[13:57:11.505859] val: [4/5] eta: 0:00:00 loss: 0.3350 (0.3885) time: 0.3652 data: 0.3340 max mem: 9672 +[13:57:11.630507] val: Total time: 0:00:01 (0.3903 s / it) +[13:57:11.642684] val loss: 0.38845452964305877 +[13:57:11.642946] Accuracy: 0.8417, F1 Score: 0.7871, ROC AUC: 0.9784, Hamming Loss: 0.0633, + Jaccard Score: 0.6617, Precision: 0.8179, Recall: 0.7798, + Average Precision: 0.8968, Kappa: 0.7759, Score: 0.8471 +[13:57:11.685997] Best epoch = 28, Best score = 0.8596 +[13:57:11.927953] log_dir: ./output_logs/retfound +[13:57:13.668184] Epoch: [31] [ 0/30] eta: 0:00:52 lr: 0.000289 loss: 0.7178 (0.7178) time: 1.7392 data: 1.5948 max mem: 9672 +[13:57:16.538160] Epoch: [31] [20/30] eta: 0:00:02 lr: 0.000272 loss: 0.5606 (0.5714) time: 0.1435 data: 0.0001 max mem: 9672 +[13:57:17.829232] Epoch: [31] [29/30] eta: 0:00:00 lr: 0.000265 loss: 0.5779 (0.5695) time: 0.1432 data: 0.0001 max mem: 9672 +[13:57:17.980667] Epoch: [31] Total time: 0:00:06 (0.2018 s / it) +[13:57:17.989213] Averaged stats: lr: 0.000265 loss: 0.5779 (0.5695) +[13:57:19.696181] val: [0/5] eta: 0:00:08 loss: 0.1097 (0.1097) time: 1.6954 data: 1.6610 max mem: 9672 +[13:57:19.817708] val: [4/5] eta: 0:00:00 loss: 0.4424 (0.4829) time: 0.3633 data: 0.3323 max mem: 9672 +[13:57:19.944332] val: Total time: 0:00:01 (0.3888 s / it) +[13:57:19.956543] val loss: 0.4829346999526024 +[13:57:19.956708] Accuracy: 0.8345, F1 Score: 0.6727, ROC AUC: 0.9749, Hamming Loss: 0.0662, + Jaccard Score: 0.5570, Precision: 0.7961, Recall: 0.6748, + Average Precision: 0.8377, Kappa: 0.7656, Score: 0.8044 +[13:57:19.994002] Best epoch = 28, Best score = 0.8596 +[13:57:20.238022] log_dir: ./output_logs/retfound +[13:57:22.045106] Epoch: [32] [ 0/30] eta: 0:00:54 lr: 0.000264 loss: 0.5741 (0.5741) time: 1.8061 data: 1.6604 max mem: 9672 +[13:57:24.912240] Epoch: [32] [20/30] eta: 0:00:02 lr: 0.000248 loss: 0.5943 (0.5526) time: 0.1433 data: 0.0001 max mem: 9672 +[13:57:26.211508] Epoch: [32] [29/30] eta: 0:00:00 lr: 0.000241 loss: 0.5082 (0.5344) time: 0.1435 data: 0.0001 max mem: 9672 +[13:57:26.356401] Epoch: [32] Total time: 0:00:06 (0.2039 s / it) +[13:57:26.365136] Averaged stats: lr: 0.000241 loss: 0.5082 (0.5344) +[13:57:28.015531] val: [0/5] eta: 0:00:08 loss: 0.2951 (0.2951) time: 1.6379 data: 1.6027 max mem: 9672 +[13:57:28.136874] val: [4/5] eta: 0:00:00 loss: 0.3170 (0.5373) time: 0.3517 data: 0.3207 max mem: 9672 +[13:57:28.267249] val: Total time: 0:00:01 (0.3781 s / it) +[13:57:28.280778] val loss: 0.5373015165328979 +[13:57:28.281039] Accuracy: 0.8201, F1 Score: 0.6644, ROC AUC: 0.9746, Hamming Loss: 0.0719, + Jaccard Score: 0.5436, Precision: 0.7939, Recall: 0.6635, + Average Precision: 0.8338, Kappa: 0.7444, Score: 0.7944 +[13:57:28.319384] Best epoch = 28, Best score = 0.8596 +[13:57:28.582891] log_dir: ./output_logs/retfound +[13:57:30.412672] Epoch: [33] [ 0/30] eta: 0:00:54 lr: 0.000240 loss: 0.6181 (0.6181) time: 1.8287 data: 1.6845 max mem: 9672 +[13:57:33.303157] Epoch: [33] [20/30] eta: 0:00:02 lr: 0.000224 loss: 0.5347 (0.5315) time: 0.1445 data: 0.0002 max mem: 9672 +[13:57:34.599446] Epoch: [33] [29/30] eta: 0:00:00 lr: 0.000217 loss: 0.5326 (0.5516) time: 0.1447 data: 0.0001 max mem: 9672 +[13:57:34.745895] Epoch: [33] Total time: 0:00:06 (0.2054 s / it) +[13:57:34.755648] Averaged stats: lr: 0.000217 loss: 0.5326 (0.5516) +[13:57:36.406340] val: [0/5] eta: 0:00:08 loss: 0.3652 (0.3652) time: 1.6352 data: 1.6004 max mem: 9672 +[13:57:36.528051] val: [4/5] eta: 0:00:00 loss: 0.3652 (0.4447) time: 0.3513 data: 0.3202 max mem: 9672 +[13:57:36.652112] val: Total time: 0:00:01 (0.3763 s / it) +[13:57:36.664519] val loss: 0.4446850061416626 +[13:57:36.664727] Accuracy: 0.8417, F1 Score: 0.7880, ROC AUC: 0.9774, Hamming Loss: 0.0633, + Jaccard Score: 0.6629, Precision: 0.8285, Recall: 0.7662, + Average Precision: 0.8879, Kappa: 0.7757, Score: 0.8471 +[13:57:36.698505] Best epoch = 28, Best score = 0.8596 +[13:57:36.934507] log_dir: ./output_logs/retfound +[13:57:38.685784] Epoch: [34] [ 0/30] eta: 0:00:52 lr: 0.000217 loss: 0.5896 (0.5896) time: 1.7502 data: 1.6067 max mem: 9672 +[13:57:41.590230] Epoch: [34] [20/30] eta: 0:00:02 lr: 0.000201 loss: 0.5393 (0.5706) time: 0.1452 data: 0.0001 max mem: 9672 +[13:57:42.894586] Epoch: [34] [29/30] eta: 0:00:00 lr: 0.000194 loss: 0.5329 (0.5663) time: 0.1450 data: 0.0001 max mem: 9672 +[13:57:43.034984] Epoch: [34] Total time: 0:00:06 (0.2033 s / it) +[13:57:43.044483] Averaged stats: lr: 0.000194 loss: 0.5329 (0.5663) +[13:57:44.725912] val: [0/5] eta: 0:00:08 loss: 0.2379 (0.2379) time: 1.6642 data: 1.6268 max mem: 9672 +[13:57:44.848474] val: [4/5] eta: 0:00:00 loss: 0.3058 (0.5136) time: 0.3572 data: 0.3255 max mem: 9672 +[13:57:44.975472] val: Total time: 0:00:01 (0.3829 s / it) +[13:57:44.987507] val loss: 0.5135580390691757 +[13:57:44.987679] Accuracy: 0.8417, F1 Score: 0.7586, ROC AUC: 0.9738, Hamming Loss: 0.0633, + Jaccard Score: 0.6297, Precision: 0.8160, Recall: 0.7371, + Average Precision: 0.8327, Kappa: 0.7755, Score: 0.8360 +[13:57:45.025106] Best epoch = 28, Best score = 0.8596 +[13:57:45.268509] log_dir: ./output_logs/retfound +[13:57:47.000712] Epoch: [35] [ 0/30] eta: 0:00:51 lr: 0.000194 loss: 0.4422 (0.4422) time: 1.7312 data: 1.5972 max mem: 9672 +[13:57:49.900592] Epoch: [35] [20/30] eta: 0:00:02 lr: 0.000179 loss: 0.5515 (0.5603) time: 0.1450 data: 0.0002 max mem: 9672 +[13:57:51.197589] Epoch: [35] [29/30] eta: 0:00:00 lr: 0.000172 loss: 0.5571 (0.5563) time: 0.1444 data: 0.0001 max mem: 9672 +[13:57:51.349570] Epoch: [35] Total time: 0:00:06 (0.2027 s / it) +[13:57:51.357687] Averaged stats: lr: 0.000172 loss: 0.5571 (0.5563) +[13:57:53.065654] val: [0/5] eta: 0:00:08 loss: 0.2287 (0.2287) time: 1.6956 data: 1.6590 max mem: 9672 +[13:57:53.187766] val: [4/5] eta: 0:00:00 loss: 0.3087 (0.4394) time: 0.3634 data: 0.3319 max mem: 9672 +[13:57:53.316717] val: Total time: 0:00:01 (0.3894 s / it) +[13:57:53.328743] val loss: 0.43940497636795045 +[13:57:53.328992] Accuracy: 0.8417, F1 Score: 0.7878, ROC AUC: 0.9757, Hamming Loss: 0.0633, + Jaccard Score: 0.6627, Precision: 0.8242, Recall: 0.7664, + Average Precision: 0.8456, Kappa: 0.7758, Score: 0.8465 +[13:57:53.380337] Best epoch = 28, Best score = 0.8596 +[13:57:53.608899] log_dir: ./output_logs/retfound +[13:57:55.342387] Epoch: [36] [ 0/30] eta: 0:00:51 lr: 0.000171 loss: 0.4897 (0.4897) time: 1.7323 data: 1.5823 max mem: 9672 +[13:57:58.212434] Epoch: [36] [20/30] eta: 0:00:02 lr: 0.000157 loss: 0.4342 (0.5063) time: 0.1435 data: 0.0001 max mem: 9672 +[13:57:59.112566] Epoch: [36] [29/30] eta: 0:00:00 lr: 0.000151 loss: 0.4342 (0.4991) time: 0.1240 data: 0.0001 max mem: 9672 +[13:57:59.239394] Epoch: [36] Total time: 0:00:05 (0.1877 s / it) +[13:57:59.247534] Averaged stats: lr: 0.000151 loss: 0.4342 (0.4991) +[13:58:00.989324] val: [0/5] eta: 0:00:08 loss: 0.3408 (0.3408) time: 1.7308 data: 1.6954 max mem: 9672 +[13:58:01.121227] val: [4/5] eta: 0:00:00 loss: 0.3521 (0.4637) time: 0.3724 data: 0.3411 max mem: 9672 +[13:58:01.244893] val: Total time: 0:00:01 (0.3974 s / it) +[13:58:01.256914] val loss: 0.4637205421924591 +[13:58:01.257070] Accuracy: 0.8345, F1 Score: 0.7926, ROC AUC: 0.9759, Hamming Loss: 0.0662, + Jaccard Score: 0.6648, Precision: 0.8233, Recall: 0.7838, + Average Precision: 0.8636, Kappa: 0.7666, Score: 0.8450 +[13:58:01.295895] Best epoch = 28, Best score = 0.8596 +[13:58:01.542542] log_dir: ./output_logs/retfound +[13:58:03.293901] Epoch: [37] [ 0/30] eta: 0:00:52 lr: 0.000150 loss: 0.7642 (0.7642) time: 1.7503 data: 1.6046 max mem: 9672 +[13:58:06.161990] Epoch: [37] [20/30] eta: 0:00:02 lr: 0.000136 loss: 0.4683 (0.5044) time: 0.1434 data: 0.0001 max mem: 9672 +[13:58:07.459928] Epoch: [37] [29/30] eta: 0:00:00 lr: 0.000130 loss: 0.4476 (0.5092) time: 0.1437 data: 0.0001 max mem: 9672 +[13:58:07.603890] Epoch: [37] Total time: 0:00:06 (0.2020 s / it) +[13:58:07.613572] Averaged stats: lr: 0.000130 loss: 0.4476 (0.5092) +[13:58:09.277291] val: [0/5] eta: 0:00:08 loss: 0.2952 (0.2952) time: 1.6517 data: 1.6158 max mem: 9672 +[13:58:09.398479] val: [4/5] eta: 0:00:00 loss: 0.3085 (0.5188) time: 0.3545 data: 0.3233 max mem: 9672 +[13:58:09.521990] val: Total time: 0:00:01 (0.3794 s / it) +[13:58:09.534122] val loss: 0.5187703669071198 +[13:58:09.534292] Accuracy: 0.8417, F1 Score: 0.7642, ROC AUC: 0.9745, Hamming Loss: 0.0633, + Jaccard Score: 0.6338, Precision: 0.8321, Recall: 0.7342, + Average Precision: 0.8352, Kappa: 0.7741, Score: 0.8376 +[13:58:09.572077] Best epoch = 28, Best score = 0.8596 +[13:58:09.822229] log_dir: ./output_logs/retfound +[13:58:11.636221] Epoch: [38] [ 0/30] eta: 0:00:54 lr: 0.000130 loss: 0.5857 (0.5857) time: 1.8130 data: 1.6670 max mem: 9672 +[13:58:14.501351] Epoch: [38] [20/30] eta: 0:00:02 lr: 0.000117 loss: 0.4936 (0.5329) time: 0.1432 data: 0.0001 max mem: 9672 +[13:58:15.795511] Epoch: [38] [29/30] eta: 0:00:00 lr: 0.000111 loss: 0.5265 (0.5353) time: 0.1433 data: 0.0001 max mem: 9672 +[13:58:15.943173] Epoch: [38] Total time: 0:00:06 (0.2040 s / it) +[13:58:15.951053] Averaged stats: lr: 0.000111 loss: 0.5265 (0.5353) +[13:58:17.641781] val: [0/5] eta: 0:00:08 loss: 0.2033 (0.2033) time: 1.6735 data: 1.6368 max mem: 9672 +[13:58:17.763288] val: [4/5] eta: 0:00:00 loss: 0.3322 (0.4631) time: 0.3589 data: 0.3275 max mem: 9672 +[13:58:17.891955] val: Total time: 0:00:01 (0.3848 s / it) +[13:58:17.904981] val loss: 0.4630659490823746 +[13:58:17.905194] Accuracy: 0.8561, F1 Score: 0.7959, ROC AUC: 0.9756, Hamming Loss: 0.0576, + Jaccard Score: 0.6757, Precision: 0.8312, Recall: 0.7747, + Average Precision: 0.8426, Kappa: 0.7962, Score: 0.8559 +[13:58:17.942324] Best epoch = 28, Best score = 0.8596 +[13:58:18.182294] log_dir: ./output_logs/retfound +[13:58:20.058116] Epoch: [39] [ 0/30] eta: 0:00:56 lr: 0.000110 loss: 0.5891 (0.5891) time: 1.8748 data: 1.7287 max mem: 9672 +[13:58:22.926598] Epoch: [39] [20/30] eta: 0:00:02 lr: 0.000098 loss: 0.4994 (0.5286) time: 0.1434 data: 0.0001 max mem: 9672 +[13:58:24.219747] Epoch: [39] [29/30] eta: 0:00:00 lr: 0.000093 loss: 0.4994 (0.5242) time: 0.1436 data: 0.0001 max mem: 9672 +[13:58:24.372240] Epoch: [39] Total time: 0:00:06 (0.2063 s / it) +[13:58:24.380635] Averaged stats: lr: 0.000093 loss: 0.4994 (0.5242) +[13:58:26.087433] val: [0/5] eta: 0:00:08 loss: 0.2377 (0.2377) time: 1.6907 data: 1.6586 max mem: 9672 +[13:58:26.209661] val: [4/5] eta: 0:00:00 loss: 0.2829 (0.4451) time: 0.3625 data: 0.3318 max mem: 9672 +[13:58:26.337691] val: Total time: 0:00:01 (0.3883 s / it) +[13:58:26.349886] val loss: 0.4450608491897583 +[13:58:26.350113] Accuracy: 0.8345, F1 Score: 0.7763, ROC AUC: 0.9763, Hamming Loss: 0.0662, + Jaccard Score: 0.6498, Precision: 0.8086, Recall: 0.7595, + Average Precision: 0.8608, Kappa: 0.7660, Score: 0.8395 +[13:58:26.387123] Best epoch = 28, Best score = 0.8596 +[13:58:26.660443] log_dir: ./output_logs/retfound +[13:58:28.460381] Epoch: [40] [ 0/30] eta: 0:00:53 lr: 0.000092 loss: 0.6664 (0.6664) time: 1.7988 data: 1.6512 max mem: 9672 +[13:58:31.317435] Epoch: [40] [20/30] eta: 0:00:02 lr: 0.000081 loss: 0.5129 (0.5372) time: 0.1428 data: 0.0002 max mem: 9672 +[13:58:32.614893] Epoch: [40] [29/30] eta: 0:00:00 lr: 0.000076 loss: 0.5182 (0.5311) time: 0.1434 data: 0.0001 max mem: 9672 +[13:58:32.768755] Epoch: [40] Total time: 0:00:06 (0.2036 s / it) +[13:58:32.778515] Averaged stats: lr: 0.000076 loss: 0.5182 (0.5311) +[13:58:34.411931] val: [0/5] eta: 0:00:08 loss: 0.2454 (0.2454) time: 1.6224 data: 1.5865 max mem: 9672 +[13:58:34.535551] val: [4/5] eta: 0:00:00 loss: 0.2735 (0.4650) time: 0.3491 data: 0.3174 max mem: 9672 +[13:58:34.658553] val: Total time: 0:00:01 (0.3740 s / it) +[13:58:34.670910] val loss: 0.46497651040554044 +[13:58:34.671114] Accuracy: 0.8489, F1 Score: 0.7978, ROC AUC: 0.9762, Hamming Loss: 0.0604, + Jaccard Score: 0.6739, Precision: 0.8483, Recall: 0.7649, + Average Precision: 0.8466, Kappa: 0.7841, Score: 0.8527 +[13:58:34.710015] Best epoch = 28, Best score = 0.8596 +[13:58:34.952957] log_dir: ./output_logs/retfound +[13:58:36.692808] Epoch: [41] [ 0/30] eta: 0:00:52 lr: 0.000076 loss: 0.7429 (0.7429) time: 1.7388 data: 1.5918 max mem: 9672 +[13:58:39.560735] Epoch: [41] [20/30] eta: 0:00:02 lr: 0.000065 loss: 0.4338 (0.5210) time: 0.1434 data: 0.0002 max mem: 9672 +[13:58:40.850980] Epoch: [41] [29/30] eta: 0:00:00 lr: 0.000061 loss: 0.5795 (0.5278) time: 0.1435 data: 0.0001 max mem: 9672 +[13:58:41.011927] Epoch: [41] Total time: 0:00:06 (0.2020 s / it) +[13:58:41.020152] Averaged stats: lr: 0.000061 loss: 0.5795 (0.5278) +[13:58:42.647825] val: [0/5] eta: 0:00:08 loss: 0.2996 (0.2996) time: 1.6143 data: 1.5793 max mem: 9672 +[13:58:42.769229] val: [4/5] eta: 0:00:00 loss: 0.3080 (0.4951) time: 0.3470 data: 0.3160 max mem: 9672 +[13:58:42.900246] val: Total time: 0:00:01 (0.3735 s / it) +[13:58:42.912531] val loss: 0.49512178897857667 +[13:58:42.912718] Accuracy: 0.8561, F1 Score: 0.8038, ROC AUC: 0.9752, Hamming Loss: 0.0576, + Jaccard Score: 0.6826, Precision: 0.8566, Recall: 0.7716, + Average Precision: 0.8409, Kappa: 0.7946, Score: 0.8579 +[13:58:42.960060] Best epoch = 28, Best score = 0.8596 +[13:58:43.182035] log_dir: ./output_logs/retfound +[13:58:45.012261] Epoch: [42] [ 0/30] eta: 0:00:54 lr: 0.000061 loss: 0.4524 (0.4524) time: 1.8291 data: 1.6827 max mem: 9672 +[13:58:47.883665] Epoch: [42] [20/30] eta: 0:00:02 lr: 0.000051 loss: 0.4534 (0.4788) time: 0.1435 data: 0.0002 max mem: 9672 +[13:58:49.182289] Epoch: [42] [29/30] eta: 0:00:00 lr: 0.000047 loss: 0.5025 (0.5098) time: 0.1435 data: 0.0001 max mem: 9672 +[13:58:49.332329] Epoch: [42] Total time: 0:00:06 (0.2050 s / it) +[13:58:49.342311] Averaged stats: lr: 0.000047 loss: 0.5025 (0.5098) +[13:58:51.006986] val: [0/5] eta: 0:00:08 loss: 0.1963 (0.1963) time: 1.6487 data: 1.6121 max mem: 9672 +[13:58:51.144813] val: [4/5] eta: 0:00:00 loss: 0.3283 (0.4631) time: 0.3572 data: 0.3247 max mem: 9672 +[13:58:51.275832] val: Total time: 0:00:01 (0.3836 s / it) +[13:58:51.287976] val loss: 0.4630673497915268 +[13:58:51.288177] Accuracy: 0.8561, F1 Score: 0.7959, ROC AUC: 0.9756, Hamming Loss: 0.0576, + Jaccard Score: 0.6757, Precision: 0.8312, Recall: 0.7747, + Average Precision: 0.8395, Kappa: 0.7962, Score: 0.8559 +[13:58:51.332444] Best epoch = 28, Best score = 0.8596 +[13:58:51.565304] log_dir: ./output_logs/retfound +[13:58:53.275619] Epoch: [43] [ 0/30] eta: 0:00:51 lr: 0.000047 loss: 0.3933 (0.3933) time: 1.7093 data: 1.5620 max mem: 9672 +[13:58:56.142514] Epoch: [43] [20/30] eta: 0:00:02 lr: 0.000039 loss: 0.5281 (0.5271) time: 0.1433 data: 0.0001 max mem: 9672 +[13:58:57.428010] Epoch: [43] [29/30] eta: 0:00:00 lr: 0.000035 loss: 0.4666 (0.5050) time: 0.1432 data: 0.0001 max mem: 9672 +[13:58:57.575513] Epoch: [43] Total time: 0:00:06 (0.2003 s / it) +[13:58:57.583620] Averaged stats: lr: 0.000035 loss: 0.4666 (0.5050) +[13:58:59.294451] val: [0/5] eta: 0:00:08 loss: 0.2372 (0.2372) time: 1.6951 data: 1.6585 max mem: 9672 +[13:58:59.481067] val: [4/5] eta: 0:00:00 loss: 0.2808 (0.4627) time: 0.3762 data: 0.3444 max mem: 9672 +[13:58:59.622840] val: Total time: 0:00:02 (0.4048 s / it) +[13:58:59.653865] val loss: 0.4626917392015457 +[13:58:59.654096] Accuracy: 0.8705, F1 Score: 0.8063, ROC AUC: 0.9759, Hamming Loss: 0.0518, + Jaccard Score: 0.6915, Precision: 0.8468, Recall: 0.7827, + Average Precision: 0.8405, Kappa: 0.8160, Score: 0.8661 +[13:59:01.513949] Best epoch = 43, Best score = 0.8661 +[13:59:01.581672] log_dir: ./output_logs/retfound +[13:59:03.370360] Epoch: [44] [ 0/30] eta: 0:00:53 lr: 0.000035 loss: 0.3384 (0.3384) time: 1.7876 data: 1.6370 max mem: 9672 +[13:59:06.231050] Epoch: [44] [20/30] eta: 0:00:02 lr: 0.000028 loss: 0.4521 (0.4865) time: 0.1430 data: 0.0002 max mem: 9672 +[13:59:07.526450] Epoch: [44] [29/30] eta: 0:00:00 lr: 0.000025 loss: 0.4839 (0.5106) time: 0.1439 data: 0.0001 max mem: 9672 +[13:59:07.674891] Epoch: [44] Total time: 0:00:06 (0.2031 s / it) +[13:59:07.683995] Averaged stats: lr: 0.000025 loss: 0.4839 (0.5106) +[13:59:09.313426] val: [0/5] eta: 0:00:08 loss: 0.2472 (0.2472) time: 1.6188 data: 1.5799 max mem: 9672 +[13:59:09.436225] val: [4/5] eta: 0:00:00 loss: 0.2778 (0.4610) time: 0.3482 data: 0.3161 max mem: 9672 +[13:59:09.563222] val: Total time: 0:00:01 (0.3738 s / it) +[13:59:09.576160] val loss: 0.46102696359157563 +[13:59:09.576351] Accuracy: 0.8561, F1 Score: 0.7906, ROC AUC: 0.9766, Hamming Loss: 0.0576, + Jaccard Score: 0.6714, Precision: 0.8286, Recall: 0.7716, + Average Precision: 0.8614, Kappa: 0.7959, Score: 0.8543 +[13:59:09.615835] Best epoch = 43, Best score = 0.8661 +[13:59:09.861236] log_dir: ./output_logs/retfound +[13:59:11.708328] Epoch: [45] [ 0/30] eta: 0:00:55 lr: 0.000025 loss: 0.9127 (0.9127) time: 1.8460 data: 1.6982 max mem: 9672 +[13:59:14.591719] Epoch: [45] [20/30] eta: 0:00:02 lr: 0.000019 loss: 0.4937 (0.5218) time: 0.1441 data: 0.0002 max mem: 9672 +[13:59:15.881239] Epoch: [45] [29/30] eta: 0:00:00 lr: 0.000017 loss: 0.4402 (0.4954) time: 0.1441 data: 0.0001 max mem: 9672 +[13:59:16.019431] Epoch: [45] Total time: 0:00:06 (0.2053 s / it) +[13:59:16.028123] Averaged stats: lr: 0.000017 loss: 0.4402 (0.4954) +[13:59:17.725137] val: [0/5] eta: 0:00:08 loss: 0.2357 (0.2357) time: 1.6850 data: 1.6484 max mem: 9672 +[13:59:17.846972] val: [4/5] eta: 0:00:00 loss: 0.2915 (0.4523) time: 0.3613 data: 0.3298 max mem: 9672 +[13:59:17.972168] val: Total time: 0:00:01 (0.3865 s / it) +[13:59:17.984386] val loss: 0.45225479304790495 +[13:59:17.984588] Accuracy: 0.8705, F1 Score: 0.8063, ROC AUC: 0.9767, Hamming Loss: 0.0518, + Jaccard Score: 0.6915, Precision: 0.8468, Recall: 0.7827, + Average Precision: 0.8618, Kappa: 0.8160, Score: 0.8663 +[13:59:19.631071] Best epoch = 45, Best score = 0.8663 +[13:59:19.689659] log_dir: ./output_logs/retfound +[13:59:21.413767] Epoch: [46] [ 0/30] eta: 0:00:51 lr: 0.000016 loss: 0.6036 (0.6036) time: 1.7231 data: 1.5761 max mem: 9672 +[13:59:24.287466] Epoch: [46] [20/30] eta: 0:00:02 lr: 0.000012 loss: 0.4638 (0.4814) time: 0.1436 data: 0.0001 max mem: 9672 +[13:59:25.573930] Epoch: [46] [29/30] eta: 0:00:00 lr: 0.000010 loss: 0.4969 (0.4830) time: 0.1433 data: 0.0001 max mem: 9672 +[13:59:25.734908] Epoch: [46] Total time: 0:00:06 (0.2015 s / it) +[13:59:25.744227] Averaged stats: lr: 0.000010 loss: 0.4969 (0.4830) +[13:59:27.477688] val: [0/5] eta: 0:00:08 loss: 0.2483 (0.2483) time: 1.7214 data: 1.6860 max mem: 9672 +[13:59:27.599676] val: [4/5] eta: 0:00:00 loss: 0.2824 (0.4585) time: 0.3686 data: 0.3373 max mem: 9672 +[13:59:27.725987] val: Total time: 0:00:01 (0.3941 s / it) +[13:59:27.740230] val loss: 0.4584751546382904 +[13:59:27.740421] Accuracy: 0.8633, F1 Score: 0.8022, ROC AUC: 0.9765, Hamming Loss: 0.0547, + Jaccard Score: 0.6848, Precision: 0.8439, Recall: 0.7785, + Average Precision: 0.8614, Kappa: 0.8058, Score: 0.8615 +[13:59:27.789710] Best epoch = 45, Best score = 0.8663 +[13:59:28.000983] log_dir: ./output_logs/retfound +[13:59:29.803312] Epoch: [47] [ 0/30] eta: 0:00:54 lr: 0.000010 loss: 0.6809 (0.6809) time: 1.8008 data: 1.6555 max mem: 9672 +[13:59:32.674737] Epoch: [47] [20/30] eta: 0:00:02 lr: 0.000006 loss: 0.4881 (0.5189) time: 0.1435 data: 0.0002 max mem: 9672 +[13:59:33.970416] Epoch: [47] [29/30] eta: 0:00:00 lr: 0.000005 loss: 0.4469 (0.4906) time: 0.1437 data: 0.0002 max mem: 9672 +[13:59:34.123010] Epoch: [47] Total time: 0:00:06 (0.2041 s / it) +[13:59:34.132463] Averaged stats: lr: 0.000005 loss: 0.4469 (0.4906) +[13:59:35.870020] val: [0/5] eta: 0:00:08 loss: 0.2402 (0.2402) time: 1.7204 data: 1.6839 max mem: 9672 +[13:59:35.991948] val: [4/5] eta: 0:00:00 loss: 0.2931 (0.4598) time: 0.3684 data: 0.3369 max mem: 9672 +[13:59:36.118917] val: Total time: 0:00:01 (0.3940 s / it) +[13:59:36.131251] val loss: 0.4597661018371582 +[13:59:36.131439] Accuracy: 0.8705, F1 Score: 0.8063, ROC AUC: 0.9763, Hamming Loss: 0.0518, + Jaccard Score: 0.6915, Precision: 0.8468, Recall: 0.7827, + Average Precision: 0.8601, Kappa: 0.8160, Score: 0.8662 +[13:59:36.170955] Best epoch = 45, Best score = 0.8663 +[13:59:36.407125] log_dir: ./output_logs/retfound +[13:59:38.272373] Epoch: [48] [ 0/30] eta: 0:00:55 lr: 0.000005 loss: 0.4629 (0.4629) time: 1.8642 data: 1.7171 max mem: 9672 +[13:59:41.140203] Epoch: [48] [20/30] eta: 0:00:02 lr: 0.000003 loss: 0.5332 (0.5433) time: 0.1433 data: 0.0001 max mem: 9672 +[13:59:42.427806] Epoch: [48] [29/30] eta: 0:00:00 lr: 0.000002 loss: 0.5256 (0.5370) time: 0.1431 data: 0.0002 max mem: 9672 +[13:59:42.554747] Epoch: [48] Total time: 0:00:06 (0.2049 s / it) +[13:59:42.564486] Averaged stats: lr: 0.000002 loss: 0.5256 (0.5370) +[13:59:44.281606] val: [0/5] eta: 0:00:08 loss: 0.2386 (0.2386) time: 1.7018 data: 1.6661 max mem: 9672 +[13:59:44.403826] val: [4/5] eta: 0:00:00 loss: 0.3007 (0.4612) time: 0.3647 data: 0.3333 max mem: 9672 +[13:59:44.534838] val: Total time: 0:00:01 (0.3911 s / it) +[13:59:44.547584] val loss: 0.46119038164615633 +[13:59:44.547766] Accuracy: 0.8633, F1 Score: 0.8012, ROC AUC: 0.9761, Hamming Loss: 0.0547, + Jaccard Score: 0.6837, Precision: 0.8405, Recall: 0.7786, + Average Precision: 0.8589, Kappa: 0.8061, Score: 0.8611 +[13:59:44.585350] Best epoch = 45, Best score = 0.8663 +[13:59:44.842751] log_dir: ./output_logs/retfound +[13:59:46.560249] Epoch: [49] [ 0/30] eta: 0:00:51 lr: 0.000002 loss: 0.4151 (0.4151) time: 1.7164 data: 1.5690 max mem: 9672 +[13:59:49.433464] Epoch: [49] [20/30] eta: 0:00:02 lr: 0.000001 loss: 0.4393 (0.4540) time: 0.1436 data: 0.0002 max mem: 9672 +[13:59:50.726523] Epoch: [49] [29/30] eta: 0:00:00 lr: 0.000001 loss: 0.4320 (0.4606) time: 0.1440 data: 0.0001 max mem: 9672 +[13:59:50.863986] Epoch: [49] Total time: 0:00:06 (0.2007 s / it) +[13:59:50.872130] Averaged stats: lr: 0.000001 loss: 0.4320 (0.4606) +[13:59:52.490465] val: [0/5] eta: 0:00:08 loss: 0.2375 (0.2375) time: 1.6072 data: 1.5705 max mem: 9672 +[13:59:52.656931] val: [4/5] eta: 0:00:00 loss: 0.3007 (0.4611) time: 0.3546 data: 0.3231 max mem: 9672 +[13:59:52.782558] val: Total time: 0:00:01 (0.3800 s / it) +[13:59:52.794804] val loss: 0.461053204536438 +[13:59:52.794994] Accuracy: 0.8633, F1 Score: 0.8012, ROC AUC: 0.9761, Hamming Loss: 0.0547, + Jaccard Score: 0.6837, Precision: 0.8405, Recall: 0.7786, + Average Precision: 0.8589, Kappa: 0.8061, Score: 0.8611 +[13:59:52.844134] Best epoch = 45, Best score = 0.8663 +[13:59:56.347721] Test with the best model, epoch = 45: +[13:59:57.879516] test: [0/9] eta: 0:00:13 loss: 0.1087 (0.1087) time: 1.5216 data: 1.4859 max mem: 9672 +[13:59:58.251311] test: [8/9] eta: 0:00:00 loss: 0.3671 (0.4798) time: 0.2103 data: 0.1664 max mem: 9672 +[13:59:58.342502] test: Total time: 0:00:01 (0.2206 s / it) +[13:59:58.357167] val loss: 0.47984142270353103 +[13:59:58.357287] Accuracy: 0.8566, F1 Score: 0.7602, ROC AUC: 0.9673, Hamming Loss: 0.0573, + Jaccard Score: 0.6355, Precision: 0.8011, Recall: 0.7423, + Average Precision: 0.8271, Kappa: 0.7967, Score: 0.8414 +[13:59:59.420735] Training time 0:07:20 +[rank0]:[W615 13:59:59.921969549 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/mmac/retfound acc=0.8566 auroc_macro_ovr=0.9672777830979756 f1_macro=0.7602 qwk=0.9263793516080929 diff --git a/results/mmac/vit/confusion_matrix.png b/results/mmac/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..b699619df69ae8367749388103d58f6db453b89c --- /dev/null +++ b/results/mmac/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e05bfd2c1f133e4895d20f90ea7ae878fde2398f1be0a2dddc6bb64795d9225 +size 99207 diff --git a/results/mmac/vit/log.csv b/results/mmac/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..b3567533dde2abf71cd302ea6266dd2927e3010f --- /dev/null +++ b/results/mmac/vit/log.csv @@ -0,0 +1,43 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,1.9697898944218954,0.6402877697841727,0.8399916108867366,0.5997217167651084,0.00015555555555555556 +1,1.459381878376007,0.5971223021582733,0.911755439055802,0.5931713226492026,0.0003222222222222222 +2,1.2024412035942078,0.5971223021582733,0.8911204002116628,0.5901292300926588,0.000488888888888889 +3,1.1377753337224326,0.6690647482014388,0.9384907262817557,0.6614924291483542,0.0004995136524488283 +4,0.9252972324689229,0.6762589928057554,0.9261378833766699,0.6714212349961506,0.0004979153985213637 +5,0.7758009274800618,0.6330935251798561,0.9398263272751215,0.6216816873956308,0.000495209894241724 +6,0.8057701071103414,0.7050359712230215,0.9555209023396735,0.7194544842485809,0.0004914092230487934 +7,0.713987159729004,0.7122302158273381,0.9513451744291777,0.7048699224358509,0.0004865303596627461 +8,0.6414933562278747,0.7338129496402878,0.961404189823946,0.732938189575691,0.00048059509427182646 +9,0.69733025431633,0.7266187050359713,0.9488629636574931,0.7291571902190285,0.00047362993521207807 +10,0.7299656987190246,0.4316546762589928,0.8067283599355157,0.48298979320757446,0.0004656659905746797 +11,0.7574182828267415,0.7410071942446043,0.939746097030904,0.7523992189884074,0.00045673882926965906 +12,0.7559382716814677,0.6187050359712231,0.9243902071339198,0.632648201467556,0.00044688832216650565 +13,0.5406759818394978,0.7194244604316546,0.9429466065147942,0.7086668537271285,0.00043615846402118917 +14,0.5643691162268321,0.7553956834532374,0.9599252113369303,0.7467786577672975,0.0004245971769848993 +15,0.543748656908671,0.762589928057554,0.9444677115872345,0.7362872803189641,0.0004122560965720853 +16,0.5359903256098429,0.762589928057554,0.9483555513423287,0.758362491117753,0.00039919034104371357 +17,0.5025801579157512,0.7985611510791367,0.9537687214119674,0.7646682543030133,0.00038545826523573424 +18,0.4450039515892665,0.7769784172661871,0.9587610210490688,0.7729688443651077,0.00037112119993221794 +19,0.414046315352122,0.7697841726618705,0.962154152562501,0.7494594940723914,0.00035624317794718564 +20,0.5068672060966491,0.8273381294964028,0.9707468226038358,0.8305866770418815,0.0003408906481385174 +21,0.4274222403764725,0.7913669064748201,0.9600082567304501,0.7844930396750424,0.0003251321786312219 +22,0.3088470141092936,0.841726618705036,0.9439402673062253,0.8336517946489845,0.00030903815057554555 +23,0.5666604439417521,0.7553956834532374,0.9605132271582881,0.7575487780680011,0.0002926804438076697 +24,0.3989462087551753,0.8345323741007195,0.968342581495939,0.8302377440162338,0.0002761321158169134 +25,0.3718212107817332,0.7985611510791367,0.9580084197195943,0.8055498210014272,0.00025946707545325117 +26,0.29139841298262276,0.7985611510791367,0.9611947485579841,0.7864221317613387,0.0002427597528324471 +27,0.21604471951723098,0.7410071942446043,0.9637663770850186,0.7589768049065141,0.0002260847669130901 +28,0.3191804145773252,0.7985611510791367,0.9660175566086895,0.7997780559289073,0.00020951659223021038 +29,0.22799710681041083,0.8201438848920863,0.9673910467863021,0.8089899847840069,0.00019312922627392725 +30,0.19453658163547516,0.8273381294964028,0.9694249259656311,0.8247059732453345,0.00017699585899869536 +31,0.16031322826941807,0.841726618705036,0.9682369138941604,0.8383688710666325,0.00016118854593919998 +32,0.1253216301401456,0.7985611510791367,0.9664279301633462,0.7772398582875929,0.00014577788639284883 +33,0.11546132092674573,0.7985611510791367,0.9689549718710984,0.7778584956190278,0.00013083270810617013 +34,0.10676546841859817,0.8129496402877698,0.9690874941179581,0.7881281773566468,0.00011641975987338946 +35,0.09784428800145785,0.8201438848920863,0.9694266201059044,0.8073749923300081,0.00010260341342011111 +36,0.10016541654864947,0.8129496402877698,0.9663051634506132,0.810768137363249,8.944537590356428e-05 +37,0.12820614899198216,0.7985611510791367,0.9627217663217145,0.7798175658478547,7.700441431346016e-05 +38,0.11299030097822348,0.8201438848920863,0.9671248101277918,0.8088530852234607,6.533609300434956e-05 +39,0.09062570532162985,0.7985611510791367,0.970129018778876,0.7873554598943472,5.449252553172707e-05 +40,0.07882252857089042,0.8345323741007195,0.9713202184702056,0.8093246821295814,4.4522141900241427e-05 +41,0.07385999610026678,0.8201438848920863,0.9701326170012974,0.8094289051073953,3.546947226353833e-05 diff --git a/results/mmac/vit/metrics.json b/results/mmac/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..34a1f3ec2aa34237c0ccacfa768ddc95ff6948b5 --- /dev/null +++ b/results/mmac/vit/metrics.json @@ -0,0 +1,71 @@ +{ + "n_test": 279, + "n_classes": 5, + "task": "multiclass", + "accuracy": 0.8243727598566308, + "balanced_accuracy": 0.7345030988512782, + "precision_macro": 0.7522028862478777, + "recall_macro": 0.7345030988512782, + "f1_macro": 0.7389990992554395, + "precision_weighted": 0.8287257881957757, + "recall_weighted": 0.8243727598566308, + "f1_weighted": 0.8254127375902228, + "cohen_kappa": 0.7526282457251425, + "quadratic_weighted_kappa": 0.9051554367702912, + "mcc": 0.7529299002581141, + "auroc_macro_ovr": 0.9511929367879477, + "auroc_weighted_ovr": 0.9564827764454924, + "auprc_macro": 0.7602940855355467, + "auroc_per_class": { + "0": 0.9912767644726408, + "1": 0.9325177584846093, + "2": 0.9454673115930723, + "3": 0.9318181818181819, + "4": 0.9548846675712348 + }, + "per_class": { + "0": { + "precision": 0.9375, + "recall": 0.9278350515463918, + "f1-score": 0.9326424870466321, + "support": 97.0 + }, + "1": { + "precision": 0.8315789473684211, + "recall": 0.8061224489795918, + "f1-score": 0.8186528497409327, + "support": 98.0 + }, + "2": { + "precision": 0.7419354838709677, + "recall": 0.7931034482758621, + "f1-score": 0.7666666666666667, + "support": 58.0 + }, + "3": { + "precision": 0.5, + "recall": 0.6, + "f1-score": 0.5454545454545454, + "support": 15.0 + }, + "4": { + "precision": 0.75, + "recall": 0.5454545454545454, + "f1-score": 0.631578947368421, + "support": 11.0 + }, + "accuracy": 0.8243727598566308, + "macro avg": { + "precision": 0.7522028862478777, + "recall": 0.7345030988512782, + "f1-score": 0.7389990992554395, + "support": 279.0 + }, + "weighted avg": { + "precision": 0.8287257881957757, + "recall": 0.8243727598566308, + "f1-score": 0.8254127375902228, + "support": 279.0 + } + } +} \ No newline at end of file diff --git a/results/mmac/vit/pr.png b/results/mmac/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a4615bb6825da155e45a919fa606db28ed62c34a --- /dev/null +++ b/results/mmac/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b9c25a87e7a32125545ec8320a4ca5bdbb99812289e4f13f08d29b368af331db +size 92483 diff --git a/results/mmac/vit/roc.png b/results/mmac/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..cf897c743a6485e49907e9a232e4a17276123ab0 --- /dev/null +++ b/results/mmac/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:388ff38b9eb63a5d0b51fffd999d750bde51b1a50ed03ba07163fcd9f9cff59b +size 82624 diff --git a/results/mmac/vit/test_pred.npz b/results/mmac/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..cd4b70f135e3e1111f64e2bc85639de5d62ebbaa --- /dev/null +++ b/results/mmac/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7cd9b551025173102277d59de30718e52b73415ab55f51e28398d5c4d8c1c93c +size 8322 diff --git a/results/mmac/vit/train.log b/results/mmac/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0f9731c2fab4b9b9ce84f3ff4f5ae6cd323e0241 --- /dev/null +++ b/results/mmac/vit/train.log @@ -0,0 +1,184 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=973 val=139 test=279 classes=['0', '1', '2', '3', '4'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=2.2916 val_acc=0.1942 val_auc=0.6648 score=0.2969 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=1.7981 val_acc=0.4676 val_auc=0.8465 score=0.4936 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=1.4788 val_acc=0.5683 val_auc=0.9057 score=0.5961 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=1.3819 val_acc=0.6691 val_auc=0.9357 score=0.6803 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=1.2328 val_acc=0.7338 val_auc=0.9513 score=0.7367 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=1.1908 val_acc=0.7698 val_auc=0.9320 score=0.7630 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=1.1252 val_acc=0.7698 val_auc=0.9438 score=0.7787 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=1.0735 val_acc=0.8058 val_auc=0.9208 score=0.7871 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=1.0729 val_acc=0.7266 val_auc=0.9230 score=0.7242 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=1.0058 val_acc=0.7770 val_auc=0.9418 score=0.7637 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.9737 val_acc=0.8201 val_auc=0.8782 score=0.7930 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.9828 val_acc=0.7986 val_auc=0.8924 score=0.7687 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.9450 val_acc=0.8129 val_auc=0.9133 score=0.7976 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.9361 val_acc=0.8129 val_auc=0.8777 score=0.7748 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.9053 val_acc=0.8129 val_auc=0.8802 score=0.7668 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.9047 val_acc=0.8273 val_auc=0.9036 score=0.8014 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.8933 val_acc=0.8345 val_auc=0.8931 score=0.8035 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.8871 val_acc=0.8273 val_auc=0.9034 score=0.8013 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.8732 val_acc=0.8058 val_auc=0.9062 score=0.7755 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.9069 val_acc=0.8058 val_auc=0.9316 score=0.7966 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.8762 val_acc=0.8129 val_auc=0.9272 score=0.7984 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.8494 val_acc=0.8273 val_auc=0.9088 score=0.7976 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.8448 val_acc=0.8201 val_auc=0.9397 score=0.8049 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.8436 val_acc=0.8201 val_auc=0.9246 score=0.7932 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.8428 val_acc=0.8345 val_auc=0.9409 score=0.8338 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.8320 val_acc=0.8273 val_auc=0.9381 score=0.8252 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.8551 val_acc=0.8345 val_auc=0.9395 score=0.8337 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.8135 val_acc=0.8201 val_auc=0.9244 score=0.8155 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.8054 val_acc=0.8129 val_auc=0.9169 score=0.7890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep29 loss=0.8224 val_acc=0.8058 val_auc=0.9396 score=0.8009 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep30 loss=0.8167 val_acc=0.8129 val_auc=0.9254 score=0.7993 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep31 loss=0.8145 val_acc=0.8273 val_auc=0.9173 score=0.7990 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep32 loss=0.8020 val_acc=0.8201 val_auc=0.9260 score=0.8161 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep33 loss=0.8091 val_acc=0.8201 val_auc=0.9284 score=0.8166 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep34 loss=0.7974 val_acc=0.8345 val_auc=0.9258 score=0.8287 +[vit] early stop at ep34 (best ep24 score=0.8338) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=24 best_val_score=0.8338 -> saved test_pred.npz (279 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/mmac/vit acc=0.8172 auroc_macro_ovr=0.8950096039884677 f1_macro=0.7093 qwk=0.8420202835662748 diff --git a/results/papila/resnet/confusion_matrix.png b/results/papila/resnet/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..315f6d5b49bbef41ad7f01c4fe8d475867ac0006 --- /dev/null +++ b/results/papila/resnet/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5d540f973ca1fe879d63a1019f85ee2107bc683576651d53411181cf83ec3c36 +size 72526 diff --git a/results/papila/resnet/log.csv b/results/papila/resnet/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..b47c5f19750e379a60e535ef4cec5fd501a26e86 --- /dev/null +++ b/results/papila/resnet/log.csv @@ -0,0 +1,32 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.6922180950641632,0.3333333333333333,0.425,0.2499896222498962,0.000125 +1,0.689469650387764,0.5952380952380952,0.49843750000000003,0.25498668078429826,0.0002916666666666667 +2,0.6794771105051041,0.7380952380952381,0.625,0.43564169173274997,0.0004583333333333333 +3,0.6663116663694382,0.8095238095238095,0.6124999999999999,0.49982439335887613,0.0004996859161456965 +4,0.6346924901008606,0.7380952380952381,0.6125,0.4969460385467029,0.0004982915790812436 +5,0.6034166216850281,0.5952380952380952,0.65,0.37658635851813566,0.0004957883115509159 +6,0.5465360134840012,0.6190476190476191,0.6609375,0.32258805193998147,0.0004921872937551814 +7,0.4897657036781311,0.7380952380952381,0.7312500000000001,0.5069176624053334,0.000487504608713676 +8,0.4183604121208191,0.5714285714285714,0.55,0.3804621848739496,0.00048176117043453436 +9,0.3654043823480606,0.42857142857142855,0.5062500000000001,0.2833392280963018,0.000474982630507352 +10,0.31567204371094704,0.5714285714285714,0.59375,0.3950455182072829,0.00046719926353695914 +11,0.26756494492292404,0.47619047619047616,0.678125,0.3866373479771656,0.00045844583192968674 +12,0.20010125637054443,0.35714285714285715,0.790625,0.4065211232431438,0.00044876143063602076 +13,0.20714566484093666,0.5238095238095238,0.7875,0.4987133300348403,0.00043818931254306284 +14,0.13108717650175095,0.6190476190476191,0.7875000000000001,0.5304299454493253,0.00042677669529663686 +15,0.11574918404221535,0.6666666666666666,0.746875,0.5317655148362447,0.0004145745504158204 +16,0.10436330921947956,0.7380952380952381,0.740625,0.5856465653156565,0.0004016373756417668 +17,0.10611409042030573,0.7380952380952381,0.740625,0.5856465653156565,0.00038802295153756415 +18,0.07898017158731818,0.8095238095238095,0.7000000000000001,0.594160272804774,0.0003737920834262134 +19,0.05685942014679313,0.8095238095238095,0.684375,0.5889519394714408,0.0003590083298192957 +20,0.09716262901201844,0.8095238095238095,0.78125,0.6812292063247072,0.0003437377185492303 +21,0.04607221717014909,0.7857142857142857,0.753125,0.6321379118112501,0.0003280484518729466 +22,0.041541534941643476,0.8095238095238095,0.7,0.6178044426366573,0.00031201060186404833 +23,0.024707847274839878,0.7857142857142857,0.7,0.5405250205086136,0.0002956957974539226 +24,0.03069802187383175,0.7857142857142857,0.684375,0.49902103145941173,0.0002791769045195441 +25,0.017025835812091827,0.8333333333333334,0.6968749999999999,0.6193666826178106,0.0002625277004467798 +26,0.019359020981937647,0.8571428571428571,0.725,0.6789535379369139,0.00024582254462267474 +27,0.008794337161816657,0.8333333333333334,0.70625,0.6453112397124849,0.00022913604632837759 +28,0.014210807625204325,0.8571428571428571,0.709375,0.6737452046035806,0.00021254273151597963 +29,0.010004262439906597,0.8571428571428571,0.703125,0.6716618712702472,0.00019611670995752162 +30,0.010085121961310506,0.8571428571428571,0.6906249999999999,0.6674952046035806,0.00017993134425276094 diff --git a/results/papila/resnet/metrics.json b/results/papila/resnet/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..6a0016a59ff5338805160174a90547ec6a47bced --- /dev/null +++ b/results/papila/resnet/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8809523809523809, + "balanced_accuracy": 0.7352941176470589, + "precision_macro": 0.845945945945946, + "recall_macro": 0.7352941176470589, + "f1_macro": 0.772481040086674, + "precision_weighted": 0.8743886743886743, + "recall_weighted": 0.8809523809523809, + "f1_weighted": 0.8697312077593767, + "cohen_kappa": 0.5493562231759657, + "quadratic_weighted_kappa": 0.5493562231759657, + "mcc": 0.5706103612971936, + "auroc": 0.7941176470588235, + "auprc": 0.7048294789225769, + "sensitivity": 0.5, + "specificity": 0.9705882352941176, + "precision_pos": 0.8, + "f1_pos": 0.6153846153846154, + "per_class": { + "0": { + "precision": 0.8918918918918919, + "recall": 0.9705882352941176, + "f1-score": 0.9295774647887324, + "support": 68.0 + }, + "1": { + "precision": 0.8, + "recall": 0.5, + "f1-score": 0.6153846153846154, + "support": 16.0 + }, + "accuracy": 0.8809523809523809, + "macro avg": { + "precision": 0.845945945945946, + "recall": 0.7352941176470589, + "f1-score": 0.772481040086674, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8743886743886743, + "recall": 0.8809523809523809, + "f1-score": 0.8697312077593767, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/papila/resnet/pr.png b/results/papila/resnet/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..d673787723b43b4ff47aa0afa676d05e904e3c24 --- /dev/null +++ b/results/papila/resnet/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ccb29914a1b88ad21c246ffc99528bcf7a57cb1271de4967fc8ae1960a211bd +size 44801 diff --git a/results/papila/resnet/roc.png b/results/papila/resnet/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..b0f7bcc95294ccf978ca6be1bea8604088fc0d9f --- /dev/null +++ b/results/papila/resnet/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:522fe4c7ecf695db453cf93bb58dd5a780eb94b9d3fa31c87255e2d77bddf275 +size 57086 diff --git a/results/papila/resnet/test_pred.npz b/results/papila/resnet/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..6a505333f1b93fdf07a1f30e485919904d0da570 --- /dev/null +++ b/results/papila/resnet/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c06dc28f0648018063be414c8348f344c833e66812bf77a23f4b61b640c7e23 +size 1854 diff --git a/results/papila/resnet/train.log b/results/papila/resnet/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f34f303316e4c35af31506cd43d29b91b1eaec9f --- /dev/null +++ b/results/papila/resnet/train.log @@ -0,0 +1,163 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:114: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[resnet] train=294 val=42 test=84 classes=['0', '1'] +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep0 loss=0.6922 val_acc=0.3333 val_auc=0.4250 score=0.2500 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep1 loss=0.6895 val_acc=0.5952 val_auc=0.4984 score=0.2550 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep2 loss=0.6795 val_acc=0.7381 val_auc=0.6250 score=0.4356 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep3 loss=0.6663 val_acc=0.8095 val_auc=0.6125 score=0.4998 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep4 loss=0.6347 val_acc=0.7381 val_auc=0.6125 score=0.4969 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep5 loss=0.6034 val_acc=0.5952 val_auc=0.6500 score=0.3766 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep6 loss=0.5465 val_acc=0.6190 val_auc=0.6609 score=0.3226 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep7 loss=0.4898 val_acc=0.7381 val_auc=0.7313 score=0.5069 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep8 loss=0.4184 val_acc=0.5714 val_auc=0.5500 score=0.3805 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep9 loss=0.3654 val_acc=0.4286 val_auc=0.5063 score=0.2833 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep10 loss=0.3157 val_acc=0.5714 val_auc=0.5938 score=0.3950 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep11 loss=0.2676 val_acc=0.4762 val_auc=0.6781 score=0.3866 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep12 loss=0.2001 val_acc=0.3571 val_auc=0.7906 score=0.4065 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep13 loss=0.2071 val_acc=0.5238 val_auc=0.7875 score=0.4987 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep14 loss=0.1311 val_acc=0.6190 val_auc=0.7875 score=0.5304 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep15 loss=0.1157 val_acc=0.6667 val_auc=0.7469 score=0.5318 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep16 loss=0.1044 val_acc=0.7381 val_auc=0.7406 score=0.5856 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep17 loss=0.1061 val_acc=0.7381 val_auc=0.7406 score=0.5856 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep18 loss=0.0790 val_acc=0.8095 val_auc=0.7000 score=0.5942 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep19 loss=0.0569 val_acc=0.8095 val_auc=0.6844 score=0.5890 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep20 loss=0.0972 val_acc=0.8095 val_auc=0.7812 score=0.6812 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep21 loss=0.0461 val_acc=0.7857 val_auc=0.7531 score=0.6321 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep22 loss=0.0415 val_acc=0.8095 val_auc=0.7000 score=0.6178 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep23 loss=0.0247 val_acc=0.7857 val_auc=0.7000 score=0.5405 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep24 loss=0.0307 val_acc=0.7857 val_auc=0.6844 score=0.4990 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep25 loss=0.0170 val_acc=0.8333 val_auc=0.6969 score=0.6194 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep26 loss=0.0194 val_acc=0.8571 val_auc=0.7250 score=0.6790 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep27 loss=0.0088 val_acc=0.8333 val_auc=0.7063 score=0.6453 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep28 loss=0.0142 val_acc=0.8571 val_auc=0.7094 score=0.6737 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep29 loss=0.0100 val_acc=0.8571 val_auc=0.7031 score=0.6717 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:136: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] ep30 loss=0.0101 val_acc=0.8571 val_auc=0.6906 score=0.6675 +[resnet] early stop at ep30 (best ep20 score=0.6812) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[resnet] DONE best_ep=20 best_val_score=0.6812 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/papila/resnet acc=0.8810 auroc=0.7941176470588235 f1_macro=0.7725 qwk=0.5493562231759657 diff --git a/results/papila/retfound/confusion_matrix.png b/results/papila/retfound/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..ac0fad06654ce85a88b195d14352067b13a9b547 --- /dev/null +++ b/results/papila/retfound/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfb2b0d9931914bfb7b6864539b77eb6062e8faacbf80f642185893ca80e37ac +size 74572 diff --git a/results/papila/retfound/confusion_matrix_test.jpg b/results/papila/retfound/confusion_matrix_test.jpg new file mode 100644 index 0000000000000000000000000000000000000000..bb78e1f3cd4bbca66196dee1305d5bd217471669 --- /dev/null +++ b/results/papila/retfound/confusion_matrix_test.jpg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f47a6c410ec7a3629543c9c4cb1093b2b9635932c2ab9c8ee089a837df3e300 +size 258603 diff --git a/results/papila/retfound/log.txt b/results/papila/retfound/log.txt new file mode 100644 index 0000000000000000000000000000000000000000..5177c358e09ebc9ed7d60591cc8f079a13f8fa59 --- /dev/null +++ b/results/papila/retfound/log.txt @@ -0,0 +1,50 @@ +{"train_lr": 2.777777777777778e-05, "train_loss": 0.6827528211805556, "epoch": 0, "n_parameters": 303303682} +{"train_lr": 9.027777777777779e-05, "train_loss": 0.5992067125108507, "epoch": 1, "n_parameters": 303303682} +{"train_lr": 0.00015277777777777777, "train_loss": 0.5172894795735677, "epoch": 2, "n_parameters": 303303682} +{"train_lr": 0.00021527777777777778, "train_loss": 0.5202977922227647, "epoch": 3, "n_parameters": 303303682} +{"train_lr": 0.0002777777777777778, "train_loss": 0.4997452629937066, "epoch": 4, "n_parameters": 303303682} +{"train_lr": 0.0003402777777777778, "train_loss": 0.5041298336452908, "epoch": 5, "n_parameters": 303303682} +{"train_lr": 0.0004027777777777778, "train_loss": 0.4900501039293077, "epoch": 6, "n_parameters": 303303682} +{"train_lr": 0.0004652777777777778, "train_loss": 0.5020377900865343, "epoch": 7, "n_parameters": 303303682} +{"train_lr": 0.0005277777777777777, "train_loss": 0.4713812934027778, "epoch": 8, "n_parameters": 303303682} +{"train_lr": 0.0005902777777777778, "train_loss": 0.44863936636183, "epoch": 9, "n_parameters": 303303682} +{"train_lr": 0.0006247307916747557, "train_loss": 0.45878227551778156, "epoch": 10, "n_parameters": 303303682} +{"train_lr": 0.0006229157334469918, "train_loss": 0.4269802040523953, "epoch": 11, "n_parameters": 303303682} +{"train_lr": 0.0006191899416650879, "train_loss": 0.4496183395385742, "epoch": 12, "n_parameters": 303303682} +{"train_lr": 0.0006135763870743313, "train_loss": 0.4470519489712185, "epoch": 13, "n_parameters": 303303682} +{"train_lr": 0.0006061096791054699, "train_loss": 0.45556622081332737, "epoch": 14, "n_parameters": 303303682} +{"train_lr": 0.0005968358524960642, "train_loss": 0.3853343062930637, "epoch": 15, "n_parameters": 303303682} +{"train_lr": 0.0005858120834710217, "train_loss": 0.4626990424262153, "epoch": 16, "n_parameters": 303303682} +{"train_lr": 0.0005731063372321567, "train_loss": 0.4492563141716851, "epoch": 17, "n_parameters": 303303682} +{"train_lr": 0.0005587969489301213, "train_loss": 0.42685553762647843, "epoch": 18, "n_parameters": 303303682} +{"train_lr": 0.0005429721407021519, "train_loss": 0.39957908789316815, "epoch": 19, "n_parameters": 303303682} +{"train_lr": 0.0005257294777532569, "train_loss": 0.38024022844102645, "epoch": 20, "n_parameters": 303303682} +{"train_lr": 0.0005071752668342834, "train_loss": 0.3989546100298564, "epoch": 21, "n_parameters": 303303682} +{"train_lr": 0.00048742390082544794, "train_loss": 0.3949376742045085, "epoch": 22, "n_parameters": 303303682} +{"train_lr": 0.0004665971534661909, "train_loss": 0.3674582905239529, "epoch": 23, "n_parameters": 303303682} +{"train_lr": 0.0004448234285795783, "train_loss": 0.3995681140157912, "epoch": 24, "n_parameters": 303303682} +{"train_lr": 0.0004222369684200342, "train_loss": 0.3284359508090549, "epoch": 25, "n_parameters": 303303682} +{"train_lr": 0.00039897702602520257, "train_loss": 0.34738753901587593, "epoch": 26, "n_parameters": 303303682} +{"train_lr": 0.0003751870066746631, "train_loss": 0.39103036456637913, "epoch": 27, "n_parameters": 303303682} +{"train_lr": 0.00035101358374869636, "train_loss": 0.41759975088967216, "epoch": 28, "n_parameters": 303303682} +{"train_lr": 0.0003266057944381195, "train_loss": 0.3715149561564128, "epoch": 29, "n_parameters": 303303682} +{"train_lr": 0.0003021141208804459, "train_loss": 0.3839879764450921, "epoch": 30, "n_parameters": 303303682} +{"train_lr": 0.0002776895623874664, "train_loss": 0.34202857149971855, "epoch": 31, "n_parameters": 303303682} +{"train_lr": 0.0002534827044842803, "train_loss": 0.3478359712494744, "epoch": 32, "n_parameters": 303303682} +{"train_lr": 0.00022964279049945762, "train_loss": 0.39175425635443795, "epoch": 33, "n_parameters": 303303682} +{"train_lr": 0.00020631680143029074, "train_loss": 0.34169335497750175, "epoch": 34, "n_parameters": 303303682} +{"train_lr": 0.00018364854975606965, "train_loss": 0.37303222550286186, "epoch": 35, "n_parameters": 303303682} +{"train_lr": 0.00016177779278632443, "train_loss": 0.3513450225194295, "epoch": 36, "n_parameters": 303303682} +{"train_lr": 0.00014083937101053823, "train_loss": 0.3313828508059184, "epoch": 37, "n_parameters": 303303682} +{"train_lr": 0.00012096237676169061, "train_loss": 0.3487516575389438, "epoch": 38, "n_parameters": 303303682} +{"train_lr": 0.00010226935831909702, "train_loss": 0.32215451531940037, "epoch": 39, "n_parameters": 303303682} +{"train_lr": 8.48755643575155e-05, "train_loss": 0.35140036212073433, "epoch": 40, "n_parameters": 303303682} +{"train_lr": 6.888823340074255e-05, "train_loss": 0.319819880856408, "epoch": 41, "n_parameters": 303303682} +{"train_lr": 5.4405932660453715e-05, "train_loss": 0.3057285149892171, "epoch": 42, "n_parameters": 303303682} +{"train_lr": 4.1517950336566605e-05, "train_loss": 0.3327277766333686, "epoch": 43, "n_parameters": 303303682} +{"train_lr": 3.0303745125797026e-05, "train_loss": 0.34663278195593095, "epoch": 44, "n_parameters": 303303682} +{"train_lr": 2.0832456332370918e-05, "train_loss": 0.37178613079918754, "epoch": 45, "n_parameters": 303303682} +{"train_lr": 1.3162477601222926e-05, "train_loss": 0.3343882891866896, "epoch": 46, "n_parameters": 303303682} +{"train_lr": 7.341096901757339e-06, "train_loss": 0.36498680379655624, "epoch": 47, "n_parameters": 303303682} +{"train_lr": 3.404204981791873e-06, "train_loss": 0.3455526265833113, "epoch": 48, "n_parameters": 303303682} +{"train_lr": 1.3760740891625011e-06, "train_loss": 0.32421498828464085, "epoch": 49, "n_parameters": 303303682} diff --git a/results/papila/retfound/metrics.json b/results/papila/retfound/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..381827cdf3d3dc0b25d94392a9ba5ef26d6496e5 --- /dev/null +++ b/results/papila/retfound/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.8333333333333334, + "balanced_accuracy": 0.7536764705882353, + "precision_macro": 0.7323232323232323, + "recall_macro": 0.7536764705882353, + "f1_macro": 0.7418788410886743, + "precision_weighted": 0.8417508417508417, + "recall_weighted": 0.8333333333333334, + "f1_weighted": 0.8369915130231197, + "cohen_kappa": 0.4842105263157894, + "quadratic_weighted_kappa": 0.4842105263157894, + "mcc": 0.48553038055886144, + "auroc": 0.8373161764705882, + "auprc": 0.6334668074576429, + "sensitivity": 0.625, + "specificity": 0.8823529411764706, + "precision_pos": 0.5555555555555556, + "f1_pos": 0.5882352941176471, + "per_class": { + "0": { + "precision": 0.9090909090909091, + "recall": 0.8823529411764706, + "f1-score": 0.8955223880597015, + "support": 68.0 + }, + "1": { + "precision": 0.5555555555555556, + "recall": 0.625, + "f1-score": 0.5882352941176471, + "support": 16.0 + }, + "accuracy": 0.8333333333333334, + "macro avg": { + "precision": 0.7323232323232323, + "recall": 0.7536764705882353, + "f1-score": 0.7418788410886743, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8417508417508417, + "recall": 0.8333333333333334, + "f1-score": 0.8369915130231197, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/papila/retfound/metrics_test.csv b/results/papila/retfound/metrics_test.csv new file mode 100644 index 0000000000000000000000000000000000000000..3b498a846f59c6dbc4ece55587c9a674e54ca0fb --- /dev/null +++ b/results/papila/retfound/metrics_test.csv @@ -0,0 +1,2 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.47513073682785034,0.8333333333333334,0.7418788410886743,0.8370863970588236,0.16666666666666666,0.6137387387387387,0.7323232323232323,0.7536764705882353,0.7821000271484148,0.4842105263157894 diff --git a/results/papila/retfound/metrics_val.csv b/results/papila/retfound/metrics_val.csv new file mode 100644 index 0000000000000000000000000000000000000000..def94f45820ffdc07fbd0514143047248b8ba3f9 --- /dev/null +++ b/results/papila/retfound/metrics_val.csv @@ -0,0 +1,51 @@ +val_loss,accuracy,f1,roc_auc,hamming,jaccard,precision,recall,average_precision,kappa +0.6945861876010895,0.7619047619047619,0.43243243243243246,0.68671875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6626308049825587,0.0 +0.7484893798828125,0.7619047619047619,0.43243243243243246,0.77265625,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7154166661741241,0.0 +0.9287281036376953,0.7619047619047619,0.43243243243243246,0.80859375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.7687674774801041,0.0 +0.8795062899589539,0.7619047619047619,0.43243243243243246,0.79375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.750414290631294,0.0 +0.8250083923339844,0.7619047619047619,0.43243243243243246,0.75546875,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6976555889940705,0.0 +0.8058834075927734,0.7619047619047619,0.43243243243243246,0.7093750000000001,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6559102729397122,0.0 +0.7665120959281921,0.7619047619047619,0.43243243243243246,0.734375,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6563890461269495,0.0 +0.8527358770370483,0.7619047619047619,0.43243243243243246,0.75,0.23809523809523808,0.38095238095238093,0.38095238095238093,0.5,0.6717796242170622,0.0 +0.6245525479316711,0.7142857142857143,0.5367647058823529,0.75,0.2857142857142857,0.4214285714285714,0.5555555555555556,0.5375,0.6939900141268103,0.08695652173913038 +0.7788131833076477,0.7619047619047619,0.6139705882352942,0.79375,0.23809523809523808,0.4871794871794872,0.6527777777777778,0.603125,0.730369188055292,0.23913043478260865 +0.5484565794467926,0.7619047619047619,0.671875,0.815625,0.23809523809523808,0.5315315315315315,0.671875,0.671875,0.7518565475904775,0.34375 +1.277198076248169,0.7857142857142857,0.590465872156013,0.85,0.21428571428571427,0.4784090909090909,0.7307692307692307,0.584375,0.80314870255797,0.22222222222222232 +0.9774696230888367,0.8333333333333334,0.7159420289855072,0.84453125,0.16666666666666666,0.5897129186602871,0.818918918918919,0.684375,0.8015047226759677,0.44528301886792454 +0.8751531839370728,0.8095238095238095,0.6911764705882353,0.86796875,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.8321278862142879,0.3913043478260869 +0.7527441382408142,0.8095238095238095,0.7375,0.8531249999999999,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8224014286197165,0.475 +0.950438380241394,0.8333333333333334,0.7418788410886743,0.859375,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8066661561331343,0.4878048780487805 +1.0000706911087036,0.8095238095238095,0.6911764705882353,0.84375,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.7957254531473945,0.3913043478260869 +0.8677769899368286,0.8095238095238095,0.7171717171717171,0.815625,0.19047619047619047,0.5841995841995842,0.7389705882352942,0.703125,0.75446910044971,0.43624161073825496 +1.1338821053504944,0.7857142857142857,0.6681299385425812,0.8070312499999999,0.21428571428571427,0.535425101214575,0.7,0.653125,0.7543627853796022,0.3414634146341464 +0.6370832324028015,0.7857142857142857,0.7142857142857142,0.825,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.785276188218651,0.42900302114803623 +0.6147848665714264,0.7857142857142857,0.7142857142857142,0.8531249999999999,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8130521054771913,0.42900302114803623 +0.6647389531135559,0.8571428571428571,0.7878787878787878,0.859375,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.830052265583969,0.5771812080536913 +0.7310305237770081,0.8333333333333334,0.7418788410886743,0.840625,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.8170925996103228,0.4878048780487805 +0.8558298349380493,0.8095238095238095,0.7375,0.828125,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8103413120078523,0.475 +0.7550559043884277,0.8095238095238095,0.7375,0.821875,0.19047619047619047,0.6031746031746031,0.7375,0.7375,0.8070984081914674,0.475 +0.757858157157898,0.8571428571428571,0.7878787878787878,0.815625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7927907195509698,0.5771812080536913 +0.659028172492981,0.7380952380952381,0.6707056307911619,0.82890625,0.2619047619047619,0.5236928104575164,0.6618037135278514,0.690625,0.7990907690725269,0.34560906515580736 +1.1868010759353638,0.8095238095238095,0.6911764705882353,0.8179687499999999,0.19047619047619047,0.5614035087719298,0.75,0.66875,0.79050604102645,0.3913043478260869 +0.6475030779838562,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.8007156258962138,0.5242718446601942 +0.5518777370452881,0.7857142857142857,0.7142857142857142,0.84375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8078735127959484,0.42900302114803623 +0.650142639875412,0.7857142857142857,0.7142857142857142,0.84453125,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8077895392728555,0.42900302114803623 +0.6480240523815155,0.7857142857142857,0.7142857142857142,0.84609375,0.21428571428571427,0.575,0.7082111436950147,0.721875,0.8087281862237307,0.42900302114803623 +0.812717080116272,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8008065294158258,0.5771812080536913 +0.7290194034576416,0.8571428571428571,0.7878787878787878,0.8359375,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.8024211127491591,0.5771812080536913 +0.9215718805789948,0.8333333333333334,0.7418788410886743,0.828125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7963100577502364,0.4878048780487805 +0.9986901581287384,0.8333333333333334,0.7418788410886743,0.83125,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7981890016514915,0.4878048780487805 +0.8881509304046631,0.8333333333333334,0.7418788410886743,0.825,0.16666666666666666,0.6137387387387387,0.7857142857142857,0.71875,0.7945901607993612,0.4878048780487805 +0.7914236485958099,0.8571428571428571,0.7878787878787878,0.82890625,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929611826223955,0.5771812080536913 +0.7529699206352234,0.8571428571428571,0.7878787878787878,0.8343750000000001,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7963039314782308,0.5771812080536913 +0.7332981824874878,0.8333333333333334,0.7619433198380567,0.83359375,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7957310148115639,0.5242718446601942 +0.7276108860969543,0.8333333333333334,0.7619433198380567,0.83125,0.16666666666666666,0.6335470085470085,0.7727272727272727,0.753125,0.7945614241682891,0.5242718446601942 +0.7322454154491425,0.8571428571428571,0.7878787878787878,0.83125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7945614241682891,0.5771812080536913 +0.7544125616550446,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7750901579856873,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7742511332035065,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7764365673065186,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7929513475101896,0.5771812080536913 +0.772107869386673,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7689108848571777,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.769738107919693,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 +0.7703390717506409,0.8571428571428571,0.7878787878787878,0.828125,0.14285714285714285,0.6666666666666667,0.8161764705882353,0.76875,0.7928743770668398,0.5771812080536913 diff --git a/results/papila/retfound/pr.png b/results/papila/retfound/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..567bef62299e8292c76bc9859243b2c14a03e919 --- /dev/null +++ b/results/papila/retfound/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da1e69338a15119bd42de1548ecd850a11f9fa7010485387bbacc23cc853bd79 +size 52261 diff --git a/results/papila/retfound/roc.png b/results/papila/retfound/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..6c1790997e7d1da62528b52c94485f129cb23717 --- /dev/null +++ b/results/papila/retfound/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:703e193e0dfa2492530843c855420951e3e1b0af10a25f4df2ae2fa4e96aeaf5 +size 57625 diff --git a/results/papila/retfound/test_pred.npz b/results/papila/retfound/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..3d702b56e1fa0fa773506bb7a507fe8432118b5c --- /dev/null +++ b/results/papila/retfound/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:288713ad75144f9b05b057382287103c78c0443830c6b60555e9c81300d12da3 +size 1518 diff --git a/results/papila/retfound/train.log b/results/papila/retfound/train.log new file mode 100644 index 0000000000000000000000000000000000000000..117789330998cdd4ae3783666ce2863f41ab2799 --- /dev/null +++ b/results/papila/retfound/train.log @@ -0,0 +1,733 @@ +| distributed init (rank 0): env://, gpu 0 +[rank0]:[W615 13:55:10.354084031 ProcessGroupNCCL.cpp:4115] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call init_process_group() with a device_id. +[13:55:11.545923] job dir: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound +[13:55:11.546174] Namespace(batch_size=32, +epochs=50, +accum_iter=1, +model='RETFound_mae', +model_arch='retfound_mae', +input_size=224, +drop_path=0.2, +global_pool=True, +clip_grad=None, +weight_decay=0.05, +lr=None, +blr=0.005, +layer_decay=0.65, +min_lr=1e-06, +warmup_epochs=10, +color_jitter=None, +aa='rand-m9-mstd0.5-inc1', +smoothing=0.1, +reprob=0.25, +remode='pixel', +recount=1, +resplit=False, +mixup=0.0, +cutmix=0.0, +cutmix_minmax=None, +mixup_prob=1.0, +mixup_switch_prob=0.5, +mixup_mode='batch', +finetune='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth', +task='retfound', +adaptation='finetune', +data_path='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Dataset/Glaucoma/papila-retinal-fundus-images', +nb_classes=2, +output_dir='/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/papila', +log_dir='./output_logs', +dataratio='1.0', +stratified=False, +device='cuda', +seed=0, +resume='', +start_epoch=0, +eval=False, +dist_eval=False, +num_workers=10, +pin_mem=True, +world_size=1, +local_rank=-1, +dist_on_itp=False, +dist_url='env://', +savemodel=True, +norm='IMAGENET', +enhance=False, +datasets_seed=2026, +rank=0, +gpu=0, +distributed=True, +dist_backend='nccl') +[13:55:14.573869] Preparing to load pre-trained weights: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:55:16.251568] Loaded pre-trained checkpoint from: /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/RETFound/RETFound_mae_natureCFP/RETFound_mae_natureCFP.pth +[13:55:16.420278] Sampler_train = +[13:55:16.463318] len of train_set: 288 +[13:55:17.078876] [Adaptation] Full fine-tuning: training all parameters. +[13:55:17.079907] number of trainable params (M): 303.30 +[13:55:17.079997] base lr: 5.00e-03 +[13:55:17.080075] actual lr: 6.25e-04 +[13:55:17.080136] accumulate grad iterations: 1 +[13:55:17.080194] effective batch size: 32 +[13:55:17.082842] criterion = CrossEntropyLoss() +[13:55:17.082929] Start training for 50 epochs +[13:55:17.084626] log_dir: ./output_logs/retfound +[13:55:20.483583] Epoch: [0] [0/9] eta: 0:00:30 lr: 0.000000 loss: 0.6927 (0.6927) time: 3.3980 data: 2.6294 max mem: 7340 +[13:55:21.698288] Epoch: [0] [8/9] eta: 0:00:00 lr: 0.000056 loss: 0.6903 (0.6828) time: 0.5125 data: 0.2922 max mem: 9671 +[13:55:21.776250] Epoch: [0] Total time: 0:00:04 (0.5213 s / it) +[13:55:21.785382] Averaged stats: lr: 0.000056 loss: 0.6903 (0.6828) +[13:55:25.001921] val: [0/2] eta: 0:00:06 loss: 0.6260 (0.6260) time: 3.2048 data: 3.1583 max mem: 9671 +[13:55:25.098474] val: [1/2] eta: 0:00:01 loss: 0.6260 (0.6946) time: 1.6504 data: 1.5792 max mem: 9671 +[13:55:25.171147] val: Total time: 0:00:03 (1.6873 s / it) +[13:55:25.181945] val loss: 0.6945861876010895 +[13:55:25.182144] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.6867, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6626, Kappa: 0.0000, Score: 0.3731 +[13:55:27.022786] Best epoch = 0, Best score = 0.3731 +[13:55:27.098560] log_dir: ./output_logs/retfound +[13:55:29.697178] Epoch: [1] [0/9] eta: 0:00:23 lr: 0.000063 loss: 0.6313 (0.6313) time: 2.5976 data: 2.4267 max mem: 9671 +[13:55:30.839408] Epoch: [1] [8/9] eta: 0:00:00 lr: 0.000118 loss: 0.5999 (0.5992) time: 0.4154 data: 0.2698 max mem: 9671 +[13:55:30.917177] Epoch: [1] Total time: 0:00:03 (0.4243 s / it) +[13:55:30.926020] Averaged stats: lr: 0.000118 loss: 0.5999 (0.5992) +[13:55:33.857949] val: [0/2] eta: 0:00:05 loss: 0.3876 (0.3876) time: 2.9186 data: 2.8833 max mem: 9671 +[13:55:33.874062] val: [1/2] eta: 0:00:01 loss: 0.3876 (0.7485) time: 1.4671 data: 1.4417 max mem: 9671 +[13:55:33.958338] val: Total time: 0:00:03 (1.5099 s / it) +[13:55:33.968242] val loss: 0.7484893798828125 +[13:55:33.968472] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7727, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7154, Kappa: 0.0000, Score: 0.4017 +[13:55:35.783728] Best epoch = 1, Best score = 0.4017 +[13:55:35.848656] log_dir: ./output_logs/retfound +[13:55:38.418568] Epoch: [2] [0/9] eta: 0:00:23 lr: 0.000125 loss: 0.5245 (0.5245) time: 2.5689 data: 2.4255 max mem: 9671 +[13:55:39.562471] Epoch: [2] [8/9] eta: 0:00:00 lr: 0.000181 loss: 0.5136 (0.5173) time: 0.4124 data: 0.2696 max mem: 9671 +[13:55:39.639048] Epoch: [2] Total time: 0:00:03 (0.4211 s / it) +[13:55:39.648442] Averaged stats: lr: 0.000181 loss: 0.5136 (0.5173) +[13:55:42.614814] val: [0/2] eta: 0:00:05 loss: 0.1880 (0.1880) time: 2.9498 data: 2.9133 max mem: 9671 +[13:55:42.631221] val: [1/2] eta: 0:00:01 loss: 0.1880 (0.9287) time: 1.4828 data: 1.4567 max mem: 9671 +[13:55:42.704105] val: Total time: 0:00:03 (1.5199 s / it) +[13:55:42.713139] val loss: 0.9287281036376953 +[13:55:42.713413] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.8086, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7688, Kappa: 0.0000, Score: 0.4137 +[13:55:44.913551] Best epoch = 2, Best score = 0.4137 +[13:55:44.981623] log_dir: ./output_logs/retfound +[13:55:47.444600] Epoch: [3] [0/9] eta: 0:00:22 lr: 0.000188 loss: 0.4505 (0.4505) time: 2.4620 data: 2.3148 max mem: 9671 +[13:55:48.589818] Epoch: [3] [8/9] eta: 0:00:00 lr: 0.000243 loss: 0.5323 (0.5203) time: 0.4007 data: 0.2573 max mem: 9671 +[13:55:48.672258] Epoch: [3] Total time: 0:00:03 (0.4101 s / it) +[13:55:48.680811] Averaged stats: lr: 0.000243 loss: 0.5323 (0.5203) +[13:55:51.433727] val: [0/2] eta: 0:00:05 loss: 0.2116 (0.2116) time: 2.7358 data: 2.7011 max mem: 9671 +[13:55:51.450054] val: [1/2] eta: 0:00:01 loss: 0.2116 (0.8795) time: 1.3758 data: 1.3506 max mem: 9671 +[13:55:51.524234] val: Total time: 0:00:02 (1.4136 s / it) +[13:55:51.533785] val loss: 0.8795062899589539 +[13:55:51.533991] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7937, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.7504, Kappa: 0.0000, Score: 0.4087 +[13:55:51.570998] Best epoch = 2, Best score = 0.4137 +[13:55:51.851112] log_dir: ./output_logs/retfound +[13:55:54.258521] Epoch: [4] [0/9] eta: 0:00:21 lr: 0.000250 loss: 0.5317 (0.5317) time: 2.4062 data: 2.2633 max mem: 9671 +[13:55:55.407332] Epoch: [4] [8/9] eta: 0:00:00 lr: 0.000306 loss: 0.5317 (0.4997) time: 0.3949 data: 0.2516 max mem: 9671 +[13:55:55.486196] Epoch: [4] Total time: 0:00:03 (0.4039 s / it) +[13:55:55.487406] Averaged stats: lr: 0.000306 loss: 0.5317 (0.4997) +[13:55:58.286097] val: [0/2] eta: 0:00:05 loss: 0.2462 (0.2462) time: 2.7812 data: 2.7467 max mem: 9671 +[13:55:58.302366] val: [1/2] eta: 0:00:01 loss: 0.2462 (0.8250) time: 1.3984 data: 1.3734 max mem: 9671 +[13:55:58.374272] val: Total time: 0:00:02 (1.4350 s / it) +[13:55:58.383235] val loss: 0.8250083923339844 +[13:55:58.383420] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7555, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6977, Kappa: 0.0000, Score: 0.3960 +[13:55:58.429296] Best epoch = 2, Best score = 0.4137 +[13:55:58.699189] log_dir: ./output_logs/retfound +[13:56:01.207850] Epoch: [5] [0/9] eta: 0:00:22 lr: 0.000313 loss: 0.4567 (0.4567) time: 2.5066 data: 2.3624 max mem: 9671 +[13:56:02.348379] Epoch: [5] [8/9] eta: 0:00:00 lr: 0.000368 loss: 0.4888 (0.5041) time: 0.4051 data: 0.2626 max mem: 9671 +[13:56:02.426998] Epoch: [5] Total time: 0:00:03 (0.4142 s / it) +[13:56:02.434829] Averaged stats: lr: 0.000368 loss: 0.4888 (0.5041) +[13:56:05.333761] val: [0/2] eta: 0:00:05 loss: 0.2470 (0.2470) time: 2.8821 data: 2.8476 max mem: 9671 +[13:56:05.350065] val: [1/2] eta: 0:00:01 loss: 0.2470 (0.8059) time: 1.4489 data: 1.4239 max mem: 9671 +[13:56:05.423618] val: Total time: 0:00:02 (1.4864 s / it) +[13:56:05.433504] val loss: 0.8058834075927734 +[13:56:05.433693] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7094, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6559, Kappa: 0.0000, Score: 0.3806 +[13:56:05.481081] Best epoch = 2, Best score = 0.4137 +[13:56:05.732462] log_dir: ./output_logs/retfound +[13:56:08.242069] Epoch: [6] [0/9] eta: 0:00:22 lr: 0.000375 loss: 0.5451 (0.5451) time: 2.5084 data: 2.3573 max mem: 9671 +[13:56:09.383518] Epoch: [6] [8/9] eta: 0:00:00 lr: 0.000431 loss: 0.5040 (0.4901) time: 0.4054 data: 0.2621 max mem: 9671 +[13:56:09.461017] Epoch: [6] Total time: 0:00:03 (0.4143 s / it) +[13:56:09.470038] Averaged stats: lr: 0.000431 loss: 0.5040 (0.4901) +[13:56:12.367358] val: [0/2] eta: 0:00:05 loss: 0.2326 (0.2326) time: 2.8852 data: 2.8515 max mem: 9671 +[13:56:12.383652] val: [1/2] eta: 0:00:01 loss: 0.2326 (0.7665) time: 1.4505 data: 1.4258 max mem: 9671 +[13:56:12.470044] val: Total time: 0:00:02 (1.4943 s / it) +[13:56:12.479101] val loss: 0.7665120959281921 +[13:56:12.479281] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7344, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6564, Kappa: 0.0000, Score: 0.3889 +[13:56:12.521325] Best epoch = 2, Best score = 0.4137 +[13:56:12.827271] log_dir: ./output_logs/retfound +[13:56:15.222740] Epoch: [7] [0/9] eta: 0:00:21 lr: 0.000438 loss: 0.4717 (0.4717) time: 2.3944 data: 2.2490 max mem: 9671 +[13:56:16.360957] Epoch: [7] [8/9] eta: 0:00:00 lr: 0.000493 loss: 0.4926 (0.5020) time: 0.3924 data: 0.2500 max mem: 9671 +[13:56:16.445032] Epoch: [7] Total time: 0:00:03 (0.4020 s / it) +[13:56:16.453966] Averaged stats: lr: 0.000493 loss: 0.4926 (0.5020) +[13:56:19.260872] val: [0/2] eta: 0:00:05 loss: 0.1709 (0.1709) time: 2.7936 data: 2.7580 max mem: 9671 +[13:56:19.276982] val: [1/2] eta: 0:00:01 loss: 0.1709 (0.8527) time: 1.4046 data: 1.3791 max mem: 9671 +[13:56:19.349674] val: Total time: 0:00:02 (1.4416 s / it) +[13:56:19.358501] val loss: 0.8527358770370483 +[13:56:19.358688] Accuracy: 0.7619, F1 Score: 0.4324, ROC AUC: 0.7500, Hamming Loss: 0.2381, + Jaccard Score: 0.3810, Precision: 0.3810, Recall: 0.5000, + Average Precision: 0.6718, Kappa: 0.0000, Score: 0.3941 +[13:56:19.400024] Best epoch = 2, Best score = 0.4137 +[13:56:19.646187] log_dir: ./output_logs/retfound +[13:56:22.140708] Epoch: [8] [0/9] eta: 0:00:22 lr: 0.000500 loss: 0.4438 (0.4438) time: 2.4934 data: 2.3450 max mem: 9671 +[13:56:23.287137] Epoch: [8] [8/9] eta: 0:00:00 lr: 0.000556 loss: 0.4622 (0.4714) time: 0.4043 data: 0.2607 max mem: 9671 +[13:56:23.365739] Epoch: [8] Total time: 0:00:03 (0.4133 s / it) +[13:56:23.375656] Averaged stats: lr: 0.000556 loss: 0.4622 (0.4714) +[13:56:26.235775] val: [0/2] eta: 0:00:05 loss: 0.4000 (0.4000) time: 2.8455 data: 2.8096 max mem: 9671 +[13:56:26.252167] val: [1/2] eta: 0:00:01 loss: 0.4000 (0.6246) time: 1.4306 data: 1.4048 max mem: 9671 +[13:56:26.324512] val: Total time: 0:00:02 (1.4675 s / it) +[13:56:26.334480] val loss: 0.6245525479316711 +[13:56:26.334719] Accuracy: 0.7143, F1 Score: 0.5368, ROC AUC: 0.7500, Hamming Loss: 0.2857, + Jaccard Score: 0.4214, Precision: 0.5556, Recall: 0.5375, + Average Precision: 0.6940, Kappa: 0.0870, Score: 0.4579 +[13:56:28.197922] Best epoch = 8, Best score = 0.4579 +[13:56:28.266079] log_dir: ./output_logs/retfound +[13:56:30.722639] Epoch: [9] [0/9] eta: 0:00:22 lr: 0.000562 loss: 0.5492 (0.5492) time: 2.4555 data: 2.3100 max mem: 9671 +[13:56:31.872352] Epoch: [9] [8/9] eta: 0:00:00 lr: 0.000618 loss: 0.4048 (0.4486) time: 0.4005 data: 0.2567 max mem: 9671 +[13:56:31.944295] Epoch: [9] Total time: 0:00:03 (0.4087 s / it) +[13:56:31.948068] Averaged stats: lr: 0.000618 loss: 0.4048 (0.4486) +[13:56:34.779051] val: [0/2] eta: 0:00:05 loss: 0.1913 (0.1913) time: 2.8157 data: 2.7803 max mem: 9671 +[13:56:34.790787] val: [1/2] eta: 0:00:01 loss: 0.1913 (0.7788) time: 1.4134 data: 1.3902 max mem: 9671 +[13:56:34.868713] val: Total time: 0:00:02 (1.4531 s / it) +[13:56:34.877641] val loss: 0.7788131833076477 +[13:56:34.877829] Accuracy: 0.7619, F1 Score: 0.6140, ROC AUC: 0.7937, Hamming Loss: 0.2381, + Jaccard Score: 0.4872, Precision: 0.6528, Recall: 0.6031, + Average Precision: 0.7304, Kappa: 0.2391, Score: 0.5490 +[13:56:36.708335] Best epoch = 9, Best score = 0.5490 +[13:56:36.771816] log_dir: ./output_logs/retfound +[13:56:39.205278] Epoch: [10] [0/9] eta: 0:00:21 lr: 0.000625 loss: 0.4722 (0.4722) time: 2.4325 data: 2.2868 max mem: 9671 +[13:56:40.351172] Epoch: [10] [8/9] eta: 0:00:00 lr: 0.000624 loss: 0.4722 (0.4588) time: 0.3975 data: 0.2542 max mem: 9671 +[13:56:40.427787] Epoch: [10] Total time: 0:00:03 (0.4062 s / it) +[13:56:40.437538] Averaged stats: lr: 0.000624 loss: 0.4722 (0.4588) +[13:56:43.195614] val: [0/2] eta: 0:00:05 loss: 0.4142 (0.4142) time: 2.7416 data: 2.7056 max mem: 9671 +[13:56:43.212199] val: [1/2] eta: 0:00:01 loss: 0.4142 (0.5485) time: 1.3788 data: 1.3529 max mem: 9671 +[13:56:43.284403] val: Total time: 0:00:02 (1.4155 s / it) +[13:56:43.294236] val loss: 0.5484565794467926 +[13:56:43.294435] Accuracy: 0.7619, F1 Score: 0.6719, ROC AUC: 0.8156, Hamming Loss: 0.2381, + Jaccard Score: 0.5315, Precision: 0.6719, Recall: 0.6719, + Average Precision: 0.7519, Kappa: 0.3438, Score: 0.6104 +[13:56:45.177430] Best epoch = 10, Best score = 0.6104 +[13:56:45.239251] log_dir: ./output_logs/retfound +[13:56:47.584440] Epoch: [11] [0/9] eta: 0:00:21 lr: 0.000624 loss: 0.3691 (0.3691) time: 2.3443 data: 2.1984 max mem: 9671 +[13:56:48.721132] Epoch: [11] [8/9] eta: 0:00:00 lr: 0.000622 loss: 0.4582 (0.4270) time: 0.3867 data: 0.2444 max mem: 9671 +[13:56:48.802658] Epoch: [11] Total time: 0:00:03 (0.3959 s / it) +[13:56:48.811779] Averaged stats: lr: 0.000622 loss: 0.4582 (0.4270) +[13:56:51.628811] val: [0/2] eta: 0:00:05 loss: 0.0844 (0.0844) time: 2.7990 data: 2.7626 max mem: 9671 +[13:56:51.645186] val: [1/2] eta: 0:00:01 loss: 0.0844 (1.2772) time: 1.4074 data: 1.3813 max mem: 9671 +[13:56:51.715710] val: Total time: 0:00:02 (1.4434 s / it) +[13:56:51.725488] val loss: 1.277198076248169 +[13:56:51.725677] Accuracy: 0.7857, F1 Score: 0.5905, ROC AUC: 0.8500, Hamming Loss: 0.2143, + Jaccard Score: 0.4784, Precision: 0.7308, Recall: 0.5844, + Average Precision: 0.8031, Kappa: 0.2222, Score: 0.5542 +[13:56:51.764844] Best epoch = 10, Best score = 0.6104 +[13:56:51.998162] log_dir: ./output_logs/retfound +[13:56:54.265414] Epoch: [12] [0/9] eta: 0:00:20 lr: 0.000621 loss: 0.5461 (0.5461) time: 2.2660 data: 2.1209 max mem: 9671 +[13:56:55.449631] Epoch: [12] [8/9] eta: 0:00:00 lr: 0.000617 loss: 0.4572 (0.4496) time: 0.3833 data: 0.2401 max mem: 9671 +[13:56:55.522998] Epoch: [12] Total time: 0:00:03 (0.3916 s / it) +[13:56:55.532154] Averaged stats: lr: 0.000617 loss: 0.4572 (0.4496) +[13:56:58.196655] val: [0/2] eta: 0:00:05 loss: 0.1181 (0.1181) time: 2.6465 data: 2.6106 max mem: 9671 +[13:56:58.212855] val: [1/2] eta: 0:00:01 loss: 0.1181 (0.9775) time: 1.3310 data: 1.3054 max mem: 9671 +[13:56:58.285744] val: Total time: 0:00:02 (1.3682 s / it) +[13:56:58.294762] val loss: 0.9774696230888367 +[13:56:58.294985] Accuracy: 0.8333, F1 Score: 0.7159, ROC AUC: 0.8445, Hamming Loss: 0.1667, + Jaccard Score: 0.5897, Precision: 0.8189, Recall: 0.6844, + Average Precision: 0.8015, Kappa: 0.4453, Score: 0.6686 +[13:57:00.094053] Best epoch = 12, Best score = 0.6686 +[13:57:00.180619] log_dir: ./output_logs/retfound +[13:57:02.617042] Epoch: [13] [0/9] eta: 0:00:21 lr: 0.000616 loss: 0.6730 (0.6730) time: 2.4354 data: 2.2968 max mem: 9671 +[13:57:03.757411] Epoch: [13] [8/9] eta: 0:00:00 lr: 0.000611 loss: 0.4303 (0.4471) time: 0.3972 data: 0.2553 max mem: 9671 +[13:57:03.839957] Epoch: [13] Total time: 0:00:03 (0.4066 s / it) +[13:57:03.848309] Averaged stats: lr: 0.000611 loss: 0.4303 (0.4471) +[13:57:06.928618] val: [0/2] eta: 0:00:06 loss: 0.1100 (0.1100) time: 3.0661 data: 3.0289 max mem: 9671 +[13:57:06.947128] val: [1/2] eta: 0:00:01 loss: 0.1100 (0.8752) time: 1.5419 data: 1.5146 max mem: 9671 +[13:57:07.020355] val: Total time: 0:00:03 (1.5794 s / it) +[13:57:07.034973] val loss: 0.8751531839370728 +[13:57:07.035285] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8680, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.8321, Kappa: 0.3913, Score: 0.6501 +[13:57:07.075865] Best epoch = 12, Best score = 0.6686 +[13:57:07.341394] log_dir: ./output_logs/retfound +[13:57:10.073758] Epoch: [14] [0/9] eta: 0:00:24 lr: 0.000610 loss: 0.3318 (0.3318) time: 2.7310 data: 2.5772 max mem: 9671 +[13:57:11.216598] Epoch: [14] [8/9] eta: 0:00:00 lr: 0.000602 loss: 0.4897 (0.4556) time: 0.4303 data: 0.2865 max mem: 9671 +[13:57:11.296103] Epoch: [14] Total time: 0:00:03 (0.4394 s / it) +[13:57:11.305060] Averaged stats: lr: 0.000602 loss: 0.4897 (0.4556) +[13:57:14.083553] val: [0/2] eta: 0:00:05 loss: 0.3347 (0.3347) time: 2.7619 data: 2.7251 max mem: 9671 +[13:57:14.099808] val: [1/2] eta: 0:00:01 loss: 0.3347 (0.7527) time: 1.3888 data: 1.3626 max mem: 9671 +[13:57:14.176037] val: Total time: 0:00:02 (1.4276 s / it) +[13:57:14.189734] val loss: 0.7527441382408142 +[13:57:14.189973] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8531, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8224, Kappa: 0.4750, Score: 0.6885 +[13:57:16.033160] Best epoch = 14, Best score = 0.6885 +[13:57:16.096444] log_dir: ./output_logs/retfound +[13:57:18.491973] Epoch: [15] [0/9] eta: 0:00:21 lr: 0.000601 loss: 0.4373 (0.4373) time: 2.3947 data: 2.2497 max mem: 9671 +[13:57:19.627140] Epoch: [15] [8/9] eta: 0:00:00 lr: 0.000592 loss: 0.3949 (0.3853) time: 0.3921 data: 0.2502 max mem: 9671 +[13:57:19.709985] Epoch: [15] Total time: 0:00:03 (0.4015 s / it) +[13:57:19.719888] Averaged stats: lr: 0.000592 loss: 0.3949 (0.3853) +[13:57:22.573012] val: [0/2] eta: 0:00:05 loss: 0.2078 (0.2078) time: 2.8347 data: 2.7999 max mem: 9671 +[13:57:22.591807] val: [1/2] eta: 0:00:01 loss: 0.2078 (0.9504) time: 1.4265 data: 1.4000 max mem: 9671 +[13:57:22.669636] val: Total time: 0:00:02 (1.4661 s / it) +[13:57:22.679520] val loss: 0.950438380241394 +[13:57:22.679745] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8594, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8067, Kappa: 0.4878, Score: 0.6964 +[13:57:24.471275] Best epoch = 15, Best score = 0.6964 +[13:57:24.538978] log_dir: ./output_logs/retfound +[13:57:26.880706] Epoch: [16] [0/9] eta: 0:00:21 lr: 0.000591 loss: 0.6279 (0.6279) time: 2.3405 data: 2.1914 max mem: 9671 +[13:57:28.020407] Epoch: [16] [8/9] eta: 0:00:00 lr: 0.000580 loss: 0.4289 (0.4627) time: 0.3866 data: 0.2436 max mem: 9671 +[13:57:28.093996] Epoch: [16] Total time: 0:00:03 (0.3950 s / it) +[13:57:28.102566] Averaged stats: lr: 0.000580 loss: 0.4289 (0.4627) +[13:57:31.046611] val: [0/2] eta: 0:00:05 loss: 0.1866 (0.1866) time: 2.9269 data: 2.8919 max mem: 9671 +[13:57:31.062877] val: [1/2] eta: 0:00:01 loss: 0.1866 (1.0001) time: 1.4713 data: 1.4460 max mem: 9671 +[13:57:31.131714] val: Total time: 0:00:03 (1.5064 s / it) +[13:57:31.141318] val loss: 1.0000706911087036 +[13:57:31.141591] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8438, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7957, Kappa: 0.3913, Score: 0.6421 +[13:57:31.179114] Best epoch = 15, Best score = 0.6964 +[13:57:31.486042] log_dir: ./output_logs/retfound +[13:57:33.970724] Epoch: [17] [0/9] eta: 0:00:22 lr: 0.000579 loss: 0.4267 (0.4267) time: 2.4834 data: 2.3348 max mem: 9671 +[13:57:35.107316] Epoch: [17] [8/9] eta: 0:00:00 lr: 0.000567 loss: 0.4068 (0.4493) time: 0.4021 data: 0.2595 max mem: 9671 +[13:57:35.182115] Epoch: [17] Total time: 0:00:03 (0.4107 s / it) +[13:57:35.191768] Averaged stats: lr: 0.000567 loss: 0.4068 (0.4493) +[13:57:37.977767] val: [0/2] eta: 0:00:05 loss: 0.2248 (0.2248) time: 2.7726 data: 2.7379 max mem: 9671 +[13:57:37.994005] val: [1/2] eta: 0:00:01 loss: 0.2248 (0.8678) time: 1.3941 data: 1.3690 max mem: 9671 +[13:57:38.062326] val: Total time: 0:00:02 (1.4289 s / it) +[13:57:38.071325] val loss: 0.8677769899368286 +[13:57:38.071526] Accuracy: 0.8095, F1 Score: 0.7172, ROC AUC: 0.8156, Hamming Loss: 0.1905, + Jaccard Score: 0.5842, Precision: 0.7390, Recall: 0.7031, + Average Precision: 0.7545, Kappa: 0.4362, Score: 0.6563 +[13:57:38.112988] Best epoch = 15, Best score = 0.6964 +[13:57:38.352388] log_dir: ./output_logs/retfound +[13:57:40.753554] Epoch: [18] [0/9] eta: 0:00:21 lr: 0.000565 loss: 0.4027 (0.4027) time: 2.4001 data: 2.2544 max mem: 9671 +[13:57:41.894373] Epoch: [18] [8/9] eta: 0:00:00 lr: 0.000552 loss: 0.4138 (0.4269) time: 0.3933 data: 0.2505 max mem: 9671 +[13:57:41.978641] Epoch: [18] Total time: 0:00:03 (0.4029 s / it) +[13:57:41.988404] Averaged stats: lr: 0.000552 loss: 0.4138 (0.4269) +[13:57:44.782008] val: [0/2] eta: 0:00:05 loss: 0.2075 (0.2075) time: 2.7798 data: 2.7448 max mem: 9671 +[13:57:44.800921] val: [1/2] eta: 0:00:01 loss: 0.2075 (1.1339) time: 1.3989 data: 1.3724 max mem: 9671 +[13:57:44.881360] val: Total time: 0:00:02 (1.4400 s / it) +[13:57:44.890232] val loss: 1.1338821053504944 +[13:57:44.890489] Accuracy: 0.7857, F1 Score: 0.6681, ROC AUC: 0.8070, Hamming Loss: 0.2143, + Jaccard Score: 0.5354, Precision: 0.7000, Recall: 0.6531, + Average Precision: 0.7544, Kappa: 0.3415, Score: 0.6055 +[13:57:44.932203] Best epoch = 15, Best score = 0.6964 +[13:57:45.195248] log_dir: ./output_logs/retfound +[13:57:47.537293] Epoch: [19] [0/9] eta: 0:00:21 lr: 0.000550 loss: 0.5522 (0.5522) time: 2.3410 data: 2.1942 max mem: 9671 +[13:57:48.679929] Epoch: [19] [8/9] eta: 0:00:00 lr: 0.000536 loss: 0.4279 (0.3996) time: 0.3870 data: 0.2439 max mem: 9671 +[13:57:48.752678] Epoch: [19] Total time: 0:00:03 (0.3953 s / it) +[13:57:48.762155] Averaged stats: lr: 0.000536 loss: 0.4279 (0.3996) +[13:57:51.764772] val: [0/2] eta: 0:00:05 loss: 0.4255 (0.4255) time: 2.9867 data: 2.9523 max mem: 9671 +[13:57:51.781134] val: [1/2] eta: 0:00:01 loss: 0.4255 (0.6371) time: 1.5012 data: 1.4762 max mem: 9671 +[13:57:51.859642] val: Total time: 0:00:03 (1.5412 s / it) +[13:57:51.868560] val loss: 0.6370832324028015 +[13:57:51.868761] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8250, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.7853, Kappa: 0.4290, Score: 0.6561 +[13:57:51.910602] Best epoch = 15, Best score = 0.6964 +[13:57:52.189764] log_dir: ./output_logs/retfound +[13:57:54.680371] Epoch: [20] [0/9] eta: 0:00:22 lr: 0.000534 loss: 0.3223 (0.3223) time: 2.4896 data: 2.3417 max mem: 9671 +[13:57:55.818299] Epoch: [20] [8/9] eta: 0:00:00 lr: 0.000518 loss: 0.3818 (0.3802) time: 0.4029 data: 0.2603 max mem: 9671 +[13:57:55.901398] Epoch: [20] Total time: 0:00:03 (0.4124 s / it) +[13:57:55.910348] Averaged stats: lr: 0.000518 loss: 0.3818 (0.3802) +[13:57:58.832345] val: [0/2] eta: 0:00:05 loss: 0.3854 (0.3854) time: 2.9037 data: 2.8690 max mem: 9671 +[13:57:58.848574] val: [1/2] eta: 0:00:01 loss: 0.3854 (0.6148) time: 1.4597 data: 1.4346 max mem: 9671 +[13:57:58.920551] val: Total time: 0:00:02 (1.4964 s / it) +[13:57:58.929424] val loss: 0.6147848665714264 +[13:57:58.929617] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8531, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8131, Kappa: 0.4290, Score: 0.6655 +[13:57:58.969708] Best epoch = 15, Best score = 0.6964 +[13:57:59.240245] log_dir: ./output_logs/retfound +[13:58:01.815043] Epoch: [21] [0/9] eta: 0:00:23 lr: 0.000516 loss: 0.4815 (0.4815) time: 2.5737 data: 2.4276 max mem: 9671 +[13:58:02.959166] Epoch: [21] [8/9] eta: 0:00:00 lr: 0.000499 loss: 0.3667 (0.3990) time: 0.4130 data: 0.2698 max mem: 9671 +[13:58:03.040258] Epoch: [21] Total time: 0:00:03 (0.4222 s / it) +[13:58:03.049863] Averaged stats: lr: 0.000499 loss: 0.3667 (0.3990) +[13:58:05.994263] val: [0/2] eta: 0:00:05 loss: 0.3414 (0.3414) time: 2.9258 data: 2.8906 max mem: 9671 +[13:58:06.010635] val: [1/2] eta: 0:00:01 loss: 0.3414 (0.6647) time: 1.4708 data: 1.4454 max mem: 9671 +[13:58:06.084263] val: Total time: 0:00:03 (1.5083 s / it) +[13:58:06.093218] val loss: 0.6647389531135559 +[13:58:06.093465] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8594, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8301, Kappa: 0.5772, Score: 0.7415 +[13:58:07.867128] Best epoch = 21, Best score = 0.7415 +[13:58:07.932332] log_dir: ./output_logs/retfound +[13:58:10.532107] Epoch: [22] [0/9] eta: 0:00:23 lr: 0.000496 loss: 0.3751 (0.3751) time: 2.5988 data: 2.4529 max mem: 9671 +[13:58:11.673140] Epoch: [22] [8/9] eta: 0:00:00 lr: 0.000478 loss: 0.4090 (0.3949) time: 0.4154 data: 0.2726 max mem: 9671 +[13:58:11.742760] Epoch: [22] Total time: 0:00:03 (0.4234 s / it) +[13:58:11.751279] Averaged stats: lr: 0.000478 loss: 0.4090 (0.3949) +[13:58:14.501400] val: [0/2] eta: 0:00:05 loss: 0.2182 (0.2182) time: 2.7419 data: 2.7078 max mem: 9671 +[13:58:14.518124] val: [1/2] eta: 0:00:01 loss: 0.2182 (0.7310) time: 1.3790 data: 1.3540 max mem: 9671 +[13:58:14.593911] val: Total time: 0:00:02 (1.4176 s / it) +[13:58:14.603455] val loss: 0.7310305237770081 +[13:58:14.603616] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8406, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.8171, Kappa: 0.4878, Score: 0.6901 +[13:58:14.643765] Best epoch = 21, Best score = 0.7415 +[13:58:14.910244] log_dir: ./output_logs/retfound +[13:58:17.541475] Epoch: [23] [0/9] eta: 0:00:23 lr: 0.000476 loss: 0.4109 (0.4109) time: 2.6302 data: 2.4844 max mem: 9671 +[13:58:18.687482] Epoch: [23] [8/9] eta: 0:00:00 lr: 0.000457 loss: 0.3560 (0.3675) time: 0.4195 data: 0.2762 max mem: 9671 +[13:58:18.763670] Epoch: [23] Total time: 0:00:03 (0.4281 s / it) +[13:58:18.771747] Averaged stats: lr: 0.000457 loss: 0.3560 (0.3675) +[13:58:21.597030] val: [0/2] eta: 0:00:05 loss: 0.4150 (0.4150) time: 2.8102 data: 2.7756 max mem: 9671 +[13:58:21.613241] val: [1/2] eta: 0:00:01 loss: 0.4150 (0.8558) time: 1.4129 data: 1.3879 max mem: 9671 +[13:58:21.687499] val: Total time: 0:00:02 (1.4507 s / it) +[13:58:21.696265] val loss: 0.8558298349380493 +[13:58:21.696449] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8281, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8103, Kappa: 0.4750, Score: 0.6802 +[13:58:21.736272] Best epoch = 21, Best score = 0.7415 +[13:58:22.010331] log_dir: ./output_logs/retfound +[13:58:24.392135] Epoch: [24] [0/9] eta: 0:00:21 lr: 0.000455 loss: 0.3960 (0.3960) time: 2.3808 data: 2.2343 max mem: 9671 +[13:58:25.526646] Epoch: [24] [8/9] eta: 0:00:00 lr: 0.000435 loss: 0.3916 (0.3996) time: 0.3905 data: 0.2483 max mem: 9671 +[13:58:25.599367] Epoch: [24] Total time: 0:00:03 (0.3988 s / it) +[13:58:25.609065] Averaged stats: lr: 0.000435 loss: 0.3916 (0.3996) +[13:58:28.429120] val: [0/2] eta: 0:00:05 loss: 0.3662 (0.3662) time: 2.8069 data: 2.7725 max mem: 9671 +[13:58:28.445609] val: [1/2] eta: 0:00:01 loss: 0.3662 (0.7551) time: 1.4114 data: 1.3863 max mem: 9671 +[13:58:28.519847] val: Total time: 0:00:02 (1.4492 s / it) +[13:58:28.528783] val loss: 0.7550559043884277 +[13:58:28.528984] Accuracy: 0.8095, F1 Score: 0.7375, ROC AUC: 0.8219, Hamming Loss: 0.1905, + Jaccard Score: 0.6032, Precision: 0.7375, Recall: 0.7375, + Average Precision: 0.8071, Kappa: 0.4750, Score: 0.6781 +[13:58:28.569941] Best epoch = 21, Best score = 0.7415 +[13:58:28.829847] log_dir: ./output_logs/retfound +[13:58:31.281887] Epoch: [25] [0/9] eta: 0:00:22 lr: 0.000432 loss: 0.4206 (0.4206) time: 2.4509 data: 2.3048 max mem: 9671 +[13:58:32.435921] Epoch: [25] [8/9] eta: 0:00:00 lr: 0.000412 loss: 0.3320 (0.3284) time: 0.4004 data: 0.2573 max mem: 9671 +[13:58:32.513854] Epoch: [25] Total time: 0:00:03 (0.4093 s / it) +[13:58:32.523035] Averaged stats: lr: 0.000412 loss: 0.3320 (0.3284) +[13:58:35.370803] val: [0/2] eta: 0:00:05 loss: 0.2804 (0.2804) time: 2.8296 data: 2.7951 max mem: 9671 +[13:58:35.389176] val: [1/2] eta: 0:00:01 loss: 0.2804 (0.7579) time: 1.4237 data: 1.3976 max mem: 9671 +[13:58:35.463851] val: Total time: 0:00:02 (1.4618 s / it) +[13:58:35.472754] val loss: 0.757858157157898 +[13:58:35.472981] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8156, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7928, Kappa: 0.5772, Score: 0.7269 +[13:58:35.517340] Best epoch = 21, Best score = 0.7415 +[13:58:35.790271] log_dir: ./output_logs/retfound +[13:58:38.027605] Epoch: [26] [0/9] eta: 0:00:20 lr: 0.000409 loss: 0.2213 (0.2213) time: 2.2361 data: 2.0892 max mem: 9671 +[13:58:39.304195] Epoch: [26] [8/9] eta: 0:00:00 lr: 0.000388 loss: 0.3055 (0.3474) time: 0.3902 data: 0.2471 max mem: 9671 +[13:58:39.382243] Epoch: [26] Total time: 0:00:03 (0.3991 s / it) +[13:58:39.390777] Averaged stats: lr: 0.000388 loss: 0.3055 (0.3474) +[13:58:42.191556] val: [0/2] eta: 0:00:05 loss: 0.5462 (0.5462) time: 2.7837 data: 2.7490 max mem: 9671 +[13:58:42.208224] val: [1/2] eta: 0:00:01 loss: 0.5462 (0.6590) time: 1.3999 data: 1.3746 max mem: 9671 +[13:58:42.282801] val: Total time: 0:00:02 (1.4378 s / it) +[13:58:42.292037] val loss: 0.659028172492981 +[13:58:42.292220] Accuracy: 0.7381, F1 Score: 0.6707, ROC AUC: 0.8289, Hamming Loss: 0.2619, + Jaccard Score: 0.5237, Precision: 0.6618, Recall: 0.6906, + Average Precision: 0.7991, Kappa: 0.3456, Score: 0.6151 +[13:58:42.337095] Best epoch = 21, Best score = 0.7415 +[13:58:42.589917] log_dir: ./output_logs/retfound +[13:58:45.235098] Epoch: [27] [0/9] eta: 0:00:23 lr: 0.000386 loss: 0.2824 (0.2824) time: 2.6441 data: 2.4990 max mem: 9671 +[13:58:46.371987] Epoch: [27] [8/9] eta: 0:00:00 lr: 0.000364 loss: 0.3541 (0.3910) time: 0.4200 data: 0.2778 max mem: 9671 +[13:58:46.446941] Epoch: [27] Total time: 0:00:03 (0.4285 s / it) +[13:58:46.456383] Averaged stats: lr: 0.000364 loss: 0.3541 (0.3910) +[13:58:49.427904] val: [0/2] eta: 0:00:05 loss: 0.1278 (0.1278) time: 2.9537 data: 2.9193 max mem: 9671 +[13:58:49.443963] val: [1/2] eta: 0:00:01 loss: 0.1278 (1.1868) time: 1.4846 data: 1.4597 max mem: 9671 +[13:58:49.514609] val: Total time: 0:00:03 (1.5206 s / it) +[13:58:49.525887] val loss: 1.1868010759353638 +[13:58:49.526088] Accuracy: 0.8095, F1 Score: 0.6912, ROC AUC: 0.8180, Hamming Loss: 0.1905, + Jaccard Score: 0.5614, Precision: 0.7500, Recall: 0.6687, + Average Precision: 0.7905, Kappa: 0.3913, Score: 0.6335 +[13:58:49.566732] Best epoch = 21, Best score = 0.7415 +[13:58:49.830786] log_dir: ./output_logs/retfound +[13:58:52.362526] Epoch: [28] [0/9] eta: 0:00:22 lr: 0.000362 loss: 0.3359 (0.3359) time: 2.5308 data: 2.3845 max mem: 9671 +[13:58:53.506596] Epoch: [28] [8/9] eta: 0:00:00 lr: 0.000340 loss: 0.3616 (0.4176) time: 0.4082 data: 0.2650 max mem: 9671 +[13:58:53.588213] Epoch: [28] Total time: 0:00:03 (0.4175 s / it) +[13:58:53.597361] Averaged stats: lr: 0.000340 loss: 0.3616 (0.4176) +[13:58:56.391561] val: [0/2] eta: 0:00:05 loss: 0.2967 (0.2967) time: 2.7809 data: 2.7469 max mem: 9671 +[13:58:56.407810] val: [1/2] eta: 0:00:01 loss: 0.2967 (0.6475) time: 1.3983 data: 1.3735 max mem: 9671 +[13:58:56.477920] val: Total time: 0:00:02 (1.4341 s / it) +[13:58:56.486927] val loss: 0.6475030779838562 +[13:58:56.487144] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.8007, Kappa: 0.5243, Score: 0.7058 +[13:58:56.530298] Best epoch = 21, Best score = 0.7415 +[13:58:56.793185] log_dir: ./output_logs/retfound +[13:58:59.322198] Epoch: [29] [0/9] eta: 0:00:22 lr: 0.000337 loss: 0.3300 (0.3300) time: 2.5278 data: 2.3767 max mem: 9671 +[13:59:00.456608] Epoch: [29] [8/9] eta: 0:00:00 lr: 0.000316 loss: 0.3625 (0.3715) time: 0.4068 data: 0.2641 max mem: 9671 +[13:59:00.534007] Epoch: [29] Total time: 0:00:03 (0.4156 s / it) +[13:59:00.543467] Averaged stats: lr: 0.000316 loss: 0.3625 (0.3715) +[13:59:03.430469] val: [0/2] eta: 0:00:05 loss: 0.4714 (0.4714) time: 2.8732 data: 2.8365 max mem: 9671 +[13:59:03.448990] val: [1/2] eta: 0:00:01 loss: 0.4714 (0.5519) time: 1.4456 data: 1.4183 max mem: 9671 +[13:59:03.525270] val: Total time: 0:00:02 (1.4845 s / it) +[13:59:03.535022] val loss: 0.5518777370452881 +[13:59:03.535275] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8438, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8079, Kappa: 0.4290, Score: 0.6623 +[13:59:03.575141] Best epoch = 21, Best score = 0.7415 +[13:59:03.845049] log_dir: ./output_logs/retfound +[13:59:06.408334] Epoch: [30] [0/9] eta: 0:00:23 lr: 0.000313 loss: 0.3485 (0.3485) time: 2.5622 data: 2.4176 max mem: 9671 +[13:59:07.552719] Epoch: [30] [8/9] eta: 0:00:00 lr: 0.000291 loss: 0.4160 (0.3840) time: 0.4118 data: 0.2687 max mem: 9671 +[13:59:07.631875] Epoch: [30] Total time: 0:00:03 (0.4207 s / it) +[13:59:07.633885] Averaged stats: lr: 0.000291 loss: 0.4160 (0.3840) +[13:59:10.523531] val: [0/2] eta: 0:00:05 loss: 0.4225 (0.4225) time: 2.8728 data: 2.8367 max mem: 9671 +[13:59:10.540114] val: [1/2] eta: 0:00:01 loss: 0.4225 (0.6501) time: 1.4443 data: 1.4184 max mem: 9671 +[13:59:10.618593] val: Total time: 0:00:02 (1.4843 s / it) +[13:59:10.633248] val loss: 0.650142639875412 +[13:59:10.633465] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8445, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8078, Kappa: 0.4290, Score: 0.6626 +[13:59:10.678107] Best epoch = 21, Best score = 0.7415 +[13:59:10.944166] log_dir: ./output_logs/retfound +[13:59:13.533616] Epoch: [31] [0/9] eta: 0:00:23 lr: 0.000289 loss: 0.4108 (0.4108) time: 2.5882 data: 2.4425 max mem: 9671 +[13:59:14.675847] Epoch: [31] [8/9] eta: 0:00:00 lr: 0.000267 loss: 0.3838 (0.3420) time: 0.4144 data: 0.2715 max mem: 9671 +[13:59:14.787143] Epoch: [31] Total time: 0:00:03 (0.4270 s / it) +[13:59:14.796905] Averaged stats: lr: 0.000267 loss: 0.3838 (0.3420) +[13:59:17.725311] val: [0/2] eta: 0:00:05 loss: 0.3892 (0.3892) time: 2.9205 data: 2.8846 max mem: 9671 +[13:59:17.741687] val: [1/2] eta: 0:00:01 loss: 0.3892 (0.6480) time: 1.4682 data: 1.4423 max mem: 9671 +[13:59:17.814687] val: Total time: 0:00:03 (1.5053 s / it) +[13:59:17.823484] val loss: 0.6480240523815155 +[13:59:17.823721] Accuracy: 0.7857, F1 Score: 0.7143, ROC AUC: 0.8461, Hamming Loss: 0.2143, + Jaccard Score: 0.5750, Precision: 0.7082, Recall: 0.7219, + Average Precision: 0.8087, Kappa: 0.4290, Score: 0.6631 +[13:59:17.869149] Best epoch = 21, Best score = 0.7415 +[13:59:18.120688] log_dir: ./output_logs/retfound +[13:59:20.619704] Epoch: [32] [0/9] eta: 0:00:22 lr: 0.000264 loss: 0.3498 (0.3498) time: 2.4980 data: 2.3460 max mem: 9671 +[13:59:21.759104] Epoch: [32] [8/9] eta: 0:00:00 lr: 0.000243 loss: 0.3191 (0.3478) time: 0.4041 data: 0.2608 max mem: 9671 +[13:59:21.837055] Epoch: [32] Total time: 0:00:03 (0.4129 s / it) +[13:59:21.845640] Averaged stats: lr: 0.000243 loss: 0.3191 (0.3478) +[13:59:24.724699] val: [0/2] eta: 0:00:05 loss: 0.3005 (0.3005) time: 2.8598 data: 2.8253 max mem: 9671 +[13:59:24.741269] val: [1/2] eta: 0:00:01 loss: 0.3005 (0.8127) time: 1.4379 data: 1.4127 max mem: 9671 +[13:59:24.816720] val: Total time: 0:00:02 (1.4762 s / it) +[13:59:24.825914] val loss: 0.812717080116272 +[13:59:24.826199] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8313, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8008, Kappa: 0.5772, Score: 0.7321 +[13:59:24.872694] Best epoch = 21, Best score = 0.7415 +[13:59:25.100828] log_dir: ./output_logs/retfound +[13:59:27.656257] Epoch: [33] [0/9] eta: 0:00:22 lr: 0.000240 loss: 0.5475 (0.5475) time: 2.5545 data: 2.4070 max mem: 9671 +[13:59:28.800392] Epoch: [33] [8/9] eta: 0:00:00 lr: 0.000219 loss: 0.4062 (0.3918) time: 0.4109 data: 0.2676 max mem: 9671 +[13:59:28.876030] Epoch: [33] Total time: 0:00:03 (0.4194 s / it) +[13:59:28.884839] Averaged stats: lr: 0.000219 loss: 0.4062 (0.3918) +[13:59:31.792914] val: [0/2] eta: 0:00:05 loss: 0.3385 (0.3385) time: 2.8919 data: 2.8570 max mem: 9671 +[13:59:31.809980] val: [1/2] eta: 0:00:01 loss: 0.3385 (0.7290) time: 1.4539 data: 1.4286 max mem: 9671 +[13:59:31.889582] val: Total time: 0:00:02 (1.4947 s / it) +[13:59:31.902008] val loss: 0.7290194034576416 +[13:59:31.902216] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8359, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.8024, Kappa: 0.5772, Score: 0.7337 +[13:59:31.946691] Best epoch = 21, Best score = 0.7415 +[13:59:32.174822] log_dir: ./output_logs/retfound +[13:59:34.714819] Epoch: [34] [0/9] eta: 0:00:22 lr: 0.000217 loss: 0.4328 (0.4328) time: 2.5390 data: 2.3931 max mem: 9671 +[13:59:35.858604] Epoch: [34] [8/9] eta: 0:00:00 lr: 0.000196 loss: 0.3536 (0.3417) time: 0.4091 data: 0.2660 max mem: 9671 +[13:59:35.931279] Epoch: [34] Total time: 0:00:03 (0.4174 s / it) +[13:59:35.941014] Averaged stats: lr: 0.000196 loss: 0.3536 (0.3417) +[13:59:38.883357] val: [0/2] eta: 0:00:05 loss: 0.2532 (0.2532) time: 2.9301 data: 2.8956 max mem: 9671 +[13:59:38.899597] val: [1/2] eta: 0:00:01 loss: 0.2532 (0.9216) time: 1.4729 data: 1.4479 max mem: 9671 +[13:59:38.981666] val: Total time: 0:00:03 (1.5147 s / it) +[13:59:38.990761] val loss: 0.9215718805789948 +[13:59:38.990961] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8281, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7963, Kappa: 0.4878, Score: 0.6859 +[13:59:39.031691] Best epoch = 21, Best score = 0.7415 +[13:59:39.296894] log_dir: ./output_logs/retfound +[13:59:41.883076] Epoch: [35] [0/9] eta: 0:00:23 lr: 0.000194 loss: 0.4965 (0.4965) time: 2.5851 data: 2.4379 max mem: 9671 +[13:59:43.218900] Epoch: [35] [8/9] eta: 0:00:00 lr: 0.000174 loss: 0.3498 (0.3730) time: 0.4355 data: 0.2913 max mem: 9671 +[13:59:43.302868] Epoch: [35] Total time: 0:00:04 (0.4451 s / it) +[13:59:43.311191] Averaged stats: lr: 0.000174 loss: 0.3498 (0.3730) +[13:59:46.285769] val: [0/2] eta: 0:00:05 loss: 0.2596 (0.2596) time: 2.9620 data: 2.9450 max mem: 9671 +[13:59:46.295821] val: [1/2] eta: 0:00:01 loss: 0.2596 (0.9987) time: 1.4857 data: 1.4726 max mem: 9671 +[13:59:46.372587] val: Total time: 0:00:03 (1.5248 s / it) +[13:59:46.384374] val loss: 0.9986901581287384 +[13:59:46.384613] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7982, Kappa: 0.4878, Score: 0.6870 +[13:59:46.425983] Best epoch = 21, Best score = 0.7415 +[13:59:46.669067] log_dir: ./output_logs/retfound +[13:59:49.104666] Epoch: [36] [0/9] eta: 0:00:21 lr: 0.000171 loss: 0.4374 (0.4374) time: 2.4345 data: 2.2896 max mem: 9671 +[13:59:50.247080] Epoch: [36] [8/9] eta: 0:00:00 lr: 0.000152 loss: 0.3730 (0.3513) time: 0.3973 data: 0.2545 max mem: 9671 +[13:59:50.326246] Epoch: [36] Total time: 0:00:03 (0.4063 s / it) +[13:59:50.335682] Averaged stats: lr: 0.000152 loss: 0.3730 (0.3513) +[13:59:53.317117] val: [0/2] eta: 0:00:05 loss: 0.2994 (0.2994) time: 2.9638 data: 2.9280 max mem: 9671 +[13:59:53.333518] val: [1/2] eta: 0:00:01 loss: 0.2994 (0.8882) time: 1.4898 data: 1.4641 max mem: 9671 +[13:59:53.408640] val: Total time: 0:00:03 (1.5280 s / it) +[13:59:53.417651] val loss: 0.8881509304046631 +[13:59:53.417848] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8250, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7857, Recall: 0.7188, + Average Precision: 0.7946, Kappa: 0.4878, Score: 0.6849 +[13:59:53.458968] Best epoch = 21, Best score = 0.7415 +[13:59:53.711333] log_dir: ./output_logs/retfound +[13:59:56.682602] Epoch: [37] [0/9] eta: 0:00:26 lr: 0.000150 loss: 0.3205 (0.3205) time: 2.9702 data: 2.8266 max mem: 9671 +[13:59:57.828941] Epoch: [37] [8/9] eta: 0:00:00 lr: 0.000132 loss: 0.3006 (0.3314) time: 0.4573 data: 0.3142 max mem: 9671 +[13:59:57.904776] Epoch: [37] Total time: 0:00:04 (0.4659 s / it) +[13:59:57.913923] Averaged stats: lr: 0.000132 loss: 0.3006 (0.3314) +[14:00:00.714969] val: [0/2] eta: 0:00:05 loss: 0.3475 (0.3475) time: 2.7883 data: 2.7642 max mem: 9671 +[14:00:00.731268] val: [1/2] eta: 0:00:01 loss: 0.3475 (0.7914) time: 1.4020 data: 1.3821 max mem: 9671 +[14:00:00.817320] val: Total time: 0:00:02 (1.4457 s / it) +[14:00:00.826098] val loss: 0.7914236485958099 +[14:00:00.826280] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8289, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7930, Kappa: 0.5772, Score: 0.7313 +[14:00:00.866591] Best epoch = 21, Best score = 0.7415 +[14:00:01.134921] log_dir: ./output_logs/retfound +[14:00:03.673848] Epoch: [38] [0/9] eta: 0:00:22 lr: 0.000130 loss: 0.3957 (0.3957) time: 2.5380 data: 2.3918 max mem: 9671 +[14:00:04.880790] Epoch: [38] [8/9] eta: 0:00:00 lr: 0.000112 loss: 0.3663 (0.3488) time: 0.4160 data: 0.2725 max mem: 9671 +[14:00:04.954368] Epoch: [38] Total time: 0:00:03 (0.4244 s / it) +[14:00:04.963204] Averaged stats: lr: 0.000112 loss: 0.3663 (0.3488) +[14:00:07.958343] val: [0/2] eta: 0:00:05 loss: 0.3682 (0.3682) time: 2.9803 data: 2.9521 max mem: 9671 +[14:00:07.976891] val: [1/2] eta: 0:00:01 loss: 0.3682 (0.7530) time: 1.4991 data: 1.4761 max mem: 9671 +[14:00:08.058040] val: Total time: 0:00:03 (1.5404 s / it) +[14:00:08.069990] val loss: 0.7529699206352234 +[14:00:08.070233] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8344, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7963, Kappa: 0.5772, Score: 0.7331 +[14:00:08.112158] Best epoch = 21, Best score = 0.7415 +[14:00:08.380612] log_dir: ./output_logs/retfound +[14:00:11.117604] Epoch: [39] [0/9] eta: 0:00:24 lr: 0.000110 loss: 0.2798 (0.2798) time: 2.7354 data: 2.5875 max mem: 9671 +[14:00:12.262705] Epoch: [39] [8/9] eta: 0:00:00 lr: 0.000094 loss: 0.2936 (0.3222) time: 0.4311 data: 0.2876 max mem: 9671 +[14:00:12.333021] Epoch: [39] Total time: 0:00:03 (0.4391 s / it) +[14:00:12.340942] Averaged stats: lr: 0.000094 loss: 0.2936 (0.3222) +[14:00:15.123378] val: [0/2] eta: 0:00:05 loss: 0.3959 (0.3959) time: 2.7665 data: 2.7304 max mem: 9671 +[14:00:15.142380] val: [1/2] eta: 0:00:01 loss: 0.3959 (0.7333) time: 1.3924 data: 1.3652 max mem: 9671 +[14:00:15.215378] val: Total time: 0:00:02 (1.4297 s / it) +[14:00:15.226444] val loss: 0.7332981824874878 +[14:00:15.226730] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8336, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.7957, Kappa: 0.5243, Score: 0.7066 +[14:00:15.270808] Best epoch = 21, Best score = 0.7415 +[14:00:15.530227] log_dir: ./output_logs/retfound +[14:00:18.168880] Epoch: [40] [0/9] eta: 0:00:23 lr: 0.000092 loss: 0.2801 (0.2801) time: 2.6370 data: 2.4912 max mem: 9671 +[14:00:19.315479] Epoch: [40] [8/9] eta: 0:00:00 lr: 0.000078 loss: 0.3197 (0.3514) time: 0.4203 data: 0.2769 max mem: 9671 +[14:00:19.397593] Epoch: [40] Total time: 0:00:03 (0.4297 s / it) +[14:00:19.406129] Averaged stats: lr: 0.000078 loss: 0.3197 (0.3514) +[14:00:22.320793] val: [0/2] eta: 0:00:05 loss: 0.4082 (0.4082) time: 2.8970 data: 2.8611 max mem: 9671 +[14:00:22.337327] val: [1/2] eta: 0:00:01 loss: 0.4082 (0.7276) time: 1.4565 data: 1.4306 max mem: 9671 +[14:00:22.409201] val: Total time: 0:00:02 (1.4931 s / it) +[14:00:22.418274] val loss: 0.7276108860969543 +[14:00:22.418464] Accuracy: 0.8333, F1 Score: 0.7619, ROC AUC: 0.8313, Hamming Loss: 0.1667, + Jaccard Score: 0.6335, Precision: 0.7727, Recall: 0.7531, + Average Precision: 0.7946, Kappa: 0.5243, Score: 0.7058 +[14:00:22.466218] Best epoch = 21, Best score = 0.7415 +[14:00:22.714090] log_dir: ./output_logs/retfound +[14:00:25.129674] Epoch: [41] [0/9] eta: 0:00:21 lr: 0.000076 loss: 0.4471 (0.4471) time: 2.4143 data: 2.2654 max mem: 9671 +[14:00:26.275167] Epoch: [41] [8/9] eta: 0:00:00 lr: 0.000062 loss: 0.3231 (0.3198) time: 0.3954 data: 0.2518 max mem: 9671 +[14:00:26.351811] Epoch: [41] Total time: 0:00:03 (0.4042 s / it) +[14:00:26.353629] Averaged stats: lr: 0.000062 loss: 0.3231 (0.3198) +[14:00:29.331471] val: [0/2] eta: 0:00:05 loss: 0.4002 (0.4002) time: 2.9612 data: 2.9271 max mem: 9671 +[14:00:29.350308] val: [1/2] eta: 0:00:01 loss: 0.4002 (0.7322) time: 1.4897 data: 1.4636 max mem: 9671 +[14:00:29.426257] val: Total time: 0:00:03 (1.5284 s / it) +[14:00:29.435263] val loss: 0.7322454154491425 +[14:00:29.435594] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8313, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7946, Kappa: 0.5772, Score: 0.7321 +[14:00:29.490297] Best epoch = 21, Best score = 0.7415 +[14:00:29.791148] log_dir: ./output_logs/retfound +[14:00:32.398071] Epoch: [42] [0/9] eta: 0:00:23 lr: 0.000061 loss: 0.2169 (0.2169) time: 2.6059 data: 2.4601 max mem: 9671 +[14:00:33.540062] Epoch: [42] [8/9] eta: 0:00:00 lr: 0.000048 loss: 0.2901 (0.3057) time: 0.4163 data: 0.2734 max mem: 9671 +[14:00:33.614982] Epoch: [42] Total time: 0:00:03 (0.4249 s / it) +[14:00:33.623546] Averaged stats: lr: 0.000048 loss: 0.2901 (0.3057) +[14:00:36.459137] val: [0/2] eta: 0:00:05 loss: 0.3896 (0.3896) time: 2.8187 data: 2.7847 max mem: 9671 +[14:00:36.475402] val: [1/2] eta: 0:00:01 loss: 0.3896 (0.7544) time: 1.4172 data: 1.3924 max mem: 9671 +[14:00:36.549164] val: Total time: 0:00:02 (1.4548 s / it) +[14:00:36.558338] val loss: 0.7544125616550446 +[14:00:36.558621] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:00:36.604471] Best epoch = 21, Best score = 0.7415 +[14:00:36.860913] log_dir: ./output_logs/retfound +[14:00:39.324364] Epoch: [43] [0/9] eta: 0:00:22 lr: 0.000047 loss: 0.2695 (0.2695) time: 2.4624 data: 2.3150 max mem: 9671 +[14:00:40.466225] Epoch: [43] [8/9] eta: 0:00:00 lr: 0.000036 loss: 0.3200 (0.3327) time: 0.4004 data: 0.2573 max mem: 9671 +[14:00:40.540759] Epoch: [43] Total time: 0:00:03 (0.4089 s / it) +[14:00:40.550402] Averaged stats: lr: 0.000036 loss: 0.3200 (0.3327) +[14:00:43.516079] val: [0/2] eta: 0:00:05 loss: 0.3908 (0.3908) time: 2.9527 data: 2.9163 max mem: 9671 +[14:00:43.532667] val: [1/2] eta: 0:00:01 loss: 0.3908 (0.7751) time: 1.4843 data: 1.4582 max mem: 9671 +[14:00:43.606683] val: Total time: 0:00:03 (1.5220 s / it) +[14:00:43.615660] val loss: 0.7750901579856873 +[14:00:43.615877] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:00:43.666142] Best epoch = 21, Best score = 0.7415 +[14:00:43.920698] log_dir: ./output_logs/retfound +[14:00:46.519695] Epoch: [44] [0/9] eta: 0:00:23 lr: 0.000035 loss: 0.3314 (0.3314) time: 2.5979 data: 2.4510 max mem: 9671 +[14:00:47.263723] Epoch: [44] [8/9] eta: 0:00:00 lr: 0.000026 loss: 0.3314 (0.3466) time: 0.3712 data: 0.2724 max mem: 9671 +[14:00:47.333573] Epoch: [44] Total time: 0:00:03 (0.3792 s / it) +[14:00:47.341987] Averaged stats: lr: 0.000026 loss: 0.3314 (0.3466) +[14:00:50.276033] val: [0/2] eta: 0:00:05 loss: 0.4027 (0.4027) time: 2.9185 data: 2.8822 max mem: 9671 +[14:00:50.294754] val: [1/2] eta: 0:00:01 loss: 0.4027 (0.7743) time: 1.4682 data: 1.4412 max mem: 9671 +[14:00:50.366882] val: Total time: 0:00:03 (1.5052 s / it) +[14:00:50.375812] val loss: 0.7742511332035065 +[14:00:50.376009] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:00:50.419862] Best epoch = 21, Best score = 0.7415 +[14:00:50.685421] log_dir: ./output_logs/retfound +[14:00:53.206130] Epoch: [45] [0/9] eta: 0:00:22 lr: 0.000025 loss: 0.3766 (0.3766) time: 2.5195 data: 2.3727 max mem: 9671 +[14:00:54.350632] Epoch: [45] [8/9] eta: 0:00:00 lr: 0.000017 loss: 0.3525 (0.3718) time: 0.4070 data: 0.2637 max mem: 9671 +[14:00:54.421065] Epoch: [45] Total time: 0:00:03 (0.4151 s / it) +[14:00:54.424963] Averaged stats: lr: 0.000017 loss: 0.3525 (0.3718) +[14:00:57.436702] val: [0/2] eta: 0:00:05 loss: 0.4050 (0.4050) time: 2.9933 data: 2.9587 max mem: 9671 +[14:00:57.452941] val: [1/2] eta: 0:00:01 loss: 0.4050 (0.7764) time: 1.5045 data: 1.4794 max mem: 9671 +[14:00:57.521231] val: Total time: 0:00:03 (1.5393 s / it) +[14:00:57.530017] val loss: 0.7764365673065186 +[14:00:57.530216] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7930, Kappa: 0.5772, Score: 0.7311 +[14:00:57.570015] Best epoch = 21, Best score = 0.7415 +[14:00:57.837358] log_dir: ./output_logs/retfound +[14:01:00.401349] Epoch: [46] [0/9] eta: 0:00:23 lr: 0.000016 loss: 0.3260 (0.3260) time: 2.5630 data: 2.4155 max mem: 9671 +[14:01:01.545660] Epoch: [46] [8/9] eta: 0:00:00 lr: 0.000010 loss: 0.3260 (0.3344) time: 0.4118 data: 0.2685 max mem: 9671 +[14:01:01.616241] Epoch: [46] Total time: 0:00:03 (0.4199 s / it) +[14:01:01.625729] Averaged stats: lr: 0.000010 loss: 0.3260 (0.3344) +[14:01:04.385939] val: [0/2] eta: 0:00:05 loss: 0.4028 (0.4028) time: 2.7439 data: 2.7091 max mem: 9671 +[14:01:04.402428] val: [1/2] eta: 0:00:01 loss: 0.4028 (0.7721) time: 1.3799 data: 1.3546 max mem: 9671 +[14:01:04.483639] val: Total time: 0:00:02 (1.4211 s / it) +[14:01:04.492882] val loss: 0.772107869386673 +[14:01:04.493080] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:01:04.532256] Best epoch = 21, Best score = 0.7415 +[14:01:04.797862] log_dir: ./output_logs/retfound +[14:01:07.296607] Epoch: [47] [0/9] eta: 0:00:22 lr: 0.000010 loss: 0.3762 (0.3762) time: 2.4977 data: 2.3523 max mem: 9671 +[14:01:08.439141] Epoch: [47] [8/9] eta: 0:00:00 lr: 0.000005 loss: 0.3166 (0.3650) time: 0.4044 data: 0.2614 max mem: 9671 +[14:01:08.513734] Epoch: [47] Total time: 0:00:03 (0.4129 s / it) +[14:01:08.521525] Averaged stats: lr: 0.000005 loss: 0.3166 (0.3650) +[14:01:11.442986] val: [0/2] eta: 0:00:05 loss: 0.4020 (0.4020) time: 2.8991 data: 2.8638 max mem: 9671 +[14:01:11.459465] val: [1/2] eta: 0:00:01 loss: 0.4020 (0.7689) time: 1.4575 data: 1.4319 max mem: 9671 +[14:01:11.533193] val: Total time: 0:00:02 (1.4950 s / it) +[14:01:11.542841] val loss: 0.7689108848571777 +[14:01:11.543093] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:01:11.587198] Best epoch = 21, Best score = 0.7415 +[14:01:11.842466] log_dir: ./output_logs/retfound +[14:01:14.428434] Epoch: [48] [0/9] eta: 0:00:23 lr: 0.000005 loss: 0.3463 (0.3463) time: 2.5849 data: 2.4371 max mem: 9671 +[14:01:15.574029] Epoch: [48] [8/9] eta: 0:00:00 lr: 0.000002 loss: 0.3568 (0.3456) time: 0.4144 data: 0.2709 max mem: 9671 +[14:01:15.659614] Epoch: [48] Total time: 0:00:03 (0.4241 s / it) +[14:01:15.668721] Averaged stats: lr: 0.000002 loss: 0.3568 (0.3456) +[14:01:18.463637] val: [0/2] eta: 0:00:05 loss: 0.3997 (0.3997) time: 2.7771 data: 2.7428 max mem: 9671 +[14:01:18.482294] val: [1/2] eta: 0:00:01 loss: 0.3997 (0.7697) time: 1.3975 data: 1.3715 max mem: 9671 +[14:01:18.553495] val: Total time: 0:00:02 (1.4339 s / it) +[14:01:18.562380] val loss: 0.769738107919693 +[14:01:18.562627] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:01:18.608173] Best epoch = 21, Best score = 0.7415 +[14:01:18.840970] log_dir: ./output_logs/retfound +[14:01:21.431223] Epoch: [49] [0/9] eta: 0:00:23 lr: 0.000002 loss: 0.3057 (0.3057) time: 2.5890 data: 2.4441 max mem: 9671 +[14:01:22.579700] Epoch: [49] [8/9] eta: 0:00:00 lr: 0.000001 loss: 0.3238 (0.3242) time: 0.4152 data: 0.2717 max mem: 9671 +[14:01:22.656234] Epoch: [49] Total time: 0:00:03 (0.4239 s / it) +[14:01:22.664951] Averaged stats: lr: 0.000001 loss: 0.3238 (0.3242) +[14:01:25.611253] val: [0/2] eta: 0:00:05 loss: 0.3984 (0.3984) time: 2.9235 data: 2.8870 max mem: 9671 +[14:01:25.627543] val: [1/2] eta: 0:00:01 loss: 0.3984 (0.7703) time: 1.4696 data: 1.4436 max mem: 9671 +[14:01:25.700416] val: Total time: 0:00:03 (1.5067 s / it) +[14:01:25.709160] val loss: 0.7703390717506409 +[14:01:25.709397] Accuracy: 0.8571, F1 Score: 0.7879, ROC AUC: 0.8281, Hamming Loss: 0.1429, + Jaccard Score: 0.6667, Precision: 0.8162, Recall: 0.7688, + Average Precision: 0.7929, Kappa: 0.5772, Score: 0.7311 +[14:01:25.754380] Best epoch = 21, Best score = 0.7415 +[14:01:30.010906] Test with the best model, epoch = 21: +[14:01:32.754312] test: [0/3] eta: 0:00:08 loss: 0.3573 (0.3573) time: 2.7323 data: 2.6972 max mem: 9671 +[14:01:32.929988] test: [2/3] eta: 0:00:00 loss: 0.3573 (0.4751) time: 0.9691 data: 0.8992 max mem: 9671 +[14:01:33.008842] test: Total time: 0:00:02 (0.9959 s / it) +[14:01:33.020336] val loss: 0.47513073682785034 +[14:01:33.020470] Accuracy: 0.8333, F1 Score: 0.7419, ROC AUC: 0.8371, Hamming Loss: 0.1667, + Jaccard Score: 0.6137, Precision: 0.7323, Recall: 0.7537, + Average Precision: 0.7821, Kappa: 0.4842, Score: 0.6877 +[14:01:33.783024] Training time 0:06:16 +[rank0]:[W615 14:01:34.199065011 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator()) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/papila/retfound acc=0.8333 auroc=0.8373161764705882 f1_macro=0.7419 qwk=0.4842105263157894 diff --git a/results/papila/vit/confusion_matrix.png b/results/papila/vit/confusion_matrix.png new file mode 100644 index 0000000000000000000000000000000000000000..f40a1551c4b96a2f0b620479cc865f469b1bd4e1 --- /dev/null +++ b/results/papila/vit/confusion_matrix.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27f4ffb4f4a539debbfc987395c1999c389c1adbc49f6f2fc9065c069227fe56 +size 73078 diff --git a/results/papila/vit/log.csv b/results/papila/vit/log.csv new file mode 100644 index 0000000000000000000000000000000000000000..ce68bd8cef0bffa9e5989839d31502d27b226910 --- /dev/null +++ b/results/papila/vit/log.csv @@ -0,0 +1,30 @@ +epoch,train_loss,val_acc,val_auc,val_score,lr +0,0.8081949055194855,0.38095238095238093,0.553125,0.32084904394735353,5.545808836528073e-08 +1,0.7463895827531815,0.6904761904761905,0.69375,0.5318601694915255,1.29402206185655e-07 +2,0.684089183807373,0.2857142857142857,0.78125,0.35694758672699844,2.0334632400602932e-07 +3,0.6925588548183441,0.7380952380952381,0.721875,0.5793965653156564,2.7729044182640365e-07 +4,0.646022841334343,0.5238095238095238,0.7124999999999999,0.4441021345246959,3.512345596467779e-07 +5,0.6103057861328125,0.5714285714285714,0.753125,0.5036502544055598,3.694672444929302e-07 +6,0.5391916632652283,0.5476190476190477,0.765625,0.49254615505996213,3.683426684987483e-07 +7,0.5554562211036682,0.5952380952380952,0.784375,0.5296914201409342,3.663241848222764e-07 +8,0.5353901088237762,0.6428571428571429,0.78125,0.5261474644139035,3.6342162731308893e-07 +9,0.5195781961083412,0.6904761904761905,0.784375,0.5620685028248588,3.5964913693960667e-07 +10,0.5137002021074295,0.6428571428571429,0.8125,0.5859409589478831,3.550250928957199e-07 +11,0.5435433834791183,0.6904761904761905,0.80625,0.6041263127115114,3.495720230592029e-07 +12,0.5050823017954826,0.6666666666666666,0.815625,0.5331997863247863,3.4331649423815874e-07 +13,0.5033866092562675,0.7380952380952381,0.859375,0.6728023928071764,3.362889827402e-07 +14,0.4588450863957405,0.7619047619047619,0.88125,0.6862313201211271,3.285237258949416e-07 +15,0.44873566925525665,0.7380952380952381,0.859375,0.6433848870056497,3.2005855525317277e-07 +16,0.42817943543195724,0.7380952380952381,0.859375,0.6433848870056497,3.1093471227534435e-07 +17,0.46573343873023987,0.8095238095238095,0.865625,0.7235844017094016,3.0119664740731875e-07 +18,0.4151291847229004,0.8095238095238095,0.88125,0.728792735042735,2.908918035222662e-07 +19,0.4494623392820358,0.7857142857142857,0.86875,0.6879721898037188,2.800703847837553e-07 +20,0.4139930456876755,0.7619047619047619,0.840625,0.6573450854700854,2.687851120561149e-07 +21,0.40749695897102356,0.7380952380952381,0.815625,0.6106465653156564,2.570909660536824e-07 +22,0.40808261930942535,0.7619047619047619,0.821875,0.6664396534544604,2.450449194802903e-07 +23,0.4032032936811447,0.7380952380952381,0.815625,0.6106465653156564,2.3270565946397963e-07 +24,0.40627045184373856,0.7380952380952381,0.815625,0.6106465653156564,2.2013330163921197e-07 +25,0.3788785859942436,0.7380952380952381,0.790625,0.6023132319823231,2.0738909726954043e-07 +26,0.4226858466863632,0.7619047619047619,0.7718750000000001,0.5712138866291551,1.9453513483761127e-07 +27,0.3840523436665535,0.7142857142857143,0.78125,0.558302142820394,1.8163403755631687e-07 +28,0.3793400377035141,0.7380952380952381,0.78125,0.5531960385467029,1.6874865827479806e-07 diff --git a/results/papila/vit/metrics.json b/results/papila/vit/metrics.json new file mode 100644 index 0000000000000000000000000000000000000000..0cc8ef3739d1d1749dce7989beef7cd71e6acc91 --- /dev/null +++ b/results/papila/vit/metrics.json @@ -0,0 +1,49 @@ +{ + "n_test": 84, + "n_classes": 2, + "task": "binary", + "accuracy": 0.75, + "balanced_accuracy": 0.7022058823529411, + "precision_macro": 0.6491525423728814, + "recall_macro": 0.7022058823529411, + "f1_macro": 0.6612252736700595, + "precision_weighted": 0.8033898305084746, + "recall_weighted": 0.75, + "f1_weighted": 0.7685807566737085, + "cohen_kappa": 0.33282904689863835, + "quadratic_weighted_kappa": 0.33282904689863835, + "mcc": 0.3473299378728699, + "auroc": 0.7849264705882353, + "auprc": 0.5894183392973115, + "sensitivity": 0.625, + "specificity": 0.7794117647058824, + "precision_pos": 0.4, + "f1_pos": 0.4878048780487805, + "per_class": { + "0": { + "precision": 0.8983050847457628, + "recall": 0.7794117647058824, + "f1-score": 0.8346456692913385, + "support": 68.0 + }, + "1": { + "precision": 0.4, + "recall": 0.625, + "f1-score": 0.4878048780487805, + "support": 16.0 + }, + "accuracy": 0.75, + "macro avg": { + "precision": 0.6491525423728814, + "recall": 0.7022058823529411, + "f1-score": 0.6612252736700595, + "support": 84.0 + }, + "weighted avg": { + "precision": 0.8033898305084746, + "recall": 0.75, + "f1-score": 0.7685807566737085, + "support": 84.0 + } + } +} \ No newline at end of file diff --git a/results/papila/vit/pr.png b/results/papila/vit/pr.png new file mode 100644 index 0000000000000000000000000000000000000000..a103b55caac4fc2eb8c4d001a8654bda6534474b --- /dev/null +++ b/results/papila/vit/pr.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6fe3680bdd3b1a11889ce1112c7916bb49503f03877f55757061e7ece3d1a591 +size 48629 diff --git a/results/papila/vit/roc.png b/results/papila/vit/roc.png new file mode 100644 index 0000000000000000000000000000000000000000..afae9df8b9cddf511ed3a568506a50daa8e8e961 --- /dev/null +++ b/results/papila/vit/roc.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b8bee8ac5f985f7ba6a106dbbc229eaf5b1647dfc2870d38fa996b35b97178a +size 57359 diff --git a/results/papila/vit/test_pred.npz b/results/papila/vit/test_pred.npz new file mode 100644 index 0000000000000000000000000000000000000000..959718674514d2e52f123d6e6b6a8afd86d1df1e --- /dev/null +++ b/results/papila/vit/test_pred.npz @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:793dde3491b2fe8d25ef8123d6a8ae58d8dfb2a380dfce74b22024499db7532c +size 1854 diff --git a/results/papila/vit/train.log b/results/papila/vit/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e74051694462daff987ba44e128c91c422869d02 --- /dev/null +++ b/results/papila/vit/train.log @@ -0,0 +1,154 @@ +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:154: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead. + scaler = torch.cuda.amp.GradScaler() +[vit] train=294 val=42 test=84 classes=['0', '1'] +[vit] optim groups=28 layer_decay=0.65 drop_path=0.1 ls=0.1 lr=0.0001 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep0 loss=0.8082 val_acc=0.3810 val_auc=0.5531 score=0.3208 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep1 loss=0.7464 val_acc=0.6905 val_auc=0.6937 score=0.5319 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep2 loss=0.6841 val_acc=0.2857 val_auc=0.7812 score=0.3569 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep3 loss=0.6926 val_acc=0.7381 val_auc=0.7219 score=0.5794 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep4 loss=0.6460 val_acc=0.5238 val_auc=0.7125 score=0.4441 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep5 loss=0.6103 val_acc=0.5714 val_auc=0.7531 score=0.5037 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep6 loss=0.5392 val_acc=0.5476 val_auc=0.7656 score=0.4925 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep7 loss=0.5555 val_acc=0.5952 val_auc=0.7844 score=0.5297 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep8 loss=0.5354 val_acc=0.6429 val_auc=0.7812 score=0.5261 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep9 loss=0.5196 val_acc=0.6905 val_auc=0.7844 score=0.5621 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep10 loss=0.5137 val_acc=0.6429 val_auc=0.8125 score=0.5859 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep11 loss=0.5435 val_acc=0.6905 val_auc=0.8063 score=0.6041 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep12 loss=0.5051 val_acc=0.6667 val_auc=0.8156 score=0.5332 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep13 loss=0.5034 val_acc=0.7381 val_auc=0.8594 score=0.6728 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep14 loss=0.4588 val_acc=0.7619 val_auc=0.8812 score=0.6862 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep15 loss=0.4487 val_acc=0.7381 val_auc=0.8594 score=0.6434 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep16 loss=0.4282 val_acc=0.7381 val_auc=0.8594 score=0.6434 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep17 loss=0.4657 val_acc=0.8095 val_auc=0.8656 score=0.7236 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep18 loss=0.4151 val_acc=0.8095 val_auc=0.8812 score=0.7288 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep19 loss=0.4495 val_acc=0.7857 val_auc=0.8688 score=0.6880 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep20 loss=0.4140 val_acc=0.7619 val_auc=0.8406 score=0.6573 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep21 loss=0.4075 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep22 loss=0.4081 val_acc=0.7619 val_auc=0.8219 score=0.6664 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep23 loss=0.4032 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep24 loss=0.4063 val_acc=0.7381 val_auc=0.8156 score=0.6106 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep25 loss=0.3789 val_acc=0.7381 val_auc=0.7906 score=0.6023 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep26 loss=0.4227 val_acc=0.7619 val_auc=0.7719 score=0.5712 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep27 loss=0.3841 val_acc=0.7143 val_auc=0.7812 score=0.5583 +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:176: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] ep28 loss=0.3793 val_acc=0.7381 val_auc=0.7812 score=0.5532 +[vit] early stop at ep28 (best ep18 score=0.7288) +/mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/Code/train_cnn_vit.py:58: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead. + with torch.cuda.amp.autocast(): +[vit] DONE best_ep=18 best_val_score=0.7288 -> saved test_pred.npz (84 samples) +[evaluate] /mnt/tidal-alsh-share2/dataset/qinshengqian/research/c3/GPT-Image/results/papila/vit acc=0.7500 auroc=0.7849264705882353 f1_macro=0.6612 qwk=0.33282904689863835 diff --git a/results/report.html b/results/report.html new file mode 100644 index 0000000000000000000000000000000000000000..2528d67306792482716774fa3191a9dadbd2357c --- /dev/null +++ b/results/report.html @@ -0,0 +1,312 @@ + + +眼底图像分类 Benchmark + +
+

眼底图像分类 Benchmark · RetFound vs ResNet vs ViT

+

7 个数据集 · 4 个疾病方向 · 三模型（RetFound ViT-L / ResNet-50 / ViT-B/16，均预训练后全参数微调）

+

每个数据集含：采集背景（FOV/设备/来源/分辨率）· 类别分布（按 split）· 模型性能（指标表 + 柱状图）· 混淆矩阵/ROC

+

评估协议：输入 224 · 官方划分优先（否则 7:1:2 分层）· val 选最优→测 test · 指标由统一脚本计算

+
+
+ +
+

近视 · Myopic Maculopathy

+ +
+

MMAC 2023

+ 5-class grade 0–4 · 划分 train/val/test = 973/139/279 · 总计 1391 张
+
📷 采集背景：彩色眼底照（非散瞳）｜FOV：45°（设备标称，论文正文未印）｜设备：Topcon TRC-NW400（单一设备）｜来源：上海健康医学中心 + 上海市第六人民医院（中国，均为中国人群）｜分辨率：未公开｜标注：META-PM 5 级，双医师分级（κ=0.91），单设备单人群为其局限。
+

类别分布 · Class distribution（按 split）

+
+
Split0·grade_01·grade_12·grade_23·grade_34·grade_4合计
train3393432025039973
val48492976139
test9798581511279
合计48449028972561391
+
+
+

模型性能 · Performance

+
+
ModelAccuracyBal-Accmacro-AUROCQWKF1-macroPrec-macroRec-macroKappa
RetFound (ViT-L, CFP)0.85660.74230.96730.92640.76020.80110.74230.7967
ResNet-500.82440.72590.94450.87890.73600.74840.72590.7510
ViT-B/160.82440.73450.95120.90520.73900.75220.73450.7526
+
+
+

每类指标 · Per-class metrics

+
+
ClassSupportRetFound (ViT-L, CFP)ResNet-50ViT-B/16
RecallF1AUROCRecallF1AUROCRecallF1AUROC
0·grade_0970.9180.9320.9900.8660.8890.9810.9280.9330.991
1·grade_1980.8780.8600.9610.8570.8280.9480.8060.8190.933
2·grade_2580.8620.8400.9700.8280.8210.9590.7930.7670.945
3·grade_3150.6000.5810.9220.5330.5710.8640.6000.5450.932
4·grade_4110.4550.5880.9920.5450.5710.9720.5450.6320.955
+
+
+ +
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+
+

AMD · Age-related Macular Degeneration

+ +
+

ADAM

+ binary AMD / Non-AMD · 划分 train/val/test = 280/40/80 · 总计 400 张
+
📷 采集背景：彩色眼底照｜FOV：未标注（仅说明取景中心为视盘 / 黄斑 / 两者中点）｜设备：Zeiss Visucam 500（2124×2056，824 张）+ Canon CR-2（1444×1444，376 张）｜来源：中山眼科中心（中国·广州）｜Training400：89 AMD / 311 非 AMD（AMD 被刻意过采样，非真实患病率）。
+

类别分布 · Class distribution（按 split）

+
+
Split0·Non-AMD1·AMD合计
train21862280
val31940
test621880
合计31189400
+
+
+

模型性能 · Performance

+
+
ModelAccuracyAUROCAUPRCF1SensitivitySpecificityKappaMCC
RetFound (ViT-L, CFP)0.92500.95160.92140.89250.83330.95160.78490.7849
ResNet-500.82500.91940.81600.76670.72220.85480.53490.5397
ViT-B/160.91250.93190.88540.87700.83330.93550.75400.7544
+
+
+ +

数据稀缺性分析 · Data-scarcity experiment

+
训练数据按类别分层抽样至 100/50/25/10/5%，保持 val/test 完整。PAPILA 随数据量下降最快，最适合作合成数据增广实验。
+ +
+

100% · 280 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.92500.95160.92140.89250.83330.9516
resnet0.73750.77960.59670.68370.72220.7419
vit0.91250.93190.88540.87700.83330.9355
+
+
+
+
+

50% · 140 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.92500.94440.92350.89250.83330.9516
resnet0.87500.92470.83860.79490.55560.9677
vit0.86250.90050.80940.80660.72220.9032
+
+
+
+
+

25% · 70 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.87500.92110.86940.83330.83330.8871
resnet0.82500.84950.66010.74910.61110.8871
vit0.81250.83510.65210.75400.72220.8387
+
+
+
+
+

10% · 28 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.77500.84090.62660.43660.00001.0000
resnet0.77500.77060.55180.68940.55560.8387
vit0.80000.82210.66030.70120.50000.8871
+
+
+
+
+

5% · 14 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.77500.80060.56570.43660.00001.0000
resnet0.82500.78410.56180.67850.33330.9677
vit0.71250.76610.43600.63430.55560.7581
+
+
+
+
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+
+

青光眼 · Glaucoma

+ +
+

AIROGS (EyePACS-AIROGS-light)

+ binary RG / NRG · 划分 train/val/test = 5000/540/1000 · 总计 6540 张
+
📷 采集背景：彩色眼底照，源自 EyePACS 远程筛查平台（美国约 500 个点、60071 人、多种族）｜设备：多相机混用（Optovue iCam100≈26%、Topcon NW200/400≈20%、Canon CR1/CR2/DGI、Centervue、Nidek、Crystalvue，约 21% 未知）｜FOV / 分辨率：因多设备未统一｜原为糖网筛查图后重标青光眼；全集 RG 仅约 3%（极不平衡），本「light」子集已平衡为 3270/3270。
+

类别分布 · Class distribution（按 split）

+
+
Split0·NRG1·RG合计
train250025005000
val270270540
test5005001000
合计327032706540
+
+
+

模型性能 · Performance

+
+
ModelAccuracyAUROCAUPRCF1SensitivitySpecificityKappaMCC
RetFound (ViT-L, CFP)0.90800.97080.97150.90800.89200.92400.81600.8164
ResNet-500.90000.96140.95960.90000.90800.89200.80000.8001
ViT-B/160.90000.96000.96250.90000.90600.89400.80000.8001
+
+
+ +

数据稀缺性分析 · Data-scarcity experiment

+
训练数据按类别分层抽样至 100/50/25/10/5%，保持 val/test 完整。PAPILA 随数据量下降最快，最适合作合成数据增广实验。
+ +
+

100% · 5000 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.90800.97080.97150.90800.89200.9240
resnet0.89800.96400.96320.89800.89400.9020
vit0.87300.94520.94140.87300.88400.8620
+
+
+
+
+

50% · 2500 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.89800.96450.96610.89800.89600.9000
resnet0.87900.94430.93920.87900.87800.8800
vit0.86300.93910.93950.86300.86200.8640
+
+
+
+
+

25% · 1250 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.88100.95430.95710.88100.87200.8900
resnet0.85500.93910.93800.85500.85200.8580
vit0.83200.91500.91930.83170.87400.7900
+
+
+
+
+

10% · 500 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.85100.93290.93780.85100.84000.8620
resnet0.81900.89150.87840.81880.78600.8520
vit0.80000.89020.89070.80000.80000.8000
+
+
+
+
+

5% · 250 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.83400.90930.92120.83400.85000.8180
resnet0.77400.85520.85520.77380.80000.7480
vit0.76600.84690.84820.76600.78000.7520
+
+
+
+
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+

PAPILA

+ binary glaucoma / healthy · 划分 train/val/test = 294/42/84 · 总计 420 张
+
📷 采集背景：彩色眼底照，以视盘为中心｜FOV：30°｜设备：Topcon TRC-NW400（非散瞳）｜分辨率：2576×1934 JPEG｜来源：Reina Sofía 大学医院（西班牙·Murcia，2018–2020）｜244 人双眼共 488 张（healthy/glaucoma/suspect，本项目已剔除 suspect → 420）｜附临床数据与视盘/视杯分割。
+

类别分布 · Class distribution（按 split）

+
+
Split0·healthy1·glaucoma合计
train23361294
val321042
test681684
合计33387420
+
+
+

模型性能 · Performance

+
+
ModelAccuracyAUROCAUPRCF1SensitivitySpecificityKappaMCC
RetFound (ViT-L, CFP)0.83330.83730.63350.74190.62500.88240.48420.4855
ResNet-500.88100.79410.70480.77250.50000.97060.54940.5706
ViT-B/160.75000.78490.58940.66120.62500.77940.33280.3473
+
+
+ +

数据稀缺性分析 · Data-scarcity experiment

+
训练数据按类别分层抽样至 100/50/25/10/5%，保持 val/test 完整。PAPILA 随数据量下降最快，最适合作合成数据增广实验。
+ +
+

100% · 294 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.83330.83730.63350.74190.62500.8824
resnet0.88100.77210.61820.77250.50000.9706
vit0.75000.78490.58940.66120.62500.7794
+
+
+
+
+

50% · 146 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.79760.80880.51920.70540.62500.8382
resnet0.71430.71420.37520.64080.68750.7206
vit0.69050.72980.47140.62080.68750.6912
+
+
+
+
+

25% · 73 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.75000.74220.38650.63610.50000.8088
resnet0.70240.62410.32830.52860.25000.8088
vit0.60710.64430.41170.53540.56250.6176
+
+
+
+
+

10% · 29 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.80950.63880.36410.44740.00001.0000
resnet0.48810.60940.26860.46050.68750.4412
vit0.79760.63790.39060.67950.50000.8676
+
+
+
+
+

5% · 15 训练样本

val/test 保持完整
+
+
ModelAccAUROCAUPRCF1SensSpec
retfound0.80950.68060.36790.44740.00001.0000
resnet0.82140.56480.35190.62220.25000.9559
vit0.71430.70310.41070.61900.56250.7500
+
+
+
+
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+
+

DR · Diabetic Retinopathy

+ +
+

IDRiD

+ 5-class grade 0–4 · 划分 train/val/test = 318/45/92 · 总计 455 张
+
📷 采集背景：彩色眼底照｜FOV：50°｜设备：Kowa VX-10α（散瞳，托吡卡胺 0.5%）｜分辨率：4288×2848 JPG｜来源：印度 Nanded（Maharashtra）眼科诊所，2009–2017｜全集 516 张（本项目有标签 455 张）｜DR 0–4（ICDR）+ 黄斑水肿风险分级。
+

类别分布 · Class distribution（按 split）

+
+
Split0·grade_01·grade_12·grade_23·grade_34·grade_4合计
train90151095945318
val132168645
test26531171392
合计129221568464455
+
+
+

模型性能 · Performance

+
+
ModelAccuracyBal-Accmacro-AUROCQWKF1-macroPrec-macroRec-macroKappa
RetFound (ViT-L, CFP)0.68480.56570.90690.87820.55100.54000.56570.5737
ResNet-500.61960.55970.88810.83910.55150.56420.55970.4997
ViT-B/160.65220.65440.86460.83560.61920.60840.65440.5399
+
+
+

每类指标 · Per-class metrics

+
+
ClassSupportRetFound (ViT-L, CFP)ResNet-50ViT-B/16
RecallF1AUROCRecallF1AUROCRecallF1AUROC
0·grade_0260.9620.8930.9890.8850.8850.9870.9230.8730.985
1·grade_150.0000.0000.9290.2000.2500.8990.8000.5710.853
2·grade_2310.6450.6560.8600.4520.5190.8230.5810.6100.812
3·grade_3170.5290.4860.8310.6470.5120.8280.3530.3750.786
4·grade_4130.6920.7200.9260.6150.5930.9030.6150.6670.887
+
+
+ +
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+

APTOS-2019

+ 5-class grade 0–4 · 划分 train/val/test = 2930/366/366 · 总计 3662 张
+
📷 采集背景：彩色眼底照｜设备 / FOV / 分辨率：均未公开（多诊所、多相机、跨时间采集，异质性大）｜来源：Aravind 眼科医院（印度），乡村远程筛查｜训练集 3662 张，DR 0–4（ICDR）｜真实世界噪声明显（伪影 / 失焦 / 过曝欠曝 / 标签噪声）。
+

类别分布 · Class distribution（按 split）

+
+
Split0·grade_01·grade_12·grade_23·grade_34·grade_4合计
train14343008081542342930
val172401042228366
test19930871733366
合计18053709991932953662
+
+
+

模型性能 · Performance

+
+
ModelAccuracyBal-Accmacro-AUROCQWKF1-macroPrec-macroRec-macroKappa
RetFound (ViT-L, CFP)0.83610.63230.94780.90560.64950.69550.63230.7384
ResNet-500.81690.62380.91880.86000.61970.62860.62380.7092
ViT-B/160.79230.61700.91650.87480.60450.61930.61700.6751
+
+
+

每类指标 · Per-class metrics

+
+
ClassSupportRetFound (ViT-L, CFP)ResNet-50ViT-B/16
RecallF1AUROCRecallF1AUROCRecallF1AUROC
0·grade_01990.9800.9870.9980.9850.9820.9990.9750.9800.995
1·grade_1300.5330.5250.9370.6000.5710.9270.6000.5000.877
2·grade_2870.8390.7560.9490.7590.7500.9490.6440.6830.914
3·grade_3170.2940.3120.9100.4120.3500.8510.4120.3040.908
4·grade_4330.5150.6670.9460.3640.4440.8680.4550.5560.887
+
+
+ +
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+

DeepDRiD

+ 5-class grade 0–4 · 划分 train/val/test = 1200/400/400 · 总计 2000 张
+
📷 采集背景：彩色眼底照（常规，非超广角）｜设备：Topcon 非散瞳（具体型号未公开）｜FOV≈45–60°、分辨率≈1956×1934（来自补充材料，中等可信）｜来源：上海市第六人民医院（中国）糖尿病筛查队列｜2000 张 / 500 人，每眼双视野（视盘中心 + 黄斑中心）｜DR 0–4 + 图像质量标注。
+

类别分布 · Class distribution（按 split）

+
+
Split0·grade_01·grade_12·grade_23·grade_34·grade_4合计
train539141234214721200
val17446926820400
test20036727220400
合计9132233983541122000
+
+
+

模型性能 · Performance

+
+
ModelAccuracyBal-Accmacro-AUROCQWKF1-macroPrec-macroRec-macroKappa
RetFound (ViT-L, CFP)0.75250.63760.92710.84420.65340.69200.63760.6339
ResNet-500.70250.53110.87600.82700.53420.55700.53110.5605
ViT-B/160.70750.59520.89440.82420.59450.62800.59520.5836
+
+
+

每类指标 · Per-class metrics

+
+
ClassSupportRetFound (ViT-L, CFP)ResNet-50ViT-B/16
RecallF1AUROCRecallF1AUROCRecallF1AUROC
0·grade_02000.8350.8430.9590.8500.8670.9380.7650.8360.955
1·grade_1360.3330.3120.8530.1670.1760.8010.3060.2470.800
2·grade_2720.7220.6840.9330.6390.5790.8810.7080.6460.907
3·grade_3720.8470.8470.9820.7500.7150.9400.8470.7770.938
4·grade_4200.4500.5810.9100.2500.3330.8190.3500.4670.872
+
+
+ +
详细图：混淆矩阵 / ROC 曲线

混淆矩阵

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16

ROC

RetFound (ViT-L, CFP)
ResNet-50
ViT-B/16
+
+
+ +
\ No newline at end of file diff --git a/results/summary.csv b/results/summary.csv new file mode 100644 index 0000000000000000000000000000000000000000..68f4fc4822d54a151254dea5b02e19fd7ebbb5e3 --- /dev/null +++ b/results/summary.csv @@ -0,0 +1,22 @@ +dataset,model,task,n_test,accuracy,balanced_accuracy,f1_macro,precision_macro,recall_macro,cohen_kappa,quadratic_weighted_kappa,mcc,auroc,auprc,sensitivity,specificity,auroc_macro_ovr,auprc_macro +adam,resnet,binary,80.0000,0.8250,0.7885,0.7667,0.7524,0.7885,0.5349,0.5349,0.5397,0.9194,0.8160,0.7222,0.8548,, +adam,retfound,binary,80.0000,0.9250,0.8925,0.8925,0.8925,0.8925,0.7849,0.7849,0.7849,0.9516,0.9214,0.8333,0.9516,, +adam,vit,binary,80.0000,0.9125,0.8844,0.8770,0.8701,0.8844,0.7540,0.7540,0.7544,0.9319,0.8854,0.8333,0.9355,, +airogs,resnet,binary,1000.0000,0.9000,0.9000,0.9000,0.9001,0.9000,0.8000,0.8000,0.8001,0.9614,0.9596,0.9080,0.8920,, +airogs,retfound,binary,1000.0000,0.9080,0.9080,0.9080,0.9084,0.9080,0.8160,0.8160,0.8164,0.9708,0.9715,0.8920,0.9240,, +airogs,vit,binary,1000.0000,0.9000,0.9000,0.9000,0.9001,0.9000,0.8000,0.8000,0.8001,0.9600,0.9625,0.9060,0.8940,, +aptos,resnet,multiclass,366.0000,0.8169,0.6238,0.6197,0.6286,0.6238,0.7092,0.8600,0.7101,,,,,0.9188,0.6550 +aptos,retfound,multiclass,366.0000,0.8361,0.6323,0.6495,0.6955,0.6323,0.7384,0.9056,0.7411,,,,,0.9478,0.6934 +aptos,vit,multiclass,366.0000,0.7923,0.6170,0.6045,0.6193,0.6170,0.6751,0.8748,0.6772,,,,,0.9165,0.6056 +deepdrid,resnet,multiclass,400.0000,0.7025,0.5311,0.5342,0.5570,0.5311,0.5605,0.8270,0.5616,,,,,0.8760,0.5810 +deepdrid,retfound,multiclass,400.0000,0.7525,0.6376,0.6534,0.6920,0.6376,0.6339,0.8442,0.6344,,,,,0.9271,0.7323 +deepdrid,vit,multiclass,400.0000,0.7075,0.5952,0.5945,0.6280,0.5952,0.5836,0.8242,0.5888,,,,,0.8944,0.6585 +idrid,resnet,multiclass,92.0000,0.6196,0.5597,0.5515,0.5642,0.5597,0.4997,0.8391,0.5056,,,,,0.8881,0.6336 +idrid,retfound,multiclass,92.0000,0.6848,0.5657,0.5510,0.5400,0.5657,0.5737,0.8782,0.5762,,,,,0.9069,0.6865 +idrid,vit,multiclass,92.0000,0.6522,0.6544,0.6192,0.6084,0.6544,0.5399,0.8356,0.5417,,,,,0.8646,0.6594 +mmac,resnet,multiclass,279.0000,0.8244,0.7259,0.7360,0.7484,0.7259,0.7510,0.8789,0.7515,,,,,0.9445,0.7624 +mmac,retfound,multiclass,279.0000,0.8566,0.7423,0.7602,0.8011,0.7423,0.7967,0.9264,0.7971,,,,,0.9673,0.8271 +mmac,vit,multiclass,279.0000,0.8244,0.7345,0.7390,0.7522,0.7345,0.7526,0.9052,0.7529,,,,,0.9512,0.7603 +papila,resnet,binary,84.0000,0.8810,0.7353,0.7725,0.8459,0.7353,0.5494,0.5494,0.5706,0.7941,0.7048,0.5000,0.9706,, +papila,retfound,binary,84.0000,0.8333,0.7537,0.7419,0.7323,0.7537,0.4842,0.4842,0.4855,0.8373,0.6335,0.6250,0.8824,, +papila,vit,binary,84.0000,0.7500,0.7022,0.6612,0.6492,0.7022,0.3328,0.3328,0.3473,0.7849,0.5894,0.6250,0.7794,,