[2021-03-20 14:40:23] Created experiment 0:
[2021-03-20 14:40:23]  - Model: vgg-pretrained
[2021-03-20 14:40:23]  - Acquisition function: least-confidence
[2021-03-20 14:40:23] Loading cifar10 test set...
[2021-03-20 14:40:24] Experiment repeat 1/1
[2021-03-20 14:40:24] Seeded: 5
[2021-03-20 14:40:24] Using 2.00% labels of the dataset (1000/50000)
[2021-03-20 14:40:24] Randomly labelled 1000/50000
[2021-03-20 14:40:24] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-20 14:40:24] Seeded: 6
[2021-03-20 14:40:24] Running: experiment 0
[2021-03-20 14:40:24] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-20 14:40:24] Creating pretrained=True VGG16...
[2021-03-20 14:40:25] No parameter reset since we are using a pretrained model.
[2021-03-20 14:40:25] vgg-pretrained: initialized 15245130 parameters.
[2021-03-20 14:40:25] Creating trainer with model on device: cuda
[2021-03-20 14:40:30] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-20 14:40:52] Training accuracy: 0.9826
[2021-03-20 14:40:52] Testing on 10000 data points...
[2021-03-20 14:40:54] Test score for 1000 training labels: 0.7171
[2021-03-20 14:40:54] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:40:54] Found 50000 unlabelled features.
[2021-03-20 14:41:04] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-20 14:41:04] Creating trainer with model on device: cuda
[2021-03-20 14:41:04] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-20 14:41:12] Training accuracy: 0.9401
[2021-03-20 14:41:12] Testing on 10000 data points...
[2021-03-20 14:41:14] Test score for 1400 training labels: 0.7481
[2021-03-20 14:41:14] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:41:14] Found 50000 unlabelled features.
[2021-03-20 14:41:24] Showing labelled: True (1781/50000 visible, 19 redundant)
[2021-03-20 14:41:24] Creating trainer with model on device: cuda
[2021-03-20 14:41:24] Training vgg-pretrained across 1800 data points in cifar10...
[2021-03-20 14:41:35] Training accuracy: 0.9588
[2021-03-20 14:41:35] Testing on 10000 data points...
[2021-03-20 14:41:37] Test score for 1800 training labels: 0.7665
[2021-03-20 14:41:37] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:41:37] Found 50000 unlabelled features.
[2021-03-20 14:41:46] Showing labelled: True (2169/50000 visible, 31 redundant)
[2021-03-20 14:41:46] Creating trainer with model on device: cuda
[2021-03-20 14:41:46] Training vgg-pretrained across 2200 data points in cifar10...
[2021-03-20 14:42:00] Training accuracy: 0.9785
[2021-03-20 14:42:00] Testing on 10000 data points...
[2021-03-20 14:42:02] Test score for 2200 training labels: 0.7785
[2021-03-20 14:42:02] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:42:02] Found 50000 unlabelled features.
[2021-03-20 14:42:11] Showing labelled: True (2558/50000 visible, 42 redundant)
[2021-03-20 14:42:11] Creating trainer with model on device: cuda
[2021-03-20 14:42:11] Training vgg-pretrained across 2600 data points in cifar10...
[2021-03-20 14:42:27] Training accuracy: 0.9932
[2021-03-20 14:42:27] Testing on 10000 data points...
[2021-03-20 14:42:29] Test score for 2600 training labels: 0.7869
[2021-03-20 14:42:29] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:42:29] Found 50000 unlabelled features.
[2021-03-20 14:42:38] Showing labelled: True (2955/50000 visible, 45 redundant)
[2021-03-20 14:42:38] Creating trainer with model on device: cuda
[2021-03-20 14:42:38] Training vgg-pretrained across 3000 data points in cifar10...
[2021-03-20 14:42:56] Training accuracy: 0.9908
[2021-03-20 14:42:56] Testing on 10000 data points...
[2021-03-20 14:42:58] Test score for 3000 training labels: 0.8006
[2021-03-20 14:42:58] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:42:58] Found 50000 unlabelled features.
[2021-03-20 14:43:08] Showing labelled: True (3353/50000 visible, 47 redundant)
[2021-03-20 14:43:08] Creating trainer with model on device: cuda
[2021-03-20 14:43:08] Training vgg-pretrained across 3400 data points in cifar10...
[2021-03-20 14:43:28] Training accuracy: 0.9833
[2021-03-20 14:43:28] Testing on 10000 data points...
[2021-03-20 14:43:30] Test score for 3400 training labels: 0.7974
[2021-03-20 14:43:30] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:43:30] Found 50000 unlabelled features.
[2021-03-20 14:43:40] Showing labelled: True (3737/50000 visible, 63 redundant)
[2021-03-20 14:43:40] Creating trainer with model on device: cuda
[2021-03-20 14:43:40] Training vgg-pretrained across 3800 data points in cifar10...
[2021-03-20 14:44:02] Training accuracy: 0.9992
[2021-03-20 14:44:02] Testing on 10000 data points...
[2021-03-20 14:44:04] Test score for 3800 training labels: 0.8147
[2021-03-20 14:44:04] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:44:04] Found 50000 unlabelled features.
[2021-03-20 14:44:14] Showing labelled: True (4136/50000 visible, 64 redundant)
[2021-03-20 14:44:14] Creating trainer with model on device: cuda
[2021-03-20 14:44:14] Training vgg-pretrained across 4200 data points in cifar10...
[2021-03-20 14:44:39] Training accuracy: 0.9989
[2021-03-20 14:44:39] Testing on 10000 data points...
[2021-03-20 14:44:41] Test score for 4200 training labels: 0.8249
[2021-03-20 14:44:41] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:44:41] Found 50000 unlabelled features.
[2021-03-20 14:44:51] Showing labelled: True (4534/50000 visible, 66 redundant)
[2021-03-20 14:44:51] Creating trainer with model on device: cuda
[2021-03-20 14:44:51] Training vgg-pretrained across 4600 data points in cifar10...
[2021-03-20 14:45:18] Training accuracy: 0.9999
[2021-03-20 14:45:18] Testing on 10000 data points...
[2021-03-20 14:45:20] Test score for 4600 training labels: 0.8299
[2021-03-20 14:45:20] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-20 14:45:20] Found 50000 unlabelled features.
[2021-03-20 14:45:30] Showing labelled: True (4933/50000 visible, 67 redundant)
[2021-03-20 14:45:30] Creating trainer with model on device: cuda
[2021-03-20 14:45:30] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-20 14:46:00] Training accuracy: 0.9998
[2021-03-20 14:46:00] Testing on 10000 data points...
[2021-03-20 14:46:02] Test score for 5000 training labels: 0.8328
[2021-03-20 14:46:02] Updated results: ../results/cifar10/vgg_pretrained_lc_nolsmooth/results.json
