[2021-03-22 00:53:27] Created experiment 0:
[2021-03-22 00:53:27]  - Model: vgg-pt
[2021-03-22 00:53:27]  - Acquisition function: random
[2021-03-22 00:53:27] Created experiment 1:
[2021-03-22 00:53:27]  - Model: vgg-pt
[2021-03-22 00:53:27]  - Acquisition function: greedy-coreset
[2021-03-22 00:53:27] Created experiment 2:
[2021-03-22 00:53:27]  - Model: vgg-pt
[2021-03-22 00:53:27]  - Acquisition function: least-confidence
[2021-03-22 00:53:27] Created experiment 3:
[2021-03-22 00:53:27]  - Model: vgg-pt
[2021-03-22 00:53:27]  - Acquisition function: lc-beam-pweighted-coreset (beams=10)
[2021-03-22 00:53:27] Loading cifar10 test set...
[2021-03-22 00:53:28] Experiment repeat 1/5
[2021-03-22 00:53:28] Seeded: 5
[2021-03-22 00:53:28] Using 10.00% labels of the dataset (5000/50000)
[2021-03-22 00:53:28] Randomly labelled 5000/50000
[2021-03-22 00:53:28] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 00:53:28] Seeded: 6
[2021-03-22 00:53:28] Running: experiment 0
[2021-03-22 00:53:28] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 00:53:28] Creating pretrained=True VGG16...
[2021-03-22 00:53:30] No parameter reset since we are using a pretrained model.
[2021-03-22 00:53:30] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 00:53:30] Creating trainer with model on device: cuda
[2021-03-22 00:53:35] Training with expected score >= 0.9900
[2021-03-22 00:53:35] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 00:55:26] Training accuracy: 0.9981
[2021-03-22 00:55:26] Testing on 10000 data points...
[2021-03-22 00:55:28] Test score for 5000 training labels: 0.8063
[2021-03-22 00:55:28] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 00:55:28] Found 50000 unlabelled features.
[2021-03-22 00:55:33] Showing labelled: True (9512/50000 visible, 488 redundant)
[2021-03-22 00:55:33] Creating trainer with model on device: cuda
[2021-03-22 00:55:33] Training with expected score >= 0.9900
[2021-03-22 00:55:33] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 00:57:03] Training accuracy: 0.9951
[2021-03-22 00:57:03] Testing on 10000 data points...
[2021-03-22 00:57:05] Test score for 10000 training labels: 0.8305
[2021-03-22 00:57:05] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 00:57:05] Found 50000 unlabelled features.
[2021-03-22 00:57:10] Showing labelled: True (13552/50000 visible, 1448 redundant)
[2021-03-22 00:57:10] Creating trainer with model on device: cuda
[2021-03-22 00:57:10] Training with expected score >= 0.9900
[2021-03-22 00:57:10] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 00:59:25] Training accuracy: 0.9993
[2021-03-22 00:59:25] Testing on 10000 data points...
[2021-03-22 00:59:27] Test score for 15000 training labels: 0.8431
[2021-03-22 00:59:27] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 00:59:27] Found 50000 unlabelled features.
[2021-03-22 00:59:33] Showing labelled: True (17219/50000 visible, 2781 redundant)
[2021-03-22 00:59:33] Creating trainer with model on device: cuda
[2021-03-22 00:59:33] Training with expected score >= 0.9900
[2021-03-22 00:59:33] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 01:02:32] Training accuracy: 0.9993
[2021-03-22 01:02:32] Testing on 10000 data points...
[2021-03-22 01:02:34] Test score for 20000 training labels: 0.8547
[2021-03-22 01:02:34] Seeded: 6
[2021-03-22 01:02:34] Running: experiment 1
[2021-03-22 01:02:34] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:02:34] Creating pretrained=True VGG16...
[2021-03-22 01:02:36] No parameter reset since we are using a pretrained model.
[2021-03-22 01:02:36] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 01:02:36] Creating trainer with model on device: cuda
[2021-03-22 01:02:36] Training with expected score >= 0.9900
[2021-03-22 01:02:36] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 01:04:28] Training accuracy: 0.9996
[2021-03-22 01:04:28] Testing on 10000 data points...
[2021-03-22 01:04:30] Test score for 5000 training labels: 0.8028
[2021-03-22 01:04:30] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:04:30] Found 50000 unlabelled features.
[2021-03-22 01:04:40] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:04:54] Searching for coresets greedily...
[2021-03-22 01:05:04] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 01:05:04] Creating trainer with model on device: cuda
[2021-03-22 01:05:04] Training with expected score >= 0.9900
[2021-03-22 01:05:04] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 01:06:34] Training accuracy: 0.9965
[2021-03-22 01:06:34] Testing on 10000 data points...
[2021-03-22 01:06:36] Test score for 10000 training labels: 0.8435
[2021-03-22 01:06:36] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:06:36] Found 50000 unlabelled features.
[2021-03-22 01:06:48] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:07:16] Searching for coresets greedily...
[2021-03-22 01:07:25] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 01:07:25] Creating trainer with model on device: cuda
[2021-03-22 01:07:25] Training with expected score >= 0.9900
[2021-03-22 01:07:25] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 01:09:40] Training accuracy: 0.9975
[2021-03-22 01:09:40] Testing on 10000 data points...
[2021-03-22 01:09:42] Test score for 15000 training labels: 0.8649
[2021-03-22 01:09:42] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:09:42] Found 50000 unlabelled features.
[2021-03-22 01:09:54] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:10:37] Searching for coresets greedily...
[2021-03-22 01:10:46] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 01:10:46] Creating trainer with model on device: cuda
[2021-03-22 01:10:46] Training with expected score >= 0.9900
[2021-03-22 01:10:46] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 01:13:45] Training accuracy: 0.9996
[2021-03-22 01:13:45] Testing on 10000 data points...
[2021-03-22 01:13:47] Test score for 20000 training labels: 0.8760
[2021-03-22 01:13:47] Seeded: 6
[2021-03-22 01:13:47] Running: experiment 2
[2021-03-22 01:13:47] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:13:47] Creating pretrained=True VGG16...
[2021-03-22 01:13:49] No parameter reset since we are using a pretrained model.
[2021-03-22 01:13:49] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 01:13:49] Creating trainer with model on device: cuda
[2021-03-22 01:13:49] Training with expected score >= 0.9900
[2021-03-22 01:13:49] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 01:15:41] Training accuracy: 0.9994
[2021-03-22 01:15:41] Testing on 10000 data points...
[2021-03-22 01:15:43] Test score for 5000 training labels: 0.8054
[2021-03-22 01:15:43] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:15:43] Found 50000 unlabelled features.
[2021-03-22 01:15:52] Showing labelled: True (9992/50000 visible, 8 redundant)
[2021-03-22 01:15:52] Creating trainer with model on device: cuda
[2021-03-22 01:15:52] Training with expected score >= 0.9900
[2021-03-22 01:15:52] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 01:17:22] Training accuracy: 0.9918
[2021-03-22 01:17:22] Testing on 10000 data points...
[2021-03-22 01:17:24] Test score for 10000 training labels: 0.8550
[2021-03-22 01:17:24] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:17:24] Found 50000 unlabelled features.
[2021-03-22 01:17:34] Showing labelled: True (14527/50000 visible, 473 redundant)
[2021-03-22 01:17:34] Creating trainer with model on device: cuda
[2021-03-22 01:17:34] Training with expected score >= 0.9900
[2021-03-22 01:17:34] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 01:19:48] Training accuracy: 0.9987
[2021-03-22 01:19:48] Testing on 10000 data points...
[2021-03-22 01:19:50] Test score for 15000 training labels: 0.8746
[2021-03-22 01:19:50] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:19:50] Found 50000 unlabelled features.
[2021-03-22 01:20:00] Showing labelled: True (17849/50000 visible, 2151 redundant)
[2021-03-22 01:20:00] Creating trainer with model on device: cuda
[2021-03-22 01:20:00] Training with expected score >= 0.9900
[2021-03-22 01:20:00] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 01:23:00] Training accuracy: 0.9995
[2021-03-22 01:23:00] Testing on 10000 data points...
[2021-03-22 01:23:02] Test score for 20000 training labels: 0.8823
[2021-03-22 01:23:02] Seeded: 6
[2021-03-22 01:23:02] Running: experiment 3
[2021-03-22 01:23:02] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:23:02] Creating pretrained=True VGG16...
[2021-03-22 01:23:03] No parameter reset since we are using a pretrained model.
[2021-03-22 01:23:03] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 01:23:03] Creating trainer with model on device: cuda
[2021-03-22 01:23:03] Training with expected score >= 0.9900
[2021-03-22 01:23:03] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 01:24:56] Training accuracy: 0.9943
[2021-03-22 01:24:56] Testing on 10000 data points...
[2021-03-22 01:24:58] Test score for 5000 training labels: 0.7999
[2021-03-22 01:24:58] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:24:58] Found 50000 unlabelled features.
[2021-03-22 01:25:09] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:25:23] Searching for coresets with 10 beams...
[2021-03-22 01:28:06] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 01:28:06] Creating trainer with model on device: cuda
[2021-03-22 01:28:06] Training with expected score >= 0.9900
[2021-03-22 01:28:06] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 01:29:37] Training accuracy: 0.9925
[2021-03-22 01:29:37] Testing on 10000 data points...
[2021-03-22 01:29:39] Test score for 10000 training labels: 0.8473
[2021-03-22 01:29:39] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:29:39] Found 50000 unlabelled features.
[2021-03-22 01:29:51] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:30:19] Searching for coresets with 10 beams...
[2021-03-22 01:33:02] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 01:33:02] Creating trainer with model on device: cuda
[2021-03-22 01:33:02] Training with expected score >= 0.9900
[2021-03-22 01:33:02] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 01:35:18] Training accuracy: 0.9980
[2021-03-22 01:35:18] Testing on 10000 data points...
[2021-03-22 01:35:20] Test score for 15000 training labels: 0.8724
[2021-03-22 01:35:20] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:35:20] Found 50000 unlabelled features.
[2021-03-22 01:35:33] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:36:15] Searching for coresets with 10 beams...
[2021-03-22 01:38:59] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 01:38:59] Creating trainer with model on device: cuda
[2021-03-22 01:38:59] Training with expected score >= 0.9900
[2021-03-22 01:38:59] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 01:42:00] Training accuracy: 0.9992
[2021-03-22 01:42:00] Testing on 10000 data points...
[2021-03-22 01:42:02] Test score for 20000 training labels: 0.8847
[2021-03-22 01:42:02] Experiment repeat 2/5
[2021-03-22 01:42:02] Seeded: 6
[2021-03-22 01:42:02] Using 10.00% labels of the dataset (5000/50000)
[2021-03-22 01:42:02] Randomly labelled 5000/50000
[2021-03-22 01:42:02] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:42:02] Seeded: 7
[2021-03-22 01:42:02] Running: experiment 0
[2021-03-22 01:42:02] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:42:02] Creating pretrained=True VGG16...
[2021-03-22 01:42:04] No parameter reset since we are using a pretrained model.
[2021-03-22 01:42:04] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 01:42:04] Creating trainer with model on device: cuda
[2021-03-22 01:42:04] Training with expected score >= 0.9900
[2021-03-22 01:42:04] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 01:43:56] Training accuracy: 0.9989
[2021-03-22 01:43:56] Testing on 10000 data points...
[2021-03-22 01:43:58] Test score for 5000 training labels: 0.8106
[2021-03-22 01:43:58] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:43:58] Found 50000 unlabelled features.
[2021-03-22 01:44:04] Showing labelled: True (9529/50000 visible, 471 redundant)
[2021-03-22 01:44:04] Creating trainer with model on device: cuda
[2021-03-22 01:44:04] Training with expected score >= 0.9900
[2021-03-22 01:44:04] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 01:45:34] Training accuracy: 0.9991
[2021-03-22 01:45:34] Testing on 10000 data points...
[2021-03-22 01:45:36] Test score for 10000 training labels: 0.8345
[2021-03-22 01:45:36] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:45:36] Found 50000 unlabelled features.
[2021-03-22 01:45:41] Showing labelled: True (13568/50000 visible, 1432 redundant)
[2021-03-22 01:45:41] Creating trainer with model on device: cuda
[2021-03-22 01:45:41] Training with expected score >= 0.9900
[2021-03-22 01:45:41] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 01:47:57] Training accuracy: 0.9964
[2021-03-22 01:47:57] Testing on 10000 data points...
[2021-03-22 01:47:59] Test score for 15000 training labels: 0.8462
[2021-03-22 01:47:59] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:47:59] Found 50000 unlabelled features.
[2021-03-22 01:48:04] Showing labelled: True (17156/50000 visible, 2844 redundant)
[2021-03-22 01:48:04] Creating trainer with model on device: cuda
[2021-03-22 01:48:04] Training with expected score >= 0.9900
[2021-03-22 01:48:04] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 01:51:04] Training accuracy: 0.9997
[2021-03-22 01:51:04] Testing on 10000 data points...
[2021-03-22 01:51:06] Test score for 20000 training labels: 0.8544
[2021-03-22 01:51:06] Seeded: 7
[2021-03-22 01:51:06] Running: experiment 1
[2021-03-22 01:51:06] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 01:51:06] Creating pretrained=True VGG16...
[2021-03-22 01:51:08] No parameter reset since we are using a pretrained model.
[2021-03-22 01:51:08] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 01:51:08] Creating trainer with model on device: cuda
[2021-03-22 01:51:08] Training with expected score >= 0.9900
[2021-03-22 01:51:08] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 01:53:00] Training accuracy: 0.9990
[2021-03-22 01:53:00] Testing on 10000 data points...
[2021-03-22 01:53:02] Test score for 5000 training labels: 0.8117
[2021-03-22 01:53:02] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:53:02] Found 50000 unlabelled features.
[2021-03-22 01:53:12] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:53:26] Searching for coresets greedily...
[2021-03-22 01:53:36] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 01:53:36] Creating trainer with model on device: cuda
[2021-03-22 01:53:36] Training with expected score >= 0.9900
[2021-03-22 01:53:36] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 01:55:06] Training accuracy: 0.9948
[2021-03-22 01:55:06] Testing on 10000 data points...
[2021-03-22 01:55:08] Test score for 10000 training labels: 0.8445
[2021-03-22 01:55:08] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:55:08] Found 50000 unlabelled features.
[2021-03-22 01:55:19] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:55:48] Searching for coresets greedily...
[2021-03-22 01:55:57] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 01:55:57] Creating trainer with model on device: cuda
[2021-03-22 01:55:57] Training with expected score >= 0.9900
[2021-03-22 01:55:57] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 01:58:12] Training accuracy: 0.9987
[2021-03-22 01:58:12] Testing on 10000 data points...
[2021-03-22 01:58:14] Test score for 15000 training labels: 0.8641
[2021-03-22 01:58:14] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 01:58:14] Found 50000 unlabelled features.
[2021-03-22 01:58:27] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 01:59:09] Searching for coresets greedily...
[2021-03-22 01:59:18] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 01:59:18] Creating trainer with model on device: cuda
[2021-03-22 01:59:18] Training with expected score >= 0.9900
[2021-03-22 01:59:18] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 02:02:18] Training accuracy: 0.9998
[2021-03-22 02:02:18] Testing on 10000 data points...
[2021-03-22 02:02:20] Test score for 20000 training labels: 0.8773
[2021-03-22 02:02:20] Seeded: 7
[2021-03-22 02:02:20] Running: experiment 2
[2021-03-22 02:02:20] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:02:20] Creating pretrained=True VGG16...
[2021-03-22 02:02:22] No parameter reset since we are using a pretrained model.
[2021-03-22 02:02:22] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 02:02:22] Creating trainer with model on device: cuda
[2021-03-22 02:02:22] Training with expected score >= 0.9900
[2021-03-22 02:02:22] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 02:04:14] Training accuracy: 0.9991
[2021-03-22 02:04:14] Testing on 10000 data points...
[2021-03-22 02:04:16] Test score for 5000 training labels: 0.8132
[2021-03-22 02:04:16] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:04:16] Found 50000 unlabelled features.
[2021-03-22 02:04:25] Showing labelled: True (9991/50000 visible, 9 redundant)
[2021-03-22 02:04:25] Creating trainer with model on device: cuda
[2021-03-22 02:04:25] Training with expected score >= 0.9900
[2021-03-22 02:04:25] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 02:05:56] Training accuracy: 0.9909
[2021-03-22 02:05:56] Testing on 10000 data points...
[2021-03-22 02:05:57] Test score for 10000 training labels: 0.8533
[2021-03-22 02:05:58] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:05:58] Found 50000 unlabelled features.
[2021-03-22 02:06:07] Showing labelled: True (14416/50000 visible, 584 redundant)
[2021-03-22 02:06:07] Creating trainer with model on device: cuda
[2021-03-22 02:06:07] Training with expected score >= 0.9900
[2021-03-22 02:06:07] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 02:08:22] Training accuracy: 0.9975
[2021-03-22 02:08:22] Testing on 10000 data points...
[2021-03-22 02:08:24] Test score for 15000 training labels: 0.8743
[2021-03-22 02:08:24] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:08:24] Found 50000 unlabelled features.
[2021-03-22 02:08:33] Showing labelled: True (17851/50000 visible, 2149 redundant)
[2021-03-22 02:08:33] Creating trainer with model on device: cuda
[2021-03-22 02:08:33] Training with expected score >= 0.9900
[2021-03-22 02:08:33] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 02:11:33] Training accuracy: 0.9990
[2021-03-22 02:11:33] Testing on 10000 data points...
[2021-03-22 02:11:35] Test score for 20000 training labels: 0.8822
[2021-03-22 02:11:35] Seeded: 7
[2021-03-22 02:11:35] Running: experiment 3
[2021-03-22 02:11:35] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:11:35] Creating pretrained=True VGG16...
[2021-03-22 02:11:37] No parameter reset since we are using a pretrained model.
[2021-03-22 02:11:37] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 02:11:37] Creating trainer with model on device: cuda
[2021-03-22 02:11:37] Training with expected score >= 0.9900
[2021-03-22 02:11:37] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 02:13:29] Training accuracy: 0.9986
[2021-03-22 02:13:29] Testing on 10000 data points...
[2021-03-22 02:13:31] Test score for 5000 training labels: 0.8120
[2021-03-22 02:13:31] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:13:31] Found 50000 unlabelled features.
[2021-03-22 02:13:41] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:13:56] Searching for coresets with 10 beams...
[2021-03-22 02:16:39] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 02:16:39] Creating trainer with model on device: cuda
[2021-03-22 02:16:39] Training with expected score >= 0.9900
[2021-03-22 02:16:39] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 02:18:10] Training accuracy: 0.9943
[2021-03-22 02:18:10] Testing on 10000 data points...
[2021-03-22 02:18:11] Test score for 10000 training labels: 0.8560
[2021-03-22 02:18:11] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:18:11] Found 50000 unlabelled features.
[2021-03-22 02:18:23] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:18:51] Searching for coresets with 10 beams...
[2021-03-22 02:21:35] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 02:21:35] Creating trainer with model on device: cuda
[2021-03-22 02:21:35] Training with expected score >= 0.9900
[2021-03-22 02:21:35] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 02:23:50] Training accuracy: 0.9992
[2021-03-22 02:23:50] Testing on 10000 data points...
[2021-03-22 02:23:52] Test score for 15000 training labels: 0.8765
[2021-03-22 02:23:52] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:23:52] Found 50000 unlabelled features.
[2021-03-22 02:24:05] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:24:47] Searching for coresets with 10 beams...
[2021-03-22 02:27:31] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 02:27:31] Creating trainer with model on device: cuda
[2021-03-22 02:27:31] Training with expected score >= 0.9900
[2021-03-22 02:27:31] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 02:30:31] Training accuracy: 0.9989
[2021-03-22 02:30:31] Testing on 10000 data points...
[2021-03-22 02:30:33] Test score for 20000 training labels: 0.8857
[2021-03-22 02:30:33] Experiment repeat 3/5
[2021-03-22 02:30:33] Seeded: 7
[2021-03-22 02:30:33] Using 10.00% labels of the dataset (5000/50000)
[2021-03-22 02:30:33] Randomly labelled 5000/50000
[2021-03-22 02:30:33] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:30:33] Seeded: 8
[2021-03-22 02:30:33] Running: experiment 0
[2021-03-22 02:30:33] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:30:33] Creating pretrained=True VGG16...
[2021-03-22 02:30:35] No parameter reset since we are using a pretrained model.
[2021-03-22 02:30:35] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 02:30:35] Creating trainer with model on device: cuda
[2021-03-22 02:30:35] Training with expected score >= 0.9900
[2021-03-22 02:30:35] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 02:32:27] Training accuracy: 0.9988
[2021-03-22 02:32:27] Testing on 10000 data points...
[2021-03-22 02:32:29] Test score for 5000 training labels: 0.8123
[2021-03-22 02:32:29] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:32:29] Found 50000 unlabelled features.
[2021-03-22 02:32:35] Showing labelled: True (9506/50000 visible, 494 redundant)
[2021-03-22 02:32:35] Creating trainer with model on device: cuda
[2021-03-22 02:32:35] Training with expected score >= 0.9900
[2021-03-22 02:32:35] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 02:34:05] Training accuracy: 0.9984
[2021-03-22 02:34:05] Testing on 10000 data points...
[2021-03-22 02:34:07] Test score for 10000 training labels: 0.8332
[2021-03-22 02:34:07] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:34:07] Found 50000 unlabelled features.
[2021-03-22 02:34:12] Showing labelled: True (13561/50000 visible, 1439 redundant)
[2021-03-22 02:34:12] Creating trainer with model on device: cuda
[2021-03-22 02:34:12] Training with expected score >= 0.9900
[2021-03-22 02:34:12] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 02:36:28] Training accuracy: 0.9989
[2021-03-22 02:36:28] Testing on 10000 data points...
[2021-03-22 02:36:30] Test score for 15000 training labels: 0.8475
[2021-03-22 02:36:30] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:36:30] Found 50000 unlabelled features.
[2021-03-22 02:36:36] Showing labelled: True (17247/50000 visible, 2753 redundant)
[2021-03-22 02:36:36] Creating trainer with model on device: cuda
[2021-03-22 02:36:36] Training with expected score >= 0.9900
[2021-03-22 02:36:36] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 02:39:36] Training accuracy: 1.0000
[2021-03-22 02:39:36] Testing on 10000 data points...
[2021-03-22 02:39:38] Test score for 20000 training labels: 0.8574
[2021-03-22 02:39:38] Seeded: 8
[2021-03-22 02:39:38] Running: experiment 1
[2021-03-22 02:39:38] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:39:38] Creating pretrained=True VGG16...
[2021-03-22 02:39:40] No parameter reset since we are using a pretrained model.
[2021-03-22 02:39:40] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 02:39:40] Creating trainer with model on device: cuda
[2021-03-22 02:39:40] Training with expected score >= 0.9900
[2021-03-22 02:39:40] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 02:41:32] Training accuracy: 0.9993
[2021-03-22 02:41:32] Testing on 10000 data points...
[2021-03-22 02:41:34] Test score for 5000 training labels: 0.8158
[2021-03-22 02:41:34] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:41:34] Found 50000 unlabelled features.
[2021-03-22 02:41:45] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:41:59] Searching for coresets greedily...
[2021-03-22 02:42:09] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 02:42:09] Creating trainer with model on device: cuda
[2021-03-22 02:42:09] Training with expected score >= 0.9900
[2021-03-22 02:42:09] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 02:43:39] Training accuracy: 0.9968
[2021-03-22 02:43:39] Testing on 10000 data points...
[2021-03-22 02:43:41] Test score for 10000 training labels: 0.8463
[2021-03-22 02:43:41] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:43:41] Found 50000 unlabelled features.
[2021-03-22 02:43:52] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:44:21] Searching for coresets greedily...
[2021-03-22 02:44:30] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 02:44:30] Creating trainer with model on device: cuda
[2021-03-22 02:44:30] Training with expected score >= 0.9900
[2021-03-22 02:44:30] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 02:46:45] Training accuracy: 0.9964
[2021-03-22 02:46:45] Testing on 10000 data points...
[2021-03-22 02:46:47] Test score for 15000 training labels: 0.8630
[2021-03-22 02:46:47] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:46:47] Found 50000 unlabelled features.
[2021-03-22 02:47:00] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 02:47:42] Searching for coresets greedily...
[2021-03-22 02:47:52] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 02:47:52] Creating trainer with model on device: cuda
[2021-03-22 02:47:52] Training with expected score >= 0.9900
[2021-03-22 02:47:52] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 02:50:52] Training accuracy: 0.9985
[2021-03-22 02:50:52] Testing on 10000 data points...
[2021-03-22 02:50:53] Test score for 20000 training labels: 0.8813
[2021-03-22 02:50:53] Seeded: 8
[2021-03-22 02:50:53] Running: experiment 2
[2021-03-22 02:50:53] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 02:50:53] Creating pretrained=True VGG16...
[2021-03-22 02:50:55] No parameter reset since we are using a pretrained model.
[2021-03-22 02:50:55] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 02:50:55] Creating trainer with model on device: cuda
[2021-03-22 02:50:55] Training with expected score >= 0.9900
[2021-03-22 02:50:55] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 02:52:47] Training accuracy: 0.9992
[2021-03-22 02:52:47] Testing on 10000 data points...
[2021-03-22 02:52:49] Test score for 5000 training labels: 0.8138
[2021-03-22 02:52:49] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:52:49] Found 50000 unlabelled features.
[2021-03-22 02:52:59] Showing labelled: True (9985/50000 visible, 15 redundant)
[2021-03-22 02:52:59] Creating trainer with model on device: cuda
[2021-03-22 02:52:59] Training with expected score >= 0.9900
[2021-03-22 02:52:59] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 02:54:29] Training accuracy: 0.9954
[2021-03-22 02:54:29] Testing on 10000 data points...
[2021-03-22 02:54:31] Test score for 10000 training labels: 0.8593
[2021-03-22 02:54:31] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:54:31] Found 50000 unlabelled features.
[2021-03-22 02:54:40] Showing labelled: True (14635/50000 visible, 365 redundant)
[2021-03-22 02:54:40] Creating trainer with model on device: cuda
[2021-03-22 02:54:40] Training with expected score >= 0.9900
[2021-03-22 02:54:40] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 02:56:55] Training accuracy: 0.9994
[2021-03-22 02:56:55] Testing on 10000 data points...
[2021-03-22 02:56:57] Test score for 15000 training labels: 0.8745
[2021-03-22 02:56:57] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 02:56:57] Found 50000 unlabelled features.
[2021-03-22 02:57:07] Showing labelled: True (17912/50000 visible, 2088 redundant)
[2021-03-22 02:57:07] Creating trainer with model on device: cuda
[2021-03-22 02:57:07] Training with expected score >= 0.9900
[2021-03-22 02:57:07] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 03:00:07] Training accuracy: 0.9987
[2021-03-22 03:00:07] Testing on 10000 data points...
[2021-03-22 03:00:09] Test score for 20000 training labels: 0.8851
[2021-03-22 03:00:09] Seeded: 8
[2021-03-22 03:00:09] Running: experiment 3
[2021-03-22 03:00:09] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:00:09] Creating pretrained=True VGG16...
[2021-03-22 03:00:10] No parameter reset since we are using a pretrained model.
[2021-03-22 03:00:10] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 03:00:10] Creating trainer with model on device: cuda
[2021-03-22 03:00:10] Training with expected score >= 0.9900
[2021-03-22 03:00:10] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 03:02:03] Training accuracy: 0.9991
[2021-03-22 03:02:03] Testing on 10000 data points...
[2021-03-22 03:02:04] Test score for 5000 training labels: 0.8114
[2021-03-22 03:02:05] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:02:05] Found 50000 unlabelled features.
[2021-03-22 03:02:15] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:02:29] Searching for coresets with 10 beams...
[2021-03-22 03:05:12] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 03:05:12] Creating trainer with model on device: cuda
[2021-03-22 03:05:12] Training with expected score >= 0.9900
[2021-03-22 03:05:12] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 03:06:42] Training accuracy: 0.9968
[2021-03-22 03:06:42] Testing on 10000 data points...
[2021-03-22 03:06:44] Test score for 10000 training labels: 0.8584
[2021-03-22 03:06:44] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:06:44] Found 50000 unlabelled features.
[2021-03-22 03:06:56] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:07:24] Searching for coresets with 10 beams...
[2021-03-22 03:10:07] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 03:10:07] Creating trainer with model on device: cuda
[2021-03-22 03:10:07] Training with expected score >= 0.9900
[2021-03-22 03:10:07] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 03:12:22] Training accuracy: 0.9989
[2021-03-22 03:12:22] Testing on 10000 data points...
[2021-03-22 03:12:24] Test score for 15000 training labels: 0.8771
[2021-03-22 03:12:24] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:12:24] Found 50000 unlabelled features.
[2021-03-22 03:12:37] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:13:20] Searching for coresets with 10 beams...
[2021-03-22 03:16:03] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 03:16:03] Creating trainer with model on device: cuda
[2021-03-22 03:16:03] Training with expected score >= 0.9900
[2021-03-22 03:16:03] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 03:19:03] Training accuracy: 0.9983
[2021-03-22 03:19:03] Testing on 10000 data points...
[2021-03-22 03:19:05] Test score for 20000 training labels: 0.8865
[2021-03-22 03:19:05] Experiment repeat 4/5
[2021-03-22 03:19:05] Seeded: 8
[2021-03-22 03:19:05] Using 10.00% labels of the dataset (5000/50000)
[2021-03-22 03:19:05] Randomly labelled 5000/50000
[2021-03-22 03:19:05] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:19:05] Seeded: 9
[2021-03-22 03:19:05] Running: experiment 0
[2021-03-22 03:19:05] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:19:05] Creating pretrained=True VGG16...
[2021-03-22 03:19:07] No parameter reset since we are using a pretrained model.
[2021-03-22 03:19:07] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 03:19:07] Creating trainer with model on device: cuda
[2021-03-22 03:19:07] Training with expected score >= 0.9900
[2021-03-22 03:19:07] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 03:20:59] Training accuracy: 0.9991
[2021-03-22 03:20:59] Testing on 10000 data points...
[2021-03-22 03:21:01] Test score for 5000 training labels: 0.8056
[2021-03-22 03:21:01] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:21:01] Found 50000 unlabelled features.
[2021-03-22 03:21:07] Showing labelled: True (9469/50000 visible, 531 redundant)
[2021-03-22 03:21:07] Creating trainer with model on device: cuda
[2021-03-22 03:21:07] Training with expected score >= 0.9900
[2021-03-22 03:21:07] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 03:22:37] Training accuracy: 0.9971
[2021-03-22 03:22:37] Testing on 10000 data points...
[2021-03-22 03:22:39] Test score for 10000 training labels: 0.8328
[2021-03-22 03:22:39] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:22:39] Found 50000 unlabelled features.
[2021-03-22 03:22:45] Showing labelled: True (13522/50000 visible, 1478 redundant)
[2021-03-22 03:22:45] Creating trainer with model on device: cuda
[2021-03-22 03:22:45] Training with expected score >= 0.9900
[2021-03-22 03:22:45] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 03:25:00] Training accuracy: 0.9999
[2021-03-22 03:25:00] Testing on 10000 data points...
[2021-03-22 03:25:02] Test score for 15000 training labels: 0.8448
[2021-03-22 03:25:02] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:25:02] Found 50000 unlabelled features.
[2021-03-22 03:25:08] Showing labelled: True (17159/50000 visible, 2841 redundant)
[2021-03-22 03:25:08] Creating trainer with model on device: cuda
[2021-03-22 03:25:08] Training with expected score >= 0.9900
[2021-03-22 03:25:08] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 03:28:08] Training accuracy: 0.9997
[2021-03-22 03:28:08] Testing on 10000 data points...
[2021-03-22 03:28:10] Test score for 20000 training labels: 0.8515
[2021-03-22 03:28:10] Seeded: 9
[2021-03-22 03:28:10] Running: experiment 1
[2021-03-22 03:28:10] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:28:10] Creating pretrained=True VGG16...
[2021-03-22 03:28:11] No parameter reset since we are using a pretrained model.
[2021-03-22 03:28:11] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 03:28:11] Creating trainer with model on device: cuda
[2021-03-22 03:28:11] Training with expected score >= 0.9900
[2021-03-22 03:28:11] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 03:30:04] Training accuracy: 0.9996
[2021-03-22 03:30:04] Testing on 10000 data points...
[2021-03-22 03:30:06] Test score for 5000 training labels: 0.8066
[2021-03-22 03:30:06] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:30:06] Found 50000 unlabelled features.
[2021-03-22 03:30:16] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:30:30] Searching for coresets greedily...
[2021-03-22 03:30:40] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 03:30:40] Creating trainer with model on device: cuda
[2021-03-22 03:30:40] Training with expected score >= 0.9900
[2021-03-22 03:30:40] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 03:32:10] Training accuracy: 0.9958
[2021-03-22 03:32:10] Testing on 10000 data points...
[2021-03-22 03:32:12] Test score for 10000 training labels: 0.8429
[2021-03-22 03:32:12] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:32:12] Found 50000 unlabelled features.
[2021-03-22 03:32:24] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:32:52] Searching for coresets greedily...
[2021-03-22 03:33:02] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 03:33:02] Creating trainer with model on device: cuda
[2021-03-22 03:33:02] Training with expected score >= 0.9900
[2021-03-22 03:33:02] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 03:35:17] Training accuracy: 0.9980
[2021-03-22 03:35:17] Testing on 10000 data points...
[2021-03-22 03:35:19] Test score for 15000 training labels: 0.8651
[2021-03-22 03:35:19] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:35:19] Found 50000 unlabelled features.
[2021-03-22 03:35:32] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:36:14] Searching for coresets greedily...
[2021-03-22 03:36:24] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 03:36:24] Creating trainer with model on device: cuda
[2021-03-22 03:36:24] Training with expected score >= 0.9900
[2021-03-22 03:36:24] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 03:39:24] Training accuracy: 0.9998
[2021-03-22 03:39:24] Testing on 10000 data points...
[2021-03-22 03:39:26] Test score for 20000 training labels: 0.8768
[2021-03-22 03:39:26] Seeded: 9
[2021-03-22 03:39:26] Running: experiment 2
[2021-03-22 03:39:26] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:39:26] Creating pretrained=True VGG16...
[2021-03-22 03:39:27] No parameter reset since we are using a pretrained model.
[2021-03-22 03:39:27] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 03:39:27] Creating trainer with model on device: cuda
[2021-03-22 03:39:27] Training with expected score >= 0.9900
[2021-03-22 03:39:27] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 03:41:20] Training accuracy: 0.9983
[2021-03-22 03:41:20] Testing on 10000 data points...
[2021-03-22 03:41:22] Test score for 5000 training labels: 0.8063
[2021-03-22 03:41:22] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:41:22] Found 50000 unlabelled features.
[2021-03-22 03:41:32] Showing labelled: True (9985/50000 visible, 15 redundant)
[2021-03-22 03:41:32] Creating trainer with model on device: cuda
[2021-03-22 03:41:32] Training with expected score >= 0.9900
[2021-03-22 03:41:32] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 03:43:02] Training accuracy: 0.9952
[2021-03-22 03:43:02] Testing on 10000 data points...
[2021-03-22 03:43:04] Test score for 10000 training labels: 0.8500
[2021-03-22 03:43:04] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:43:04] Found 50000 unlabelled features.
[2021-03-22 03:43:14] Showing labelled: True (14648/50000 visible, 352 redundant)
[2021-03-22 03:43:14] Creating trainer with model on device: cuda
[2021-03-22 03:43:14] Training with expected score >= 0.9900
[2021-03-22 03:43:14] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 03:45:29] Training accuracy: 0.9986
[2021-03-22 03:45:29] Testing on 10000 data points...
[2021-03-22 03:45:31] Test score for 15000 training labels: 0.8730
[2021-03-22 03:45:31] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:45:31] Found 50000 unlabelled features.
[2021-03-22 03:45:40] Showing labelled: True (17953/50000 visible, 2047 redundant)
[2021-03-22 03:45:40] Creating trainer with model on device: cuda
[2021-03-22 03:45:40] Training with expected score >= 0.9900
[2021-03-22 03:45:40] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 03:48:40] Training accuracy: 0.9996
[2021-03-22 03:48:40] Testing on 10000 data points...
[2021-03-22 03:48:42] Test score for 20000 training labels: 0.8841
[2021-03-22 03:48:42] Seeded: 9
[2021-03-22 03:48:42] Running: experiment 3
[2021-03-22 03:48:42] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 03:48:42] Creating pretrained=True VGG16...
[2021-03-22 03:48:44] No parameter reset since we are using a pretrained model.
[2021-03-22 03:48:44] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 03:48:44] Creating trainer with model on device: cuda
[2021-03-22 03:48:44] Training with expected score >= 0.9900
[2021-03-22 03:48:44] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 03:50:36] Training accuracy: 0.9993
[2021-03-22 03:50:36] Testing on 10000 data points...
[2021-03-22 03:50:38] Test score for 5000 training labels: 0.8050
[2021-03-22 03:50:38] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:50:38] Found 50000 unlabelled features.
[2021-03-22 03:50:49] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:51:03] Searching for coresets with 10 beams...
[2021-03-22 03:53:46] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 03:53:46] Creating trainer with model on device: cuda
[2021-03-22 03:53:46] Training with expected score >= 0.9900
[2021-03-22 03:53:46] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 03:55:17] Training accuracy: 0.9953
[2021-03-22 03:55:17] Testing on 10000 data points...
[2021-03-22 03:55:19] Test score for 10000 training labels: 0.8561
[2021-03-22 03:55:19] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 03:55:19] Found 50000 unlabelled features.
[2021-03-22 03:55:31] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 03:55:59] Searching for coresets with 10 beams...
[2021-03-22 03:58:43] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 03:58:43] Creating trainer with model on device: cuda
[2021-03-22 03:58:43] Training with expected score >= 0.9900
[2021-03-22 03:58:43] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 04:00:58] Training accuracy: 0.9976
[2021-03-22 04:00:58] Testing on 10000 data points...
[2021-03-22 04:01:00] Test score for 15000 training labels: 0.8726
[2021-03-22 04:01:00] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:01:00] Found 50000 unlabelled features.
[2021-03-22 04:01:13] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:01:55] Searching for coresets with 10 beams...
[2021-03-22 04:04:38] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 04:04:38] Creating trainer with model on device: cuda
[2021-03-22 04:04:38] Training with expected score >= 0.9900
[2021-03-22 04:04:38] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 04:07:38] Training accuracy: 0.9984
[2021-03-22 04:07:38] Testing on 10000 data points...
[2021-03-22 04:07:40] Test score for 20000 training labels: 0.8876
[2021-03-22 04:07:40] Experiment repeat 5/5
[2021-03-22 04:07:40] Seeded: 9
[2021-03-22 04:07:40] Using 10.00% labels of the dataset (5000/50000)
[2021-03-22 04:07:40] Randomly labelled 5000/50000
[2021-03-22 04:07:40] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 04:07:40] Seeded: 10
[2021-03-22 04:07:40] Running: experiment 0
[2021-03-22 04:07:40] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 04:07:40] Creating pretrained=True VGG16...
[2021-03-22 04:07:42] No parameter reset since we are using a pretrained model.
[2021-03-22 04:07:42] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 04:07:42] Creating trainer with model on device: cuda
[2021-03-22 04:07:42] Training with expected score >= 0.9900
[2021-03-22 04:07:42] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 04:09:34] Training accuracy: 0.9994
[2021-03-22 04:09:34] Testing on 10000 data points...
[2021-03-22 04:09:36] Test score for 5000 training labels: 0.8050
[2021-03-22 04:09:36] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:09:36] Found 50000 unlabelled features.
[2021-03-22 04:09:42] Showing labelled: True (9518/50000 visible, 482 redundant)
[2021-03-22 04:09:42] Creating trainer with model on device: cuda
[2021-03-22 04:09:42] Training with expected score >= 0.9900
[2021-03-22 04:09:42] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 04:11:12] Training accuracy: 0.9980
[2021-03-22 04:11:12] Testing on 10000 data points...
[2021-03-22 04:11:14] Test score for 10000 training labels: 0.8333
[2021-03-22 04:11:14] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:11:14] Found 50000 unlabelled features.
[2021-03-22 04:11:20] Showing labelled: True (13557/50000 visible, 1443 redundant)
[2021-03-22 04:11:20] Creating trainer with model on device: cuda
[2021-03-22 04:11:20] Training with expected score >= 0.9900
[2021-03-22 04:11:20] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 04:13:35] Training accuracy: 0.9967
[2021-03-22 04:13:35] Testing on 10000 data points...
[2021-03-22 04:13:37] Test score for 15000 training labels: 0.8467
[2021-03-22 04:13:37] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:13:37] Found 50000 unlabelled features.
[2021-03-22 04:13:42] Showing labelled: True (17167/50000 visible, 2833 redundant)
[2021-03-22 04:13:42] Creating trainer with model on device: cuda
[2021-03-22 04:13:42] Training with expected score >= 0.9900
[2021-03-22 04:13:42] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 04:16:42] Training accuracy: 0.9991
[2021-03-22 04:16:42] Testing on 10000 data points...
[2021-03-22 04:16:44] Test score for 20000 training labels: 0.8520
[2021-03-22 04:16:44] Seeded: 10
[2021-03-22 04:16:44] Running: experiment 1
[2021-03-22 04:16:44] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 04:16:44] Creating pretrained=True VGG16...
[2021-03-22 04:16:46] No parameter reset since we are using a pretrained model.
[2021-03-22 04:16:46] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 04:16:46] Creating trainer with model on device: cuda
[2021-03-22 04:16:46] Training with expected score >= 0.9900
[2021-03-22 04:16:46] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 04:18:38] Training accuracy: 0.9994
[2021-03-22 04:18:38] Testing on 10000 data points...
[2021-03-22 04:18:40] Test score for 5000 training labels: 0.8082
[2021-03-22 04:18:40] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:18:40] Found 50000 unlabelled features.
[2021-03-22 04:18:51] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:19:05] Searching for coresets greedily...
[2021-03-22 04:19:15] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 04:19:15] Creating trainer with model on device: cuda
[2021-03-22 04:19:15] Training with expected score >= 0.9900
[2021-03-22 04:19:15] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 04:20:45] Training accuracy: 0.9978
[2021-03-22 04:20:45] Testing on 10000 data points...
[2021-03-22 04:20:47] Test score for 10000 training labels: 0.8447
[2021-03-22 04:20:47] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:20:47] Found 50000 unlabelled features.
[2021-03-22 04:20:58] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:21:26] Searching for coresets greedily...
[2021-03-22 04:21:36] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 04:21:36] Creating trainer with model on device: cuda
[2021-03-22 04:21:36] Training with expected score >= 0.9900
[2021-03-22 04:21:36] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 04:23:51] Training accuracy: 0.9993
[2021-03-22 04:23:51] Testing on 10000 data points...
[2021-03-22 04:23:53] Test score for 15000 training labels: 0.8673
[2021-03-22 04:23:53] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:23:53] Found 50000 unlabelled features.
[2021-03-22 04:24:06] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:24:48] Searching for coresets greedily...
[2021-03-22 04:24:58] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 04:24:58] Creating trainer with model on device: cuda
[2021-03-22 04:24:58] Training with expected score >= 0.9900
[2021-03-22 04:24:58] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 04:27:58] Training accuracy: 0.9983
[2021-03-22 04:27:58] Testing on 10000 data points...
[2021-03-22 04:28:00] Test score for 20000 training labels: 0.8782
[2021-03-22 04:28:00] Seeded: 10
[2021-03-22 04:28:00] Running: experiment 2
[2021-03-22 04:28:00] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 04:28:00] Creating pretrained=True VGG16...
[2021-03-22 04:28:01] No parameter reset since we are using a pretrained model.
[2021-03-22 04:28:01] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 04:28:01] Creating trainer with model on device: cuda
[2021-03-22 04:28:01] Training with expected score >= 0.9900
[2021-03-22 04:28:01] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 04:29:54] Training accuracy: 0.9946
[2021-03-22 04:29:54] Testing on 10000 data points...
[2021-03-22 04:29:56] Test score for 5000 training labels: 0.8016
[2021-03-22 04:29:56] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:29:56] Found 50000 unlabelled features.
[2021-03-22 04:30:05] Showing labelled: True (9955/50000 visible, 45 redundant)
[2021-03-22 04:30:05] Creating trainer with model on device: cuda
[2021-03-22 04:30:05] Training with expected score >= 0.9900
[2021-03-22 04:30:05] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 04:31:35] Training accuracy: 0.9941
[2021-03-22 04:31:35] Testing on 10000 data points...
[2021-03-22 04:31:37] Test score for 10000 training labels: 0.8537
[2021-03-22 04:31:37] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:31:37] Found 50000 unlabelled features.
[2021-03-22 04:31:47] Showing labelled: True (14482/50000 visible, 518 redundant)
[2021-03-22 04:31:47] Creating trainer with model on device: cuda
[2021-03-22 04:31:47] Training with expected score >= 0.9900
[2021-03-22 04:31:47] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 04:34:02] Training accuracy: 0.9994
[2021-03-22 04:34:02] Testing on 10000 data points...
[2021-03-22 04:34:04] Test score for 15000 training labels: 0.8743
[2021-03-22 04:34:04] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:34:04] Found 50000 unlabelled features.
[2021-03-22 04:34:14] Showing labelled: True (17931/50000 visible, 2069 redundant)
[2021-03-22 04:34:14] Creating trainer with model on device: cuda
[2021-03-22 04:34:14] Training with expected score >= 0.9900
[2021-03-22 04:34:14] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 04:37:14] Training accuracy: 0.9990
[2021-03-22 04:37:14] Testing on 10000 data points...
[2021-03-22 04:37:15] Test score for 20000 training labels: 0.8829
[2021-03-22 04:37:15] Seeded: 10
[2021-03-22 04:37:15] Running: experiment 3
[2021-03-22 04:37:15] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-22 04:37:15] Creating pretrained=True VGG16...
[2021-03-22 04:37:17] No parameter reset since we are using a pretrained model.
[2021-03-22 04:37:17] vgg-pretrained: initialized 15245130 parameters.
[2021-03-22 04:37:17] Creating trainer with model on device: cuda
[2021-03-22 04:37:17] Training with expected score >= 0.9900
[2021-03-22 04:37:17] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-22 04:39:10] Training accuracy: 0.9995
[2021-03-22 04:39:10] Testing on 10000 data points...
[2021-03-22 04:39:11] Test score for 5000 training labels: 0.8082
[2021-03-22 04:39:12] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:39:12] Found 50000 unlabelled features.
[2021-03-22 04:39:22] Computing distance between 5000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:39:36] Searching for coresets with 10 beams...
[2021-03-22 04:42:20] Showing labelled: True (10000/50000 visible, 0 redundant)
[2021-03-22 04:42:20] Creating trainer with model on device: cuda
[2021-03-22 04:42:20] Training with expected score >= 0.9900
[2021-03-22 04:42:20] Training vgg-pretrained across 10000 data points in cifar10...
[2021-03-22 04:43:50] Training accuracy: 0.9952
[2021-03-22 04:43:50] Testing on 10000 data points...
[2021-03-22 04:43:52] Test score for 10000 training labels: 0.8569
[2021-03-22 04:43:52] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:43:52] Found 50000 unlabelled features.
[2021-03-22 04:44:04] Computing distance between 10000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:44:32] Searching for coresets with 10 beams...
[2021-03-22 04:47:16] Showing labelled: True (15000/50000 visible, 0 redundant)
[2021-03-22 04:47:16] Creating trainer with model on device: cuda
[2021-03-22 04:47:16] Training with expected score >= 0.9900
[2021-03-22 04:47:16] Training vgg-pretrained across 15000 data points in cifar10...
[2021-03-22 04:49:31] Training accuracy: 0.9982
[2021-03-22 04:49:31] Testing on 10000 data points...
[2021-03-22 04:49:33] Test score for 15000 training labels: 0.8768
[2021-03-22 04:49:33] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-22 04:49:33] Found 50000 unlabelled features.
[2021-03-22 04:49:46] Computing distance between 15000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-22 04:50:29] Searching for coresets with 10 beams...
[2021-03-22 04:53:12] Showing labelled: True (20000/50000 visible, 0 redundant)
[2021-03-22 04:53:12] Creating trainer with model on device: cuda
[2021-03-22 04:53:12] Training with expected score >= 0.9900
[2021-03-22 04:53:12] Training vgg-pretrained across 20000 data points in cifar10...
[2021-03-22 04:56:13] Training accuracy: 0.9984
[2021-03-22 04:56:13] Testing on 10000 data points...
[2021-03-22 04:56:15] Test score for 20000 training labels: 0.8870
[2021-03-22 04:56:15] Updated results: ../results/cifar10/main/results.json
