[2021-03-21 00:43:19] Created experiment 0:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: random
[2021-03-21 00:43:19] Created experiment 1:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: greedy-coreset
[2021-03-21 00:43:19] Created experiment 2:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: least-confidence
[2021-03-21 00:43:19] Created experiment 3:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: lc-beam-coreset
[2021-03-21 00:43:19] Created experiment 4:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: lc-beam-coreset
[2021-03-21 00:43:19] Created experiment 5:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: lc-beam-pweighted-coreset
[2021-03-21 00:43:19] Created experiment 6:
[2021-03-21 00:43:19]  - Model: vgg-pretrained
[2021-03-21 00:43:19]  - Acquisition function: lc-beam-pweighted-relconf-optimcore
[2021-03-21 00:43:19] Loading cifar10 test set...
[2021-03-21 00:43:20] Experiment repeat 1/1
[2021-03-21 00:43:20] Seeded: 5
[2021-03-21 00:43:20] Using 2.00% labels of the dataset (1000/50000)
[2021-03-21 00:43:20] Randomly labelled 1000/50000
[2021-03-21 00:43:20] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:43:20] Seeded: 6
[2021-03-21 00:43:20] Running: experiment 0
[2021-03-21 00:43:20] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:43:20] Creating pretrained=True VGG16...
[2021-03-21 00:43:22] No parameter reset since we are using a pretrained model.
[2021-03-21 00:43:22] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:43:22] Creating trainer with model on device: cuda
[2021-03-21 00:43:27] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:43:27] Training accuracy: 0.0992
[2021-03-21 00:43:27] Testing on 10000 data points...
[2021-03-21 00:43:29] Test score for 1000 training labels: 0.0907
[2021-03-21 00:43:29] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:43:29] Found 50000 unlabelled features.
[2021-03-21 00:43:35] Showing labelled: True (1382/50000 visible, 18 redundant)
[2021-03-21 00:43:35] Creating trainer with model on device: cuda
[2021-03-21 00:43:35] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:43:36] Training accuracy: 0.0735
[2021-03-21 00:43:36] Testing on 10000 data points...
[2021-03-21 00:43:38] Test score for 1400 training labels: 0.0916
[2021-03-21 00:43:38] Seeded: 6
[2021-03-21 00:43:38] Running: experiment 1
[2021-03-21 00:43:38] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:43:38] Creating pretrained=True VGG16...
[2021-03-21 00:43:39] No parameter reset since we are using a pretrained model.
[2021-03-21 00:43:39] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:43:39] Creating trainer with model on device: cuda
[2021-03-21 00:43:39] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:43:40] Training accuracy: 0.0992
[2021-03-21 00:43:40] Testing on 10000 data points...
[2021-03-21 00:43:42] Test score for 1000 training labels: 0.0907
[2021-03-21 00:43:42] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:43:42] Found 50000 unlabelled features.
[2021-03-21 00:43:52] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 00:43:55] Searching for coresets greedily...
[2021-03-21 00:43:55] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 00:43:55] Creating trainer with model on device: cuda
[2021-03-21 00:43:55] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:43:56] Training accuracy: 0.0845
[2021-03-21 00:43:56] Testing on 10000 data points...
[2021-03-21 00:43:58] Test score for 1400 training labels: 0.0921
[2021-03-21 00:43:58] Seeded: 6
[2021-03-21 00:43:58] Running: experiment 2
[2021-03-21 00:43:58] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:43:58] Creating pretrained=True VGG16...
[2021-03-21 00:44:00] No parameter reset since we are using a pretrained model.
[2021-03-21 00:44:00] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:44:00] Creating trainer with model on device: cuda
[2021-03-21 00:44:00] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:44:01] Training accuracy: 0.0992
[2021-03-21 00:44:01] Testing on 10000 data points...
[2021-03-21 00:44:03] Test score for 1000 training labels: 0.0907
[2021-03-21 00:44:03] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:44:03] Found 50000 unlabelled features.
[2021-03-21 00:44:12] Showing labelled: True (1386/50000 visible, 14 redundant)
[2021-03-21 00:44:12] Creating trainer with model on device: cuda
[2021-03-21 00:44:12] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:44:13] Training accuracy: 0.0918
[2021-03-21 00:44:13] Testing on 10000 data points...
[2021-03-21 00:44:15] Test score for 1400 training labels: 0.0914
[2021-03-21 00:44:15] Seeded: 6
[2021-03-21 00:44:15] Running: experiment 3
[2021-03-21 00:44:15] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:44:15] Creating pretrained=True VGG16...
[2021-03-21 00:44:17] No parameter reset since we are using a pretrained model.
[2021-03-21 00:44:17] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:44:17] Creating trainer with model on device: cuda
[2021-03-21 00:44:17] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:44:17] Training accuracy: 0.0992
[2021-03-21 00:44:17] Testing on 10000 data points...
[2021-03-21 00:44:19] Test score for 1000 training labels: 0.0907
[2021-03-21 00:44:19] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:44:19] Found 50000 unlabelled features.
[2021-03-21 00:44:29] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 00:44:32] Searching for coresets with 1 beams...
[2021-03-21 00:44:33] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 00:44:33] Creating trainer with model on device: cuda
[2021-03-21 00:44:33] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:44:34] Training accuracy: 0.0845
[2021-03-21 00:44:34] Testing on 10000 data points...
[2021-03-21 00:44:36] Test score for 1400 training labels: 0.0921
[2021-03-21 00:44:36] Seeded: 6
[2021-03-21 00:44:36] Running: experiment 4
[2021-03-21 00:44:36] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:44:36] Creating pretrained=True VGG16...
[2021-03-21 00:44:38] No parameter reset since we are using a pretrained model.
[2021-03-21 00:44:38] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:44:38] Creating trainer with model on device: cuda
[2021-03-21 00:44:38] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:44:38] Training accuracy: 0.0992
[2021-03-21 00:44:38] Testing on 10000 data points...
[2021-03-21 00:44:40] Test score for 1000 training labels: 0.0907
[2021-03-21 00:44:40] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:44:40] Found 50000 unlabelled features.
[2021-03-21 00:44:50] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 00:44:53] Searching for coresets with 20 beams...
[2021-03-21 00:45:22] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 00:45:22] Creating trainer with model on device: cuda
[2021-03-21 00:45:22] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:45:23] Training accuracy: 0.0751
[2021-03-21 00:45:23] Testing on 10000 data points...
[2021-03-21 00:45:25] Test score for 1400 training labels: 0.0918
[2021-03-21 00:45:25] Seeded: 6
[2021-03-21 00:45:25] Running: experiment 5
[2021-03-21 00:45:25] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:45:25] Creating pretrained=True VGG16...
[2021-03-21 00:45:26] No parameter reset since we are using a pretrained model.
[2021-03-21 00:45:26] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:45:26] Creating trainer with model on device: cuda
[2021-03-21 00:45:26] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:45:27] Training accuracy: 0.0992
[2021-03-21 00:45:27] Testing on 10000 data points...
[2021-03-21 00:45:29] Test score for 1000 training labels: 0.0907
[2021-03-21 00:45:29] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:45:29] Found 50000 unlabelled features.
[2021-03-21 00:45:39] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 00:45:41] Searching for coresets with 2 beams...
[2021-03-21 00:45:44] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 00:45:44] Creating trainer with model on device: cuda
[2021-03-21 00:45:44] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:45:45] Training accuracy: 0.0864
[2021-03-21 00:45:45] Testing on 10000 data points...
[2021-03-21 00:45:47] Test score for 1400 training labels: 0.0916
[2021-03-21 00:45:47] Seeded: 6
[2021-03-21 00:45:47] Running: experiment 6
[2021-03-21 00:45:47] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 00:45:47] Creating pretrained=True VGG16...
[2021-03-21 00:45:49] No parameter reset since we are using a pretrained model.
[2021-03-21 00:45:49] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 00:45:49] Creating trainer with model on device: cuda
[2021-03-21 00:45:49] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 00:45:50] Training accuracy: 0.0992
[2021-03-21 00:45:50] Testing on 10000 data points...
[2021-03-21 00:45:52] Test score for 1000 training labels: 0.0907
[2021-03-21 00:45:52] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 00:45:52] Found 50000 unlabelled features.
[2021-03-21 00:46:01] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 00:46:04] Searching for coresets with 2 beams...
[2021-03-21 00:46:07] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 00:46:07] Creating trainer with model on device: cuda
[2021-03-21 00:46:07] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 00:46:08] Training accuracy: 0.0848
[2021-03-21 00:46:08] Testing on 10000 data points...
[2021-03-21 00:46:10] Test score for 1400 training labels: 0.0919
[2021-03-21 00:46:10] Updated results: ../results/cifar10/main-test/results.json
