[2021-03-21 17:55:29] Created experiment 0:
[2021-03-21 17:55:29]  - Model: vgg-pt
[2021-03-21 17:55:29]  - Acquisition function: lc-beam-pweighted-coreset (beams=20)
[2021-03-21 17:55:29] Loading cifar10 test set...
[2021-03-21 17:55:30] Experiment repeat 1/1
[2021-03-21 17:55:30] Seeded: 9
[2021-03-21 17:55:30] Using 2.00% labels of the dataset (1000/50000)
[2021-03-21 17:55:31] Randomly labelled 1000/50000
[2021-03-21 17:55:31] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 17:55:31] Seeded: 10
[2021-03-21 17:55:31] Running: experiment 0
[2021-03-21 17:55:31] Showing labelled: True (1000/50000 visible, 0 redundant)
[2021-03-21 17:55:31] Creating pretrained=True VGG16...
[2021-03-21 17:55:32] No parameter reset since we are using a pretrained model.
[2021-03-21 17:55:32] vgg-pretrained: initialized 15245130 parameters.
[2021-03-21 17:55:32] Creating trainer with model on device: cuda
[2021-03-21 17:55:37] Training with expected score >= 0.9900
[2021-03-21 17:55:37] Training vgg-pretrained across 1000 data points in cifar10...
[2021-03-21 17:56:36] Training accuracy: 1.0000
[2021-03-21 17:56:36] Testing on 10000 data points...
[2021-03-21 17:56:38] Test score for 1000 training labels: 0.7335
[2021-03-21 17:56:38] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 17:56:38] Found 50000 unlabelled features.
[2021-03-21 17:56:49] Computing distance between 1000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 17:56:52] Searching for coresets with 20 beams...
[2021-03-21 17:57:21] Showing labelled: True (1400/50000 visible, 0 redundant)
[2021-03-21 17:57:21] Creating trainer with model on device: cuda
[2021-03-21 17:57:21] Training with expected score >= 0.9900
[2021-03-21 17:57:21] Training vgg-pretrained across 1400 data points in cifar10...
[2021-03-21 17:57:41] Training accuracy: 0.9927
[2021-03-21 17:57:41] Testing on 10000 data points...
[2021-03-21 17:57:43] Test score for 1400 training labels: 0.7414
[2021-03-21 17:57:43] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 17:57:43] Found 50000 unlabelled features.
[2021-03-21 17:57:54] Computing distance between 1400 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 17:57:59] Searching for coresets with 20 beams...
[2021-03-21 17:58:27] Showing labelled: True (1800/50000 visible, 0 redundant)
[2021-03-21 17:58:27] Creating trainer with model on device: cuda
[2021-03-21 17:58:27] Training with expected score >= 0.9900
[2021-03-21 17:58:27] Training vgg-pretrained across 1800 data points in cifar10...
[2021-03-21 17:58:54] Training accuracy: 0.9986
[2021-03-21 17:58:54] Testing on 10000 data points...
[2021-03-21 17:58:56] Test score for 1800 training labels: 0.7670
[2021-03-21 17:58:56] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 17:58:56] Found 50000 unlabelled features.
[2021-03-21 17:59:06] Computing distance between 1800 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 17:59:12] Searching for coresets with 20 beams...
[2021-03-21 17:59:41] Showing labelled: True (2200/50000 visible, 0 redundant)
[2021-03-21 17:59:41] Creating trainer with model on device: cuda
[2021-03-21 17:59:41] Training with expected score >= 0.9900
[2021-03-21 17:59:41] Training vgg-pretrained across 2200 data points in cifar10...
[2021-03-21 18:00:15] Training accuracy: 0.9985
[2021-03-21 18:00:15] Testing on 10000 data points...
[2021-03-21 18:00:17] Test score for 2200 training labels: 0.7779
[2021-03-21 18:00:17] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:00:17] Found 50000 unlabelled features.
[2021-03-21 18:00:27] Computing distance between 2200 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:00:33] Searching for coresets with 20 beams...
[2021-03-21 18:01:02] Showing labelled: True (2600/50000 visible, 0 redundant)
[2021-03-21 18:01:02] Creating trainer with model on device: cuda
[2021-03-21 18:01:02] Training with expected score >= 0.9900
[2021-03-21 18:01:02] Training vgg-pretrained across 2600 data points in cifar10...
[2021-03-21 18:01:41] Training accuracy: 0.9992
[2021-03-21 18:01:41] Testing on 10000 data points...
[2021-03-21 18:01:43] Test score for 2600 training labels: 0.7941
[2021-03-21 18:01:43] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:01:43] Found 50000 unlabelled features.
[2021-03-21 18:01:54] Computing distance between 2600 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:02:02] Searching for coresets with 20 beams...
[2021-03-21 18:02:31] Showing labelled: True (3000/50000 visible, 0 redundant)
[2021-03-21 18:02:31] Creating trainer with model on device: cuda
[2021-03-21 18:02:31] Training with expected score >= 0.9900
[2021-03-21 18:02:31] Training vgg-pretrained across 3000 data points in cifar10...
[2021-03-21 18:03:16] Training accuracy: 0.9942
[2021-03-21 18:03:16] Testing on 10000 data points...
[2021-03-21 18:03:18] Test score for 3000 training labels: 0.7926
[2021-03-21 18:03:18] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:03:18] Found 50000 unlabelled features.
[2021-03-21 18:03:29] Computing distance between 3000 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:03:38] Searching for coresets with 20 beams...
[2021-03-21 18:04:07] Showing labelled: True (3400/50000 visible, 0 redundant)
[2021-03-21 18:04:07] Creating trainer with model on device: cuda
[2021-03-21 18:04:07] Training with expected score >= 0.9900
[2021-03-21 18:04:07] Training vgg-pretrained across 3400 data points in cifar10...
[2021-03-21 18:04:59] Training accuracy: 0.9975
[2021-03-21 18:04:59] Testing on 10000 data points...
[2021-03-21 18:05:02] Test score for 3400 training labels: 0.7913
[2021-03-21 18:05:02] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:05:02] Found 50000 unlabelled features.
[2021-03-21 18:05:13] Computing distance between 3400 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:05:23] Searching for coresets with 20 beams...
[2021-03-21 18:05:52] Showing labelled: True (3800/50000 visible, 0 redundant)
[2021-03-21 18:05:52] Creating trainer with model on device: cuda
[2021-03-21 18:05:52] Training with expected score >= 0.9900
[2021-03-21 18:05:52] Training vgg-pretrained across 3800 data points in cifar10...
[2021-03-21 18:06:51] Training accuracy: 0.9994
[2021-03-21 18:06:51] Testing on 10000 data points...
[2021-03-21 18:06:53] Test score for 3800 training labels: 0.8093
[2021-03-21 18:06:53] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:06:53] Found 50000 unlabelled features.
[2021-03-21 18:07:04] Computing distance between 3800 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:07:15] Searching for coresets with 20 beams...
[2021-03-21 18:07:44] Showing labelled: True (4200/50000 visible, 0 redundant)
[2021-03-21 18:07:44] Creating trainer with model on device: cuda
[2021-03-21 18:07:44] Training with expected score >= 0.9900
[2021-03-21 18:07:44] Training vgg-pretrained across 4200 data points in cifar10...
[2021-03-21 18:08:47] Training accuracy: 0.9999
[2021-03-21 18:08:47] Testing on 10000 data points...
[2021-03-21 18:08:49] Test score for 4200 training labels: 0.8184
[2021-03-21 18:08:49] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:08:49] Found 50000 unlabelled features.
[2021-03-21 18:09:01] Computing distance between 4200 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:09:13] Searching for coresets with 20 beams...
[2021-03-21 18:09:42] Showing labelled: True (4600/50000 visible, 0 redundant)
[2021-03-21 18:09:42] Creating trainer with model on device: cuda
[2021-03-21 18:09:42] Training with expected score >= 0.9900
[2021-03-21 18:09:42] Training vgg-pretrained across 4600 data points in cifar10...
[2021-03-21 18:10:50] Training accuracy: 1.0000
[2021-03-21 18:10:50] Testing on 10000 data points...
[2021-03-21 18:10:53] Test score for 4600 training labels: 0.8314
[2021-03-21 18:10:53] Showing labelled: False (50000/50000 visible, 0 redundant)
[2021-03-21 18:10:53] Found 50000 unlabelled features.
[2021-03-21 18:11:04] Computing distance between 4600 labelled and 50000 unlabelled vectors of length 512...
[2021-03-21 18:11:17] Searching for coresets with 20 beams...
[2021-03-21 18:11:46] Showing labelled: True (5000/50000 visible, 0 redundant)
[2021-03-21 18:11:46] Creating trainer with model on device: cuda
[2021-03-21 18:11:46] Training with expected score >= 0.9900
[2021-03-21 18:11:46] Training vgg-pretrained across 5000 data points in cifar10...
[2021-03-21 18:13:02] Training accuracy: 0.9999
[2021-03-21 18:13:02] Testing on 10000 data points...
[2021-03-21 18:13:04] Test score for 5000 training labels: 0.8333
[2021-03-21 18:13:04] Updated results: ../results/cifar10/main_redo_seed9_lc_beam_pweighted_coreset/results.json
