[2021-03-17 14:51:25] Seeded: 5
[2021-03-17 14:51:25] Created experiment 0:
[2021-03-17 14:51:25]  - Model: simple-cnn
[2021-03-17 14:51:25]  - Acquisition function: lc-beam-coreset
[2021-03-17 14:51:25] Created experiment 1:
[2021-03-17 14:51:25]  - Model: simple-cnn
[2021-03-17 14:51:25]  - Acquisition function: greedy-coreset
[2021-03-17 14:51:25] Created experiment 2:
[2021-03-17 14:51:25]  - Model: simple-cnn
[2021-03-17 14:51:25]  - Acquisition function: random
[2021-03-17 14:51:25] Created experiment 3:
[2021-03-17 14:51:25]  - Model: simple-cnn
[2021-03-17 14:51:25]  - Acquisition function: least-confidence
[2021-03-17 14:51:25] Loading mnist test set...
[2021-03-17 14:51:25] Experiment repeat 1/1
[2021-03-17 14:51:25] Using 0.33% labels of the dataset (200/60000)
[2021-03-17 14:51:25] Randomly labelled 200/60000
[2021-03-17 14:51:25] Showing labelled: True (200/60000 visible)
[2021-03-17 14:51:25] Running: experiment 0
[2021-03-17 14:51:25] Showing labelled: True (200/60000 visible)
[2021-03-17 14:51:25] simple-cnn: initialized 98314 parameters.
[2021-03-17 14:51:25] Creating trainer with model on device: cuda
[2021-03-17 14:51:30] Training simple-cnn across 200 data points in mnist...
[2021-03-17 14:51:48] Training accuracy: 0.7759
[2021-03-17 14:51:48] Testing on 10000 data points...
[2021-03-17 14:51:49] Test score for 200 training labels: 0.6552
[2021-03-17 14:51:49] Showing labelled: False (59800/60000 visible)
[2021-03-17 14:51:49] Found 59800 unlabelled features.
[2021-03-17 14:51:56] Computing distance between 200 labelled and 59800 unlabelled vectors of length 64...
[2021-03-17 14:52:02] Searching for coresets with 20 beams...
[2021-03-17 14:53:48] Showing labelled: True (400/60000 visible)
[2021-03-17 14:53:48] Training simple-cnn across 400 data points in mnist...
[2021-03-17 14:54:22] Training accuracy: 0.9965
[2021-03-17 14:54:22] Testing on 10000 data points...
[2021-03-17 14:54:23] Test score for 400 training labels: 0.8110
[2021-03-17 14:54:23] Showing labelled: False (59600/60000 visible)
[2021-03-17 14:54:23] Found 59600 unlabelled features.
[2021-03-17 14:54:29] Computing distance between 400 labelled and 59600 unlabelled vectors of length 64...
[2021-03-17 14:54:37] Searching for coresets with 20 beams...
[2021-03-17 14:56:23] Showing labelled: True (600/60000 visible)
[2021-03-17 14:56:23] Training simple-cnn across 600 data points in mnist...
[2021-03-17 14:57:14] Training accuracy: 0.9985
[2021-03-17 14:57:14] Testing on 10000 data points...
[2021-03-17 14:57:15] Test score for 600 training labels: 0.8905
[2021-03-17 14:57:15] Showing labelled: False (59400/60000 visible)
[2021-03-17 14:57:15] Found 59400 unlabelled features.
[2021-03-17 14:57:21] Computing distance between 600 labelled and 59400 unlabelled vectors of length 64...
[2021-03-17 14:57:37] Searching for coresets with 20 beams...
[2021-03-17 14:59:22] Showing labelled: True (800/60000 visible)
[2021-03-17 14:59:22] Training simple-cnn across 800 data points in mnist...
[2021-03-17 15:00:31] Training accuracy: 0.9988
[2021-03-17 15:00:31] Testing on 10000 data points...
[2021-03-17 15:00:32] Test score for 800 training labels: 0.9252
[2021-03-17 15:00:32] Showing labelled: False (59200/60000 visible)
[2021-03-17 15:00:32] Found 59200 unlabelled features.
[2021-03-17 15:00:38] Computing distance between 800 labelled and 59200 unlabelled vectors of length 64...
[2021-03-17 15:00:54] Searching for coresets with 20 beams...
[2021-03-17 15:02:38] Showing labelled: True (1000/60000 visible)
[2021-03-17 15:02:38] Training simple-cnn across 1000 data points in mnist...
[2021-03-17 15:04:04] Training accuracy: 0.9985
[2021-03-17 15:04:04] Testing on 10000 data points...
[2021-03-17 15:04:05] Test score for 1000 training labels: 0.9425
[2021-03-17 15:04:05] Running: experiment 1
[2021-03-17 15:04:05] Showing labelled: True (200/60000 visible)
[2021-03-17 15:04:05] simple-cnn: initialized 98314 parameters.
[2021-03-17 15:04:05] Creating trainer with model on device: cuda
[2021-03-17 15:04:05] Training simple-cnn across 200 data points in mnist...
[2021-03-17 15:04:23] Training accuracy: 0.7855
[2021-03-17 15:04:23] Testing on 10000 data points...
[2021-03-17 15:04:24] Test score for 200 training labels: 0.6673
[2021-03-17 15:04:24] Showing labelled: False (59800/60000 visible)
[2021-03-17 15:04:24] Found 59800 unlabelled features.
[2021-03-17 15:04:30] Computing distance between 200 labelled and 59800 unlabelled vectors of length 64...
[2021-03-17 15:04:35] Searching for coresets greedily...
[2021-03-17 15:04:36] Showing labelled: True (400/60000 visible)
[2021-03-17 15:04:36] Training simple-cnn across 400 data points in mnist...
[2021-03-17 15:05:11] Training accuracy: 0.9966
[2021-03-17 15:05:11] Testing on 10000 data points...
[2021-03-17 15:05:12] Test score for 400 training labels: 0.8321
[2021-03-17 15:05:12] Showing labelled: False (59600/60000 visible)
[2021-03-17 15:05:12] Found 59600 unlabelled features.
[2021-03-17 15:05:18] Computing distance between 400 labelled and 59600 unlabelled vectors of length 64...
[2021-03-17 15:05:29] Searching for coresets greedily...
[2021-03-17 15:05:31] Showing labelled: True (600/60000 visible)
[2021-03-17 15:05:31] Training simple-cnn across 600 data points in mnist...
[2021-03-17 15:06:25] Training accuracy: 0.9985
[2021-03-17 15:06:25] Testing on 10000 data points...
[2021-03-17 15:06:26] Test score for 600 training labels: 0.8825
[2021-03-17 15:06:26] Showing labelled: False (59400/60000 visible)
[2021-03-17 15:06:26] Found 59400 unlabelled features.
[2021-03-17 15:06:33] Computing distance between 600 labelled and 59400 unlabelled vectors of length 64...
[2021-03-17 15:06:47] Searching for coresets greedily...
[2021-03-17 15:06:49] Showing labelled: True (800/60000 visible)
[2021-03-17 15:06:49] Training simple-cnn across 800 data points in mnist...
[2021-03-17 15:07:59] Training accuracy: 0.9989
[2021-03-17 15:07:59] Testing on 10000 data points...
[2021-03-17 15:08:00] Test score for 800 training labels: 0.9122
[2021-03-17 15:08:00] Showing labelled: False (59200/60000 visible)
[2021-03-17 15:08:00] Found 59200 unlabelled features.
[2021-03-17 15:08:07] Computing distance between 800 labelled and 59200 unlabelled vectors of length 64...
[2021-03-17 15:08:28] Searching for coresets greedily...
[2021-03-17 15:08:30] Showing labelled: True (1000/60000 visible)
[2021-03-17 15:08:30] Training simple-cnn across 1000 data points in mnist...
[2021-03-17 15:09:58] Training accuracy: 0.9985
[2021-03-17 15:09:58] Testing on 10000 data points...
[2021-03-17 15:09:59] Test score for 1000 training labels: 0.9406
[2021-03-17 15:09:59] Running: experiment 2
[2021-03-17 15:09:59] Showing labelled: True (200/60000 visible)
[2021-03-17 15:09:59] simple-cnn: initialized 98314 parameters.
[2021-03-17 15:09:59] Creating trainer with model on device: cuda
[2021-03-17 15:09:59] Training simple-cnn across 200 data points in mnist...
[2021-03-17 15:10:19] Training accuracy: 0.7728
[2021-03-17 15:10:19] Testing on 10000 data points...
[2021-03-17 15:10:20] Test score for 200 training labels: 0.6858
[2021-03-17 15:10:20] Showing labelled: False (59800/60000 visible)
[2021-03-17 15:10:20] Found 59800 unlabelled features.
[2021-03-17 15:10:25] Showing labelled: True (400/60000 visible)
[2021-03-17 15:10:25] Training simple-cnn across 400 data points in mnist...
[2021-03-17 15:11:00] Training accuracy: 0.9988
[2021-03-17 15:11:00] Testing on 10000 data points...
[2021-03-17 15:11:01] Test score for 400 training labels: 0.8613
[2021-03-17 15:11:01] Showing labelled: False (59600/60000 visible)
[2021-03-17 15:11:01] Found 59600 unlabelled features.
[2021-03-17 15:11:07] Showing labelled: True (600/60000 visible)
[2021-03-17 15:11:07] Training simple-cnn across 600 data points in mnist...
[2021-03-17 15:11:59] Training accuracy: 0.9992
[2021-03-17 15:11:59] Testing on 10000 data points...
[2021-03-17 15:12:00] Test score for 600 training labels: 0.8905
[2021-03-17 15:12:00] Showing labelled: False (59400/60000 visible)
[2021-03-17 15:12:00] Found 59400 unlabelled features.
[2021-03-17 15:12:06] Showing labelled: True (800/60000 visible)
[2021-03-17 15:12:06] Training simple-cnn across 800 data points in mnist...
[2021-03-17 15:13:17] Training accuracy: 0.9997
[2021-03-17 15:13:17] Testing on 10000 data points...
[2021-03-17 15:13:18] Test score for 800 training labels: 0.9157
[2021-03-17 15:13:18] Showing labelled: False (59200/60000 visible)
[2021-03-17 15:13:18] Found 59200 unlabelled features.
[2021-03-17 15:13:23] Showing labelled: True (1000/60000 visible)
[2021-03-17 15:13:23] Training simple-cnn across 1000 data points in mnist...
[2021-03-17 15:14:53] Training accuracy: 0.9990
[2021-03-17 15:14:53] Testing on 10000 data points...
[2021-03-17 15:14:54] Test score for 1000 training labels: 0.9349
[2021-03-17 15:14:54] Running: experiment 3
[2021-03-17 15:14:54] Showing labelled: True (200/60000 visible)
[2021-03-17 15:14:54] simple-cnn: initialized 98314 parameters.
[2021-03-17 15:14:54] Creating trainer with model on device: cuda
[2021-03-17 15:14:54] Training simple-cnn across 200 data points in mnist...
[2021-03-17 15:15:12] Training accuracy: 0.7608
[2021-03-17 15:15:12] Testing on 10000 data points...
[2021-03-17 15:15:13] Test score for 200 training labels: 0.6980
[2021-03-17 15:15:13] Showing labelled: False (59800/60000 visible)
[2021-03-17 15:15:13] Found 59800 unlabelled features.
[2021-03-17 15:15:19] Showing labelled: True (400/60000 visible)
[2021-03-17 15:15:19] Training simple-cnn across 400 data points in mnist...
[2021-03-17 15:15:56] Training accuracy: 0.9976
[2021-03-17 15:15:56] Testing on 10000 data points...
[2021-03-17 15:15:57] Test score for 400 training labels: 0.8631
[2021-03-17 15:15:57] Showing labelled: False (59600/60000 visible)
[2021-03-17 15:15:57] Found 59600 unlabelled features.
[2021-03-17 15:16:03] Showing labelled: True (600/60000 visible)
[2021-03-17 15:16:03] Training simple-cnn across 600 data points in mnist...
[2021-03-17 15:16:55] Training accuracy: 0.9980
[2021-03-17 15:16:55] Testing on 10000 data points...
[2021-03-17 15:16:56] Test score for 600 training labels: 0.9173
[2021-03-17 15:16:56] Showing labelled: False (59400/60000 visible)
[2021-03-17 15:16:56] Found 59400 unlabelled features.
[2021-03-17 15:17:03] Showing labelled: True (800/60000 visible)
[2021-03-17 15:17:03] Training simple-cnn across 800 data points in mnist...
[2021-03-17 15:18:14] Training accuracy: 0.9984
[2021-03-17 15:18:14] Testing on 10000 data points...
[2021-03-17 15:18:15] Test score for 800 training labels: 0.9413
[2021-03-17 15:18:15] Showing labelled: False (59200/60000 visible)
[2021-03-17 15:18:15] Found 59200 unlabelled features.
[2021-03-17 15:18:21] Showing labelled: True (1000/60000 visible)
[2021-03-17 15:18:21] Training simple-cnn across 1000 data points in mnist...
[2021-03-17 15:19:48] Training accuracy: 0.9979
[2021-03-17 15:19:48] Testing on 10000 data points...
[2021-03-17 15:19:49] Test score for 1000 training labels: 0.9580
[2021-03-17 15:19:49] Updated results: ../results/mnist/lc_beam_coreset_crossentropy_beam20/results.json
