SystemyDialogowe/slot-model/training.log
2022-05-18 00:06:14 +02:00

187 lines
14 KiB
Plaintext

2022-05-17 23:52:57,432 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,433 Model: "SequenceTagger(
(embeddings): StackedEmbeddings(
(list_embedding_0): WordEmbeddings('pl')
(list_embedding_1): FlairEmbeddings(
(lm): LanguageModel(
(drop): Dropout(p=0.25, inplace=False)
(encoder): Embedding(1602, 100)
(rnn): LSTM(100, 2048)
(decoder): Linear(in_features=2048, out_features=1602, bias=True)
)
)
(list_embedding_2): FlairEmbeddings(
(lm): LanguageModel(
(drop): Dropout(p=0.25, inplace=False)
(encoder): Embedding(1602, 100)
(rnn): LSTM(100, 2048)
(decoder): Linear(in_features=2048, out_features=1602, bias=True)
)
)
(list_embedding_3): CharacterEmbeddings(
(char_embedding): Embedding(275, 25)
(char_rnn): LSTM(25, 25, bidirectional=True)
)
)
(word_dropout): WordDropout(p=0.05)
(locked_dropout): LockedDropout(p=0.5)
(embedding2nn): Linear(in_features=4446, out_features=4446, bias=True)
(rnn): LSTM(4446, 256, batch_first=True, bidirectional=True)
(linear): Linear(in_features=512, out_features=12, bias=True)
(beta): 1.0
(weights): None
(weight_tensor) None
)"
2022-05-17 23:52:57,434 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,435 Corpus: "Corpus: 194 train + 22 dev + 33 test sentences"
2022-05-17 23:52:57,435 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,435 Parameters:
2022-05-17 23:52:57,436 - learning_rate: "0.1"
2022-05-17 23:52:57,437 - mini_batch_size: "32"
2022-05-17 23:52:57,437 - patience: "3"
2022-05-17 23:52:57,437 - anneal_factor: "0.5"
2022-05-17 23:52:57,438 - max_epochs: "10"
2022-05-17 23:52:57,439 - shuffle: "True"
2022-05-17 23:52:57,440 - train_with_dev: "False"
2022-05-17 23:52:57,440 - batch_growth_annealing: "False"
2022-05-17 23:52:57,441 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,441 Model training base path: "slot-model"
2022-05-17 23:52:57,442 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,443 Device: cpu
2022-05-17 23:52:57,443 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:57,444 Embeddings storage mode: cpu
2022-05-17 23:52:57,446 ----------------------------------------------------------------------------------------------------
2022-05-17 23:52:59,206 epoch 1 - iter 1/7 - loss 16.77810669 - samples/sec: 18.23 - lr: 0.100000
2022-05-17 23:53:01,036 epoch 1 - iter 2/7 - loss 15.17136908 - samples/sec: 17.51 - lr: 0.100000
2022-05-17 23:53:02,450 epoch 1 - iter 3/7 - loss 13.45863914 - samples/sec: 22.63 - lr: 0.100000
2022-05-17 23:53:04,163 epoch 1 - iter 4/7 - loss 11.81387305 - samples/sec: 18.70 - lr: 0.100000
2022-05-17 23:53:06,030 epoch 1 - iter 5/7 - loss 10.41218300 - samples/sec: 17.14 - lr: 0.100000
2022-05-17 23:53:07,655 epoch 1 - iter 6/7 - loss 9.20362504 - samples/sec: 19.70 - lr: 0.100000
2022-05-17 23:53:07,968 epoch 1 - iter 7/7 - loss 8.10721644 - samples/sec: 102.61 - lr: 0.100000
2022-05-17 23:53:07,969 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:07,970 EPOCH 1 done: loss 8.1072 - lr 0.1000000
2022-05-17 23:53:09,606 DEV : loss 3.991352081298828 - score 0.2
2022-05-17 23:53:09,607 BAD EPOCHS (no improvement): 0
2022-05-17 23:53:14,975 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:15,484 epoch 2 - iter 1/7 - loss 3.58558130 - samples/sec: 63.03 - lr: 0.100000
2022-05-17 23:53:15,865 epoch 2 - iter 2/7 - loss 3.12797976 - samples/sec: 84.20 - lr: 0.100000
2022-05-17 23:53:16,267 epoch 2 - iter 3/7 - loss 2.60615242 - samples/sec: 79.80 - lr: 0.100000
2022-05-17 23:53:16,738 epoch 2 - iter 4/7 - loss 2.71958175 - samples/sec: 68.18 - lr: 0.100000
2022-05-17 23:53:17,170 epoch 2 - iter 5/7 - loss 2.70331609 - samples/sec: 74.26 - lr: 0.100000
2022-05-17 23:53:17,603 epoch 2 - iter 6/7 - loss 2.51522466 - samples/sec: 74.01 - lr: 0.100000
2022-05-17 23:53:17,748 epoch 2 - iter 7/7 - loss 2.19215042 - samples/sec: 221.61 - lr: 0.100000
2022-05-17 23:53:17,749 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:17,750 EPOCH 2 done: loss 2.1922 - lr 0.1000000
2022-05-17 23:53:17,844 DEV : loss 3.9842920303344727 - score 0.3636
2022-05-17 23:53:17,846 BAD EPOCHS (no improvement): 0
2022-05-17 23:53:22,865 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:23,305 epoch 3 - iter 1/7 - loss 2.19582605 - samples/sec: 72.76 - lr: 0.100000
2022-05-17 23:53:23,741 epoch 3 - iter 2/7 - loss 1.85529530 - samples/sec: 73.58 - lr: 0.100000
2022-05-17 23:53:24,212 epoch 3 - iter 3/7 - loss 1.91948136 - samples/sec: 68.09 - lr: 0.100000
2022-05-17 23:53:24,717 epoch 3 - iter 4/7 - loss 2.11527669 - samples/sec: 63.50 - lr: 0.100000
2022-05-17 23:53:25,129 epoch 3 - iter 5/7 - loss 2.12587404 - samples/sec: 77.75 - lr: 0.100000
2022-05-17 23:53:25,630 epoch 3 - iter 6/7 - loss 2.01592445 - samples/sec: 63.92 - lr: 0.100000
2022-05-17 23:53:25,755 epoch 3 - iter 7/7 - loss 1.73551549 - samples/sec: 258.75 - lr: 0.100000
2022-05-17 23:53:25,756 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:25,757 EPOCH 3 done: loss 1.7355 - lr 0.1000000
2022-05-17 23:53:25,854 DEV : loss 3.3194284439086914 - score 0.3077
2022-05-17 23:53:25,855 BAD EPOCHS (no improvement): 1
2022-05-17 23:53:25,856 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:26,274 epoch 4 - iter 1/7 - loss 1.46010232 - samples/sec: 76.66 - lr: 0.100000
2022-05-17 23:53:26,734 epoch 4 - iter 2/7 - loss 1.18807647 - samples/sec: 69.66 - lr: 0.100000
2022-05-17 23:53:27,229 epoch 4 - iter 3/7 - loss 1.33144226 - samples/sec: 64.87 - lr: 0.100000
2022-05-17 23:53:27,775 epoch 4 - iter 4/7 - loss 1.64428358 - samples/sec: 58.69 - lr: 0.100000
2022-05-17 23:53:28,243 epoch 4 - iter 5/7 - loss 1.62551130 - samples/sec: 68.71 - lr: 0.100000
2022-05-17 23:53:28,727 epoch 4 - iter 6/7 - loss 1.74551653 - samples/sec: 66.25 - lr: 0.100000
2022-05-17 23:53:28,856 epoch 4 - iter 7/7 - loss 1.53921426 - samples/sec: 248.73 - lr: 0.100000
2022-05-17 23:53:28,857 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:28,858 EPOCH 4 done: loss 1.5392 - lr 0.1000000
2022-05-17 23:53:28,962 DEV : loss 2.8986825942993164 - score 0.2857
2022-05-17 23:53:28,963 BAD EPOCHS (no improvement): 2
2022-05-17 23:53:28,965 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:29,417 epoch 5 - iter 1/7 - loss 1.72827125 - samples/sec: 70.90 - lr: 0.100000
2022-05-17 23:53:29,902 epoch 5 - iter 2/7 - loss 1.51951337 - samples/sec: 66.07 - lr: 0.100000
2022-05-17 23:53:30,355 epoch 5 - iter 3/7 - loss 1.55555471 - samples/sec: 70.83 - lr: 0.100000
2022-05-17 23:53:30,840 epoch 5 - iter 4/7 - loss 1.31492138 - samples/sec: 66.16 - lr: 0.100000
2022-05-17 23:53:31,257 epoch 5 - iter 5/7 - loss 1.46497860 - samples/sec: 76.92 - lr: 0.100000
2022-05-17 23:53:31,768 epoch 5 - iter 6/7 - loss 1.60987592 - samples/sec: 62.75 - lr: 0.100000
2022-05-17 23:53:31,929 epoch 5 - iter 7/7 - loss 2.72113044 - samples/sec: 200.53 - lr: 0.100000
2022-05-17 23:53:31,930 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:31,931 EPOCH 5 done: loss 2.7211 - lr 0.1000000
2022-05-17 23:53:32,024 DEV : loss 2.766446590423584 - score 0.3077
2022-05-17 23:53:32,025 BAD EPOCHS (no improvement): 3
2022-05-17 23:53:32,026 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:32,475 epoch 6 - iter 1/7 - loss 1.68398678 - samples/sec: 71.62 - lr: 0.100000
2022-05-17 23:53:32,971 epoch 6 - iter 2/7 - loss 1.67541099 - samples/sec: 64.62 - lr: 0.100000
2022-05-17 23:53:33,400 epoch 6 - iter 3/7 - loss 1.58060956 - samples/sec: 74.78 - lr: 0.100000
2022-05-17 23:53:33,878 epoch 6 - iter 4/7 - loss 1.55456299 - samples/sec: 66.92 - lr: 0.100000
2022-05-17 23:53:34,278 epoch 6 - iter 5/7 - loss 1.50003145 - samples/sec: 80.28 - lr: 0.100000
2022-05-17 23:53:34,813 epoch 6 - iter 6/7 - loss 1.46878848 - samples/sec: 60.04 - lr: 0.100000
2022-05-17 23:53:34,951 epoch 6 - iter 7/7 - loss 1.66172016 - samples/sec: 233.22 - lr: 0.100000
2022-05-17 23:53:34,952 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:34,952 EPOCH 6 done: loss 1.6617 - lr 0.1000000
2022-05-17 23:53:35,040 DEV : loss 2.2595832347869873 - score 0.2857
2022-05-17 23:53:35,041 BAD EPOCHS (no improvement): 4
2022-05-17 23:53:35,043 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:35,461 epoch 7 - iter 1/7 - loss 1.14667833 - samples/sec: 76.93 - lr: 0.050000
2022-05-17 23:53:35,976 epoch 7 - iter 2/7 - loss 1.11618459 - samples/sec: 62.22 - lr: 0.050000
2022-05-17 23:53:36,416 epoch 7 - iter 3/7 - loss 1.24378494 - samples/sec: 72.88 - lr: 0.050000
2022-05-17 23:53:36,880 epoch 7 - iter 4/7 - loss 1.31663331 - samples/sec: 69.14 - lr: 0.050000
2022-05-17 23:53:37,298 epoch 7 - iter 5/7 - loss 1.39581544 - samples/sec: 76.75 - lr: 0.050000
2022-05-17 23:53:37,714 epoch 7 - iter 6/7 - loss 1.34690581 - samples/sec: 77.09 - lr: 0.050000
2022-05-17 23:53:37,860 epoch 7 - iter 7/7 - loss 1.46004195 - samples/sec: 220.36 - lr: 0.050000
2022-05-17 23:53:37,861 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:37,861 EPOCH 7 done: loss 1.4600 - lr 0.0500000
2022-05-17 23:53:37,954 DEV : loss 2.200728416442871 - score 0.2857
2022-05-17 23:53:37,955 BAD EPOCHS (no improvement): 1
2022-05-17 23:53:37,956 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:38,423 epoch 8 - iter 1/7 - loss 1.14459288 - samples/sec: 68.83 - lr: 0.050000
2022-05-17 23:53:38,805 epoch 8 - iter 2/7 - loss 0.95714736 - samples/sec: 83.88 - lr: 0.050000
2022-05-17 23:53:39,302 epoch 8 - iter 3/7 - loss 1.17704646 - samples/sec: 64.42 - lr: 0.050000
2022-05-17 23:53:39,781 epoch 8 - iter 4/7 - loss 1.29963121 - samples/sec: 66.92 - lr: 0.050000
2022-05-17 23:53:40,256 epoch 8 - iter 5/7 - loss 1.34262223 - samples/sec: 67.59 - lr: 0.050000
2022-05-17 23:53:40,704 epoch 8 - iter 6/7 - loss 1.33356750 - samples/sec: 71.53 - lr: 0.050000
2022-05-17 23:53:40,846 epoch 8 - iter 7/7 - loss 1.20113390 - samples/sec: 226.59 - lr: 0.050000
2022-05-17 23:53:40,847 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:40,848 EPOCH 8 done: loss 1.2011 - lr 0.0500000
2022-05-17 23:53:40,941 DEV : loss 2.4227261543273926 - score 0.2857
2022-05-17 23:53:40,942 BAD EPOCHS (no improvement): 2
2022-05-17 23:53:40,943 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:41,389 epoch 9 - iter 1/7 - loss 1.12297106 - samples/sec: 71.73 - lr: 0.050000
2022-05-17 23:53:41,800 epoch 9 - iter 2/7 - loss 0.92356640 - samples/sec: 78.01 - lr: 0.050000
2022-05-17 23:53:42,249 epoch 9 - iter 3/7 - loss 1.02407436 - samples/sec: 71.37 - lr: 0.050000
2022-05-17 23:53:42,667 epoch 9 - iter 4/7 - loss 1.04805315 - samples/sec: 76.71 - lr: 0.050000
2022-05-17 23:53:43,215 epoch 9 - iter 5/7 - loss 1.33371143 - samples/sec: 58.59 - lr: 0.050000
2022-05-17 23:53:43,661 epoch 9 - iter 6/7 - loss 1.27829826 - samples/sec: 71.89 - lr: 0.050000
2022-05-17 23:53:43,796 epoch 9 - iter 7/7 - loss 1.10260926 - samples/sec: 240.25 - lr: 0.050000
2022-05-17 23:53:43,797 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:43,798 EPOCH 9 done: loss 1.1026 - lr 0.0500000
2022-05-17 23:53:43,895 DEV : loss 2.1707162857055664 - score 0.3077
2022-05-17 23:53:43,896 BAD EPOCHS (no improvement): 3
2022-05-17 23:53:43,903 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:44,338 epoch 10 - iter 1/7 - loss 1.34320462 - samples/sec: 73.74 - lr: 0.050000
2022-05-17 23:53:44,808 epoch 10 - iter 2/7 - loss 0.96772069 - samples/sec: 68.25 - lr: 0.050000
2022-05-17 23:53:45,207 epoch 10 - iter 3/7 - loss 1.06257542 - samples/sec: 80.34 - lr: 0.050000
2022-05-17 23:53:45,729 epoch 10 - iter 4/7 - loss 0.92318819 - samples/sec: 61.50 - lr: 0.050000
2022-05-17 23:53:46,202 epoch 10 - iter 5/7 - loss 1.08295707 - samples/sec: 67.82 - lr: 0.050000
2022-05-17 23:53:46,707 epoch 10 - iter 6/7 - loss 1.18012399 - samples/sec: 63.49 - lr: 0.050000
2022-05-17 23:53:46,841 epoch 10 - iter 7/7 - loss 1.01267667 - samples/sec: 239.34 - lr: 0.050000
2022-05-17 23:53:46,842 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:46,842 EPOCH 10 done: loss 1.0127 - lr 0.0500000
2022-05-17 23:53:46,942 DEV : loss 1.9863343238830566 - score 0.3077
2022-05-17 23:53:46,943 BAD EPOCHS (no improvement): 4
2022-05-17 23:53:51,951 ----------------------------------------------------------------------------------------------------
2022-05-17 23:53:51,952 Testing using best model ...
2022-05-17 23:53:51,953 loading file slot-model\best-model.pt
2022-05-17 23:53:57,745 0.8000 0.2667 0.4000
2022-05-17 23:53:57,746
Results:
- F1-score (micro) 0.4000
- F1-score (macro) 0.2424
By class:
date tp: 0 - fp: 0 - fn: 4 - precision: 0.0000 - recall: 0.0000 - f1-score: 0.0000
quantity tp: 4 - fp: 1 - fn: 2 - precision: 0.8000 - recall: 0.6667 - f1-score: 0.7273
time tp: 0 - fp: 0 - fn: 5 - precision: 0.0000 - recall: 0.0000 - f1-score: 0.0000
2022-05-17 23:53:57,747 ----------------------------------------------------------------------------------------------------