aitech-sd-lab/intent-model-pl/training.log
2022-06-15 11:45:20 +02:00

345 lines
28 KiB
Plaintext

2022-06-08 12:34:22,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,080 Model: "TextClassifier(
(loss_function): CrossEntropyLoss()
(document_embeddings): DocumentRNNEmbeddings(
(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_reprojection_map): Linear(in_features=4446, out_features=4446, bias=True)
(rnn): GRU(4446, 512, batch_first=True)
(dropout): Dropout(p=0.5, inplace=False)
)
(decoder): Linear(in_features=512, out_features=12, bias=True)
(weights): None
(weight_tensor) None
)"
2022-06-08 12:34:22,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,080 Corpus: "Corpus: 230 train + 26 dev + 28 test sentences"
2022-06-08 12:34:22,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,080 Parameters:
2022-06-08 12:34:22,080 - learning_rate: "0.1"
2022-06-08 12:34:22,080 - mini_batch_size: "32"
2022-06-08 12:34:22,080 - patience: "5"
2022-06-08 12:34:22,080 - anneal_factor: "0.5"
2022-06-08 12:34:22,080 - max_epochs: "20"
2022-06-08 12:34:22,080 - shuffle: "True"
2022-06-08 12:34:22,080 - train_with_dev: "False"
2022-06-08 12:34:22,080 - batch_growth_annealing: "False"
2022-06-08 12:34:22,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,080 Model training base path: "intent-model-pl"
2022-06-08 12:34:22,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,081 Device: cpu
2022-06-08 12:34:22,081 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:22,081 Embeddings storage mode: cpu
2022-06-08 12:34:22,081 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:27,032 epoch 1 - iter 1/8 - loss 0.07898223 - samples/sec: 6.80 - lr: 0.100000
2022-06-08 12:34:29,165 epoch 1 - iter 2/8 - loss 0.06796387 - samples/sec: 15.00 - lr: 0.100000
2022-06-08 12:34:30,824 epoch 1 - iter 3/8 - loss 0.06508271 - samples/sec: 19.67 - lr: 0.100000
2022-06-08 12:34:32,885 epoch 1 - iter 4/8 - loss 0.06386624 - samples/sec: 15.53 - lr: 0.100000
2022-06-08 12:34:34,449 epoch 1 - iter 5/8 - loss 0.06001042 - samples/sec: 20.46 - lr: 0.100000
2022-06-08 12:34:36,511 epoch 1 - iter 6/8 - loss 0.06037725 - samples/sec: 15.53 - lr: 0.100000
2022-06-08 12:34:38,340 epoch 1 - iter 7/8 - loss 0.05774200 - samples/sec: 17.50 - lr: 0.100000
2022-06-08 12:34:39,567 epoch 1 - iter 8/8 - loss 0.06147691 - samples/sec: 26.09 - lr: 0.100000
2022-06-08 12:34:39,662 ----------------------------------------------------------------------------------------------------
2022-06-08 12:34:39,662 EPOCH 1 done: loss 0.0615 - lr 0.1000000
2022-06-08 12:34:41,854 DEV : loss 0.05763816088438034 - f1-score (micro avg) 0.5385
2022-06-08 12:34:41,857 BAD EPOCHS (no improvement): 0
2022-06-08 12:34:41,857 saving best model
2022-06-08 12:34:59,881 ----------------------------------------------------------------------------------------------------
2022-06-08 12:35:05,515 epoch 2 - iter 1/8 - loss 0.04383654 - samples/sec: 5.94 - lr: 0.100000
2022-06-08 12:35:07,644 epoch 2 - iter 2/8 - loss 0.04492071 - samples/sec: 15.03 - lr: 0.100000
2022-06-08 12:35:09,446 epoch 2 - iter 3/8 - loss 0.04089480 - samples/sec: 17.76 - lr: 0.100000
2022-06-08 12:35:11,628 epoch 2 - iter 4/8 - loss 0.03963978 - samples/sec: 14.66 - lr: 0.100000
2022-06-08 12:35:13,633 epoch 2 - iter 5/8 - loss 0.04254628 - samples/sec: 15.96 - lr: 0.100000
2022-06-08 12:35:15,634 epoch 2 - iter 6/8 - loss 0.04305776 - samples/sec: 16.00 - lr: 0.100000
2022-06-08 12:35:18,469 epoch 2 - iter 7/8 - loss 0.04148617 - samples/sec: 11.29 - lr: 0.100000
2022-06-08 12:35:19,106 epoch 2 - iter 8/8 - loss 0.04477506 - samples/sec: 50.28 - lr: 0.100000
2022-06-08 12:35:19,207 ----------------------------------------------------------------------------------------------------
2022-06-08 12:35:19,207 EPOCH 2 done: loss 0.0448 - lr 0.1000000
2022-06-08 12:35:21,500 DEV : loss 0.042724136263132095 - f1-score (micro avg) 0.6538
2022-06-08 12:35:21,504 BAD EPOCHS (no improvement): 0
2022-06-08 12:35:21,504 saving best model
2022-06-08 12:35:34,450 ----------------------------------------------------------------------------------------------------
2022-06-08 12:35:37,771 epoch 3 - iter 1/8 - loss 0.03886844 - samples/sec: 10.54 - lr: 0.100000
2022-06-08 12:35:40,125 epoch 3 - iter 2/8 - loss 0.03984906 - samples/sec: 13.74 - lr: 0.100000
2022-06-08 12:35:42,112 epoch 3 - iter 3/8 - loss 0.04055306 - samples/sec: 16.11 - lr: 0.100000
2022-06-08 12:35:44,306 epoch 3 - iter 4/8 - loss 0.03952748 - samples/sec: 14.58 - lr: 0.100000
2022-06-08 12:35:48,966 epoch 3 - iter 5/8 - loss 0.03757689 - samples/sec: 6.87 - lr: 0.100000
2022-06-08 12:35:51,148 epoch 3 - iter 6/8 - loss 0.03498927 - samples/sec: 14.68 - lr: 0.100000
2022-06-08 12:35:53,949 epoch 3 - iter 7/8 - loss 0.03476784 - samples/sec: 11.45 - lr: 0.100000
2022-06-08 12:35:54,306 epoch 3 - iter 8/8 - loss 0.03608819 - samples/sec: 89.68 - lr: 0.100000
2022-06-08 12:35:54,436 ----------------------------------------------------------------------------------------------------
2022-06-08 12:35:54,436 EPOCH 3 done: loss 0.0361 - lr 0.1000000
2022-06-08 12:35:56,784 DEV : loss 0.05470799654722214 - f1-score (micro avg) 0.5
2022-06-08 12:35:56,822 BAD EPOCHS (no improvement): 1
2022-06-08 12:35:56,823 ----------------------------------------------------------------------------------------------------
2022-06-08 12:35:59,330 epoch 4 - iter 1/8 - loss 0.03909241 - samples/sec: 14.34 - lr: 0.100000
2022-06-08 12:36:01,466 epoch 4 - iter 2/8 - loss 0.03235384 - samples/sec: 14.99 - lr: 0.100000
2022-06-08 12:36:03,705 epoch 4 - iter 3/8 - loss 0.02995395 - samples/sec: 14.30 - lr: 0.100000
2022-06-08 12:36:05,365 epoch 4 - iter 4/8 - loss 0.02825506 - samples/sec: 19.29 - lr: 0.100000
2022-06-08 12:36:06,965 epoch 4 - iter 5/8 - loss 0.02953058 - samples/sec: 20.02 - lr: 0.100000
2022-06-08 12:36:11,659 epoch 4 - iter 6/8 - loss 0.02911502 - samples/sec: 6.82 - lr: 0.100000
2022-06-08 12:36:13,855 epoch 4 - iter 7/8 - loss 0.02968962 - samples/sec: 14.58 - lr: 0.100000
2022-06-08 12:36:14,483 epoch 4 - iter 8/8 - loss 0.03153667 - samples/sec: 50.95 - lr: 0.100000
2022-06-08 12:36:14,574 ----------------------------------------------------------------------------------------------------
2022-06-08 12:36:14,574 EPOCH 4 done: loss 0.0315 - lr 0.1000000
2022-06-08 12:36:16,935 DEV : loss 0.0489649623632431 - f1-score (micro avg) 0.6538
2022-06-08 12:36:16,939 BAD EPOCHS (no improvement): 2
2022-06-08 12:36:16,939 ----------------------------------------------------------------------------------------------------
2022-06-08 12:36:19,337 epoch 5 - iter 1/8 - loss 0.03982311 - samples/sec: 15.05 - lr: 0.100000
2022-06-08 12:36:21,627 epoch 5 - iter 2/8 - loss 0.03406263 - samples/sec: 13.99 - lr: 0.100000
2022-06-08 12:36:23,271 epoch 5 - iter 3/8 - loss 0.03089139 - samples/sec: 19.48 - lr: 0.100000
2022-06-08 12:36:25,152 epoch 5 - iter 4/8 - loss 0.03117696 - samples/sec: 17.01 - lr: 0.100000
2022-06-08 12:36:27,375 epoch 5 - iter 5/8 - loss 0.02918866 - samples/sec: 14.40 - lr: 0.100000
2022-06-08 12:36:30,199 epoch 5 - iter 6/8 - loss 0.02907407 - samples/sec: 11.34 - lr: 0.100000
2022-06-08 12:36:34,868 epoch 5 - iter 7/8 - loss 0.02830181 - samples/sec: 6.85 - lr: 0.100000
2022-06-08 12:36:35,598 epoch 5 - iter 8/8 - loss 0.02879909 - samples/sec: 43.85 - lr: 0.100000
2022-06-08 12:36:35,702 ----------------------------------------------------------------------------------------------------
2022-06-08 12:36:35,702 EPOCH 5 done: loss 0.0288 - lr 0.1000000
2022-06-08 12:36:38,102 DEV : loss 0.03488617390394211 - f1-score (micro avg) 0.6923
2022-06-08 12:36:38,105 BAD EPOCHS (no improvement): 0
2022-06-08 12:36:38,105 saving best model
2022-06-08 12:36:51,917 ----------------------------------------------------------------------------------------------------
2022-06-08 12:36:55,625 epoch 6 - iter 1/8 - loss 0.01752557 - samples/sec: 9.46 - lr: 0.100000
2022-06-08 12:37:00,500 epoch 6 - iter 2/8 - loss 0.02358356 - samples/sec: 6.57 - lr: 0.100000
2022-06-08 12:37:02,609 epoch 6 - iter 3/8 - loss 0.02323116 - samples/sec: 15.18 - lr: 0.100000
2022-06-08 12:37:04,450 epoch 6 - iter 4/8 - loss 0.02282404 - samples/sec: 17.38 - lr: 0.100000
2022-06-08 12:37:06,344 epoch 6 - iter 5/8 - loss 0.02280068 - samples/sec: 16.89 - lr: 0.100000
2022-06-08 12:37:09,226 epoch 6 - iter 6/8 - loss 0.02309799 - samples/sec: 11.10 - lr: 0.100000
2022-06-08 12:37:11,370 epoch 6 - iter 7/8 - loss 0.02368303 - samples/sec: 14.96 - lr: 0.100000
2022-06-08 12:37:11,908 epoch 6 - iter 8/8 - loss 0.02837071 - samples/sec: 59.45 - lr: 0.100000
2022-06-08 12:37:12,080 ----------------------------------------------------------------------------------------------------
2022-06-08 12:37:12,080 EPOCH 6 done: loss 0.0284 - lr 0.1000000
2022-06-08 12:37:14,505 DEV : loss 0.03307553753256798 - f1-score (micro avg) 0.7692
2022-06-08 12:37:14,508 BAD EPOCHS (no improvement): 0
2022-06-08 12:37:14,509 saving best model
2022-06-08 12:37:25,596 ----------------------------------------------------------------------------------------------------
2022-06-08 12:37:29,075 epoch 7 - iter 1/8 - loss 0.01517198 - samples/sec: 11.87 - lr: 0.100000
2022-06-08 12:37:30,842 epoch 7 - iter 2/8 - loss 0.01896675 - samples/sec: 18.12 - lr: 0.100000
2022-06-08 12:37:32,529 epoch 7 - iter 3/8 - loss 0.02089164 - samples/sec: 18.98 - lr: 0.100000
2022-06-08 12:37:37,295 epoch 7 - iter 4/8 - loss 0.02471227 - samples/sec: 6.72 - lr: 0.100000
2022-06-08 12:37:40,119 epoch 7 - iter 5/8 - loss 0.02357146 - samples/sec: 11.33 - lr: 0.100000
2022-06-08 12:37:41,682 epoch 7 - iter 6/8 - loss 0.02472030 - samples/sec: 20.47 - lr: 0.100000
2022-06-08 12:37:43,828 epoch 7 - iter 7/8 - loss 0.02383381 - samples/sec: 14.93 - lr: 0.100000
2022-06-08 12:37:44,515 epoch 7 - iter 8/8 - loss 0.02485561 - samples/sec: 46.60 - lr: 0.100000
2022-06-08 12:37:44,657 ----------------------------------------------------------------------------------------------------
2022-06-08 12:37:44,657 EPOCH 7 done: loss 0.0249 - lr 0.1000000
2022-06-08 12:37:47,057 DEV : loss 0.03743937239050865 - f1-score (micro avg) 0.6923
2022-06-08 12:37:47,108 BAD EPOCHS (no improvement): 1
2022-06-08 12:37:47,110 ----------------------------------------------------------------------------------------------------
2022-06-08 12:37:49,740 epoch 8 - iter 1/8 - loss 0.02161453 - samples/sec: 13.59 - lr: 0.100000
2022-06-08 12:37:51,890 epoch 8 - iter 2/8 - loss 0.02138850 - samples/sec: 14.89 - lr: 0.100000
2022-06-08 12:37:54,754 epoch 8 - iter 3/8 - loss 0.02179432 - samples/sec: 11.17 - lr: 0.100000
2022-06-08 12:37:56,975 epoch 8 - iter 4/8 - loss 0.02093382 - samples/sec: 14.41 - lr: 0.100000
2022-06-08 12:37:58,994 epoch 8 - iter 5/8 - loss 0.02126111 - samples/sec: 15.86 - lr: 0.100000
2022-06-08 12:38:00,519 epoch 8 - iter 6/8 - loss 0.02195439 - samples/sec: 20.99 - lr: 0.100000
2022-06-08 12:38:05,196 epoch 8 - iter 7/8 - loss 0.02149169 - samples/sec: 6.84 - lr: 0.100000
2022-06-08 12:38:06,027 epoch 8 - iter 8/8 - loss 0.02449182 - samples/sec: 38.53 - lr: 0.100000
2022-06-08 12:38:06,137 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:06,137 EPOCH 8 done: loss 0.0245 - lr 0.1000000
2022-06-08 12:38:08,492 DEV : loss 0.037434812635183334 - f1-score (micro avg) 0.7308
2022-06-08 12:38:08,494 BAD EPOCHS (no improvement): 2
2022-06-08 12:38:08,494 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:10,396 epoch 9 - iter 1/8 - loss 0.02186348 - samples/sec: 19.79 - lr: 0.100000
2022-06-08 12:38:13,232 epoch 9 - iter 2/8 - loss 0.02077455 - samples/sec: 11.29 - lr: 0.100000
2022-06-08 12:38:15,307 epoch 9 - iter 3/8 - loss 0.02100625 - samples/sec: 15.58 - lr: 0.100000
2022-06-08 12:38:19,952 epoch 9 - iter 4/8 - loss 0.02074756 - samples/sec: 6.89 - lr: 0.100000
2022-06-08 12:38:22,174 epoch 9 - iter 5/8 - loss 0.02067224 - samples/sec: 14.41 - lr: 0.100000
2022-06-08 12:38:24,347 epoch 9 - iter 6/8 - loss 0.02019721 - samples/sec: 14.74 - lr: 0.100000
2022-06-08 12:38:25,990 epoch 9 - iter 7/8 - loss 0.01983759 - samples/sec: 19.48 - lr: 0.100000
2022-06-08 12:38:26,600 epoch 9 - iter 8/8 - loss 0.02107237 - samples/sec: 52.47 - lr: 0.100000
2022-06-08 12:38:26,713 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:26,713 EPOCH 9 done: loss 0.0211 - lr 0.1000000
2022-06-08 12:38:29,103 DEV : loss 0.042992427945137024 - f1-score (micro avg) 0.6923
2022-06-08 12:38:29,105 BAD EPOCHS (no improvement): 3
2022-06-08 12:38:29,106 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:31,434 epoch 10 - iter 1/8 - loss 0.02094492 - samples/sec: 16.03 - lr: 0.100000
2022-06-08 12:38:36,102 epoch 10 - iter 2/8 - loss 0.02262537 - samples/sec: 6.86 - lr: 0.100000
2022-06-08 12:38:37,698 epoch 10 - iter 3/8 - loss 0.02049491 - samples/sec: 20.06 - lr: 0.100000
2022-06-08 12:38:39,710 epoch 10 - iter 4/8 - loss 0.01806376 - samples/sec: 15.91 - lr: 0.100000
2022-06-08 12:38:42,522 epoch 10 - iter 5/8 - loss 0.01732504 - samples/sec: 11.38 - lr: 0.100000
2022-06-08 12:38:44,752 epoch 10 - iter 6/8 - loss 0.01746574 - samples/sec: 14.53 - lr: 0.100000
2022-06-08 12:38:46,731 epoch 10 - iter 7/8 - loss 0.01723046 - samples/sec: 16.17 - lr: 0.100000
2022-06-08 12:38:47,614 epoch 10 - iter 8/8 - loss 0.01904527 - samples/sec: 36.25 - lr: 0.100000
2022-06-08 12:38:47,732 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:47,732 EPOCH 10 done: loss 0.0190 - lr 0.1000000
2022-06-08 12:38:50,129 DEV : loss 0.038346242159605026 - f1-score (micro avg) 0.6923
2022-06-08 12:38:50,132 BAD EPOCHS (no improvement): 4
2022-06-08 12:38:50,132 ----------------------------------------------------------------------------------------------------
2022-06-08 12:38:53,363 epoch 11 - iter 1/8 - loss 0.02434092 - samples/sec: 10.90 - lr: 0.100000
2022-06-08 12:38:55,224 epoch 11 - iter 2/8 - loss 0.02028142 - samples/sec: 17.20 - lr: 0.100000
2022-06-08 12:38:57,343 epoch 11 - iter 3/8 - loss 0.01786905 - samples/sec: 15.36 - lr: 0.100000
2022-06-08 12:38:59,556 epoch 11 - iter 4/8 - loss 0.01878600 - samples/sec: 14.46 - lr: 0.100000
2022-06-08 12:39:01,231 epoch 11 - iter 5/8 - loss 0.01823605 - samples/sec: 19.11 - lr: 0.100000
2022-06-08 12:39:05,875 epoch 11 - iter 6/8 - loss 0.01741535 - samples/sec: 6.90 - lr: 0.100000
2022-06-08 12:39:07,860 epoch 11 - iter 7/8 - loss 0.01748662 - samples/sec: 16.12 - lr: 0.100000
2022-06-08 12:39:08,500 epoch 11 - iter 8/8 - loss 0.01859635 - samples/sec: 50.01 - lr: 0.100000
2022-06-08 12:39:08,609 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:08,609 EPOCH 11 done: loss 0.0186 - lr 0.1000000
2022-06-08 12:39:11,091 DEV : loss 0.034892842173576355 - f1-score (micro avg) 0.6923
2022-06-08 12:39:11,100 BAD EPOCHS (no improvement): 5
2022-06-08 12:39:11,100 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:13,750 epoch 12 - iter 1/8 - loss 0.01634524 - samples/sec: 14.71 - lr: 0.100000
2022-06-08 12:39:15,623 epoch 12 - iter 2/8 - loss 0.01762738 - samples/sec: 17.09 - lr: 0.100000
2022-06-08 12:39:17,598 epoch 12 - iter 3/8 - loss 0.01626732 - samples/sec: 16.21 - lr: 0.100000
2022-06-08 12:39:19,816 epoch 12 - iter 4/8 - loss 0.01639530 - samples/sec: 14.43 - lr: 0.100000
2022-06-08 12:39:22,573 epoch 12 - iter 5/8 - loss 0.01589918 - samples/sec: 11.61 - lr: 0.100000
2022-06-08 12:39:24,629 epoch 12 - iter 6/8 - loss 0.01590672 - samples/sec: 15.62 - lr: 0.100000
2022-06-08 12:39:29,193 epoch 12 - iter 7/8 - loss 0.01647368 - samples/sec: 7.01 - lr: 0.100000
2022-06-08 12:39:29,931 epoch 12 - iter 8/8 - loss 0.01770140 - samples/sec: 43.39 - lr: 0.100000
2022-06-08 12:39:30,066 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:30,066 EPOCH 12 done: loss 0.0177 - lr 0.1000000
2022-06-08 12:39:32,443 DEV : loss 0.034579355269670486 - f1-score (micro avg) 0.7308
2022-06-08 12:39:32,446 BAD EPOCHS (no improvement): 6
2022-06-08 12:39:32,446 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:34,865 epoch 13 - iter 1/8 - loss 0.01002872 - samples/sec: 15.16 - lr: 0.050000
2022-06-08 12:39:39,522 epoch 13 - iter 2/8 - loss 0.00905764 - samples/sec: 6.87 - lr: 0.050000
2022-06-08 12:39:41,432 epoch 13 - iter 3/8 - loss 0.01199489 - samples/sec: 16.81 - lr: 0.050000
2022-06-08 12:39:43,623 epoch 13 - iter 4/8 - loss 0.01235516 - samples/sec: 14.61 - lr: 0.050000
2022-06-08 12:39:45,664 epoch 13 - iter 5/8 - loss 0.01314778 - samples/sec: 15.69 - lr: 0.050000
2022-06-08 12:39:48,505 epoch 13 - iter 6/8 - loss 0.01353356 - samples/sec: 11.26 - lr: 0.050000
2022-06-08 12:39:50,685 epoch 13 - iter 7/8 - loss 0.01315784 - samples/sec: 14.68 - lr: 0.050000
2022-06-08 12:39:51,159 epoch 13 - iter 8/8 - loss 0.01533346 - samples/sec: 67.63 - lr: 0.050000
2022-06-08 12:39:51,274 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:51,274 EPOCH 13 done: loss 0.0153 - lr 0.0500000
2022-06-08 12:39:53,706 DEV : loss 0.04151434451341629 - f1-score (micro avg) 0.6154
2022-06-08 12:39:53,709 BAD EPOCHS (no improvement): 1
2022-06-08 12:39:53,709 ----------------------------------------------------------------------------------------------------
2022-06-08 12:39:56,144 epoch 14 - iter 1/8 - loss 0.01341827 - samples/sec: 15.44 - lr: 0.050000
2022-06-08 12:40:00,982 epoch 14 - iter 2/8 - loss 0.01319934 - samples/sec: 6.62 - lr: 0.050000
2022-06-08 12:40:03,230 epoch 14 - iter 3/8 - loss 0.01336550 - samples/sec: 14.24 - lr: 0.050000
2022-06-08 12:40:05,286 epoch 14 - iter 4/8 - loss 0.01466392 - samples/sec: 15.57 - lr: 0.050000
2022-06-08 12:40:06,621 epoch 14 - iter 5/8 - loss 0.01446051 - samples/sec: 23.97 - lr: 0.050000
2022-06-08 12:40:08,331 epoch 14 - iter 6/8 - loss 0.01459049 - samples/sec: 18.75 - lr: 0.050000
2022-06-08 12:40:10,554 epoch 14 - iter 7/8 - loss 0.01388460 - samples/sec: 14.40 - lr: 0.050000
2022-06-08 12:40:11,158 epoch 14 - iter 8/8 - loss 0.01660707 - samples/sec: 53.05 - lr: 0.050000
2022-06-08 12:40:11,276 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:11,277 EPOCH 14 done: loss 0.0166 - lr 0.0500000
2022-06-08 12:40:13,718 DEV : loss 0.03577208146452904 - f1-score (micro avg) 0.6923
2022-06-08 12:40:13,721 BAD EPOCHS (no improvement): 2
2022-06-08 12:40:13,721 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:16,105 epoch 15 - iter 1/8 - loss 0.01129489 - samples/sec: 15.44 - lr: 0.050000
2022-06-08 12:40:17,658 epoch 15 - iter 2/8 - loss 0.01163308 - samples/sec: 20.63 - lr: 0.050000
2022-06-08 12:40:22,263 epoch 15 - iter 3/8 - loss 0.01246105 - samples/sec: 6.95 - lr: 0.050000
2022-06-08 12:40:25,085 epoch 15 - iter 4/8 - loss 0.01287149 - samples/sec: 11.36 - lr: 0.050000
2022-06-08 12:40:27,276 epoch 15 - iter 5/8 - loss 0.01303117 - samples/sec: 14.61 - lr: 0.050000
2022-06-08 12:40:29,424 epoch 15 - iter 6/8 - loss 0.01248200 - samples/sec: 14.91 - lr: 0.050000
2022-06-08 12:40:31,238 epoch 15 - iter 7/8 - loss 0.01238609 - samples/sec: 17.64 - lr: 0.050000
2022-06-08 12:40:31,809 epoch 15 - iter 8/8 - loss 0.01345632 - samples/sec: 56.11 - lr: 0.050000
2022-06-08 12:40:31,937 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:31,937 EPOCH 15 done: loss 0.0135 - lr 0.0500000
2022-06-08 12:40:34,369 DEV : loss 0.03413332253694534 - f1-score (micro avg) 0.6923
2022-06-08 12:40:34,371 BAD EPOCHS (no improvement): 3
2022-06-08 12:40:34,372 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:37,084 epoch 16 - iter 1/8 - loss 0.00571280 - samples/sec: 13.68 - lr: 0.050000
2022-06-08 12:40:39,898 epoch 16 - iter 2/8 - loss 0.00735429 - samples/sec: 11.37 - lr: 0.050000
2022-06-08 12:40:44,599 epoch 16 - iter 3/8 - loss 0.00981462 - samples/sec: 6.81 - lr: 0.050000
2022-06-08 12:40:46,399 epoch 16 - iter 4/8 - loss 0.01109150 - samples/sec: 17.78 - lr: 0.050000
2022-06-08 12:40:48,038 epoch 16 - iter 5/8 - loss 0.01045035 - samples/sec: 19.58 - lr: 0.050000
2022-06-08 12:40:49,922 epoch 16 - iter 6/8 - loss 0.01143379 - samples/sec: 16.99 - lr: 0.050000
2022-06-08 12:40:52,117 epoch 16 - iter 7/8 - loss 0.01223135 - samples/sec: 14.58 - lr: 0.050000
2022-06-08 12:40:52,574 epoch 16 - iter 8/8 - loss 0.01220234 - samples/sec: 70.09 - lr: 0.050000
2022-06-08 12:40:52,709 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:52,709 EPOCH 16 done: loss 0.0122 - lr 0.0500000
2022-06-08 12:40:55,283 DEV : loss 0.036424651741981506 - f1-score (micro avg) 0.6538
2022-06-08 12:40:55,286 BAD EPOCHS (no improvement): 4
2022-06-08 12:40:55,286 ----------------------------------------------------------------------------------------------------
2022-06-08 12:40:57,804 epoch 17 - iter 1/8 - loss 0.01031059 - samples/sec: 14.68 - lr: 0.050000
2022-06-08 12:41:02,494 epoch 17 - iter 2/8 - loss 0.00809774 - samples/sec: 6.83 - lr: 0.050000
2022-06-08 12:41:04,653 epoch 17 - iter 3/8 - loss 0.00930187 - samples/sec: 14.86 - lr: 0.050000
2022-06-08 12:41:06,891 epoch 17 - iter 4/8 - loss 0.00949802 - samples/sec: 14.30 - lr: 0.050000
2022-06-08 12:41:08,513 epoch 17 - iter 5/8 - loss 0.00988633 - samples/sec: 19.74 - lr: 0.050000
2022-06-08 12:41:11,296 epoch 17 - iter 6/8 - loss 0.01026497 - samples/sec: 11.50 - lr: 0.050000
2022-06-08 12:41:13,285 epoch 17 - iter 7/8 - loss 0.01118278 - samples/sec: 16.10 - lr: 0.050000
2022-06-08 12:41:13,854 epoch 17 - iter 8/8 - loss 0.01508908 - samples/sec: 56.24 - lr: 0.050000
2022-06-08 12:41:13,993 ----------------------------------------------------------------------------------------------------
2022-06-08 12:41:13,993 EPOCH 17 done: loss 0.0151 - lr 0.0500000
2022-06-08 12:41:16,440 DEV : loss 0.043076448142528534 - f1-score (micro avg) 0.5769
2022-06-08 12:41:16,443 BAD EPOCHS (no improvement): 5
2022-06-08 12:41:16,443 ----------------------------------------------------------------------------------------------------
2022-06-08 12:41:21,548 epoch 18 - iter 1/8 - loss 0.01487586 - samples/sec: 6.70 - lr: 0.050000
2022-06-08 12:41:23,000 epoch 18 - iter 2/8 - loss 0.01476814 - samples/sec: 22.11 - lr: 0.050000
2022-06-08 12:41:25,283 epoch 18 - iter 3/8 - loss 0.01292397 - samples/sec: 14.28 - lr: 0.050000
2022-06-08 12:41:27,254 epoch 18 - iter 4/8 - loss 0.01281850 - samples/sec: 16.24 - lr: 0.050000
2022-06-08 12:41:30,065 epoch 18 - iter 5/8 - loss 0.01216846 - samples/sec: 11.39 - lr: 0.050000
2022-06-08 12:41:31,727 epoch 18 - iter 6/8 - loss 0.01235817 - samples/sec: 19.25 - lr: 0.050000
2022-06-08 12:41:33,795 epoch 18 - iter 7/8 - loss 0.01231274 - samples/sec: 15.48 - lr: 0.050000
2022-06-08 12:41:34,503 epoch 18 - iter 8/8 - loss 0.01456582 - samples/sec: 45.18 - lr: 0.050000
2022-06-08 12:41:34,645 ----------------------------------------------------------------------------------------------------
2022-06-08 12:41:34,645 EPOCH 18 done: loss 0.0146 - lr 0.0500000
2022-06-08 12:41:37,113 DEV : loss 0.03268644958734512 - f1-score (micro avg) 0.6538
2022-06-08 12:41:37,116 BAD EPOCHS (no improvement): 6
2022-06-08 12:41:37,116 ----------------------------------------------------------------------------------------------------
2022-06-08 12:41:39,579 epoch 19 - iter 1/8 - loss 0.01510787 - samples/sec: 15.47 - lr: 0.025000
2022-06-08 12:41:41,456 epoch 19 - iter 2/8 - loss 0.01197568 - samples/sec: 17.08 - lr: 0.025000
2022-06-08 12:41:43,215 epoch 19 - iter 3/8 - loss 0.01018611 - samples/sec: 18.19 - lr: 0.025000
2022-06-08 12:41:45,412 epoch 19 - iter 4/8 - loss 0.01054903 - samples/sec: 14.57 - lr: 0.025000
2022-06-08 12:41:47,599 epoch 19 - iter 5/8 - loss 0.01028886 - samples/sec: 14.63 - lr: 0.025000
2022-06-08 12:41:52,295 epoch 19 - iter 6/8 - loss 0.01072677 - samples/sec: 6.86 - lr: 0.025000
2022-06-08 12:41:54,351 epoch 19 - iter 7/8 - loss 0.01028150 - samples/sec: 15.56 - lr: 0.025000
2022-06-08 12:41:54,991 epoch 19 - iter 8/8 - loss 0.01098372 - samples/sec: 50.06 - lr: 0.025000
2022-06-08 12:41:55,126 ----------------------------------------------------------------------------------------------------
2022-06-08 12:41:55,126 EPOCH 19 done: loss 0.0110 - lr 0.0250000
2022-06-08 12:41:57,608 DEV : loss 0.03280213102698326 - f1-score (micro avg) 0.7308
2022-06-08 12:41:57,611 BAD EPOCHS (no improvement): 1
2022-06-08 12:41:57,611 ----------------------------------------------------------------------------------------------------
2022-06-08 12:42:00,143 epoch 20 - iter 1/8 - loss 0.01120961 - samples/sec: 14.57 - lr: 0.025000
2022-06-08 12:42:02,137 epoch 20 - iter 2/8 - loss 0.00992019 - samples/sec: 16.09 - lr: 0.025000
2022-06-08 12:42:04,354 epoch 20 - iter 3/8 - loss 0.01115414 - samples/sec: 14.45 - lr: 0.025000
2022-06-08 12:42:09,083 epoch 20 - iter 4/8 - loss 0.01073034 - samples/sec: 6.79 - lr: 0.025000
2022-06-08 12:42:11,268 epoch 20 - iter 5/8 - loss 0.01123121 - samples/sec: 14.65 - lr: 0.025000
2022-06-08 12:42:12,959 epoch 20 - iter 6/8 - loss 0.01051536 - samples/sec: 18.94 - lr: 0.025000
2022-06-08 12:42:15,770 epoch 20 - iter 7/8 - loss 0.01079541 - samples/sec: 11.38 - lr: 0.025000
2022-06-08 12:42:16,382 epoch 20 - iter 8/8 - loss 0.01113883 - samples/sec: 52.33 - lr: 0.025000
2022-06-08 12:42:16,523 ----------------------------------------------------------------------------------------------------
2022-06-08 12:42:16,524 EPOCH 20 done: loss 0.0111 - lr 0.0250000
2022-06-08 12:42:18,991 DEV : loss 0.034038737416267395 - f1-score (micro avg) 0.7308
2022-06-08 12:42:18,994 BAD EPOCHS (no improvement): 2
2022-06-08 12:42:33,998 ----------------------------------------------------------------------------------------------------
2022-06-08 12:42:34,011 loading file intent-model-pl/best-model.pt
2022-06-08 12:42:53,520 0.6786 0.6786 0.6786 0.6786
2022-06-08 12:42:53,520
Results:
- F-score (micro) 0.6786
- F-score (macro) 0.3456
- Accuracy 0.6786
By class:
precision recall f1-score support
inform 0.7647 0.8667 0.8125 15
null 0.2500 0.3333 0.2857 3
request 0.6667 0.6667 0.6667 3
hello 1.0000 1.0000 1.0000 3
ack 0.0000 0.0000 0.0000 2
deny 0.0000 0.0000 0.0000 1
reqmore 0.0000 0.0000 0.0000 1
affirm 0.0000 0.0000 0.0000 0
micro avg 0.6786 0.6786 0.6786 28
macro avg 0.3352 0.3583 0.3456 28
weighted avg 0.6150 0.6786 0.6445 28
samples avg 0.6786 0.6786 0.6786 28
2022-06-08 12:42:53,520 ----------------------------------------------------------------------------------------------------