2022-04-03 20:45:35,951 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:35,958 Model: "SequenceTagger( (embeddings): StackedEmbeddings( (list_embedding_0): WordEmbeddings('fa') (list_embedding_1): FlairEmbeddings( (lm): LanguageModel( (drop): Dropout(p=0.1, inplace=False) (encoder): Embedding(5105, 100) (rnn): LSTM(100, 2048) (decoder): Linear(in_features=2048, out_features=5105, bias=True) ) ) (list_embedding_2): FlairEmbeddings( (lm): LanguageModel( (drop): Dropout(p=0.1, inplace=False) (encoder): Embedding(5105, 100) (rnn): LSTM(100, 2048) (decoder): Linear(in_features=2048, out_features=5105, bias=True) ) ) ) (word_dropout): WordDropout(p=0.05) (locked_dropout): LockedDropout(p=0.5) (embedding2nn): Linear(in_features=4396, out_features=4396, bias=True) (rnn): LSTM(4396, 256, batch_first=True, bidirectional=True) (linear): Linear(in_features=512, out_features=17, bias=True) (beta): 1.0 (weights): None (weight_tensor) None )" 2022-04-03 20:45:35,962 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:35,967 Corpus: "Corpus: 23060 train + 4070 dev + 4150 test sentences" 2022-04-03 20:45:35,971 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:35,973 Parameters: 2022-04-03 20:45:35,975 - learning_rate: "0.05" 2022-04-03 20:45:35,977 - mini_batch_size: "4" 2022-04-03 20:45:35,980 - patience: "3" 2022-04-03 20:45:35,982 - anneal_factor: "0.5" 2022-04-03 20:45:35,985 - max_epochs: "40" 2022-04-03 20:45:35,988 - shuffle: "True" 2022-04-03 20:45:35,991 - train_with_dev: "False" 2022-04-03 20:45:35,996 - batch_growth_annealing: "False" 2022-04-03 20:45:35,998 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:36,001 Model training base path: "/content/gdrive/MyDrive/project/data/ner/model2" 2022-04-03 20:45:36,004 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:36,006 Device: cuda:0 2022-04-03 20:45:36,007 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:45:36,009 Embeddings storage mode: none 2022-04-03 20:45:36,559 ---------------------------------------------------------------------------------------------------- 2022-04-03 20:49:55,248 epoch 40 - iter 576/5765 - loss 0.05129424 - samples/sec: 8.91 - lr: 0.050000 2022-04-03 20:54:12,817 epoch 40 - iter 1152/5765 - loss 0.05045109 - samples/sec: 8.98 - lr: 0.050000 2022-04-03 20:58:35,265 epoch 40 - iter 1728/5765 - loss 0.05189116 - samples/sec: 8.81 - lr: 0.050000 2022-04-03 21:03:11,325 epoch 40 - iter 2304/5765 - loss 0.05151945 - samples/sec: 8.38 - lr: 0.050000 2022-04-03 21:07:46,802 epoch 40 - iter 2880/5765 - loss 0.05105861 - samples/sec: 8.40 - lr: 0.050000 2022-04-03 21:12:16,061 epoch 40 - iter 3456/5765 - loss 0.05160696 - samples/sec: 8.59 - lr: 0.050000 2022-04-03 21:16:46,997 epoch 40 - iter 4032/5765 - loss 0.05158343 - samples/sec: 8.54 - lr: 0.050000 2022-04-03 21:21:12,246 epoch 40 - iter 4608/5765 - loss 0.05160290 - samples/sec: 8.72 - lr: 0.050000 2022-04-03 21:25:34,335 epoch 40 - iter 5184/5765 - loss 0.05188003 - samples/sec: 8.83 - lr: 0.050000 2022-04-03 21:30:00,227 epoch 40 - iter 5760/5765 - loss 0.05183257 - samples/sec: 8.70 - lr: 0.050000 2022-04-03 21:30:03,367 ---------------------------------------------------------------------------------------------------- 2022-04-03 21:30:03,370 EPOCH 40 done: loss 0.0519 - lr 0.0500000 2022-04-03 21:36:15,762 DEV : loss 0.05283118411898613 - f1-score (micro avg) 0.828 2022-04-03 21:36:15,836 BAD EPOCHS (no improvement): 0 2022-04-03 21:36:18,064 saving best model 2022-04-03 21:36:29,253 ---------------------------------------------------------------------------------------------------- 2022-04-03 21:36:29,271 loading file /content/gdrive/MyDrive/project/data/ner/model2/best-model.pt 2022-04-03 21:43:00,026 0.8616 0.82 0.8403 0.7357 2022-04-03 21:43:00,030 Results: - F-score (micro) 0.8403 - F-score (macro) 0.8656 - Accuracy 0.7357 By class: precision recall f1-score support LOC 0.8789 0.8589 0.8688 4083 ORG 0.8390 0.7653 0.8005 3166 PER 0.8395 0.8169 0.8280 2741 DAT 0.8648 0.7957 0.8288 1150 MON 0.9758 0.9020 0.9374 357 TIM 0.8500 0.8193 0.8344 166 PCT 0.9615 0.9615 0.9615 156 micro avg 0.8616 0.8200 0.8403 11819 macro avg 0.8871 0.8456 0.8656 11819 weighted avg 0.8613 0.8200 0.8400 11819 samples avg 0.7357 0.7357 0.7357 11819 2022-04-03 21:43:00,035 ----------------------------------------------------------------------------------------------------