2023-10-10 10:17:25,583 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,586 Model: "SequenceTagger( (embeddings): ByT5Embeddings( (model): T5EncoderModel( (shared): Embedding(384, 1472) (encoder): T5Stack( (embed_tokens): Embedding(384, 1472) (block): ModuleList( (0): T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) (relative_attention_bias): Embedding(32, 6) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (1-11): 11 x T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) ) (final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (locked_dropout): LockedDropout(p=0.5) (linear): Linear(in_features=1472, out_features=17, bias=True) (loss_function): CrossEntropyLoss() )" 2023-10-10 10:17:25,586 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,587 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences - NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator 2023-10-10 10:17:25,587 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,587 Train: 20847 sentences 2023-10-10 10:17:25,587 (train_with_dev=False, train_with_test=False) 2023-10-10 10:17:25,587 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,587 Training Params: 2023-10-10 10:17:25,587 - learning_rate: "0.00015" 2023-10-10 10:17:25,587 - mini_batch_size: "8" 2023-10-10 10:17:25,587 - max_epochs: "10" 2023-10-10 10:17:25,587 - shuffle: "True" 2023-10-10 10:17:25,587 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,587 Plugins: 2023-10-10 10:17:25,588 - TensorboardLogger 2023-10-10 10:17:25,588 - LinearScheduler | warmup_fraction: '0.1' 2023-10-10 10:17:25,588 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,588 Final evaluation on model from best epoch (best-model.pt) 2023-10-10 10:17:25,588 - metric: "('micro avg', 'f1-score')" 2023-10-10 10:17:25,588 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,588 Computation: 2023-10-10 10:17:25,588 - compute on device: cuda:0 2023-10-10 10:17:25,588 - embedding storage: none 2023-10-10 10:17:25,588 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,588 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-2" 2023-10-10 10:17:25,588 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,588 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:17:25,589 Logging anything other than scalars to TensorBoard is currently not supported. 2023-10-10 10:19:50,358 epoch 1 - iter 260/2606 - loss 2.82858377 - time (sec): 144.77 - samples/sec: 246.69 - lr: 0.000015 - momentum: 0.000000 2023-10-10 10:22:07,314 epoch 1 - iter 520/2606 - loss 2.59379089 - time (sec): 281.72 - samples/sec: 255.43 - lr: 0.000030 - momentum: 0.000000 2023-10-10 10:24:28,436 epoch 1 - iter 780/2606 - loss 2.18946733 - time (sec): 422.84 - samples/sec: 258.37 - lr: 0.000045 - momentum: 0.000000 2023-10-10 10:26:51,376 epoch 1 - iter 1040/2606 - loss 1.78347214 - time (sec): 565.78 - samples/sec: 262.46 - lr: 0.000060 - momentum: 0.000000 2023-10-10 10:29:09,559 epoch 1 - iter 1300/2606 - loss 1.51733285 - time (sec): 703.97 - samples/sec: 263.06 - lr: 0.000075 - momentum: 0.000000 2023-10-10 10:31:37,873 epoch 1 - iter 1560/2606 - loss 1.33965070 - time (sec): 852.28 - samples/sec: 261.90 - lr: 0.000090 - momentum: 0.000000 2023-10-10 10:33:58,911 epoch 1 - iter 1820/2606 - loss 1.20517898 - time (sec): 993.32 - samples/sec: 261.43 - lr: 0.000105 - momentum: 0.000000 2023-10-10 10:36:19,665 epoch 1 - iter 2080/2606 - loss 1.09396343 - time (sec): 1134.07 - samples/sec: 260.27 - lr: 0.000120 - momentum: 0.000000 2023-10-10 10:38:40,561 epoch 1 - iter 2340/2606 - loss 1.00927714 - time (sec): 1274.97 - samples/sec: 258.05 - lr: 0.000135 - momentum: 0.000000 2023-10-10 10:41:03,325 epoch 1 - iter 2600/2606 - loss 0.92825533 - time (sec): 1417.73 - samples/sec: 258.37 - lr: 0.000150 - momentum: 0.000000 2023-10-10 10:41:06,639 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:41:06,639 EPOCH 1 done: loss 0.9266 - lr: 0.000150 2023-10-10 10:41:45,319 DEV : loss 0.13984432816505432 - f1-score (micro avg) 0.2605 2023-10-10 10:41:45,373 saving best model 2023-10-10 10:41:46,388 ---------------------------------------------------------------------------------------------------- 2023-10-10 10:44:03,782 epoch 2 - iter 260/2606 - loss 0.20048946 - time (sec): 137.39 - samples/sec: 254.87 - lr: 0.000148 - momentum: 0.000000 2023-10-10 10:46:21,256 epoch 2 - iter 520/2606 - loss 0.19474593 - time (sec): 274.86 - samples/sec: 257.90 - lr: 0.000147 - momentum: 0.000000 2023-10-10 10:48:42,738 epoch 2 - iter 780/2606 - loss 0.18435066 - time (sec): 416.35 - samples/sec: 258.36 - lr: 0.000145 - momentum: 0.000000 2023-10-10 10:51:06,768 epoch 2 - iter 1040/2606 - loss 0.17821735 - time (sec): 560.38 - samples/sec: 254.46 - lr: 0.000143 - momentum: 0.000000 2023-10-10 10:53:29,090 epoch 2 - iter 1300/2606 - loss 0.17222389 - time (sec): 702.70 - samples/sec: 255.82 - lr: 0.000142 - momentum: 0.000000 2023-10-10 10:55:53,869 epoch 2 - iter 1560/2606 - loss 0.16461177 - time (sec): 847.48 - samples/sec: 256.54 - lr: 0.000140 - momentum: 0.000000 2023-10-10 10:58:21,709 epoch 2 - iter 1820/2606 - loss 0.16129296 - time (sec): 995.32 - samples/sec: 256.21 - lr: 0.000138 - momentum: 0.000000 2023-10-10 11:00:47,042 epoch 2 - iter 2080/2606 - loss 0.15850392 - time (sec): 1140.65 - samples/sec: 254.82 - lr: 0.000137 - momentum: 0.000000 2023-10-10 11:03:09,654 epoch 2 - iter 2340/2606 - loss 0.15546215 - time (sec): 1283.26 - samples/sec: 254.16 - lr: 0.000135 - momentum: 0.000000 2023-10-10 11:05:34,501 epoch 2 - iter 2600/2606 - loss 0.15151244 - time (sec): 1428.11 - samples/sec: 256.78 - lr: 0.000133 - momentum: 0.000000 2023-10-10 11:05:37,581 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:05:37,582 EPOCH 2 done: loss 0.1514 - lr: 0.000133 2023-10-10 11:06:18,193 DEV : loss 0.12324459105730057 - f1-score (micro avg) 0.3194 2023-10-10 11:06:18,251 saving best model 2023-10-10 11:06:20,967 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:08:38,954 epoch 3 - iter 260/2606 - loss 0.08453196 - time (sec): 137.98 - samples/sec: 257.85 - lr: 0.000132 - momentum: 0.000000 2023-10-10 11:11:00,136 epoch 3 - iter 520/2606 - loss 0.08884281 - time (sec): 279.16 - samples/sec: 259.92 - lr: 0.000130 - momentum: 0.000000 2023-10-10 11:13:22,749 epoch 3 - iter 780/2606 - loss 0.09171291 - time (sec): 421.78 - samples/sec: 259.93 - lr: 0.000128 - momentum: 0.000000 2023-10-10 11:15:41,613 epoch 3 - iter 1040/2606 - loss 0.09501175 - time (sec): 560.64 - samples/sec: 259.52 - lr: 0.000127 - momentum: 0.000000 2023-10-10 11:18:01,854 epoch 3 - iter 1300/2606 - loss 0.09753731 - time (sec): 700.88 - samples/sec: 264.60 - lr: 0.000125 - momentum: 0.000000 2023-10-10 11:20:24,416 epoch 3 - iter 1560/2606 - loss 0.09540397 - time (sec): 843.45 - samples/sec: 264.50 - lr: 0.000123 - momentum: 0.000000 2023-10-10 11:22:44,671 epoch 3 - iter 1820/2606 - loss 0.09431741 - time (sec): 983.70 - samples/sec: 264.06 - lr: 0.000122 - momentum: 0.000000 2023-10-10 11:25:05,740 epoch 3 - iter 2080/2606 - loss 0.09273873 - time (sec): 1124.77 - samples/sec: 263.24 - lr: 0.000120 - momentum: 0.000000 2023-10-10 11:27:19,392 epoch 3 - iter 2340/2606 - loss 0.09140459 - time (sec): 1258.42 - samples/sec: 261.51 - lr: 0.000118 - momentum: 0.000000 2023-10-10 11:29:46,096 epoch 3 - iter 2600/2606 - loss 0.09148228 - time (sec): 1405.13 - samples/sec: 260.87 - lr: 0.000117 - momentum: 0.000000 2023-10-10 11:29:49,569 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:29:49,569 EPOCH 3 done: loss 0.0914 - lr: 0.000117 2023-10-10 11:30:33,329 DEV : loss 0.16454216837882996 - f1-score (micro avg) 0.375 2023-10-10 11:30:33,407 saving best model 2023-10-10 11:30:36,290 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:33:00,647 epoch 4 - iter 260/2606 - loss 0.05238673 - time (sec): 144.35 - samples/sec: 255.20 - lr: 0.000115 - momentum: 0.000000 2023-10-10 11:35:25,933 epoch 4 - iter 520/2606 - loss 0.06195508 - time (sec): 289.64 - samples/sec: 263.40 - lr: 0.000113 - momentum: 0.000000 2023-10-10 11:37:44,551 epoch 4 - iter 780/2606 - loss 0.06048133 - time (sec): 428.26 - samples/sec: 263.11 - lr: 0.000112 - momentum: 0.000000 2023-10-10 11:40:01,800 epoch 4 - iter 1040/2606 - loss 0.06150890 - time (sec): 565.51 - samples/sec: 262.28 - lr: 0.000110 - momentum: 0.000000 2023-10-10 11:42:24,722 epoch 4 - iter 1300/2606 - loss 0.05978791 - time (sec): 708.43 - samples/sec: 261.76 - lr: 0.000108 - momentum: 0.000000 2023-10-10 11:44:44,516 epoch 4 - iter 1560/2606 - loss 0.06398599 - time (sec): 848.22 - samples/sec: 261.95 - lr: 0.000107 - momentum: 0.000000 2023-10-10 11:47:03,146 epoch 4 - iter 1820/2606 - loss 0.06393966 - time (sec): 986.85 - samples/sec: 263.46 - lr: 0.000105 - momentum: 0.000000 2023-10-10 11:49:18,081 epoch 4 - iter 2080/2606 - loss 0.06373753 - time (sec): 1121.79 - samples/sec: 262.88 - lr: 0.000103 - momentum: 0.000000 2023-10-10 11:51:36,877 epoch 4 - iter 2340/2606 - loss 0.06554623 - time (sec): 1260.58 - samples/sec: 262.80 - lr: 0.000102 - momentum: 0.000000 2023-10-10 11:53:53,323 epoch 4 - iter 2600/2606 - loss 0.06560053 - time (sec): 1397.03 - samples/sec: 262.60 - lr: 0.000100 - momentum: 0.000000 2023-10-10 11:53:56,216 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:53:56,216 EPOCH 4 done: loss 0.0656 - lr: 0.000100 2023-10-10 11:54:36,789 DEV : loss 0.26484549045562744 - f1-score (micro avg) 0.3304 2023-10-10 11:54:36,848 ---------------------------------------------------------------------------------------------------- 2023-10-10 11:56:53,975 epoch 5 - iter 260/2606 - loss 0.04025488 - time (sec): 137.12 - samples/sec: 250.50 - lr: 0.000098 - momentum: 0.000000 2023-10-10 11:59:13,175 epoch 5 - iter 520/2606 - loss 0.04564273 - time (sec): 276.32 - samples/sec: 254.62 - lr: 0.000097 - momentum: 0.000000 2023-10-10 12:01:31,458 epoch 5 - iter 780/2606 - loss 0.04191303 - time (sec): 414.61 - samples/sec: 261.94 - lr: 0.000095 - momentum: 0.000000 2023-10-10 12:03:47,948 epoch 5 - iter 1040/2606 - loss 0.04370480 - time (sec): 551.10 - samples/sec: 263.07 - lr: 0.000093 - momentum: 0.000000 2023-10-10 12:06:09,643 epoch 5 - iter 1300/2606 - loss 0.04555479 - time (sec): 692.79 - samples/sec: 263.14 - lr: 0.000092 - momentum: 0.000000 2023-10-10 12:08:27,869 epoch 5 - iter 1560/2606 - loss 0.04638323 - time (sec): 831.02 - samples/sec: 263.60 - lr: 0.000090 - momentum: 0.000000 2023-10-10 12:10:44,931 epoch 5 - iter 1820/2606 - loss 0.04605700 - time (sec): 968.08 - samples/sec: 262.91 - lr: 0.000088 - momentum: 0.000000 2023-10-10 12:13:06,008 epoch 5 - iter 2080/2606 - loss 0.04618736 - time (sec): 1109.16 - samples/sec: 264.32 - lr: 0.000087 - momentum: 0.000000 2023-10-10 12:15:28,125 epoch 5 - iter 2340/2606 - loss 0.04664304 - time (sec): 1251.27 - samples/sec: 264.95 - lr: 0.000085 - momentum: 0.000000 2023-10-10 12:17:46,232 epoch 5 - iter 2600/2606 - loss 0.04668505 - time (sec): 1389.38 - samples/sec: 263.95 - lr: 0.000083 - momentum: 0.000000 2023-10-10 12:17:49,292 ---------------------------------------------------------------------------------------------------- 2023-10-10 12:17:49,292 EPOCH 5 done: loss 0.0467 - lr: 0.000083 2023-10-10 12:18:32,521 DEV : loss 0.29650354385375977 - f1-score (micro avg) 0.3622 2023-10-10 12:18:32,581 ---------------------------------------------------------------------------------------------------- 2023-10-10 12:20:49,821 epoch 6 - iter 260/2606 - loss 0.03241520 - time (sec): 137.24 - samples/sec: 254.95 - lr: 0.000082 - momentum: 0.000000 2023-10-10 12:23:05,715 epoch 6 - iter 520/2606 - loss 0.02994548 - time (sec): 273.13 - samples/sec: 256.33 - lr: 0.000080 - momentum: 0.000000 2023-10-10 12:25:25,839 epoch 6 - iter 780/2606 - loss 0.03207625 - time (sec): 413.25 - samples/sec: 261.76 - lr: 0.000078 - momentum: 0.000000 2023-10-10 12:27:48,917 epoch 6 - iter 1040/2606 - loss 0.03228089 - time (sec): 556.33 - samples/sec: 263.64 - lr: 0.000077 - momentum: 0.000000 2023-10-10 12:30:11,973 epoch 6 - iter 1300/2606 - loss 0.03163962 - time (sec): 699.39 - samples/sec: 265.05 - lr: 0.000075 - momentum: 0.000000 2023-10-10 12:32:30,584 epoch 6 - iter 1560/2606 - loss 0.03281868 - time (sec): 838.00 - samples/sec: 263.89 - lr: 0.000073 - momentum: 0.000000 2023-10-10 12:34:50,290 epoch 6 - iter 1820/2606 - loss 0.03179360 - time (sec): 977.71 - samples/sec: 263.53 - lr: 0.000072 - momentum: 0.000000 2023-10-10 12:37:07,793 epoch 6 - iter 2080/2606 - loss 0.03208471 - time (sec): 1115.21 - samples/sec: 264.32 - lr: 0.000070 - momentum: 0.000000 2023-10-10 12:39:24,116 epoch 6 - iter 2340/2606 - loss 0.03214342 - time (sec): 1251.53 - samples/sec: 263.15 - lr: 0.000068 - momentum: 0.000000 2023-10-10 12:41:43,053 epoch 6 - iter 2600/2606 - loss 0.03446797 - time (sec): 1390.47 - samples/sec: 263.73 - lr: 0.000067 - momentum: 0.000000 2023-10-10 12:41:45,990 ---------------------------------------------------------------------------------------------------- 2023-10-10 12:41:45,991 EPOCH 6 done: loss 0.0344 - lr: 0.000067 2023-10-10 12:42:28,894 DEV : loss 0.33373507857322693 - f1-score (micro avg) 0.378 2023-10-10 12:42:28,972 saving best model 2023-10-10 12:42:31,708 ---------------------------------------------------------------------------------------------------- 2023-10-10 12:44:49,513 epoch 7 - iter 260/2606 - loss 0.02101641 - time (sec): 137.80 - samples/sec: 257.43 - lr: 0.000065 - momentum: 0.000000 2023-10-10 12:47:08,074 epoch 7 - iter 520/2606 - loss 0.01944662 - time (sec): 276.36 - samples/sec: 258.39 - lr: 0.000063 - momentum: 0.000000 2023-10-10 12:49:26,954 epoch 7 - iter 780/2606 - loss 0.02172094 - time (sec): 415.24 - samples/sec: 259.97 - lr: 0.000062 - momentum: 0.000000 2023-10-10 12:51:45,464 epoch 7 - iter 1040/2606 - loss 0.02238459 - time (sec): 553.75 - samples/sec: 260.69 - lr: 0.000060 - momentum: 0.000000 2023-10-10 12:54:07,890 epoch 7 - iter 1300/2606 - loss 0.02535700 - time (sec): 696.18 - samples/sec: 262.29 - lr: 0.000058 - momentum: 0.000000 2023-10-10 12:56:26,291 epoch 7 - iter 1560/2606 - loss 0.02596897 - time (sec): 834.58 - samples/sec: 262.39 - lr: 0.000057 - momentum: 0.000000 2023-10-10 12:58:45,212 epoch 7 - iter 1820/2606 - loss 0.02683384 - time (sec): 973.50 - samples/sec: 262.42 - lr: 0.000055 - momentum: 0.000000 2023-10-10 13:01:01,249 epoch 7 - iter 2080/2606 - loss 0.02612235 - time (sec): 1109.54 - samples/sec: 260.72 - lr: 0.000053 - momentum: 0.000000 2023-10-10 13:03:21,969 epoch 7 - iter 2340/2606 - loss 0.02627648 - time (sec): 1250.26 - samples/sec: 261.53 - lr: 0.000052 - momentum: 0.000000 2023-10-10 13:05:45,810 epoch 7 - iter 2600/2606 - loss 0.02563448 - time (sec): 1394.10 - samples/sec: 262.94 - lr: 0.000050 - momentum: 0.000000 2023-10-10 13:05:48,990 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:05:48,990 EPOCH 7 done: loss 0.0256 - lr: 0.000050 2023-10-10 13:06:31,953 DEV : loss 0.3978530466556549 - f1-score (micro avg) 0.3722 2023-10-10 13:06:32,018 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:08:55,328 epoch 8 - iter 260/2606 - loss 0.01412281 - time (sec): 143.31 - samples/sec: 277.93 - lr: 0.000048 - momentum: 0.000000 2023-10-10 13:11:15,605 epoch 8 - iter 520/2606 - loss 0.01664019 - time (sec): 283.58 - samples/sec: 270.88 - lr: 0.000047 - momentum: 0.000000 2023-10-10 13:13:32,923 epoch 8 - iter 780/2606 - loss 0.01697467 - time (sec): 420.90 - samples/sec: 266.21 - lr: 0.000045 - momentum: 0.000000 2023-10-10 13:15:54,238 epoch 8 - iter 1040/2606 - loss 0.01697407 - time (sec): 562.22 - samples/sec: 265.97 - lr: 0.000043 - momentum: 0.000000 2023-10-10 13:18:13,799 epoch 8 - iter 1300/2606 - loss 0.01628044 - time (sec): 701.78 - samples/sec: 263.85 - lr: 0.000042 - momentum: 0.000000 2023-10-10 13:20:35,746 epoch 8 - iter 1560/2606 - loss 0.01692923 - time (sec): 843.73 - samples/sec: 263.69 - lr: 0.000040 - momentum: 0.000000 2023-10-10 13:22:55,645 epoch 8 - iter 1820/2606 - loss 0.01698632 - time (sec): 983.62 - samples/sec: 263.30 - lr: 0.000038 - momentum: 0.000000 2023-10-10 13:25:13,795 epoch 8 - iter 2080/2606 - loss 0.01667188 - time (sec): 1121.77 - samples/sec: 261.61 - lr: 0.000037 - momentum: 0.000000 2023-10-10 13:27:29,647 epoch 8 - iter 2340/2606 - loss 0.01736529 - time (sec): 1257.63 - samples/sec: 260.13 - lr: 0.000035 - momentum: 0.000000 2023-10-10 13:29:50,730 epoch 8 - iter 2600/2606 - loss 0.01789555 - time (sec): 1398.71 - samples/sec: 262.23 - lr: 0.000033 - momentum: 0.000000 2023-10-10 13:29:53,639 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:29:53,640 EPOCH 8 done: loss 0.0179 - lr: 0.000033 2023-10-10 13:30:36,471 DEV : loss 0.40461236238479614 - f1-score (micro avg) 0.3935 2023-10-10 13:30:36,533 saving best model 2023-10-10 13:30:39,231 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:32:59,878 epoch 9 - iter 260/2606 - loss 0.01216931 - time (sec): 140.64 - samples/sec: 263.28 - lr: 0.000032 - momentum: 0.000000 2023-10-10 13:35:25,145 epoch 9 - iter 520/2606 - loss 0.01085860 - time (sec): 285.91 - samples/sec: 270.57 - lr: 0.000030 - momentum: 0.000000 2023-10-10 13:37:44,303 epoch 9 - iter 780/2606 - loss 0.01126911 - time (sec): 425.07 - samples/sec: 266.36 - lr: 0.000028 - momentum: 0.000000 2023-10-10 13:39:59,037 epoch 9 - iter 1040/2606 - loss 0.01140504 - time (sec): 559.80 - samples/sec: 261.97 - lr: 0.000027 - momentum: 0.000000 2023-10-10 13:42:20,737 epoch 9 - iter 1300/2606 - loss 0.01208397 - time (sec): 701.50 - samples/sec: 264.42 - lr: 0.000025 - momentum: 0.000000 2023-10-10 13:44:42,081 epoch 9 - iter 1560/2606 - loss 0.01245932 - time (sec): 842.85 - samples/sec: 261.13 - lr: 0.000023 - momentum: 0.000000 2023-10-10 13:46:58,383 epoch 9 - iter 1820/2606 - loss 0.01254583 - time (sec): 979.15 - samples/sec: 260.67 - lr: 0.000022 - momentum: 0.000000 2023-10-10 13:49:17,990 epoch 9 - iter 2080/2606 - loss 0.01291335 - time (sec): 1118.75 - samples/sec: 260.52 - lr: 0.000020 - momentum: 0.000000 2023-10-10 13:51:38,722 epoch 9 - iter 2340/2606 - loss 0.01272261 - time (sec): 1259.49 - samples/sec: 260.05 - lr: 0.000018 - momentum: 0.000000 2023-10-10 13:53:58,619 epoch 9 - iter 2600/2606 - loss 0.01295044 - time (sec): 1399.38 - samples/sec: 262.18 - lr: 0.000017 - momentum: 0.000000 2023-10-10 13:54:01,480 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:54:01,480 EPOCH 9 done: loss 0.0129 - lr: 0.000017 2023-10-10 13:54:42,752 DEV : loss 0.4415739178657532 - f1-score (micro avg) 0.3754 2023-10-10 13:54:42,814 ---------------------------------------------------------------------------------------------------- 2023-10-10 13:57:04,272 epoch 10 - iter 260/2606 - loss 0.00790204 - time (sec): 141.46 - samples/sec: 270.22 - lr: 0.000015 - momentum: 0.000000 2023-10-10 13:59:26,749 epoch 10 - iter 520/2606 - loss 0.00758597 - time (sec): 283.93 - samples/sec: 270.20 - lr: 0.000013 - momentum: 0.000000 2023-10-10 14:01:45,507 epoch 10 - iter 780/2606 - loss 0.00856811 - time (sec): 422.69 - samples/sec: 266.23 - lr: 0.000012 - momentum: 0.000000 2023-10-10 14:04:03,408 epoch 10 - iter 1040/2606 - loss 0.00841758 - time (sec): 560.59 - samples/sec: 259.01 - lr: 0.000010 - momentum: 0.000000 2023-10-10 14:06:23,118 epoch 10 - iter 1300/2606 - loss 0.00819062 - time (sec): 700.30 - samples/sec: 260.67 - lr: 0.000008 - momentum: 0.000000 2023-10-10 14:08:41,954 epoch 10 - iter 1560/2606 - loss 0.00817874 - time (sec): 839.14 - samples/sec: 260.44 - lr: 0.000007 - momentum: 0.000000 2023-10-10 14:11:00,012 epoch 10 - iter 1820/2606 - loss 0.00826103 - time (sec): 977.20 - samples/sec: 261.61 - lr: 0.000005 - momentum: 0.000000 2023-10-10 14:13:22,211 epoch 10 - iter 2080/2606 - loss 0.00858032 - time (sec): 1119.40 - samples/sec: 263.35 - lr: 0.000003 - momentum: 0.000000 2023-10-10 14:15:40,101 epoch 10 - iter 2340/2606 - loss 0.00841120 - time (sec): 1257.28 - samples/sec: 262.99 - lr: 0.000002 - momentum: 0.000000 2023-10-10 14:17:58,169 epoch 10 - iter 2600/2606 - loss 0.00852789 - time (sec): 1395.35 - samples/sec: 262.65 - lr: 0.000000 - momentum: 0.000000 2023-10-10 14:18:01,398 ---------------------------------------------------------------------------------------------------- 2023-10-10 14:18:01,398 EPOCH 10 done: loss 0.0085 - lr: 0.000000 2023-10-10 14:18:43,420 DEV : loss 0.46130311489105225 - f1-score (micro avg) 0.3804 2023-10-10 14:18:44,392 ---------------------------------------------------------------------------------------------------- 2023-10-10 14:18:44,394 Loading model from best epoch ... 2023-10-10 14:18:48,948 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd 2023-10-10 14:20:38,998 Results: - F-score (micro) 0.4462 - F-score (macro) 0.3145 - Accuracy 0.2916 By class: precision recall f1-score support LOC 0.4555 0.5231 0.4870 1214 PER 0.4229 0.4517 0.4369 808 ORG 0.3343 0.3343 0.3343 353 HumanProd 0.0000 0.0000 0.0000 15 micro avg 0.4266 0.4678 0.4462 2390 macro avg 0.3032 0.3273 0.3145 2390 weighted avg 0.4237 0.4678 0.4444 2390 2023-10-10 14:20:38,999 ----------------------------------------------------------------------------------------------------