File size: 24,064 Bytes
dcccb85 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 |
2023-10-15 19:50:34,450 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(32001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 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
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Train: 20847 sentences
2023-10-15 19:50:34,451 (train_with_dev=False, train_with_test=False)
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Training Params:
2023-10-15 19:50:34,451 - learning_rate: "5e-05"
2023-10-15 19:50:34,451 - mini_batch_size: "8"
2023-10-15 19:50:34,451 - max_epochs: "10"
2023-10-15 19:50:34,451 - shuffle: "True"
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Plugins:
2023-10-15 19:50:34,451 - LinearScheduler | warmup_fraction: '0.1'
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Final evaluation on model from best epoch (best-model.pt)
2023-10-15 19:50:34,451 - metric: "('micro avg', 'f1-score')"
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,451 Computation:
2023-10-15 19:50:34,451 - compute on device: cuda:0
2023-10-15 19:50:34,451 - embedding storage: none
2023-10-15 19:50:34,451 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,452 Model training base path: "hmbench-newseye/de-dbmdz/bert-base-historic-multilingual-cased-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-4"
2023-10-15 19:50:34,452 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:34,452 ----------------------------------------------------------------------------------------------------
2023-10-15 19:50:52,747 epoch 1 - iter 260/2606 - loss 1.59028644 - time (sec): 18.29 - samples/sec: 1909.58 - lr: 0.000005 - momentum: 0.000000
2023-10-15 19:51:12,202 epoch 1 - iter 520/2606 - loss 0.98518790 - time (sec): 37.75 - samples/sec: 1945.90 - lr: 0.000010 - momentum: 0.000000
2023-10-15 19:51:30,947 epoch 1 - iter 780/2606 - loss 0.76092996 - time (sec): 56.49 - samples/sec: 1924.76 - lr: 0.000015 - momentum: 0.000000
2023-10-15 19:51:50,886 epoch 1 - iter 1040/2606 - loss 0.63417270 - time (sec): 76.43 - samples/sec: 1893.42 - lr: 0.000020 - momentum: 0.000000
2023-10-15 19:52:10,166 epoch 1 - iter 1300/2606 - loss 0.55309956 - time (sec): 95.71 - samples/sec: 1915.70 - lr: 0.000025 - momentum: 0.000000
2023-10-15 19:52:29,070 epoch 1 - iter 1560/2606 - loss 0.49758201 - time (sec): 114.62 - samples/sec: 1905.72 - lr: 0.000030 - momentum: 0.000000
2023-10-15 19:52:48,118 epoch 1 - iter 1820/2606 - loss 0.45440578 - time (sec): 133.67 - samples/sec: 1904.65 - lr: 0.000035 - momentum: 0.000000
2023-10-15 19:53:07,758 epoch 1 - iter 2080/2606 - loss 0.41855391 - time (sec): 153.30 - samples/sec: 1900.97 - lr: 0.000040 - momentum: 0.000000
2023-10-15 19:53:26,472 epoch 1 - iter 2340/2606 - loss 0.39560327 - time (sec): 172.02 - samples/sec: 1905.23 - lr: 0.000045 - momentum: 0.000000
2023-10-15 19:53:46,488 epoch 1 - iter 2600/2606 - loss 0.37125077 - time (sec): 192.03 - samples/sec: 1909.56 - lr: 0.000050 - momentum: 0.000000
2023-10-15 19:53:46,912 ----------------------------------------------------------------------------------------------------
2023-10-15 19:53:46,912 EPOCH 1 done: loss 0.3708 - lr: 0.000050
2023-10-15 19:53:53,679 DEV : loss 0.09866511821746826 - f1-score (micro avg) 0.332
2023-10-15 19:53:53,707 saving best model
2023-10-15 19:53:54,083 ----------------------------------------------------------------------------------------------------
2023-10-15 19:54:13,897 epoch 2 - iter 260/2606 - loss 0.16979246 - time (sec): 19.81 - samples/sec: 1913.91 - lr: 0.000049 - momentum: 0.000000
2023-10-15 19:54:33,718 epoch 2 - iter 520/2606 - loss 0.15396119 - time (sec): 39.63 - samples/sec: 1921.04 - lr: 0.000049 - momentum: 0.000000
2023-10-15 19:54:52,579 epoch 2 - iter 780/2606 - loss 0.15278405 - time (sec): 58.49 - samples/sec: 1938.50 - lr: 0.000048 - momentum: 0.000000
2023-10-15 19:55:11,345 epoch 2 - iter 1040/2606 - loss 0.15627393 - time (sec): 77.26 - samples/sec: 1914.12 - lr: 0.000048 - momentum: 0.000000
2023-10-15 19:55:30,228 epoch 2 - iter 1300/2606 - loss 0.15531609 - time (sec): 96.14 - samples/sec: 1926.26 - lr: 0.000047 - momentum: 0.000000
2023-10-15 19:55:48,789 epoch 2 - iter 1560/2606 - loss 0.15197189 - time (sec): 114.70 - samples/sec: 1925.11 - lr: 0.000047 - momentum: 0.000000
2023-10-15 19:56:08,319 epoch 2 - iter 1820/2606 - loss 0.14969900 - time (sec): 134.23 - samples/sec: 1933.14 - lr: 0.000046 - momentum: 0.000000
2023-10-15 19:56:26,763 epoch 2 - iter 2080/2606 - loss 0.15359140 - time (sec): 152.68 - samples/sec: 1928.53 - lr: 0.000046 - momentum: 0.000000
2023-10-15 19:56:46,738 epoch 2 - iter 2340/2606 - loss 0.15309562 - time (sec): 172.65 - samples/sec: 1927.59 - lr: 0.000045 - momentum: 0.000000
2023-10-15 19:57:05,105 epoch 2 - iter 2600/2606 - loss 0.15242289 - time (sec): 191.02 - samples/sec: 1921.72 - lr: 0.000044 - momentum: 0.000000
2023-10-15 19:57:05,409 ----------------------------------------------------------------------------------------------------
2023-10-15 19:57:05,409 EPOCH 2 done: loss 0.1524 - lr: 0.000044
2023-10-15 19:57:15,626 DEV : loss 0.1584077775478363 - f1-score (micro avg) 0.3406
2023-10-15 19:57:15,657 saving best model
2023-10-15 19:57:16,211 ----------------------------------------------------------------------------------------------------
2023-10-15 19:57:35,187 epoch 3 - iter 260/2606 - loss 0.11299931 - time (sec): 18.97 - samples/sec: 1921.95 - lr: 0.000044 - momentum: 0.000000
2023-10-15 19:57:54,309 epoch 3 - iter 520/2606 - loss 0.10767341 - time (sec): 38.09 - samples/sec: 1918.22 - lr: 0.000043 - momentum: 0.000000
2023-10-15 19:58:15,019 epoch 3 - iter 780/2606 - loss 0.10899205 - time (sec): 58.80 - samples/sec: 1875.68 - lr: 0.000043 - momentum: 0.000000
2023-10-15 19:58:35,731 epoch 3 - iter 1040/2606 - loss 0.11063881 - time (sec): 79.52 - samples/sec: 1865.12 - lr: 0.000042 - momentum: 0.000000
2023-10-15 19:58:55,613 epoch 3 - iter 1300/2606 - loss 0.10698292 - time (sec): 99.40 - samples/sec: 1857.48 - lr: 0.000042 - momentum: 0.000000
2023-10-15 19:59:14,663 epoch 3 - iter 1560/2606 - loss 0.10679134 - time (sec): 118.45 - samples/sec: 1858.88 - lr: 0.000041 - momentum: 0.000000
2023-10-15 19:59:33,448 epoch 3 - iter 1820/2606 - loss 0.10893641 - time (sec): 137.23 - samples/sec: 1867.67 - lr: 0.000041 - momentum: 0.000000
2023-10-15 19:59:53,465 epoch 3 - iter 2080/2606 - loss 0.10710962 - time (sec): 157.25 - samples/sec: 1877.55 - lr: 0.000040 - momentum: 0.000000
2023-10-15 20:00:12,483 epoch 3 - iter 2340/2606 - loss 0.10687887 - time (sec): 176.27 - samples/sec: 1883.24 - lr: 0.000039 - momentum: 0.000000
2023-10-15 20:00:30,949 epoch 3 - iter 2600/2606 - loss 0.10778076 - time (sec): 194.74 - samples/sec: 1882.60 - lr: 0.000039 - momentum: 0.000000
2023-10-15 20:00:31,324 ----------------------------------------------------------------------------------------------------
2023-10-15 20:00:31,324 EPOCH 3 done: loss 0.1078 - lr: 0.000039
2023-10-15 20:00:40,591 DEV : loss 0.15240149199962616 - f1-score (micro avg) 0.4084
2023-10-15 20:00:40,626 saving best model
2023-10-15 20:00:41,208 ----------------------------------------------------------------------------------------------------
2023-10-15 20:01:00,629 epoch 4 - iter 260/2606 - loss 0.08035620 - time (sec): 19.41 - samples/sec: 1887.95 - lr: 0.000038 - momentum: 0.000000
2023-10-15 20:01:18,964 epoch 4 - iter 520/2606 - loss 0.07715406 - time (sec): 37.75 - samples/sec: 1884.39 - lr: 0.000038 - momentum: 0.000000
2023-10-15 20:01:37,159 epoch 4 - iter 780/2606 - loss 0.08039817 - time (sec): 55.94 - samples/sec: 1918.31 - lr: 0.000037 - momentum: 0.000000
2023-10-15 20:01:56,554 epoch 4 - iter 1040/2606 - loss 0.07833131 - time (sec): 75.34 - samples/sec: 1914.94 - lr: 0.000037 - momentum: 0.000000
2023-10-15 20:02:14,628 epoch 4 - iter 1300/2606 - loss 0.07761709 - time (sec): 93.41 - samples/sec: 1924.78 - lr: 0.000036 - momentum: 0.000000
2023-10-15 20:02:32,876 epoch 4 - iter 1560/2606 - loss 0.07793021 - time (sec): 111.66 - samples/sec: 1930.25 - lr: 0.000036 - momentum: 0.000000
2023-10-15 20:02:52,296 epoch 4 - iter 1820/2606 - loss 0.07911638 - time (sec): 131.08 - samples/sec: 1936.60 - lr: 0.000035 - momentum: 0.000000
2023-10-15 20:03:10,776 epoch 4 - iter 2080/2606 - loss 0.07888385 - time (sec): 149.56 - samples/sec: 1932.16 - lr: 0.000034 - momentum: 0.000000
2023-10-15 20:03:29,996 epoch 4 - iter 2340/2606 - loss 0.07940196 - time (sec): 168.78 - samples/sec: 1941.28 - lr: 0.000034 - momentum: 0.000000
2023-10-15 20:03:50,163 epoch 4 - iter 2600/2606 - loss 0.07772595 - time (sec): 188.95 - samples/sec: 1939.73 - lr: 0.000033 - momentum: 0.000000
2023-10-15 20:03:50,679 ----------------------------------------------------------------------------------------------------
2023-10-15 20:03:50,679 EPOCH 4 done: loss 0.0776 - lr: 0.000033
2023-10-15 20:03:59,788 DEV : loss 0.24223537743091583 - f1-score (micro avg) 0.3895
2023-10-15 20:03:59,816 ----------------------------------------------------------------------------------------------------
2023-10-15 20:04:17,856 epoch 5 - iter 260/2606 - loss 0.05320156 - time (sec): 18.04 - samples/sec: 1901.00 - lr: 0.000033 - momentum: 0.000000
2023-10-15 20:04:36,525 epoch 5 - iter 520/2606 - loss 0.06004399 - time (sec): 36.71 - samples/sec: 1892.24 - lr: 0.000032 - momentum: 0.000000
2023-10-15 20:04:55,064 epoch 5 - iter 780/2606 - loss 0.06361788 - time (sec): 55.25 - samples/sec: 1900.04 - lr: 0.000032 - momentum: 0.000000
2023-10-15 20:05:13,980 epoch 5 - iter 1040/2606 - loss 0.06308322 - time (sec): 74.16 - samples/sec: 1916.92 - lr: 0.000031 - momentum: 0.000000
2023-10-15 20:05:33,591 epoch 5 - iter 1300/2606 - loss 0.06120360 - time (sec): 93.77 - samples/sec: 1917.50 - lr: 0.000031 - momentum: 0.000000
2023-10-15 20:05:52,267 epoch 5 - iter 1560/2606 - loss 0.06016703 - time (sec): 112.45 - samples/sec: 1913.29 - lr: 0.000030 - momentum: 0.000000
2023-10-15 20:06:11,407 epoch 5 - iter 1820/2606 - loss 0.06035800 - time (sec): 131.59 - samples/sec: 1920.61 - lr: 0.000029 - momentum: 0.000000
2023-10-15 20:06:30,790 epoch 5 - iter 2080/2606 - loss 0.05910496 - time (sec): 150.97 - samples/sec: 1928.30 - lr: 0.000029 - momentum: 0.000000
2023-10-15 20:06:50,194 epoch 5 - iter 2340/2606 - loss 0.05814657 - time (sec): 170.38 - samples/sec: 1929.74 - lr: 0.000028 - momentum: 0.000000
2023-10-15 20:07:09,957 epoch 5 - iter 2600/2606 - loss 0.05838664 - time (sec): 190.14 - samples/sec: 1927.76 - lr: 0.000028 - momentum: 0.000000
2023-10-15 20:07:10,400 ----------------------------------------------------------------------------------------------------
2023-10-15 20:07:10,401 EPOCH 5 done: loss 0.0584 - lr: 0.000028
2023-10-15 20:07:18,725 DEV : loss 0.3234298527240753 - f1-score (micro avg) 0.3501
2023-10-15 20:07:18,755 ----------------------------------------------------------------------------------------------------
2023-10-15 20:07:38,043 epoch 6 - iter 260/2606 - loss 0.04968909 - time (sec): 19.29 - samples/sec: 1943.44 - lr: 0.000027 - momentum: 0.000000
2023-10-15 20:07:58,295 epoch 6 - iter 520/2606 - loss 0.04591580 - time (sec): 39.54 - samples/sec: 1911.03 - lr: 0.000027 - momentum: 0.000000
2023-10-15 20:08:16,820 epoch 6 - iter 780/2606 - loss 0.04497115 - time (sec): 58.06 - samples/sec: 1917.34 - lr: 0.000026 - momentum: 0.000000
2023-10-15 20:08:36,677 epoch 6 - iter 1040/2606 - loss 0.04222920 - time (sec): 77.92 - samples/sec: 1932.37 - lr: 0.000026 - momentum: 0.000000
2023-10-15 20:08:55,221 epoch 6 - iter 1300/2606 - loss 0.04176238 - time (sec): 96.46 - samples/sec: 1935.86 - lr: 0.000025 - momentum: 0.000000
2023-10-15 20:09:14,166 epoch 6 - iter 1560/2606 - loss 0.04201919 - time (sec): 115.41 - samples/sec: 1930.35 - lr: 0.000024 - momentum: 0.000000
2023-10-15 20:09:32,322 epoch 6 - iter 1820/2606 - loss 0.04211214 - time (sec): 133.57 - samples/sec: 1928.39 - lr: 0.000024 - momentum: 0.000000
2023-10-15 20:09:51,338 epoch 6 - iter 2080/2606 - loss 0.04276177 - time (sec): 152.58 - samples/sec: 1919.57 - lr: 0.000023 - momentum: 0.000000
2023-10-15 20:10:09,800 epoch 6 - iter 2340/2606 - loss 0.04345221 - time (sec): 171.04 - samples/sec: 1915.62 - lr: 0.000023 - momentum: 0.000000
2023-10-15 20:10:29,580 epoch 6 - iter 2600/2606 - loss 0.04434131 - time (sec): 190.82 - samples/sec: 1918.41 - lr: 0.000022 - momentum: 0.000000
2023-10-15 20:10:30,220 ----------------------------------------------------------------------------------------------------
2023-10-15 20:10:30,220 EPOCH 6 done: loss 0.0442 - lr: 0.000022
2023-10-15 20:10:38,461 DEV : loss 0.360347181558609 - f1-score (micro avg) 0.3831
2023-10-15 20:10:38,489 ----------------------------------------------------------------------------------------------------
2023-10-15 20:10:57,550 epoch 7 - iter 260/2606 - loss 0.03605703 - time (sec): 19.06 - samples/sec: 1967.84 - lr: 0.000022 - momentum: 0.000000
2023-10-15 20:11:17,462 epoch 7 - iter 520/2606 - loss 0.03233024 - time (sec): 38.97 - samples/sec: 1942.69 - lr: 0.000021 - momentum: 0.000000
2023-10-15 20:11:36,061 epoch 7 - iter 780/2606 - loss 0.03422854 - time (sec): 57.57 - samples/sec: 1930.25 - lr: 0.000021 - momentum: 0.000000
2023-10-15 20:11:55,421 epoch 7 - iter 1040/2606 - loss 0.03509936 - time (sec): 76.93 - samples/sec: 1887.52 - lr: 0.000020 - momentum: 0.000000
2023-10-15 20:12:15,268 epoch 7 - iter 1300/2606 - loss 0.03481716 - time (sec): 96.78 - samples/sec: 1892.73 - lr: 0.000019 - momentum: 0.000000
2023-10-15 20:12:33,527 epoch 7 - iter 1560/2606 - loss 0.03453842 - time (sec): 115.04 - samples/sec: 1897.80 - lr: 0.000019 - momentum: 0.000000
2023-10-15 20:12:53,305 epoch 7 - iter 1820/2606 - loss 0.03453028 - time (sec): 134.81 - samples/sec: 1910.27 - lr: 0.000018 - momentum: 0.000000
2023-10-15 20:13:11,312 epoch 7 - iter 2080/2606 - loss 0.03351666 - time (sec): 152.82 - samples/sec: 1910.11 - lr: 0.000018 - momentum: 0.000000
2023-10-15 20:13:30,700 epoch 7 - iter 2340/2606 - loss 0.03294077 - time (sec): 172.21 - samples/sec: 1917.56 - lr: 0.000017 - momentum: 0.000000
2023-10-15 20:13:49,306 epoch 7 - iter 2600/2606 - loss 0.03257111 - time (sec): 190.82 - samples/sec: 1920.47 - lr: 0.000017 - momentum: 0.000000
2023-10-15 20:13:49,775 ----------------------------------------------------------------------------------------------------
2023-10-15 20:13:49,775 EPOCH 7 done: loss 0.0326 - lr: 0.000017
2023-10-15 20:13:57,989 DEV : loss 0.37199294567108154 - f1-score (micro avg) 0.3858
2023-10-15 20:13:58,018 ----------------------------------------------------------------------------------------------------
2023-10-15 20:14:16,046 epoch 8 - iter 260/2606 - loss 0.02342308 - time (sec): 18.03 - samples/sec: 1889.83 - lr: 0.000016 - momentum: 0.000000
2023-10-15 20:14:35,427 epoch 8 - iter 520/2606 - loss 0.02364228 - time (sec): 37.41 - samples/sec: 1945.44 - lr: 0.000016 - momentum: 0.000000
2023-10-15 20:14:54,448 epoch 8 - iter 780/2606 - loss 0.02356941 - time (sec): 56.43 - samples/sec: 1926.14 - lr: 0.000015 - momentum: 0.000000
2023-10-15 20:15:13,454 epoch 8 - iter 1040/2606 - loss 0.02491977 - time (sec): 75.44 - samples/sec: 1918.27 - lr: 0.000014 - momentum: 0.000000
2023-10-15 20:15:32,549 epoch 8 - iter 1300/2606 - loss 0.02428659 - time (sec): 94.53 - samples/sec: 1932.58 - lr: 0.000014 - momentum: 0.000000
2023-10-15 20:15:52,160 epoch 8 - iter 1560/2606 - loss 0.02366101 - time (sec): 114.14 - samples/sec: 1931.16 - lr: 0.000013 - momentum: 0.000000
2023-10-15 20:16:10,784 epoch 8 - iter 1820/2606 - loss 0.02343244 - time (sec): 132.77 - samples/sec: 1938.37 - lr: 0.000013 - momentum: 0.000000
2023-10-15 20:16:30,328 epoch 8 - iter 2080/2606 - loss 0.02351442 - time (sec): 152.31 - samples/sec: 1925.81 - lr: 0.000012 - momentum: 0.000000
2023-10-15 20:16:49,792 epoch 8 - iter 2340/2606 - loss 0.02300749 - time (sec): 171.77 - samples/sec: 1923.52 - lr: 0.000012 - momentum: 0.000000
2023-10-15 20:17:08,499 epoch 8 - iter 2600/2606 - loss 0.02311972 - time (sec): 190.48 - samples/sec: 1924.21 - lr: 0.000011 - momentum: 0.000000
2023-10-15 20:17:08,968 ----------------------------------------------------------------------------------------------------
2023-10-15 20:17:08,968 EPOCH 8 done: loss 0.0231 - lr: 0.000011
2023-10-15 20:17:17,209 DEV : loss 0.38435670733451843 - f1-score (micro avg) 0.3878
2023-10-15 20:17:17,237 ----------------------------------------------------------------------------------------------------
2023-10-15 20:17:36,880 epoch 9 - iter 260/2606 - loss 0.01429002 - time (sec): 19.64 - samples/sec: 2021.11 - lr: 0.000011 - momentum: 0.000000
2023-10-15 20:17:56,464 epoch 9 - iter 520/2606 - loss 0.01483258 - time (sec): 39.23 - samples/sec: 1987.77 - lr: 0.000010 - momentum: 0.000000
2023-10-15 20:18:15,715 epoch 9 - iter 780/2606 - loss 0.01447465 - time (sec): 58.48 - samples/sec: 1973.65 - lr: 0.000009 - momentum: 0.000000
2023-10-15 20:18:34,051 epoch 9 - iter 1040/2606 - loss 0.01472942 - time (sec): 76.81 - samples/sec: 1971.06 - lr: 0.000009 - momentum: 0.000000
2023-10-15 20:18:52,832 epoch 9 - iter 1300/2606 - loss 0.01528011 - time (sec): 95.59 - samples/sec: 1965.67 - lr: 0.000008 - momentum: 0.000000
2023-10-15 20:19:10,916 epoch 9 - iter 1560/2606 - loss 0.01518566 - time (sec): 113.68 - samples/sec: 1940.69 - lr: 0.000008 - momentum: 0.000000
2023-10-15 20:19:30,000 epoch 9 - iter 1820/2606 - loss 0.01575329 - time (sec): 132.76 - samples/sec: 1940.91 - lr: 0.000007 - momentum: 0.000000
2023-10-15 20:19:48,598 epoch 9 - iter 2080/2606 - loss 0.01545732 - time (sec): 151.36 - samples/sec: 1943.70 - lr: 0.000007 - momentum: 0.000000
2023-10-15 20:20:07,608 epoch 9 - iter 2340/2606 - loss 0.01535373 - time (sec): 170.37 - samples/sec: 1939.88 - lr: 0.000006 - momentum: 0.000000
2023-10-15 20:20:26,963 epoch 9 - iter 2600/2606 - loss 0.01534271 - time (sec): 189.72 - samples/sec: 1933.83 - lr: 0.000006 - momentum: 0.000000
2023-10-15 20:20:27,294 ----------------------------------------------------------------------------------------------------
2023-10-15 20:20:27,295 EPOCH 9 done: loss 0.0153 - lr: 0.000006
2023-10-15 20:20:35,652 DEV : loss 0.44299206137657166 - f1-score (micro avg) 0.3749
2023-10-15 20:20:35,697 ----------------------------------------------------------------------------------------------------
2023-10-15 20:20:53,742 epoch 10 - iter 260/2606 - loss 0.01007017 - time (sec): 18.04 - samples/sec: 1966.73 - lr: 0.000005 - momentum: 0.000000
2023-10-15 20:21:11,849 epoch 10 - iter 520/2606 - loss 0.00872718 - time (sec): 36.15 - samples/sec: 1916.23 - lr: 0.000004 - momentum: 0.000000
2023-10-15 20:21:30,620 epoch 10 - iter 780/2606 - loss 0.00877023 - time (sec): 54.92 - samples/sec: 1938.58 - lr: 0.000004 - momentum: 0.000000
2023-10-15 20:21:48,942 epoch 10 - iter 1040/2606 - loss 0.00900546 - time (sec): 73.24 - samples/sec: 1944.56 - lr: 0.000003 - momentum: 0.000000
2023-10-15 20:22:07,349 epoch 10 - iter 1300/2606 - loss 0.00925451 - time (sec): 91.65 - samples/sec: 1947.96 - lr: 0.000003 - momentum: 0.000000
2023-10-15 20:22:26,240 epoch 10 - iter 1560/2606 - loss 0.00955119 - time (sec): 110.54 - samples/sec: 1945.38 - lr: 0.000002 - momentum: 0.000000
2023-10-15 20:22:45,017 epoch 10 - iter 1820/2606 - loss 0.00958418 - time (sec): 129.32 - samples/sec: 1949.64 - lr: 0.000002 - momentum: 0.000000
2023-10-15 20:23:04,810 epoch 10 - iter 2080/2606 - loss 0.00978376 - time (sec): 149.11 - samples/sec: 1952.87 - lr: 0.000001 - momentum: 0.000000
2023-10-15 20:23:23,622 epoch 10 - iter 2340/2606 - loss 0.00995251 - time (sec): 167.92 - samples/sec: 1946.93 - lr: 0.000001 - momentum: 0.000000
2023-10-15 20:23:43,809 epoch 10 - iter 2600/2606 - loss 0.01001595 - time (sec): 188.11 - samples/sec: 1948.65 - lr: 0.000000 - momentum: 0.000000
2023-10-15 20:23:44,199 ----------------------------------------------------------------------------------------------------
2023-10-15 20:23:44,200 EPOCH 10 done: loss 0.0100 - lr: 0.000000
2023-10-15 20:23:53,258 DEV : loss 0.4734904170036316 - f1-score (micro avg) 0.3815
2023-10-15 20:23:53,688 ----------------------------------------------------------------------------------------------------
2023-10-15 20:23:53,689 Loading model from best epoch ...
2023-10-15 20:23:55,224 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
2023-10-15 20:24:11,965
Results:
- F-score (micro) 0.446
- F-score (macro) 0.2767
- Accuracy 0.2916
By class:
precision recall f1-score support
LOC 0.4970 0.6063 0.5462 1214
PER 0.4008 0.3651 0.3821 808
ORG 0.2091 0.1558 0.1786 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4379 0.4544 0.4460 2390
macro avg 0.2767 0.2818 0.2767 2390
weighted avg 0.4188 0.4544 0.4330 2390
2023-10-15 20:24:11,966 ----------------------------------------------------------------------------------------------------
|