File size: 24,075 Bytes
6e98dcd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
2023-10-17 09:02:10,577 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,579 Model: "SequenceTagger(
  (embeddings): TransformerWordEmbeddings(
    (model): ElectraModel(
      (embeddings): ElectraEmbeddings(
        (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): ElectraEncoder(
        (layer): ModuleList(
          (0-11): 12 x ElectraLayer(
            (attention): ElectraAttention(
              (self): ElectraSelfAttention(
                (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): ElectraSelfOutput(
                (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): ElectraIntermediate(
              (dense): Linear(in_features=768, out_features=3072, bias=True)
              (intermediate_act_fn): GELUActivation()
            )
            (output): ElectraOutput(
              (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)
            )
          )
        )
      )
    )
  )
  (locked_dropout): LockedDropout(p=0.5)
  (linear): Linear(in_features=768, out_features=13, bias=True)
  (loss_function): CrossEntropyLoss()
)"
2023-10-17 09:02:10,579 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,579 MultiCorpus: 14465 train + 1392 dev + 2432 test sentences
 - NER_HIPE_2022 Corpus: 14465 train + 1392 dev + 2432 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/letemps/fr/with_doc_seperator
2023-10-17 09:02:10,579 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,579 Train:  14465 sentences
2023-10-17 09:02:10,579         (train_with_dev=False, train_with_test=False)
2023-10-17 09:02:10,579 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,579 Training Params:
2023-10-17 09:02:10,579  - learning_rate: "3e-05" 
2023-10-17 09:02:10,579  - mini_batch_size: "4"
2023-10-17 09:02:10,579  - max_epochs: "10"
2023-10-17 09:02:10,579  - shuffle: "True"
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 Plugins:
2023-10-17 09:02:10,580  - TensorboardLogger
2023-10-17 09:02:10,580  - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 09:02:10,580  - metric: "('micro avg', 'f1-score')"
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 Computation:
2023-10-17 09:02:10,580  - compute on device: cuda:0
2023-10-17 09:02:10,580  - embedding storage: none
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 Model training base path: "hmbench-letemps/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 ----------------------------------------------------------------------------------------------------
2023-10-17 09:02:10,580 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 09:02:34,601 epoch 1 - iter 361/3617 - loss 1.57318286 - time (sec): 24.02 - samples/sec: 1618.04 - lr: 0.000003 - momentum: 0.000000
2023-10-17 09:02:57,319 epoch 1 - iter 722/3617 - loss 0.91675384 - time (sec): 46.74 - samples/sec: 1634.35 - lr: 0.000006 - momentum: 0.000000
2023-10-17 09:03:20,631 epoch 1 - iter 1083/3617 - loss 0.66327805 - time (sec): 70.05 - samples/sec: 1635.60 - lr: 0.000009 - momentum: 0.000000
2023-10-17 09:03:43,792 epoch 1 - iter 1444/3617 - loss 0.53397507 - time (sec): 93.21 - samples/sec: 1641.10 - lr: 0.000012 - momentum: 0.000000
2023-10-17 09:04:07,330 epoch 1 - iter 1805/3617 - loss 0.45416962 - time (sec): 116.75 - samples/sec: 1629.84 - lr: 0.000015 - momentum: 0.000000
2023-10-17 09:04:28,950 epoch 1 - iter 2166/3617 - loss 0.40035588 - time (sec): 138.37 - samples/sec: 1642.28 - lr: 0.000018 - momentum: 0.000000
2023-10-17 09:04:50,503 epoch 1 - iter 2527/3617 - loss 0.36212873 - time (sec): 159.92 - samples/sec: 1653.79 - lr: 0.000021 - momentum: 0.000000
2023-10-17 09:05:13,205 epoch 1 - iter 2888/3617 - loss 0.33159814 - time (sec): 182.62 - samples/sec: 1663.85 - lr: 0.000024 - momentum: 0.000000
2023-10-17 09:05:35,839 epoch 1 - iter 3249/3617 - loss 0.30611244 - time (sec): 205.26 - samples/sec: 1667.31 - lr: 0.000027 - momentum: 0.000000
2023-10-17 09:05:56,932 epoch 1 - iter 3610/3617 - loss 0.28747179 - time (sec): 226.35 - samples/sec: 1674.81 - lr: 0.000030 - momentum: 0.000000
2023-10-17 09:05:57,348 ----------------------------------------------------------------------------------------------------
2023-10-17 09:05:57,349 EPOCH 1 done: loss 0.2873 - lr: 0.000030
2023-10-17 09:06:03,627 DEV : loss 0.12736186385154724 - f1-score (micro avg)  0.5729
2023-10-17 09:06:03,669 saving best model
2023-10-17 09:06:04,170 ----------------------------------------------------------------------------------------------------
2023-10-17 09:06:27,927 epoch 2 - iter 361/3617 - loss 0.10706765 - time (sec): 23.76 - samples/sec: 1541.66 - lr: 0.000030 - momentum: 0.000000
2023-10-17 09:06:52,938 epoch 2 - iter 722/3617 - loss 0.10073542 - time (sec): 48.77 - samples/sec: 1557.77 - lr: 0.000029 - momentum: 0.000000
2023-10-17 09:07:16,610 epoch 2 - iter 1083/3617 - loss 0.10019335 - time (sec): 72.44 - samples/sec: 1589.80 - lr: 0.000029 - momentum: 0.000000
2023-10-17 09:07:39,159 epoch 2 - iter 1444/3617 - loss 0.10373908 - time (sec): 94.99 - samples/sec: 1596.28 - lr: 0.000029 - momentum: 0.000000
2023-10-17 09:08:02,807 epoch 2 - iter 1805/3617 - loss 0.10168351 - time (sec): 118.64 - samples/sec: 1585.32 - lr: 0.000028 - momentum: 0.000000
2023-10-17 09:08:26,939 epoch 2 - iter 2166/3617 - loss 0.10048563 - time (sec): 142.77 - samples/sec: 1582.24 - lr: 0.000028 - momentum: 0.000000
2023-10-17 09:08:48,749 epoch 2 - iter 2527/3617 - loss 0.09886610 - time (sec): 164.58 - samples/sec: 1601.65 - lr: 0.000028 - momentum: 0.000000
2023-10-17 09:09:10,359 epoch 2 - iter 2888/3617 - loss 0.10012403 - time (sec): 186.19 - samples/sec: 1621.82 - lr: 0.000027 - momentum: 0.000000
2023-10-17 09:09:31,958 epoch 2 - iter 3249/3617 - loss 0.09783095 - time (sec): 207.79 - samples/sec: 1650.26 - lr: 0.000027 - momentum: 0.000000
2023-10-17 09:09:53,444 epoch 2 - iter 3610/3617 - loss 0.09771129 - time (sec): 229.27 - samples/sec: 1654.59 - lr: 0.000027 - momentum: 0.000000
2023-10-17 09:09:53,849 ----------------------------------------------------------------------------------------------------
2023-10-17 09:09:53,850 EPOCH 2 done: loss 0.0977 - lr: 0.000027
2023-10-17 09:10:00,951 DEV : loss 0.1343551129102707 - f1-score (micro avg)  0.6562
2023-10-17 09:10:00,995 saving best model
2023-10-17 09:10:01,589 ----------------------------------------------------------------------------------------------------
2023-10-17 09:10:23,891 epoch 3 - iter 361/3617 - loss 0.07791442 - time (sec): 22.30 - samples/sec: 1689.74 - lr: 0.000026 - momentum: 0.000000
2023-10-17 09:10:48,004 epoch 3 - iter 722/3617 - loss 0.07169122 - time (sec): 46.41 - samples/sec: 1664.02 - lr: 0.000026 - momentum: 0.000000
2023-10-17 09:11:12,224 epoch 3 - iter 1083/3617 - loss 0.07458371 - time (sec): 70.63 - samples/sec: 1628.39 - lr: 0.000026 - momentum: 0.000000
2023-10-17 09:11:34,685 epoch 3 - iter 1444/3617 - loss 0.07550290 - time (sec): 93.09 - samples/sec: 1658.33 - lr: 0.000025 - momentum: 0.000000
2023-10-17 09:11:57,759 epoch 3 - iter 1805/3617 - loss 0.07532046 - time (sec): 116.17 - samples/sec: 1655.01 - lr: 0.000025 - momentum: 0.000000
2023-10-17 09:12:20,983 epoch 3 - iter 2166/3617 - loss 0.07566162 - time (sec): 139.39 - samples/sec: 1655.63 - lr: 0.000025 - momentum: 0.000000
2023-10-17 09:12:44,462 epoch 3 - iter 2527/3617 - loss 0.07561865 - time (sec): 162.87 - samples/sec: 1649.55 - lr: 0.000024 - momentum: 0.000000
2023-10-17 09:13:08,282 epoch 3 - iter 2888/3617 - loss 0.07691066 - time (sec): 186.69 - samples/sec: 1634.52 - lr: 0.000024 - momentum: 0.000000
2023-10-17 09:13:30,472 epoch 3 - iter 3249/3617 - loss 0.07626978 - time (sec): 208.88 - samples/sec: 1640.52 - lr: 0.000024 - momentum: 0.000000
2023-10-17 09:13:52,448 epoch 3 - iter 3610/3617 - loss 0.07614469 - time (sec): 230.86 - samples/sec: 1642.45 - lr: 0.000023 - momentum: 0.000000
2023-10-17 09:13:52,876 ----------------------------------------------------------------------------------------------------
2023-10-17 09:13:52,876 EPOCH 3 done: loss 0.0762 - lr: 0.000023
2023-10-17 09:13:59,262 DEV : loss 0.183589369058609 - f1-score (micro avg)  0.6295
2023-10-17 09:13:59,306 ----------------------------------------------------------------------------------------------------
2023-10-17 09:14:22,434 epoch 4 - iter 361/3617 - loss 0.05314260 - time (sec): 23.13 - samples/sec: 1683.44 - lr: 0.000023 - momentum: 0.000000
2023-10-17 09:14:44,723 epoch 4 - iter 722/3617 - loss 0.04744612 - time (sec): 45.42 - samples/sec: 1691.58 - lr: 0.000023 - momentum: 0.000000
2023-10-17 09:15:07,522 epoch 4 - iter 1083/3617 - loss 0.05024746 - time (sec): 68.21 - samples/sec: 1705.54 - lr: 0.000022 - momentum: 0.000000
2023-10-17 09:15:30,144 epoch 4 - iter 1444/3617 - loss 0.05060285 - time (sec): 90.84 - samples/sec: 1687.44 - lr: 0.000022 - momentum: 0.000000
2023-10-17 09:15:51,988 epoch 4 - iter 1805/3617 - loss 0.05209485 - time (sec): 112.68 - samples/sec: 1703.02 - lr: 0.000022 - momentum: 0.000000
2023-10-17 09:16:13,057 epoch 4 - iter 2166/3617 - loss 0.05270590 - time (sec): 133.75 - samples/sec: 1703.34 - lr: 0.000021 - momentum: 0.000000
2023-10-17 09:16:35,238 epoch 4 - iter 2527/3617 - loss 0.05313400 - time (sec): 155.93 - samples/sec: 1704.57 - lr: 0.000021 - momentum: 0.000000
2023-10-17 09:16:57,109 epoch 4 - iter 2888/3617 - loss 0.05411486 - time (sec): 177.80 - samples/sec: 1712.35 - lr: 0.000021 - momentum: 0.000000
2023-10-17 09:17:19,686 epoch 4 - iter 3249/3617 - loss 0.05477012 - time (sec): 200.38 - samples/sec: 1703.16 - lr: 0.000020 - momentum: 0.000000
2023-10-17 09:17:41,730 epoch 4 - iter 3610/3617 - loss 0.05488383 - time (sec): 222.42 - samples/sec: 1704.66 - lr: 0.000020 - momentum: 0.000000
2023-10-17 09:17:42,130 ----------------------------------------------------------------------------------------------------
2023-10-17 09:17:42,130 EPOCH 4 done: loss 0.0548 - lr: 0.000020
2023-10-17 09:17:48,498 DEV : loss 0.23047274351119995 - f1-score (micro avg)  0.6346
2023-10-17 09:17:48,542 ----------------------------------------------------------------------------------------------------
2023-10-17 09:18:11,645 epoch 5 - iter 361/3617 - loss 0.03877276 - time (sec): 23.10 - samples/sec: 1648.99 - lr: 0.000020 - momentum: 0.000000
2023-10-17 09:18:35,842 epoch 5 - iter 722/3617 - loss 0.03893797 - time (sec): 47.30 - samples/sec: 1573.53 - lr: 0.000019 - momentum: 0.000000
2023-10-17 09:19:00,608 epoch 5 - iter 1083/3617 - loss 0.03738715 - time (sec): 72.06 - samples/sec: 1566.11 - lr: 0.000019 - momentum: 0.000000
2023-10-17 09:19:24,105 epoch 5 - iter 1444/3617 - loss 0.03521727 - time (sec): 95.56 - samples/sec: 1570.21 - lr: 0.000019 - momentum: 0.000000
2023-10-17 09:19:49,600 epoch 5 - iter 1805/3617 - loss 0.03658251 - time (sec): 121.06 - samples/sec: 1563.80 - lr: 0.000018 - momentum: 0.000000
2023-10-17 09:20:12,773 epoch 5 - iter 2166/3617 - loss 0.03666109 - time (sec): 144.23 - samples/sec: 1576.19 - lr: 0.000018 - momentum: 0.000000
2023-10-17 09:20:35,723 epoch 5 - iter 2527/3617 - loss 0.03560606 - time (sec): 167.18 - samples/sec: 1592.90 - lr: 0.000018 - momentum: 0.000000
2023-10-17 09:20:59,453 epoch 5 - iter 2888/3617 - loss 0.03567818 - time (sec): 190.91 - samples/sec: 1591.22 - lr: 0.000017 - momentum: 0.000000
2023-10-17 09:21:21,835 epoch 5 - iter 3249/3617 - loss 0.03843292 - time (sec): 213.29 - samples/sec: 1598.18 - lr: 0.000017 - momentum: 0.000000
2023-10-17 09:21:45,332 epoch 5 - iter 3610/3617 - loss 0.03827749 - time (sec): 236.79 - samples/sec: 1601.48 - lr: 0.000017 - momentum: 0.000000
2023-10-17 09:21:45,766 ----------------------------------------------------------------------------------------------------
2023-10-17 09:21:45,766 EPOCH 5 done: loss 0.0386 - lr: 0.000017
2023-10-17 09:21:52,102 DEV : loss 0.30182531476020813 - f1-score (micro avg)  0.655
2023-10-17 09:21:52,151 ----------------------------------------------------------------------------------------------------
2023-10-17 09:22:13,654 epoch 6 - iter 361/3617 - loss 0.02971622 - time (sec): 21.50 - samples/sec: 1768.48 - lr: 0.000016 - momentum: 0.000000
2023-10-17 09:22:35,210 epoch 6 - iter 722/3617 - loss 0.02596294 - time (sec): 43.06 - samples/sec: 1771.12 - lr: 0.000016 - momentum: 0.000000
2023-10-17 09:22:56,691 epoch 6 - iter 1083/3617 - loss 0.02699636 - time (sec): 64.54 - samples/sec: 1753.12 - lr: 0.000016 - momentum: 0.000000
2023-10-17 09:23:18,245 epoch 6 - iter 1444/3617 - loss 0.02615175 - time (sec): 86.09 - samples/sec: 1759.27 - lr: 0.000015 - momentum: 0.000000
2023-10-17 09:23:39,965 epoch 6 - iter 1805/3617 - loss 0.02525762 - time (sec): 107.81 - samples/sec: 1757.56 - lr: 0.000015 - momentum: 0.000000
2023-10-17 09:24:03,115 epoch 6 - iter 2166/3617 - loss 0.02578467 - time (sec): 130.96 - samples/sec: 1734.42 - lr: 0.000015 - momentum: 0.000000
2023-10-17 09:24:25,467 epoch 6 - iter 2527/3617 - loss 0.02698358 - time (sec): 153.31 - samples/sec: 1728.18 - lr: 0.000014 - momentum: 0.000000
2023-10-17 09:24:47,331 epoch 6 - iter 2888/3617 - loss 0.02700357 - time (sec): 175.18 - samples/sec: 1728.11 - lr: 0.000014 - momentum: 0.000000
2023-10-17 09:25:09,745 epoch 6 - iter 3249/3617 - loss 0.02824365 - time (sec): 197.59 - samples/sec: 1727.15 - lr: 0.000014 - momentum: 0.000000
2023-10-17 09:25:31,716 epoch 6 - iter 3610/3617 - loss 0.02780229 - time (sec): 219.56 - samples/sec: 1726.51 - lr: 0.000013 - momentum: 0.000000
2023-10-17 09:25:32,140 ----------------------------------------------------------------------------------------------------
2023-10-17 09:25:32,140 EPOCH 6 done: loss 0.0277 - lr: 0.000013
2023-10-17 09:25:39,197 DEV : loss 0.35149648785591125 - f1-score (micro avg)  0.6416
2023-10-17 09:25:39,238 ----------------------------------------------------------------------------------------------------
2023-10-17 09:26:03,014 epoch 7 - iter 361/3617 - loss 0.01810299 - time (sec): 23.77 - samples/sec: 1687.02 - lr: 0.000013 - momentum: 0.000000
2023-10-17 09:26:26,361 epoch 7 - iter 722/3617 - loss 0.01752033 - time (sec): 47.12 - samples/sec: 1674.89 - lr: 0.000013 - momentum: 0.000000
2023-10-17 09:26:49,545 epoch 7 - iter 1083/3617 - loss 0.01868235 - time (sec): 70.31 - samples/sec: 1650.25 - lr: 0.000012 - momentum: 0.000000
2023-10-17 09:27:14,249 epoch 7 - iter 1444/3617 - loss 0.01923882 - time (sec): 95.01 - samples/sec: 1609.58 - lr: 0.000012 - momentum: 0.000000
2023-10-17 09:27:37,499 epoch 7 - iter 1805/3617 - loss 0.01853291 - time (sec): 118.26 - samples/sec: 1604.87 - lr: 0.000012 - momentum: 0.000000
2023-10-17 09:28:00,694 epoch 7 - iter 2166/3617 - loss 0.01937851 - time (sec): 141.45 - samples/sec: 1599.75 - lr: 0.000011 - momentum: 0.000000
2023-10-17 09:28:23,889 epoch 7 - iter 2527/3617 - loss 0.01913223 - time (sec): 164.65 - samples/sec: 1608.56 - lr: 0.000011 - momentum: 0.000000
2023-10-17 09:28:48,438 epoch 7 - iter 2888/3617 - loss 0.01935787 - time (sec): 189.20 - samples/sec: 1603.18 - lr: 0.000011 - momentum: 0.000000
2023-10-17 09:29:12,391 epoch 7 - iter 3249/3617 - loss 0.01962716 - time (sec): 213.15 - samples/sec: 1600.38 - lr: 0.000010 - momentum: 0.000000
2023-10-17 09:29:34,243 epoch 7 - iter 3610/3617 - loss 0.01941165 - time (sec): 235.00 - samples/sec: 1614.35 - lr: 0.000010 - momentum: 0.000000
2023-10-17 09:29:34,669 ----------------------------------------------------------------------------------------------------
2023-10-17 09:29:34,670 EPOCH 7 done: loss 0.0194 - lr: 0.000010
2023-10-17 09:29:41,014 DEV : loss 0.34877660870552063 - f1-score (micro avg)  0.6485
2023-10-17 09:29:41,056 ----------------------------------------------------------------------------------------------------
2023-10-17 09:30:02,968 epoch 8 - iter 361/3617 - loss 0.01268390 - time (sec): 21.91 - samples/sec: 1703.57 - lr: 0.000010 - momentum: 0.000000
2023-10-17 09:30:25,299 epoch 8 - iter 722/3617 - loss 0.01099667 - time (sec): 44.24 - samples/sec: 1711.21 - lr: 0.000009 - momentum: 0.000000
2023-10-17 09:30:48,107 epoch 8 - iter 1083/3617 - loss 0.01320618 - time (sec): 67.05 - samples/sec: 1694.10 - lr: 0.000009 - momentum: 0.000000
2023-10-17 09:31:11,789 epoch 8 - iter 1444/3617 - loss 0.01438812 - time (sec): 90.73 - samples/sec: 1666.77 - lr: 0.000009 - momentum: 0.000000
2023-10-17 09:31:33,254 epoch 8 - iter 1805/3617 - loss 0.01526598 - time (sec): 112.20 - samples/sec: 1689.19 - lr: 0.000008 - momentum: 0.000000
2023-10-17 09:31:54,672 epoch 8 - iter 2166/3617 - loss 0.01461418 - time (sec): 133.61 - samples/sec: 1697.90 - lr: 0.000008 - momentum: 0.000000
2023-10-17 09:32:16,127 epoch 8 - iter 2527/3617 - loss 0.01418461 - time (sec): 155.07 - samples/sec: 1709.65 - lr: 0.000008 - momentum: 0.000000
2023-10-17 09:32:37,708 epoch 8 - iter 2888/3617 - loss 0.01390302 - time (sec): 176.65 - samples/sec: 1717.72 - lr: 0.000007 - momentum: 0.000000
2023-10-17 09:33:01,690 epoch 8 - iter 3249/3617 - loss 0.01320267 - time (sec): 200.63 - samples/sec: 1699.52 - lr: 0.000007 - momentum: 0.000000
2023-10-17 09:33:24,874 epoch 8 - iter 3610/3617 - loss 0.01285080 - time (sec): 223.82 - samples/sec: 1693.91 - lr: 0.000007 - momentum: 0.000000
2023-10-17 09:33:25,290 ----------------------------------------------------------------------------------------------------
2023-10-17 09:33:25,290 EPOCH 8 done: loss 0.0129 - lr: 0.000007
2023-10-17 09:33:31,720 DEV : loss 0.38916581869125366 - f1-score (micro avg)  0.6549
2023-10-17 09:33:31,761 ----------------------------------------------------------------------------------------------------
2023-10-17 09:33:56,673 epoch 9 - iter 361/3617 - loss 0.01001985 - time (sec): 24.91 - samples/sec: 1557.19 - lr: 0.000006 - momentum: 0.000000
2023-10-17 09:34:20,402 epoch 9 - iter 722/3617 - loss 0.00865293 - time (sec): 48.64 - samples/sec: 1545.49 - lr: 0.000006 - momentum: 0.000000
2023-10-17 09:34:44,683 epoch 9 - iter 1083/3617 - loss 0.00726475 - time (sec): 72.92 - samples/sec: 1550.75 - lr: 0.000006 - momentum: 0.000000
2023-10-17 09:35:07,954 epoch 9 - iter 1444/3617 - loss 0.00785104 - time (sec): 96.19 - samples/sec: 1579.21 - lr: 0.000005 - momentum: 0.000000
2023-10-17 09:35:32,051 epoch 9 - iter 1805/3617 - loss 0.00836511 - time (sec): 120.29 - samples/sec: 1579.08 - lr: 0.000005 - momentum: 0.000000
2023-10-17 09:35:54,176 epoch 9 - iter 2166/3617 - loss 0.00795807 - time (sec): 142.41 - samples/sec: 1601.95 - lr: 0.000005 - momentum: 0.000000
2023-10-17 09:36:16,583 epoch 9 - iter 2527/3617 - loss 0.00807517 - time (sec): 164.82 - samples/sec: 1611.25 - lr: 0.000004 - momentum: 0.000000
2023-10-17 09:36:39,198 epoch 9 - iter 2888/3617 - loss 0.00797982 - time (sec): 187.44 - samples/sec: 1620.56 - lr: 0.000004 - momentum: 0.000000
2023-10-17 09:37:02,573 epoch 9 - iter 3249/3617 - loss 0.00787163 - time (sec): 210.81 - samples/sec: 1624.12 - lr: 0.000004 - momentum: 0.000000
2023-10-17 09:37:26,353 epoch 9 - iter 3610/3617 - loss 0.00821338 - time (sec): 234.59 - samples/sec: 1617.36 - lr: 0.000003 - momentum: 0.000000
2023-10-17 09:37:26,818 ----------------------------------------------------------------------------------------------------
2023-10-17 09:37:26,818 EPOCH 9 done: loss 0.0082 - lr: 0.000003
2023-10-17 09:37:34,515 DEV : loss 0.3819396495819092 - f1-score (micro avg)  0.6596
2023-10-17 09:37:34,561 saving best model
2023-10-17 09:37:35,193 ----------------------------------------------------------------------------------------------------
2023-10-17 09:37:59,413 epoch 10 - iter 361/3617 - loss 0.00781618 - time (sec): 24.22 - samples/sec: 1593.10 - lr: 0.000003 - momentum: 0.000000
2023-10-17 09:38:22,354 epoch 10 - iter 722/3617 - loss 0.00571085 - time (sec): 47.16 - samples/sec: 1608.77 - lr: 0.000003 - momentum: 0.000000
2023-10-17 09:38:45,290 epoch 10 - iter 1083/3617 - loss 0.00491489 - time (sec): 70.10 - samples/sec: 1617.50 - lr: 0.000002 - momentum: 0.000000
2023-10-17 09:39:08,269 epoch 10 - iter 1444/3617 - loss 0.00460967 - time (sec): 93.07 - samples/sec: 1618.29 - lr: 0.000002 - momentum: 0.000000
2023-10-17 09:39:30,478 epoch 10 - iter 1805/3617 - loss 0.00449936 - time (sec): 115.28 - samples/sec: 1622.89 - lr: 0.000002 - momentum: 0.000000
2023-10-17 09:39:54,643 epoch 10 - iter 2166/3617 - loss 0.00457182 - time (sec): 139.45 - samples/sec: 1617.06 - lr: 0.000001 - momentum: 0.000000
2023-10-17 09:40:17,739 epoch 10 - iter 2527/3617 - loss 0.00425884 - time (sec): 162.54 - samples/sec: 1632.20 - lr: 0.000001 - momentum: 0.000000
2023-10-17 09:40:40,411 epoch 10 - iter 2888/3617 - loss 0.00465737 - time (sec): 185.22 - samples/sec: 1632.18 - lr: 0.000001 - momentum: 0.000000
2023-10-17 09:41:03,801 epoch 10 - iter 3249/3617 - loss 0.00458577 - time (sec): 208.61 - samples/sec: 1633.83 - lr: 0.000000 - momentum: 0.000000
2023-10-17 09:41:26,934 epoch 10 - iter 3610/3617 - loss 0.00474640 - time (sec): 231.74 - samples/sec: 1636.82 - lr: 0.000000 - momentum: 0.000000
2023-10-17 09:41:27,381 ----------------------------------------------------------------------------------------------------
2023-10-17 09:41:27,381 EPOCH 10 done: loss 0.0047 - lr: 0.000000
2023-10-17 09:41:33,760 DEV : loss 0.40882349014282227 - f1-score (micro avg)  0.66
2023-10-17 09:41:33,801 saving best model
2023-10-17 09:41:34,823 ----------------------------------------------------------------------------------------------------
2023-10-17 09:41:34,824 Loading model from best epoch ...
2023-10-17 09:41:36,920 SequenceTagger predicts: Dictionary with 13 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org
2023-10-17 09:41:46,180 
Results:
- F-score (micro) 0.6469
- F-score (macro) 0.5083
- Accuracy 0.4891

By class:
              precision    recall  f1-score   support

         loc     0.6288    0.7766    0.6949       591
        pers     0.5766    0.7591    0.6554       357
         org     0.1857    0.1646    0.1745        79

   micro avg     0.5850    0.7235    0.6469      1027
   macro avg     0.4637    0.5668    0.5083      1027
weighted avg     0.5766    0.7235    0.6411      1027

2023-10-17 09:41:46,180 ----------------------------------------------------------------------------------------------------