Hugging Face's logo

hmbert-tiny
/
flair-icdar-fr

Token Classification
Flair
PyTorch
TensorBoard
French
sequence-tagger-model
Model card Files Metrics Community
1
stefan-it commited on
Commit
eb0ba09
•
1 Parent(s): f0079bd

Upload folder using huggingface_hub

Browse files
Files changed (6) hide show
  1. best-model.pt +3 -0
  2. dev.tsv +0 -0
  3. loss.tsv +11 -0
  4. runs/events.out.tfevents.1697664318.46dc0c540dd0.3341.13 +3 -0
  5. test.tsv +0 -0
  6. training.log +245 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f37cfbe7e1895f25b66ae3a8a01e0d5cca89e6f5b2968bf5085176a44d7ca8d
3
+ size 19045922
dev.tsv ADDED
The diff for this file is too large to render. See raw diff
 
loss.tsv ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ EPOCH TIMESTAMP LEARNING_RATE TRAIN_LOSS DEV_LOSS DEV_PRECISION DEV_RECALL DEV_F1 DEV_ACCURACY
2
+ 1 21:25:50 0.0000 0.8594 0.2119 0.3472 0.2557 0.2945 0.1885
3
+ 2 21:26:22 0.0000 0.2546 0.1665 0.3849 0.4276 0.4051 0.2773
4
+ 3 21:26:55 0.0000 0.2046 0.1502 0.5165 0.5486 0.5321 0.3871
5
+ 4 21:27:28 0.0000 0.1797 0.1447 0.5793 0.5577 0.5683 0.4206
6
+ 5 21:28:00 0.0000 0.1629 0.1397 0.5904 0.6131 0.6016 0.4574
7
+ 6 21:28:32 0.0000 0.1536 0.1435 0.6024 0.6256 0.6138 0.4698
8
+ 7 21:29:04 0.0000 0.1432 0.1461 0.6077 0.6290 0.6181 0.4724
9
+ 8 21:29:37 0.0000 0.1354 0.1463 0.5910 0.6357 0.6125 0.4679
10
+ 9 21:30:09 0.0000 0.1302 0.1465 0.5956 0.6482 0.6208 0.4767
11
+ 10 21:30:42 0.0000 0.1267 0.1485 0.5973 0.6391 0.6175 0.4724
runs/events.out.tfevents.1697664318.46dc0c540dd0.3341.13 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27ff5615eacfa5c88fa16d9156012558280eb457efae93b70a709431c4d23e74
3
+ size 1108164
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,245 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-18 21:25:18,822 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-18 21:25:18,823 Model: "SequenceTagger(
3
+ (embeddings): TransformerWordEmbeddings(
4
+ (model): BertModel(
5
+ (embeddings): BertEmbeddings(
6
+ (word_embeddings): Embedding(32001, 128)
7
+ (position_embeddings): Embedding(512, 128)
8
+ (token_type_embeddings): Embedding(2, 128)
9
+ (LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
10
+ (dropout): Dropout(p=0.1, inplace=False)
11
+ )
12
+ (encoder): BertEncoder(
13
+ (layer): ModuleList(
14
+ (0-1): 2 x BertLayer(
15
+ (attention): BertAttention(
16
+ (self): BertSelfAttention(
17
+ (query): Linear(in_features=128, out_features=128, bias=True)
18
+ (key): Linear(in_features=128, out_features=128, bias=True)
19
+ (value): Linear(in_features=128, out_features=128, bias=True)
20
+ (dropout): Dropout(p=0.1, inplace=False)
21
+ )
22
+ (output): BertSelfOutput(
23
+ (dense): Linear(in_features=128, out_features=128, bias=True)
24
+ (LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
25
+ (dropout): Dropout(p=0.1, inplace=False)
26
+ )
27
+ )
28
+ (intermediate): BertIntermediate(
29
+ (dense): Linear(in_features=128, out_features=512, bias=True)
30
+ (intermediate_act_fn): GELUActivation()
31
+ )
32
+ (output): BertOutput(
33
+ (dense): Linear(in_features=512, out_features=128, bias=True)
34
+ (LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
35
+ (dropout): Dropout(p=0.1, inplace=False)
36
+ )
37
+ )
38
+ )
39
+ )
40
+ (pooler): BertPooler(
41
+ (dense): Linear(in_features=128, out_features=128, bias=True)
42
+ (activation): Tanh()
43
+ )
44
+ )
45
+ )
46
+ (locked_dropout): LockedDropout(p=0.5)
47
+ (linear): Linear(in_features=128, out_features=13, bias=True)
48
+ (loss_function): CrossEntropyLoss()
49
+ )"
50
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
51
+ 2023-10-18 21:25:18,823 MultiCorpus: 7936 train + 992 dev + 992 test sentences
52
+ - NER_ICDAR_EUROPEANA Corpus: 7936 train + 992 dev + 992 test sentences - /root/.flair/datasets/ner_icdar_europeana/fr
53
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
54
+ 2023-10-18 21:25:18,823 Train: 7936 sentences
55
+ 2023-10-18 21:25:18,823 (train_with_dev=False, train_with_test=False)
56
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
57
+ 2023-10-18 21:25:18,823 Training Params:
58
+ 2023-10-18 21:25:18,823 - learning_rate: "5e-05"
59
+ 2023-10-18 21:25:18,823 - mini_batch_size: "4"
60
+ 2023-10-18 21:25:18,823 - max_epochs: "10"
61
+ 2023-10-18 21:25:18,823 - shuffle: "True"
62
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
63
+ 2023-10-18 21:25:18,823 Plugins:
64
+ 2023-10-18 21:25:18,823 - TensorboardLogger
65
+ 2023-10-18 21:25:18,823 - LinearScheduler | warmup_fraction: '0.1'
66
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
67
+ 2023-10-18 21:25:18,823 Final evaluation on model from best epoch (best-model.pt)
68
+ 2023-10-18 21:25:18,823 - metric: "('micro avg', 'f1-score')"
69
+ 2023-10-18 21:25:18,823 ----------------------------------------------------------------------------------------------------
70
+ 2023-10-18 21:25:18,823 Computation:
71
+ 2023-10-18 21:25:18,824 - compute on device: cuda:0
72
+ 2023-10-18 21:25:18,824 - embedding storage: none
73
+ 2023-10-18 21:25:18,824 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-18 21:25:18,824 Model training base path: "hmbench-icdar/fr-dbmdz/bert-tiny-historic-multilingual-cased-bs4-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-4"
75
+ 2023-10-18 21:25:18,824 ----------------------------------------------------------------------------------------------------
76
+ 2023-10-18 21:25:18,824 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-18 21:25:18,824 Logging anything other than scalars to TensorBoard is currently not supported.
78
+ 2023-10-18 21:25:21,871 epoch 1 - iter 198/1984 - loss 3.05001127 - time (sec): 3.05 - samples/sec: 5386.70 - lr: 0.000005 - momentum: 0.000000
79
+ 2023-10-18 21:25:24,628 epoch 1 - iter 396/1984 - loss 2.47724953 - time (sec): 5.80 - samples/sec: 5922.14 - lr: 0.000010 - momentum: 0.000000
80
+ 2023-10-18 21:25:27,667 epoch 1 - iter 594/1984 - loss 1.93619435 - time (sec): 8.84 - samples/sec: 5709.43 - lr: 0.000015 - momentum: 0.000000
81
+ 2023-10-18 21:25:30,706 epoch 1 - iter 792/1984 - loss 1.58401951 - time (sec): 11.88 - samples/sec: 5618.92 - lr: 0.000020 - momentum: 0.000000
82
+ 2023-10-18 21:25:33,758 epoch 1 - iter 990/1984 - loss 1.35401175 - time (sec): 14.93 - samples/sec: 5563.79 - lr: 0.000025 - momentum: 0.000000
83
+ 2023-10-18 21:25:36,794 epoch 1 - iter 1188/1984 - loss 1.20410373 - time (sec): 17.97 - samples/sec: 5524.66 - lr: 0.000030 - momentum: 0.000000
84
+ 2023-10-18 21:25:39,832 epoch 1 - iter 1386/1984 - loss 1.09303606 - time (sec): 21.01 - samples/sec: 5470.81 - lr: 0.000035 - momentum: 0.000000
85
+ 2023-10-18 21:25:42,910 epoch 1 - iter 1584/1984 - loss 0.99807468 - time (sec): 24.09 - samples/sec: 5451.75 - lr: 0.000040 - momentum: 0.000000
86
+ 2023-10-18 21:25:45,923 epoch 1 - iter 1782/1984 - loss 0.92290600 - time (sec): 27.10 - samples/sec: 5442.29 - lr: 0.000045 - momentum: 0.000000
87
+ 2023-10-18 21:25:49,008 epoch 1 - iter 1980/1984 - loss 0.86043295 - time (sec): 30.18 - samples/sec: 5425.81 - lr: 0.000050 - momentum: 0.000000
88
+ 2023-10-18 21:25:49,068 ----------------------------------------------------------------------------------------------------
89
+ 2023-10-18 21:25:49,068 EPOCH 1 done: loss 0.8594 - lr: 0.000050
90
+ 2023-10-18 21:25:50,560 DEV : loss 0.21192067861557007 - f1-score (micro avg) 0.2945
91
+ 2023-10-18 21:25:50,578 saving best model
92
+ 2023-10-18 21:25:50,613 ----------------------------------------------------------------------------------------------------
93
+ 2023-10-18 21:25:53,727 epoch 2 - iter 198/1984 - loss 0.30950366 - time (sec): 3.11 - samples/sec: 4927.98 - lr: 0.000049 - momentum: 0.000000
94
+ 2023-10-18 21:25:56,796 epoch 2 - iter 396/1984 - loss 0.28151013 - time (sec): 6.18 - samples/sec: 5145.74 - lr: 0.000049 - momentum: 0.000000
95
+ 2023-10-18 21:25:59,865 epoch 2 - iter 594/1984 - loss 0.28431083 - time (sec): 9.25 - samples/sec: 5342.97 - lr: 0.000048 - momentum: 0.000000
96
+ 2023-10-18 21:26:02,940 epoch 2 - iter 792/1984 - loss 0.27867074 - time (sec): 12.33 - samples/sec: 5398.12 - lr: 0.000048 - momentum: 0.000000
97
+ 2023-10-18 21:26:06,023 epoch 2 - iter 990/1984 - loss 0.27356192 - time (sec): 15.41 - samples/sec: 5362.82 - lr: 0.000047 - momentum: 0.000000
98
+ 2023-10-18 21:26:09,040 epoch 2 - iter 1188/1984 - loss 0.26808423 - time (sec): 18.43 - samples/sec: 5380.47 - lr: 0.000047 - momentum: 0.000000
99
+ 2023-10-18 21:26:11,822 epoch 2 - iter 1386/1984 - loss 0.26113663 - time (sec): 21.21 - samples/sec: 5451.69 - lr: 0.000046 - momentum: 0.000000
100
+ 2023-10-18 21:26:14,900 epoch 2 - iter 1584/1984 - loss 0.25788671 - time (sec): 24.29 - samples/sec: 5438.15 - lr: 0.000046 - momentum: 0.000000
101
+ 2023-10-18 21:26:17,947 epoch 2 - iter 1782/1984 - loss 0.25525919 - time (sec): 27.33 - samples/sec: 5421.86 - lr: 0.000045 - momentum: 0.000000
102
+ 2023-10-18 21:26:20,762 epoch 2 - iter 1980/1984 - loss 0.25444187 - time (sec): 30.15 - samples/sec: 5429.33 - lr: 0.000044 - momentum: 0.000000
103
+ 2023-10-18 21:26:20,822 ----------------------------------------------------------------------------------------------------
104
+ 2023-10-18 21:26:20,822 EPOCH 2 done: loss 0.2546 - lr: 0.000044
105
+ 2023-10-18 21:26:22,634 DEV : loss 0.16650356352329254 - f1-score (micro avg) 0.4051
106
+ 2023-10-18 21:26:22,654 saving best model
107
+ 2023-10-18 21:26:22,690 ----------------------------------------------------------------------------------------------------
108
+ 2023-10-18 21:26:25,787 epoch 3 - iter 198/1984 - loss 0.21161120 - time (sec): 3.10 - samples/sec: 5483.96 - lr: 0.000044 - momentum: 0.000000
109
+ 2023-10-18 21:26:28,837 epoch 3 - iter 396/1984 - loss 0.20257772 - time (sec): 6.15 - samples/sec: 5328.22 - lr: 0.000043 - momentum: 0.000000
110
+ 2023-10-18 21:26:31,767 epoch 3 - iter 594/1984 - loss 0.21363359 - time (sec): 9.08 - samples/sec: 5582.96 - lr: 0.000043 - momentum: 0.000000
111
+ 2023-10-18 21:26:34,842 epoch 3 - iter 792/1984 - loss 0.20789207 - time (sec): 12.15 - samples/sec: 5470.71 - lr: 0.000042 - momentum: 0.000000
112
+ 2023-10-18 21:26:37,879 epoch 3 - iter 990/1984 - loss 0.21055604 - time (sec): 15.19 - samples/sec: 5444.07 - lr: 0.000042 - momentum: 0.000000
113
+ 2023-10-18 21:26:40,954 epoch 3 - iter 1188/1984 - loss 0.20723027 - time (sec): 18.26 - samples/sec: 5444.07 - lr: 0.000041 - momentum: 0.000000
114
+ 2023-10-18 21:26:44,031 epoch 3 - iter 1386/1984 - loss 0.20657276 - time (sec): 21.34 - samples/sec: 5424.20 - lr: 0.000041 - momentum: 0.000000
115
+ 2023-10-18 21:26:47,055 epoch 3 - iter 1584/1984 - loss 0.20547288 - time (sec): 24.36 - samples/sec: 5434.88 - lr: 0.000040 - momentum: 0.000000
116
+ 2023-10-18 21:26:50,070 epoch 3 - iter 1782/1984 - loss 0.20632094 - time (sec): 27.38 - samples/sec: 5413.63 - lr: 0.000039 - momentum: 0.000000
117
+ 2023-10-18 21:26:53,106 epoch 3 - iter 1980/1984 - loss 0.20495296 - time (sec): 30.41 - samples/sec: 5376.32 - lr: 0.000039 - momentum: 0.000000
118
+ 2023-10-18 21:26:53,176 ----------------------------------------------------------------------------------------------------
119
+ 2023-10-18 21:26:53,176 EPOCH 3 done: loss 0.2046 - lr: 0.000039
120
+ 2023-10-18 21:26:55,394 DEV : loss 0.15022146701812744 - f1-score (micro avg) 0.5321
121
+ 2023-10-18 21:26:55,413 saving best model
122
+ 2023-10-18 21:26:55,447 ----------------------------------------------------------------------------------------------------
123
+ 2023-10-18 21:26:58,566 epoch 4 - iter 198/1984 - loss 0.18075544 - time (sec): 3.12 - samples/sec: 5291.79 - lr: 0.000038 - momentum: 0.000000
124
+ 2023-10-18 21:27:01,634 epoch 4 - iter 396/1984 - loss 0.18240152 - time (sec): 6.19 - samples/sec: 5143.93 - lr: 0.000038 - momentum: 0.000000
125
+ 2023-10-18 21:27:04,650 epoch 4 - iter 594/1984 - loss 0.17441944 - time (sec): 9.20 - samples/sec: 5216.19 - lr: 0.000037 - momentum: 0.000000
126
+ 2023-10-18 21:27:07,674 epoch 4 - iter 792/1984 - loss 0.17622746 - time (sec): 12.23 - samples/sec: 5214.64 - lr: 0.000037 - momentum: 0.000000
127
+ 2023-10-18 21:27:10,762 epoch 4 - iter 990/1984 - loss 0.17870370 - time (sec): 15.31 - samples/sec: 5266.42 - lr: 0.000036 - momentum: 0.000000
128
+ 2023-10-18 21:27:13,782 epoch 4 - iter 1188/1984 - loss 0.17730408 - time (sec): 18.33 - samples/sec: 5322.13 - lr: 0.000036 - momentum: 0.000000
129
+ 2023-10-18 21:27:16,890 epoch 4 - iter 1386/1984 - loss 0.17390393 - time (sec): 21.44 - samples/sec: 5403.97 - lr: 0.000035 - momentum: 0.000000
130
+ 2023-10-18 21:27:20,088 epoch 4 - iter 1584/1984 - loss 0.17431675 - time (sec): 24.64 - samples/sec: 5366.83 - lr: 0.000034 - momentum: 0.000000
131
+ 2023-10-18 21:27:23,148 epoch 4 - iter 1782/1984 - loss 0.17801776 - time (sec): 27.70 - samples/sec: 5333.04 - lr: 0.000034 - momentum: 0.000000
132
+ 2023-10-18 21:27:26,231 epoch 4 - iter 1980/1984 - loss 0.17957717 - time (sec): 30.78 - samples/sec: 5317.52 - lr: 0.000033 - momentum: 0.000000
133
+ 2023-10-18 21:27:26,291 ----------------------------------------------------------------------------------------------------
134
+ 2023-10-18 21:27:26,291 EPOCH 4 done: loss 0.1797 - lr: 0.000033
135
+ 2023-10-18 21:27:28,115 DEV : loss 0.14466138184070587 - f1-score (micro avg) 0.5683
136
+ 2023-10-18 21:27:28,133 saving best model
137
+ 2023-10-18 21:27:28,167 ----------------------------------------------------------------------------------------------------
138
+ 2023-10-18 21:27:31,281 epoch 5 - iter 198/1984 - loss 0.18093443 - time (sec): 3.11 - samples/sec: 5540.53 - lr: 0.000033 - momentum: 0.000000
139
+ 2023-10-18 21:27:34,301 epoch 5 - iter 396/1984 - loss 0.17976008 - time (sec): 6.13 - samples/sec: 5491.12 - lr: 0.000032 - momentum: 0.000000
140
+ 2023-10-18 21:27:37,345 epoch 5 - iter 594/1984 - loss 0.17027389 - time (sec): 9.18 - samples/sec: 5444.29 - lr: 0.000032 - momentum: 0.000000
141
+ 2023-10-18 21:27:40,347 epoch 5 - iter 792/1984 - loss 0.16509799 - time (sec): 12.18 - samples/sec: 5422.24 - lr: 0.000031 - momentum: 0.000000
142
+ 2023-10-18 21:27:43,351 epoch 5 - iter 990/1984 - loss 0.16344421 - time (sec): 15.18 - samples/sec: 5364.28 - lr: 0.000031 - momentum: 0.000000
143
+ 2023-10-18 21:27:46,391 epoch 5 - iter 1188/1984 - loss 0.16463321 - time (sec): 18.22 - samples/sec: 5341.67 - lr: 0.000030 - momentum: 0.000000
144
+ 2023-10-18 21:27:49,456 epoch 5 - iter 1386/1984 - loss 0.16486788 - time (sec): 21.29 - samples/sec: 5405.39 - lr: 0.000029 - momentum: 0.000000
145
+ 2023-10-18 21:27:52,493 epoch 5 - iter 1584/1984 - loss 0.16396812 - time (sec): 24.33 - samples/sec: 5388.54 - lr: 0.000029 - momentum: 0.000000
146
+ 2023-10-18 21:27:55,556 epoch 5 - iter 1782/1984 - loss 0.16439615 - time (sec): 27.39 - samples/sec: 5372.41 - lr: 0.000028 - momentum: 0.000000
147
+ 2023-10-18 21:27:58,492 epoch 5 - iter 1980/1984 - loss 0.16305676 - time (sec): 30.32 - samples/sec: 5399.14 - lr: 0.000028 - momentum: 0.000000
148
+ 2023-10-18 21:27:58,549 ----------------------------------------------------------------------------------------------------
149
+ 2023-10-18 21:27:58,549 EPOCH 5 done: loss 0.1629 - lr: 0.000028
150
+ 2023-10-18 21:28:00,391 DEV : loss 0.1396612972021103 - f1-score (micro avg) 0.6016
151
+ 2023-10-18 21:28:00,410 saving best model
152
+ 2023-10-18 21:28:00,443 ----------------------------------------------------------------------------------------------------
153
+ 2023-10-18 21:28:03,439 epoch 6 - iter 198/1984 - loss 0.16107360 - time (sec): 3.00 - samples/sec: 5139.73 - lr: 0.000027 - momentum: 0.000000
154
+ 2023-10-18 21:28:06,549 epoch 6 - iter 396/1984 - loss 0.15421664 - time (sec): 6.11 - samples/sec: 5364.46 - lr: 0.000027 - momentum: 0.000000
155
+ 2023-10-18 21:28:09,674 epoch 6 - iter 594/1984 - loss 0.15915795 - time (sec): 9.23 - samples/sec: 5330.25 - lr: 0.000026 - momentum: 0.000000
156
+ 2023-10-18 21:28:12,713 epoch 6 - iter 792/1984 - loss 0.15438849 - time (sec): 12.27 - samples/sec: 5364.46 - lr: 0.000026 - momentum: 0.000000
157
+ 2023-10-18 21:28:15,664 epoch 6 - iter 990/1984 - loss 0.15292486 - time (sec): 15.22 - samples/sec: 5399.57 - lr: 0.000025 - momentum: 0.000000
158
+ 2023-10-18 21:28:18,518 epoch 6 - iter 1188/1984 - loss 0.15391287 - time (sec): 18.07 - samples/sec: 5466.01 - lr: 0.000024 - momentum: 0.000000
159
+ 2023-10-18 21:28:21,258 epoch 6 - iter 1386/1984 - loss 0.15391731 - time (sec): 20.81 - samples/sec: 5532.31 - lr: 0.000024 - momentum: 0.000000
160
+ 2023-10-18 21:28:24,192 epoch 6 - iter 1584/1984 - loss 0.15364826 - time (sec): 23.75 - samples/sec: 5517.99 - lr: 0.000023 - momentum: 0.000000
161
+ 2023-10-18 21:28:27,215 epoch 6 - iter 1782/1984 - loss 0.15355381 - time (sec): 26.77 - samples/sec: 5506.02 - lr: 0.000023 - momentum: 0.000000
162
+ 2023-10-18 21:28:30,278 epoch 6 - iter 1980/1984 - loss 0.15357759 - time (sec): 29.84 - samples/sec: 5488.25 - lr: 0.000022 - momentum: 0.000000
163
+ 2023-10-18 21:28:30,343 ----------------------------------------------------------------------------------------------------
164
+ 2023-10-18 21:28:30,343 EPOCH 6 done: loss 0.1536 - lr: 0.000022
165
+ 2023-10-18 21:28:32,160 DEV : loss 0.14352329075336456 - f1-score (micro avg) 0.6138
166
+ 2023-10-18 21:28:32,179 saving best model
167
+ 2023-10-18 21:28:32,213 ----------------------------------------------------------------------------------------------------
168
+ 2023-10-18 21:28:35,313 epoch 7 - iter 198/1984 - loss 0.14010170 - time (sec): 3.10 - samples/sec: 5495.13 - lr: 0.000022 - momentum: 0.000000
169
+ 2023-10-18 21:28:38,375 epoch 7 - iter 396/1984 - loss 0.14331514 - time (sec): 6.16 - samples/sec: 5523.70 - lr: 0.000021 - momentum: 0.000000
170
+ 2023-10-18 21:28:41,433 epoch 7 - iter 594/1984 - loss 0.14376651 - time (sec): 9.22 - samples/sec: 5353.76 - lr: 0.000021 - momentum: 0.000000
171
+ 2023-10-18 21:28:44,525 epoch 7 - iter 792/1984 - loss 0.14327862 - time (sec): 12.31 - samples/sec: 5209.88 - lr: 0.000020 - momentum: 0.000000
172
+ 2023-10-18 21:28:47,677 epoch 7 - iter 990/1984 - loss 0.14545548 - time (sec): 15.46 - samples/sec: 5195.18 - lr: 0.000019 - momentum: 0.000000
173
+ 2023-10-18 21:28:50,714 epoch 7 - iter 1188/1984 - loss 0.14341143 - time (sec): 18.50 - samples/sec: 5255.03 - lr: 0.000019 - momentum: 0.000000
174
+ 2023-10-18 21:28:53,759 epoch 7 - iter 1386/1984 - loss 0.14255578 - time (sec): 21.54 - samples/sec: 5272.56 - lr: 0.000018 - momentum: 0.000000
175
+ 2023-10-18 21:28:56,904 epoch 7 - iter 1584/1984 - loss 0.14220705 - time (sec): 24.69 - samples/sec: 5262.27 - lr: 0.000018 - momentum: 0.000000
176
+ 2023-10-18 21:28:59,957 epoch 7 - iter 1782/1984 - loss 0.14227082 - time (sec): 27.74 - samples/sec: 5273.02 - lr: 0.000017 - momentum: 0.000000
177
+ 2023-10-18 21:29:03,032 epoch 7 - iter 1980/1984 - loss 0.14322288 - time (sec): 30.82 - samples/sec: 5315.07 - lr: 0.000017 - momentum: 0.000000
178
+ 2023-10-18 21:29:03,091 ----------------------------------------------------------------------------------------------------
179
+ 2023-10-18 21:29:03,091 EPOCH 7 done: loss 0.1432 - lr: 0.000017
180
+ 2023-10-18 21:29:04,914 DEV : loss 0.1460934579372406 - f1-score (micro avg) 0.6181
181
+ 2023-10-18 21:29:04,933 saving best model
182
+ 2023-10-18 21:29:04,967 ----------------------------------------------------------------------------------------------------
183
+ 2023-10-18 21:29:07,991 epoch 8 - iter 198/1984 - loss 0.15508407 - time (sec): 3.02 - samples/sec: 5680.49 - lr: 0.000016 - momentum: 0.000000
184
+ 2023-10-18 21:29:11,034 epoch 8 - iter 396/1984 - loss 0.14857609 - time (sec): 6.07 - samples/sec: 5447.32 - lr: 0.000016 - momentum: 0.000000
185
+ 2023-10-18 21:29:14,053 epoch 8 - iter 594/1984 - loss 0.14831674 - time (sec): 9.09 - samples/sec: 5539.95 - lr: 0.000015 - momentum: 0.000000
186
+ 2023-10-18 21:29:16,965 epoch 8 - iter 792/1984 - loss 0.14431698 - time (sec): 12.00 - samples/sec: 5580.93 - lr: 0.000014 - momentum: 0.000000
187
+ 2023-10-18 21:29:19,967 epoch 8 - iter 990/1984 - loss 0.14072716 - time (sec): 15.00 - samples/sec: 5527.79 - lr: 0.000014 - momentum: 0.000000
188
+ 2023-10-18 21:29:23,046 epoch 8 - iter 1188/1984 - loss 0.13730687 - time (sec): 18.08 - samples/sec: 5511.35 - lr: 0.000013 - momentum: 0.000000
189
+ 2023-10-18 21:29:26,169 epoch 8 - iter 1386/1984 - loss 0.13802672 - time (sec): 21.20 - samples/sec: 5447.35 - lr: 0.000013 - momentum: 0.000000
190
+ 2023-10-18 21:29:29,261 epoch 8 - iter 1584/1984 - loss 0.13646616 - time (sec): 24.29 - samples/sec: 5454.78 - lr: 0.000012 - momentum: 0.000000
191
+ 2023-10-18 21:29:32,269 epoch 8 - iter 1782/1984 - loss 0.13626809 - time (sec): 27.30 - samples/sec: 5441.22 - lr: 0.000012 - momentum: 0.000000
192
+ 2023-10-18 21:29:35,338 epoch 8 - iter 1980/1984 - loss 0.13563883 - time (sec): 30.37 - samples/sec: 5385.09 - lr: 0.000011 - momentum: 0.000000
193
+ 2023-10-18 21:29:35,401 ----------------------------------------------------------------------------------------------------
194
+ 2023-10-18 21:29:35,401 EPOCH 8 done: loss 0.1354 - lr: 0.000011
195
+ 2023-10-18 21:29:37,660 DEV : loss 0.14632448554039001 - f1-score (micro avg) 0.6125
196
+ 2023-10-18 21:29:37,681 ----------------------------------------------------------------------------------------------------
197
+ 2023-10-18 21:29:40,791 epoch 9 - iter 198/1984 - loss 0.12436606 - time (sec): 3.11 - samples/sec: 4996.85 - lr: 0.000011 - momentum: 0.000000
198
+ 2023-10-18 21:29:43,807 epoch 9 - iter 396/1984 - loss 0.12130140 - time (sec): 6.12 - samples/sec: 5243.22 - lr: 0.000010 - momentum: 0.000000
199
+ 2023-10-18 21:29:46,923 epoch 9 - iter 594/1984 - loss 0.12036860 - time (sec): 9.24 - samples/sec: 5222.59 - lr: 0.000009 - momentum: 0.000000
200
+ 2023-10-18 21:29:49,986 epoch 9 - iter 792/1984 - loss 0.12390287 - time (sec): 12.30 - samples/sec: 5269.18 - lr: 0.000009 - momentum: 0.000000
201
+ 2023-10-18 21:29:52,890 epoch 9 - iter 990/1984 - loss 0.12172877 - time (sec): 15.21 - samples/sec: 5357.59 - lr: 0.000008 - momentum: 0.000000
202
+ 2023-10-18 21:29:55,968 epoch 9 - iter 1188/1984 - loss 0.12359398 - time (sec): 18.29 - samples/sec: 5348.49 - lr: 0.000008 - momentum: 0.000000
203
+ 2023-10-18 21:29:59,087 epoch 9 - iter 1386/1984 - loss 0.12601590 - time (sec): 21.41 - samples/sec: 5335.60 - lr: 0.000007 - momentum: 0.000000
204
+ 2023-10-18 21:30:02,163 epoch 9 - iter 1584/1984 - loss 0.12731249 - time (sec): 24.48 - samples/sec: 5329.19 - lr: 0.000007 - momentum: 0.000000
205
+ 2023-10-18 21:30:05,015 epoch 9 - iter 1782/1984 - loss 0.13000702 - time (sec): 27.33 - samples/sec: 5411.53 - lr: 0.000006 - momentum: 0.000000
206
+ 2023-10-18 21:30:07,908 epoch 9 - iter 1980/1984 - loss 0.13026706 - time (sec): 30.23 - samples/sec: 5415.84 - lr: 0.000006 - momentum: 0.000000
207
+ 2023-10-18 21:30:07,967 ----------------------------------------------------------------------------------------------------
208
+ 2023-10-18 21:30:07,968 EPOCH 9 done: loss 0.1302 - lr: 0.000006
209
+ 2023-10-18 21:30:09,795 DEV : loss 0.14653360843658447 - f1-score (micro avg) 0.6208
210
+ 2023-10-18 21:30:09,816 saving best model
211
+ 2023-10-18 21:30:09,851 ----------------------------------------------------------------------------------------------------
212
+ 2023-10-18 21:30:13,178 epoch 10 - iter 198/1984 - loss 0.10287722 - time (sec): 3.33 - samples/sec: 4987.47 - lr: 0.000005 - momentum: 0.000000
213
+ 2023-10-18 21:30:16,230 epoch 10 - iter 396/1984 - loss 0.11633178 - time (sec): 6.38 - samples/sec: 5153.20 - lr: 0.000004 - momentum: 0.000000
214
+ 2023-10-18 21:30:19,312 epoch 10 - iter 594/1984 - loss 0.12476201 - time (sec): 9.46 - samples/sec: 5193.09 - lr: 0.000004 - momentum: 0.000000
215
+ 2023-10-18 21:30:22,352 epoch 10 - iter 792/1984 - loss 0.12536802 - time (sec): 12.50 - samples/sec: 5209.88 - lr: 0.000003 - momentum: 0.000000
216
+ 2023-10-18 21:30:25,421 epoch 10 - iter 990/1984 - loss 0.12695762 - time (sec): 15.57 - samples/sec: 5294.90 - lr: 0.000003 - momentum: 0.000000
217
+ 2023-10-18 21:30:28,513 epoch 10 - iter 1188/1984 - loss 0.12707600 - time (sec): 18.66 - samples/sec: 5283.15 - lr: 0.000002 - momentum: 0.000000
218
+ 2023-10-18 21:30:31,586 epoch 10 - iter 1386/1984 - loss 0.12684283 - time (sec): 21.73 - samples/sec: 5283.84 - lr: 0.000002 - momentum: 0.000000
219
+ 2023-10-18 21:30:34,749 epoch 10 - iter 1584/1984 - loss 0.12638192 - time (sec): 24.90 - samples/sec: 5252.68 - lr: 0.000001 - momentum: 0.000000
220
+ 2023-10-18 21:30:37,835 epoch 10 - iter 1782/1984 - loss 0.12660082 - time (sec): 27.98 - samples/sec: 5271.15 - lr: 0.000001 - momentum: 0.000000
221
+ 2023-10-18 21:30:40,835 epoch 10 - iter 1980/1984 - loss 0.12678588 - time (sec): 30.98 - samples/sec: 5285.51 - lr: 0.000000 - momentum: 0.000000
222
+ 2023-10-18 21:30:40,894 ----------------------------------------------------------------------------------------------------
223
+ 2023-10-18 21:30:40,894 EPOCH 10 done: loss 0.1267 - lr: 0.000000
224
+ 2023-10-18 21:30:42,746 DEV : loss 0.1484779566526413 - f1-score (micro avg) 0.6175
225
+ 2023-10-18 21:30:42,793 ----------------------------------------------------------------------------------------------------
226
+ 2023-10-18 21:30:42,793 Loading model from best epoch ...
227
+ 2023-10-18 21:30:42,875 SequenceTagger predicts: Dictionary with 13 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG
228
+ 2023-10-18 21:30:44,397
229
+ Results:
230
+ - F-score (micro) 0.6425
231
+ - F-score (macro) 0.4918
232
+ - Accuracy 0.5084
233
+
234
+ By class:
235
+ precision recall f1-score support
236
+
237
+ LOC 0.7192 0.7664 0.7421 655
238
+ PER 0.4534 0.6547 0.5358 223
239
+ ORG 0.3778 0.1339 0.1977 127
240
+
241
+ micro avg 0.6244 0.6617 0.6425 1005
242
+ macro avg 0.5168 0.5183 0.4918 1005
243
+ weighted avg 0.6171 0.6617 0.6275 1005
244
+
245
+ 2023-10-18 21:30:44,397 ----------------------------------------------------------------------------------------------------