Hugging Face's logo

stefan-it
/
hmbench-newseye-de-hmbert-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1

Token Classification
Flair
PyTorch
German
sequence-tagger-model
Model card Files Community
2
stefan-it commited on
Commit
97e16d5
•
1 Parent(s): 86be455

Upload folder using huggingface_hub

Browse files
Files changed (5) hide show
  1. best-model.pt +3 -0
  2. dev.tsv +0 -0
  3. loss.tsv +11 -0
  4. test.tsv +0 -0
  5. training.log +239 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed09f83229829f604b6f44690d94d537bdd4916a92bbb37ac8e127dbfa85aea9
3
+ size 443323527
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 12:35:35 0.0000 0.4018 0.1225 0.2074 0.2765 0.2370 0.1347
3
+ 2 12:40:03 0.0000 0.1744 0.1554 0.2755 0.5682 0.3711 0.2292
4
+ 3 12:44:32 0.0000 0.1165 0.3027 0.2323 0.6383 0.3406 0.2062
5
+ 4 12:48:57 0.0000 0.0809 0.2266 0.2685 0.4811 0.3446 0.2089
6
+ 5 12:53:21 0.0000 0.0633 0.2984 0.2451 0.5890 0.3461 0.2107
7
+ 6 12:57:49 0.0000 0.0425 0.3900 0.2718 0.5777 0.3697 0.2280
8
+ 7 13:02:10 0.0000 0.0323 0.4293 0.2582 0.5795 0.3573 0.2189
9
+ 8 13:06:30 0.0000 0.0216 0.4999 0.2488 0.5833 0.3488 0.2127
10
+ 9 13:10:56 0.0000 0.0164 0.4991 0.2629 0.6004 0.3656 0.2251
11
+ 10 13:15:19 0.0000 0.0110 0.5127 0.2716 0.5966 0.3732 0.2309
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-15 12:31:16,764 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-15 12:31:16,765 Model: "SequenceTagger(
3
+ (embeddings): TransformerWordEmbeddings(
4
+ (model): BertModel(
5
+ (embeddings): BertEmbeddings(
6
+ (word_embeddings): Embedding(32001, 768)
7
+ (position_embeddings): Embedding(512, 768)
8
+ (token_type_embeddings): Embedding(2, 768)
9
+ (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
10
+ (dropout): Dropout(p=0.1, inplace=False)
11
+ )
12
+ (encoder): BertEncoder(
13
+ (layer): ModuleList(
14
+ (0-11): 12 x BertLayer(
15
+ (attention): BertAttention(
16
+ (self): BertSelfAttention(
17
+ (query): Linear(in_features=768, out_features=768, bias=True)
18
+ (key): Linear(in_features=768, out_features=768, bias=True)
19
+ (value): Linear(in_features=768, out_features=768, bias=True)
20
+ (dropout): Dropout(p=0.1, inplace=False)
21
+ )
22
+ (output): BertSelfOutput(
23
+ (dense): Linear(in_features=768, out_features=768, bias=True)
24
+ (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
25
+ (dropout): Dropout(p=0.1, inplace=False)
26
+ )
27
+ )
28
+ (intermediate): BertIntermediate(
29
+ (dense): Linear(in_features=768, out_features=3072, bias=True)
30
+ (intermediate_act_fn): GELUActivation()
31
+ )
32
+ (output): BertOutput(
33
+ (dense): Linear(in_features=3072, out_features=768, bias=True)
34
+ (LayerNorm): LayerNorm((768,), 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=768, out_features=768, bias=True)
42
+ (activation): Tanh()
43
+ )
44
+ )
45
+ )
46
+ (locked_dropout): LockedDropout(p=0.5)
47
+ (linear): Linear(in_features=768, out_features=17, bias=True)
48
+ (loss_function): CrossEntropyLoss()
49
+ )"
50
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
51
+ 2023-10-15 12:31:16,765 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
52
+ - NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
53
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
54
+ 2023-10-15 12:31:16,765 Train: 20847 sentences
55
+ 2023-10-15 12:31:16,765 (train_with_dev=False, train_with_test=False)
56
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
57
+ 2023-10-15 12:31:16,765 Training Params:
58
+ 2023-10-15 12:31:16,765 - learning_rate: "3e-05"
59
+ 2023-10-15 12:31:16,765 - mini_batch_size: "4"
60
+ 2023-10-15 12:31:16,765 - max_epochs: "10"
61
+ 2023-10-15 12:31:16,765 - shuffle: "True"
62
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
63
+ 2023-10-15 12:31:16,765 Plugins:
64
+ 2023-10-15 12:31:16,765 - LinearScheduler | warmup_fraction: '0.1'
65
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
66
+ 2023-10-15 12:31:16,765 Final evaluation on model from best epoch (best-model.pt)
67
+ 2023-10-15 12:31:16,765 - metric: "('micro avg', 'f1-score')"
68
+ 2023-10-15 12:31:16,765 ----------------------------------------------------------------------------------------------------
69
+ 2023-10-15 12:31:16,765 Computation:
70
+ 2023-10-15 12:31:16,765 - compute on device: cuda:0
71
+ 2023-10-15 12:31:16,766 - embedding storage: none
72
+ 2023-10-15 12:31:16,766 ----------------------------------------------------------------------------------------------------
73
+ 2023-10-15 12:31:16,766 Model training base path: "hmbench-newseye/de-dbmdz/bert-base-historic-multilingual-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
74
+ 2023-10-15 12:31:16,766 ----------------------------------------------------------------------------------------------------
75
+ 2023-10-15 12:31:16,766 ----------------------------------------------------------------------------------------------------
76
+ 2023-10-15 12:31:42,416 epoch 1 - iter 521/5212 - loss 1.65107690 - time (sec): 25.65 - samples/sec: 1428.48 - lr: 0.000003 - momentum: 0.000000
77
+ 2023-10-15 12:32:07,452 epoch 1 - iter 1042/5212 - loss 1.04881765 - time (sec): 50.69 - samples/sec: 1445.79 - lr: 0.000006 - momentum: 0.000000
78
+ 2023-10-15 12:32:33,017 epoch 1 - iter 1563/5212 - loss 0.78409455 - time (sec): 76.25 - samples/sec: 1475.39 - lr: 0.000009 - momentum: 0.000000
79
+ 2023-10-15 12:32:58,619 epoch 1 - iter 2084/5212 - loss 0.66200413 - time (sec): 101.85 - samples/sec: 1457.97 - lr: 0.000012 - momentum: 0.000000
80
+ 2023-10-15 12:33:24,941 epoch 1 - iter 2605/5212 - loss 0.56564184 - time (sec): 128.17 - samples/sec: 1472.11 - lr: 0.000015 - momentum: 0.000000
81
+ 2023-10-15 12:33:50,287 epoch 1 - iter 3126/5212 - loss 0.51031280 - time (sec): 153.52 - samples/sec: 1469.15 - lr: 0.000018 - momentum: 0.000000
82
+ 2023-10-15 12:34:14,962 epoch 1 - iter 3647/5212 - loss 0.47660046 - time (sec): 178.20 - samples/sec: 1454.65 - lr: 0.000021 - momentum: 0.000000
83
+ 2023-10-15 12:34:40,269 epoch 1 - iter 4168/5212 - loss 0.44308784 - time (sec): 203.50 - samples/sec: 1461.60 - lr: 0.000024 - momentum: 0.000000
84
+ 2023-10-15 12:35:05,226 epoch 1 - iter 4689/5212 - loss 0.41975678 - time (sec): 228.46 - samples/sec: 1459.22 - lr: 0.000027 - momentum: 0.000000
85
+ 2023-10-15 12:35:29,725 epoch 1 - iter 5210/5212 - loss 0.40193518 - time (sec): 252.96 - samples/sec: 1452.18 - lr: 0.000030 - momentum: 0.000000
86
+ 2023-10-15 12:35:29,815 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-15 12:35:29,815 EPOCH 1 done: loss 0.4018 - lr: 0.000030
88
+ 2023-10-15 12:35:35,446 DEV : loss 0.12250665575265884 - f1-score (micro avg) 0.237
89
+ 2023-10-15 12:35:35,472 saving best model
90
+ 2023-10-15 12:35:35,862 ----------------------------------------------------------------------------------------------------
91
+ 2023-10-15 12:36:01,334 epoch 2 - iter 521/5212 - loss 0.17665338 - time (sec): 25.47 - samples/sec: 1495.90 - lr: 0.000030 - momentum: 0.000000
92
+ 2023-10-15 12:36:26,317 epoch 2 - iter 1042/5212 - loss 0.18260145 - time (sec): 50.45 - samples/sec: 1425.12 - lr: 0.000029 - momentum: 0.000000
93
+ 2023-10-15 12:36:51,509 epoch 2 - iter 1563/5212 - loss 0.18236373 - time (sec): 75.65 - samples/sec: 1440.60 - lr: 0.000029 - momentum: 0.000000
94
+ 2023-10-15 12:37:16,813 epoch 2 - iter 2084/5212 - loss 0.18783782 - time (sec): 100.95 - samples/sec: 1456.84 - lr: 0.000029 - momentum: 0.000000
95
+ 2023-10-15 12:37:42,017 epoch 2 - iter 2605/5212 - loss 0.18436223 - time (sec): 126.15 - samples/sec: 1460.36 - lr: 0.000028 - momentum: 0.000000
96
+ 2023-10-15 12:38:06,008 epoch 2 - iter 3126/5212 - loss 0.18087201 - time (sec): 150.14 - samples/sec: 1449.38 - lr: 0.000028 - momentum: 0.000000
97
+ 2023-10-15 12:38:34,760 epoch 2 - iter 3647/5212 - loss 0.18105599 - time (sec): 178.90 - samples/sec: 1430.39 - lr: 0.000028 - momentum: 0.000000
98
+ 2023-10-15 12:39:00,663 epoch 2 - iter 4168/5212 - loss 0.17564865 - time (sec): 204.80 - samples/sec: 1433.44 - lr: 0.000027 - momentum: 0.000000
99
+ 2023-10-15 12:39:29,513 epoch 2 - iter 4689/5212 - loss 0.17426936 - time (sec): 233.65 - samples/sec: 1428.02 - lr: 0.000027 - momentum: 0.000000
100
+ 2023-10-15 12:39:54,646 epoch 2 - iter 5210/5212 - loss 0.17441380 - time (sec): 258.78 - samples/sec: 1419.35 - lr: 0.000027 - momentum: 0.000000
101
+ 2023-10-15 12:39:54,738 ----------------------------------------------------------------------------------------------------
102
+ 2023-10-15 12:39:54,738 EPOCH 2 done: loss 0.1744 - lr: 0.000027
103
+ 2023-10-15 12:40:03,882 DEV : loss 0.15540799498558044 - f1-score (micro avg) 0.3711
104
+ 2023-10-15 12:40:03,908 saving best model
105
+ 2023-10-15 12:40:04,385 ----------------------------------------------------------------------------------------------------
106
+ 2023-10-15 12:40:30,445 epoch 3 - iter 521/5212 - loss 0.12465846 - time (sec): 26.05 - samples/sec: 1449.20 - lr: 0.000026 - momentum: 0.000000
107
+ 2023-10-15 12:40:57,362 epoch 3 - iter 1042/5212 - loss 0.12448844 - time (sec): 52.97 - samples/sec: 1383.88 - lr: 0.000026 - momentum: 0.000000
108
+ 2023-10-15 12:41:23,372 epoch 3 - iter 1563/5212 - loss 0.11976469 - time (sec): 78.98 - samples/sec: 1365.31 - lr: 0.000026 - momentum: 0.000000
109
+ 2023-10-15 12:41:48,900 epoch 3 - iter 2084/5212 - loss 0.12448630 - time (sec): 104.51 - samples/sec: 1375.60 - lr: 0.000025 - momentum: 0.000000
110
+ 2023-10-15 12:42:14,444 epoch 3 - iter 2605/5212 - loss 0.12482318 - time (sec): 130.05 - samples/sec: 1388.34 - lr: 0.000025 - momentum: 0.000000
111
+ 2023-10-15 12:42:39,582 epoch 3 - iter 3126/5212 - loss 0.12114157 - time (sec): 155.19 - samples/sec: 1404.67 - lr: 0.000025 - momentum: 0.000000
112
+ 2023-10-15 12:43:06,537 epoch 3 - iter 3647/5212 - loss 0.11954176 - time (sec): 182.14 - samples/sec: 1412.98 - lr: 0.000024 - momentum: 0.000000
113
+ 2023-10-15 12:43:31,875 epoch 3 - iter 4168/5212 - loss 0.11725032 - time (sec): 207.48 - samples/sec: 1412.38 - lr: 0.000024 - momentum: 0.000000
114
+ 2023-10-15 12:43:57,560 epoch 3 - iter 4689/5212 - loss 0.11753155 - time (sec): 233.17 - samples/sec: 1419.48 - lr: 0.000024 - momentum: 0.000000
115
+ 2023-10-15 12:44:22,998 epoch 3 - iter 5210/5212 - loss 0.11654077 - time (sec): 258.61 - samples/sec: 1420.61 - lr: 0.000023 - momentum: 0.000000
116
+ 2023-10-15 12:44:23,086 ----------------------------------------------------------------------------------------------------
117
+ 2023-10-15 12:44:23,086 EPOCH 3 done: loss 0.1165 - lr: 0.000023
118
+ 2023-10-15 12:44:32,250 DEV : loss 0.30272355675697327 - f1-score (micro avg) 0.3406
119
+ 2023-10-15 12:44:32,276 ----------------------------------------------------------------------------------------------------
120
+ 2023-10-15 12:44:57,267 epoch 4 - iter 521/5212 - loss 0.08646191 - time (sec): 24.99 - samples/sec: 1410.48 - lr: 0.000023 - momentum: 0.000000
121
+ 2023-10-15 12:45:23,235 epoch 4 - iter 1042/5212 - loss 0.07591061 - time (sec): 50.96 - samples/sec: 1388.32 - lr: 0.000023 - momentum: 0.000000
122
+ 2023-10-15 12:45:49,155 epoch 4 - iter 1563/5212 - loss 0.08070237 - time (sec): 76.88 - samples/sec: 1418.84 - lr: 0.000022 - momentum: 0.000000
123
+ 2023-10-15 12:46:14,293 epoch 4 - iter 2084/5212 - loss 0.08273905 - time (sec): 102.02 - samples/sec: 1431.21 - lr: 0.000022 - momentum: 0.000000
124
+ 2023-10-15 12:46:39,323 epoch 4 - iter 2605/5212 - loss 0.08254731 - time (sec): 127.05 - samples/sec: 1425.41 - lr: 0.000022 - momentum: 0.000000
125
+ 2023-10-15 12:47:05,403 epoch 4 - iter 3126/5212 - loss 0.08232248 - time (sec): 153.13 - samples/sec: 1431.72 - lr: 0.000021 - momentum: 0.000000
126
+ 2023-10-15 12:47:31,031 epoch 4 - iter 3647/5212 - loss 0.08272838 - time (sec): 178.75 - samples/sec: 1429.12 - lr: 0.000021 - momentum: 0.000000
127
+ 2023-10-15 12:47:56,348 epoch 4 - iter 4168/5212 - loss 0.08217946 - time (sec): 204.07 - samples/sec: 1437.95 - lr: 0.000021 - momentum: 0.000000
128
+ 2023-10-15 12:48:21,768 epoch 4 - iter 4689/5212 - loss 0.08130461 - time (sec): 229.49 - samples/sec: 1431.79 - lr: 0.000020 - momentum: 0.000000
129
+ 2023-10-15 12:48:47,792 epoch 4 - iter 5210/5212 - loss 0.08086252 - time (sec): 255.51 - samples/sec: 1437.58 - lr: 0.000020 - momentum: 0.000000
130
+ 2023-10-15 12:48:47,885 ----------------------------------------------------------------------------------------------------
131
+ 2023-10-15 12:48:47,885 EPOCH 4 done: loss 0.0809 - lr: 0.000020
132
+ 2023-10-15 12:48:57,739 DEV : loss 0.22659707069396973 - f1-score (micro avg) 0.3446
133
+ 2023-10-15 12:48:57,775 ----------------------------------------------------------------------------------------------------
134
+ 2023-10-15 12:49:24,168 epoch 5 - iter 521/5212 - loss 0.06166189 - time (sec): 26.39 - samples/sec: 1361.81 - lr: 0.000020 - momentum: 0.000000
135
+ 2023-10-15 12:49:48,880 epoch 5 - iter 1042/5212 - loss 0.05657890 - time (sec): 51.10 - samples/sec: 1384.21 - lr: 0.000019 - momentum: 0.000000
136
+ 2023-10-15 12:50:14,462 epoch 5 - iter 1563/5212 - loss 0.06599464 - time (sec): 76.68 - samples/sec: 1396.70 - lr: 0.000019 - momentum: 0.000000
137
+ 2023-10-15 12:50:40,037 epoch 5 - iter 2084/5212 - loss 0.06735536 - time (sec): 102.26 - samples/sec: 1407.44 - lr: 0.000019 - momentum: 0.000000
138
+ 2023-10-15 12:51:05,650 epoch 5 - iter 2605/5212 - loss 0.06616069 - time (sec): 127.87 - samples/sec: 1406.81 - lr: 0.000018 - momentum: 0.000000
139
+ 2023-10-15 12:51:31,087 epoch 5 - iter 3126/5212 - loss 0.06521237 - time (sec): 153.31 - samples/sec: 1420.29 - lr: 0.000018 - momentum: 0.000000
140
+ 2023-10-15 12:51:56,491 epoch 5 - iter 3647/5212 - loss 0.06386659 - time (sec): 178.71 - samples/sec: 1428.75 - lr: 0.000018 - momentum: 0.000000
141
+ 2023-10-15 12:52:22,206 epoch 5 - iter 4168/5212 - loss 0.06327760 - time (sec): 204.43 - samples/sec: 1441.60 - lr: 0.000017 - momentum: 0.000000
142
+ 2023-10-15 12:52:47,551 epoch 5 - iter 4689/5212 - loss 0.06392307 - time (sec): 229.77 - samples/sec: 1450.51 - lr: 0.000017 - momentum: 0.000000
143
+ 2023-10-15 12:53:12,302 epoch 5 - iter 5210/5212 - loss 0.06327392 - time (sec): 254.52 - samples/sec: 1443.39 - lr: 0.000017 - momentum: 0.000000
144
+ 2023-10-15 12:53:12,393 ----------------------------------------------------------------------------------------------------
145
+ 2023-10-15 12:53:12,393 EPOCH 5 done: loss 0.0633 - lr: 0.000017
146
+ 2023-10-15 12:53:21,387 DEV : loss 0.29836305975914 - f1-score (micro avg) 0.3461
147
+ 2023-10-15 12:53:21,415 ----------------------------------------------------------------------------------------------------
148
+ 2023-10-15 12:53:46,882 epoch 6 - iter 521/5212 - loss 0.03648971 - time (sec): 25.47 - samples/sec: 1543.84 - lr: 0.000016 - momentum: 0.000000
149
+ 2023-10-15 12:54:12,031 epoch 6 - iter 1042/5212 - loss 0.03491365 - time (sec): 50.62 - samples/sec: 1509.08 - lr: 0.000016 - momentum: 0.000000
150
+ 2023-10-15 12:54:37,381 epoch 6 - iter 1563/5212 - loss 0.03950362 - time (sec): 75.97 - samples/sec: 1525.17 - lr: 0.000016 - momentum: 0.000000
151
+ 2023-10-15 12:55:01,670 epoch 6 - iter 2084/5212 - loss 0.04166816 - time (sec): 100.25 - samples/sec: 1500.88 - lr: 0.000015 - momentum: 0.000000
152
+ 2023-10-15 12:55:26,198 epoch 6 - iter 2605/5212 - loss 0.04137428 - time (sec): 124.78 - samples/sec: 1468.52 - lr: 0.000015 - momentum: 0.000000
153
+ 2023-10-15 12:55:51,598 epoch 6 - iter 3126/5212 - loss 0.04172807 - time (sec): 150.18 - samples/sec: 1473.04 - lr: 0.000015 - momentum: 0.000000
154
+ 2023-10-15 12:56:18,660 epoch 6 - iter 3647/5212 - loss 0.04148080 - time (sec): 177.24 - samples/sec: 1436.93 - lr: 0.000014 - momentum: 0.000000
155
+ 2023-10-15 12:56:48,113 epoch 6 - iter 4168/5212 - loss 0.04185072 - time (sec): 206.70 - samples/sec: 1425.39 - lr: 0.000014 - momentum: 0.000000
156
+ 2023-10-15 12:57:14,874 epoch 6 - iter 4689/5212 - loss 0.04234827 - time (sec): 233.46 - samples/sec: 1423.26 - lr: 0.000014 - momentum: 0.000000
157
+ 2023-10-15 12:57:39,407 epoch 6 - iter 5210/5212 - loss 0.04248019 - time (sec): 257.99 - samples/sec: 1423.91 - lr: 0.000013 - momentum: 0.000000
158
+ 2023-10-15 12:57:39,496 ----------------------------------------------------------------------------------------------------
159
+ 2023-10-15 12:57:39,496 EPOCH 6 done: loss 0.0425 - lr: 0.000013
160
+ 2023-10-15 12:57:49,189 DEV : loss 0.38996848464012146 - f1-score (micro avg) 0.3697
161
+ 2023-10-15 12:57:49,225 ----------------------------------------------------------------------------------------------------
162
+ 2023-10-15 12:58:14,088 epoch 7 - iter 521/5212 - loss 0.03048326 - time (sec): 24.86 - samples/sec: 1418.67 - lr: 0.000013 - momentum: 0.000000
163
+ 2023-10-15 12:58:41,269 epoch 7 - iter 1042/5212 - loss 0.02866313 - time (sec): 52.04 - samples/sec: 1386.42 - lr: 0.000013 - momentum: 0.000000
164
+ 2023-10-15 12:59:06,440 epoch 7 - iter 1563/5212 - loss 0.03616746 - time (sec): 77.21 - samples/sec: 1419.60 - lr: 0.000012 - momentum: 0.000000
165
+ 2023-10-15 12:59:31,369 epoch 7 - iter 2084/5212 - loss 0.03622811 - time (sec): 102.14 - samples/sec: 1435.71 - lr: 0.000012 - momentum: 0.000000
166
+ 2023-10-15 12:59:56,319 epoch 7 - iter 2605/5212 - loss 0.03386848 - time (sec): 127.09 - samples/sec: 1440.15 - lr: 0.000012 - momentum: 0.000000
167
+ 2023-10-15 13:00:21,524 epoch 7 - iter 3126/5212 - loss 0.03403511 - time (sec): 152.29 - samples/sec: 1443.79 - lr: 0.000011 - momentum: 0.000000
168
+ 2023-10-15 13:00:47,278 epoch 7 - iter 3647/5212 - loss 0.03296196 - time (sec): 178.05 - samples/sec: 1452.00 - lr: 0.000011 - momentum: 0.000000
169
+ 2023-10-15 13:01:12,245 epoch 7 - iter 4168/5212 - loss 0.03210474 - time (sec): 203.02 - samples/sec: 1450.16 - lr: 0.000011 - momentum: 0.000000
170
+ 2023-10-15 13:01:37,095 epoch 7 - iter 4689/5212 - loss 0.03179437 - time (sec): 227.87 - samples/sec: 1449.99 - lr: 0.000010 - momentum: 0.000000
171
+ 2023-10-15 13:02:01,984 epoch 7 - iter 5210/5212 - loss 0.03231694 - time (sec): 252.75 - samples/sec: 1452.59 - lr: 0.000010 - momentum: 0.000000
172
+ 2023-10-15 13:02:02,089 ----------------------------------------------------------------------------------------------------
173
+ 2023-10-15 13:02:02,089 EPOCH 7 done: loss 0.0323 - lr: 0.000010
174
+ 2023-10-15 13:02:10,410 DEV : loss 0.4292934238910675 - f1-score (micro avg) 0.3573
175
+ 2023-10-15 13:02:10,437 ----------------------------------------------------------------------------------------------------
176
+ 2023-10-15 13:02:36,582 epoch 8 - iter 521/5212 - loss 0.01718034 - time (sec): 26.14 - samples/sec: 1431.13 - lr: 0.000010 - momentum: 0.000000
177
+ 2023-10-15 13:03:01,497 epoch 8 - iter 1042/5212 - loss 0.02296624 - time (sec): 51.06 - samples/sec: 1422.71 - lr: 0.000009 - momentum: 0.000000
178
+ 2023-10-15 13:03:26,494 epoch 8 - iter 1563/5212 - loss 0.02050494 - time (sec): 76.06 - samples/sec: 1438.11 - lr: 0.000009 - momentum: 0.000000
179
+ 2023-10-15 13:03:51,390 epoch 8 - iter 2084/5212 - loss 0.02052023 - time (sec): 100.95 - samples/sec: 1434.50 - lr: 0.000009 - momentum: 0.000000
180
+ 2023-10-15 13:04:16,470 epoch 8 - iter 2605/5212 - loss 0.02085292 - time (sec): 126.03 - samples/sec: 1435.46 - lr: 0.000008 - momentum: 0.000000
181
+ 2023-10-15 13:04:42,255 epoch 8 - iter 3126/5212 - loss 0.02097252 - time (sec): 151.82 - samples/sec: 1453.03 - lr: 0.000008 - momentum: 0.000000
182
+ 2023-10-15 13:05:07,098 epoch 8 - iter 3647/5212 - loss 0.02219826 - time (sec): 176.66 - samples/sec: 1460.28 - lr: 0.000008 - momentum: 0.000000
183
+ 2023-10-15 13:05:32,019 epoch 8 - iter 4168/5212 - loss 0.02231714 - time (sec): 201.58 - samples/sec: 1453.62 - lr: 0.000007 - momentum: 0.000000
184
+ 2023-10-15 13:05:57,035 epoch 8 - iter 4689/5212 - loss 0.02169589 - time (sec): 226.60 - samples/sec: 1456.33 - lr: 0.000007 - momentum: 0.000000
185
+ 2023-10-15 13:06:22,253 epoch 8 - iter 5210/5212 - loss 0.02163248 - time (sec): 251.81 - samples/sec: 1458.53 - lr: 0.000007 - momentum: 0.000000
186
+ 2023-10-15 13:06:22,352 ----------------------------------------------------------------------------------------------------
187
+ 2023-10-15 13:06:22,352 EPOCH 8 done: loss 0.0216 - lr: 0.000007
188
+ 2023-10-15 13:06:30,833 DEV : loss 0.49988579750061035 - f1-score (micro avg) 0.3488
189
+ 2023-10-15 13:06:30,892 ----------------------------------------------------------------------------------------------------
190
+ 2023-10-15 13:06:57,558 epoch 9 - iter 521/5212 - loss 0.01767213 - time (sec): 26.66 - samples/sec: 1480.44 - lr: 0.000006 - momentum: 0.000000
191
+ 2023-10-15 13:07:22,775 epoch 9 - iter 1042/5212 - loss 0.01672506 - time (sec): 51.88 - samples/sec: 1484.59 - lr: 0.000006 - momentum: 0.000000
192
+ 2023-10-15 13:07:47,441 epoch 9 - iter 1563/5212 - loss 0.01737499 - time (sec): 76.55 - samples/sec: 1445.12 - lr: 0.000006 - momentum: 0.000000
193
+ 2023-10-15 13:08:13,548 epoch 9 - iter 2084/5212 - loss 0.01745047 - time (sec): 102.65 - samples/sec: 1444.23 - lr: 0.000005 - momentum: 0.000000
194
+ 2023-10-15 13:08:38,703 epoch 9 - iter 2605/5212 - loss 0.01740294 - time (sec): 127.81 - samples/sec: 1446.33 - lr: 0.000005 - momentum: 0.000000
195
+ 2023-10-15 13:09:04,305 epoch 9 - iter 3126/5212 - loss 0.01646355 - time (sec): 153.41 - samples/sec: 1437.32 - lr: 0.000005 - momentum: 0.000000
196
+ 2023-10-15 13:09:30,090 epoch 9 - iter 3647/5212 - loss 0.01604700 - time (sec): 179.20 - samples/sec: 1432.55 - lr: 0.000004 - momentum: 0.000000
197
+ 2023-10-15 13:09:55,749 epoch 9 - iter 4168/5212 - loss 0.01657625 - time (sec): 204.85 - samples/sec: 1431.01 - lr: 0.000004 - momentum: 0.000000
198
+ 2023-10-15 13:10:20,975 epoch 9 - iter 4689/5212 - loss 0.01653058 - time (sec): 230.08 - samples/sec: 1422.36 - lr: 0.000004 - momentum: 0.000000
199
+ 2023-10-15 13:10:48,149 epoch 9 - iter 5210/5212 - loss 0.01636217 - time (sec): 257.25 - samples/sec: 1427.84 - lr: 0.000003 - momentum: 0.000000
200
+ 2023-10-15 13:10:48,244 ----------------------------------------------------------------------------------------------------
201
+ 2023-10-15 13:10:48,244 EPOCH 9 done: loss 0.0164 - lr: 0.000003
202
+ 2023-10-15 13:10:56,680 DEV : loss 0.49910324811935425 - f1-score (micro avg) 0.3656
203
+ 2023-10-15 13:10:56,713 ----------------------------------------------------------------------------------------------------
204
+ 2023-10-15 13:11:23,057 epoch 10 - iter 521/5212 - loss 0.00940569 - time (sec): 26.34 - samples/sec: 1454.27 - lr: 0.000003 - momentum: 0.000000
205
+ 2023-10-15 13:11:48,413 epoch 10 - iter 1042/5212 - loss 0.00987505 - time (sec): 51.70 - samples/sec: 1470.96 - lr: 0.000003 - momentum: 0.000000
206
+ 2023-10-15 13:12:14,305 epoch 10 - iter 1563/5212 - loss 0.01023767 - time (sec): 77.59 - samples/sec: 1483.17 - lr: 0.000002 - momentum: 0.000000
207
+ 2023-10-15 13:12:39,460 epoch 10 - iter 2084/5212 - loss 0.01032580 - time (sec): 102.75 - samples/sec: 1482.32 - lr: 0.000002 - momentum: 0.000000
208
+ 2023-10-15 13:13:04,679 epoch 10 - iter 2605/5212 - loss 0.01108425 - time (sec): 127.96 - samples/sec: 1472.25 - lr: 0.000002 - momentum: 0.000000
209
+ 2023-10-15 13:13:29,443 epoch 10 - iter 3126/5212 - loss 0.01123512 - time (sec): 152.73 - samples/sec: 1451.98 - lr: 0.000001 - momentum: 0.000000
210
+ 2023-10-15 13:13:54,441 epoch 10 - iter 3647/5212 - loss 0.01086175 - time (sec): 177.73 - samples/sec: 1450.39 - lr: 0.000001 - momentum: 0.000000
211
+ 2023-10-15 13:14:19,603 epoch 10 - iter 4168/5212 - loss 0.01095131 - time (sec): 202.89 - samples/sec: 1451.33 - lr: 0.000001 - momentum: 0.000000
212
+ 2023-10-15 13:14:45,781 epoch 10 - iter 4689/5212 - loss 0.01104708 - time (sec): 229.07 - samples/sec: 1450.56 - lr: 0.000000 - momentum: 0.000000
213
+ 2023-10-15 13:15:10,852 epoch 10 - iter 5210/5212 - loss 0.01095808 - time (sec): 254.14 - samples/sec: 1445.55 - lr: 0.000000 - momentum: 0.000000
214
+ 2023-10-15 13:15:10,942 ----------------------------------------------------------------------------------------------------
215
+ 2023-10-15 13:15:10,942 EPOCH 10 done: loss 0.0110 - lr: 0.000000
216
+ 2023-10-15 13:15:19,421 DEV : loss 0.5126752257347107 - f1-score (micro avg) 0.3732
217
+ 2023-10-15 13:15:19,464 saving best model
218
+ 2023-10-15 13:15:20,314 ----------------------------------------------------------------------------------------------------
219
+ 2023-10-15 13:15:20,315 Loading model from best epoch ...
220
+ 2023-10-15 13:15:21,741 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
221
+ 2023-10-15 13:15:38,600
222
+ Results:
223
+ - F-score (micro) 0.4948
224
+ - F-score (macro) 0.3328
225
+ - Accuracy 0.3335
226
+
227
+ By class:
228
+ precision recall f1-score support
229
+
230
+ LOC 0.5075 0.6680 0.5768 1214
231
+ PER 0.4158 0.4889 0.4494 808
232
+ ORG 0.2912 0.3201 0.3050 353
233
+ HumanProd 0.0000 0.0000 0.0000 15
234
+
235
+ micro avg 0.4485 0.5519 0.4948 2390
236
+ macro avg 0.3036 0.3693 0.3328 2390
237
+ weighted avg 0.4414 0.5519 0.4900 2390
238
+
239
+ 2023-10-15 13:15:38,600 ----------------------------------------------------------------------------------------------------