Hugging Face's logo

beki
/
flair-pii-distilbert

Token Classification
Flair
PyTorch
English
sequence-tagger-model
Model card Files
xet
 Community
1
beki commited on
Commit
89f55b6
·
1 Parent(s): 6c8ebaa

Upload 5 files

Browse files
Files changed (6) hide show
  1. .gitattributes +1 -0
  2. dev.tsv +3 -0
  3. loss.tsv +21 -0
  4. test.tsv +0 -0
  5. training.log +491 -0
  6. weights.txt +0 -0
.gitattributes CHANGED
@@ -30,3 +30,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
30
  *.zip filter=lfs diff=lfs merge=lfs -text
31
  *.zst filter=lfs diff=lfs merge=lfs -text
32
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
30
  *.zip filter=lfs diff=lfs merge=lfs -text
31
  *.zst filter=lfs diff=lfs merge=lfs -text
32
  *tfevents* filter=lfs diff=lfs merge=lfs -text
33
+ dev.tsv filter=lfs diff=lfs merge=lfs -text
dev.tsv ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92f892c2879578f5b2652a0f0b22c97409b93dfc4526e0746c224cad4de177ff
3
+ size 16231537
loss.tsv ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ EPOCH TIMESTAMP BAD_EPOCHS LEARNING_RATE TRAIN_LOSS DEV_LOSS DEV_PRECISION DEV_RECALL DEV_F1 DEV_ACCURACY
2
+ 1 16:13:31 4 0.0000 0.5167920140669576 0.07261496782302856 0.6999 0.7008 0.7003 0.5529
3
+ 2 17:48:56 4 0.0000 0.19001624853523155 0.019461628049612045 0.9127 0.9391 0.9258 0.8711
4
+ 3 19:24:46 4 0.0000 0.17175931267209962 0.0132540138438344 0.9353 0.9548 0.9449 0.9035
5
+ 4 21:00:51 4 0.0000 0.1651708900129011 0.011177059262990952 0.9487 0.9583 0.9535 0.9179
6
+ 5 22:35:13 4 0.0000 0.16130743654362578 0.011488113552331924 0.9462 0.9631 0.9546 0.9199
7
+ 6 00:10:46 4 0.0000 0.15800901282123106 0.010615515522658825 0.9527 0.9649 0.9587 0.9273
8
+ 7 01:46:49 4 0.0000 0.1551098387415943 0.010866315104067326 0.9523 0.9673 0.9597 0.9289
9
+ 8 03:22:53 4 0.0000 0.1532771505682582 0.010451321490108967 0.9578 0.9657 0.9617 0.9324
10
+ 9 04:58:17 4 0.0000 0.15132318660084354 0.01064694207161665 0.9543 0.9677 0.9609 0.9309
11
+ 10 06:33:35 4 0.0000 0.14953294716354223 0.010687584988772869 0.9559 0.9689 0.9623 0.9336
12
+ 11 08:08:47 4 0.0000 0.14839159017834289 0.010935463011264801 0.9559 0.9683 0.962 0.9329
13
+ 12 09:43:47 4 0.0000 0.14699742678882574 0.011056484654545784 0.9587 0.9682 0.9634 0.9355
14
+ 13 11:14:57 4 0.0000 0.14604769509499 0.011409671977162361 0.9569 0.9687 0.9628 0.9342
15
+ 14 12:44:53 4 0.0000 0.14518390266473472 0.011419754475355148 0.9577 0.9697 0.9637 0.936
16
+ 15 14:24:07 4 0.0000 0.1443252341730405 0.011627680622041225 0.9582 0.9693 0.9637 0.9359
17
+ 16 16:05:36 4 0.0000 0.1435298321123121 0.011783876456320286 0.9601 0.9688 0.9644 0.9373
18
+ 17 17:46:47 4 0.0000 0.14356430696292566 0.011797642335295677 0.9595 0.9691 0.9643 0.9371
19
+ 18 19:28:18 4 0.0000 0.1423554343151879 0.011939478106796741 0.9588 0.9693 0.964 0.9365
20
+ 19 21:09:43 4 0.0000 0.14288157071177124 0.012016847729682922 0.9594 0.9692 0.9643 0.937
21
+ 20 22:51:01 4 0.0000 0.14238437937592288 0.012119622901082039 0.9589 0.9691 0.964 0.9366
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,491 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2022-10-09 14:35:47,018 ----------------------------------------------------------------------------------------------------
2
+ 2022-10-09 14:35:47,019 Model: "SequenceTagger(
3
+ (embeddings): StackedEmbeddings(
4
+ (list_embedding_0): TransformerWordEmbeddings(
5
+ (model): DistilBertModel(
6
+ (embeddings): Embeddings(
7
+ (word_embeddings): Embedding(28996, 768, padding_idx=0)
8
+ (position_embeddings): Embedding(512, 768)
9
+ (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
10
+ (dropout): Dropout(p=0.1, inplace=False)
11
+ )
12
+ (transformer): Transformer(
13
+ (layer): ModuleList(
14
+ (0): TransformerBlock(
15
+ (attention): MultiHeadSelfAttention(
16
+ (dropout): Dropout(p=0.1, inplace=False)
17
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
18
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
19
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
20
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
21
+ )
22
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
23
+ (ffn): FFN(
24
+ (dropout): Dropout(p=0.1, inplace=False)
25
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
26
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
27
+ (activation): GELUActivation()
28
+ )
29
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
30
+ )
31
+ (1): TransformerBlock(
32
+ (attention): MultiHeadSelfAttention(
33
+ (dropout): Dropout(p=0.1, inplace=False)
34
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
35
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
36
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
37
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
38
+ )
39
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
40
+ (ffn): FFN(
41
+ (dropout): Dropout(p=0.1, inplace=False)
42
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
43
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
44
+ (activation): GELUActivation()
45
+ )
46
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
47
+ )
48
+ (2): TransformerBlock(
49
+ (attention): MultiHeadSelfAttention(
50
+ (dropout): Dropout(p=0.1, inplace=False)
51
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
52
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
53
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
54
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
55
+ )
56
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
57
+ (ffn): FFN(
58
+ (dropout): Dropout(p=0.1, inplace=False)
59
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
60
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
61
+ (activation): GELUActivation()
62
+ )
63
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
64
+ )
65
+ (3): TransformerBlock(
66
+ (attention): MultiHeadSelfAttention(
67
+ (dropout): Dropout(p=0.1, inplace=False)
68
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
69
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
70
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
71
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
72
+ )
73
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
74
+ (ffn): FFN(
75
+ (dropout): Dropout(p=0.1, inplace=False)
76
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
77
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
78
+ (activation): GELUActivation()
79
+ )
80
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
81
+ )
82
+ (4): TransformerBlock(
83
+ (attention): MultiHeadSelfAttention(
84
+ (dropout): Dropout(p=0.1, inplace=False)
85
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
86
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
87
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
88
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
89
+ )
90
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
91
+ (ffn): FFN(
92
+ (dropout): Dropout(p=0.1, inplace=False)
93
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
94
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
95
+ (activation): GELUActivation()
96
+ )
97
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
98
+ )
99
+ (5): TransformerBlock(
100
+ (attention): MultiHeadSelfAttention(
101
+ (dropout): Dropout(p=0.1, inplace=False)
102
+ (q_lin): Linear(in_features=768, out_features=768, bias=True)
103
+ (k_lin): Linear(in_features=768, out_features=768, bias=True)
104
+ (v_lin): Linear(in_features=768, out_features=768, bias=True)
105
+ (out_lin): Linear(in_features=768, out_features=768, bias=True)
106
+ )
107
+ (sa_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
108
+ (ffn): FFN(
109
+ (dropout): Dropout(p=0.1, inplace=False)
110
+ (lin1): Linear(in_features=768, out_features=3072, bias=True)
111
+ (lin2): Linear(in_features=3072, out_features=768, bias=True)
112
+ (activation): GELUActivation()
113
+ )
114
+ (output_layer_norm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
115
+ )
116
+ )
117
+ )
118
+ )
119
+ )
120
+ )
121
+ (word_dropout): WordDropout(p=0.05)
122
+ (locked_dropout): LockedDropout(p=0.5)
123
+ (linear): Linear(in_features=768, out_features=21, bias=True)
124
+ (loss_function): CrossEntropyLoss()
125
+ )"
126
+ 2022-10-09 14:35:47,020 ----------------------------------------------------------------------------------------------------
127
+ 2022-10-09 14:35:47,020 Corpus: "MultiCorpus: 126439 train + 28967 dev + 17625 test sentences
128
+ - ColumnCorpus Corpus: 14896 train + 3444 dev + 3679 test sentences - ./
129
+ - ColumnCorpus Corpus: 1491 train + 166 dev + 184 test sentences - ./
130
+ - ColumnCorpus Corpus: 65087 train + 18419 dev + 9176 test sentences - ./datasets
131
+ - ColumnCorpus Corpus: 44965 train + 6938 dev + 4586 test sentences - ./"
132
+ 2022-10-09 14:35:47,020 ----------------------------------------------------------------------------------------------------
133
+ 2022-10-09 14:35:47,020 Parameters:
134
+ 2022-10-09 14:35:47,020 - learning_rate: "0.000005"
135
+ 2022-10-09 14:35:47,020 - mini_batch_size: "32"
136
+ 2022-10-09 14:35:47,020 - patience: "3"
137
+ 2022-10-09 14:35:47,020 - anneal_factor: "0.5"
138
+ 2022-10-09 14:35:47,020 - max_epochs: "20"
139
+ 2022-10-09 14:35:47,020 - shuffle: "True"
140
+ 2022-10-09 14:35:47,020 - train_with_dev: "False"
141
+ 2022-10-09 14:35:47,021 - batch_growth_annealing: "False"
142
+ 2022-10-09 14:35:47,021 ----------------------------------------------------------------------------------------------------
143
+ 2022-10-09 14:35:47,021 Model training base path: "resources/taggers/privy-flair-transformers"
144
+ 2022-10-09 14:35:47,021 ----------------------------------------------------------------------------------------------------
145
+ 2022-10-09 14:35:47,021 Device: cuda:0
146
+ 2022-10-09 14:35:47,021 ----------------------------------------------------------------------------------------------------
147
+ 2022-10-09 14:35:47,021 Embeddings storage mode: none
148
+ 2022-10-09 14:35:47,021 ----------------------------------------------------------------------------------------------------
149
+ 2022-10-09 14:41:45,282 epoch 1 - iter 395/3952 - loss 3.32419044 - samples/sec: 35.64 - lr: 0.000000
150
+ 2022-10-09 14:50:42,225 epoch 1 - iter 790/3952 - loss 1.82346877 - samples/sec: 24.23 - lr: 0.000000
151
+ 2022-10-09 15:00:44,300 epoch 1 - iter 1185/3952 - loss 1.06483796 - samples/sec: 21.66 - lr: 0.000001
152
+ 2022-10-09 15:10:53,476 epoch 1 - iter 1580/3952 - loss 0.79311831 - samples/sec: 21.46 - lr: 0.000001
153
+ 2022-10-09 15:20:53,647 epoch 1 - iter 1975/3952 - loss 0.65220017 - samples/sec: 21.79 - lr: 0.000001
154
+ 2022-10-09 15:30:48,260 epoch 1 - iter 2370/3952 - loss 0.56201630 - samples/sec: 21.92 - lr: 0.000001
155
+ 2022-10-09 15:38:37,611 epoch 1 - iter 2765/3952 - loss 0.53726885 - samples/sec: 27.75 - lr: 0.000002
156
+ 2022-10-09 15:44:53,320 epoch 1 - iter 3160/3952 - loss 0.53328468 - samples/sec: 34.63 - lr: 0.000002
157
+ 2022-10-09 15:51:16,972 epoch 1 - iter 3555/3952 - loss 0.52470503 - samples/sec: 33.80 - lr: 0.000002
158
+ 2022-10-09 15:57:35,830 epoch 1 - iter 3950/3952 - loss 0.51681052 - samples/sec: 34.26 - lr: 0.000002
159
+ 2022-10-09 15:57:37,275 ----------------------------------------------------------------------------------------------------
160
+ 2022-10-09 15:57:37,275 EPOCH 1 done: loss 0.5168 - lr 0.000002
161
+ 2022-10-09 16:12:59,483 Evaluating as a multi-label problem: False
162
+ 2022-10-09 16:12:59,975 DEV : loss 0.07261496782302856 - f1-score (micro avg) 0.7003
163
+ 2022-10-09 16:13:31,047 BAD EPOCHS (no improvement): 4
164
+ 2022-10-09 16:13:31,940 ----------------------------------------------------------------------------------------------------
165
+ 2022-10-09 16:21:39,781 epoch 2 - iter 395/3952 - loss 0.21218652 - samples/sec: 26.89 - lr: 0.000003
166
+ 2022-10-09 16:29:46,206 epoch 2 - iter 790/3952 - loss 0.20655105 - samples/sec: 26.79 - lr: 0.000003
167
+ 2022-10-09 16:37:52,936 epoch 2 - iter 1185/3952 - loss 0.20259102 - samples/sec: 26.73 - lr: 0.000003
168
+ 2022-10-09 16:45:58,507 epoch 2 - iter 1580/3952 - loss 0.20005535 - samples/sec: 26.81 - lr: 0.000003
169
+ 2022-10-09 16:53:52,122 epoch 2 - iter 1975/3952 - loss 0.19747189 - samples/sec: 27.49 - lr: 0.000004
170
+ 2022-10-09 17:01:39,634 epoch 2 - iter 2370/3952 - loss 0.19566392 - samples/sec: 27.76 - lr: 0.000004
171
+ 2022-10-09 17:09:30,471 epoch 2 - iter 2765/3952 - loss 0.19386487 - samples/sec: 27.73 - lr: 0.000004
172
+ 2022-10-09 17:17:22,096 epoch 2 - iter 3160/3952 - loss 0.19249352 - samples/sec: 27.64 - lr: 0.000004
173
+ 2022-10-09 17:25:14,518 epoch 2 - iter 3555/3952 - loss 0.19133590 - samples/sec: 27.53 - lr: 0.000005
174
+ 2022-10-09 17:33:11,367 epoch 2 - iter 3950/3952 - loss 0.19002868 - samples/sec: 27.17 - lr: 0.000005
175
+ 2022-10-09 17:33:13,051 ----------------------------------------------------------------------------------------------------
176
+ 2022-10-09 17:33:13,051 EPOCH 2 done: loss 0.1900 - lr 0.000005
177
+ 2022-10-09 17:48:24,882 Evaluating as a multi-label problem: False
178
+ 2022-10-09 17:48:25,337 DEV : loss 0.019461628049612045 - f1-score (micro avg) 0.9258
179
+ 2022-10-09 17:48:56,512 BAD EPOCHS (no improvement): 4
180
+ 2022-10-09 17:48:57,426 ----------------------------------------------------------------------------------------------------
181
+ 2022-10-09 17:57:02,355 epoch 3 - iter 395/3952 - loss 0.17847621 - samples/sec: 26.73 - lr: 0.000005
182
+ 2022-10-09 18:05:03,813 epoch 3 - iter 790/3952 - loss 0.17591453 - samples/sec: 27.01 - lr: 0.000005
183
+ 2022-10-09 18:13:04,328 epoch 3 - iter 1185/3952 - loss 0.17539734 - samples/sec: 27.16 - lr: 0.000005
184
+ 2022-10-09 18:21:04,749 epoch 3 - iter 1580/3952 - loss 0.17471956 - samples/sec: 27.12 - lr: 0.000005
185
+ 2022-10-09 18:29:08,781 epoch 3 - iter 1975/3952 - loss 0.17411210 - samples/sec: 26.88 - lr: 0.000005
186
+ 2022-10-09 18:37:08,694 epoch 3 - iter 2370/3952 - loss 0.17357470 - samples/sec: 27.08 - lr: 0.000005
187
+ 2022-10-09 18:45:04,417 epoch 3 - iter 2765/3952 - loss 0.17312875 - samples/sec: 27.38 - lr: 0.000005
188
+ 2022-10-09 18:53:02,063 epoch 3 - iter 3160/3952 - loss 0.17256571 - samples/sec: 27.23 - lr: 0.000005
189
+ 2022-10-09 19:01:03,502 epoch 3 - iter 3555/3952 - loss 0.17219397 - samples/sec: 26.99 - lr: 0.000005
190
+ 2022-10-09 19:09:01,525 epoch 3 - iter 3950/3952 - loss 0.17175673 - samples/sec: 27.18 - lr: 0.000005
191
+ 2022-10-09 19:09:02,937 ----------------------------------------------------------------------------------------------------
192
+ 2022-10-09 19:09:02,938 EPOCH 3 done: loss 0.1718 - lr 0.000005
193
+ 2022-10-09 19:24:14,202 Evaluating as a multi-label problem: False
194
+ 2022-10-09 19:24:14,636 DEV : loss 0.0132540138438344 - f1-score (micro avg) 0.9449
195
+ 2022-10-09 19:24:46,251 BAD EPOCHS (no improvement): 4
196
+ 2022-10-09 19:24:47,160 ----------------------------------------------------------------------------------------------------
197
+ 2022-10-09 19:32:49,924 epoch 4 - iter 395/3952 - loss 0.16804626 - samples/sec: 27.13 - lr: 0.000005
198
+ 2022-10-09 19:40:52,439 epoch 4 - iter 790/3952 - loss 0.16663174 - samples/sec: 26.93 - lr: 0.000005
199
+ 2022-10-09 19:48:52,963 epoch 4 - iter 1185/3952 - loss 0.16647828 - samples/sec: 26.96 - lr: 0.000005
200
+ 2022-10-09 19:56:51,613 epoch 4 - iter 1580/3952 - loss 0.16639047 - samples/sec: 27.16 - lr: 0.000005
201
+ 2022-10-09 20:04:52,753 epoch 4 - iter 1975/3952 - loss 0.16657475 - samples/sec: 27.08 - lr: 0.000005
202
+ 2022-10-09 20:12:54,756 epoch 4 - iter 2370/3952 - loss 0.16632582 - samples/sec: 27.01 - lr: 0.000005
203
+ 2022-10-09 20:20:58,248 epoch 4 - iter 2765/3952 - loss 0.16578692 - samples/sec: 26.90 - lr: 0.000005
204
+ 2022-10-09 20:28:56,132 epoch 4 - iter 3160/3952 - loss 0.16538230 - samples/sec: 27.31 - lr: 0.000005
205
+ 2022-10-09 20:37:02,675 epoch 4 - iter 3555/3952 - loss 0.16519531 - samples/sec: 26.71 - lr: 0.000004
206
+ 2022-10-09 20:45:04,375 epoch 4 - iter 3950/3952 - loss 0.16516842 - samples/sec: 27.05 - lr: 0.000004
207
+ 2022-10-09 20:45:05,820 ----------------------------------------------------------------------------------------------------
208
+ 2022-10-09 20:45:05,821 EPOCH 4 done: loss 0.1652 - lr 0.000004
209
+ 2022-10-09 21:00:18,495 Evaluating as a multi-label problem: False
210
+ 2022-10-09 21:00:18,914 DEV : loss 0.011177059262990952 - f1-score (micro avg) 0.9535
211
+ 2022-10-09 21:00:51,130 BAD EPOCHS (no improvement): 4
212
+ 2022-10-09 21:00:52,018 ----------------------------------------------------------------------------------------------------
213
+ 2022-10-09 21:08:44,713 epoch 5 - iter 395/3952 - loss 0.16331808 - samples/sec: 27.41 - lr: 0.000004
214
+ 2022-10-09 21:16:38,390 epoch 5 - iter 790/3952 - loss 0.16221079 - samples/sec: 27.45 - lr: 0.000004
215
+ 2022-10-09 21:24:27,598 epoch 5 - iter 1185/3952 - loss 0.16205464 - samples/sec: 27.72 - lr: 0.000004
216
+ 2022-10-09 21:32:21,639 epoch 5 - iter 1580/3952 - loss 0.16189961 - samples/sec: 27.42 - lr: 0.000004
217
+ 2022-10-09 21:40:09,976 epoch 5 - iter 1975/3952 - loss 0.16206946 - samples/sec: 27.79 - lr: 0.000004
218
+ 2022-10-09 21:48:02,577 epoch 5 - iter 2370/3952 - loss 0.16196815 - samples/sec: 27.47 - lr: 0.000004
219
+ 2022-10-09 21:55:56,886 epoch 5 - iter 2765/3952 - loss 0.16172381 - samples/sec: 27.34 - lr: 0.000004
220
+ 2022-10-09 22:03:49,873 epoch 5 - iter 3160/3952 - loss 0.16156487 - samples/sec: 27.47 - lr: 0.000004
221
+ 2022-10-09 22:11:41,572 epoch 5 - iter 3555/3952 - loss 0.16147326 - samples/sec: 27.64 - lr: 0.000004
222
+ 2022-10-09 22:19:35,682 epoch 5 - iter 3950/3952 - loss 0.16130607 - samples/sec: 27.55 - lr: 0.000004
223
+ 2022-10-09 22:19:37,189 ----------------------------------------------------------------------------------------------------
224
+ 2022-10-09 22:19:37,189 EPOCH 5 done: loss 0.1613 - lr 0.000004
225
+ 2022-10-09 22:34:40,766 Evaluating as a multi-label problem: False
226
+ 2022-10-09 22:34:41,185 DEV : loss 0.011488113552331924 - f1-score (micro avg) 0.9546
227
+ 2022-10-09 22:35:13,419 BAD EPOCHS (no improvement): 4
228
+ 2022-10-09 22:35:14,308 ----------------------------------------------------------------------------------------------------
229
+ 2022-10-09 22:43:06,921 epoch 6 - iter 395/3952 - loss 0.16063155 - samples/sec: 27.36 - lr: 0.000004
230
+ 2022-10-09 22:50:59,349 epoch 6 - iter 790/3952 - loss 0.15984298 - samples/sec: 27.55 - lr: 0.000004
231
+ 2022-10-09 22:58:46,877 epoch 6 - iter 1185/3952 - loss 0.15946325 - samples/sec: 27.80 - lr: 0.000004
232
+ 2022-10-09 23:06:50,203 epoch 6 - iter 1580/3952 - loss 0.15917470 - samples/sec: 26.82 - lr: 0.000004
233
+ 2022-10-09 23:14:51,992 epoch 6 - iter 1975/3952 - loss 0.15882030 - samples/sec: 27.09 - lr: 0.000004
234
+ 2022-10-09 23:22:52,029 epoch 6 - iter 2370/3952 - loss 0.15876178 - samples/sec: 27.05 - lr: 0.000004
235
+ 2022-10-09 23:30:58,678 epoch 6 - iter 2765/3952 - loss 0.15864742 - samples/sec: 26.76 - lr: 0.000004
236
+ 2022-10-09 23:38:57,773 epoch 6 - iter 3160/3952 - loss 0.15842630 - samples/sec: 27.18 - lr: 0.000004
237
+ 2022-10-09 23:46:58,724 epoch 6 - iter 3555/3952 - loss 0.15814869 - samples/sec: 27.02 - lr: 0.000004
238
+ 2022-10-09 23:55:00,979 epoch 6 - iter 3950/3952 - loss 0.15800367 - samples/sec: 27.09 - lr: 0.000004
239
+ 2022-10-09 23:55:02,321 ----------------------------------------------------------------------------------------------------
240
+ 2022-10-09 23:55:02,321 EPOCH 6 done: loss 0.1580 - lr 0.000004
241
+ 2022-10-10 00:10:14,450 Evaluating as a multi-label problem: False
242
+ 2022-10-10 00:10:14,910 DEV : loss 0.010615515522658825 - f1-score (micro avg) 0.9587
243
+ 2022-10-10 00:10:46,276 BAD EPOCHS (no improvement): 4
244
+ 2022-10-10 00:10:47,232 ----------------------------------------------------------------------------------------------------
245
+ 2022-10-10 00:18:56,370 epoch 7 - iter 395/3952 - loss 0.15533572 - samples/sec: 26.74 - lr: 0.000004
246
+ 2022-10-10 00:26:53,533 epoch 7 - iter 790/3952 - loss 0.15567018 - samples/sec: 27.29 - lr: 0.000004
247
+ 2022-10-10 00:34:55,929 epoch 7 - iter 1185/3952 - loss 0.15559902 - samples/sec: 26.92 - lr: 0.000004
248
+ 2022-10-10 00:42:56,064 epoch 7 - iter 1580/3952 - loss 0.15526644 - samples/sec: 27.04 - lr: 0.000004
249
+ 2022-10-10 00:50:56,575 epoch 7 - iter 1975/3952 - loss 0.15532544 - samples/sec: 27.04 - lr: 0.000004
250
+ 2022-10-10 00:58:55,726 epoch 7 - iter 2370/3952 - loss 0.15538178 - samples/sec: 27.14 - lr: 0.000004
251
+ 2022-10-10 01:06:54,255 epoch 7 - iter 2765/3952 - loss 0.15537470 - samples/sec: 27.15 - lr: 0.000004
252
+ 2022-10-10 01:14:59,643 epoch 7 - iter 3160/3952 - loss 0.15531628 - samples/sec: 26.79 - lr: 0.000004
253
+ 2022-10-10 01:23:03,037 epoch 7 - iter 3555/3952 - loss 0.15533451 - samples/sec: 26.86 - lr: 0.000004
254
+ 2022-10-10 01:31:03,511 epoch 7 - iter 3950/3952 - loss 0.15511299 - samples/sec: 26.97 - lr: 0.000004
255
+ 2022-10-10 01:31:05,040 ----------------------------------------------------------------------------------------------------
256
+ 2022-10-10 01:31:05,041 EPOCH 7 done: loss 0.1551 - lr 0.000004
257
+ 2022-10-10 01:46:17,630 Evaluating as a multi-label problem: False
258
+ 2022-10-10 01:46:18,057 DEV : loss 0.010866315104067326 - f1-score (micro avg) 0.9597
259
+ 2022-10-10 01:46:49,834 BAD EPOCHS (no improvement): 4
260
+ 2022-10-10 01:46:50,741 ----------------------------------------------------------------------------------------------------
261
+ 2022-10-10 01:54:49,387 epoch 8 - iter 395/3952 - loss 0.15339956 - samples/sec: 27.34 - lr: 0.000004
262
+ 2022-10-10 02:02:54,436 epoch 8 - iter 790/3952 - loss 0.15357118 - samples/sec: 26.88 - lr: 0.000004
263
+ 2022-10-10 02:10:57,380 epoch 8 - iter 1185/3952 - loss 0.15383618 - samples/sec: 26.86 - lr: 0.000004
264
+ 2022-10-10 02:18:57,413 epoch 8 - iter 1580/3952 - loss 0.15388423 - samples/sec: 27.15 - lr: 0.000004
265
+ 2022-10-10 02:26:58,665 epoch 8 - iter 1975/3952 - loss 0.15366022 - samples/sec: 26.96 - lr: 0.000003
266
+ 2022-10-10 02:35:00,936 epoch 8 - iter 2370/3952 - loss 0.15388824 - samples/sec: 26.92 - lr: 0.000003
267
+ 2022-10-10 02:43:03,179 epoch 8 - iter 2765/3952 - loss 0.15380049 - samples/sec: 27.06 - lr: 0.000003
268
+ 2022-10-10 02:51:07,445 epoch 8 - iter 3160/3952 - loss 0.15356183 - samples/sec: 26.93 - lr: 0.000003
269
+ 2022-10-10 02:59:09,568 epoch 8 - iter 3555/3952 - loss 0.15337591 - samples/sec: 26.91 - lr: 0.000003
270
+ 2022-10-10 03:07:06,249 epoch 8 - iter 3950/3952 - loss 0.15327199 - samples/sec: 27.26 - lr: 0.000003
271
+ 2022-10-10 03:07:07,508 ----------------------------------------------------------------------------------------------------
272
+ 2022-10-10 03:07:07,509 EPOCH 8 done: loss 0.1533 - lr 0.000003
273
+ 2022-10-10 03:22:20,421 Evaluating as a multi-label problem: False
274
+ 2022-10-10 03:22:20,849 DEV : loss 0.010451321490108967 - f1-score (micro avg) 0.9617
275
+ 2022-10-10 03:22:53,399 BAD EPOCHS (no improvement): 4
276
+ 2022-10-10 03:22:55,354 ----------------------------------------------------------------------------------------------------
277
+ 2022-10-10 03:31:03,911 epoch 9 - iter 395/3952 - loss 0.15095455 - samples/sec: 26.52 - lr: 0.000003
278
+ 2022-10-10 03:39:03,919 epoch 9 - iter 790/3952 - loss 0.15100488 - samples/sec: 27.07 - lr: 0.000003
279
+ 2022-10-10 03:46:57,642 epoch 9 - iter 1185/3952 - loss 0.15141407 - samples/sec: 27.49 - lr: 0.000003
280
+ 2022-10-10 03:54:55,677 epoch 9 - iter 1580/3952 - loss 0.15153248 - samples/sec: 27.33 - lr: 0.000003
281
+ 2022-10-10 04:02:55,192 epoch 9 - iter 1975/3952 - loss 0.15137991 - samples/sec: 27.30 - lr: 0.000003
282
+ 2022-10-10 04:10:56,499 epoch 9 - iter 2370/3952 - loss 0.15134929 - samples/sec: 27.05 - lr: 0.000003
283
+ 2022-10-10 04:18:51,998 epoch 9 - iter 2765/3952 - loss 0.15139573 - samples/sec: 27.48 - lr: 0.000003
284
+ 2022-10-10 04:26:48,529 epoch 9 - iter 3160/3952 - loss 0.15141239 - samples/sec: 27.31 - lr: 0.000003
285
+ 2022-10-10 04:34:41,608 epoch 9 - iter 3555/3952 - loss 0.15135720 - samples/sec: 27.53 - lr: 0.000003
286
+ 2022-10-10 04:42:37,267 epoch 9 - iter 3950/3952 - loss 0.15132694 - samples/sec: 27.23 - lr: 0.000003
287
+ 2022-10-10 04:42:38,593 ----------------------------------------------------------------------------------------------------
288
+ 2022-10-10 04:42:38,594 EPOCH 9 done: loss 0.1513 - lr 0.000003
289
+ 2022-10-10 04:57:46,312 Evaluating as a multi-label problem: False
290
+ 2022-10-10 04:57:46,749 DEV : loss 0.01064694207161665 - f1-score (micro avg) 0.9609
291
+ 2022-10-10 04:58:17,984 BAD EPOCHS (no improvement): 4
292
+ 2022-10-10 04:58:18,878 ----------------------------------------------------------------------------------------------------
293
+ 2022-10-10 05:06:19,341 epoch 10 - iter 395/3952 - loss 0.14934098 - samples/sec: 27.00 - lr: 0.000003
294
+ 2022-10-10 05:14:13,052 epoch 10 - iter 790/3952 - loss 0.15047359 - samples/sec: 27.54 - lr: 0.000003
295
+ 2022-10-10 05:22:09,904 epoch 10 - iter 1185/3952 - loss 0.15005411 - samples/sec: 27.27 - lr: 0.000003
296
+ 2022-10-10 05:30:10,047 epoch 10 - iter 1580/3952 - loss 0.14970562 - samples/sec: 27.14 - lr: 0.000003
297
+ 2022-10-10 05:38:05,869 epoch 10 - iter 1975/3952 - loss 0.14954158 - samples/sec: 27.29 - lr: 0.000003
298
+ 2022-10-10 05:46:05,902 epoch 10 - iter 2370/3952 - loss 0.14932048 - samples/sec: 27.11 - lr: 0.000003
299
+ 2022-10-10 05:54:05,041 epoch 10 - iter 2765/3952 - loss 0.14927630 - samples/sec: 27.19 - lr: 0.000003
300
+ 2022-10-10 06:02:04,693 epoch 10 - iter 3160/3952 - loss 0.14935304 - samples/sec: 27.16 - lr: 0.000003
301
+ 2022-10-10 06:10:01,212 epoch 10 - iter 3555/3952 - loss 0.14941757 - samples/sec: 27.25 - lr: 0.000003
302
+ 2022-10-10 06:17:54,179 epoch 10 - iter 3950/3952 - loss 0.14953843 - samples/sec: 27.53 - lr: 0.000003
303
+ 2022-10-10 06:17:55,747 ----------------------------------------------------------------------------------------------------
304
+ 2022-10-10 06:17:55,747 EPOCH 10 done: loss 0.1495 - lr 0.000003
305
+ 2022-10-10 06:33:03,662 Evaluating as a multi-label problem: False
306
+ 2022-10-10 06:33:04,089 DEV : loss 0.010687584988772869 - f1-score (micro avg) 0.9623
307
+ 2022-10-10 06:33:35,248 BAD EPOCHS (no improvement): 4
308
+ 2022-10-10 06:33:36,135 ----------------------------------------------------------------------------------------------------
309
+ 2022-10-10 06:41:36,387 epoch 11 - iter 395/3952 - loss 0.14722548 - samples/sec: 27.24 - lr: 0.000003
310
+ 2022-10-10 06:49:32,701 epoch 11 - iter 790/3952 - loss 0.14792717 - samples/sec: 27.36 - lr: 0.000003
311
+ 2022-10-10 06:57:28,372 epoch 11 - iter 1185/3952 - loss 0.14804400 - samples/sec: 27.34 - lr: 0.000003
312
+ 2022-10-10 07:05:28,768 epoch 11 - iter 1580/3952 - loss 0.14822560 - samples/sec: 27.11 - lr: 0.000003
313
+ 2022-10-10 07:13:27,055 epoch 11 - iter 1975/3952 - loss 0.14845261 - samples/sec: 27.25 - lr: 0.000003
314
+ 2022-10-10 07:21:21,803 epoch 11 - iter 2370/3952 - loss 0.14860234 - samples/sec: 27.39 - lr: 0.000003
315
+ 2022-10-10 07:29:18,530 epoch 11 - iter 2765/3952 - loss 0.14881168 - samples/sec: 27.27 - lr: 0.000003
316
+ 2022-10-10 07:37:14,641 epoch 11 - iter 3160/3952 - loss 0.14859987 - samples/sec: 27.27 - lr: 0.000003
317
+ 2022-10-10 07:45:11,011 epoch 11 - iter 3555/3952 - loss 0.14841785 - samples/sec: 27.30 - lr: 0.000003
318
+ 2022-10-10 07:53:06,062 epoch 11 - iter 3950/3952 - loss 0.14839159 - samples/sec: 27.46 - lr: 0.000003
319
+ 2022-10-10 07:53:07,694 ----------------------------------------------------------------------------------------------------
320
+ 2022-10-10 07:53:07,694 EPOCH 11 done: loss 0.1484 - lr 0.000003
321
+ 2022-10-10 08:08:15,642 Evaluating as a multi-label problem: False
322
+ 2022-10-10 08:08:16,078 DEV : loss 0.010935463011264801 - f1-score (micro avg) 0.962
323
+ 2022-10-10 08:08:47,374 BAD EPOCHS (no improvement): 4
324
+ 2022-10-10 08:08:48,267 ----------------------------------------------------------------------------------------------------
325
+ 2022-10-10 08:16:43,768 epoch 12 - iter 395/3952 - loss 0.14779592 - samples/sec: 27.35 - lr: 0.000002
326
+ 2022-10-10 08:24:44,096 epoch 12 - iter 790/3952 - loss 0.14727136 - samples/sec: 27.09 - lr: 0.000002
327
+ 2022-10-10 08:32:40,480 epoch 12 - iter 1185/3952 - loss 0.14742119 - samples/sec: 27.30 - lr: 0.000002
328
+ 2022-10-10 08:40:34,998 epoch 12 - iter 1580/3952 - loss 0.14735918 - samples/sec: 27.41 - lr: 0.000002
329
+ 2022-10-10 08:48:29,447 epoch 12 - iter 1975/3952 - loss 0.14739904 - samples/sec: 27.37 - lr: 0.000002
330
+ 2022-10-10 08:56:21,930 epoch 12 - iter 2370/3952 - loss 0.14746441 - samples/sec: 27.64 - lr: 0.000002
331
+ 2022-10-10 09:04:20,566 epoch 12 - iter 2765/3952 - loss 0.14727131 - samples/sec: 27.27 - lr: 0.000002
332
+ 2022-10-10 09:12:17,286 epoch 12 - iter 3160/3952 - loss 0.14733990 - samples/sec: 27.30 - lr: 0.000002
333
+ 2022-10-10 09:20:14,749 epoch 12 - iter 3555/3952 - loss 0.14706041 - samples/sec: 27.28 - lr: 0.000002
334
+ 2022-10-10 09:28:08,079 epoch 12 - iter 3950/3952 - loss 0.14700556 - samples/sec: 27.52 - lr: 0.000002
335
+ 2022-10-10 09:28:09,718 ----------------------------------------------------------------------------------------------------
336
+ 2022-10-10 09:28:09,718 EPOCH 12 done: loss 0.1470 - lr 0.000002
337
+ 2022-10-10 09:43:14,910 Evaluating as a multi-label problem: False
338
+ 2022-10-10 09:43:15,334 DEV : loss 0.011056484654545784 - f1-score (micro avg) 0.9634
339
+ 2022-10-10 09:43:47,802 BAD EPOCHS (no improvement): 4
340
+ 2022-10-10 09:43:48,705 ----------------------------------------------------------------------------------------------------
341
+ 2022-10-10 09:51:46,179 epoch 13 - iter 395/3952 - loss 0.14506338 - samples/sec: 27.10 - lr: 0.000002
342
+ 2022-10-10 09:59:43,717 epoch 13 - iter 790/3952 - loss 0.14619048 - samples/sec: 27.23 - lr: 0.000002
343
+ 2022-10-10 10:07:33,958 epoch 13 - iter 1185/3952 - loss 0.14639748 - samples/sec: 27.70 - lr: 0.000002
344
+ 2022-10-10 10:15:01,211 epoch 13 - iter 1580/3952 - loss 0.14615405 - samples/sec: 29.14 - lr: 0.000002
345
+ 2022-10-10 10:22:17,577 epoch 13 - iter 1975/3952 - loss 0.14620482 - samples/sec: 29.92 - lr: 0.000002
346
+ 2022-10-10 10:29:43,376 epoch 13 - iter 2370/3952 - loss 0.14616699 - samples/sec: 29.29 - lr: 0.000002
347
+ 2022-10-10 10:37:06,729 epoch 13 - iter 2765/3952 - loss 0.14605036 - samples/sec: 29.39 - lr: 0.000002
348
+ 2022-10-10 10:44:37,315 epoch 13 - iter 3160/3952 - loss 0.14597794 - samples/sec: 28.97 - lr: 0.000002
349
+ 2022-10-10 10:52:01,383 epoch 13 - iter 3555/3952 - loss 0.14602289 - samples/sec: 29.36 - lr: 0.000002
350
+ 2022-10-10 10:59:26,413 epoch 13 - iter 3950/3952 - loss 0.14605007 - samples/sec: 29.23 - lr: 0.000002
351
+ 2022-10-10 10:59:27,781 ----------------------------------------------------------------------------------------------------
352
+ 2022-10-10 10:59:27,782 EPOCH 13 done: loss 0.1460 - lr 0.000002
353
+ 2022-10-10 11:14:26,437 Evaluating as a multi-label problem: False
354
+ 2022-10-10 11:14:26,846 DEV : loss 0.011409671977162361 - f1-score (micro avg) 0.9628
355
+ 2022-10-10 11:14:57,380 BAD EPOCHS (no improvement): 4
356
+ 2022-10-10 11:14:58,218 ----------------------------------------------------------------------------------------------------
357
+ 2022-10-10 11:22:28,388 epoch 14 - iter 395/3952 - loss 0.14532304 - samples/sec: 29.11 - lr: 0.000002
358
+ 2022-10-10 11:29:54,824 epoch 14 - iter 790/3952 - loss 0.14560920 - samples/sec: 29.11 - lr: 0.000002
359
+ 2022-10-10 11:37:22,235 epoch 14 - iter 1185/3952 - loss 0.14518057 - samples/sec: 29.05 - lr: 0.000002
360
+ 2022-10-10 11:44:50,891 epoch 14 - iter 1580/3952 - loss 0.14527092 - samples/sec: 28.98 - lr: 0.000002
361
+ 2022-10-10 11:52:18,549 epoch 14 - iter 1975/3952 - loss 0.14511930 - samples/sec: 29.20 - lr: 0.000002
362
+ 2022-10-10 11:59:56,465 epoch 14 - iter 2370/3952 - loss 0.14523496 - samples/sec: 28.44 - lr: 0.000002
363
+ 2022-10-10 12:07:18,925 epoch 14 - iter 2765/3952 - loss 0.14524068 - samples/sec: 29.46 - lr: 0.000002
364
+ 2022-10-10 12:14:42,038 epoch 14 - iter 3160/3952 - loss 0.14516594 - samples/sec: 29.36 - lr: 0.000002
365
+ 2022-10-10 12:22:07,540 epoch 14 - iter 3555/3952 - loss 0.14526955 - samples/sec: 29.18 - lr: 0.000002
366
+ 2022-10-10 12:29:36,124 epoch 14 - iter 3950/3952 - loss 0.14518783 - samples/sec: 29.17 - lr: 0.000002
367
+ 2022-10-10 12:29:37,533 ----------------------------------------------------------------------------------------------------
368
+ 2022-10-10 12:29:37,533 EPOCH 14 done: loss 0.1452 - lr 0.000002
369
+ 2022-10-10 12:44:22,577 Evaluating as a multi-label problem: False
370
+ 2022-10-10 12:44:22,990 DEV : loss 0.011419754475355148 - f1-score (micro avg) 0.9637
371
+ 2022-10-10 12:44:53,663 BAD EPOCHS (no improvement): 4
372
+ 2022-10-10 12:44:54,557 ----------------------------------------------------------------------------------------------------
373
+ 2022-10-10 12:52:25,951 epoch 15 - iter 395/3952 - loss 0.14181725 - samples/sec: 28.90 - lr: 0.000002
374
+ 2022-10-10 12:59:53,144 epoch 15 - iter 790/3952 - loss 0.14383060 - samples/sec: 29.20 - lr: 0.000002
375
+ 2022-10-10 13:08:18,071 epoch 15 - iter 1185/3952 - loss 0.14395256 - samples/sec: 25.82 - lr: 0.000002
376
+ 2022-10-10 13:16:43,174 epoch 15 - iter 1580/3952 - loss 0.14433998 - samples/sec: 25.78 - lr: 0.000002
377
+ 2022-10-10 13:25:14,818 epoch 15 - iter 1975/3952 - loss 0.14428390 - samples/sec: 25.37 - lr: 0.000002
378
+ 2022-10-10 13:33:46,506 epoch 15 - iter 2370/3952 - loss 0.14440542 - samples/sec: 25.41 - lr: 0.000002
379
+ 2022-10-10 13:42:09,041 epoch 15 - iter 2765/3952 - loss 0.14445593 - samples/sec: 26.01 - lr: 0.000001
380
+ 2022-10-10 13:50:39,620 epoch 15 - iter 3160/3952 - loss 0.14456461 - samples/sec: 25.44 - lr: 0.000001
381
+ 2022-10-10 13:59:09,404 epoch 15 - iter 3555/3952 - loss 0.14444586 - samples/sec: 25.61 - lr: 0.000001
382
+ 2022-10-10 14:07:41,706 epoch 15 - iter 3950/3952 - loss 0.14432217 - samples/sec: 25.45 - lr: 0.000001
383
+ 2022-10-10 14:07:43,149 ----------------------------------------------------------------------------------------------------
384
+ 2022-10-10 14:07:43,150 EPOCH 15 done: loss 0.1443 - lr 0.000001
385
+ 2022-10-10 14:23:33,181 Evaluating as a multi-label problem: False
386
+ 2022-10-10 14:23:33,654 DEV : loss 0.011627680622041225 - f1-score (micro avg) 0.9637
387
+ 2022-10-10 14:24:07,996 BAD EPOCHS (no improvement): 4
388
+ 2022-10-10 14:24:09,032 ----------------------------------------------------------------------------------------------------
389
+ 2022-10-10 14:32:40,414 epoch 16 - iter 395/3952 - loss 0.14350737 - samples/sec: 25.61 - lr: 0.000001
390
+ 2022-10-10 14:41:10,956 epoch 16 - iter 790/3952 - loss 0.14341419 - samples/sec: 25.59 - lr: 0.000001
391
+ 2022-10-10 14:49:40,914 epoch 16 - iter 1185/3952 - loss 0.14370127 - samples/sec: 25.52 - lr: 0.000001
392
+ 2022-10-10 14:58:09,406 epoch 16 - iter 1580/3952 - loss 0.14378459 - samples/sec: 25.57 - lr: 0.000001
393
+ 2022-10-10 15:06:40,193 epoch 16 - iter 1975/3952 - loss 0.14360404 - samples/sec: 25.52 - lr: 0.000001
394
+ 2022-10-10 15:15:11,603 epoch 16 - iter 2370/3952 - loss 0.14360062 - samples/sec: 25.44 - lr: 0.000001
395
+ 2022-10-10 15:23:44,499 epoch 16 - iter 2765/3952 - loss 0.14356139 - samples/sec: 25.37 - lr: 0.000001
396
+ 2022-10-10 15:32:14,460 epoch 16 - iter 3160/3952 - loss 0.14361871 - samples/sec: 25.48 - lr: 0.000001
397
+ 2022-10-10 15:40:46,346 epoch 16 - iter 3555/3952 - loss 0.14360176 - samples/sec: 25.51 - lr: 0.000001
398
+ 2022-10-10 15:49:16,072 epoch 16 - iter 3950/3952 - loss 0.14352181 - samples/sec: 25.55 - lr: 0.000001
399
+ 2022-10-10 15:49:18,082 ----------------------------------------------------------------------------------------------------
400
+ 2022-10-10 15:49:18,082 EPOCH 16 done: loss 0.1435 - lr 0.000001
401
+ 2022-10-10 16:05:01,512 Evaluating as a multi-label problem: False
402
+ 2022-10-10 16:05:01,984 DEV : loss 0.011783876456320286 - f1-score (micro avg) 0.9644
403
+ 2022-10-10 16:05:36,459 BAD EPOCHS (no improvement): 4
404
+ 2022-10-10 16:05:37,421 ----------------------------------------------------------------------------------------------------
405
+ 2022-10-10 16:14:08,530 epoch 17 - iter 395/3952 - loss 0.14367645 - samples/sec: 25.33 - lr: 0.000001
406
+ 2022-10-10 16:22:34,521 epoch 17 - iter 790/3952 - loss 0.14312751 - samples/sec: 25.71 - lr: 0.000001
407
+ 2022-10-10 16:31:01,690 epoch 17 - iter 1185/3952 - loss 0.14363484 - samples/sec: 25.68 - lr: 0.000001
408
+ 2022-10-10 16:39:26,318 epoch 17 - iter 1580/3952 - loss 0.14329122 - samples/sec: 25.77 - lr: 0.000001
409
+ 2022-10-10 16:47:51,245 epoch 17 - iter 1975/3952 - loss 0.14338973 - samples/sec: 25.84 - lr: 0.000001
410
+ 2022-10-10 16:56:18,671 epoch 17 - iter 2370/3952 - loss 0.14364105 - samples/sec: 25.62 - lr: 0.000001
411
+ 2022-10-10 17:04:48,817 epoch 17 - iter 2765/3952 - loss 0.14374600 - samples/sec: 25.48 - lr: 0.000001
412
+ 2022-10-10 17:13:21,802 epoch 17 - iter 3160/3952 - loss 0.14369645 - samples/sec: 25.31 - lr: 0.000001
413
+ 2022-10-10 17:21:51,309 epoch 17 - iter 3555/3952 - loss 0.14360598 - samples/sec: 25.59 - lr: 0.000001
414
+ 2022-10-10 17:30:20,509 epoch 17 - iter 3950/3952 - loss 0.14356029 - samples/sec: 25.54 - lr: 0.000001
415
+ 2022-10-10 17:30:22,113 ----------------------------------------------------------------------------------------------------
416
+ 2022-10-10 17:30:22,114 EPOCH 17 done: loss 0.1436 - lr 0.000001
417
+ 2022-10-10 17:46:12,566 Evaluating as a multi-label problem: False
418
+ 2022-10-10 17:46:13,046 DEV : loss 0.011797642335295677 - f1-score (micro avg) 0.9643
419
+ 2022-10-10 17:46:47,683 BAD EPOCHS (no improvement): 4
420
+ 2022-10-10 17:46:48,723 ----------------------------------------------------------------------------------------------------
421
+ 2022-10-10 17:55:28,142 epoch 18 - iter 395/3952 - loss 0.14306617 - samples/sec: 25.20 - lr: 0.000001
422
+ 2022-10-10 18:03:57,902 epoch 18 - iter 790/3952 - loss 0.14196615 - samples/sec: 25.53 - lr: 0.000001
423
+ 2022-10-10 18:12:31,453 epoch 18 - iter 1185/3952 - loss 0.14182625 - samples/sec: 25.38 - lr: 0.000001
424
+ 2022-10-10 18:20:57,991 epoch 18 - iter 1580/3952 - loss 0.14185926 - samples/sec: 25.62 - lr: 0.000001
425
+ 2022-10-10 18:29:28,131 epoch 18 - iter 1975/3952 - loss 0.14207068 - samples/sec: 25.46 - lr: 0.000001
426
+ 2022-10-10 18:37:54,888 epoch 18 - iter 2370/3952 - loss 0.14229279 - samples/sec: 25.71 - lr: 0.000001
427
+ 2022-10-10 18:46:22,698 epoch 18 - iter 2765/3952 - loss 0.14234187 - samples/sec: 25.65 - lr: 0.000001
428
+ 2022-10-10 18:54:50,839 epoch 18 - iter 3160/3952 - loss 0.14240556 - samples/sec: 25.65 - lr: 0.000001
429
+ 2022-10-10 19:03:22,482 epoch 18 - iter 3555/3952 - loss 0.14233153 - samples/sec: 25.48 - lr: 0.000001
430
+ 2022-10-10 19:11:53,854 epoch 18 - iter 3950/3952 - loss 0.14236278 - samples/sec: 25.30 - lr: 0.000001
431
+ 2022-10-10 19:11:56,073 ----------------------------------------------------------------------------------------------------
432
+ 2022-10-10 19:11:56,074 EPOCH 18 done: loss 0.1424 - lr 0.000001
433
+ 2022-10-10 19:27:45,449 Evaluating as a multi-label problem: False
434
+ 2022-10-10 19:27:45,930 DEV : loss 0.011939478106796741 - f1-score (micro avg) 0.964
435
+ 2022-10-10 19:28:18,875 BAD EPOCHS (no improvement): 4
436
+ 2022-10-10 19:28:19,941 ----------------------------------------------------------------------------------------------------
437
+ 2022-10-10 19:36:53,864 epoch 19 - iter 395/3952 - loss 0.14362086 - samples/sec: 25.29 - lr: 0.000001
438
+ 2022-10-10 19:45:24,479 epoch 19 - iter 790/3952 - loss 0.14325958 - samples/sec: 25.49 - lr: 0.000001
439
+ 2022-10-10 19:53:54,808 epoch 19 - iter 1185/3952 - loss 0.14310735 - samples/sec: 25.48 - lr: 0.000000
440
+ 2022-10-10 20:02:24,384 epoch 19 - iter 1580/3952 - loss 0.14293734 - samples/sec: 25.47 - lr: 0.000000
441
+ 2022-10-10 20:10:51,221 epoch 19 - iter 1975/3952 - loss 0.14306481 - samples/sec: 25.77 - lr: 0.000000
442
+ 2022-10-10 20:19:18,624 epoch 19 - iter 2370/3952 - loss 0.14291352 - samples/sec: 25.72 - lr: 0.000000
443
+ 2022-10-10 20:27:46,259 epoch 19 - iter 2765/3952 - loss 0.14298740 - samples/sec: 25.60 - lr: 0.000000
444
+ 2022-10-10 20:36:16,560 epoch 19 - iter 3160/3952 - loss 0.14288623 - samples/sec: 25.52 - lr: 0.000000
445
+ 2022-10-10 20:44:47,260 epoch 19 - iter 3555/3952 - loss 0.14282900 - samples/sec: 25.45 - lr: 0.000000
446
+ 2022-10-10 20:53:18,466 epoch 19 - iter 3950/3952 - loss 0.14288617 - samples/sec: 25.54 - lr: 0.000000
447
+ 2022-10-10 20:53:19,964 ----------------------------------------------------------------------------------------------------
448
+ 2022-10-10 20:53:19,964 EPOCH 19 done: loss 0.1429 - lr 0.000000
449
+ 2022-10-10 21:09:08,715 Evaluating as a multi-label problem: False
450
+ 2022-10-10 21:09:09,202 DEV : loss 0.012016847729682922 - f1-score (micro avg) 0.9643
451
+ 2022-10-10 21:09:43,778 BAD EPOCHS (no improvement): 4
452
+ 2022-10-10 21:09:44,810 ----------------------------------------------------------------------------------------------------
453
+ 2022-10-10 21:18:11,781 epoch 20 - iter 395/3952 - loss 0.14263110 - samples/sec: 25.65 - lr: 0.000000
454
+ 2022-10-10 21:26:40,891 epoch 20 - iter 790/3952 - loss 0.14225428 - samples/sec: 25.60 - lr: 0.000000
455
+ 2022-10-10 21:35:08,495 epoch 20 - iter 1185/3952 - loss 0.14205051 - samples/sec: 25.66 - lr: 0.000000
456
+ 2022-10-10 21:43:34,108 epoch 20 - iter 1580/3952 - loss 0.14228947 - samples/sec: 25.71 - lr: 0.000000
457
+ 2022-10-10 21:52:11,211 epoch 20 - iter 1975/3952 - loss 0.14209594 - samples/sec: 25.19 - lr: 0.000000
458
+ 2022-10-10 22:00:41,644 epoch 20 - iter 2370/3952 - loss 0.14227931 - samples/sec: 25.63 - lr: 0.000000
459
+ 2022-10-10 22:09:10,266 epoch 20 - iter 2765/3952 - loss 0.14254834 - samples/sec: 25.65 - lr: 0.000000
460
+ 2022-10-10 22:17:38,261 epoch 20 - iter 3160/3952 - loss 0.14259954 - samples/sec: 25.71 - lr: 0.000000
461
+ 2022-10-10 22:26:05,321 epoch 20 - iter 3555/3952 - loss 0.14252244 - samples/sec: 25.59 - lr: 0.000000
462
+ 2022-10-10 22:34:35,781 epoch 20 - iter 3950/3952 - loss 0.14238758 - samples/sec: 25.47 - lr: 0.000000
463
+ 2022-10-10 22:34:37,421 ----------------------------------------------------------------------------------------------------
464
+ 2022-10-10 22:34:37,422 EPOCH 20 done: loss 0.1424 - lr 0.000000
465
+ 2022-10-10 22:50:27,724 Evaluating as a multi-label problem: False
466
+ 2022-10-10 22:50:28,207 DEV : loss 0.012119622901082039 - f1-score (micro avg) 0.964
467
+ 2022-10-10 22:51:01,203 BAD EPOCHS (no improvement): 4
468
+ 2022-10-10 22:51:03,269 ----------------------------------------------------------------------------------------------------
469
+ 2022-10-10 22:51:03,271 Testing using last state of model ...
470
+ 2022-10-10 22:59:53,131 Evaluating as a multi-label problem: False
471
+ 2022-10-10 22:59:53,392 0.945 0.9596 0.9522 0.9179
472
+ 2022-10-10 22:59:53,392
473
+ Results:
474
+ - F-score (micro) 0.9522
475
+ - F-score (macro) 0.9468
476
+ - Accuracy 0.9179
477
+
478
+ By class:
479
+ precision recall f1-score support
480
+
481
+ LOC 0.9643 0.9671 0.9657 11823
482
+ PER 0.9722 0.9736 0.9729 7836
483
+ DATE_TIME 0.9152 0.9458 0.9303 4746
484
+ ORG 0.8720 0.9196 0.8952 4565
485
+ NRP 0.9633 0.9766 0.9699 2905
486
+
487
+ micro avg 0.9450 0.9596 0.9522 31875
488
+ macro avg 0.9374 0.9565 0.9468 31875
489
+ weighted avg 0.9456 0.9596 0.9525 31875
490
+
491
+ 2022-10-10 22:59:53,392 ----------------------------------------------------------------------------------------------------
weights.txt ADDED
File without changes