Hugging Face's logo

stefan-it
/
hmbench-newseye-fr-hmteams-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1

Token Classification
Flair
PyTorch
TensorBoard
French
sequence-tagger-model
Model card Files Metrics Community
stefan-it commited on
Commit
7ae32e2
·
1 Parent(s): a912e3d

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.1697546319.0468bd9609d6.7281.2 +3 -0
  5. test.tsv +0 -0
  6. training.log +241 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c9f22bb3e77a4c3a4dcb02906029ff09c7541e8aed55ff9230426e1904b4d76
3
+ size 440954373
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 12:40:10 0.0000 0.4994 0.1228 0.7043 0.7714 0.7364 0.6000
3
+ 2 12:41:44 0.0000 0.1162 0.1234 0.7516 0.8190 0.7839 0.6623
4
+ 3 12:43:17 0.0000 0.0825 0.1487 0.8230 0.7973 0.8100 0.6968
5
+ 4 12:44:51 0.0000 0.0557 0.1733 0.7710 0.8245 0.7968 0.6756
6
+ 5 12:46:24 0.0000 0.0440 0.1758 0.8031 0.8327 0.8176 0.7083
7
+ 6 12:47:59 0.0000 0.0328 0.1848 0.8117 0.8272 0.8194 0.7161
8
+ 7 12:49:32 0.0000 0.0234 0.1932 0.8214 0.8259 0.8236 0.7166
9
+ 8 12:51:05 0.0000 0.0169 0.1982 0.8016 0.8408 0.8207 0.7169
10
+ 9 12:52:37 0.0000 0.0118 0.2068 0.8074 0.8272 0.8172 0.7078
11
+ 10 12:54:10 0.0000 0.0085 0.2104 0.8233 0.8367 0.8300 0.7252
runs/events.out.tfevents.1697546319.0468bd9609d6.7281.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be7b4f7662068f5909cf87c734d9fc671384abf1d442a5b3c6d609b13e53316e
3
+ size 999862
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,241 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-17 12:38:39,938 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-17 12:38:39,939 Model: "SequenceTagger(
3
+ (embeddings): TransformerWordEmbeddings(
4
+ (model): ElectraModel(
5
+ (embeddings): ElectraEmbeddings(
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): ElectraEncoder(
13
+ (layer): ModuleList(
14
+ (0-11): 12 x ElectraLayer(
15
+ (attention): ElectraAttention(
16
+ (self): ElectraSelfAttention(
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): ElectraSelfOutput(
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): ElectraIntermediate(
29
+ (dense): Linear(in_features=768, out_features=3072, bias=True)
30
+ (intermediate_act_fn): GELUActivation()
31
+ )
32
+ (output): ElectraOutput(
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
+ )
41
+ )
42
+ (locked_dropout): LockedDropout(p=0.5)
43
+ (linear): Linear(in_features=768, out_features=17, bias=True)
44
+ (loss_function): CrossEntropyLoss()
45
+ )"
46
+ 2023-10-17 12:38:39,939 ----------------------------------------------------------------------------------------------------
47
+ 2023-10-17 12:38:39,940 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
48
+ - NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
49
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
50
+ 2023-10-17 12:38:39,940 Train: 7142 sentences
51
+ 2023-10-17 12:38:39,940 (train_with_dev=False, train_with_test=False)
52
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
53
+ 2023-10-17 12:38:39,940 Training Params:
54
+ 2023-10-17 12:38:39,940 - learning_rate: "3e-05"
55
+ 2023-10-17 12:38:39,940 - mini_batch_size: "4"
56
+ 2023-10-17 12:38:39,940 - max_epochs: "10"
57
+ 2023-10-17 12:38:39,940 - shuffle: "True"
58
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
59
+ 2023-10-17 12:38:39,940 Plugins:
60
+ 2023-10-17 12:38:39,940 - TensorboardLogger
61
+ 2023-10-17 12:38:39,940 - LinearScheduler | warmup_fraction: '0.1'
62
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
63
+ 2023-10-17 12:38:39,940 Final evaluation on model from best epoch (best-model.pt)
64
+ 2023-10-17 12:38:39,940 - metric: "('micro avg', 'f1-score')"
65
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
66
+ 2023-10-17 12:38:39,940 Computation:
67
+ 2023-10-17 12:38:39,940 - compute on device: cuda:0
68
+ 2023-10-17 12:38:39,940 - embedding storage: none
69
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
70
+ 2023-10-17 12:38:39,940 Model training base path: "hmbench-newseye/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
71
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
72
+ 2023-10-17 12:38:39,940 ----------------------------------------------------------------------------------------------------
73
+ 2023-10-17 12:38:39,940 Logging anything other than scalars to TensorBoard is currently not supported.
74
+ 2023-10-17 12:38:48,805 epoch 1 - iter 178/1786 - loss 2.58419137 - time (sec): 8.86 - samples/sec: 2764.45 - lr: 0.000003 - momentum: 0.000000
75
+ 2023-10-17 12:38:57,446 epoch 1 - iter 356/1786 - loss 1.60835755 - time (sec): 17.50 - samples/sec: 2837.81 - lr: 0.000006 - momentum: 0.000000
76
+ 2023-10-17 12:39:06,220 epoch 1 - iter 534/1786 - loss 1.19630048 - time (sec): 26.28 - samples/sec: 2857.21 - lr: 0.000009 - momentum: 0.000000
77
+ 2023-10-17 12:39:14,892 epoch 1 - iter 712/1786 - loss 0.98846743 - time (sec): 34.95 - samples/sec: 2812.48 - lr: 0.000012 - momentum: 0.000000
78
+ 2023-10-17 12:39:23,628 epoch 1 - iter 890/1786 - loss 0.84098822 - time (sec): 43.69 - samples/sec: 2818.02 - lr: 0.000015 - momentum: 0.000000
79
+ 2023-10-17 12:39:32,454 epoch 1 - iter 1068/1786 - loss 0.73229552 - time (sec): 52.51 - samples/sec: 2824.31 - lr: 0.000018 - momentum: 0.000000
80
+ 2023-10-17 12:39:41,095 epoch 1 - iter 1246/1786 - loss 0.65469828 - time (sec): 61.15 - samples/sec: 2822.73 - lr: 0.000021 - momentum: 0.000000
81
+ 2023-10-17 12:39:49,717 epoch 1 - iter 1424/1786 - loss 0.58726625 - time (sec): 69.78 - samples/sec: 2840.73 - lr: 0.000024 - momentum: 0.000000
82
+ 2023-10-17 12:39:58,326 epoch 1 - iter 1602/1786 - loss 0.54044553 - time (sec): 78.38 - samples/sec: 2838.36 - lr: 0.000027 - momentum: 0.000000
83
+ 2023-10-17 12:40:07,261 epoch 1 - iter 1780/1786 - loss 0.50050960 - time (sec): 87.32 - samples/sec: 2839.87 - lr: 0.000030 - momentum: 0.000000
84
+ 2023-10-17 12:40:07,548 ----------------------------------------------------------------------------------------------------
85
+ 2023-10-17 12:40:07,548 EPOCH 1 done: loss 0.4994 - lr: 0.000030
86
+ 2023-10-17 12:40:10,607 DEV : loss 0.12284992635250092 - f1-score (micro avg) 0.7364
87
+ 2023-10-17 12:40:10,624 saving best model
88
+ 2023-10-17 12:40:10,972 ----------------------------------------------------------------------------------------------------
89
+ 2023-10-17 12:40:19,706 epoch 2 - iter 178/1786 - loss 0.14323297 - time (sec): 8.73 - samples/sec: 2696.00 - lr: 0.000030 - momentum: 0.000000
90
+ 2023-10-17 12:40:28,597 epoch 2 - iter 356/1786 - loss 0.13101019 - time (sec): 17.62 - samples/sec: 2759.03 - lr: 0.000029 - momentum: 0.000000
91
+ 2023-10-17 12:40:37,665 epoch 2 - iter 534/1786 - loss 0.12875450 - time (sec): 26.69 - samples/sec: 2766.30 - lr: 0.000029 - momentum: 0.000000
92
+ 2023-10-17 12:40:46,600 epoch 2 - iter 712/1786 - loss 0.12509880 - time (sec): 35.63 - samples/sec: 2767.43 - lr: 0.000029 - momentum: 0.000000
93
+ 2023-10-17 12:40:55,119 epoch 2 - iter 890/1786 - loss 0.12118468 - time (sec): 44.15 - samples/sec: 2790.56 - lr: 0.000028 - momentum: 0.000000
94
+ 2023-10-17 12:41:03,960 epoch 2 - iter 1068/1786 - loss 0.12100919 - time (sec): 52.99 - samples/sec: 2800.47 - lr: 0.000028 - momentum: 0.000000
95
+ 2023-10-17 12:41:12,795 epoch 2 - iter 1246/1786 - loss 0.12037138 - time (sec): 61.82 - samples/sec: 2787.10 - lr: 0.000028 - momentum: 0.000000
96
+ 2023-10-17 12:41:21,556 epoch 2 - iter 1424/1786 - loss 0.11693455 - time (sec): 70.58 - samples/sec: 2796.51 - lr: 0.000027 - momentum: 0.000000
97
+ 2023-10-17 12:41:30,702 epoch 2 - iter 1602/1786 - loss 0.11676725 - time (sec): 79.73 - samples/sec: 2820.70 - lr: 0.000027 - momentum: 0.000000
98
+ 2023-10-17 12:41:39,400 epoch 2 - iter 1780/1786 - loss 0.11614911 - time (sec): 88.43 - samples/sec: 2804.09 - lr: 0.000027 - momentum: 0.000000
99
+ 2023-10-17 12:41:39,695 ----------------------------------------------------------------------------------------------------
100
+ 2023-10-17 12:41:39,695 EPOCH 2 done: loss 0.1162 - lr: 0.000027
101
+ 2023-10-17 12:41:44,377 DEV : loss 0.12338940799236298 - f1-score (micro avg) 0.7839
102
+ 2023-10-17 12:41:44,393 saving best model
103
+ 2023-10-17 12:41:44,872 ----------------------------------------------------------------------------------------------------
104
+ 2023-10-17 12:41:53,888 epoch 3 - iter 178/1786 - loss 0.07449215 - time (sec): 9.01 - samples/sec: 2885.63 - lr: 0.000026 - momentum: 0.000000
105
+ 2023-10-17 12:42:02,741 epoch 3 - iter 356/1786 - loss 0.07691580 - time (sec): 17.87 - samples/sec: 2828.11 - lr: 0.000026 - momentum: 0.000000
106
+ 2023-10-17 12:42:11,936 epoch 3 - iter 534/1786 - loss 0.07682986 - time (sec): 27.06 - samples/sec: 2816.30 - lr: 0.000026 - momentum: 0.000000
107
+ 2023-10-17 12:42:20,884 epoch 3 - iter 712/1786 - loss 0.07692302 - time (sec): 36.01 - samples/sec: 2858.19 - lr: 0.000025 - momentum: 0.000000
108
+ 2023-10-17 12:42:29,900 epoch 3 - iter 890/1786 - loss 0.07803257 - time (sec): 45.03 - samples/sec: 2847.16 - lr: 0.000025 - momentum: 0.000000
109
+ 2023-10-17 12:42:38,522 epoch 3 - iter 1068/1786 - loss 0.08020088 - time (sec): 53.65 - samples/sec: 2835.54 - lr: 0.000025 - momentum: 0.000000
110
+ 2023-10-17 12:42:47,357 epoch 3 - iter 1246/1786 - loss 0.08256246 - time (sec): 62.48 - samples/sec: 2828.02 - lr: 0.000024 - momentum: 0.000000
111
+ 2023-10-17 12:42:55,914 epoch 3 - iter 1424/1786 - loss 0.08154790 - time (sec): 71.04 - samples/sec: 2824.51 - lr: 0.000024 - momentum: 0.000000
112
+ 2023-10-17 12:43:04,852 epoch 3 - iter 1602/1786 - loss 0.08168387 - time (sec): 79.98 - samples/sec: 2813.12 - lr: 0.000024 - momentum: 0.000000
113
+ 2023-10-17 12:43:13,481 epoch 3 - iter 1780/1786 - loss 0.08241782 - time (sec): 88.61 - samples/sec: 2798.91 - lr: 0.000023 - momentum: 0.000000
114
+ 2023-10-17 12:43:13,791 ----------------------------------------------------------------------------------------------------
115
+ 2023-10-17 12:43:13,791 EPOCH 3 done: loss 0.0825 - lr: 0.000023
116
+ 2023-10-17 12:43:17,926 DEV : loss 0.14865444600582123 - f1-score (micro avg) 0.81
117
+ 2023-10-17 12:43:17,943 saving best model
118
+ 2023-10-17 12:43:18,414 ----------------------------------------------------------------------------------------------------
119
+ 2023-10-17 12:43:27,142 epoch 4 - iter 178/1786 - loss 0.05085513 - time (sec): 8.73 - samples/sec: 2870.18 - lr: 0.000023 - momentum: 0.000000
120
+ 2023-10-17 12:43:35,981 epoch 4 - iter 356/1786 - loss 0.05118122 - time (sec): 17.56 - samples/sec: 2838.14 - lr: 0.000023 - momentum: 0.000000
121
+ 2023-10-17 12:43:45,038 epoch 4 - iter 534/1786 - loss 0.05465020 - time (sec): 26.62 - samples/sec: 2834.92 - lr: 0.000022 - momentum: 0.000000
122
+ 2023-10-17 12:43:53,763 epoch 4 - iter 712/1786 - loss 0.05592902 - time (sec): 35.35 - samples/sec: 2823.21 - lr: 0.000022 - momentum: 0.000000
123
+ 2023-10-17 12:44:02,634 epoch 4 - iter 890/1786 - loss 0.05561688 - time (sec): 44.22 - samples/sec: 2823.23 - lr: 0.000022 - momentum: 0.000000
124
+ 2023-10-17 12:44:11,294 epoch 4 - iter 1068/1786 - loss 0.05734254 - time (sec): 52.88 - samples/sec: 2847.91 - lr: 0.000021 - momentum: 0.000000
125
+ 2023-10-17 12:44:20,737 epoch 4 - iter 1246/1786 - loss 0.05604538 - time (sec): 62.32 - samples/sec: 2818.67 - lr: 0.000021 - momentum: 0.000000
126
+ 2023-10-17 12:44:29,480 epoch 4 - iter 1424/1786 - loss 0.05505510 - time (sec): 71.06 - samples/sec: 2795.24 - lr: 0.000021 - momentum: 0.000000
127
+ 2023-10-17 12:44:38,262 epoch 4 - iter 1602/1786 - loss 0.05522979 - time (sec): 79.85 - samples/sec: 2789.62 - lr: 0.000020 - momentum: 0.000000
128
+ 2023-10-17 12:44:47,144 epoch 4 - iter 1780/1786 - loss 0.05563942 - time (sec): 88.73 - samples/sec: 2796.45 - lr: 0.000020 - momentum: 0.000000
129
+ 2023-10-17 12:44:47,417 ----------------------------------------------------------------------------------------------------
130
+ 2023-10-17 12:44:47,418 EPOCH 4 done: loss 0.0557 - lr: 0.000020
131
+ 2023-10-17 12:44:51,659 DEV : loss 0.1732509285211563 - f1-score (micro avg) 0.7968
132
+ 2023-10-17 12:44:51,675 ----------------------------------------------------------------------------------------------------
133
+ 2023-10-17 12:45:00,231 epoch 5 - iter 178/1786 - loss 0.04800649 - time (sec): 8.56 - samples/sec: 2861.03 - lr: 0.000020 - momentum: 0.000000
134
+ 2023-10-17 12:45:08,672 epoch 5 - iter 356/1786 - loss 0.04479618 - time (sec): 17.00 - samples/sec: 2806.48 - lr: 0.000019 - momentum: 0.000000
135
+ 2023-10-17 12:45:17,652 epoch 5 - iter 534/1786 - loss 0.04797973 - time (sec): 25.98 - samples/sec: 2798.29 - lr: 0.000019 - momentum: 0.000000
136
+ 2023-10-17 12:45:26,563 epoch 5 - iter 712/1786 - loss 0.04688139 - time (sec): 34.89 - samples/sec: 2802.90 - lr: 0.000019 - momentum: 0.000000
137
+ 2023-10-17 12:45:35,499 epoch 5 - iter 890/1786 - loss 0.04767877 - time (sec): 43.82 - samples/sec: 2840.66 - lr: 0.000018 - momentum: 0.000000
138
+ 2023-10-17 12:45:44,398 epoch 5 - iter 1068/1786 - loss 0.04614346 - time (sec): 52.72 - samples/sec: 2830.04 - lr: 0.000018 - momentum: 0.000000
139
+ 2023-10-17 12:45:53,388 epoch 5 - iter 1246/1786 - loss 0.04538086 - time (sec): 61.71 - samples/sec: 2834.72 - lr: 0.000018 - momentum: 0.000000
140
+ 2023-10-17 12:46:02,100 epoch 5 - iter 1424/1786 - loss 0.04455120 - time (sec): 70.42 - samples/sec: 2835.78 - lr: 0.000017 - momentum: 0.000000
141
+ 2023-10-17 12:46:11,083 epoch 5 - iter 1602/1786 - loss 0.04412066 - time (sec): 79.41 - samples/sec: 2831.98 - lr: 0.000017 - momentum: 0.000000
142
+ 2023-10-17 12:46:19,606 epoch 5 - iter 1780/1786 - loss 0.04400326 - time (sec): 87.93 - samples/sec: 2821.87 - lr: 0.000017 - momentum: 0.000000
143
+ 2023-10-17 12:46:19,874 ----------------------------------------------------------------------------------------------------
144
+ 2023-10-17 12:46:19,874 EPOCH 5 done: loss 0.0440 - lr: 0.000017
145
+ 2023-10-17 12:46:24,615 DEV : loss 0.17575478553771973 - f1-score (micro avg) 0.8176
146
+ 2023-10-17 12:46:24,631 saving best model
147
+ 2023-10-17 12:46:25,082 ----------------------------------------------------------------------------------------------------
148
+ 2023-10-17 12:46:34,054 epoch 6 - iter 178/1786 - loss 0.03021896 - time (sec): 8.97 - samples/sec: 2804.62 - lr: 0.000016 - momentum: 0.000000
149
+ 2023-10-17 12:46:42,641 epoch 6 - iter 356/1786 - loss 0.02818833 - time (sec): 17.55 - samples/sec: 2907.95 - lr: 0.000016 - momentum: 0.000000
150
+ 2023-10-17 12:46:51,341 epoch 6 - iter 534/1786 - loss 0.03091021 - time (sec): 26.25 - samples/sec: 2877.64 - lr: 0.000016 - momentum: 0.000000
151
+ 2023-10-17 12:47:00,933 epoch 6 - iter 712/1786 - loss 0.03157363 - time (sec): 35.85 - samples/sec: 2792.54 - lr: 0.000015 - momentum: 0.000000
152
+ 2023-10-17 12:47:10,770 epoch 6 - iter 890/1786 - loss 0.03226007 - time (sec): 45.68 - samples/sec: 2747.95 - lr: 0.000015 - momentum: 0.000000
153
+ 2023-10-17 12:47:19,864 epoch 6 - iter 1068/1786 - loss 0.03366693 - time (sec): 54.78 - samples/sec: 2773.82 - lr: 0.000015 - momentum: 0.000000
154
+ 2023-10-17 12:47:28,608 epoch 6 - iter 1246/1786 - loss 0.03408422 - time (sec): 63.52 - samples/sec: 2776.37 - lr: 0.000014 - momentum: 0.000000
155
+ 2023-10-17 12:47:37,057 epoch 6 - iter 1424/1786 - loss 0.03358057 - time (sec): 71.97 - samples/sec: 2780.65 - lr: 0.000014 - momentum: 0.000000
156
+ 2023-10-17 12:47:45,656 epoch 6 - iter 1602/1786 - loss 0.03267451 - time (sec): 80.57 - samples/sec: 2774.99 - lr: 0.000014 - momentum: 0.000000
157
+ 2023-10-17 12:47:54,484 epoch 6 - iter 1780/1786 - loss 0.03278461 - time (sec): 89.40 - samples/sec: 2774.25 - lr: 0.000013 - momentum: 0.000000
158
+ 2023-10-17 12:47:54,777 ----------------------------------------------------------------------------------------------------
159
+ 2023-10-17 12:47:54,777 EPOCH 6 done: loss 0.0328 - lr: 0.000013
160
+ 2023-10-17 12:47:59,058 DEV : loss 0.18483255803585052 - f1-score (micro avg) 0.8194
161
+ 2023-10-17 12:47:59,076 saving best model
162
+ 2023-10-17 12:47:59,571 ----------------------------------------------------------------------------------------------------
163
+ 2023-10-17 12:48:08,260 epoch 7 - iter 178/1786 - loss 0.01835655 - time (sec): 8.69 - samples/sec: 2684.07 - lr: 0.000013 - momentum: 0.000000
164
+ 2023-10-17 12:48:17,298 epoch 7 - iter 356/1786 - loss 0.02049370 - time (sec): 17.73 - samples/sec: 2810.28 - lr: 0.000013 - momentum: 0.000000
165
+ 2023-10-17 12:48:26,095 epoch 7 - iter 534/1786 - loss 0.02307035 - time (sec): 26.52 - samples/sec: 2782.84 - lr: 0.000012 - momentum: 0.000000
166
+ 2023-10-17 12:48:35,199 epoch 7 - iter 712/1786 - loss 0.02161123 - time (sec): 35.63 - samples/sec: 2825.58 - lr: 0.000012 - momentum: 0.000000
167
+ 2023-10-17 12:48:43,775 epoch 7 - iter 890/1786 - loss 0.02268236 - time (sec): 44.20 - samples/sec: 2853.93 - lr: 0.000012 - momentum: 0.000000
168
+ 2023-10-17 12:48:52,265 epoch 7 - iter 1068/1786 - loss 0.02277251 - time (sec): 52.69 - samples/sec: 2819.47 - lr: 0.000011 - momentum: 0.000000
169
+ 2023-10-17 12:49:00,956 epoch 7 - iter 1246/1786 - loss 0.02387286 - time (sec): 61.38 - samples/sec: 2802.17 - lr: 0.000011 - momentum: 0.000000
170
+ 2023-10-17 12:49:09,676 epoch 7 - iter 1424/1786 - loss 0.02364284 - time (sec): 70.10 - samples/sec: 2798.53 - lr: 0.000011 - momentum: 0.000000
171
+ 2023-10-17 12:49:18,563 epoch 7 - iter 1602/1786 - loss 0.02334329 - time (sec): 78.99 - samples/sec: 2812.50 - lr: 0.000010 - momentum: 0.000000
172
+ 2023-10-17 12:49:28,118 epoch 7 - iter 1780/1786 - loss 0.02349736 - time (sec): 88.55 - samples/sec: 2802.48 - lr: 0.000010 - momentum: 0.000000
173
+ 2023-10-17 12:49:28,393 ----------------------------------------------------------------------------------------------------
174
+ 2023-10-17 12:49:28,394 EPOCH 7 done: loss 0.0234 - lr: 0.000010
175
+ 2023-10-17 12:49:32,541 DEV : loss 0.19320163130760193 - f1-score (micro avg) 0.8236
176
+ 2023-10-17 12:49:32,559 saving best model
177
+ 2023-10-17 12:49:33,026 ----------------------------------------------------------------------------------------------------
178
+ 2023-10-17 12:49:41,645 epoch 8 - iter 178/1786 - loss 0.01560280 - time (sec): 8.62 - samples/sec: 3039.75 - lr: 0.000010 - momentum: 0.000000
179
+ 2023-10-17 12:49:50,439 epoch 8 - iter 356/1786 - loss 0.01462824 - time (sec): 17.41 - samples/sec: 2947.26 - lr: 0.000009 - momentum: 0.000000
180
+ 2023-10-17 12:49:59,265 epoch 8 - iter 534/1786 - loss 0.01544668 - time (sec): 26.24 - samples/sec: 2902.15 - lr: 0.000009 - momentum: 0.000000
181
+ 2023-10-17 12:50:07,965 epoch 8 - iter 712/1786 - loss 0.01648773 - time (sec): 34.94 - samples/sec: 2856.18 - lr: 0.000009 - momentum: 0.000000
182
+ 2023-10-17 12:50:16,852 epoch 8 - iter 890/1786 - loss 0.01642399 - time (sec): 43.82 - samples/sec: 2847.43 - lr: 0.000008 - momentum: 0.000000
183
+ 2023-10-17 12:50:25,719 epoch 8 - iter 1068/1786 - loss 0.01759485 - time (sec): 52.69 - samples/sec: 2849.73 - lr: 0.000008 - momentum: 0.000000
184
+ 2023-10-17 12:50:34,770 epoch 8 - iter 1246/1786 - loss 0.01639187 - time (sec): 61.74 - samples/sec: 2845.88 - lr: 0.000008 - momentum: 0.000000
185
+ 2023-10-17 12:50:43,717 epoch 8 - iter 1424/1786 - loss 0.01647365 - time (sec): 70.69 - samples/sec: 2863.89 - lr: 0.000007 - momentum: 0.000000
186
+ 2023-10-17 12:50:52,381 epoch 8 - iter 1602/1786 - loss 0.01678055 - time (sec): 79.35 - samples/sec: 2847.54 - lr: 0.000007 - momentum: 0.000000
187
+ 2023-10-17 12:51:00,775 epoch 8 - iter 1780/1786 - loss 0.01689960 - time (sec): 87.75 - samples/sec: 2827.07 - lr: 0.000007 - momentum: 0.000000
188
+ 2023-10-17 12:51:01,071 ----------------------------------------------------------------------------------------------------
189
+ 2023-10-17 12:51:01,071 EPOCH 8 done: loss 0.0169 - lr: 0.000007
190
+ 2023-10-17 12:51:05,268 DEV : loss 0.19824054837226868 - f1-score (micro avg) 0.8207
191
+ 2023-10-17 12:51:05,284 ----------------------------------------------------------------------------------------------------
192
+ 2023-10-17 12:51:14,686 epoch 9 - iter 178/1786 - loss 0.01052900 - time (sec): 9.40 - samples/sec: 2720.90 - lr: 0.000006 - momentum: 0.000000
193
+ 2023-10-17 12:51:23,649 epoch 9 - iter 356/1786 - loss 0.01469236 - time (sec): 18.36 - samples/sec: 2771.46 - lr: 0.000006 - momentum: 0.000000
194
+ 2023-10-17 12:51:32,564 epoch 9 - iter 534/1786 - loss 0.01355989 - time (sec): 27.28 - samples/sec: 2805.94 - lr: 0.000006 - momentum: 0.000000
195
+ 2023-10-17 12:51:40,994 epoch 9 - iter 712/1786 - loss 0.01249401 - time (sec): 35.71 - samples/sec: 2814.57 - lr: 0.000005 - momentum: 0.000000
196
+ 2023-10-17 12:51:49,608 epoch 9 - iter 890/1786 - loss 0.01257532 - time (sec): 44.32 - samples/sec: 2829.89 - lr: 0.000005 - momentum: 0.000000
197
+ 2023-10-17 12:51:58,268 epoch 9 - iter 1068/1786 - loss 0.01305676 - time (sec): 52.98 - samples/sec: 2831.92 - lr: 0.000005 - momentum: 0.000000
198
+ 2023-10-17 12:52:07,045 epoch 9 - iter 1246/1786 - loss 0.01263776 - time (sec): 61.76 - samples/sec: 2824.51 - lr: 0.000004 - momentum: 0.000000
199
+ 2023-10-17 12:52:15,718 epoch 9 - iter 1424/1786 - loss 0.01215768 - time (sec): 70.43 - samples/sec: 2808.67 - lr: 0.000004 - momentum: 0.000000
200
+ 2023-10-17 12:52:24,522 epoch 9 - iter 1602/1786 - loss 0.01172614 - time (sec): 79.24 - samples/sec: 2815.65 - lr: 0.000004 - momentum: 0.000000
201
+ 2023-10-17 12:52:33,450 epoch 9 - iter 1780/1786 - loss 0.01183726 - time (sec): 88.16 - samples/sec: 2814.13 - lr: 0.000003 - momentum: 0.000000
202
+ 2023-10-17 12:52:33,742 ----------------------------------------------------------------------------------------------------
203
+ 2023-10-17 12:52:33,742 EPOCH 9 done: loss 0.0118 - lr: 0.000003
204
+ 2023-10-17 12:52:37,926 DEV : loss 0.20676946640014648 - f1-score (micro avg) 0.8172
205
+ 2023-10-17 12:52:37,944 ----------------------------------------------------------------------------------------------------
206
+ 2023-10-17 12:52:46,932 epoch 10 - iter 178/1786 - loss 0.01278157 - time (sec): 8.99 - samples/sec: 2814.97 - lr: 0.000003 - momentum: 0.000000
207
+ 2023-10-17 12:52:55,908 epoch 10 - iter 356/1786 - loss 0.01285362 - time (sec): 17.96 - samples/sec: 2786.82 - lr: 0.000003 - momentum: 0.000000
208
+ 2023-10-17 12:53:04,725 epoch 10 - iter 534/1786 - loss 0.01038423 - time (sec): 26.78 - samples/sec: 2775.89 - lr: 0.000002 - momentum: 0.000000
209
+ 2023-10-17 12:53:13,650 epoch 10 - iter 712/1786 - loss 0.01035496 - time (sec): 35.70 - samples/sec: 2804.32 - lr: 0.000002 - momentum: 0.000000
210
+ 2023-10-17 12:53:22,361 epoch 10 - iter 890/1786 - loss 0.00954459 - time (sec): 44.42 - samples/sec: 2809.79 - lr: 0.000002 - momentum: 0.000000
211
+ 2023-10-17 12:53:30,976 epoch 10 - iter 1068/1786 - loss 0.00937681 - time (sec): 53.03 - samples/sec: 2799.34 - lr: 0.000001 - momentum: 0.000000
212
+ 2023-10-17 12:53:39,358 epoch 10 - iter 1246/1786 - loss 0.00888181 - time (sec): 61.41 - samples/sec: 2807.33 - lr: 0.000001 - momentum: 0.000000
213
+ 2023-10-17 12:53:47,865 epoch 10 - iter 1424/1786 - loss 0.00852283 - time (sec): 69.92 - samples/sec: 2809.14 - lr: 0.000001 - momentum: 0.000000
214
+ 2023-10-17 12:53:56,420 epoch 10 - iter 1602/1786 - loss 0.00837085 - time (sec): 78.47 - samples/sec: 2821.52 - lr: 0.000000 - momentum: 0.000000
215
+ 2023-10-17 12:54:05,123 epoch 10 - iter 1780/1786 - loss 0.00851493 - time (sec): 87.18 - samples/sec: 2846.18 - lr: 0.000000 - momentum: 0.000000
216
+ 2023-10-17 12:54:05,392 ----------------------------------------------------------------------------------------------------
217
+ 2023-10-17 12:54:05,392 EPOCH 10 done: loss 0.0085 - lr: 0.000000
218
+ 2023-10-17 12:54:10,017 DEV : loss 0.21040573716163635 - f1-score (micro avg) 0.83
219
+ 2023-10-17 12:54:10,033 saving best model
220
+ 2023-10-17 12:54:10,788 ----------------------------------------------------------------------------------------------------
221
+ 2023-10-17 12:54:10,789 Loading model from best epoch ...
222
+ 2023-10-17 12:54:12,118 SequenceTagger predicts: Dictionary with 17 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, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
223
+ 2023-10-17 12:54:21,698
224
+ Results:
225
+ - F-score (micro) 0.7085
226
+ - F-score (macro) 0.642
227
+ - Accuracy 0.5666
228
+
229
+ By class:
230
+ precision recall f1-score support
231
+
232
+ LOC 0.7307 0.7014 0.7158 1095
233
+ PER 0.7868 0.7915 0.7892 1012
234
+ ORG 0.4543 0.5574 0.5006 357
235
+ HumanProd 0.4286 0.8182 0.5625 33
236
+
237
+ micro avg 0.6984 0.7189 0.7085 2497
238
+ macro avg 0.6001 0.7171 0.6420 2497
239
+ weighted avg 0.7100 0.7189 0.7127 2497
240
+
241
+ 2023-10-17 12:54:21,698 ----------------------------------------------------------------------------------------------------