Hugging Face's logo

stefan-it
/
hmbench-icdar-fr-hmbyt5-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-1

Token Classification
Flair
PyTorch
TensorBoard
French
sequence-tagger-model
Model card Files Metrics Community
2
stefan-it commited on
Commit
5708ee7
•
1 Parent(s): 8a86a03

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. best-model.pt +3 -0
  2. dev.tsv +0 -0
  3. final-model.pt +3 -0
  4. loss.tsv +11 -0
  5. runs/events.out.tfevents.1697114616.de2e83fddbee.1952.3 +3 -0
  6. test.tsv +0 -0
  7. training.log +260 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d129491608f083514048693ccd42cad98962753b31328cf117fce5ab45a7dec1
3
+ size 870793839
dev.tsv ADDED
The diff for this file is too large to render. See raw diff
 
final-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65aad3892bbc4ee0c2aea086ee0b534ad4a65d48ed06e3c88db3085089b97f59
3
+ size 870793956
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 12:53:19 0.0002 0.8384 0.1406 0.6283 0.6731 0.6499 0.5143
3
+ 2 13:03:09 0.0001 0.1223 0.0859 0.7242 0.7308 0.7275 0.5943
4
+ 3 13:13:08 0.0001 0.0760 0.1018 0.7353 0.7919 0.7625 0.6375
5
+ 4 13:23:11 0.0001 0.0560 0.1271 0.7370 0.7828 0.7592 0.6331
6
+ 5 13:32:48 0.0001 0.0425 0.1447 0.7183 0.7760 0.7461 0.6236
7
+ 6 13:42:19 0.0001 0.0297 0.1701 0.7544 0.7681 0.7612 0.6316
8
+ 7 13:51:59 0.0001 0.0205 0.2086 0.7330 0.7670 0.7496 0.6192
9
+ 8 14:01:58 0.0000 0.0150 0.2162 0.7462 0.7681 0.7570 0.6270
10
+ 9 14:11:55 0.0000 0.0107 0.2269 0.7505 0.7726 0.7614 0.6342
11
+ 10 14:21:51 0.0000 0.0085 0.2328 0.7478 0.7681 0.7578 0.6287
runs/events.out.tfevents.1697114616.de2e83fddbee.1952.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1c2629a82ff4d310c0c48400a5b64cab3ac1ce34c62ca400ad09432ea39203f
3
+ size 1108164
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,260 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-12 12:43:36,931 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-12 12:43:36,933 Model: "SequenceTagger(
3
+ (embeddings): ByT5Embeddings(
4
+ (model): T5EncoderModel(
5
+ (shared): Embedding(384, 1472)
6
+ (encoder): T5Stack(
7
+ (embed_tokens): Embedding(384, 1472)
8
+ (block): ModuleList(
9
+ (0): T5Block(
10
+ (layer): ModuleList(
11
+ (0): T5LayerSelfAttention(
12
+ (SelfAttention): T5Attention(
13
+ (q): Linear(in_features=1472, out_features=384, bias=False)
14
+ (k): Linear(in_features=1472, out_features=384, bias=False)
15
+ (v): Linear(in_features=1472, out_features=384, bias=False)
16
+ (o): Linear(in_features=384, out_features=1472, bias=False)
17
+ (relative_attention_bias): Embedding(32, 6)
18
+ )
19
+ (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
20
+ (dropout): Dropout(p=0.1, inplace=False)
21
+ )
22
+ (1): T5LayerFF(
23
+ (DenseReluDense): T5DenseGatedActDense(
24
+ (wi_0): Linear(in_features=1472, out_features=3584, bias=False)
25
+ (wi_1): Linear(in_features=1472, out_features=3584, bias=False)
26
+ (wo): Linear(in_features=3584, out_features=1472, bias=False)
27
+ (dropout): Dropout(p=0.1, inplace=False)
28
+ (act): NewGELUActivation()
29
+ )
30
+ (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
31
+ (dropout): Dropout(p=0.1, inplace=False)
32
+ )
33
+ )
34
+ )
35
+ (1-11): 11 x T5Block(
36
+ (layer): ModuleList(
37
+ (0): T5LayerSelfAttention(
38
+ (SelfAttention): T5Attention(
39
+ (q): Linear(in_features=1472, out_features=384, bias=False)
40
+ (k): Linear(in_features=1472, out_features=384, bias=False)
41
+ (v): Linear(in_features=1472, out_features=384, bias=False)
42
+ (o): Linear(in_features=384, out_features=1472, bias=False)
43
+ )
44
+ (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
45
+ (dropout): Dropout(p=0.1, inplace=False)
46
+ )
47
+ (1): T5LayerFF(
48
+ (DenseReluDense): T5DenseGatedActDense(
49
+ (wi_0): Linear(in_features=1472, out_features=3584, bias=False)
50
+ (wi_1): Linear(in_features=1472, out_features=3584, bias=False)
51
+ (wo): Linear(in_features=3584, out_features=1472, bias=False)
52
+ (dropout): Dropout(p=0.1, inplace=False)
53
+ (act): NewGELUActivation()
54
+ )
55
+ (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
56
+ (dropout): Dropout(p=0.1, inplace=False)
57
+ )
58
+ )
59
+ )
60
+ )
61
+ (final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
62
+ (dropout): Dropout(p=0.1, inplace=False)
63
+ )
64
+ )
65
+ )
66
+ (locked_dropout): LockedDropout(p=0.5)
67
+ (linear): Linear(in_features=1472, out_features=13, bias=True)
68
+ (loss_function): CrossEntropyLoss()
69
+ )"
70
+ 2023-10-12 12:43:36,933 ----------------------------------------------------------------------------------------------------
71
+ 2023-10-12 12:43:36,933 MultiCorpus: 7936 train + 992 dev + 992 test sentences
72
+ - NER_ICDAR_EUROPEANA Corpus: 7936 train + 992 dev + 992 test sentences - /root/.flair/datasets/ner_icdar_europeana/fr
73
+ 2023-10-12 12:43:36,933 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-12 12:43:36,933 Train: 7936 sentences
75
+ 2023-10-12 12:43:36,933 (train_with_dev=False, train_with_test=False)
76
+ 2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-12 12:43:36,934 Training Params:
78
+ 2023-10-12 12:43:36,934 - learning_rate: "0.00016"
79
+ 2023-10-12 12:43:36,934 - mini_batch_size: "4"
80
+ 2023-10-12 12:43:36,934 - max_epochs: "10"
81
+ 2023-10-12 12:43:36,934 - shuffle: "True"
82
+ 2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
83
+ 2023-10-12 12:43:36,934 Plugins:
84
+ 2023-10-12 12:43:36,934 - TensorboardLogger
85
+ 2023-10-12 12:43:36,934 - LinearScheduler | warmup_fraction: '0.1'
86
+ 2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-12 12:43:36,934 Final evaluation on model from best epoch (best-model.pt)
88
+ 2023-10-12 12:43:36,934 - metric: "('micro avg', 'f1-score')"
89
+ 2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
90
+ 2023-10-12 12:43:36,934 Computation:
91
+ 2023-10-12 12:43:36,935 - compute on device: cuda:0
92
+ 2023-10-12 12:43:36,935 - embedding storage: none
93
+ 2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
94
+ 2023-10-12 12:43:36,935 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-1"
95
+ 2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
96
+ 2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
97
+ 2023-10-12 12:43:36,935 Logging anything other than scalars to TensorBoard is currently not supported.
98
+ 2023-10-12 12:44:29,444 epoch 1 - iter 198/1984 - loss 2.57872388 - time (sec): 52.51 - samples/sec: 294.57 - lr: 0.000016 - momentum: 0.000000
99
+ 2023-10-12 12:45:23,159 epoch 1 - iter 396/1984 - loss 2.42977020 - time (sec): 106.22 - samples/sec: 296.79 - lr: 0.000032 - momentum: 0.000000
100
+ 2023-10-12 12:46:17,601 epoch 1 - iter 594/1984 - loss 2.09651161 - time (sec): 160.66 - samples/sec: 298.18 - lr: 0.000048 - momentum: 0.000000
101
+ 2023-10-12 12:47:19,176 epoch 1 - iter 792/1984 - loss 1.74658842 - time (sec): 222.24 - samples/sec: 289.43 - lr: 0.000064 - momentum: 0.000000
102
+ 2023-10-12 12:48:12,650 epoch 1 - iter 990/1984 - loss 1.47898744 - time (sec): 275.71 - samples/sec: 291.16 - lr: 0.000080 - momentum: 0.000000
103
+ 2023-10-12 12:49:03,780 epoch 1 - iter 1188/1984 - loss 1.27073991 - time (sec): 326.84 - samples/sec: 296.79 - lr: 0.000096 - momentum: 0.000000
104
+ 2023-10-12 12:49:58,979 epoch 1 - iter 1386/1984 - loss 1.11087775 - time (sec): 382.04 - samples/sec: 299.34 - lr: 0.000112 - momentum: 0.000000
105
+ 2023-10-12 12:51:00,244 epoch 1 - iter 1584/1984 - loss 0.99932054 - time (sec): 443.31 - samples/sec: 295.65 - lr: 0.000128 - momentum: 0.000000
106
+ 2023-10-12 12:51:58,406 epoch 1 - iter 1782/1984 - loss 0.91329740 - time (sec): 501.47 - samples/sec: 293.81 - lr: 0.000144 - momentum: 0.000000
107
+ 2023-10-12 12:52:52,746 epoch 1 - iter 1980/1984 - loss 0.83963990 - time (sec): 555.81 - samples/sec: 294.49 - lr: 0.000160 - momentum: 0.000000
108
+ 2023-10-12 12:52:53,861 ----------------------------------------------------------------------------------------------------
109
+ 2023-10-12 12:52:53,862 EPOCH 1 done: loss 0.8384 - lr: 0.000160
110
+ 2023-10-12 12:53:19,383 DEV : loss 0.14055970311164856 - f1-score (micro avg) 0.6499
111
+ 2023-10-12 12:53:19,431 saving best model
112
+ 2023-10-12 12:53:20,360 ----------------------------------------------------------------------------------------------------
113
+ 2023-10-12 12:54:14,361 epoch 2 - iter 198/1984 - loss 0.16528251 - time (sec): 54.00 - samples/sec: 305.04 - lr: 0.000158 - momentum: 0.000000
114
+ 2023-10-12 12:55:10,170 epoch 2 - iter 396/1984 - loss 0.14401493 - time (sec): 109.81 - samples/sec: 297.71 - lr: 0.000156 - momentum: 0.000000
115
+ 2023-10-12 12:56:08,489 epoch 2 - iter 594/1984 - loss 0.13710523 - time (sec): 168.13 - samples/sec: 291.73 - lr: 0.000155 - momentum: 0.000000
116
+ 2023-10-12 12:57:06,323 epoch 2 - iter 792/1984 - loss 0.13661458 - time (sec): 225.96 - samples/sec: 290.86 - lr: 0.000153 - momentum: 0.000000
117
+ 2023-10-12 12:58:01,146 epoch 2 - iter 990/1984 - loss 0.13287502 - time (sec): 280.78 - samples/sec: 293.36 - lr: 0.000151 - momentum: 0.000000
118
+ 2023-10-12 12:58:55,564 epoch 2 - iter 1188/1984 - loss 0.13028398 - time (sec): 335.20 - samples/sec: 293.98 - lr: 0.000149 - momentum: 0.000000
119
+ 2023-10-12 12:59:49,914 epoch 2 - iter 1386/1984 - loss 0.13061150 - time (sec): 389.55 - samples/sec: 295.42 - lr: 0.000148 - momentum: 0.000000
120
+ 2023-10-12 13:00:43,318 epoch 2 - iter 1584/1984 - loss 0.12716756 - time (sec): 442.96 - samples/sec: 295.30 - lr: 0.000146 - momentum: 0.000000
121
+ 2023-10-12 13:01:41,149 epoch 2 - iter 1782/1984 - loss 0.12425762 - time (sec): 500.79 - samples/sec: 293.51 - lr: 0.000144 - momentum: 0.000000
122
+ 2023-10-12 13:02:40,598 epoch 2 - iter 1980/1984 - loss 0.12247253 - time (sec): 560.24 - samples/sec: 291.85 - lr: 0.000142 - momentum: 0.000000
123
+ 2023-10-12 13:02:41,892 ----------------------------------------------------------------------------------------------------
124
+ 2023-10-12 13:02:41,893 EPOCH 2 done: loss 0.1223 - lr: 0.000142
125
+ 2023-10-12 13:03:09,818 DEV : loss 0.0859101414680481 - f1-score (micro avg) 0.7275
126
+ 2023-10-12 13:03:09,869 saving best model
127
+ 2023-10-12 13:03:10,961 ----------------------------------------------------------------------------------------------------
128
+ 2023-10-12 13:04:04,131 epoch 3 - iter 198/1984 - loss 0.07669844 - time (sec): 53.17 - samples/sec: 296.03 - lr: 0.000140 - momentum: 0.000000
129
+ 2023-10-12 13:04:57,976 epoch 3 - iter 396/1984 - loss 0.07779340 - time (sec): 107.01 - samples/sec: 298.62 - lr: 0.000139 - momentum: 0.000000
130
+ 2023-10-12 13:05:53,748 epoch 3 - iter 594/1984 - loss 0.07883378 - time (sec): 162.78 - samples/sec: 299.80 - lr: 0.000137 - momentum: 0.000000
131
+ 2023-10-12 13:06:53,717 epoch 3 - iter 792/1984 - loss 0.07827758 - time (sec): 222.75 - samples/sec: 292.63 - lr: 0.000135 - momentum: 0.000000
132
+ 2023-10-12 13:07:52,154 epoch 3 - iter 990/1984 - loss 0.07681269 - time (sec): 281.19 - samples/sec: 289.45 - lr: 0.000133 - momentum: 0.000000
133
+ 2023-10-12 13:08:51,767 epoch 3 - iter 1188/1984 - loss 0.07663353 - time (sec): 340.80 - samples/sec: 285.97 - lr: 0.000132 - momentum: 0.000000
134
+ 2023-10-12 13:09:47,186 epoch 3 - iter 1386/1984 - loss 0.07734261 - time (sec): 396.22 - samples/sec: 285.72 - lr: 0.000130 - momentum: 0.000000
135
+ 2023-10-12 13:10:44,118 epoch 3 - iter 1584/1984 - loss 0.07604814 - time (sec): 453.15 - samples/sec: 288.78 - lr: 0.000128 - momentum: 0.000000
136
+ 2023-10-12 13:11:40,292 epoch 3 - iter 1782/1984 - loss 0.07555546 - time (sec): 509.33 - samples/sec: 290.07 - lr: 0.000126 - momentum: 0.000000
137
+ 2023-10-12 13:12:38,011 epoch 3 - iter 1980/1984 - loss 0.07603113 - time (sec): 567.05 - samples/sec: 288.67 - lr: 0.000125 - momentum: 0.000000
138
+ 2023-10-12 13:12:39,271 ----------------------------------------------------------------------------------------------------
139
+ 2023-10-12 13:12:39,271 EPOCH 3 done: loss 0.0760 - lr: 0.000125
140
+ 2023-10-12 13:13:08,663 DEV : loss 0.10180744528770447 - f1-score (micro avg) 0.7625
141
+ 2023-10-12 13:13:08,710 saving best model
142
+ 2023-10-12 13:13:11,483 ----------------------------------------------------------------------------------------------------
143
+ 2023-10-12 13:14:09,654 epoch 4 - iter 198/1984 - loss 0.04829137 - time (sec): 58.16 - samples/sec: 294.70 - lr: 0.000123 - momentum: 0.000000
144
+ 2023-10-12 13:15:02,659 epoch 4 - iter 396/1984 - loss 0.05173970 - time (sec): 111.17 - samples/sec: 306.37 - lr: 0.000121 - momentum: 0.000000
145
+ 2023-10-12 13:15:57,996 epoch 4 - iter 594/1984 - loss 0.05651052 - time (sec): 166.51 - samples/sec: 300.16 - lr: 0.000119 - momentum: 0.000000
146
+ 2023-10-12 13:16:53,481 epoch 4 - iter 792/1984 - loss 0.05517049 - time (sec): 221.99 - samples/sec: 296.01 - lr: 0.000117 - momentum: 0.000000
147
+ 2023-10-12 13:17:52,233 epoch 4 - iter 990/1984 - loss 0.05634728 - time (sec): 280.74 - samples/sec: 291.99 - lr: 0.000116 - momentum: 0.000000
148
+ 2023-10-12 13:18:49,209 epoch 4 - iter 1188/1984 - loss 0.05607999 - time (sec): 337.72 - samples/sec: 290.31 - lr: 0.000114 - momentum: 0.000000
149
+ 2023-10-12 13:19:46,175 epoch 4 - iter 1386/1984 - loss 0.05551788 - time (sec): 394.69 - samples/sec: 290.32 - lr: 0.000112 - momentum: 0.000000
150
+ 2023-10-12 13:20:44,641 epoch 4 - iter 1584/1984 - loss 0.05700501 - time (sec): 453.15 - samples/sec: 288.52 - lr: 0.000110 - momentum: 0.000000
151
+ 2023-10-12 13:21:46,416 epoch 4 - iter 1782/1984 - loss 0.05565357 - time (sec): 514.93 - samples/sec: 286.92 - lr: 0.000109 - momentum: 0.000000
152
+ 2023-10-12 13:22:42,025 epoch 4 - iter 1980/1984 - loss 0.05607850 - time (sec): 570.53 - samples/sec: 287.00 - lr: 0.000107 - momentum: 0.000000
153
+ 2023-10-12 13:22:43,130 ----------------------------------------------------------------------------------------------------
154
+ 2023-10-12 13:22:43,130 EPOCH 4 done: loss 0.0560 - lr: 0.000107
155
+ 2023-10-12 13:23:11,653 DEV : loss 0.12707574665546417 - f1-score (micro avg) 0.7592
156
+ 2023-10-12 13:23:11,704 ----------------------------------------------------------------------------------------------------
157
+ 2023-10-12 13:24:12,175 epoch 5 - iter 198/1984 - loss 0.04138412 - time (sec): 60.47 - samples/sec: 267.17 - lr: 0.000105 - momentum: 0.000000
158
+ 2023-10-12 13:25:08,068 epoch 5 - iter 396/1984 - loss 0.03535050 - time (sec): 116.36 - samples/sec: 277.92 - lr: 0.000103 - momentum: 0.000000
159
+ 2023-10-12 13:26:04,766 epoch 5 - iter 594/1984 - loss 0.03648738 - time (sec): 173.06 - samples/sec: 281.12 - lr: 0.000101 - momentum: 0.000000
160
+ 2023-10-12 13:27:01,501 epoch 5 - iter 792/1984 - loss 0.03686215 - time (sec): 229.79 - samples/sec: 283.29 - lr: 0.000100 - momentum: 0.000000
161
+ 2023-10-12 13:27:53,816 epoch 5 - iter 990/1984 - loss 0.03625039 - time (sec): 282.11 - samples/sec: 288.19 - lr: 0.000098 - momentum: 0.000000
162
+ 2023-10-12 13:28:46,678 epoch 5 - iter 1188/1984 - loss 0.03960505 - time (sec): 334.97 - samples/sec: 292.34 - lr: 0.000096 - momentum: 0.000000
163
+ 2023-10-12 13:29:40,681 epoch 5 - iter 1386/1984 - loss 0.04001738 - time (sec): 388.97 - samples/sec: 294.20 - lr: 0.000094 - momentum: 0.000000
164
+ 2023-10-12 13:30:35,002 epoch 5 - iter 1584/1984 - loss 0.04082094 - time (sec): 443.29 - samples/sec: 296.26 - lr: 0.000093 - momentum: 0.000000
165
+ 2023-10-12 13:31:28,676 epoch 5 - iter 1782/1984 - loss 0.04148610 - time (sec): 496.97 - samples/sec: 297.49 - lr: 0.000091 - momentum: 0.000000
166
+ 2023-10-12 13:32:21,692 epoch 5 - iter 1980/1984 - loss 0.04251132 - time (sec): 549.99 - samples/sec: 297.50 - lr: 0.000089 - momentum: 0.000000
167
+ 2023-10-12 13:32:22,804 ----------------------------------------------------------------------------------------------------
168
+ 2023-10-12 13:32:22,804 EPOCH 5 done: loss 0.0425 - lr: 0.000089
169
+ 2023-10-12 13:32:48,621 DEV : loss 0.14472655951976776 - f1-score (micro avg) 0.7461
170
+ 2023-10-12 13:32:48,680 ----------------------------------------------------------------------------------------------------
171
+ 2023-10-12 13:33:42,335 epoch 6 - iter 198/1984 - loss 0.02627233 - time (sec): 53.65 - samples/sec: 291.97 - lr: 0.000087 - momentum: 0.000000
172
+ 2023-10-12 13:34:34,598 epoch 6 - iter 396/1984 - loss 0.02760424 - time (sec): 105.92 - samples/sec: 302.33 - lr: 0.000085 - momentum: 0.000000
173
+ 2023-10-12 13:35:26,843 epoch 6 - iter 594/1984 - loss 0.02865242 - time (sec): 158.16 - samples/sec: 304.60 - lr: 0.000084 - momentum: 0.000000
174
+ 2023-10-12 13:36:21,745 epoch 6 - iter 792/1984 - loss 0.02818054 - time (sec): 213.06 - samples/sec: 305.87 - lr: 0.000082 - momentum: 0.000000
175
+ 2023-10-12 13:37:16,009 epoch 6 - iter 990/1984 - loss 0.02691623 - time (sec): 267.33 - samples/sec: 303.51 - lr: 0.000080 - momentum: 0.000000
176
+ 2023-10-12 13:38:10,624 epoch 6 - iter 1188/1984 - loss 0.02670693 - time (sec): 321.94 - samples/sec: 303.88 - lr: 0.000078 - momentum: 0.000000
177
+ 2023-10-12 13:39:05,104 epoch 6 - iter 1386/1984 - loss 0.02728730 - time (sec): 376.42 - samples/sec: 304.70 - lr: 0.000077 - momentum: 0.000000
178
+ 2023-10-12 13:40:00,408 epoch 6 - iter 1584/1984 - loss 0.02914735 - time (sec): 431.73 - samples/sec: 302.80 - lr: 0.000075 - momentum: 0.000000
179
+ 2023-10-12 13:40:54,644 epoch 6 - iter 1782/1984 - loss 0.02922904 - time (sec): 485.96 - samples/sec: 303.18 - lr: 0.000073 - momentum: 0.000000
180
+ 2023-10-12 13:41:50,958 epoch 6 - iter 1980/1984 - loss 0.02961781 - time (sec): 542.28 - samples/sec: 301.72 - lr: 0.000071 - momentum: 0.000000
181
+ 2023-10-12 13:41:52,135 ----------------------------------------------------------------------------------------------------
182
+ 2023-10-12 13:41:52,136 EPOCH 6 done: loss 0.0297 - lr: 0.000071
183
+ 2023-10-12 13:42:19,073 DEV : loss 0.17014847695827484 - f1-score (micro avg) 0.7612
184
+ 2023-10-12 13:42:19,119 ----------------------------------------------------------------------------------------------------
185
+ 2023-10-12 13:43:11,962 epoch 7 - iter 198/1984 - loss 0.01757743 - time (sec): 52.84 - samples/sec: 308.61 - lr: 0.000069 - momentum: 0.000000
186
+ 2023-10-12 13:44:03,780 epoch 7 - iter 396/1984 - loss 0.02000202 - time (sec): 104.66 - samples/sec: 315.03 - lr: 0.000068 - momentum: 0.000000
187
+ 2023-10-12 13:44:57,934 epoch 7 - iter 594/1984 - loss 0.01900677 - time (sec): 158.81 - samples/sec: 308.16 - lr: 0.000066 - momentum: 0.000000
188
+ 2023-10-12 13:45:54,285 epoch 7 - iter 792/1984 - loss 0.02132227 - time (sec): 215.16 - samples/sec: 305.10 - lr: 0.000064 - momentum: 0.000000
189
+ 2023-10-12 13:46:51,268 epoch 7 - iter 990/1984 - loss 0.02065637 - time (sec): 272.15 - samples/sec: 299.69 - lr: 0.000062 - momentum: 0.000000
190
+ 2023-10-12 13:47:50,086 epoch 7 - iter 1188/1984 - loss 0.02083419 - time (sec): 330.96 - samples/sec: 295.66 - lr: 0.000061 - momentum: 0.000000
191
+ 2023-10-12 13:48:47,692 epoch 7 - iter 1386/1984 - loss 0.02064694 - time (sec): 388.57 - samples/sec: 295.32 - lr: 0.000059 - momentum: 0.000000
192
+ 2023-10-12 13:49:41,352 epoch 7 - iter 1584/1984 - loss 0.02141756 - time (sec): 442.23 - samples/sec: 293.14 - lr: 0.000057 - momentum: 0.000000
193
+ 2023-10-12 13:50:36,232 epoch 7 - iter 1782/1984 - loss 0.02092148 - time (sec): 497.11 - samples/sec: 294.32 - lr: 0.000055 - momentum: 0.000000
194
+ 2023-10-12 13:51:31,622 epoch 7 - iter 1980/1984 - loss 0.02052943 - time (sec): 552.50 - samples/sec: 296.11 - lr: 0.000053 - momentum: 0.000000
195
+ 2023-10-12 13:51:32,742 ----------------------------------------------------------------------------------------------------
196
+ 2023-10-12 13:51:32,742 EPOCH 7 done: loss 0.0205 - lr: 0.000053
197
+ 2023-10-12 13:51:59,673 DEV : loss 0.2085985541343689 - f1-score (micro avg) 0.7496
198
+ 2023-10-12 13:51:59,718 ----------------------------------------------------------------------------------------------------
199
+ 2023-10-12 13:52:54,707 epoch 8 - iter 198/1984 - loss 0.01667937 - time (sec): 54.99 - samples/sec: 307.84 - lr: 0.000052 - momentum: 0.000000
200
+ 2023-10-12 13:53:50,334 epoch 8 - iter 396/1984 - loss 0.01438657 - time (sec): 110.61 - samples/sec: 291.56 - lr: 0.000050 - momentum: 0.000000
201
+ 2023-10-12 13:54:45,708 epoch 8 - iter 594/1984 - loss 0.01421150 - time (sec): 165.99 - samples/sec: 287.68 - lr: 0.000048 - momentum: 0.000000
202
+ 2023-10-12 13:55:41,536 epoch 8 - iter 792/1984 - loss 0.01432064 - time (sec): 221.82 - samples/sec: 287.50 - lr: 0.000046 - momentum: 0.000000
203
+ 2023-10-12 13:56:38,668 epoch 8 - iter 990/1984 - loss 0.01369389 - time (sec): 278.95 - samples/sec: 288.69 - lr: 0.000045 - momentum: 0.000000
204
+ 2023-10-12 13:57:37,033 epoch 8 - iter 1188/1984 - loss 0.01537915 - time (sec): 337.31 - samples/sec: 290.16 - lr: 0.000043 - momentum: 0.000000
205
+ 2023-10-12 13:58:34,687 epoch 8 - iter 1386/1984 - loss 0.01524152 - time (sec): 394.97 - samples/sec: 288.48 - lr: 0.000041 - momentum: 0.000000
206
+ 2023-10-12 13:59:30,605 epoch 8 - iter 1584/1984 - loss 0.01517160 - time (sec): 450.88 - samples/sec: 289.72 - lr: 0.000039 - momentum: 0.000000
207
+ 2023-10-12 14:00:28,078 epoch 8 - iter 1782/1984 - loss 0.01553226 - time (sec): 508.36 - samples/sec: 288.49 - lr: 0.000037 - momentum: 0.000000
208
+ 2023-10-12 14:01:27,127 epoch 8 - iter 1980/1984 - loss 0.01498313 - time (sec): 567.41 - samples/sec: 288.59 - lr: 0.000036 - momentum: 0.000000
209
+ 2023-10-12 14:01:28,169 ----------------------------------------------------------------------------------------------------
210
+ 2023-10-12 14:01:28,170 EPOCH 8 done: loss 0.0150 - lr: 0.000036
211
+ 2023-10-12 14:01:58,438 DEV : loss 0.21621058881282806 - f1-score (micro avg) 0.757
212
+ 2023-10-12 14:01:58,492 ----------------------------------------------------------------------------------------------------
213
+ 2023-10-12 14:02:56,912 epoch 9 - iter 198/1984 - loss 0.01321939 - time (sec): 58.42 - samples/sec: 294.64 - lr: 0.000034 - momentum: 0.000000
214
+ 2023-10-12 14:03:55,154 epoch 9 - iter 396/1984 - loss 0.01312927 - time (sec): 116.66 - samples/sec: 288.45 - lr: 0.000032 - momentum: 0.000000
215
+ 2023-10-12 14:04:49,193 epoch 9 - iter 594/1984 - loss 0.01098511 - time (sec): 170.70 - samples/sec: 296.55 - lr: 0.000030 - momentum: 0.000000
216
+ 2023-10-12 14:05:43,288 epoch 9 - iter 792/1984 - loss 0.01215431 - time (sec): 224.79 - samples/sec: 294.40 - lr: 0.000029 - momentum: 0.000000
217
+ 2023-10-12 14:06:43,526 epoch 9 - iter 990/1984 - loss 0.01151383 - time (sec): 285.03 - samples/sec: 290.06 - lr: 0.000027 - momentum: 0.000000
218
+ 2023-10-12 14:07:40,132 epoch 9 - iter 1188/1984 - loss 0.01094490 - time (sec): 341.64 - samples/sec: 289.99 - lr: 0.000025 - momentum: 0.000000
219
+ 2023-10-12 14:08:35,756 epoch 9 - iter 1386/1984 - loss 0.01030513 - time (sec): 397.26 - samples/sec: 291.31 - lr: 0.000023 - momentum: 0.000000
220
+ 2023-10-12 14:09:31,788 epoch 9 - iter 1584/1984 - loss 0.01081881 - time (sec): 453.29 - samples/sec: 289.50 - lr: 0.000021 - momentum: 0.000000
221
+ 2023-10-12 14:10:30,471 epoch 9 - iter 1782/1984 - loss 0.01121344 - time (sec): 511.98 - samples/sec: 287.80 - lr: 0.000020 - momentum: 0.000000
222
+ 2023-10-12 14:11:25,942 epoch 9 - iter 1980/1984 - loss 0.01074014 - time (sec): 567.45 - samples/sec: 288.38 - lr: 0.000018 - momentum: 0.000000
223
+ 2023-10-12 14:11:27,169 ----------------------------------------------------------------------------------------------------
224
+ 2023-10-12 14:11:27,169 EPOCH 9 done: loss 0.0107 - lr: 0.000018
225
+ 2023-10-12 14:11:55,640 DEV : loss 0.22689764201641083 - f1-score (micro avg) 0.7614
226
+ 2023-10-12 14:11:55,690 ----------------------------------------------------------------------------------------------------
227
+ 2023-10-12 14:12:53,298 epoch 10 - iter 198/1984 - loss 0.00665429 - time (sec): 57.61 - samples/sec: 289.71 - lr: 0.000016 - momentum: 0.000000
228
+ 2023-10-12 14:13:51,316 epoch 10 - iter 396/1984 - loss 0.00646911 - time (sec): 115.62 - samples/sec: 283.40 - lr: 0.000014 - momentum: 0.000000
229
+ 2023-10-12 14:14:49,815 epoch 10 - iter 594/1984 - loss 0.00659430 - time (sec): 174.12 - samples/sec: 283.69 - lr: 0.000013 - momentum: 0.000000
230
+ 2023-10-12 14:15:46,196 epoch 10 - iter 792/1984 - loss 0.00696637 - time (sec): 230.50 - samples/sec: 286.08 - lr: 0.000011 - momentum: 0.000000
231
+ 2023-10-12 14:16:42,436 epoch 10 - iter 990/1984 - loss 0.00672449 - time (sec): 286.74 - samples/sec: 287.82 - lr: 0.000009 - momentum: 0.000000
232
+ 2023-10-12 14:17:38,147 epoch 10 - iter 1188/1984 - loss 0.00643333 - time (sec): 342.45 - samples/sec: 287.35 - lr: 0.000007 - momentum: 0.000000
233
+ 2023-10-12 14:18:34,399 epoch 10 - iter 1386/1984 - loss 0.00719725 - time (sec): 398.71 - samples/sec: 286.32 - lr: 0.000005 - momentum: 0.000000
234
+ 2023-10-12 14:19:30,932 epoch 10 - iter 1584/1984 - loss 0.00759850 - time (sec): 455.24 - samples/sec: 286.89 - lr: 0.000004 - momentum: 0.000000
235
+ 2023-10-12 14:20:26,329 epoch 10 - iter 1782/1984 - loss 0.00756874 - time (sec): 510.64 - samples/sec: 288.34 - lr: 0.000002 - momentum: 0.000000
236
+ 2023-10-12 14:21:23,506 epoch 10 - iter 1980/1984 - loss 0.00842761 - time (sec): 567.81 - samples/sec: 288.42 - lr: 0.000000 - momentum: 0.000000
237
+ 2023-10-12 14:21:24,509 ----------------------------------------------------------------------------------------------------
238
+ 2023-10-12 14:21:24,509 EPOCH 10 done: loss 0.0085 - lr: 0.000000
239
+ 2023-10-12 14:21:51,035 DEV : loss 0.2328162044286728 - f1-score (micro avg) 0.7578
240
+ 2023-10-12 14:21:52,053 ----------------------------------------------------------------------------------------------------
241
+ 2023-10-12 14:21:52,056 Loading model from best epoch ...
242
+ 2023-10-12 14:21:56,378 SequenceTagger predicts: Dictionary with 13 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG
243
+ 2023-10-12 14:22:22,594
244
+ Results:
245
+ - F-score (micro) 0.7368
246
+ - F-score (macro) 0.6608
247
+ - Accuracy 0.6204
248
+
249
+ By class:
250
+ precision recall f1-score support
251
+
252
+ LOC 0.8380 0.7817 0.8088 655
253
+ PER 0.6341 0.8161 0.7137 223
254
+ ORG 0.4125 0.5197 0.4599 127
255
+
256
+ micro avg 0.7183 0.7562 0.7368 1005
257
+ macro avg 0.6282 0.7058 0.6608 1005
258
+ weighted avg 0.7390 0.7562 0.7436 1005
259
+
260
+ 2023-10-12 14:22:22,594 ----------------------------------------------------------------------------------------------------