Hugging Face's logo

stefan-it
/
hmbench-newseye-de-hmbyt5-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-2

Token Classification
Flair
PyTorch
TensorBoard
German
sequence-tagger-model
Model card Files Metrics Community
2
stefan-it commited on
Commit
f1ead6f
·
1 Parent(s): f43e2b8

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.1696947686.6d4c7681f95b.1253.5 +3 -0
  6. test.tsv +0 -0
  7. training.log +265 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a9bcdda6d76b12492d020ddd1a41b83d986b6d7cefa39dfaa06b48640fefa2b
3
+ size 870817519
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 sha256:54556bdc9d4caf7fa07908f8d9ff7e8415cb2040ea0c89c9bc94c18412b0af5b
3
+ size 870817636
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 14:45:51 0.0002 0.9036 0.1515 0.1719 0.3617 0.2331 0.1322
3
+ 2 15:10:10 0.0001 0.1490 0.1195 0.2659 0.4129 0.3234 0.1936
4
+ 3 15:34:30 0.0001 0.0905 0.1802 0.2708 0.5189 0.3558 0.2171
5
+ 4 15:57:41 0.0001 0.0650 0.2246 0.2708 0.5303 0.3585 0.2186
6
+ 5 16:21:27 0.0001 0.0453 0.3552 0.2472 0.5398 0.3391 0.2052
7
+ 6 16:45:00 0.0001 0.0336 0.3477 0.2933 0.5417 0.3806 0.2360
8
+ 7 17:07:04 0.0001 0.0242 0.3891 0.2914 0.5833 0.3886 0.2419
9
+ 8 17:29:14 0.0000 0.0185 0.4152 0.3009 0.5625 0.3921 0.2450
10
+ 9 17:51:27 0.0000 0.0125 0.4364 0.2928 0.5739 0.3877 0.2418
11
+ 10 18:14:05 0.0000 0.0087 0.4859 0.2820 0.5890 0.3814 0.2367
runs/events.out.tfevents.1696947686.6d4c7681f95b.1253.5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84b9ea4d4d59394461b0c0a63c3721de08dde9f4033ef384e1aeeeac90da85e6
3
+ size 1464420
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-10 14:21:26,577 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-10 14:21:26,580 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=17, bias=True)
68
+ (loss_function): CrossEntropyLoss()
69
+ )"
70
+ 2023-10-10 14:21:26,580 ----------------------------------------------------------------------------------------------------
71
+ 2023-10-10 14:21:26,580 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
72
+ - NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
73
+ 2023-10-10 14:21:26,580 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-10 14:21:26,580 Train: 20847 sentences
75
+ 2023-10-10 14:21:26,581 (train_with_dev=False, train_with_test=False)
76
+ 2023-10-10 14:21:26,581 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-10 14:21:26,581 Training Params:
78
+ 2023-10-10 14:21:26,581 - learning_rate: "0.00016"
79
+ 2023-10-10 14:21:26,581 - mini_batch_size: "8"
80
+ 2023-10-10 14:21:26,581 - max_epochs: "10"
81
+ 2023-10-10 14:21:26,581 - shuffle: "True"
82
+ 2023-10-10 14:21:26,581 ----------------------------------------------------------------------------------------------------
83
+ 2023-10-10 14:21:26,581 Plugins:
84
+ 2023-10-10 14:21:26,581 - TensorboardLogger
85
+ 2023-10-10 14:21:26,581 - LinearScheduler | warmup_fraction: '0.1'
86
+ 2023-10-10 14:21:26,581 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-10 14:21:26,581 Final evaluation on model from best epoch (best-model.pt)
88
+ 2023-10-10 14:21:26,582 - metric: "('micro avg', 'f1-score')"
89
+ 2023-10-10 14:21:26,582 ----------------------------------------------------------------------------------------------------
90
+ 2023-10-10 14:21:26,582 Computation:
91
+ 2023-10-10 14:21:26,582 - compute on device: cuda:0
92
+ 2023-10-10 14:21:26,582 - embedding storage: none
93
+ 2023-10-10 14:21:26,582 ----------------------------------------------------------------------------------------------------
94
+ 2023-10-10 14:21:26,582 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-2"
95
+ 2023-10-10 14:21:26,582 ----------------------------------------------------------------------------------------------------
96
+ 2023-10-10 14:21:26,582 ----------------------------------------------------------------------------------------------------
97
+ 2023-10-10 14:21:26,582 Logging anything other than scalars to TensorBoard is currently not supported.
98
+ 2023-10-10 14:23:45,008 epoch 1 - iter 260/2606 - loss 2.82650621 - time (sec): 138.42 - samples/sec: 258.00 - lr: 0.000016 - momentum: 0.000000
99
+ 2023-10-10 14:26:05,289 epoch 1 - iter 520/2606 - loss 2.56314673 - time (sec): 278.70 - samples/sec: 258.20 - lr: 0.000032 - momentum: 0.000000
100
+ 2023-10-10 14:28:31,209 epoch 1 - iter 780/2606 - loss 2.14030820 - time (sec): 424.62 - samples/sec: 257.28 - lr: 0.000048 - momentum: 0.000000
101
+ 2023-10-10 14:30:56,830 epoch 1 - iter 1040/2606 - loss 1.73439842 - time (sec): 570.25 - samples/sec: 260.41 - lr: 0.000064 - momentum: 0.000000
102
+ 2023-10-10 14:33:17,157 epoch 1 - iter 1300/2606 - loss 1.47736320 - time (sec): 710.57 - samples/sec: 260.61 - lr: 0.000080 - momentum: 0.000000
103
+ 2023-10-10 14:35:44,796 epoch 1 - iter 1560/2606 - loss 1.30609247 - time (sec): 858.21 - samples/sec: 260.09 - lr: 0.000096 - momentum: 0.000000
104
+ 2023-10-10 14:38:04,201 epoch 1 - iter 1820/2606 - loss 1.17487372 - time (sec): 997.62 - samples/sec: 260.31 - lr: 0.000112 - momentum: 0.000000
105
+ 2023-10-10 14:40:25,567 epoch 1 - iter 2080/2606 - loss 1.06691289 - time (sec): 1138.98 - samples/sec: 259.14 - lr: 0.000128 - momentum: 0.000000
106
+ 2023-10-10 14:42:44,023 epoch 1 - iter 2340/2606 - loss 0.98460797 - time (sec): 1277.44 - samples/sec: 257.56 - lr: 0.000144 - momentum: 0.000000
107
+ 2023-10-10 14:45:09,215 epoch 1 - iter 2600/2606 - loss 0.90554459 - time (sec): 1422.63 - samples/sec: 257.48 - lr: 0.000160 - momentum: 0.000000
108
+ 2023-10-10 14:45:12,600 ----------------------------------------------------------------------------------------------------
109
+ 2023-10-10 14:45:12,601 EPOCH 1 done: loss 0.9036 - lr: 0.000160
110
+ 2023-10-10 14:45:51,723 DEV : loss 0.15153414011001587 - f1-score (micro avg) 0.2331
111
+ 2023-10-10 14:45:51,776 saving best model
112
+ 2023-10-10 14:45:52,788 ----------------------------------------------------------------------------------------------------
113
+ 2023-10-10 14:48:11,787 epoch 2 - iter 260/2606 - loss 0.18842284 - time (sec): 139.00 - samples/sec: 251.93 - lr: 0.000158 - momentum: 0.000000
114
+ 2023-10-10 14:50:30,954 epoch 2 - iter 520/2606 - loss 0.18792004 - time (sec): 278.16 - samples/sec: 254.84 - lr: 0.000156 - momentum: 0.000000
115
+ 2023-10-10 14:52:50,250 epoch 2 - iter 780/2606 - loss 0.17880044 - time (sec): 417.46 - samples/sec: 257.67 - lr: 0.000155 - momentum: 0.000000
116
+ 2023-10-10 14:55:12,932 epoch 2 - iter 1040/2606 - loss 0.17379042 - time (sec): 560.14 - samples/sec: 254.57 - lr: 0.000153 - momentum: 0.000000
117
+ 2023-10-10 14:57:36,261 epoch 2 - iter 1300/2606 - loss 0.16829776 - time (sec): 703.47 - samples/sec: 255.54 - lr: 0.000151 - momentum: 0.000000
118
+ 2023-10-10 14:59:58,439 epoch 2 - iter 1560/2606 - loss 0.16028935 - time (sec): 845.65 - samples/sec: 257.10 - lr: 0.000149 - momentum: 0.000000
119
+ 2023-10-10 15:02:21,827 epoch 2 - iter 1820/2606 - loss 0.15744925 - time (sec): 989.04 - samples/sec: 257.84 - lr: 0.000148 - momentum: 0.000000
120
+ 2023-10-10 15:04:42,614 epoch 2 - iter 2080/2606 - loss 0.15613165 - time (sec): 1129.82 - samples/sec: 257.27 - lr: 0.000146 - momentum: 0.000000
121
+ 2023-10-10 15:07:03,205 epoch 2 - iter 2340/2606 - loss 0.15275373 - time (sec): 1270.42 - samples/sec: 256.73 - lr: 0.000144 - momentum: 0.000000
122
+ 2023-10-10 15:09:24,020 epoch 2 - iter 2600/2606 - loss 0.14918224 - time (sec): 1411.23 - samples/sec: 259.85 - lr: 0.000142 - momentum: 0.000000
123
+ 2023-10-10 15:09:26,918 ----------------------------------------------------------------------------------------------------
124
+ 2023-10-10 15:09:26,919 EPOCH 2 done: loss 0.1490 - lr: 0.000142
125
+ 2023-10-10 15:10:10,127 DEV : loss 0.11950229853391647 - f1-score (micro avg) 0.3234
126
+ 2023-10-10 15:10:10,187 saving best model
127
+ 2023-10-10 15:10:12,879 ----------------------------------------------------------------------------------------------------
128
+ 2023-10-10 15:12:28,827 epoch 3 - iter 260/2606 - loss 0.08047863 - time (sec): 135.94 - samples/sec: 261.72 - lr: 0.000140 - momentum: 0.000000
129
+ 2023-10-10 15:14:47,287 epoch 3 - iter 520/2606 - loss 0.08774055 - time (sec): 274.40 - samples/sec: 264.43 - lr: 0.000139 - momentum: 0.000000
130
+ 2023-10-10 15:17:07,350 epoch 3 - iter 780/2606 - loss 0.08904143 - time (sec): 414.47 - samples/sec: 264.51 - lr: 0.000137 - momentum: 0.000000
131
+ 2023-10-10 15:19:28,398 epoch 3 - iter 1040/2606 - loss 0.09225516 - time (sec): 555.51 - samples/sec: 261.92 - lr: 0.000135 - momentum: 0.000000
132
+ 2023-10-10 15:21:56,150 epoch 3 - iter 1300/2606 - loss 0.09497422 - time (sec): 703.27 - samples/sec: 263.70 - lr: 0.000133 - momentum: 0.000000
133
+ 2023-10-10 15:24:21,771 epoch 3 - iter 1560/2606 - loss 0.09353639 - time (sec): 848.89 - samples/sec: 262.80 - lr: 0.000132 - momentum: 0.000000
134
+ 2023-10-10 15:26:45,547 epoch 3 - iter 1820/2606 - loss 0.09255565 - time (sec): 992.66 - samples/sec: 261.67 - lr: 0.000130 - momentum: 0.000000
135
+ 2023-10-10 15:29:00,939 epoch 3 - iter 2080/2606 - loss 0.09167665 - time (sec): 1128.06 - samples/sec: 262.47 - lr: 0.000128 - momentum: 0.000000
136
+ 2023-10-10 15:31:19,016 epoch 3 - iter 2340/2606 - loss 0.09077967 - time (sec): 1266.13 - samples/sec: 259.91 - lr: 0.000126 - momentum: 0.000000
137
+ 2023-10-10 15:33:46,544 epoch 3 - iter 2600/2606 - loss 0.09046288 - time (sec): 1413.66 - samples/sec: 259.30 - lr: 0.000125 - momentum: 0.000000
138
+ 2023-10-10 15:33:49,679 ----------------------------------------------------------------------------------------------------
139
+ 2023-10-10 15:33:49,679 EPOCH 3 done: loss 0.0905 - lr: 0.000125
140
+ 2023-10-10 15:34:30,739 DEV : loss 0.18015694618225098 - f1-score (micro avg) 0.3558
141
+ 2023-10-10 15:34:30,790 saving best model
142
+ 2023-10-10 15:34:33,476 ----------------------------------------------------------------------------------------------------
143
+ 2023-10-10 15:36:51,151 epoch 4 - iter 260/2606 - loss 0.05157111 - time (sec): 137.67 - samples/sec: 267.59 - lr: 0.000123 - momentum: 0.000000
144
+ 2023-10-10 15:39:11,379 epoch 4 - iter 520/2606 - loss 0.06164379 - time (sec): 277.90 - samples/sec: 274.53 - lr: 0.000121 - momentum: 0.000000
145
+ 2023-10-10 15:41:28,765 epoch 4 - iter 780/2606 - loss 0.06041382 - time (sec): 415.28 - samples/sec: 271.33 - lr: 0.000119 - momentum: 0.000000
146
+ 2023-10-10 15:43:44,655 epoch 4 - iter 1040/2606 - loss 0.06191568 - time (sec): 551.17 - samples/sec: 269.10 - lr: 0.000117 - momentum: 0.000000
147
+ 2023-10-10 15:46:05,001 epoch 4 - iter 1300/2606 - loss 0.06056013 - time (sec): 691.52 - samples/sec: 268.16 - lr: 0.000116 - momentum: 0.000000
148
+ 2023-10-10 15:48:16,573 epoch 4 - iter 1560/2606 - loss 0.06351392 - time (sec): 823.09 - samples/sec: 269.95 - lr: 0.000114 - momentum: 0.000000
149
+ 2023-10-10 15:50:29,084 epoch 4 - iter 1820/2606 - loss 0.06300262 - time (sec): 955.60 - samples/sec: 272.07 - lr: 0.000112 - momentum: 0.000000
150
+ 2023-10-10 15:52:38,568 epoch 4 - iter 2080/2606 - loss 0.06312479 - time (sec): 1085.09 - samples/sec: 271.77 - lr: 0.000110 - momentum: 0.000000
151
+ 2023-10-10 15:54:50,406 epoch 4 - iter 2340/2606 - loss 0.06500771 - time (sec): 1216.93 - samples/sec: 272.23 - lr: 0.000109 - momentum: 0.000000
152
+ 2023-10-10 15:56:59,946 epoch 4 - iter 2600/2606 - loss 0.06498073 - time (sec): 1346.47 - samples/sec: 272.46 - lr: 0.000107 - momentum: 0.000000
153
+ 2023-10-10 15:57:02,689 ----------------------------------------------------------------------------------------------------
154
+ 2023-10-10 15:57:02,690 EPOCH 4 done: loss 0.0650 - lr: 0.000107
155
+ 2023-10-10 15:57:41,615 DEV : loss 0.22460463643074036 - f1-score (micro avg) 0.3585
156
+ 2023-10-10 15:57:41,667 saving best model
157
+ 2023-10-10 15:57:44,347 ----------------------------------------------------------------------------------------------------
158
+ 2023-10-10 15:59:52,262 epoch 5 - iter 260/2606 - loss 0.03592166 - time (sec): 127.91 - samples/sec: 268.55 - lr: 0.000105 - momentum: 0.000000
159
+ 2023-10-10 16:02:02,554 epoch 5 - iter 520/2606 - loss 0.04034587 - time (sec): 258.20 - samples/sec: 272.50 - lr: 0.000103 - momentum: 0.000000
160
+ 2023-10-10 16:04:14,414 epoch 5 - iter 780/2606 - loss 0.03986586 - time (sec): 390.06 - samples/sec: 278.42 - lr: 0.000101 - momentum: 0.000000
161
+ 2023-10-10 16:06:26,506 epoch 5 - iter 1040/2606 - loss 0.04078098 - time (sec): 522.15 - samples/sec: 277.65 - lr: 0.000100 - momentum: 0.000000
162
+ 2023-10-10 16:08:39,982 epoch 5 - iter 1300/2606 - loss 0.04342342 - time (sec): 655.63 - samples/sec: 278.05 - lr: 0.000098 - momentum: 0.000000
163
+ 2023-10-10 16:10:56,870 epoch 5 - iter 1560/2606 - loss 0.04386466 - time (sec): 792.52 - samples/sec: 276.41 - lr: 0.000096 - momentum: 0.000000
164
+ 2023-10-10 16:13:22,636 epoch 5 - iter 1820/2606 - loss 0.04404192 - time (sec): 938.28 - samples/sec: 271.26 - lr: 0.000094 - momentum: 0.000000
165
+ 2023-10-10 16:15:50,560 epoch 5 - iter 2080/2606 - loss 0.04511901 - time (sec): 1086.21 - samples/sec: 269.90 - lr: 0.000093 - momentum: 0.000000
166
+ 2023-10-10 16:18:17,914 epoch 5 - iter 2340/2606 - loss 0.04514489 - time (sec): 1233.56 - samples/sec: 268.76 - lr: 0.000091 - momentum: 0.000000
167
+ 2023-10-10 16:20:38,307 epoch 5 - iter 2600/2606 - loss 0.04534151 - time (sec): 1373.95 - samples/sec: 266.91 - lr: 0.000089 - momentum: 0.000000
168
+ 2023-10-10 16:20:41,341 ----------------------------------------------------------------------------------------------------
169
+ 2023-10-10 16:20:41,341 EPOCH 5 done: loss 0.0453 - lr: 0.000089
170
+ 2023-10-10 16:21:26,960 DEV : loss 0.3552384078502655 - f1-score (micro avg) 0.3391
171
+ 2023-10-10 16:21:27,020 ----------------------------------------------------------------------------------------------------
172
+ 2023-10-10 16:23:46,309 epoch 6 - iter 260/2606 - loss 0.03073255 - time (sec): 139.29 - samples/sec: 251.20 - lr: 0.000087 - momentum: 0.000000
173
+ 2023-10-10 16:26:06,409 epoch 6 - iter 520/2606 - loss 0.03336276 - time (sec): 279.39 - samples/sec: 250.59 - lr: 0.000085 - momentum: 0.000000
174
+ 2023-10-10 16:28:31,218 epoch 6 - iter 780/2606 - loss 0.03342619 - time (sec): 424.20 - samples/sec: 255.01 - lr: 0.000084 - momentum: 0.000000
175
+ 2023-10-10 16:30:56,809 epoch 6 - iter 1040/2606 - loss 0.03305085 - time (sec): 569.79 - samples/sec: 257.41 - lr: 0.000082 - momentum: 0.000000
176
+ 2023-10-10 16:33:15,628 epoch 6 - iter 1300/2606 - loss 0.03234751 - time (sec): 708.61 - samples/sec: 261.61 - lr: 0.000080 - momentum: 0.000000
177
+ 2023-10-10 16:35:27,531 epoch 6 - iter 1560/2606 - loss 0.03383010 - time (sec): 840.51 - samples/sec: 263.10 - lr: 0.000078 - momentum: 0.000000
178
+ 2023-10-10 16:37:38,652 epoch 6 - iter 1820/2606 - loss 0.03360923 - time (sec): 971.63 - samples/sec: 265.18 - lr: 0.000077 - momentum: 0.000000
179
+ 2023-10-10 16:39:50,837 epoch 6 - iter 2080/2606 - loss 0.03354235 - time (sec): 1103.82 - samples/sec: 267.05 - lr: 0.000075 - momentum: 0.000000
180
+ 2023-10-10 16:42:00,685 epoch 6 - iter 2340/2606 - loss 0.03417729 - time (sec): 1233.66 - samples/sec: 266.96 - lr: 0.000073 - momentum: 0.000000
181
+ 2023-10-10 16:44:19,012 epoch 6 - iter 2600/2606 - loss 0.03364077 - time (sec): 1371.99 - samples/sec: 267.28 - lr: 0.000071 - momentum: 0.000000
182
+ 2023-10-10 16:44:21,887 ----------------------------------------------------------------------------------------------------
183
+ 2023-10-10 16:44:21,888 EPOCH 6 done: loss 0.0336 - lr: 0.000071
184
+ 2023-10-10 16:45:00,806 DEV : loss 0.34769406914711 - f1-score (micro avg) 0.3806
185
+ 2023-10-10 16:45:00,863 saving best model
186
+ 2023-10-10 16:45:03,552 ----------------------------------------------------------------------------------------------------
187
+ 2023-10-10 16:47:10,382 epoch 7 - iter 260/2606 - loss 0.02507245 - time (sec): 126.82 - samples/sec: 279.71 - lr: 0.000069 - momentum: 0.000000
188
+ 2023-10-10 16:49:17,745 epoch 7 - iter 520/2606 - loss 0.02198476 - time (sec): 254.19 - samples/sec: 280.93 - lr: 0.000068 - momentum: 0.000000
189
+ 2023-10-10 16:51:24,652 epoch 7 - iter 780/2606 - loss 0.02230588 - time (sec): 381.09 - samples/sec: 283.27 - lr: 0.000066 - momentum: 0.000000
190
+ 2023-10-10 16:53:32,924 epoch 7 - iter 1040/2606 - loss 0.02469517 - time (sec): 509.37 - samples/sec: 283.41 - lr: 0.000064 - momentum: 0.000000
191
+ 2023-10-10 16:55:41,775 epoch 7 - iter 1300/2606 - loss 0.02514914 - time (sec): 638.22 - samples/sec: 286.11 - lr: 0.000062 - momentum: 0.000000
192
+ 2023-10-10 16:57:48,643 epoch 7 - iter 1560/2606 - loss 0.02586490 - time (sec): 765.09 - samples/sec: 286.23 - lr: 0.000061 - momentum: 0.000000
193
+ 2023-10-10 16:59:55,786 epoch 7 - iter 1820/2606 - loss 0.02506821 - time (sec): 892.23 - samples/sec: 286.32 - lr: 0.000059 - momentum: 0.000000
194
+ 2023-10-10 17:02:01,919 epoch 7 - iter 2080/2606 - loss 0.02485056 - time (sec): 1018.36 - samples/sec: 284.07 - lr: 0.000057 - momentum: 0.000000
195
+ 2023-10-10 17:04:11,098 epoch 7 - iter 2340/2606 - loss 0.02484702 - time (sec): 1147.54 - samples/sec: 284.94 - lr: 0.000055 - momentum: 0.000000
196
+ 2023-10-10 17:06:22,781 epoch 7 - iter 2600/2606 - loss 0.02416949 - time (sec): 1279.22 - samples/sec: 286.55 - lr: 0.000053 - momentum: 0.000000
197
+ 2023-10-10 17:06:25,669 ----------------------------------------------------------------------------------------------------
198
+ 2023-10-10 17:06:25,669 EPOCH 7 done: loss 0.0242 - lr: 0.000053
199
+ 2023-10-10 17:07:04,419 DEV : loss 0.38912469148635864 - f1-score (micro avg) 0.3886
200
+ 2023-10-10 17:07:04,492 saving best model
201
+ 2023-10-10 17:07:08,095 ----------------------------------------------------------------------------------------------------
202
+ 2023-10-10 17:09:19,922 epoch 8 - iter 260/2606 - loss 0.01590417 - time (sec): 131.82 - samples/sec: 302.15 - lr: 0.000052 - momentum: 0.000000
203
+ 2023-10-10 17:11:31,107 epoch 8 - iter 520/2606 - loss 0.01657268 - time (sec): 263.01 - samples/sec: 292.08 - lr: 0.000050 - momentum: 0.000000
204
+ 2023-10-10 17:13:38,438 epoch 8 - iter 780/2606 - loss 0.01716544 - time (sec): 390.34 - samples/sec: 287.05 - lr: 0.000048 - momentum: 0.000000
205
+ 2023-10-10 17:15:47,199 epoch 8 - iter 1040/2606 - loss 0.01681174 - time (sec): 519.10 - samples/sec: 288.07 - lr: 0.000046 - momentum: 0.000000
206
+ 2023-10-10 17:17:54,730 epoch 8 - iter 1300/2606 - loss 0.01634833 - time (sec): 646.63 - samples/sec: 286.35 - lr: 0.000045 - momentum: 0.000000
207
+ 2023-10-10 17:20:02,473 epoch 8 - iter 1560/2606 - loss 0.01711598 - time (sec): 774.37 - samples/sec: 287.31 - lr: 0.000043 - momentum: 0.000000
208
+ 2023-10-10 17:22:11,287 epoch 8 - iter 1820/2606 - loss 0.01677195 - time (sec): 903.19 - samples/sec: 286.75 - lr: 0.000041 - momentum: 0.000000
209
+ 2023-10-10 17:24:17,214 epoch 8 - iter 2080/2606 - loss 0.01710973 - time (sec): 1029.11 - samples/sec: 285.16 - lr: 0.000039 - momentum: 0.000000
210
+ 2023-10-10 17:26:22,215 epoch 8 - iter 2340/2606 - loss 0.01765144 - time (sec): 1154.12 - samples/sec: 283.46 - lr: 0.000037 - momentum: 0.000000
211
+ 2023-10-10 17:28:34,006 epoch 8 - iter 2600/2606 - loss 0.01846439 - time (sec): 1285.91 - samples/sec: 285.24 - lr: 0.000036 - momentum: 0.000000
212
+ 2023-10-10 17:28:36,682 ----------------------------------------------------------------------------------------------------
213
+ 2023-10-10 17:28:36,682 EPOCH 8 done: loss 0.0185 - lr: 0.000036
214
+ 2023-10-10 17:29:14,536 DEV : loss 0.41515445709228516 - f1-score (micro avg) 0.3921
215
+ 2023-10-10 17:29:14,587 saving best model
216
+ 2023-10-10 17:29:17,244 ----------------------------------------------------------------------------------------------------
217
+ 2023-10-10 17:31:26,353 epoch 9 - iter 260/2606 - loss 0.01719462 - time (sec): 129.10 - samples/sec: 286.82 - lr: 0.000034 - momentum: 0.000000
218
+ 2023-10-10 17:33:37,943 epoch 9 - iter 520/2606 - loss 0.01306206 - time (sec): 260.69 - samples/sec: 296.75 - lr: 0.000032 - momentum: 0.000000
219
+ 2023-10-10 17:35:45,408 epoch 9 - iter 780/2606 - loss 0.01249072 - time (sec): 388.15 - samples/sec: 291.69 - lr: 0.000030 - momentum: 0.000000
220
+ 2023-10-10 17:37:50,026 epoch 9 - iter 1040/2606 - loss 0.01153243 - time (sec): 512.77 - samples/sec: 286.00 - lr: 0.000029 - momentum: 0.000000
221
+ 2023-10-10 17:40:00,767 epoch 9 - iter 1300/2606 - loss 0.01176972 - time (sec): 643.51 - samples/sec: 288.25 - lr: 0.000027 - momentum: 0.000000
222
+ 2023-10-10 17:42:06,166 epoch 9 - iter 1560/2606 - loss 0.01199337 - time (sec): 768.91 - samples/sec: 286.24 - lr: 0.000025 - momentum: 0.000000
223
+ 2023-10-10 17:44:13,171 epoch 9 - iter 1820/2606 - loss 0.01190054 - time (sec): 895.91 - samples/sec: 284.89 - lr: 0.000023 - momentum: 0.000000
224
+ 2023-10-10 17:46:21,808 epoch 9 - iter 2080/2606 - loss 0.01247403 - time (sec): 1024.55 - samples/sec: 284.48 - lr: 0.000021 - momentum: 0.000000
225
+ 2023-10-10 17:48:30,521 epoch 9 - iter 2340/2606 - loss 0.01213440 - time (sec): 1153.26 - samples/sec: 284.01 - lr: 0.000020 - momentum: 0.000000
226
+ 2023-10-10 17:50:42,947 epoch 9 - iter 2600/2606 - loss 0.01250624 - time (sec): 1285.69 - samples/sec: 285.36 - lr: 0.000018 - momentum: 0.000000
227
+ 2023-10-10 17:50:45,623 ----------------------------------------------------------------------------------------------------
228
+ 2023-10-10 17:50:45,624 EPOCH 9 done: loss 0.0125 - lr: 0.000018
229
+ 2023-10-10 17:51:27,746 DEV : loss 0.4364360272884369 - f1-score (micro avg) 0.3877
230
+ 2023-10-10 17:51:27,812 ----------------------------------------------------------------------------------------------------
231
+ 2023-10-10 17:53:43,933 epoch 10 - iter 260/2606 - loss 0.01035287 - time (sec): 136.12 - samples/sec: 280.82 - lr: 0.000016 - momentum: 0.000000
232
+ 2023-10-10 17:55:59,905 epoch 10 - iter 520/2606 - loss 0.00933623 - time (sec): 272.09 - samples/sec: 281.96 - lr: 0.000014 - momentum: 0.000000
233
+ 2023-10-10 17:58:14,199 epoch 10 - iter 780/2606 - loss 0.00911268 - time (sec): 406.38 - samples/sec: 276.91 - lr: 0.000013 - momentum: 0.000000
234
+ 2023-10-10 18:00:25,139 epoch 10 - iter 1040/2606 - loss 0.00886118 - time (sec): 537.32 - samples/sec: 270.22 - lr: 0.000011 - momentum: 0.000000
235
+ 2023-10-10 18:02:36,223 epoch 10 - iter 1300/2606 - loss 0.00858393 - time (sec): 668.41 - samples/sec: 273.11 - lr: 0.000009 - momentum: 0.000000
236
+ 2023-10-10 18:04:45,180 epoch 10 - iter 1560/2606 - loss 0.00837399 - time (sec): 797.37 - samples/sec: 274.08 - lr: 0.000007 - momentum: 0.000000
237
+ 2023-10-10 18:06:54,526 epoch 10 - iter 1820/2606 - loss 0.00838034 - time (sec): 926.71 - samples/sec: 275.86 - lr: 0.000005 - momentum: 0.000000
238
+ 2023-10-10 18:09:05,520 epoch 10 - iter 2080/2606 - loss 0.00862970 - time (sec): 1057.71 - samples/sec: 278.71 - lr: 0.000004 - momentum: 0.000000
239
+ 2023-10-10 18:11:13,645 epoch 10 - iter 2340/2606 - loss 0.00858216 - time (sec): 1185.83 - samples/sec: 278.84 - lr: 0.000002 - momentum: 0.000000
240
+ 2023-10-10 18:13:21,604 epoch 10 - iter 2600/2606 - loss 0.00867931 - time (sec): 1313.79 - samples/sec: 278.96 - lr: 0.000000 - momentum: 0.000000
241
+ 2023-10-10 18:13:24,617 ----------------------------------------------------------------------------------------------------
242
+ 2023-10-10 18:13:24,618 EPOCH 10 done: loss 0.0087 - lr: 0.000000
243
+ 2023-10-10 18:14:05,757 DEV : loss 0.48594337701797485 - f1-score (micro avg) 0.3814
244
+ 2023-10-10 18:14:06,788 ----------------------------------------------------------------------------------------------------
245
+ 2023-10-10 18:14:06,791 Loading model from best epoch ...
246
+ 2023-10-10 18:14:10,833 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
247
+ 2023-10-10 18:15:46,590
248
+ Results:
249
+ - F-score (micro) 0.4343
250
+ - F-score (macro) 0.3054
251
+ - Accuracy 0.2807
252
+
253
+ By class:
254
+ precision recall f1-score support
255
+
256
+ LOC 0.4638 0.4901 0.4766 1214
257
+ PER 0.4027 0.4480 0.4241 808
258
+ ORG 0.3274 0.3144 0.3208 353
259
+ HumanProd 0.0000 0.0000 0.0000 15
260
+
261
+ micro avg 0.4225 0.4469 0.4343 2390
262
+ macro avg 0.2985 0.3131 0.3054 2390
263
+ weighted avg 0.4201 0.4469 0.4328 2390
264
+
265
+ 2023-10-10 18:15:46,590 ----------------------------------------------------------------------------------------------------