Hugging Face's logo

stefan-it
/
hmbench-newseye-de-hmbyt5-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-4

Token Classification
Flair
PyTorch
TensorBoard
German
sequence-tagger-model
Model card Files Metrics Community
2
stefan-it commited on
Commit
57b1f09
·
1 Parent(s): 82b26a6

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.1697077150.6d4c7681f95b.1253.14 +3 -0
  6. test.tsv +0 -0
  7. training.log +264 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d73122cc85eabb2c8e138e02799c3d71b22710b23003ce1c289facbe45ca28bb
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 sha256:531e0026cb5f9bce97aedf428c29f02d57146ae7beaaac9059ca5837f21c5c55
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 02:42:52 0.0001 0.7433 0.1264 0.2155 0.3902 0.2776 0.1612
3
+ 2 03:07:05 0.0001 0.1446 0.1417 0.2918 0.3902 0.3339 0.2010
4
+ 3 03:31:15 0.0001 0.0981 0.2601 0.2530 0.6345 0.3618 0.2224
5
+ 4 03:55:45 0.0001 0.0665 0.3102 0.2560 0.5814 0.3555 0.2171
6
+ 5 04:20:13 0.0001 0.0464 0.3113 0.3113 0.5606 0.4003 0.2519
7
+ 6 04:44:50 0.0001 0.0330 0.4072 0.2902 0.6288 0.3971 0.2494
8
+ 7 05:09:04 0.0001 0.0232 0.4082 0.3015 0.6098 0.4035 0.2543
9
+ 8 05:32:58 0.0000 0.0164 0.4336 0.3006 0.6155 0.4040 0.2543
10
+ 9 05:57:30 0.0000 0.0109 0.4758 0.2941 0.6004 0.3948 0.2475
11
+ 10 06:21:43 0.0000 0.0072 0.4922 0.2891 0.6023 0.3907 0.2441
runs/events.out.tfevents.1697077150.6d4c7681f95b.1253.14 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd916826dcdda4e92192cfe7df8b3fa0cd8f2af49a3635d4fa327312be52fbb2
3
+ size 2923780
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-12 02:19:10,693 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-12 02:19:10,695 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
+ 2023-10-12 02:19:10,696 ----------------------------------------------------------------------------------------------------
71
+ 2023-10-12 02:19:10,696 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
+ 2023-10-12 02:19:10,696 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-12 02:19:10,696 Train: 20847 sentences
75
+ 2023-10-12 02:19:10,696 (train_with_dev=False, train_with_test=False)
76
+ 2023-10-12 02:19:10,696 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-12 02:19:10,696 Training Params:
78
+ 2023-10-12 02:19:10,696 - learning_rate: "0.00015"
79
+ 2023-10-12 02:19:10,696 - mini_batch_size: "4"
80
+ 2023-10-12 02:19:10,696 - max_epochs: "10"
81
+ 2023-10-12 02:19:10,696 - shuffle: "True"
82
+ 2023-10-12 02:19:10,696 ----------------------------------------------------------------------------------------------------
83
+ 2023-10-12 02:19:10,696 Plugins:
84
+ 2023-10-12 02:19:10,697 - TensorboardLogger
85
+ 2023-10-12 02:19:10,697 - LinearScheduler | warmup_fraction: '0.1'
86
+ 2023-10-12 02:19:10,697 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-12 02:19:10,697 Final evaluation on model from best epoch (best-model.pt)
88
+ 2023-10-12 02:19:10,697 - metric: "('micro avg', 'f1-score')"
89
+ 2023-10-12 02:19:10,697 ----------------------------------------------------------------------------------------------------
90
+ 2023-10-12 02:19:10,697 Computation:
91
+ 2023-10-12 02:19:10,697 - compute on device: cuda:0
92
+ 2023-10-12 02:19:10,697 - embedding storage: none
93
+ 2023-10-12 02:19:10,697 ----------------------------------------------------------------------------------------------------
94
+ 2023-10-12 02:19:10,697 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-4"
95
+ 2023-10-12 02:19:10,697 ----------------------------------------------------------------------------------------------------
96
+ 2023-10-12 02:19:10,697 ----------------------------------------------------------------------------------------------------
97
+ 2023-10-12 02:19:10,697 Logging anything other than scalars to TensorBoard is currently not supported.
98
+ 2023-10-12 02:21:27,177 epoch 1 - iter 521/5212 - loss 2.79668095 - time (sec): 136.48 - samples/sec: 242.74 - lr: 0.000015 - momentum: 0.000000
99
+ 2023-10-12 02:23:45,273 epoch 1 - iter 1042/5212 - loss 2.35843978 - time (sec): 274.57 - samples/sec: 247.36 - lr: 0.000030 - momentum: 0.000000
100
+ 2023-10-12 02:26:04,245 epoch 1 - iter 1563/5212 - loss 1.80796410 - time (sec): 413.55 - samples/sec: 254.16 - lr: 0.000045 - momentum: 0.000000
101
+ 2023-10-12 02:28:22,402 epoch 1 - iter 2084/5212 - loss 1.45533862 - time (sec): 551.70 - samples/sec: 257.38 - lr: 0.000060 - momentum: 0.000000
102
+ 2023-10-12 02:30:41,547 epoch 1 - iter 2605/5212 - loss 1.24266578 - time (sec): 690.85 - samples/sec: 260.63 - lr: 0.000075 - momentum: 0.000000
103
+ 2023-10-12 02:32:58,209 epoch 1 - iter 3126/5212 - loss 1.09834367 - time (sec): 827.51 - samples/sec: 259.96 - lr: 0.000090 - momentum: 0.000000
104
+ 2023-10-12 02:35:17,942 epoch 1 - iter 3647/5212 - loss 0.98084623 - time (sec): 967.24 - samples/sec: 261.96 - lr: 0.000105 - momentum: 0.000000
105
+ 2023-10-12 02:37:35,810 epoch 1 - iter 4168/5212 - loss 0.88950826 - time (sec): 1105.11 - samples/sec: 262.06 - lr: 0.000120 - momentum: 0.000000
106
+ 2023-10-12 02:39:57,111 epoch 1 - iter 4689/5212 - loss 0.80746283 - time (sec): 1246.41 - samples/sec: 264.35 - lr: 0.000135 - momentum: 0.000000
107
+ 2023-10-12 02:42:16,837 epoch 1 - iter 5210/5212 - loss 0.74371560 - time (sec): 1386.14 - samples/sec: 264.92 - lr: 0.000150 - momentum: 0.000000
108
+ 2023-10-12 02:42:17,401 ----------------------------------------------------------------------------------------------------
109
+ 2023-10-12 02:42:17,402 EPOCH 1 done: loss 0.7433 - lr: 0.000150
110
+ 2023-10-12 02:42:52,120 DEV : loss 0.12636248767375946 - f1-score (micro avg) 0.2776
111
+ 2023-10-12 02:42:52,174 saving best model
112
+ 2023-10-12 02:42:53,044 ----------------------------------------------------------------------------------------------------
113
+ 2023-10-12 02:45:11,088 epoch 2 - iter 521/5212 - loss 0.17686924 - time (sec): 138.04 - samples/sec: 262.94 - lr: 0.000148 - momentum: 0.000000
114
+ 2023-10-12 02:47:30,960 epoch 2 - iter 1042/5212 - loss 0.15464838 - time (sec): 277.91 - samples/sec: 267.07 - lr: 0.000147 - momentum: 0.000000
115
+ 2023-10-12 02:49:50,501 epoch 2 - iter 1563/5212 - loss 0.15677141 - time (sec): 417.45 - samples/sec: 261.48 - lr: 0.000145 - momentum: 0.000000
116
+ 2023-10-12 02:52:16,593 epoch 2 - iter 2084/5212 - loss 0.15541495 - time (sec): 563.55 - samples/sec: 263.09 - lr: 0.000143 - momentum: 0.000000
117
+ 2023-10-12 02:54:38,805 epoch 2 - iter 2605/5212 - loss 0.15288785 - time (sec): 705.76 - samples/sec: 261.65 - lr: 0.000142 - momentum: 0.000000
118
+ 2023-10-12 02:56:58,256 epoch 2 - iter 3126/5212 - loss 0.15178570 - time (sec): 845.21 - samples/sec: 258.20 - lr: 0.000140 - momentum: 0.000000
119
+ 2023-10-12 02:59:16,224 epoch 2 - iter 3647/5212 - loss 0.15320470 - time (sec): 983.18 - samples/sec: 254.99 - lr: 0.000138 - momentum: 0.000000
120
+ 2023-10-12 03:01:39,472 epoch 2 - iter 4168/5212 - loss 0.14998413 - time (sec): 1126.43 - samples/sec: 256.63 - lr: 0.000137 - momentum: 0.000000
121
+ 2023-10-12 03:04:04,945 epoch 2 - iter 4689/5212 - loss 0.14664370 - time (sec): 1271.90 - samples/sec: 259.73 - lr: 0.000135 - momentum: 0.000000
122
+ 2023-10-12 03:06:26,106 epoch 2 - iter 5210/5212 - loss 0.14460575 - time (sec): 1413.06 - samples/sec: 259.97 - lr: 0.000133 - momentum: 0.000000
123
+ 2023-10-12 03:06:26,551 ----------------------------------------------------------------------------------------------------
124
+ 2023-10-12 03:06:26,552 EPOCH 2 done: loss 0.1446 - lr: 0.000133
125
+ 2023-10-12 03:07:05,601 DEV : loss 0.14167223870754242 - f1-score (micro avg) 0.3339
126
+ 2023-10-12 03:07:05,653 saving best model
127
+ 2023-10-12 03:07:08,265 ----------------------------------------------------------------------------------------------------
128
+ 2023-10-12 03:09:23,808 epoch 3 - iter 521/5212 - loss 0.10082822 - time (sec): 135.54 - samples/sec: 254.81 - lr: 0.000132 - momentum: 0.000000
129
+ 2023-10-12 03:11:37,979 epoch 3 - iter 1042/5212 - loss 0.09748085 - time (sec): 269.71 - samples/sec: 250.08 - lr: 0.000130 - momentum: 0.000000
130
+ 2023-10-12 03:13:58,974 epoch 3 - iter 1563/5212 - loss 0.09985524 - time (sec): 410.70 - samples/sec: 264.24 - lr: 0.000128 - momentum: 0.000000
131
+ 2023-10-12 03:16:15,894 epoch 3 - iter 2084/5212 - loss 0.09997962 - time (sec): 547.62 - samples/sec: 263.03 - lr: 0.000127 - momentum: 0.000000
132
+ 2023-10-12 03:18:33,733 epoch 3 - iter 2605/5212 - loss 0.09885177 - time (sec): 685.46 - samples/sec: 261.01 - lr: 0.000125 - momentum: 0.000000
133
+ 2023-10-12 03:20:55,408 epoch 3 - iter 3126/5212 - loss 0.09501689 - time (sec): 827.14 - samples/sec: 264.99 - lr: 0.000123 - momentum: 0.000000
134
+ 2023-10-12 03:23:18,463 epoch 3 - iter 3647/5212 - loss 0.09597740 - time (sec): 970.19 - samples/sec: 267.70 - lr: 0.000122 - momentum: 0.000000
135
+ 2023-10-12 03:25:40,852 epoch 3 - iter 4168/5212 - loss 0.09794887 - time (sec): 1112.58 - samples/sec: 263.01 - lr: 0.000120 - momentum: 0.000000
136
+ 2023-10-12 03:28:07,225 epoch 3 - iter 4689/5212 - loss 0.09904610 - time (sec): 1258.96 - samples/sec: 261.73 - lr: 0.000118 - momentum: 0.000000
137
+ 2023-10-12 03:30:34,184 epoch 3 - iter 5210/5212 - loss 0.09816484 - time (sec): 1405.91 - samples/sec: 261.22 - lr: 0.000117 - momentum: 0.000000
138
+ 2023-10-12 03:30:34,730 ----------------------------------------------------------------------------------------------------
139
+ 2023-10-12 03:30:34,731 EPOCH 3 done: loss 0.0981 - lr: 0.000117
140
+ 2023-10-12 03:31:15,205 DEV : loss 0.2600547671318054 - f1-score (micro avg) 0.3618
141
+ 2023-10-12 03:31:15,262 saving best model
142
+ 2023-10-12 03:31:17,833 ----------------------------------------------------------------------------------------------------
143
+ 2023-10-12 03:33:41,338 epoch 4 - iter 521/5212 - loss 0.06934331 - time (sec): 143.50 - samples/sec: 252.51 - lr: 0.000115 - momentum: 0.000000
144
+ 2023-10-12 03:36:05,531 epoch 4 - iter 1042/5212 - loss 0.07084255 - time (sec): 287.69 - samples/sec: 257.98 - lr: 0.000113 - momentum: 0.000000
145
+ 2023-10-12 03:38:29,695 epoch 4 - iter 1563/5212 - loss 0.06755543 - time (sec): 431.86 - samples/sec: 261.22 - lr: 0.000112 - momentum: 0.000000
146
+ 2023-10-12 03:40:54,693 epoch 4 - iter 2084/5212 - loss 0.06616838 - time (sec): 576.86 - samples/sec: 261.23 - lr: 0.000110 - momentum: 0.000000
147
+ 2023-10-12 03:43:17,865 epoch 4 - iter 2605/5212 - loss 0.06558266 - time (sec): 720.03 - samples/sec: 258.99 - lr: 0.000108 - momentum: 0.000000
148
+ 2023-10-12 03:45:41,115 epoch 4 - iter 3126/5212 - loss 0.06435344 - time (sec): 863.28 - samples/sec: 259.71 - lr: 0.000107 - momentum: 0.000000
149
+ 2023-10-12 03:48:03,559 epoch 4 - iter 3647/5212 - loss 0.06413682 - time (sec): 1005.72 - samples/sec: 259.21 - lr: 0.000105 - momentum: 0.000000
150
+ 2023-10-12 03:50:23,570 epoch 4 - iter 4168/5212 - loss 0.06596094 - time (sec): 1145.73 - samples/sec: 257.63 - lr: 0.000103 - momentum: 0.000000
151
+ 2023-10-12 03:52:45,111 epoch 4 - iter 4689/5212 - loss 0.06614742 - time (sec): 1287.27 - samples/sec: 258.43 - lr: 0.000102 - momentum: 0.000000
152
+ 2023-10-12 03:55:04,718 epoch 4 - iter 5210/5212 - loss 0.06648116 - time (sec): 1426.88 - samples/sec: 257.45 - lr: 0.000100 - momentum: 0.000000
153
+ 2023-10-12 03:55:05,152 ----------------------------------------------------------------------------------------------------
154
+ 2023-10-12 03:55:05,153 EPOCH 4 done: loss 0.0665 - lr: 0.000100
155
+ 2023-10-12 03:55:45,279 DEV : loss 0.3101561367511749 - f1-score (micro avg) 0.3555
156
+ 2023-10-12 03:55:45,331 ----------------------------------------------------------------------------------------------------
157
+ 2023-10-12 03:58:04,744 epoch 5 - iter 521/5212 - loss 0.04426358 - time (sec): 139.41 - samples/sec: 259.84 - lr: 0.000098 - momentum: 0.000000
158
+ 2023-10-12 04:00:24,001 epoch 5 - iter 1042/5212 - loss 0.04273940 - time (sec): 278.67 - samples/sec: 260.65 - lr: 0.000097 - momentum: 0.000000
159
+ 2023-10-12 04:02:41,900 epoch 5 - iter 1563/5212 - loss 0.04286202 - time (sec): 416.57 - samples/sec: 256.36 - lr: 0.000095 - momentum: 0.000000
160
+ 2023-10-12 04:05:03,372 epoch 5 - iter 2084/5212 - loss 0.04497980 - time (sec): 558.04 - samples/sec: 260.25 - lr: 0.000093 - momentum: 0.000000
161
+ 2023-10-12 04:07:19,613 epoch 5 - iter 2605/5212 - loss 0.04479044 - time (sec): 694.28 - samples/sec: 258.94 - lr: 0.000092 - momentum: 0.000000
162
+ 2023-10-12 04:09:43,942 epoch 5 - iter 3126/5212 - loss 0.04436294 - time (sec): 838.61 - samples/sec: 258.68 - lr: 0.000090 - momentum: 0.000000
163
+ 2023-10-12 04:12:11,821 epoch 5 - iter 3647/5212 - loss 0.04380965 - time (sec): 986.49 - samples/sec: 258.99 - lr: 0.000088 - momentum: 0.000000
164
+ 2023-10-12 04:14:37,186 epoch 5 - iter 4168/5212 - loss 0.04530386 - time (sec): 1131.85 - samples/sec: 258.58 - lr: 0.000087 - momentum: 0.000000
165
+ 2023-10-12 04:17:02,247 epoch 5 - iter 4689/5212 - loss 0.04665775 - time (sec): 1276.91 - samples/sec: 257.28 - lr: 0.000085 - momentum: 0.000000
166
+ 2023-10-12 04:19:31,906 epoch 5 - iter 5210/5212 - loss 0.04638068 - time (sec): 1426.57 - samples/sec: 257.51 - lr: 0.000083 - momentum: 0.000000
167
+ 2023-10-12 04:19:32,352 ----------------------------------------------------------------------------------------------------
168
+ 2023-10-12 04:19:32,353 EPOCH 5 done: loss 0.0464 - lr: 0.000083
169
+ 2023-10-12 04:20:13,213 DEV : loss 0.3113304078578949 - f1-score (micro avg) 0.4003
170
+ 2023-10-12 04:20:13,271 saving best model
171
+ 2023-10-12 04:20:14,217 ----------------------------------------------------------------------------------------------------
172
+ 2023-10-12 04:22:41,995 epoch 6 - iter 521/5212 - loss 0.02551644 - time (sec): 147.78 - samples/sec: 258.80 - lr: 0.000082 - momentum: 0.000000
173
+ 2023-10-12 04:25:08,081 epoch 6 - iter 1042/5212 - loss 0.02937621 - time (sec): 293.86 - samples/sec: 259.22 - lr: 0.000080 - momentum: 0.000000
174
+ 2023-10-12 04:27:33,632 epoch 6 - iter 1563/5212 - loss 0.03011832 - time (sec): 439.41 - samples/sec: 254.55 - lr: 0.000078 - momentum: 0.000000
175
+ 2023-10-12 04:29:57,124 epoch 6 - iter 2084/5212 - loss 0.03067784 - time (sec): 582.90 - samples/sec: 252.43 - lr: 0.000077 - momentum: 0.000000
176
+ 2023-10-12 04:32:22,709 epoch 6 - iter 2605/5212 - loss 0.03033894 - time (sec): 728.49 - samples/sec: 256.01 - lr: 0.000075 - momentum: 0.000000
177
+ 2023-10-12 04:34:46,393 epoch 6 - iter 3126/5212 - loss 0.03057800 - time (sec): 872.17 - samples/sec: 256.98 - lr: 0.000073 - momentum: 0.000000
178
+ 2023-10-12 04:37:07,410 epoch 6 - iter 3647/5212 - loss 0.03169587 - time (sec): 1013.19 - samples/sec: 256.43 - lr: 0.000072 - momentum: 0.000000
179
+ 2023-10-12 04:39:27,981 epoch 6 - iter 4168/5212 - loss 0.03188952 - time (sec): 1153.76 - samples/sec: 256.11 - lr: 0.000070 - momentum: 0.000000
180
+ 2023-10-12 04:41:50,066 epoch 6 - iter 4689/5212 - loss 0.03265092 - time (sec): 1295.85 - samples/sec: 256.18 - lr: 0.000068 - momentum: 0.000000
181
+ 2023-10-12 04:44:09,929 epoch 6 - iter 5210/5212 - loss 0.03297566 - time (sec): 1435.71 - samples/sec: 255.87 - lr: 0.000067 - momentum: 0.000000
182
+ 2023-10-12 04:44:10,371 ----------------------------------------------------------------------------------------------------
183
+ 2023-10-12 04:44:10,371 EPOCH 6 done: loss 0.0330 - lr: 0.000067
184
+ 2023-10-12 04:44:50,633 DEV : loss 0.40718281269073486 - f1-score (micro avg) 0.3971
185
+ 2023-10-12 04:44:50,685 ----------------------------------------------------------------------------------------------------
186
+ 2023-10-12 04:47:11,977 epoch 7 - iter 521/5212 - loss 0.02552297 - time (sec): 141.29 - samples/sec: 259.67 - lr: 0.000065 - momentum: 0.000000
187
+ 2023-10-12 04:49:32,146 epoch 7 - iter 1042/5212 - loss 0.02437180 - time (sec): 281.46 - samples/sec: 271.70 - lr: 0.000063 - momentum: 0.000000
188
+ 2023-10-12 04:51:48,960 epoch 7 - iter 1563/5212 - loss 0.02514062 - time (sec): 418.27 - samples/sec: 267.26 - lr: 0.000062 - momentum: 0.000000
189
+ 2023-10-12 04:54:07,753 epoch 7 - iter 2084/5212 - loss 0.02516668 - time (sec): 557.07 - samples/sec: 267.87 - lr: 0.000060 - momentum: 0.000000
190
+ 2023-10-12 04:56:27,890 epoch 7 - iter 2605/5212 - loss 0.02469412 - time (sec): 697.20 - samples/sec: 265.74 - lr: 0.000058 - momentum: 0.000000
191
+ 2023-10-12 04:58:49,869 epoch 7 - iter 3126/5212 - loss 0.02456715 - time (sec): 839.18 - samples/sec: 265.93 - lr: 0.000057 - momentum: 0.000000
192
+ 2023-10-12 05:01:09,634 epoch 7 - iter 3647/5212 - loss 0.02477855 - time (sec): 978.95 - samples/sec: 263.65 - lr: 0.000055 - momentum: 0.000000
193
+ 2023-10-12 05:03:36,700 epoch 7 - iter 4168/5212 - loss 0.02394687 - time (sec): 1126.01 - samples/sec: 262.72 - lr: 0.000053 - momentum: 0.000000
194
+ 2023-10-12 05:06:00,466 epoch 7 - iter 4689/5212 - loss 0.02356060 - time (sec): 1269.78 - samples/sec: 260.95 - lr: 0.000052 - momentum: 0.000000
195
+ 2023-10-12 05:08:24,608 epoch 7 - iter 5210/5212 - loss 0.02323141 - time (sec): 1413.92 - samples/sec: 259.78 - lr: 0.000050 - momentum: 0.000000
196
+ 2023-10-12 05:08:25,097 ----------------------------------------------------------------------------------------------------
197
+ 2023-10-12 05:08:25,097 EPOCH 7 done: loss 0.0232 - lr: 0.000050
198
+ 2023-10-12 05:09:04,885 DEV : loss 0.408222496509552 - f1-score (micro avg) 0.4035
199
+ 2023-10-12 05:09:04,936 saving best model
200
+ 2023-10-12 05:09:07,501 ----------------------------------------------------------------------------------------------------
201
+ 2023-10-12 05:11:27,797 epoch 8 - iter 521/5212 - loss 0.01537607 - time (sec): 140.29 - samples/sec: 265.21 - lr: 0.000048 - momentum: 0.000000
202
+ 2023-10-12 05:13:47,952 epoch 8 - iter 1042/5212 - loss 0.01753064 - time (sec): 280.45 - samples/sec: 271.51 - lr: 0.000047 - momentum: 0.000000
203
+ 2023-10-12 05:16:11,225 epoch 8 - iter 1563/5212 - loss 0.01696724 - time (sec): 423.72 - samples/sec: 277.40 - lr: 0.000045 - momentum: 0.000000
204
+ 2023-10-12 05:18:28,174 epoch 8 - iter 2084/5212 - loss 0.01669412 - time (sec): 560.67 - samples/sec: 273.78 - lr: 0.000043 - momentum: 0.000000
205
+ 2023-10-12 05:20:45,546 epoch 8 - iter 2605/5212 - loss 0.01686027 - time (sec): 698.04 - samples/sec: 269.95 - lr: 0.000042 - momentum: 0.000000
206
+ 2023-10-12 05:23:00,618 epoch 8 - iter 3126/5212 - loss 0.01671656 - time (sec): 833.11 - samples/sec: 267.12 - lr: 0.000040 - momentum: 0.000000
207
+ 2023-10-12 05:25:16,222 epoch 8 - iter 3647/5212 - loss 0.01600572 - time (sec): 968.72 - samples/sec: 265.31 - lr: 0.000038 - momentum: 0.000000
208
+ 2023-10-12 05:27:35,013 epoch 8 - iter 4168/5212 - loss 0.01597679 - time (sec): 1107.51 - samples/sec: 264.67 - lr: 0.000037 - momentum: 0.000000
209
+ 2023-10-12 05:29:56,551 epoch 8 - iter 4689/5212 - loss 0.01553053 - time (sec): 1249.05 - samples/sec: 263.33 - lr: 0.000035 - momentum: 0.000000
210
+ 2023-10-12 05:32:19,571 epoch 8 - iter 5210/5212 - loss 0.01639257 - time (sec): 1392.06 - samples/sec: 263.90 - lr: 0.000033 - momentum: 0.000000
211
+ 2023-10-12 05:32:19,996 ----------------------------------------------------------------------------------------------------
212
+ 2023-10-12 05:32:19,997 EPOCH 8 done: loss 0.0164 - lr: 0.000033
213
+ 2023-10-12 05:32:58,072 DEV : loss 0.4335840940475464 - f1-score (micro avg) 0.404
214
+ 2023-10-12 05:32:58,123 saving best model
215
+ 2023-10-12 05:33:00,797 ----------------------------------------------------------------------------------------------------
216
+ 2023-10-12 05:35:21,905 epoch 9 - iter 521/5212 - loss 0.01030793 - time (sec): 141.10 - samples/sec: 275.22 - lr: 0.000032 - momentum: 0.000000
217
+ 2023-10-12 05:37:41,298 epoch 9 - iter 1042/5212 - loss 0.01148381 - time (sec): 280.50 - samples/sec: 274.31 - lr: 0.000030 - momentum: 0.000000
218
+ 2023-10-12 05:40:02,195 epoch 9 - iter 1563/5212 - loss 0.01090402 - time (sec): 421.39 - samples/sec: 263.64 - lr: 0.000028 - momentum: 0.000000
219
+ 2023-10-12 05:42:24,062 epoch 9 - iter 2084/5212 - loss 0.01205054 - time (sec): 563.26 - samples/sec: 259.21 - lr: 0.000027 - momentum: 0.000000
220
+ 2023-10-12 05:44:49,921 epoch 9 - iter 2605/5212 - loss 0.01249863 - time (sec): 709.12 - samples/sec: 258.65 - lr: 0.000025 - momentum: 0.000000
221
+ 2023-10-12 05:47:12,873 epoch 9 - iter 3126/5212 - loss 0.01215066 - time (sec): 852.07 - samples/sec: 258.10 - lr: 0.000023 - momentum: 0.000000
222
+ 2023-10-12 05:49:39,769 epoch 9 - iter 3647/5212 - loss 0.01116151 - time (sec): 998.97 - samples/sec: 259.45 - lr: 0.000022 - momentum: 0.000000
223
+ 2023-10-12 05:52:01,961 epoch 9 - iter 4168/5212 - loss 0.01081521 - time (sec): 1141.16 - samples/sec: 257.32 - lr: 0.000020 - momentum: 0.000000
224
+ 2023-10-12 05:54:25,864 epoch 9 - iter 4689/5212 - loss 0.01082631 - time (sec): 1285.06 - samples/sec: 256.89 - lr: 0.000018 - momentum: 0.000000
225
+ 2023-10-12 05:56:50,574 epoch 9 - iter 5210/5212 - loss 0.01087435 - time (sec): 1429.77 - samples/sec: 256.94 - lr: 0.000017 - momentum: 0.000000
226
+ 2023-10-12 05:56:51,009 ----------------------------------------------------------------------------------------------------
227
+ 2023-10-12 05:56:51,009 EPOCH 9 done: loss 0.0109 - lr: 0.000017
228
+ 2023-10-12 05:57:30,552 DEV : loss 0.47575101256370544 - f1-score (micro avg) 0.3948
229
+ 2023-10-12 05:57:30,605 ----------------------------------------------------------------------------------------------------
230
+ 2023-10-12 05:59:52,416 epoch 10 - iter 521/5212 - loss 0.00498377 - time (sec): 141.81 - samples/sec: 252.02 - lr: 0.000015 - momentum: 0.000000
231
+ 2023-10-12 06:02:13,588 epoch 10 - iter 1042/5212 - loss 0.00681335 - time (sec): 282.98 - samples/sec: 255.31 - lr: 0.000013 - momentum: 0.000000
232
+ 2023-10-12 06:04:37,692 epoch 10 - iter 1563/5212 - loss 0.00622093 - time (sec): 427.08 - samples/sec: 259.59 - lr: 0.000012 - momentum: 0.000000
233
+ 2023-10-12 06:06:59,831 epoch 10 - iter 2084/5212 - loss 0.00629281 - time (sec): 569.22 - samples/sec: 258.24 - lr: 0.000010 - momentum: 0.000000
234
+ 2023-10-12 06:09:21,704 epoch 10 - iter 2605/5212 - loss 0.00707237 - time (sec): 711.10 - samples/sec: 257.28 - lr: 0.000008 - momentum: 0.000000
235
+ 2023-10-12 06:11:42,090 epoch 10 - iter 3126/5212 - loss 0.00730698 - time (sec): 851.48 - samples/sec: 256.91 - lr: 0.000007 - momentum: 0.000000
236
+ 2023-10-12 06:14:01,717 epoch 10 - iter 3647/5212 - loss 0.00730729 - time (sec): 991.11 - samples/sec: 258.63 - lr: 0.000005 - momentum: 0.000000
237
+ 2023-10-12 06:16:22,285 epoch 10 - iter 4168/5212 - loss 0.00690845 - time (sec): 1131.68 - samples/sec: 260.93 - lr: 0.000003 - momentum: 0.000000
238
+ 2023-10-12 06:18:43,018 epoch 10 - iter 4689/5212 - loss 0.00696219 - time (sec): 1272.41 - samples/sec: 260.74 - lr: 0.000002 - momentum: 0.000000
239
+ 2023-10-12 06:21:02,617 epoch 10 - iter 5210/5212 - loss 0.00715270 - time (sec): 1412.01 - samples/sec: 260.15 - lr: 0.000000 - momentum: 0.000000
240
+ 2023-10-12 06:21:03,064 ----------------------------------------------------------------------------------------------------
241
+ 2023-10-12 06:21:03,065 EPOCH 10 done: loss 0.0072 - lr: 0.000000
242
+ 2023-10-12 06:21:42,939 DEV : loss 0.4921533763408661 - f1-score (micro avg) 0.3907
243
+ 2023-10-12 06:21:43,893 ----------------------------------------------------------------------------------------------------
244
+ 2023-10-12 06:21:43,895 Loading model from best epoch ...
245
+ 2023-10-12 06:21:47,649 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
246
+ 2023-10-12 06:23:28,188
247
+ Results:
248
+ - F-score (micro) 0.4681
249
+ - F-score (macro) 0.3274
250
+ - Accuracy 0.3108
251
+
252
+ By class:
253
+ precision recall f1-score support
254
+
255
+ LOC 0.5033 0.5610 0.5306 1214
256
+ PER 0.4123 0.4567 0.4334 808
257
+ ORG 0.3282 0.3654 0.3458 353
258
+ HumanProd 0.0000 0.0000 0.0000 15
259
+
260
+ micro avg 0.4454 0.4933 0.4681 2390
261
+ macro avg 0.3110 0.3458 0.3274 2390
262
+ weighted avg 0.4435 0.4933 0.4671 2390
263
+
264
+ 2023-10-12 06:23:28,188 ----------------------------------------------------------------------------------------------------