Hugging Face's logo

stefan-it
/
hmbench-newseye-de-hmbyt5-bs8-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-1

Token Classification
Flair
PyTorch
TensorBoard
German
sequence-tagger-model
Model card Files Metrics Community
2
stefan-it commited on
Commit
9c93531
•
1 Parent(s): 2a8a824

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.1696871716.6d4c7681f95b.1253.0 +3 -0
  6. test.tsv +0 -0
  7. training.log +266 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9bf7e7b28552c38ca89afc8a389eae078cf859c18970949a5cede49004f9c4d1
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:dbac1255eab9b7bb902eb98d9b694efdeeec3f3c236741db9c0654aeb14d9e1c
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 17:39:08 0.0001 0.9389 0.1276 0.5275 0.0909 0.1551 0.0841
3
+ 2 18:03:10 0.0001 0.1657 0.1198 0.2745 0.5720 0.3710 0.2291
4
+ 3 18:27:40 0.0001 0.0948 0.1977 0.2712 0.6610 0.3846 0.2399
5
+ 4 18:52:16 0.0001 0.0667 0.2411 0.2721 0.6705 0.3871 0.2408
6
+ 5 19:16:30 0.0001 0.0495 0.3171 0.2790 0.6515 0.3907 0.2438
7
+ 6 19:40:33 0.0001 0.0375 0.3395 0.2861 0.6307 0.3936 0.2465
8
+ 7 20:04:40 0.0001 0.0274 0.3874 0.2909 0.6402 0.4000 0.2521
9
+ 8 20:28:55 0.0000 0.0193 0.3869 0.3035 0.6098 0.4053 0.2560
10
+ 9 20:53:13 0.0000 0.0146 0.4353 0.2910 0.6155 0.3951 0.2479
11
+ 10 21:17:20 0.0000 0.0111 0.4450 0.2878 0.6098 0.3910 0.2449
runs/events.out.tfevents.1696871716.6d4c7681f95b.1253.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:429918a5bdbf8be9a583dc47bf2be99465b595780d198fa57aaf0c25abf1e534
3
+ size 1464420
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,266 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-09 17:15:16,485 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-09 17:15:16,488 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
+ 2023-10-09 17:15:16,488 ----------------------------------------------------------------------------------------------------
71
+ 2023-10-09 17:15:16,488 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
+ 2023-10-09 17:15:16,488 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-09 17:15:16,488 Train: 20847 sentences
75
+ 2023-10-09 17:15:16,488 (train_with_dev=False, train_with_test=False)
76
+ 2023-10-09 17:15:16,489 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-09 17:15:16,489 Training Params:
78
+ 2023-10-09 17:15:16,489 - learning_rate: "0.00015"
79
+ 2023-10-09 17:15:16,489 - mini_batch_size: "8"
80
+ 2023-10-09 17:15:16,489 - max_epochs: "10"
81
+ 2023-10-09 17:15:16,489 - shuffle: "True"
82
+ 2023-10-09 17:15:16,489 ----------------------------------------------------------------------------------------------------
83
+ 2023-10-09 17:15:16,489 Plugins:
84
+ 2023-10-09 17:15:16,489 - TensorboardLogger
85
+ 2023-10-09 17:15:16,489 - LinearScheduler | warmup_fraction: '0.1'
86
+ 2023-10-09 17:15:16,489 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-09 17:15:16,489 Final evaluation on model from best epoch (best-model.pt)
88
+ 2023-10-09 17:15:16,489 - metric: "('micro avg', 'f1-score')"
89
+ 2023-10-09 17:15:16,489 ----------------------------------------------------------------------------------------------------
90
+ 2023-10-09 17:15:16,489 Computation:
91
+ 2023-10-09 17:15:16,490 - compute on device: cuda:0
92
+ 2023-10-09 17:15:16,490 - embedding storage: none
93
+ 2023-10-09 17:15:16,490 ----------------------------------------------------------------------------------------------------
94
+ 2023-10-09 17:15:16,490 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-1"
95
+ 2023-10-09 17:15:16,490 ----------------------------------------------------------------------------------------------------
96
+ 2023-10-09 17:15:16,490 ----------------------------------------------------------------------------------------------------
97
+ 2023-10-09 17:15:16,490 Logging anything other than scalars to TensorBoard is currently not supported.
98
+ 2023-10-09 17:17:37,420 epoch 1 - iter 260/2606 - loss 2.80844215 - time (sec): 140.93 - samples/sec: 279.58 - lr: 0.000015 - momentum: 0.000000
99
+ 2023-10-09 17:19:54,518 epoch 1 - iter 520/2606 - loss 2.59634928 - time (sec): 278.03 - samples/sec: 265.90 - lr: 0.000030 - momentum: 0.000000
100
+ 2023-10-09 17:22:11,733 epoch 1 - iter 780/2606 - loss 2.19837014 - time (sec): 415.24 - samples/sec: 264.99 - lr: 0.000045 - momentum: 0.000000
101
+ 2023-10-09 17:24:27,192 epoch 1 - iter 1040/2606 - loss 1.84021340 - time (sec): 550.70 - samples/sec: 261.53 - lr: 0.000060 - momentum: 0.000000
102
+ 2023-10-09 17:26:50,107 epoch 1 - iter 1300/2606 - loss 1.54998051 - time (sec): 693.62 - samples/sec: 262.04 - lr: 0.000075 - momentum: 0.000000
103
+ 2023-10-09 17:29:09,005 epoch 1 - iter 1560/2606 - loss 1.36095305 - time (sec): 832.51 - samples/sec: 264.14 - lr: 0.000090 - momentum: 0.000000
104
+ 2023-10-09 17:31:25,263 epoch 1 - iter 1820/2606 - loss 1.22756925 - time (sec): 968.77 - samples/sec: 263.17 - lr: 0.000105 - momentum: 0.000000
105
+ 2023-10-09 17:33:44,623 epoch 1 - iter 2080/2606 - loss 1.11197802 - time (sec): 1108.13 - samples/sec: 263.29 - lr: 0.000120 - momentum: 0.000000
106
+ 2023-10-09 17:36:04,875 epoch 1 - iter 2340/2606 - loss 1.01382528 - time (sec): 1248.38 - samples/sec: 264.69 - lr: 0.000135 - momentum: 0.000000
107
+ 2023-10-09 17:38:28,029 epoch 1 - iter 2600/2606 - loss 0.94022654 - time (sec): 1391.54 - samples/sec: 263.48 - lr: 0.000150 - momentum: 0.000000
108
+ 2023-10-09 17:38:31,088 ----------------------------------------------------------------------------------------------------
109
+ 2023-10-09 17:38:31,088 EPOCH 1 done: loss 0.9389 - lr: 0.000150
110
+ 2023-10-09 17:39:08,925 DEV : loss 0.12756438553333282 - f1-score (micro avg) 0.1551
111
+ 2023-10-09 17:39:08,978 saving best model
112
+ 2023-10-09 17:39:09,961 ----------------------------------------------------------------------------------------------------
113
+ 2023-10-09 17:41:29,838 epoch 2 - iter 260/2606 - loss 0.21716590 - time (sec): 139.87 - samples/sec: 284.27 - lr: 0.000148 - momentum: 0.000000
114
+ 2023-10-09 17:43:51,518 epoch 2 - iter 520/2606 - loss 0.21823044 - time (sec): 281.55 - samples/sec: 279.07 - lr: 0.000147 - momentum: 0.000000
115
+ 2023-10-09 17:46:14,404 epoch 2 - iter 780/2606 - loss 0.20602971 - time (sec): 424.44 - samples/sec: 272.97 - lr: 0.000145 - momentum: 0.000000
116
+ 2023-10-09 17:48:30,932 epoch 2 - iter 1040/2606 - loss 0.19906942 - time (sec): 560.97 - samples/sec: 267.92 - lr: 0.000143 - momentum: 0.000000
117
+ 2023-10-09 17:50:52,954 epoch 2 - iter 1300/2606 - loss 0.19435757 - time (sec): 702.99 - samples/sec: 265.12 - lr: 0.000142 - momentum: 0.000000
118
+ 2023-10-09 17:53:10,549 epoch 2 - iter 1560/2606 - loss 0.18819497 - time (sec): 840.59 - samples/sec: 264.28 - lr: 0.000140 - momentum: 0.000000
119
+ 2023-10-09 17:55:26,312 epoch 2 - iter 1820/2606 - loss 0.18214589 - time (sec): 976.35 - samples/sec: 264.47 - lr: 0.000138 - momentum: 0.000000
120
+ 2023-10-09 17:57:51,067 epoch 2 - iter 2080/2606 - loss 0.17553467 - time (sec): 1121.10 - samples/sec: 263.23 - lr: 0.000137 - momentum: 0.000000
121
+ 2023-10-09 18:00:12,053 epoch 2 - iter 2340/2606 - loss 0.17073161 - time (sec): 1262.09 - samples/sec: 263.72 - lr: 0.000135 - momentum: 0.000000
122
+ 2023-10-09 18:02:25,945 epoch 2 - iter 2600/2606 - loss 0.16588428 - time (sec): 1395.98 - samples/sec: 262.64 - lr: 0.000133 - momentum: 0.000000
123
+ 2023-10-09 18:02:28,954 ----------------------------------------------------------------------------------------------------
124
+ 2023-10-09 18:02:28,954 EPOCH 2 done: loss 0.1657 - lr: 0.000133
125
+ 2023-10-09 18:03:09,987 DEV : loss 0.11978743970394135 - f1-score (micro avg) 0.371
126
+ 2023-10-09 18:03:10,043 saving best model
127
+ 2023-10-09 18:03:12,756 ----------------------------------------------------------------------------------------------------
128
+ 2023-10-09 18:05:30,218 epoch 3 - iter 260/2606 - loss 0.09297887 - time (sec): 137.46 - samples/sec: 264.90 - lr: 0.000132 - momentum: 0.000000
129
+ 2023-10-09 18:07:49,683 epoch 3 - iter 520/2606 - loss 0.09874622 - time (sec): 276.92 - samples/sec: 257.70 - lr: 0.000130 - momentum: 0.000000
130
+ 2023-10-09 18:10:11,726 epoch 3 - iter 780/2606 - loss 0.09506556 - time (sec): 418.97 - samples/sec: 263.99 - lr: 0.000128 - momentum: 0.000000
131
+ 2023-10-09 18:12:28,824 epoch 3 - iter 1040/2606 - loss 0.09671015 - time (sec): 556.06 - samples/sec: 258.83 - lr: 0.000127 - momentum: 0.000000
132
+ 2023-10-09 18:14:46,288 epoch 3 - iter 1300/2606 - loss 0.09618722 - time (sec): 693.53 - samples/sec: 255.54 - lr: 0.000125 - momentum: 0.000000
133
+ 2023-10-09 18:17:21,963 epoch 3 - iter 1560/2606 - loss 0.09599238 - time (sec): 849.20 - samples/sec: 255.47 - lr: 0.000123 - momentum: 0.000000
134
+ 2023-10-09 18:19:49,225 epoch 3 - iter 1820/2606 - loss 0.09643398 - time (sec): 996.46 - samples/sec: 256.55 - lr: 0.000122 - momentum: 0.000000
135
+ 2023-10-09 18:22:13,095 epoch 3 - iter 2080/2606 - loss 0.09553134 - time (sec): 1140.33 - samples/sec: 257.17 - lr: 0.000120 - momentum: 0.000000
136
+ 2023-10-09 18:24:35,735 epoch 3 - iter 2340/2606 - loss 0.09565909 - time (sec): 1282.98 - samples/sec: 258.25 - lr: 0.000118 - momentum: 0.000000
137
+ 2023-10-09 18:26:55,235 epoch 3 - iter 2600/2606 - loss 0.09460902 - time (sec): 1422.47 - samples/sec: 257.86 - lr: 0.000117 - momentum: 0.000000
138
+ 2023-10-09 18:26:58,231 ----------------------------------------------------------------------------------------------------
139
+ 2023-10-09 18:26:58,232 EPOCH 3 done: loss 0.0948 - lr: 0.000117
140
+ 2023-10-09 18:27:40,221 DEV : loss 0.19768929481506348 - f1-score (micro avg) 0.3846
141
+ 2023-10-09 18:27:40,288 saving best model
142
+ 2023-10-09 18:27:43,051 ----------------------------------------------------------------------------------------------------
143
+ 2023-10-09 18:30:08,103 epoch 4 - iter 260/2606 - loss 0.05758779 - time (sec): 145.05 - samples/sec: 251.73 - lr: 0.000115 - momentum: 0.000000
144
+ 2023-10-09 18:32:28,172 epoch 4 - iter 520/2606 - loss 0.05701437 - time (sec): 285.11 - samples/sec: 252.42 - lr: 0.000113 - momentum: 0.000000
145
+ 2023-10-09 18:34:43,346 epoch 4 - iter 780/2606 - loss 0.06038270 - time (sec): 420.29 - samples/sec: 256.11 - lr: 0.000112 - momentum: 0.000000
146
+ 2023-10-09 18:37:06,586 epoch 4 - iter 1040/2606 - loss 0.06347394 - time (sec): 563.53 - samples/sec: 254.59 - lr: 0.000110 - momentum: 0.000000
147
+ 2023-10-09 18:39:25,773 epoch 4 - iter 1300/2606 - loss 0.06753057 - time (sec): 702.72 - samples/sec: 257.50 - lr: 0.000108 - momentum: 0.000000
148
+ 2023-10-09 18:41:57,299 epoch 4 - iter 1560/2606 - loss 0.06627879 - time (sec): 854.24 - samples/sec: 258.99 - lr: 0.000107 - momentum: 0.000000
149
+ 2023-10-09 18:44:27,149 epoch 4 - iter 1820/2606 - loss 0.06505761 - time (sec): 1004.09 - samples/sec: 255.23 - lr: 0.000105 - momentum: 0.000000
150
+ 2023-10-09 18:46:47,745 epoch 4 - iter 2080/2606 - loss 0.06531655 - time (sec): 1144.69 - samples/sec: 255.16 - lr: 0.000103 - momentum: 0.000000
151
+ 2023-10-09 18:49:08,031 epoch 4 - iter 2340/2606 - loss 0.06644938 - time (sec): 1284.97 - samples/sec: 256.77 - lr: 0.000102 - momentum: 0.000000
152
+ 2023-10-09 18:51:30,901 epoch 4 - iter 2600/2606 - loss 0.06675340 - time (sec): 1427.84 - samples/sec: 256.83 - lr: 0.000100 - momentum: 0.000000
153
+ 2023-10-09 18:51:33,919 ----------------------------------------------------------------------------------------------------
154
+ 2023-10-09 18:51:33,920 EPOCH 4 done: loss 0.0667 - lr: 0.000100
155
+ 2023-10-09 18:52:16,713 DEV : loss 0.2411346584558487 - f1-score (micro avg) 0.3871
156
+ 2023-10-09 18:52:16,768 saving best model
157
+ 2023-10-09 18:52:19,558 ----------------------------------------------------------------------------------------------------
158
+ 2023-10-09 18:54:37,061 epoch 5 - iter 260/2606 - loss 0.04897725 - time (sec): 137.50 - samples/sec: 250.32 - lr: 0.000098 - momentum: 0.000000
159
+ 2023-10-09 18:56:59,505 epoch 5 - iter 520/2606 - loss 0.05212024 - time (sec): 279.94 - samples/sec: 254.73 - lr: 0.000097 - momentum: 0.000000
160
+ 2023-10-09 18:59:18,730 epoch 5 - iter 780/2606 - loss 0.05140447 - time (sec): 419.17 - samples/sec: 262.04 - lr: 0.000095 - momentum: 0.000000
161
+ 2023-10-09 19:01:43,206 epoch 5 - iter 1040/2606 - loss 0.04930560 - time (sec): 563.64 - samples/sec: 263.39 - lr: 0.000093 - momentum: 0.000000
162
+ 2023-10-09 19:04:11,740 epoch 5 - iter 1300/2606 - loss 0.04940420 - time (sec): 712.18 - samples/sec: 257.16 - lr: 0.000092 - momentum: 0.000000
163
+ 2023-10-09 19:06:28,925 epoch 5 - iter 1560/2606 - loss 0.05106248 - time (sec): 849.36 - samples/sec: 259.62 - lr: 0.000090 - momentum: 0.000000
164
+ 2023-10-09 19:08:49,284 epoch 5 - iter 1820/2606 - loss 0.05104734 - time (sec): 989.72 - samples/sec: 261.69 - lr: 0.000088 - momentum: 0.000000
165
+ 2023-10-09 19:11:08,989 epoch 5 - iter 2080/2606 - loss 0.05050373 - time (sec): 1129.43 - samples/sec: 261.66 - lr: 0.000087 - momentum: 0.000000
166
+ 2023-10-09 19:13:24,149 epoch 5 - iter 2340/2606 - loss 0.04911289 - time (sec): 1264.59 - samples/sec: 260.94 - lr: 0.000085 - momentum: 0.000000
167
+ 2023-10-09 19:15:47,186 epoch 5 - iter 2600/2606 - loss 0.04960549 - time (sec): 1407.62 - samples/sec: 260.13 - lr: 0.000083 - momentum: 0.000000
168
+ 2023-10-09 19:15:50,688 ----------------------------------------------------------------------------------------------------
169
+ 2023-10-09 19:15:50,688 EPOCH 5 done: loss 0.0495 - lr: 0.000083
170
+ 2023-10-09 19:16:30,675 DEV : loss 0.3170567452907562 - f1-score (micro avg) 0.3907
171
+ 2023-10-09 19:16:30,741 saving best model
172
+ 2023-10-09 19:16:33,464 ----------------------------------------------------------------------------------------------------
173
+ 2023-10-09 19:18:54,346 epoch 6 - iter 260/2606 - loss 0.02953684 - time (sec): 140.88 - samples/sec: 260.88 - lr: 0.000082 - momentum: 0.000000
174
+ 2023-10-09 19:21:09,418 epoch 6 - iter 520/2606 - loss 0.03598038 - time (sec): 275.95 - samples/sec: 257.01 - lr: 0.000080 - momentum: 0.000000
175
+ 2023-10-09 19:23:34,781 epoch 6 - iter 780/2606 - loss 0.03404511 - time (sec): 421.31 - samples/sec: 260.71 - lr: 0.000078 - momentum: 0.000000
176
+ 2023-10-09 19:25:54,098 epoch 6 - iter 1040/2606 - loss 0.03515730 - time (sec): 560.63 - samples/sec: 259.17 - lr: 0.000077 - momentum: 0.000000
177
+ 2023-10-09 19:28:11,300 epoch 6 - iter 1300/2606 - loss 0.03485847 - time (sec): 697.83 - samples/sec: 260.53 - lr: 0.000075 - momentum: 0.000000
178
+ 2023-10-09 19:30:28,334 epoch 6 - iter 1560/2606 - loss 0.03437084 - time (sec): 834.87 - samples/sec: 258.62 - lr: 0.000073 - momentum: 0.000000
179
+ 2023-10-09 19:32:52,558 epoch 6 - iter 1820/2606 - loss 0.03481753 - time (sec): 979.09 - samples/sec: 258.14 - lr: 0.000072 - momentum: 0.000000
180
+ 2023-10-09 19:35:10,177 epoch 6 - iter 2080/2606 - loss 0.03570515 - time (sec): 1116.71 - samples/sec: 260.16 - lr: 0.000070 - momentum: 0.000000
181
+ 2023-10-09 19:37:31,616 epoch 6 - iter 2340/2606 - loss 0.03684716 - time (sec): 1258.15 - samples/sec: 261.27 - lr: 0.000068 - momentum: 0.000000
182
+ 2023-10-09 19:39:48,990 epoch 6 - iter 2600/2606 - loss 0.03755532 - time (sec): 1395.52 - samples/sec: 262.95 - lr: 0.000067 - momentum: 0.000000
183
+ 2023-10-09 19:39:51,772 ----------------------------------------------------------------------------------------------------
184
+ 2023-10-09 19:39:51,772 EPOCH 6 done: loss 0.0375 - lr: 0.000067
185
+ 2023-10-09 19:40:33,007 DEV : loss 0.33949336409568787 - f1-score (micro avg) 0.3936
186
+ 2023-10-09 19:40:33,059 saving best model
187
+ 2023-10-09 19:40:36,223 ----------------------------------------------------------------------------------------------------
188
+ 2023-10-09 19:42:58,944 epoch 7 - iter 260/2606 - loss 0.02483364 - time (sec): 142.72 - samples/sec: 268.78 - lr: 0.000065 - momentum: 0.000000
189
+ 2023-10-09 19:45:20,220 epoch 7 - iter 520/2606 - loss 0.02577290 - time (sec): 283.99 - samples/sec: 269.77 - lr: 0.000063 - momentum: 0.000000
190
+ 2023-10-09 19:47:45,096 epoch 7 - iter 780/2606 - loss 0.02603966 - time (sec): 428.87 - samples/sec: 260.08 - lr: 0.000062 - momentum: 0.000000
191
+ 2023-10-09 19:50:02,396 epoch 7 - iter 1040/2606 - loss 0.02599984 - time (sec): 566.17 - samples/sec: 262.99 - lr: 0.000060 - momentum: 0.000000
192
+ 2023-10-09 19:52:21,988 epoch 7 - iter 1300/2606 - loss 0.02629164 - time (sec): 705.76 - samples/sec: 264.61 - lr: 0.000058 - momentum: 0.000000
193
+ 2023-10-09 19:54:41,201 epoch 7 - iter 1560/2606 - loss 0.02755866 - time (sec): 844.97 - samples/sec: 264.70 - lr: 0.000057 - momentum: 0.000000
194
+ 2023-10-09 19:57:00,615 epoch 7 - iter 1820/2606 - loss 0.02652858 - time (sec): 984.39 - samples/sec: 263.55 - lr: 0.000055 - momentum: 0.000000
195
+ 2023-10-09 19:59:22,484 epoch 7 - iter 2080/2606 - loss 0.02678278 - time (sec): 1126.26 - samples/sec: 262.35 - lr: 0.000053 - momentum: 0.000000
196
+ 2023-10-09 20:01:39,845 epoch 7 - iter 2340/2606 - loss 0.02684772 - time (sec): 1263.62 - samples/sec: 262.71 - lr: 0.000052 - momentum: 0.000000
197
+ 2023-10-09 20:03:57,279 epoch 7 - iter 2600/2606 - loss 0.02739470 - time (sec): 1401.05 - samples/sec: 261.70 - lr: 0.000050 - momentum: 0.000000
198
+ 2023-10-09 20:04:00,381 ----------------------------------------------------------------------------------------------------
199
+ 2023-10-09 20:04:00,381 EPOCH 7 done: loss 0.0274 - lr: 0.000050
200
+ 2023-10-09 20:04:40,701 DEV : loss 0.38742774724960327 - f1-score (micro avg) 0.4
201
+ 2023-10-09 20:04:40,755 saving best model
202
+ 2023-10-09 20:04:43,466 ----------------------------------------------------------------------------------------------------
203
+ 2023-10-09 20:07:02,590 epoch 8 - iter 260/2606 - loss 0.01385814 - time (sec): 139.12 - samples/sec: 259.45 - lr: 0.000048 - momentum: 0.000000
204
+ 2023-10-09 20:09:25,786 epoch 8 - iter 520/2606 - loss 0.01675223 - time (sec): 282.31 - samples/sec: 258.12 - lr: 0.000047 - momentum: 0.000000
205
+ 2023-10-09 20:11:46,436 epoch 8 - iter 780/2606 - loss 0.01929813 - time (sec): 422.96 - samples/sec: 261.89 - lr: 0.000045 - momentum: 0.000000
206
+ 2023-10-09 20:14:07,638 epoch 8 - iter 1040/2606 - loss 0.01972895 - time (sec): 564.17 - samples/sec: 259.87 - lr: 0.000043 - momentum: 0.000000
207
+ 2023-10-09 20:16:28,983 epoch 8 - iter 1300/2606 - loss 0.01962523 - time (sec): 705.51 - samples/sec: 258.60 - lr: 0.000042 - momentum: 0.000000
208
+ 2023-10-09 20:18:48,035 epoch 8 - iter 1560/2606 - loss 0.01996312 - time (sec): 844.56 - samples/sec: 259.96 - lr: 0.000040 - momentum: 0.000000
209
+ 2023-10-09 20:21:05,991 epoch 8 - iter 1820/2606 - loss 0.01993597 - time (sec): 982.52 - samples/sec: 259.22 - lr: 0.000038 - momentum: 0.000000
210
+ 2023-10-09 20:23:29,379 epoch 8 - iter 2080/2606 - loss 0.01930578 - time (sec): 1125.91 - samples/sec: 260.65 - lr: 0.000037 - momentum: 0.000000
211
+ 2023-10-09 20:25:49,353 epoch 8 - iter 2340/2606 - loss 0.01916722 - time (sec): 1265.88 - samples/sec: 261.06 - lr: 0.000035 - momentum: 0.000000
212
+ 2023-10-09 20:28:08,993 epoch 8 - iter 2600/2606 - loss 0.01927053 - time (sec): 1405.52 - samples/sec: 260.85 - lr: 0.000033 - momentum: 0.000000
213
+ 2023-10-09 20:28:12,122 ----------------------------------------------------------------------------------------------------
214
+ 2023-10-09 20:28:12,123 EPOCH 8 done: loss 0.0193 - lr: 0.000033
215
+ 2023-10-09 20:28:55,401 DEV : loss 0.3869289755821228 - f1-score (micro avg) 0.4053
216
+ 2023-10-09 20:28:55,456 saving best model
217
+ 2023-10-09 20:28:58,212 ----------------------------------------------------------------------------------------------------
218
+ 2023-10-09 20:31:20,576 epoch 9 - iter 260/2606 - loss 0.01964526 - time (sec): 142.36 - samples/sec: 265.14 - lr: 0.000032 - momentum: 0.000000
219
+ 2023-10-09 20:33:49,275 epoch 9 - iter 520/2606 - loss 0.01736846 - time (sec): 291.06 - samples/sec: 259.41 - lr: 0.000030 - momentum: 0.000000
220
+ 2023-10-09 20:36:07,617 epoch 9 - iter 780/2606 - loss 0.01681947 - time (sec): 429.40 - samples/sec: 256.02 - lr: 0.000028 - momentum: 0.000000
221
+ 2023-10-09 20:38:25,705 epoch 9 - iter 1040/2606 - loss 0.01643746 - time (sec): 567.49 - samples/sec: 258.74 - lr: 0.000027 - momentum: 0.000000
222
+ 2023-10-09 20:40:47,560 epoch 9 - iter 1300/2606 - loss 0.01638627 - time (sec): 709.35 - samples/sec: 258.39 - lr: 0.000025 - momentum: 0.000000
223
+ 2023-10-09 20:43:12,690 epoch 9 - iter 1560/2606 - loss 0.01632827 - time (sec): 854.48 - samples/sec: 257.57 - lr: 0.000023 - momentum: 0.000000
224
+ 2023-10-09 20:45:29,379 epoch 9 - iter 1820/2606 - loss 0.01610700 - time (sec): 991.16 - samples/sec: 257.84 - lr: 0.000022 - momentum: 0.000000
225
+ 2023-10-09 20:47:48,266 epoch 9 - iter 2080/2606 - loss 0.01530351 - time (sec): 1130.05 - samples/sec: 257.52 - lr: 0.000020 - momentum: 0.000000
226
+ 2023-10-09 20:50:06,400 epoch 9 - iter 2340/2606 - loss 0.01476616 - time (sec): 1268.19 - samples/sec: 258.76 - lr: 0.000018 - momentum: 0.000000
227
+ 2023-10-09 20:52:29,519 epoch 9 - iter 2600/2606 - loss 0.01456160 - time (sec): 1411.30 - samples/sec: 259.57 - lr: 0.000017 - momentum: 0.000000
228
+ 2023-10-09 20:52:32,863 ----------------------------------------------------------------------------------------------------
229
+ 2023-10-09 20:52:32,863 EPOCH 9 done: loss 0.0146 - lr: 0.000017
230
+ 2023-10-09 20:53:13,595 DEV : loss 0.4352709650993347 - f1-score (micro avg) 0.3951
231
+ 2023-10-09 20:53:13,648 ----------------------------------------------------------------------------------------------------
232
+ 2023-10-09 20:55:32,333 epoch 10 - iter 260/2606 - loss 0.01430925 - time (sec): 138.68 - samples/sec: 264.73 - lr: 0.000015 - momentum: 0.000000
233
+ 2023-10-09 20:57:49,209 epoch 10 - iter 520/2606 - loss 0.01237005 - time (sec): 275.56 - samples/sec: 260.68 - lr: 0.000013 - momentum: 0.000000
234
+ 2023-10-09 21:00:07,081 epoch 10 - iter 780/2606 - loss 0.01311737 - time (sec): 413.43 - samples/sec: 253.88 - lr: 0.000012 - momentum: 0.000000
235
+ 2023-10-09 21:02:31,957 epoch 10 - iter 1040/2606 - loss 0.01208504 - time (sec): 558.31 - samples/sec: 257.55 - lr: 0.000010 - momentum: 0.000000
236
+ 2023-10-09 21:04:55,245 epoch 10 - iter 1300/2606 - loss 0.01132225 - time (sec): 701.59 - samples/sec: 263.05 - lr: 0.000008 - momentum: 0.000000
237
+ 2023-10-09 21:07:13,857 epoch 10 - iter 1560/2606 - loss 0.01149290 - time (sec): 840.21 - samples/sec: 261.50 - lr: 0.000007 - momentum: 0.000000
238
+ 2023-10-09 21:09:34,496 epoch 10 - iter 1820/2606 - loss 0.01187455 - time (sec): 980.85 - samples/sec: 262.04 - lr: 0.000005 - momentum: 0.000000
239
+ 2023-10-09 21:11:54,667 epoch 10 - iter 2080/2606 - loss 0.01114111 - time (sec): 1121.02 - samples/sec: 262.78 - lr: 0.000003 - momentum: 0.000000
240
+ 2023-10-09 21:14:15,362 epoch 10 - iter 2340/2606 - loss 0.01096729 - time (sec): 1261.71 - samples/sec: 263.22 - lr: 0.000002 - momentum: 0.000000
241
+ 2023-10-09 21:16:34,194 epoch 10 - iter 2600/2606 - loss 0.01107741 - time (sec): 1400.54 - samples/sec: 261.67 - lr: 0.000000 - momentum: 0.000000
242
+ 2023-10-09 21:16:37,457 ----------------------------------------------------------------------------------------------------
243
+ 2023-10-09 21:16:37,458 EPOCH 10 done: loss 0.0111 - lr: 0.000000
244
+ 2023-10-09 21:17:20,301 DEV : loss 0.44496238231658936 - f1-score (micro avg) 0.391
245
+ 2023-10-09 21:17:21,350 ----------------------------------------------------------------------------------------------------
246
+ 2023-10-09 21:17:21,352 Loading model from best epoch ...
247
+ 2023-10-09 21:17:26,869 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
248
+ 2023-10-09 21:19:11,975
249
+ Results:
250
+ - F-score (micro) 0.4469
251
+ - F-score (macro) 0.3035
252
+ - Accuracy 0.2922
253
+
254
+ By class:
255
+ precision recall f1-score support
256
+
257
+ LOC 0.4817 0.5313 0.5053 1214
258
+ PER 0.4103 0.4728 0.4393 808
259
+ ORG 0.2643 0.2748 0.2694 353
260
+ HumanProd 0.0000 0.0000 0.0000 15
261
+
262
+ micro avg 0.4258 0.4703 0.4469 2390
263
+ macro avg 0.2891 0.3197 0.3035 2390
264
+ weighted avg 0.4224 0.4703 0.4450 2390
265
+
266
+ 2023-10-09 21:19:11,976 ----------------------------------------------------------------------------------------------------