Hugging Face's logo

stefan-it
/
hmbench-newseye-fr-hmbyt5-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-3

Token Classification
Flair
PyTorch
TensorBoard
French
sequence-tagger-model
Model card Files Metrics Community
2
stefan-it commited on
Commit
c2b34d1
·
1 Parent(s): 153b62c

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.1697017911.de2e83fddbee.1120.9 +3 -0
  6. test.tsv +0 -0
  7. training.log +263 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c83c60a6f5035292bf4c643d5ed9b5c385eba41b9a204e49fe9f007af3a5488
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 sha256:7e7525195d9e84eba736efbc5e0508f0f9fe99a054f15204a8cb5e3cac5ab481
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 10:00:49 0.0002 1.2953 0.2372 0.5104 0.5320 0.5210 0.3760
3
+ 2 10:09:58 0.0001 0.1596 0.0998 0.7320 0.7878 0.7588 0.6356
4
+ 3 10:19:08 0.0001 0.0709 0.1063 0.7819 0.8000 0.7909 0.6712
5
+ 4 10:28:16 0.0001 0.0491 0.1310 0.7868 0.8082 0.7973 0.6765
6
+ 5 10:37:25 0.0001 0.0360 0.1431 0.7900 0.8190 0.8043 0.6864
7
+ 6 10:46:11 0.0001 0.0270 0.1740 0.7978 0.8000 0.7989 0.6806
8
+ 7 10:55:04 0.0001 0.0211 0.1712 0.7819 0.8095 0.7955 0.6800
9
+ 8 11:04:09 0.0000 0.0166 0.1897 0.7886 0.8122 0.8003 0.6838
10
+ 9 11:13:13 0.0000 0.0131 0.1976 0.7824 0.8122 0.7971 0.6792
11
+ 10 11:21:46 0.0000 0.0101 0.2045 0.7808 0.8095 0.7949 0.6761
runs/events.out.tfevents.1697017911.de2e83fddbee.1120.9 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c6a3ac55bdfb9cd314839ad8be537fce545147891b211e540721c988e2d7f16
3
+ size 502461
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,263 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-10-11 09:51:51,454 ----------------------------------------------------------------------------------------------------
2
+ 2023-10-11 09:51:51,456 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
+ 2023-10-11 09:51:51,456 ----------------------------------------------------------------------------------------------------
71
+ 2023-10-11 09:51:51,457 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
72
+ - NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
73
+ 2023-10-11 09:51:51,457 ----------------------------------------------------------------------------------------------------
74
+ 2023-10-11 09:51:51,457 Train: 7142 sentences
75
+ 2023-10-11 09:51:51,457 (train_with_dev=False, train_with_test=False)
76
+ 2023-10-11 09:51:51,457 ----------------------------------------------------------------------------------------------------
77
+ 2023-10-11 09:51:51,457 Training Params:
78
+ 2023-10-11 09:51:51,457 - learning_rate: "0.00016"
79
+ 2023-10-11 09:51:51,457 - mini_batch_size: "8"
80
+ 2023-10-11 09:51:51,457 - max_epochs: "10"
81
+ 2023-10-11 09:51:51,457 - shuffle: "True"
82
+ 2023-10-11 09:51:51,457 ----------------------------------------------------------------------------------------------------
83
+ 2023-10-11 09:51:51,457 Plugins:
84
+ 2023-10-11 09:51:51,458 - TensorboardLogger
85
+ 2023-10-11 09:51:51,458 - LinearScheduler | warmup_fraction: '0.1'
86
+ 2023-10-11 09:51:51,458 ----------------------------------------------------------------------------------------------------
87
+ 2023-10-11 09:51:51,458 Final evaluation on model from best epoch (best-model.pt)
88
+ 2023-10-11 09:51:51,458 - metric: "('micro avg', 'f1-score')"
89
+ 2023-10-11 09:51:51,458 ----------------------------------------------------------------------------------------------------
90
+ 2023-10-11 09:51:51,458 Computation:
91
+ 2023-10-11 09:51:51,458 - compute on device: cuda:0
92
+ 2023-10-11 09:51:51,458 - embedding storage: none
93
+ 2023-10-11 09:51:51,458 ----------------------------------------------------------------------------------------------------
94
+ 2023-10-11 09:51:51,458 Model training base path: "hmbench-newseye/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-3"
95
+ 2023-10-11 09:51:51,458 ----------------------------------------------------------------------------------------------------
96
+ 2023-10-11 09:51:51,458 ----------------------------------------------------------------------------------------------------
97
+ 2023-10-11 09:51:51,459 Logging anything other than scalars to TensorBoard is currently not supported.
98
+ 2023-10-11 09:52:43,650 epoch 1 - iter 89/893 - loss 2.81958198 - time (sec): 52.19 - samples/sec: 515.94 - lr: 0.000016 - momentum: 0.000000
99
+ 2023-10-11 09:53:33,684 epoch 1 - iter 178/893 - loss 2.73966772 - time (sec): 102.22 - samples/sec: 495.89 - lr: 0.000032 - momentum: 0.000000
100
+ 2023-10-11 09:54:24,264 epoch 1 - iter 267/893 - loss 2.54471735 - time (sec): 152.80 - samples/sec: 487.65 - lr: 0.000048 - momentum: 0.000000
101
+ 2023-10-11 09:55:14,301 epoch 1 - iter 356/893 - loss 2.32649706 - time (sec): 202.84 - samples/sec: 486.76 - lr: 0.000064 - momentum: 0.000000
102
+ 2023-10-11 09:56:05,159 epoch 1 - iter 445/893 - loss 2.08476844 - time (sec): 253.70 - samples/sec: 492.21 - lr: 0.000080 - momentum: 0.000000
103
+ 2023-10-11 09:56:55,867 epoch 1 - iter 534/893 - loss 1.86965509 - time (sec): 304.41 - samples/sec: 488.69 - lr: 0.000095 - momentum: 0.000000
104
+ 2023-10-11 09:57:49,598 epoch 1 - iter 623/893 - loss 1.68007860 - time (sec): 358.14 - samples/sec: 486.88 - lr: 0.000111 - momentum: 0.000000
105
+ 2023-10-11 09:58:44,484 epoch 1 - iter 712/893 - loss 1.52983259 - time (sec): 413.02 - samples/sec: 481.46 - lr: 0.000127 - momentum: 0.000000
106
+ 2023-10-11 09:59:39,435 epoch 1 - iter 801/893 - loss 1.39878794 - time (sec): 467.97 - samples/sec: 478.76 - lr: 0.000143 - momentum: 0.000000
107
+ 2023-10-11 10:00:28,196 epoch 1 - iter 890/893 - loss 1.29860971 - time (sec): 516.74 - samples/sec: 479.81 - lr: 0.000159 - momentum: 0.000000
108
+ 2023-10-11 10:00:29,735 ----------------------------------------------------------------------------------------------------
109
+ 2023-10-11 10:00:29,736 EPOCH 1 done: loss 1.2953 - lr: 0.000159
110
+ 2023-10-11 10:00:49,291 DEV : loss 0.23720598220825195 - f1-score (micro avg) 0.521
111
+ 2023-10-11 10:00:49,321 saving best model
112
+ 2023-10-11 10:00:50,224 ----------------------------------------------------------------------------------------------------
113
+ 2023-10-11 10:01:41,521 epoch 2 - iter 89/893 - loss 0.24156291 - time (sec): 51.29 - samples/sec: 511.50 - lr: 0.000158 - momentum: 0.000000
114
+ 2023-10-11 10:02:32,514 epoch 2 - iter 178/893 - loss 0.23973758 - time (sec): 102.29 - samples/sec: 503.29 - lr: 0.000156 - momentum: 0.000000
115
+ 2023-10-11 10:03:23,339 epoch 2 - iter 267/893 - loss 0.22533212 - time (sec): 153.11 - samples/sec: 492.16 - lr: 0.000155 - momentum: 0.000000
116
+ 2023-10-11 10:04:15,696 epoch 2 - iter 356/893 - loss 0.20433271 - time (sec): 205.47 - samples/sec: 490.36 - lr: 0.000153 - momentum: 0.000000
117
+ 2023-10-11 10:05:06,103 epoch 2 - iter 445/893 - loss 0.19447693 - time (sec): 255.88 - samples/sec: 486.27 - lr: 0.000151 - momentum: 0.000000
118
+ 2023-10-11 10:06:00,826 epoch 2 - iter 534/893 - loss 0.18500783 - time (sec): 310.60 - samples/sec: 483.67 - lr: 0.000149 - momentum: 0.000000
119
+ 2023-10-11 10:06:55,207 epoch 2 - iter 623/893 - loss 0.17866349 - time (sec): 364.98 - samples/sec: 478.57 - lr: 0.000148 - momentum: 0.000000
120
+ 2023-10-11 10:07:46,050 epoch 2 - iter 712/893 - loss 0.17165749 - time (sec): 415.82 - samples/sec: 476.07 - lr: 0.000146 - momentum: 0.000000
121
+ 2023-10-11 10:08:40,495 epoch 2 - iter 801/893 - loss 0.16565209 - time (sec): 470.27 - samples/sec: 473.21 - lr: 0.000144 - momentum: 0.000000
122
+ 2023-10-11 10:09:35,309 epoch 2 - iter 890/893 - loss 0.15984826 - time (sec): 525.08 - samples/sec: 472.17 - lr: 0.000142 - momentum: 0.000000
123
+ 2023-10-11 10:09:36,886 ----------------------------------------------------------------------------------------------------
124
+ 2023-10-11 10:09:36,887 EPOCH 2 done: loss 0.1596 - lr: 0.000142
125
+ 2023-10-11 10:09:58,231 DEV : loss 0.09981917589902878 - f1-score (micro avg) 0.7588
126
+ 2023-10-11 10:09:58,264 saving best model
127
+ 2023-10-11 10:10:00,886 ----------------------------------------------------------------------------------------------------
128
+ 2023-10-11 10:10:52,118 epoch 3 - iter 89/893 - loss 0.06769040 - time (sec): 51.23 - samples/sec: 465.50 - lr: 0.000140 - momentum: 0.000000
129
+ 2023-10-11 10:11:42,503 epoch 3 - iter 178/893 - loss 0.06730058 - time (sec): 101.61 - samples/sec: 481.27 - lr: 0.000139 - momentum: 0.000000
130
+ 2023-10-11 10:12:32,365 epoch 3 - iter 267/893 - loss 0.06608996 - time (sec): 151.47 - samples/sec: 483.16 - lr: 0.000137 - momentum: 0.000000
131
+ 2023-10-11 10:13:22,493 epoch 3 - iter 356/893 - loss 0.06822392 - time (sec): 201.60 - samples/sec: 487.02 - lr: 0.000135 - momentum: 0.000000
132
+ 2023-10-11 10:14:13,138 epoch 3 - iter 445/893 - loss 0.07122185 - time (sec): 252.25 - samples/sec: 487.15 - lr: 0.000133 - momentum: 0.000000
133
+ 2023-10-11 10:15:03,603 epoch 3 - iter 534/893 - loss 0.07293485 - time (sec): 302.71 - samples/sec: 487.29 - lr: 0.000132 - momentum: 0.000000
134
+ 2023-10-11 10:15:59,820 epoch 3 - iter 623/893 - loss 0.07412771 - time (sec): 358.93 - samples/sec: 485.22 - lr: 0.000130 - momentum: 0.000000
135
+ 2023-10-11 10:16:54,820 epoch 3 - iter 712/893 - loss 0.07275147 - time (sec): 413.93 - samples/sec: 478.54 - lr: 0.000128 - momentum: 0.000000
136
+ 2023-10-11 10:17:50,647 epoch 3 - iter 801/893 - loss 0.07101037 - time (sec): 469.76 - samples/sec: 475.21 - lr: 0.000126 - momentum: 0.000000
137
+ 2023-10-11 10:18:45,676 epoch 3 - iter 890/893 - loss 0.07076952 - time (sec): 524.78 - samples/sec: 472.85 - lr: 0.000125 - momentum: 0.000000
138
+ 2023-10-11 10:18:47,139 ----------------------------------------------------------------------------------------------------
139
+ 2023-10-11 10:18:47,140 EPOCH 3 done: loss 0.0709 - lr: 0.000125
140
+ 2023-10-11 10:19:08,915 DEV : loss 0.10631939768791199 - f1-score (micro avg) 0.7909
141
+ 2023-10-11 10:19:08,960 saving best model
142
+ 2023-10-11 10:19:11,681 ----------------------------------------------------------------------------------------------------
143
+ 2023-10-11 10:20:04,148 epoch 4 - iter 89/893 - loss 0.04488972 - time (sec): 52.46 - samples/sec: 457.96 - lr: 0.000123 - momentum: 0.000000
144
+ 2023-10-11 10:20:59,199 epoch 4 - iter 178/893 - loss 0.04859362 - time (sec): 107.51 - samples/sec: 455.92 - lr: 0.000121 - momentum: 0.000000
145
+ 2023-10-11 10:21:55,697 epoch 4 - iter 267/893 - loss 0.04714332 - time (sec): 164.01 - samples/sec: 461.00 - lr: 0.000119 - momentum: 0.000000
146
+ 2023-10-11 10:22:50,160 epoch 4 - iter 356/893 - loss 0.04851860 - time (sec): 218.47 - samples/sec: 458.60 - lr: 0.000117 - momentum: 0.000000
147
+ 2023-10-11 10:23:41,893 epoch 4 - iter 445/893 - loss 0.04963523 - time (sec): 270.21 - samples/sec: 465.48 - lr: 0.000116 - momentum: 0.000000
148
+ 2023-10-11 10:24:30,632 epoch 4 - iter 534/893 - loss 0.05027391 - time (sec): 318.95 - samples/sec: 466.00 - lr: 0.000114 - momentum: 0.000000
149
+ 2023-10-11 10:25:19,741 epoch 4 - iter 623/893 - loss 0.05064819 - time (sec): 368.06 - samples/sec: 472.03 - lr: 0.000112 - momentum: 0.000000
150
+ 2023-10-11 10:26:11,200 epoch 4 - iter 712/893 - loss 0.05119551 - time (sec): 419.51 - samples/sec: 471.86 - lr: 0.000110 - momentum: 0.000000
151
+ 2023-10-11 10:27:01,611 epoch 4 - iter 801/893 - loss 0.05054556 - time (sec): 469.93 - samples/sec: 474.73 - lr: 0.000109 - momentum: 0.000000
152
+ 2023-10-11 10:27:52,358 epoch 4 - iter 890/893 - loss 0.04922809 - time (sec): 520.67 - samples/sec: 476.32 - lr: 0.000107 - momentum: 0.000000
153
+ 2023-10-11 10:27:53,878 ----------------------------------------------------------------------------------------------------
154
+ 2023-10-11 10:27:53,879 EPOCH 4 done: loss 0.0491 - lr: 0.000107
155
+ 2023-10-11 10:28:16,061 DEV : loss 0.13096819818019867 - f1-score (micro avg) 0.7973
156
+ 2023-10-11 10:28:16,097 saving best model
157
+ 2023-10-11 10:28:18,714 ----------------------------------------------------------------------------------------------------
158
+ 2023-10-11 10:29:14,514 epoch 5 - iter 89/893 - loss 0.03067518 - time (sec): 55.79 - samples/sec: 450.15 - lr: 0.000105 - momentum: 0.000000
159
+ 2023-10-11 10:30:10,324 epoch 5 - iter 178/893 - loss 0.03624382 - time (sec): 111.61 - samples/sec: 453.36 - lr: 0.000103 - momentum: 0.000000
160
+ 2023-10-11 10:31:01,711 epoch 5 - iter 267/893 - loss 0.03352199 - time (sec): 162.99 - samples/sec: 464.01 - lr: 0.000101 - momentum: 0.000000
161
+ 2023-10-11 10:31:50,487 epoch 5 - iter 356/893 - loss 0.03375946 - time (sec): 211.77 - samples/sec: 470.97 - lr: 0.000100 - momentum: 0.000000
162
+ 2023-10-11 10:32:40,342 epoch 5 - iter 445/893 - loss 0.03395690 - time (sec): 261.62 - samples/sec: 472.45 - lr: 0.000098 - momentum: 0.000000
163
+ 2023-10-11 10:33:31,750 epoch 5 - iter 534/893 - loss 0.03542389 - time (sec): 313.03 - samples/sec: 471.58 - lr: 0.000096 - momentum: 0.000000
164
+ 2023-10-11 10:34:28,397 epoch 5 - iter 623/893 - loss 0.03472368 - time (sec): 369.68 - samples/sec: 467.82 - lr: 0.000094 - momentum: 0.000000
165
+ 2023-10-11 10:35:18,765 epoch 5 - iter 712/893 - loss 0.03520394 - time (sec): 420.05 - samples/sec: 469.87 - lr: 0.000093 - momentum: 0.000000
166
+ 2023-10-11 10:36:13,262 epoch 5 - iter 801/893 - loss 0.03464309 - time (sec): 474.54 - samples/sec: 468.46 - lr: 0.000091 - momentum: 0.000000
167
+ 2023-10-11 10:37:02,901 epoch 5 - iter 890/893 - loss 0.03596866 - time (sec): 524.18 - samples/sec: 473.24 - lr: 0.000089 - momentum: 0.000000
168
+ 2023-10-11 10:37:04,415 ----------------------------------------------------------------------------------------------------
169
+ 2023-10-11 10:37:04,415 EPOCH 5 done: loss 0.0360 - lr: 0.000089
170
+ 2023-10-11 10:37:25,634 DEV : loss 0.14307744801044464 - f1-score (micro avg) 0.8043
171
+ 2023-10-11 10:37:25,663 saving best model
172
+ 2023-10-11 10:37:28,263 ----------------------------------------------------------------------------------------------------
173
+ 2023-10-11 10:38:18,809 epoch 6 - iter 89/893 - loss 0.03174465 - time (sec): 50.54 - samples/sec: 514.67 - lr: 0.000087 - momentum: 0.000000
174
+ 2023-10-11 10:39:07,991 epoch 6 - iter 178/893 - loss 0.02928006 - time (sec): 99.72 - samples/sec: 497.76 - lr: 0.000085 - momentum: 0.000000
175
+ 2023-10-11 10:39:57,896 epoch 6 - iter 267/893 - loss 0.02968319 - time (sec): 149.63 - samples/sec: 492.29 - lr: 0.000084 - momentum: 0.000000
176
+ 2023-10-11 10:40:49,294 epoch 6 - iter 356/893 - loss 0.02835877 - time (sec): 201.03 - samples/sec: 493.58 - lr: 0.000082 - momentum: 0.000000
177
+ 2023-10-11 10:41:40,128 epoch 6 - iter 445/893 - loss 0.02693656 - time (sec): 251.86 - samples/sec: 491.28 - lr: 0.000080 - momentum: 0.000000
178
+ 2023-10-11 10:42:30,558 epoch 6 - iter 534/893 - loss 0.02653119 - time (sec): 302.29 - samples/sec: 486.48 - lr: 0.000078 - momentum: 0.000000
179
+ 2023-10-11 10:43:21,338 epoch 6 - iter 623/893 - loss 0.02667286 - time (sec): 353.07 - samples/sec: 486.64 - lr: 0.000077 - momentum: 0.000000
180
+ 2023-10-11 10:44:12,346 epoch 6 - iter 712/893 - loss 0.02748104 - time (sec): 404.08 - samples/sec: 490.38 - lr: 0.000075 - momentum: 0.000000
181
+ 2023-10-11 10:45:00,449 epoch 6 - iter 801/893 - loss 0.02718055 - time (sec): 452.18 - samples/sec: 493.59 - lr: 0.000073 - momentum: 0.000000
182
+ 2023-10-11 10:45:49,594 epoch 6 - iter 890/893 - loss 0.02706511 - time (sec): 501.33 - samples/sec: 494.82 - lr: 0.000071 - momentum: 0.000000
183
+ 2023-10-11 10:45:51,093 ----------------------------------------------------------------------------------------------------
184
+ 2023-10-11 10:45:51,093 EPOCH 6 done: loss 0.0270 - lr: 0.000071
185
+ 2023-10-11 10:46:11,841 DEV : loss 0.1740272492170334 - f1-score (micro avg) 0.7989
186
+ 2023-10-11 10:46:11,871 ----------------------------------------------------------------------------------------------------
187
+ 2023-10-11 10:46:59,408 epoch 7 - iter 89/893 - loss 0.02529529 - time (sec): 47.53 - samples/sec: 515.71 - lr: 0.000069 - momentum: 0.000000
188
+ 2023-10-11 10:47:48,889 epoch 7 - iter 178/893 - loss 0.02618348 - time (sec): 97.02 - samples/sec: 492.54 - lr: 0.000068 - momentum: 0.000000
189
+ 2023-10-11 10:48:38,830 epoch 7 - iter 267/893 - loss 0.02283278 - time (sec): 146.96 - samples/sec: 497.17 - lr: 0.000066 - momentum: 0.000000
190
+ 2023-10-11 10:49:27,859 epoch 7 - iter 356/893 - loss 0.02247875 - time (sec): 195.99 - samples/sec: 495.73 - lr: 0.000064 - momentum: 0.000000
191
+ 2023-10-11 10:50:20,268 epoch 7 - iter 445/893 - loss 0.02349580 - time (sec): 248.39 - samples/sec: 492.55 - lr: 0.000062 - momentum: 0.000000
192
+ 2023-10-11 10:51:11,607 epoch 7 - iter 534/893 - loss 0.02280138 - time (sec): 299.73 - samples/sec: 493.64 - lr: 0.000061 - momentum: 0.000000
193
+ 2023-10-11 10:52:02,733 epoch 7 - iter 623/893 - loss 0.02188778 - time (sec): 350.86 - samples/sec: 493.14 - lr: 0.000059 - momentum: 0.000000
194
+ 2023-10-11 10:52:55,211 epoch 7 - iter 712/893 - loss 0.02139551 - time (sec): 403.34 - samples/sec: 491.20 - lr: 0.000057 - momentum: 0.000000
195
+ 2023-10-11 10:53:47,078 epoch 7 - iter 801/893 - loss 0.02130839 - time (sec): 455.20 - samples/sec: 490.46 - lr: 0.000055 - momentum: 0.000000
196
+ 2023-10-11 10:54:40,102 epoch 7 - iter 890/893 - loss 0.02103175 - time (sec): 508.23 - samples/sec: 488.30 - lr: 0.000053 - momentum: 0.000000
197
+ 2023-10-11 10:54:41,689 ----------------------------------------------------------------------------------------------------
198
+ 2023-10-11 10:54:41,689 EPOCH 7 done: loss 0.0211 - lr: 0.000053
199
+ 2023-10-11 10:55:04,149 DEV : loss 0.17123691737651825 - f1-score (micro avg) 0.7955
200
+ 2023-10-11 10:55:04,179 ----------------------------------------------------------------------------------------------------
201
+ 2023-10-11 10:55:54,180 epoch 8 - iter 89/893 - loss 0.01527667 - time (sec): 50.00 - samples/sec: 493.94 - lr: 0.000052 - momentum: 0.000000
202
+ 2023-10-11 10:56:44,974 epoch 8 - iter 178/893 - loss 0.01586546 - time (sec): 100.79 - samples/sec: 490.48 - lr: 0.000050 - momentum: 0.000000
203
+ 2023-10-11 10:57:38,893 epoch 8 - iter 267/893 - loss 0.01414335 - time (sec): 154.71 - samples/sec: 470.32 - lr: 0.000048 - momentum: 0.000000
204
+ 2023-10-11 10:58:32,085 epoch 8 - iter 356/893 - loss 0.01357556 - time (sec): 207.90 - samples/sec: 462.77 - lr: 0.000046 - momentum: 0.000000
205
+ 2023-10-11 10:59:27,665 epoch 8 - iter 445/893 - loss 0.01506294 - time (sec): 263.48 - samples/sec: 455.96 - lr: 0.000045 - momentum: 0.000000
206
+ 2023-10-11 11:00:20,764 epoch 8 - iter 534/893 - loss 0.01614469 - time (sec): 316.58 - samples/sec: 462.89 - lr: 0.000043 - momentum: 0.000000
207
+ 2023-10-11 11:01:12,150 epoch 8 - iter 623/893 - loss 0.01576585 - time (sec): 367.97 - samples/sec: 468.57 - lr: 0.000041 - momentum: 0.000000
208
+ 2023-10-11 11:02:04,707 epoch 8 - iter 712/893 - loss 0.01631603 - time (sec): 420.53 - samples/sec: 473.12 - lr: 0.000039 - momentum: 0.000000
209
+ 2023-10-11 11:02:56,711 epoch 8 - iter 801/893 - loss 0.01698976 - time (sec): 472.53 - samples/sec: 474.98 - lr: 0.000037 - momentum: 0.000000
210
+ 2023-10-11 11:03:46,712 epoch 8 - iter 890/893 - loss 0.01660442 - time (sec): 522.53 - samples/sec: 474.80 - lr: 0.000036 - momentum: 0.000000
211
+ 2023-10-11 11:03:48,175 ----------------------------------------------------------------------------------------------------
212
+ 2023-10-11 11:03:48,176 EPOCH 8 done: loss 0.0166 - lr: 0.000036
213
+ 2023-10-11 11:04:09,326 DEV : loss 0.1897462159395218 - f1-score (micro avg) 0.8003
214
+ 2023-10-11 11:04:09,356 ----------------------------------------------------------------------------------------------------
215
+ 2023-10-11 11:04:59,464 epoch 9 - iter 89/893 - loss 0.01141607 - time (sec): 50.11 - samples/sec: 476.01 - lr: 0.000034 - momentum: 0.000000
216
+ 2023-10-11 11:05:49,221 epoch 9 - iter 178/893 - loss 0.01061287 - time (sec): 99.86 - samples/sec: 467.98 - lr: 0.000032 - momentum: 0.000000
217
+ 2023-10-11 11:06:37,768 epoch 9 - iter 267/893 - loss 0.01141794 - time (sec): 148.41 - samples/sec: 463.59 - lr: 0.000030 - momentum: 0.000000
218
+ 2023-10-11 11:07:31,411 epoch 9 - iter 356/893 - loss 0.01110773 - time (sec): 202.05 - samples/sec: 469.68 - lr: 0.000029 - momentum: 0.000000
219
+ 2023-10-11 11:08:27,558 epoch 9 - iter 445/893 - loss 0.01210552 - time (sec): 258.20 - samples/sec: 465.79 - lr: 0.000027 - momentum: 0.000000
220
+ 2023-10-11 11:09:21,450 epoch 9 - iter 534/893 - loss 0.01286214 - time (sec): 312.09 - samples/sec: 468.30 - lr: 0.000025 - momentum: 0.000000
221
+ 2023-10-11 11:10:14,841 epoch 9 - iter 623/893 - loss 0.01302278 - time (sec): 365.48 - samples/sec: 472.28 - lr: 0.000023 - momentum: 0.000000
222
+ 2023-10-11 11:11:06,931 epoch 9 - iter 712/893 - loss 0.01261530 - time (sec): 417.57 - samples/sec: 475.71 - lr: 0.000022 - momentum: 0.000000
223
+ 2023-10-11 11:11:59,563 epoch 9 - iter 801/893 - loss 0.01305653 - time (sec): 470.21 - samples/sec: 475.44 - lr: 0.000020 - momentum: 0.000000
224
+ 2023-10-11 11:12:49,808 epoch 9 - iter 890/893 - loss 0.01318959 - time (sec): 520.45 - samples/sec: 475.89 - lr: 0.000018 - momentum: 0.000000
225
+ 2023-10-11 11:12:51,530 ----------------------------------------------------------------------------------------------------
226
+ 2023-10-11 11:12:51,530 EPOCH 9 done: loss 0.0131 - lr: 0.000018
227
+ 2023-10-11 11:13:13,077 DEV : loss 0.19762861728668213 - f1-score (micro avg) 0.7971
228
+ 2023-10-11 11:13:13,107 ----------------------------------------------------------------------------------------------------
229
+ 2023-10-11 11:14:01,040 epoch 10 - iter 89/893 - loss 0.00903721 - time (sec): 47.93 - samples/sec: 520.50 - lr: 0.000016 - momentum: 0.000000
230
+ 2023-10-11 11:14:48,169 epoch 10 - iter 178/893 - loss 0.01155862 - time (sec): 95.06 - samples/sec: 506.09 - lr: 0.000014 - momentum: 0.000000
231
+ 2023-10-11 11:15:36,368 epoch 10 - iter 267/893 - loss 0.01117476 - time (sec): 143.26 - samples/sec: 507.88 - lr: 0.000013 - momentum: 0.000000
232
+ 2023-10-11 11:16:25,236 epoch 10 - iter 356/893 - loss 0.01092791 - time (sec): 192.13 - samples/sec: 511.73 - lr: 0.000011 - momentum: 0.000000
233
+ 2023-10-11 11:17:15,043 epoch 10 - iter 445/893 - loss 0.01104186 - time (sec): 241.93 - samples/sec: 513.40 - lr: 0.000009 - momentum: 0.000000
234
+ 2023-10-11 11:18:03,341 epoch 10 - iter 534/893 - loss 0.01048931 - time (sec): 290.23 - samples/sec: 512.33 - lr: 0.000007 - momentum: 0.000000
235
+ 2023-10-11 11:18:53,582 epoch 10 - iter 623/893 - loss 0.01050700 - time (sec): 340.47 - samples/sec: 507.21 - lr: 0.000006 - momentum: 0.000000
236
+ 2023-10-11 11:19:43,252 epoch 10 - iter 712/893 - loss 0.01002609 - time (sec): 390.14 - samples/sec: 506.48 - lr: 0.000004 - momentum: 0.000000
237
+ 2023-10-11 11:20:33,114 epoch 10 - iter 801/893 - loss 0.00990462 - time (sec): 440.01 - samples/sec: 505.61 - lr: 0.000002 - momentum: 0.000000
238
+ 2023-10-11 11:21:24,177 epoch 10 - iter 890/893 - loss 0.01009786 - time (sec): 491.07 - samples/sec: 505.52 - lr: 0.000000 - momentum: 0.000000
239
+ 2023-10-11 11:21:25,540 ----------------------------------------------------------------------------------------------------
240
+ 2023-10-11 11:21:25,540 EPOCH 10 done: loss 0.0101 - lr: 0.000000
241
+ 2023-10-11 11:21:46,902 DEV : loss 0.2045479267835617 - f1-score (micro avg) 0.7949
242
+ 2023-10-11 11:21:47,835 ----------------------------------------------------------------------------------------------------
243
+ 2023-10-11 11:21:47,837 Loading model from best epoch ...
244
+ 2023-10-11 11:21:51,551 SequenceTagger predicts: Dictionary with 17 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
245
+ 2023-10-11 11:23:00,255
246
+ Results:
247
+ - F-score (micro) 0.6987
248
+ - F-score (macro) 0.6131
249
+ - Accuracy 0.5529
250
+
251
+ By class:
252
+ precision recall f1-score support
253
+
254
+ LOC 0.6829 0.7315 0.7063 1095
255
+ PER 0.7810 0.7717 0.7763 1012
256
+ ORG 0.4562 0.5686 0.5062 357
257
+ HumanProd 0.3878 0.5758 0.4634 33
258
+
259
+ micro avg 0.6764 0.7225 0.6987 2497
260
+ macro avg 0.5769 0.6619 0.6131 2497
261
+ weighted avg 0.6863 0.7225 0.7029 2497
262
+
263
+ 2023-10-11 11:23:00,255 ----------------------------------------------------------------------------------------------------