Hugging Face's logo

Spaces:
NLTM
/
pos-hindi
No application file

App Files Community
divyanshusingh commited on
Commit
627e1a9
·
1 Parent(s): 5225f10

Add pos-hindi

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. test.tsv +0 -0
  6. training.log +268 -0
  7. weights.txt +0 -0
best-model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4f447009bfe2337a3e9aa5507f6cd5c8f15e51961bf4294c2d8b1e56bcc9e37
3
+ size 1543466104
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 sha256:049e6ca758a17ddf6a8fbe1250ffe009d99ea0ba75ab39ae58c5b180a74cec4e
3
+ size 1543466104
loss.tsv ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ EPOCH TIMESTAMP BAD_EPOCHS LEARNING_RATE TRAIN_LOSS DEV_LOSS DEV_PRECISION DEV_RECALL DEV_F1 DEV_ACCURACY
2
+ 1 10:58:47 0 0.1000 0.4972969775912337 0.16205480694770813 0.9458 0.9458 0.9458 0.9458
3
+ 2 10:59:40 0 0.1000 0.21439881967892635 0.11246180534362793 0.9578 0.9578 0.9578 0.9578
4
+ 3 11:00:31 0 0.1000 0.17297799300942512 0.09120035171508789 0.9647 0.9647 0.9647 0.9647
5
+ 4 11:01:25 0 0.1000 0.1503387751833925 0.08160468190908432 0.9671 0.9671 0.9671 0.9671
6
+ 5 11:02:22 0 0.1000 0.13411458622883882 0.07825736701488495 0.9683 0.9683 0.9683 0.9683
7
+ 6 11:03:15 0 0.1000 0.1262321389238882 0.0752514973282814 0.97 0.97 0.97 0.97
8
+ 7 11:04:07 1 0.1000 0.11718843640647182 0.07426313310861588 0.9696 0.9696 0.9696 0.9696
9
+ 8 11:04:54 0 0.1000 0.11085070710745507 0.07048413157463074 0.9713 0.9713 0.9713 0.9713
10
+ 9 11:05:44 0 0.1000 0.10515870564345044 0.07029640674591064 0.9719 0.9719 0.9719 0.9719
11
+ 10 11:06:35 1 0.1000 0.10052994074044588 0.07123567163944244 0.9716 0.9716 0.9716 0.9716
test.tsv ADDED
The diff for this file is too large to render. See raw diff
 
training.log ADDED
@@ -0,0 +1,268 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2022-08-24 10:57:35,091 ----------------------------------------------------------------------------------------------------
2
+ 2022-08-24 10:57:35,092 Model: "SequenceTagger(
3
+ (embeddings): StackedEmbeddings(
4
+ (list_embedding_0): WordEmbeddings(
5
+ 'hindi'
6
+ (embedding): Embedding(1000000, 300)
7
+ )
8
+ (list_embedding_1): FlairEmbeddings(
9
+ (lm): LanguageModel(
10
+ (drop): Dropout(p=0.1, inplace=False)
11
+ (encoder): Embedding(3520, 100)
12
+ (rnn): LSTM(100, 2048)
13
+ (decoder): Linear(in_features=2048, out_features=3520, bias=True)
14
+ )
15
+ )
16
+ (list_embedding_2): FlairEmbeddings(
17
+ (lm): LanguageModel(
18
+ (drop): Dropout(p=0.1, inplace=False)
19
+ (encoder): Embedding(3520, 100)
20
+ (rnn): LSTM(100, 2048)
21
+ (decoder): Linear(in_features=2048, out_features=3520, bias=True)
22
+ )
23
+ )
24
+ )
25
+ (word_dropout): WordDropout(p=0.05)
26
+ (locked_dropout): LockedDropout(p=0.5)
27
+ (embedding2nn): Linear(in_features=4396, out_features=4396, bias=True)
28
+ (rnn): LSTM(4396, 256, batch_first=True, bidirectional=True)
29
+ (linear): Linear(in_features=512, out_features=34, bias=True)
30
+ (loss_function): ViterbiLoss()
31
+ (crf): CRF()
32
+ )"
33
+ 2022-08-24 10:57:35,092 ----------------------------------------------------------------------------------------------------
34
+ 2022-08-24 10:57:35,093 Corpus: "Corpus: 13304 train + 1659 dev + 1684 test sentences"
35
+ 2022-08-24 10:57:35,093 ----------------------------------------------------------------------------------------------------
36
+ 2022-08-24 10:57:35,093 Parameters:
37
+ 2022-08-24 10:57:35,094 - learning_rate: "0.100000"
38
+ 2022-08-24 10:57:35,094 - mini_batch_size: "32"
39
+ 2022-08-24 10:57:35,094 - patience: "3"
40
+ 2022-08-24 10:57:35,095 - anneal_factor: "0.5"
41
+ 2022-08-24 10:57:35,095 - max_epochs: "10"
42
+ 2022-08-24 10:57:35,095 - shuffle: "True"
43
+ 2022-08-24 10:57:35,096 - train_with_dev: "False"
44
+ 2022-08-24 10:57:35,096 - batch_growth_annealing: "False"
45
+ 2022-08-24 10:57:35,097 ----------------------------------------------------------------------------------------------------
46
+ 2022-08-24 10:57:35,097 Model training base path: "resources/taggers"
47
+ 2022-08-24 10:57:35,097 ----------------------------------------------------------------------------------------------------
48
+ 2022-08-24 10:57:35,130 Device: cuda:0
49
+ 2022-08-24 10:57:35,131 ----------------------------------------------------------------------------------------------------
50
+ 2022-08-24 10:57:35,131 Embeddings storage mode: cpu
51
+ 2022-08-24 10:57:35,131 ----------------------------------------------------------------------------------------------------
52
+ 2022-08-24 10:57:43,221 epoch 1 - iter 41/416 - loss 1.75374784 - samples/sec: 163.39 - lr: 0.100000
53
+ 2022-08-24 10:57:48,774 epoch 1 - iter 82/416 - loss 1.21144958 - samples/sec: 236.39 - lr: 0.100000
54
+ 2022-08-24 10:57:54,449 epoch 1 - iter 123/416 - loss 0.96022365 - samples/sec: 231.30 - lr: 0.100000
55
+ 2022-08-24 10:58:00,063 epoch 1 - iter 164/416 - loss 0.81474244 - samples/sec: 233.84 - lr: 0.100000
56
+ 2022-08-24 10:58:06,071 epoch 1 - iter 205/416 - loss 0.72188458 - samples/sec: 218.46 - lr: 0.100000
57
+ 2022-08-24 10:58:12,045 epoch 1 - iter 246/416 - loss 0.65566787 - samples/sec: 222.79 - lr: 0.100000
58
+ 2022-08-24 10:58:17,908 epoch 1 - iter 287/416 - loss 0.60358770 - samples/sec: 223.89 - lr: 0.100000
59
+ 2022-08-24 10:58:24,130 epoch 1 - iter 328/416 - loss 0.56015346 - samples/sec: 210.95 - lr: 0.100000
60
+ 2022-08-24 10:58:30,431 epoch 1 - iter 369/416 - loss 0.52768132 - samples/sec: 208.36 - lr: 0.100000
61
+ 2022-08-24 10:58:36,388 epoch 1 - iter 410/416 - loss 0.50102804 - samples/sec: 220.39 - lr: 0.100000
62
+ 2022-08-24 10:58:37,307 ----------------------------------------------------------------------------------------------------
63
+ 2022-08-24 10:58:37,308 EPOCH 1 done: loss 0.4973 - lr 0.100000
64
+ 2022-08-24 10:58:47,137 Evaluating as a multi-label problem: False
65
+ 2022-08-24 10:58:47,698 DEV : loss 0.16205480694770813 - f1-score (micro avg) 0.9458
66
+ 2022-08-24 10:58:47,863 BAD EPOCHS (no improvement): 0
67
+ 2022-08-24 10:58:47,865 saving best model
68
+ 2022-08-24 10:58:51,789 ----------------------------------------------------------------------------------------------------
69
+ 2022-08-24 10:58:55,679 epoch 2 - iter 41/416 - loss 0.24514879 - samples/sec: 343.22 - lr: 0.100000
70
+ 2022-08-24 10:58:59,592 epoch 2 - iter 82/416 - loss 0.24050334 - samples/sec: 335.61 - lr: 0.100000
71
+ 2022-08-24 10:59:03,458 epoch 2 - iter 123/416 - loss 0.23347290 - samples/sec: 339.70 - lr: 0.100000
72
+ 2022-08-24 10:59:07,545 epoch 2 - iter 164/416 - loss 0.23151347 - samples/sec: 322.37 - lr: 0.100000
73
+ 2022-08-24 10:59:11,970 epoch 2 - iter 205/416 - loss 0.22632911 - samples/sec: 299.49 - lr: 0.100000
74
+ 2022-08-24 10:59:15,668 epoch 2 - iter 246/416 - loss 0.22429595 - samples/sec: 355.20 - lr: 0.100000
75
+ 2022-08-24 10:59:19,397 epoch 2 - iter 287/416 - loss 0.22215739 - samples/sec: 352.18 - lr: 0.100000
76
+ 2022-08-24 10:59:23,375 epoch 2 - iter 328/416 - loss 0.21962025 - samples/sec: 330.16 - lr: 0.100000
77
+ 2022-08-24 10:59:27,423 epoch 2 - iter 369/416 - loss 0.21729297 - samples/sec: 324.45 - lr: 0.100000
78
+ 2022-08-24 10:59:31,761 epoch 2 - iter 410/416 - loss 0.21444959 - samples/sec: 302.89 - lr: 0.100000
79
+ 2022-08-24 10:59:32,350 ----------------------------------------------------------------------------------------------------
80
+ 2022-08-24 10:59:32,350 EPOCH 2 done: loss 0.2144 - lr 0.100000
81
+ 2022-08-24 10:59:39,506 Evaluating as a multi-label problem: False
82
+ 2022-08-24 10:59:40,286 DEV : loss 0.11246180534362793 - f1-score (micro avg) 0.9578
83
+ 2022-08-24 10:59:40,475 BAD EPOCHS (no improvement): 0
84
+ 2022-08-24 10:59:40,477 saving best model
85
+ 2022-08-24 10:59:44,103 ----------------------------------------------------------------------------------------------------
86
+ 2022-08-24 10:59:48,468 epoch 3 - iter 41/416 - loss 0.18469601 - samples/sec: 301.11 - lr: 0.100000
87
+ 2022-08-24 10:59:52,504 epoch 3 - iter 82/416 - loss 0.18261678 - samples/sec: 325.54 - lr: 0.100000
88
+ 2022-08-24 10:59:56,236 epoch 3 - iter 123/416 - loss 0.18163090 - samples/sec: 352.04 - lr: 0.100000
89
+ 2022-08-24 11:00:00,041 epoch 3 - iter 164/416 - loss 0.17863873 - samples/sec: 345.20 - lr: 0.100000
90
+ 2022-08-24 11:00:03,939 epoch 3 - iter 205/416 - loss 0.17689939 - samples/sec: 336.91 - lr: 0.100000
91
+ 2022-08-24 11:00:07,952 epoch 3 - iter 246/416 - loss 0.17651204 - samples/sec: 327.26 - lr: 0.100000
92
+ 2022-08-24 11:00:11,880 epoch 3 - iter 287/416 - loss 0.17610330 - samples/sec: 334.48 - lr: 0.100000
93
+ 2022-08-24 11:00:15,861 epoch 3 - iter 328/416 - loss 0.17480990 - samples/sec: 329.95 - lr: 0.100000
94
+ 2022-08-24 11:00:19,654 epoch 3 - iter 369/416 - loss 0.17340667 - samples/sec: 346.31 - lr: 0.100000
95
+ 2022-08-24 11:00:23,590 epoch 3 - iter 410/416 - loss 0.17286643 - samples/sec: 333.65 - lr: 0.100000
96
+ 2022-08-24 11:00:24,178 ----------------------------------------------------------------------------------------------------
97
+ 2022-08-24 11:00:24,179 EPOCH 3 done: loss 0.1730 - lr 0.100000
98
+ 2022-08-24 11:00:31,344 Evaluating as a multi-label problem: False
99
+ 2022-08-24 11:00:31,586 DEV : loss 0.09120035171508789 - f1-score (micro avg) 0.9647
100
+ 2022-08-24 11:00:31,765 BAD EPOCHS (no improvement): 0
101
+ 2022-08-24 11:00:31,767 saving best model
102
+ 2022-08-24 11:00:35,385 ----------------------------------------------------------------------------------------------------
103
+ 2022-08-24 11:00:39,594 epoch 4 - iter 41/416 - loss 0.16103536 - samples/sec: 312.15 - lr: 0.100000
104
+ 2022-08-24 11:00:43,611 epoch 4 - iter 82/416 - loss 0.15827466 - samples/sec: 327.05 - lr: 0.100000
105
+ 2022-08-24 11:00:47,595 epoch 4 - iter 123/416 - loss 0.15368275 - samples/sec: 329.72 - lr: 0.100000
106
+ 2022-08-24 11:00:51,430 epoch 4 - iter 164/416 - loss 0.15380442 - samples/sec: 342.46 - lr: 0.100000
107
+ 2022-08-24 11:00:55,373 epoch 4 - iter 205/416 - loss 0.15360279 - samples/sec: 333.13 - lr: 0.100000
108
+ 2022-08-24 11:00:59,206 epoch 4 - iter 246/416 - loss 0.15416655 - samples/sec: 345.39 - lr: 0.100000
109
+ 2022-08-24 11:01:03,077 epoch 4 - iter 287/416 - loss 0.15297728 - samples/sec: 346.54 - lr: 0.100000
110
+ 2022-08-24 11:01:07,189 epoch 4 - iter 328/416 - loss 0.15273117 - samples/sec: 319.42 - lr: 0.100000
111
+ 2022-08-24 11:01:11,111 epoch 4 - iter 369/416 - loss 0.15136905 - samples/sec: 334.89 - lr: 0.100000
112
+ 2022-08-24 11:01:14,964 epoch 4 - iter 410/416 - loss 0.15052223 - samples/sec: 340.79 - lr: 0.100000
113
+ 2022-08-24 11:01:15,525 ----------------------------------------------------------------------------------------------------
114
+ 2022-08-24 11:01:15,526 EPOCH 4 done: loss 0.1503 - lr 0.100000
115
+ 2022-08-24 11:01:24,855 Evaluating as a multi-label problem: False
116
+ 2022-08-24 11:01:25,108 DEV : loss 0.08160468190908432 - f1-score (micro avg) 0.9671
117
+ 2022-08-24 11:01:25,259 BAD EPOCHS (no improvement): 0
118
+ 2022-08-24 11:01:25,261 saving best model
119
+ 2022-08-24 11:01:28,924 ----------------------------------------------------------------------------------------------------
120
+ 2022-08-24 11:01:33,139 epoch 5 - iter 41/416 - loss 0.13452886 - samples/sec: 313.17 - lr: 0.100000
121
+ 2022-08-24 11:01:36,957 epoch 5 - iter 82/416 - loss 0.13710874 - samples/sec: 344.08 - lr: 0.100000
122
+ 2022-08-24 11:01:40,703 epoch 5 - iter 123/416 - loss 0.13645113 - samples/sec: 350.66 - lr: 0.100000
123
+ 2022-08-24 11:01:44,909 epoch 5 - iter 164/416 - loss 0.13713570 - samples/sec: 312.19 - lr: 0.100000
124
+ 2022-08-24 11:01:48,859 epoch 5 - iter 205/416 - loss 0.13547128 - samples/sec: 332.51 - lr: 0.100000
125
+ 2022-08-24 11:01:52,795 epoch 5 - iter 246/416 - loss 0.13505763 - samples/sec: 333.65 - lr: 0.100000
126
+ 2022-08-24 11:01:56,601 epoch 5 - iter 287/416 - loss 0.13405671 - samples/sec: 345.07 - lr: 0.100000
127
+ 2022-08-24 11:02:00,371 epoch 5 - iter 328/416 - loss 0.13335547 - samples/sec: 348.32 - lr: 0.100000
128
+ 2022-08-24 11:02:04,122 epoch 5 - iter 369/416 - loss 0.13393736 - samples/sec: 350.13 - lr: 0.100000
129
+ 2022-08-24 11:02:07,887 epoch 5 - iter 410/416 - loss 0.13423791 - samples/sec: 348.83 - lr: 0.100000
130
+ 2022-08-24 11:02:08,432 ----------------------------------------------------------------------------------------------------
131
+ 2022-08-24 11:02:08,433 EPOCH 5 done: loss 0.1341 - lr 0.100000
132
+ 2022-08-24 11:02:22,185 Evaluating as a multi-label problem: False
133
+ 2022-08-24 11:02:22,641 DEV : loss 0.07825736701488495 - f1-score (micro avg) 0.9683
134
+ 2022-08-24 11:02:22,794 BAD EPOCHS (no improvement): 0
135
+ 2022-08-24 11:02:22,796 saving best model
136
+ 2022-08-24 11:02:26,511 ----------------------------------------------------------------------------------------------------
137
+ 2022-08-24 11:02:30,210 epoch 6 - iter 41/416 - loss 0.12471213 - samples/sec: 355.02 - lr: 0.100000
138
+ 2022-08-24 11:02:34,090 epoch 6 - iter 82/416 - loss 0.12531338 - samples/sec: 338.40 - lr: 0.100000
139
+ 2022-08-24 11:02:37,977 epoch 6 - iter 123/416 - loss 0.12700505 - samples/sec: 337.77 - lr: 0.100000
140
+ 2022-08-24 11:02:41,958 epoch 6 - iter 164/416 - loss 0.12652385 - samples/sec: 329.91 - lr: 0.100000
141
+ 2022-08-24 11:02:46,847 epoch 6 - iter 205/416 - loss 0.12700222 - samples/sec: 268.61 - lr: 0.100000
142
+ 2022-08-24 11:02:50,839 epoch 6 - iter 246/416 - loss 0.12583029 - samples/sec: 329.02 - lr: 0.100000
143
+ 2022-08-24 11:02:54,884 epoch 6 - iter 287/416 - loss 0.12590751 - samples/sec: 324.67 - lr: 0.100000
144
+ 2022-08-24 11:02:58,955 epoch 6 - iter 328/416 - loss 0.12594140 - samples/sec: 328.72 - lr: 0.100000
145
+ 2022-08-24 11:03:02,798 epoch 6 - iter 369/416 - loss 0.12619028 - samples/sec: 341.85 - lr: 0.100000
146
+ 2022-08-24 11:03:06,816 epoch 6 - iter 410/416 - loss 0.12617017 - samples/sec: 326.85 - lr: 0.100000
147
+ 2022-08-24 11:03:07,433 ----------------------------------------------------------------------------------------------------
148
+ 2022-08-24 11:03:07,434 EPOCH 6 done: loss 0.1262 - lr 0.100000
149
+ 2022-08-24 11:03:14,657 Evaluating as a multi-label problem: False
150
+ 2022-08-24 11:03:14,881 DEV : loss 0.0752514973282814 - f1-score (micro avg) 0.97
151
+ 2022-08-24 11:03:15,034 BAD EPOCHS (no improvement): 0
152
+ 2022-08-24 11:03:15,036 saving best model
153
+ 2022-08-24 11:03:18,770 ----------------------------------------------------------------------------------------------------
154
+ 2022-08-24 11:03:22,706 epoch 7 - iter 41/416 - loss 0.11739473 - samples/sec: 333.76 - lr: 0.100000
155
+ 2022-08-24 11:03:26,636 epoch 7 - iter 82/416 - loss 0.11762875 - samples/sec: 334.19 - lr: 0.100000
156
+ 2022-08-24 11:03:30,530 epoch 7 - iter 123/416 - loss 0.11795678 - samples/sec: 337.32 - lr: 0.100000
157
+ 2022-08-24 11:03:34,541 epoch 7 - iter 164/416 - loss 0.11731247 - samples/sec: 327.37 - lr: 0.100000
158
+ 2022-08-24 11:03:39,870 epoch 7 - iter 205/416 - loss 0.11730825 - samples/sec: 247.45 - lr: 0.100000
159
+ 2022-08-24 11:03:43,818 epoch 7 - iter 246/416 - loss 0.11685275 - samples/sec: 332.60 - lr: 0.100000
160
+ 2022-08-24 11:03:47,733 epoch 7 - iter 287/416 - loss 0.11654411 - samples/sec: 335.53 - lr: 0.100000
161
+ 2022-08-24 11:03:51,557 epoch 7 - iter 328/416 - loss 0.11687989 - samples/sec: 343.41 - lr: 0.100000
162
+ 2022-08-24 11:03:55,312 epoch 7 - iter 369/416 - loss 0.11710931 - samples/sec: 349.86 - lr: 0.100000
163
+ 2022-08-24 11:03:59,149 epoch 7 - iter 410/416 - loss 0.11723170 - samples/sec: 342.29 - lr: 0.100000
164
+ 2022-08-24 11:03:59,718 ----------------------------------------------------------------------------------------------------
165
+ 2022-08-24 11:03:59,719 EPOCH 7 done: loss 0.1172 - lr 0.100000
166
+ 2022-08-24 11:04:06,739 Evaluating as a multi-label problem: False
167
+ 2022-08-24 11:04:06,960 DEV : loss 0.07426313310861588 - f1-score (micro avg) 0.9696
168
+ 2022-08-24 11:04:07,109 BAD EPOCHS (no improvement): 1
169
+ 2022-08-24 11:04:07,110 ----------------------------------------------------------------------------------------------------
170
+ 2022-08-24 11:04:10,972 epoch 8 - iter 41/416 - loss 0.11078633 - samples/sec: 340.11 - lr: 0.100000
171
+ 2022-08-24 11:04:14,939 epoch 8 - iter 82/416 - loss 0.10893638 - samples/sec: 331.00 - lr: 0.100000
172
+ 2022-08-24 11:04:18,833 epoch 8 - iter 123/416 - loss 0.10944998 - samples/sec: 337.22 - lr: 0.100000
173
+ 2022-08-24 11:04:22,638 epoch 8 - iter 164/416 - loss 0.10903293 - samples/sec: 345.14 - lr: 0.100000
174
+ 2022-08-24 11:04:26,635 epoch 8 - iter 205/416 - loss 0.10899615 - samples/sec: 328.58 - lr: 0.100000
175
+ 2022-08-24 11:04:30,526 epoch 8 - iter 246/416 - loss 0.10934547 - samples/sec: 337.61 - lr: 0.100000
176
+ 2022-08-24 11:04:34,391 epoch 8 - iter 287/416 - loss 0.10995397 - samples/sec: 339.83 - lr: 0.100000
177
+ 2022-08-24 11:04:38,189 epoch 8 - iter 328/416 - loss 0.10996701 - samples/sec: 345.79 - lr: 0.100000
178
+ 2022-08-24 11:04:41,970 epoch 8 - iter 369/416 - loss 0.11080413 - samples/sec: 347.42 - lr: 0.100000
179
+ 2022-08-24 11:04:45,949 epoch 8 - iter 410/416 - loss 0.11086062 - samples/sec: 330.00 - lr: 0.100000
180
+ 2022-08-24 11:04:46,488 ----------------------------------------------------------------------------------------------------
181
+ 2022-08-24 11:04:46,489 EPOCH 8 done: loss 0.1109 - lr 0.100000
182
+ 2022-08-24 11:04:53,723 Evaluating as a multi-label problem: False
183
+ 2022-08-24 11:04:53,952 DEV : loss 0.07048413157463074 - f1-score (micro avg) 0.9713
184
+ 2022-08-24 11:04:54,108 BAD EPOCHS (no improvement): 0
185
+ 2022-08-24 11:04:54,109 saving best model
186
+ 2022-08-24 11:04:57,751 ----------------------------------------------------------------------------------------------------
187
+ 2022-08-24 11:05:01,731 epoch 9 - iter 41/416 - loss 0.10321695 - samples/sec: 330.13 - lr: 0.100000
188
+ 2022-08-24 11:05:05,885 epoch 9 - iter 82/416 - loss 0.10395837 - samples/sec: 316.14 - lr: 0.100000
189
+ 2022-08-24 11:05:09,765 epoch 9 - iter 123/416 - loss 0.10476506 - samples/sec: 339.68 - lr: 0.100000
190
+ 2022-08-24 11:05:13,707 epoch 9 - iter 164/416 - loss 0.10424004 - samples/sec: 340.40 - lr: 0.100000
191
+ 2022-08-24 11:05:17,615 epoch 9 - iter 205/416 - loss 0.10460388 - samples/sec: 336.08 - lr: 0.100000
192
+ 2022-08-24 11:05:21,505 epoch 9 - iter 246/416 - loss 0.10580129 - samples/sec: 337.66 - lr: 0.100000
193
+ 2022-08-24 11:05:25,352 epoch 9 - iter 287/416 - loss 0.10519137 - samples/sec: 341.45 - lr: 0.100000
194
+ 2022-08-24 11:05:29,167 epoch 9 - iter 328/416 - loss 0.10576453 - samples/sec: 344.33 - lr: 0.100000
195
+ 2022-08-24 11:05:33,091 epoch 9 - iter 369/416 - loss 0.10522369 - samples/sec: 334.70 - lr: 0.100000
196
+ 2022-08-24 11:05:36,953 epoch 9 - iter 410/416 - loss 0.10499612 - samples/sec: 339.97 - lr: 0.100000
197
+ 2022-08-24 11:05:37,541 ----------------------------------------------------------------------------------------------------
198
+ 2022-08-24 11:05:37,542 EPOCH 9 done: loss 0.1052 - lr 0.100000
199
+ 2022-08-24 11:05:44,490 Evaluating as a multi-label problem: False
200
+ 2022-08-24 11:05:44,720 DEV : loss 0.07029640674591064 - f1-score (micro avg) 0.9719
201
+ 2022-08-24 11:05:44,884 BAD EPOCHS (no improvement): 0
202
+ 2022-08-24 11:05:44,886 saving best model
203
+ 2022-08-24 11:05:48,465 ----------------------------------------------------------------------------------------------------
204
+ 2022-08-24 11:05:52,355 epoch 10 - iter 41/416 - loss 0.10064182 - samples/sec: 337.69 - lr: 0.100000
205
+ 2022-08-24 11:05:56,216 epoch 10 - iter 82/416 - loss 0.09818265 - samples/sec: 340.21 - lr: 0.100000
206
+ 2022-08-24 11:06:00,337 epoch 10 - iter 123/416 - loss 0.09808745 - samples/sec: 318.69 - lr: 0.100000
207
+ 2022-08-24 11:06:04,274 epoch 10 - iter 164/416 - loss 0.10072979 - samples/sec: 333.58 - lr: 0.100000
208
+ 2022-08-24 11:06:08,146 epoch 10 - iter 205/416 - loss 0.09947450 - samples/sec: 339.18 - lr: 0.100000
209
+ 2022-08-24 11:06:12,068 epoch 10 - iter 246/416 - loss 0.09907584 - samples/sec: 334.96 - lr: 0.100000
210
+ 2022-08-24 11:06:16,056 epoch 10 - iter 287/416 - loss 0.09971183 - samples/sec: 329.34 - lr: 0.100000
211
+ 2022-08-24 11:06:19,927 epoch 10 - iter 328/416 - loss 0.10065394 - samples/sec: 339.34 - lr: 0.100000
212
+ 2022-08-24 11:06:23,749 epoch 10 - iter 369/416 - loss 0.10061685 - samples/sec: 343.62 - lr: 0.100000
213
+ 2022-08-24 11:06:27,632 epoch 10 - iter 410/416 - loss 0.10070568 - samples/sec: 338.22 - lr: 0.100000
214
+ 2022-08-24 11:06:28,178 ----------------------------------------------------------------------------------------------------
215
+ 2022-08-24 11:06:28,182 EPOCH 10 done: loss 0.1005 - lr 0.100000
216
+ 2022-08-24 11:06:35,003 Evaluating as a multi-label problem: False
217
+ 2022-08-24 11:06:35,224 DEV : loss 0.07123567163944244 - f1-score (micro avg) 0.9716
218
+ 2022-08-24 11:06:35,379 BAD EPOCHS (no improvement): 1
219
+ 2022-08-24 11:06:39,298 ----------------------------------------------------------------------------------------------------
220
+ 2022-08-24 11:06:39,299 loading file resources/taggers/best-model.pt
221
+ 2022-08-24 11:06:42,251 SequenceTagger predicts: Dictionary with 34 tags: <unk>, PSP, NN, VM, NNP, SYM, VAUX, JJ, NNPC, PRP, CC, NNC, QC, NST, DEM, RP, QF, NEG, RB, QCC, QO, INTF, JJC, WQ, RDP, UNK, PRPC, NSTC, RBC, QFC, CCC, INJ, <START>, <STOP>
222
+ 2022-08-24 11:06:49,127 Evaluating as a multi-label problem: False
223
+ 2022-08-24 11:06:49,348 0.9709 0.9709 0.9709 0.9709
224
+ 2022-08-24 11:06:49,349
225
+ Results:
226
+ - F-score (micro) 0.9709
227
+ - F-score (macro) 0.8174
228
+ - Accuracy 0.9709
229
+
230
+ By class:
231
+ precision recall f1-score support
232
+
233
+ PSP 0.9971 0.9975 0.9973 7182
234
+ NN 0.9710 0.9685 0.9697 7181
235
+ VM 0.9942 0.9923 0.9933 3643
236
+ NNP 0.9400 0.9076 0.9235 2846
237
+ SYM 1.0000 1.0000 1.0000 2420
238
+ VAUX 0.9928 0.9973 0.9950 2216
239
+ JJ 0.9409 0.9560 0.9484 1933
240
+ NNPC 0.8861 0.8938 0.8899 1592
241
+ PRP 0.9851 0.9829 0.9840 1348
242
+ CC 0.9892 0.9938 0.9915 1289
243
+ NNC 0.7871 0.8778 0.8300 851
244
+ QC 0.9866 0.9933 0.9899 593
245
+ NST 1.0000 0.9960 0.9980 500
246
+ RP 0.9958 0.9754 0.9855 487
247
+ DEM 0.9622 0.9935 0.9776 461
248
+ QF 0.9668 0.9357 0.9510 280
249
+ NEG 1.0000 1.0000 1.0000 190
250
+ RB 0.9677 0.8889 0.9266 135
251
+ QCC 0.9796 0.9697 0.9746 99
252
+ QO 0.9821 0.9649 0.9735 57
253
+ JJC 0.8846 0.4792 0.6216 48
254
+ INTF 0.7576 0.9615 0.8475 26
255
+ WQ 0.9524 0.9524 0.9524 21
256
+ RDP 0.8462 0.6875 0.7586 16
257
+ UNK 0.3333 0.4286 0.3750 7
258
+ PRPC 1.0000 0.5000 0.6667 4
259
+ NSTC 1.0000 1.0000 1.0000 2
260
+ RBC 0.0000 0.0000 0.0000 1
261
+ QFC 0.0000 0.0000 0.0000 1
262
+ CCC 0.0000 0.0000 0.0000 1
263
+
264
+ accuracy 0.9709 35430
265
+ macro avg 0.8366 0.8098 0.8174 35430
266
+ weighted avg 0.9713 0.9709 0.9709 35430
267
+
268
+ 2022-08-24 11:06:49,349 ----------------------------------------------------------------------------------------------------
weights.txt ADDED
File without changes