Hugging Face's logo

fimu-docproc-research
/
standard_0.0.2_EasyOcrEngine

Model card Files Community
xmelus commited on
Commit
dc842b9
·
1 Parent(s): a9f4111

add model

Browse files
Files changed (4) hide show
  1. log_dataset.txt +8 -0
  2. log_train.txt +80 -0
  3. opt.txt +26 -0
  4. weights.pth +3 -0
log_dataset.txt ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ --------------------------------------------------------------------------------
2
+ dataset_name: fimu-docproc-research/born_digital, split: train
3
+ num samples: 424432
4
+ batch_size: 32
5
+ --------------------------------------------------------------------------------
6
+ dataset_name: fimu-docproc-research/born_digital, split: valid
7
+ num samples: 142881
8
+ --------------------------------------------------------------------------------
log_train.txt ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [10000/100000] Train loss: 0.00295, Valid loss: 0.03153, Elapsed_time: 951.36270, Infer_time: 36.37296
2
+ Current_accuracy : 98.880, Current_norm_ED : 0.9964
3
+ Best_accuracy : 98.880
4
+ --------------------------------------------------------------------------------
5
+ Ground Truth | Prediction | Confidence Score & T/F
6
+ --------------------------------------------------------------------------------
7
+ - | - | 0.9984 True
8
+ --------------------------------------------------------------------------------
9
+ [20000/100000] Train loss: 0.00134, Valid loss: 0.02905, Elapsed_time: 2604.50049, Infer_time: 36.42053
10
+ Current_accuracy : 98.980, Current_norm_ED : 0.9967
11
+ Best_accuracy : 98.980
12
+ --------------------------------------------------------------------------------
13
+ Ground Truth | Prediction | Confidence Score & T/F
14
+ --------------------------------------------------------------------------------
15
+ 14.08.2014 | 14.08.2014 | 0.7549 True
16
+ --------------------------------------------------------------------------------
17
+ [30000/100000] Train loss: 0.00085, Valid loss: 0.03176, Elapsed_time: 4256.45649, Infer_time: 36.42712
18
+ Current_accuracy : 98.945, Current_norm_ED : 0.9963
19
+ Best_accuracy : 98.980
20
+ --------------------------------------------------------------------------------
21
+ Ground Truth | Prediction | Confidence Score & T/F
22
+ --------------------------------------------------------------------------------
23
+ DPH | DPH | 0.5268 True
24
+ --------------------------------------------------------------------------------
25
+ [40000/100000] Train loss: 0.00088, Valid loss: 0.02621, Elapsed_time: 5908.42990, Infer_time: 36.18533
26
+ Current_accuracy : 98.888, Current_norm_ED : 0.9961
27
+ Best_accuracy : 98.980
28
+ --------------------------------------------------------------------------------
29
+ Ground Truth | Prediction | Confidence Score & T/F
30
+ --------------------------------------------------------------------------------
31
+ Vigantice, | Vigantice, | 0.9766 True
32
+ --------------------------------------------------------------------------------
33
+ [50000/100000] Train loss: 0.00037, Valid loss: 0.02450, Elapsed_time: 7560.27955, Infer_time: 36.28681
34
+ Current_accuracy : 98.986, Current_norm_ED : 0.9967
35
+ Best_accuracy : 98.986
36
+ --------------------------------------------------------------------------------
37
+ Ground Truth | Prediction | Confidence Score & T/F
38
+ --------------------------------------------------------------------------------
39
+ slúži | slúžm | 0.5081 False
40
+ --------------------------------------------------------------------------------
41
+ [60000/100000] Train loss: 0.00027, Valid loss: 0.02777, Elapsed_time: 9215.24257, Infer_time: 36.79374
42
+ Current_accuracy : 98.926, Current_norm_ED : 0.9966
43
+ Best_accuracy : 98.986
44
+ --------------------------------------------------------------------------------
45
+ Ground Truth | Prediction | Confidence Score & T/F
46
+ --------------------------------------------------------------------------------
47
+ s.r.o. | s.r.o. | 0.9805 True
48
+ --------------------------------------------------------------------------------
49
+ [70000/100000] Train loss: 0.00021, Valid loss: 0.02271, Elapsed_time: 10873.43919, Infer_time: 36.44267
50
+ Current_accuracy : 98.933, Current_norm_ED : 0.9963
51
+ Best_accuracy : 98.986
52
+ --------------------------------------------------------------------------------
53
+ Ground Truth | Prediction | Confidence Score & T/F
54
+ --------------------------------------------------------------------------------
55
+ Hypoalergenní | Hypoalergenní | 0.4633 True
56
+ --------------------------------------------------------------------------------
57
+ [80000/100000] Train loss: 0.00022, Valid loss: 0.02620, Elapsed_time: 12526.53949, Infer_time: 36.43973
58
+ Current_accuracy : 98.949, Current_norm_ED : 0.9964
59
+ Best_accuracy : 98.986
60
+ --------------------------------------------------------------------------------
61
+ Ground Truth | Prediction | Confidence Score & T/F
62
+ --------------------------------------------------------------------------------
63
+ Firma | Firma | 0.9593 True
64
+ --------------------------------------------------------------------------------
65
+ [90000/100000] Train loss: 0.00010, Valid loss: 0.02698, Elapsed_time: 14175.97223, Infer_time: 36.30002
66
+ Current_accuracy : 98.926, Current_norm_ED : 0.9963
67
+ Best_accuracy : 98.986
68
+ --------------------------------------------------------------------------------
69
+ Ground Truth | Prediction | Confidence Score & T/F
70
+ --------------------------------------------------------------------------------
71
+ 2014 | 2014 | 0.8803 True
72
+ --------------------------------------------------------------------------------
73
+ [100000/100000] Train loss: 0.00022, Valid loss: 0.03040, Elapsed_time: 15823.38003, Infer_time: 36.30645
74
+ Current_accuracy : 98.975, Current_norm_ED : 0.9966
75
+ Best_accuracy : 98.986
76
+ --------------------------------------------------------------------------------
77
+ Ground Truth | Prediction | Confidence Score & T/F
78
+ --------------------------------------------------------------------------------
79
+ 263,00 | 263,00 | 0.9889 True
80
+ --------------------------------------------------------------------------------
opt.txt ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ------------ Options -------------
2
+ character: !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~ªÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿĀāĂ㥹ĆćČčĎďĐđĒēĖėĘęĚěĞğĨĩĪīĮįİıĶķĹĺĻļĽľŁłŃńŅņŇňŒœŔŕŘřŚśŞşŠšŤťŨũŪūŮůŲųŸŹźŻżŽžƏƠơƯưȘșȚțə̇ḌḍḶḷṀṁṂṃṄṅṆṇṬṭẠạẢảẤấẦầẨẩẪẫẬậẮắẰằẲẳẴẵẶặẸẹẺẻẼẽẾếỀềỂểỄễỆệỈỉỊịỌọỎỏỐốỒồỔổỖỗỘộỚớỜờỞởỠỡỢợỤụỦủỨứỪừỬửỮữỰựỲỳỴỵỶỷỸỹ€
3
+ experiment_name: invoices_standard_0.0.2
4
+ hf_dataset_name: fimu-docproc-research/born_digital
5
+ workers: 6
6
+ batch_size: 32
7
+ num_iter: 100000
8
+ valInterval: 10000
9
+ optim: Adam
10
+ lr: 1.0
11
+ beta1: 0.9
12
+ rho: 0.95
13
+ eps: 1e-08
14
+ grad_clip: 5
15
+ batch_max_length: 34
16
+ imgH: 64
17
+ imgW: 512
18
+ contrast_adjust: 0.0
19
+ sensitive: True
20
+ PAD: True
21
+ data_filtering_off: False
22
+ Prediction: CTC
23
+ freeze_FeatureFxtraction: False
24
+ freeze_SequenceModeling: False
25
+ num_class: 352
26
+ ---------------------------------------
weights.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1aab42629202cb40ae3fc2099116ea34ea39c1519f128efce6d44c71e1f987cb
3
+ size 15404555