Hugging Face's logo

HueyNemud
/
icdar23-entrydetector_plaintext_breaks

Token Classification
Transformers
PyTorch
camembert
Generated from Trainer
Inference Endpoints
Model card Files Community
1
HueyNemud commited on
Commit
a76cb17
•
1 Parent(s): fddb0ee

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +32 -20
README.md CHANGED
@@ -13,13 +13,13 @@ should probably proofread and complete it, then remove this comment. -->
13
 
14
  This model is a fine-tuned version of [HueyNemud/das22-10-camembert_pretrained](https://huggingface.co/HueyNemud/das22-10-camembert_pretrained) on the None dataset.
15
  It achieves the following results on the evaluation set:
16
- - Loss: 0.0126
17
- - Ebegin: {'precision': 0.9965437788018433, 'recall': 0.9759308010530274, 'f1': 0.9861295838875166, 'number': 2659}
18
- - Eend: {'precision': 0.992786636294609, 'recall': 0.977204783258595, 'f1': 0.9849340866290018, 'number': 2676}
19
- - Overall Precision: 0.9947
20
- - Overall Recall: 0.9766
21
- - Overall F1: 0.9855
22
- - Overall Accuracy: 0.9974
23
 
24
  ## Model description
25
 
@@ -50,19 +50,31 @@ The following hyperparameters were used during training:
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
52
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
53
- | No log | 0.07 | 300 | 0.0323 | 0.9649 | 0.9892 | 0.9769 | 0.9956 |
54
- | 0.1454 | 0.14 | 600 | 0.0187 | 0.9963 | 0.9693 | 0.9826 | 0.9968 |
55
- | 0.1454 | 0.21 | 900 | 0.0254 | 0.9947 | 0.9649 | 0.9796 | 0.9961 |
56
- | 0.0174 | 0.29 | 1200 | 0.0101 | 0.9916 | 0.9899 | 0.9907 | 0.9982 |
57
- | 0.0113 | 0.36 | 1500 | 0.0082 | 0.9948 | 0.9879 | 0.9913 | 0.9983 |
58
- | 0.0113 | 0.43 | 1800 | 0.0083 | 0.9917 | 0.9903 | 0.9910 | 0.9983 |
59
- | 0.011 | 0.5 | 2100 | 0.0104 | 0.9889 | 0.9912 | 0.9900 | 0.9981 |
60
- | 0.011 | 0.57 | 2400 | 0.0080 | 0.9940 | 0.9906 | 0.9923 | 0.9985 |
61
- | 0.0105 | 0.64 | 2700 | 0.0079 | 0.9896 | 0.9911 | 0.9903 | 0.9981 |
62
- | 0.0073 | 0.72 | 3000 | 0.0075 | 0.9921 | 0.9899 | 0.9910 | 0.9983 |
63
- | 0.0073 | 0.79 | 3300 | 0.0097 | 0.9920 | 0.9902 | 0.9911 | 0.9983 |
64
- | 0.0062 | 0.86 | 3600 | 0.0089 | 0.9955 | 0.9858 | 0.9906 | 0.9982 |
65
- | 0.0062 | 0.93 | 3900 | 0.0069 | 0.9894 | 0.9910 | 0.9902 | 0.9981 |
 
 
 
 
 
 
 
 
 
 
 
 
66
 
67
 
68
  ### Framework versions
 
13
 
14
  This model is a fine-tuned version of [HueyNemud/das22-10-camembert_pretrained](https://huggingface.co/HueyNemud/das22-10-camembert_pretrained) on the None dataset.
15
  It achieves the following results on the evaluation set:
16
+ - Loss: 0.0084
17
+ - Ebegin: {'precision': 0.9988514548238897, 'recall': 0.9811959383226777, 'f1': 0.9899449819768545, 'number': 2659}
18
+ - Eend: {'precision': 0.9984726995036274, 'recall': 0.977204783258595, 'f1': 0.987724268177526, 'number': 2676}
19
+ - Overall Precision: 0.9987
20
+ - Overall Recall: 0.9792
21
+ - Overall F1: 0.9888
22
+ - Overall Accuracy: 0.9980
23
 
24
  ## Model description
25
 
 
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
52
  |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
53
+ | No log | 0.07 | 300 | 0.0367 | 0.9567 | 0.9818 | 0.9691 | 0.9948 |
54
+ | 0.1432 | 0.14 | 600 | 0.0181 | 0.9809 | 0.9811 | 0.9810 | 0.9971 |
55
+ | 0.1432 | 0.21 | 900 | 0.0111 | 0.9877 | 0.9920 | 0.9899 | 0.9981 |
56
+ | 0.0188 | 0.29 | 1200 | 0.0111 | 0.9955 | 0.9869 | 0.9912 | 0.9983 |
57
+ | 0.0121 | 0.36 | 1500 | 0.0094 | 0.9902 | 0.9899 | 0.9901 | 0.9981 |
58
+ | 0.0121 | 0.43 | 1800 | 0.0083 | 0.9914 | 0.9912 | 0.9913 | 0.9983 |
59
+ | 0.0106 | 0.5 | 2100 | 0.0078 | 0.9932 | 0.9902 | 0.9917 | 0.9984 |
60
+ | 0.0106 | 0.57 | 2400 | 0.0083 | 0.9906 | 0.9911 | 0.9909 | 0.9982 |
61
+ | 0.0105 | 0.64 | 2700 | 0.0083 | 0.9871 | 0.9927 | 0.9899 | 0.9981 |
62
+ | 0.0093 | 0.72 | 3000 | 0.0085 | 0.9938 | 0.9851 | 0.9894 | 0.9980 |
63
+ | 0.0093 | 0.79 | 3300 | 0.0075 | 0.9962 | 0.9879 | 0.9920 | 0.9985 |
64
+ | 0.0073 | 0.86 | 3600 | 0.0081 | 0.9927 | 0.9901 | 0.9914 | 0.9984 |
65
+ | 0.0073 | 0.93 | 3900 | 0.0083 | 0.9856 | 0.9923 | 0.9890 | 0.9980 |
66
+ | 0.0073 | 1.0 | 4200 | 0.0063 | 0.9936 | 0.9912 | 0.9924 | 0.9985 |
67
+ | 0.0041 | 1.07 | 4500 | 0.0063 | 0.9959 | 0.9902 | 0.9931 | 0.9987 |
68
+ | 0.0041 | 1.14 | 4800 | 0.0068 | 0.9948 | 0.9907 | 0.9928 | 0.9986 |
69
+ | 0.0048 | 1.22 | 5100 | 0.0074 | 0.9937 | 0.9905 | 0.9921 | 0.9985 |
70
+ | 0.0048 | 1.29 | 5400 | 0.0074 | 0.9912 | 0.9906 | 0.9909 | 0.9982 |
71
+ | 0.0043 | 1.36 | 5700 | 0.0070 | 0.9947 | 0.9907 | 0.9927 | 0.9986 |
72
+ | 0.0046 | 1.43 | 6000 | 0.0072 | 0.9948 | 0.9914 | 0.9931 | 0.9987 |
73
+ | 0.0046 | 1.5 | 6300 | 0.0080 | 0.9939 | 0.9915 | 0.9927 | 0.9986 |
74
+ | 0.0038 | 1.57 | 6600 | 0.0072 | 0.9939 | 0.9921 | 0.9930 | 0.9986 |
75
+ | 0.0038 | 1.65 | 6900 | 0.0061 | 0.9952 | 0.9916 | 0.9934 | 0.9987 |
76
+ | 0.0051 | 1.72 | 7200 | 0.0060 | 0.9959 | 0.9913 | 0.9936 | 0.9988 |
77
+ | 0.005 | 1.79 | 7500 | 0.0060 | 0.9959 | 0.9913 | 0.9936 | 0.9988 |
78
 
79
 
80
  ### Framework versions