HueyNemud commited on
Commit
3baca4e
1 Parent(s): 9b69472

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +84 -0
README.md ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - generated_from_trainer
4
+ model-index:
5
+ - name: icdar23-entrydetector_texttokens_breaks_indents_left_diff_right_ref
6
+ results: []
7
+ ---
8
+
9
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
+ should probably proofread and complete it, then remove this comment. -->
11
+
12
+ # icdar23-entrydetector_texttokens_breaks_indents_left_diff_right_ref
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: 0.0448
17
+ - Ebegin: {'precision': 0.9843225083986562, 'recall': 0.9788418708240535, 'f1': 0.9815745393634842, 'number': 2694}
18
+ - Eend: {'precision': 0.9872036130974784, 'recall': 0.9707623982235382, 'f1': 0.9789139764881508, 'number': 2702}
19
+ - Overall Precision: 0.9858
20
+ - Overall Recall: 0.9748
21
+ - Overall F1: 0.9802
22
+ - Overall Accuracy: 0.9860
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 0.0001
42
+ - train_batch_size: 2
43
+ - eval_batch_size: 2
44
+ - seed: 42
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - training_steps: 7500
48
+
49
+ ### Training results
50
+
51
+ | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
52
+ |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
53
+ | No log | 0.07 | 300 | 0.0868 | 0.9708 | 0.9867 | 0.9787 | 0.9858 |
54
+ | 0.31 | 0.14 | 600 | 0.0805 | 0.9890 | 0.9606 | 0.9746 | 0.9834 |
55
+ | 0.31 | 0.21 | 900 | 0.0758 | 0.9793 | 0.9340 | 0.9561 | 0.9733 |
56
+ | 0.1178 | 0.29 | 1200 | 0.0434 | 0.9845 | 0.9808 | 0.9826 | 0.9885 |
57
+ | 0.1413 | 0.36 | 1500 | 0.0635 | 0.9909 | 0.9687 | 0.9796 | 0.9867 |
58
+ | 0.1413 | 0.43 | 1800 | 0.0355 | 0.9848 | 0.9839 | 0.9844 | 0.9907 |
59
+ | 0.1699 | 0.5 | 2100 | 0.0327 | 0.9914 | 0.9843 | 0.9879 | 0.9920 |
60
+ | 0.1699 | 0.57 | 2400 | 0.0330 | 0.9904 | 0.9832 | 0.9868 | 0.9913 |
61
+ | 0.144 | 0.64 | 2700 | 0.0285 | 0.9840 | 0.9891 | 0.9865 | 0.9911 |
62
+ | 0.0958 | 0.72 | 3000 | 0.0264 | 0.9922 | 0.9836 | 0.9879 | 0.9920 |
63
+ | 0.0958 | 0.79 | 3300 | 0.0312 | 0.9912 | 0.9852 | 0.9882 | 0.9922 |
64
+ | 0.0585 | 0.86 | 3600 | 0.0296 | 0.9893 | 0.9862 | 0.9878 | 0.9919 |
65
+ | 0.0585 | 0.93 | 3900 | 0.0259 | 0.9864 | 0.9899 | 0.9881 | 0.9922 |
66
+ | 0.0478 | 1.0 | 4200 | 0.0314 | 0.9933 | 0.9649 | 0.9789 | 0.9862 |
67
+ | 0.0842 | 1.07 | 4500 | 0.0222 | 0.9887 | 0.9897 | 0.9892 | 0.9928 |
68
+ | 0.0842 | 1.14 | 4800 | 0.0189 | 0.9925 | 0.9883 | 0.9904 | 0.9937 |
69
+ | 0.075 | 1.22 | 5100 | 0.0241 | 0.9890 | 0.9898 | 0.9894 | 0.9930 |
70
+ | 0.075 | 1.29 | 5400 | 0.0242 | 0.9915 | 0.9854 | 0.9884 | 0.9924 |
71
+ | 0.0511 | 1.36 | 5700 | 0.0197 | 0.9929 | 0.9885 | 0.9907 | 0.9939 |
72
+ | 0.042 | 1.43 | 6000 | 0.0223 | 0.9936 | 0.9852 | 0.9894 | 0.9930 |
73
+ | 0.042 | 1.5 | 6300 | 0.0203 | 0.9899 | 0.9905 | 0.9902 | 0.9935 |
74
+ | 0.0596 | 1.57 | 6600 | 0.0215 | 0.9892 | 0.9914 | 0.9903 | 0.9936 |
75
+ | 0.0596 | 1.65 | 6900 | 0.0211 | 0.9922 | 0.9875 | 0.9898 | 0.9933 |
76
+ | 0.0489 | 1.72 | 7200 | 0.0212 | 0.9923 | 0.9869 | 0.9896 | 0.9931 |
77
+
78
+
79
+ ### Framework versions
80
+
81
+ - Transformers 4.26.1
82
+ - Pytorch 1.13.1+cu116
83
+ - Datasets 2.9.0
84
+ - Tokenizers 0.13.2