update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- generated_from_trainer
|
4 |
+
model-index:
|
5 |
+
- name: icdar23-entrydetector_plaintext_breaks
|
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_plaintext_breaks
|
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: 0.0119
|
17 |
+
- Ebegin: {'precision': 0.9996581196581197, 'recall': 0.9740173217854764, 'f1': 0.9866711658511895, 'number': 3002}
|
18 |
+
- Eend: {'precision': 0.9866799866799867, 'recall': 0.9876666666666667, 'f1': 0.9871730801266034, 'number': 3000}
|
19 |
+
- Overall Precision: 0.9931
|
20 |
+
- Overall Recall: 0.9808
|
21 |
+
- Overall F1: 0.9869
|
22 |
+
- Overall Accuracy: 0.9976
|
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: 6000
|
48 |
+
|
49 |
+
### Training results
|
50 |
+
|
51 |
+
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
52 |
+
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
53 |
+
| No log | 0.07 | 300 | 0.0362 | 0.9796 | 0.9789 | 0.9793 | 0.9961 |
|
54 |
+
| 0.1771 | 0.14 | 600 | 0.0178 | 0.9918 | 0.9784 | 0.9851 | 0.9971 |
|
55 |
+
| 0.1771 | 0.21 | 900 | 0.0181 | 0.9915 | 0.9760 | 0.9837 | 0.9969 |
|
56 |
+
| 0.0232 | 0.29 | 1200 | 0.0146 | 0.9878 | 0.9842 | 0.9860 | 0.9973 |
|
57 |
+
| 0.016 | 0.36 | 1500 | 0.0122 | 0.9797 | 0.9872 | 0.9835 | 0.9968 |
|
58 |
+
| 0.016 | 0.43 | 1800 | 0.0119 | 0.9927 | 0.9807 | 0.9866 | 0.9974 |
|
59 |
+
| 0.0117 | 0.5 | 2100 | 0.0114 | 0.9897 | 0.9824 | 0.9860 | 0.9973 |
|
60 |
+
| 0.0117 | 0.57 | 2400 | 0.0109 | 0.9913 | 0.9849 | 0.9881 | 0.9977 |
|
61 |
+
| 0.0105 | 0.64 | 2700 | 0.0111 | 0.9971 | 0.9746 | 0.9857 | 0.9973 |
|
62 |
+
| 0.0109 | 0.72 | 3000 | 0.0101 | 0.9887 | 0.9859 | 0.9873 | 0.9976 |
|
63 |
+
| 0.0109 | 0.79 | 3300 | 0.0115 | 0.9857 | 0.9865 | 0.9861 | 0.9973 |
|
64 |
+
|
65 |
+
|
66 |
+
### Framework versions
|
67 |
+
|
68 |
+
- Transformers 4.26.0
|
69 |
+
- Pytorch 1.13.1+cu116
|
70 |
+
- Datasets 2.9.0
|
71 |
+
- Tokenizers 0.13.2
|