File size: 11,871 Bytes
9e9d27b
339b1e7
 
 
 
 
 
 
9e9d27b
339b1e7
 
9e9d27b
339b1e7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e9d27b
 
 
 
 
 
339b1e7
 
 
 
 
 
 
 
 
6dd6afe
 
 
9e9d27b
 
 
 
 
5a532e4
9e9d27b
339b1e7
9e9d27b
339b1e7
9e9d27b
 
 
 
 
 
339b1e7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e9d27b
 
339b1e7
 
 
9e9d27b
339b1e7
9e9d27b
339b1e7
9e9d27b
 
 
339b1e7
9e9d27b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
aa550df
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
---
language:
- multilingual
- en
- de
- fr
- ja
license: mit
tags:
- object-detection
- vision
- generated_from_trainer
- DocLayNet
- COCO
- PDF
- IBM
- Financial-Reports
- Finance
- Manuals
- Scientific-Articles
- Science
- Laws
- Law
- Regulations
- Patents
- Government-Tenders
- object-detection
- image-segmentation
- token-classification
inference: false
datasets:
- pierreguillou/DocLayNet-base
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: layout-xlm-base-finetuned-with-DocLayNet-base-at-linelevel-ml384
  results:
  - task:
      name: Token Classification
      type: token-classification
    metrics:
    - name: f1
      type: f1
      value: 0.7336
    - name: accuracy
      type: accuracy
      value: 0.9373
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# Document Understanding model (finetuned LayoutXLM base at line level on DocLayNet base)

This model is a fine-tuned version of [microsoft/layoutxlm-base](https://huggingface.co/microsoft/layoutxlm-base) with the [DocLayNet base](https://huggingface.co/datasets/pierreguillou/DocLayNet-base) dataset.
It achieves the following results on the evaluation set:

- Loss: 0.2364
- Precision: 0.7260
- Recall: 0.7415
- F1: 0.7336
- Accuracy: 0.9373

## References 

### Blog posts

  - Layout XLM base
    - (03/05/2023) [Document AI | Inference APP and fine-tuning notebook for Document Understanding at line level with LayoutXLM base]()
  - LiLT base
      - (02/16/2023) [Document AI | Inference APP and fine-tuning notebook for Document Understanding at paragraph level](https://medium.com/@pierre_guillou/document-ai-inference-app-and-fine-tuning-notebook-for-document-understanding-at-paragraph-level-c18d16e53cf8)
      - (02/14/2023) [Document AI | Inference APP for Document Understanding at line level](https://medium.com/@pierre_guillou/document-ai-inference-app-for-document-understanding-at-line-level-a35bbfa98893)
      - (02/10/2023) [Document AI | Document Understanding model at line level with LiLT, Tesseract and DocLayNet dataset](https://medium.com/@pierre_guillou/document-ai-document-understanding-model-at-line-level-with-lilt-tesseract-and-doclaynet-dataset-347107a643b8)
      - (01/31/2023) [Document AI | DocLayNet image viewer APP](https://medium.com/@pierre_guillou/document-ai-doclaynet-image-viewer-app-3ac54c19956)
      - (01/27/2023) [Document AI | Processing of DocLayNet dataset to be used by layout models of the Hugging Face hub (finetuning, inference)](https://medium.com/@pierre_guillou/document-ai-processing-of-doclaynet-dataset-to-be-used-by-layout-models-of-the-hugging-face-hub-308d8bd81cdb)

### Notebooks (paragraph level)
- LiLT base
    - [Document AI | Inference APP at paragraph level with a Document Understanding model (LiLT fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/Gradio_inference_on_LiLT_model_finetuned_on_DocLayNet_base_in_any_language_at_levelparagraphs_ml512.ipynb)
    - [Document AI | Inference at paragraph level with a Document Understanding model (LiLT fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/inference_on_LiLT_model_finetuned_on_DocLayNet_base_in_any_language_at_levelparagraphs_ml512.ipynb)
    - [Document AI | Fine-tune LiLT on DocLayNet base in any language at paragraph level (chunk of 512 tokens with overlap)](https://github.com/piegu/language-models/blob/master/Fine_tune_LiLT_on_DocLayNet_base_in_any_language_at_paragraphlevel_ml_512.ipynb)

### Notebooks (line level)
- Layout XLM base
    - [Document AI | Inference at line level with a Document Understanding model (LayoutXLM base fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/inference_on_LayoutXLM_base_model_finetuned_on_DocLayNet_base_in_any_language_at_levellines_ml384.ipynb)
    - [Document AI | Inference APP at line level with a Document Understanding model (LayoutXLM base fine-tuned on DocLayNet base dataset)](https://github.com/piegu/language-models/blob/master/Gradio_inference_on_LayoutXLM_base_model_finetuned_on_DocLayNet_base_in_any_language_at_levellines_ml384.ipynb)
    - [Document AI | Fine-tune LayoutXLM base on DocLayNet base in any language at line level (chunk of 384 tokens with overlap)](https://github.com/piegu/language-models/blob/master/Fine_tune_LayoutXLM_base_on_DocLayNet_base_in_any_language_at_linelevel_ml_384.ipynb)
- LiLT base
    - [Document AI | Inference at line level with a Document Understanding model (LiLT fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/inference_on_LiLT_model_finetuned_on_DocLayNet_base_in_any_language_at_levellines_ml384.ipynb)
    - [Document AI | Inference APP at line level with a Document Understanding model (LiLT fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/Gradio_inference_on_LiLT_model_finetuned_on_DocLayNet_base_in_any_language_at_levellines_ml384.ipynb)
    - [Document AI | Fine-tune LiLT on DocLayNet base in any language at line level (chunk of 384 tokens with overlap)](https://github.com/piegu/language-models/blob/master/Fine_tune_LiLT_on_DocLayNet_base_in_any_language_at_linelevel_ml_384.ipynb)
    - [DocLayNet image viewer APP](https://github.com/piegu/language-models/blob/master/DocLayNet_image_viewer_APP.ipynb)
    - [Processing of DocLayNet dataset to be used by layout models of the Hugging Face hub (finetuning, inference)](processing_DocLayNet_dataset_to_be_used_by_layout_models_of_HF_hub.ipynb)

### APP

You can test this model with this APP in Hugging Face Spaces: [Inference APP for Document Understanding at line level (v2)](https://huggingface.co/spaces/pierreguillou/Inference-APP-Document-Understanding-at-linelevel-v2).

![Inference APP for Document Understanding at line level (v2)](https://huggingface.co/pierreguillou/layout-xlm-base-finetuned-with-DocLayNet-base-at-linelevel-ml384/resolve/main/app_layoutXLM_base_document_understanding_AI.png)

### DocLayNet dataset

[DocLayNet dataset](https://github.com/DS4SD/DocLayNet) (IBM) provides page-by-page layout segmentation ground-truth using bounding-boxes for 11 distinct class labels on 80863 unique pages from 6 document categories. 

Until today, the dataset can be downloaded through direct links or as a dataset from Hugging Face datasets:
- direct links: [doclaynet_core.zip](https://codait-cos-dax.s3.us.cloud-object-storage.appdomain.cloud/dax-doclaynet/1.0.0/DocLayNet_core.zip) (28 GiB), [doclaynet_extra.zip](https://codait-cos-dax.s3.us.cloud-object-storage.appdomain.cloud/dax-doclaynet/1.0.0/DocLayNet_extra.zip) (7.5 GiB)
- Hugging Face dataset library: [dataset DocLayNet](https://huggingface.co/datasets/ds4sd/DocLayNet)

Paper: [DocLayNet: A Large Human-Annotated Dataset for Document-Layout Analysis](https://arxiv.org/abs/2206.01062) (06/02/2022)

## Model description

The model was finetuned at **line level on chunk of 384 tokens with overlap of 128 tokens**. Thus, the model was trained with all layout and text data of all pages of the dataset.

At inference time, a calculation of best probabilities give the label to each line bounding boxes.

## Inference

See notebook: [Document AI | Inference at line level with a Document Understanding model (LayoutXLM base fine-tuned on DocLayNet dataset)](https://github.com/piegu/language-models/blob/master/inference_on_LayoutXLM_base_model_finetuned_on_DocLayNet_base_in_any_language_at_levellines_ml384.ipynb)

## Training and evaluation data

See notebook: [Document AI | Fine-tune LayoutXLM base on DocLayNet base in any language at line level (chunk of 384 tokens with overlap)](https://github.com/piegu/language-models/blob/master/Fine_tune_LayoutXLM_base_on_DocLayNet_base_in_any_language_at_linelevel_ml_384.ipynb)

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Accuracy | F1     | Validation Loss | Precision | Recall |
|:-------------:|:-----:|:----:|:--------:|:------:|:---------------:|:---------:|:------:|
| No log        | 0.12  | 300  | 0.8413   | 0.1311 | 0.5185          | 0.1437    | 0.1205 |
| 0.9231        | 0.25  | 600  | 0.8751   | 0.5031 | 0.4108          | 0.4637    | 0.5498 |
| 0.9231        | 0.37  | 900  | 0.8887   | 0.5206 | 0.3911          | 0.5076    | 0.5343 |
| 0.369         | 0.5   | 1200 | 0.8724   | 0.5365 | 0.4118          | 0.5094    | 0.5667 |
| 0.2737        | 0.62  | 1500 | 0.8960   | 0.6033 | 0.3328          | 0.6046    | 0.6020 |
| 0.2737        | 0.75  | 1800 | 0.9186   | 0.6404 | 0.2984          | 0.6062    | 0.6787 |
| 0.2542        | 0.87  | 2100 | 0.9163   | 0.6593 | 0.3115          | 0.6324    | 0.6887 |
| 0.2542        | 1.0   | 2400 | 0.9198   | 0.6537 | 0.2878          | 0.6160    | 0.6962 |
| 0.1938        | 1.12  | 2700 | 0.9165   | 0.6752 | 0.3414          | 0.6673    | 0.6833 |
| 0.1581        | 1.25  | 3000 | 0.9193   | 0.6871 | 0.3611          | 0.6868    | 0.6875 |
| 0.1581        | 1.37  | 3300 | 0.9256   | 0.6822 | 0.2763          | 0.6988    | 0.6663 |
| 0.1428        | 1.5   | 3600 | 0.9287   | 0.7084 | 0.3065          | 0.7246    | 0.6929 |
| 0.1428        | 1.62  | 3900 | 0.9194   | 0.6812 | 0.2942          | 0.6866    | 0.6760 |
| 0.1025        | 1.74  | 4200 | 0.9347   | 0.7223 | 0.2990          | 0.7315    | 0.7133 |
| 0.1225        | 1.87  | 4500 | 0.9360   | 0.7048 | 0.2729          | 0.7249    | 0.6858 |
| 0.1225        | 1.99  | 4800 | 0.9396   | 0.7222 | 0.2826          | 0.7497    | 0.6966 |
| 0.108         | 2.12  | 5100 | 0.9301   | 0.7193 | 0.3071          | 0.7022    | 0.7372 |
| 0.108         | 2.24  | 5400 | 0.9334   | 0.7243 | 0.2999          | 0.7250    | 0.7237 |
| 0.0799        | 2.37  | 5700 | 0.9382   | 0.7254 | 0.2710          | 0.7310    | 0.7198 |
| 0.0793        | 2.49  | 6000 | 0.9329   | 0.7228 | 0.3201          | 0.7352    | 0.7108 |
| 0.0793        | 2.62  | 6300 | 0.9373   | 0.7336 | 0.3035          | 0.7260    | 0.7415 |
| 0.0696        | 2.74  | 6600 | 0.9374   | 0.7275 | 0.3137          | 0.7313    | 0.7237 |
| 0.0696        | 2.87  | 6900 | 0.9381   | 0.7253 | 0.3242          | 0.7369    | 0.7142 |
| 0.0866        | 2.99  | 7200 | 0.2473   | 0.7439 | 0.7207          | 0.7321    | 0.9407 |


### Framework versions

- Transformers 4.26.1
- Pytorch 1.10.0+cu111
- Datasets 2.10.1
- Tokenizers 0.13.2

## Other models
- Line level
  - [Document Understanding model (finetuned LiLT base at line level on DocLayNet base)](https://huggingface.co/pierreguillou/lilt-xlm-roberta-base-finetuned-with-DocLayNet-base-at-linelevel-ml384) (accuracy | tokens:  85.84% - lines: 91.97%)
  - [Document Understanding model (finetuned LayoutXLM base at line level on DocLayNet base)](https://huggingface.co/pierreguillou/layout-xlm-base-finetuned-with-DocLayNet-base-at-linelevel-ml384) (accuracy | tokens:  93.73% - lines: ...)
- Paragraph level
  - [Document Understanding model (finetuned LiLT base at paragraph level on DocLayNet base)](https://huggingface.co/pierreguillou/lilt-xlm-roberta-base-finetuned-with-DocLayNet-base-at-paragraphlevel-ml512) (accuracy | tokens: 86.34% - paragraphs: 68.15%)
  - [Document Understanding model (finetuned LayoutXLM base at paragraph level on DocLayNet base)](https://huggingface.co/pierreguillou/layout-xlm-base-finetuned-with-DocLayNet-base-at-paragraphlevel-ml512) (accuracy | tokens:  96.93% - paragraphs: 86.55%)