MohamedRashad commited on
Commit
a2c7e96
1 Parent(s): 4ea8c70

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +40 -8
README.md CHANGED
@@ -12,7 +12,13 @@ datasets:
12
 
13
  # Arabic Small Nougat
14
 
15
- **Sma**ll, **Simp**le **En**d-**t**o-**En**d **Structur**ed **OC**R **fo**r **Arab**ic **boo**ks.
 
 
 
 
 
 
16
 
17
  ## How to Get Started with the Model
18
 
@@ -54,16 +60,17 @@ print(predict("path/to/page_image.jpg"))
54
 
55
  ## Bias, Risks, and Limitations
56
 
57
- <!-- This section is meant to convey both technical and sociotechnical limitations. -->
58
-
59
- [More Information Needed]
60
 
61
- ### Recommendations
62
 
63
- <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
64
 
65
- Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
66
 
 
67
 
68
  ## Model Details
69
 
@@ -71,4 +78,29 @@ Users (both direct and downstream) should be made aware of the risks, biases and
71
  - **Model type:** VisionEncoderDecoderModel
72
  - **Language(s) (NLP):** Arabic & English
73
  - **License:** GPL 3.0
74
- - **Finetuned from model:** [nougat-small](https://huggingface.co/facebook/nougat-small)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
 
13
  # Arabic Small Nougat
14
 
15
+ **Simp**le **En**d-**t**o-**En**d **Structur**ed **OC**R **fo**r **Arab**ic **boo**ks.
16
+
17
+ ## Description
18
+
19
+ The arabic-small-nougat OCR is an end-to-end structured Optical Character Recognition (OCR) system designed specifically for the Arabic language.
20
+
21
+ The model is based on the [facebook/nougat-small](https://huggingface.co/facebook/nougat-small) architecture and has been fine-tuned using the [Khatt dataset](https://huggingface.co/datasets/Fakhraddin/khatt) along with a custom dataset created for this purpose.
22
 
23
  ## How to Get Started with the Model
24
 
 
60
 
61
  ## Bias, Risks, and Limitations
62
 
63
+ 1. **Text Hallucination:** The model may occasionally generate repeated or incorrect text due to the inherent complexities of OCR tasks.
64
+ 1. **Erroneous Image Paths:** There are instances where the model outputs image paths that are not relevant to the input, indicating occasional confusion.
65
+ 1. **Context Length Constraint:** The model has a maximum context length of 2048 tokens, which may result in incomplete transcriptions for longer book pages.
66
 
67
+ ## Intended Use
68
 
69
+ The arabic-small-nougat OCR is designed for tasks that involve converting images of Arabic book pages into structured text, especially when Markdown format is desired. It is suitable for applications in the field of digitizing Arabic literature and facilitating text extraction from printed materials.
70
 
71
+ ## Ethical Considerations
72
 
73
+ It is crucial to be aware of the model's limitations, particularly in instances where accurate OCR results are critical. Users are advised to verify and review the output, especially in scenarios where precision is paramount.
74
 
75
  ## Model Details
76
 
 
78
  - **Model type:** VisionEncoderDecoderModel
79
  - **Language(s) (NLP):** Arabic & English
80
  - **License:** GPL 3.0
81
+ - **Finetuned from model:** [nougat-small](https://huggingface.co/facebook/nougat-small)
82
+
83
+ ## Acknowledgment
84
+
85
+ If you use or build upon the arabic-small-nougat OCR, please acknowledge the model developer and the open-source community for their contributions. Additionally, be sure to include a copy of the GPL 3.0 license with any redistributed or modified versions of the model.
86
+
87
+ By selecting the GPL 3.0 license, you promote the principles of open source and ensure that the benefits of the model are shared with the broader community.
88
+
89
+ ## Citation
90
+
91
+ If you find this model useful, please consider citing the original facebook/nougat-small model and the datasets used for fine-tuning, including the Khatt dataset and any details regarding the custom dataset.
92
+
93
+ ```bibtex
94
+ @misc{blecher2023nougat,
95
+ title={Nougat: Neural Optical Understanding for Academic Documents},
96
+ author={Lukas Blecher and Guillem Cucurull and Thomas Scialom and Robert Stojnic},
97
+ year={2023},
98
+ eprint={2308.13418},
99
+ archivePrefix={arXiv},
100
+ primaryClass={cs.LG}
101
+ }
102
+ ```
103
+
104
+ ## Disclaimer
105
+
106
+ The arabic-small-nougat OCR is a tool provided "as is," and the developers make no guarantees regarding its suitability for specific tasks. Users are encouraged to thoroughly evaluate the model's output for their particular use cases and requirements.