Update task category to `image-text-to-text`, add link to paper and Github repo and ocr tag

This PR updates the task category to `image-text-to-text` to reflect the multimodal nature of the dataset. It also adds a link to the paper on the Hugging Face Hub, a link to the Github repository, and the "ocr" tag.

README.md

@@ -1,43 +1,44 @@
 ---
+language:
+- es
+license: apache-2.0
+size_categories:
+- 10K<n<100K
+task_categories:
+- image-text-to-text
+pretty_name: SpanishOCR
 dataset_info:
   features:
-    - name: image
-      dtype: string
-    - name: text
-      dtype: string
+  - name: image
+    dtype: string
+  - name: text
+    dtype: string
   splits:
-    - name: train
-      num_bytes: 11860671915
-      num_examples: 13320
+  - name: train
+    num_bytes: 11860671915
+    num_examples: 13320
   download_size: 11661151472
   dataset_size: 11860671915
 configs:
-  - config_name: default
-    data_files:
-      - split: train
-        path: data/train-*
-license: apache-2.0
-language:
-- es
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
 tags:
 - finance
-pretty_name: SpanishOCR
-size_categories:
-- 10K<n<100K
-task_categories:
-- image-to-text
+- ocr
 ---
 
 ----------------------------------------------------------------
 # Dataset Card for SpanishOCR Dataset
 
+[Paper](https://huggingface.co/papers/2506.14028)
 
-### Dataset Summary
-The SpanishOCR dataset contains images derived from regulatory documents from Peru government in pdf format. This dataset is used for benchmarkingg and evaluating Large Language Models ability on converting unstructured dcuments, such as pdfs and images, into machine readable format, particularly in finance domain, where the conversion task is more complex and valuable.
+The SpanishOCR dataset contains images derived from regulatory documents from Peru government in pdf format. This dataset is used for benchmarking and evaluating Large Language Models ability on converting unstructured documents, such as pdfs and images, into machine readable format, particularly in the finance domain, where the conversion task is more complex and valuable.
 
 ### Supported Tasks
 
-- **Task:** Image-to-Text
+- **Task:** Image-Text-to-Text (OCR, Financial QA)
 - **Evaluation Metrics:** ROUGE-1
 
 ### Languages
@@ -58,7 +59,7 @@ Each instance in the SpanishOCR dataset comprises 2 fields:
 ## Dataset Creation
 
 ### Curation Rationale
-The SpanishOCR dataset was curated to support research and development on information extraction techniques and layout retain ability for unstructured documents in Spanish. By providing real-world regulatory documents in unstructured format with ground truth, the dataset seeks to address challenges in extracting informat as well as layouts and convert into machine-readable format.
+The SpanishOCR dataset was curated to support research and development on information extraction techniques and layout retain ability for unstructured documents in Spanish. By providing real-world regulatory documents in unstructured format with ground truth, the dataset seeks to address challenges in extracting information as well as layouts and convert into machine-readable format.
 
 ### Source Data
 #### Initial Data Collection and Normalization
@@ -96,6 +97,7 @@ This dataset enables AI models to extract structured information from scanned fi
 - While the dataset covers regulatory documents, it may lack sufficient variety in layout styles (e.g., handwritten notes, non-standard financial forms, embedded charts), which could limit a model’s ability to generalize to less structured or unconventional financial documents.
 
 ## Additional Information
+[Code](https://github.com/sunnymouse2022/FinBen)
 
 ### Dataset Curators
 - Yueru He
@@ -117,7 +119,6 @@ If you use this dataset, please cite:
       year={2025},
       eprint={2506.14028},
       archivePrefix={arXiv},
-      primaryClass={cs.CL},
+      primaryClass={cs.CL},\
       url={https://arxiv.org/abs/2506.14028}, 
-}
-```
+}