Dataset Viewer
Auto-converted to Parquet Duplicate
samples_evaluated
int64
per_field_accuracy
dict
macro_accuracy
float64
details
list
3
{ "vendor_name": 0.333, "vendor_tax_id": 1, "invoice_number": 0.667, "invoice_date_jalali": 1, "total_amount": 0.333 }
0.667
[ { "id": "inv_001", "fields": { "vendor_name": { "expected": "شرکت فولاد آریا صنعت", "actual": "تعنص اىرآ دالوف تكرش", "match": false }, "vendor_tax_id": { "expected": "1234567890", "actual": "1234567890", "match": true }, "invoice...

DocFlow Invoice Samples — Persian & Bilingual

Synthetic invoice dataset for evaluating DocFlow AI field extraction pipelines. Published by Aria AI Engineering Team.

Dataset Summary

Property Value
Samples 50 (synthetic, OCR-friendly)
Languages Persian (FA), English (EN)
Formats PNG images + JSON annotations
Use case Invoice OCR benchmarking, AP automation R&D
Synthetic Yes — no real PII

Fields Annotated

  • vendor_name, vendor_tax_id, invoice_number
  • invoice_date_jalali, buyer_name
  • subtotal, tax_amount, total_amount, currency
  • line_items[] with description, quantity, unit_price, total

Intended Use

  • Benchmarking DocFlow extraction accuracy
  • Fine-tuning LayoutLMv3 / Donut models (future release)
  • Integration testing for ERP export pipelines

Limitations

  • Small demo set; not representative of all Iranian invoice layouts
  • Synthetic data only — do not use for compliance certification

Citation

@dataset{docflow_invoice_samples_fa_2025,
  title={DocFlow Invoice Samples FA},
  author={Aria AI Engineering Team},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/alirezaaminzadeh/docflow-invoice-samples-fa}
}

License

Apache 2.0

Downloads last month
-

Space using alirezaaminzadeh/docflow-invoice-samples-fa 1