Donut Base German Receipts

Model Details

• Base model: naver-clova-ix/donut-base
• Dataset:
  • german receipts based on small private dataset(473 images)
• Training:
  • training based on 435 images
  • test validation with 38 images
    • accurcay on supermarked-receipts is well (~70% exact string comparison)
    • not so many unkown stores in dataset, results differ in quality
  • Image size used: 960x1280
  • Epochs: 3
  • LR: 2e-5
  • Batch-Size: 2

Input Example

bon144

Output Schema

{
  "store": "C & A Geretsried",
  "date": "26.07.2025",
  "address": "Karl-Lederer-Platz 15",
  "total": "39,99"
}

Framework versions

• Transformers 4.27.4
• Pytorch 2.7.1
• Datasets 4.0.0
• Tokenizers 0.13.3

--- license: mit language: - de base_model: - naver-clova-ix/donut-base pipeline_tag: image-to-text tags: - Transformers - donut - receipts --- > ⚠️**Warning** > > - Transformers versions above 4.28.XX may break inference result: https://github.com/huggingface/transformers/issues/39473 > - Latest versions seem to work properly (e.g 4.57.6) # Donut German Receipts Fine-tuned donut model for german receipts. ## Model Details - Base model: naver-clova-ix/donut-base - Dataset: - german receipts based on small private dataset(473 images) - Training: - training based on 435 images - test validation with 38 images - accurcay on supermarked-receipts is well (~70% exact string comparison) - not so many unkown stores in dataset, results differ in quality - Image size used: 960x1280 - Epochs: 3 - LR: 2e-5 - Batch-Size: 2 ## Input Example  ## Output Schema ```json { "store": "C & A Geretsried", "date": "26.07.2025", "address": "Karl-Lederer-Platz 15", "total": "39,99" } ``` ## Framework versions - Transformers 4.27.4 - Pytorch 2.7.1 - Datasets 4.0.0 - Tokenizers 0.13.3