README.md

metadata

license: apache-2.0
base_model: PekingU/rtdetr_v2_r101vd
tags:
  - object-detection
  - computer-vision
  - voucher-classification
  - rt-detr
  - rtdetrv2
datasets:
  - custom-voucher-dataset
metrics:
  - map
  - map_50
  - map_75
widget:
  - src: >-
      https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg
    example_title: Example Image

RT-DETRv2 Fine-tuned for Voucher Classification

This model is a fine-tuned version of PekingU/rtdetr_v2_r101vd for voucher classification and object detection.

Model Details

Model Description

• Model Type: Object Detection (RT-DETRv2)
• Base Model: PekingU/rtdetr_v2_r101vd
• Task: Multi-class voucher classification and detection
• Classes: 3 classes
  • 0: digital (digital invoices)
  • 1: fisico (physical receipts on blank pages)
  • 2: tesoreria (small on-site payment receipts)

Training Details

Training Dataset:

• Total Samples: 507
• Class Distribution:
• fisico (id: 1): 241 samples (47.5%)
• digital (id: 0): 147 samples (29.0%)
• tesoreria (id: 2): 119 samples (23.5%)

Training Configuration:

• Image Size: 800x800
• Batch Size: 24
• Learning Rate: 1.5e-05
• Weight Decay: 0.0001
• Epochs: 2
• Validation Split: 0.0

Data Processing:

• Pre-augmented dataset used (no runtime augmentation)
• External train/validation split (use create_train_val_split.py)
• Preprocessing: Resize + Normalization only

Performance Metrics

Final Evaluation Results: Dataset Information: Training Dataset:

• Digital invoices: 147 samples (29.0%)
• Fisico receipts: 241 samples (47.5%)
• Tesoreria receipts: 119 samples (23.5%)
• Total training samples: 507

Model Configuration:

• Base model: PekingU/rtdetr_v2_r101vd
• Architecture: rtdetr_v2_r101vd
• Input resolution: 800×800 pixels
• Training epochs: 2
• Batch size: 24

Training Hardware:

• GPU: NVIDIA A100-SXM4-40GB
• VRAM: 39.6 GB
• RAM: 83.5 GB
• GPU configuration: A100 optimized

Training Time: 0.0 minutes

Training Summary:

• Final training loss: 0.0000

MLflow Tracking

• MLflow Run ID: c348e8235f8c40138c05c051fc207bb6
• MLflow Experiment: RT-DETRv2_Voucher_Classification

Usage

from transformers import AutoModelForObjectDetection, AutoImageProcessor
import torch
from PIL import Image
import numpy as np

# Load model and processor
model = AutoModelForObjectDetection.from_pretrained("jnmrr/rtdetr-v2-voucher-classifier")
image_processor = AutoImageProcessor.from_pretrained("jnmrr/rtdetr-v2-voucher-classifier")

# Load and preprocess image
image = Image.open("path/to/your/voucher.jpg").convert("RGB")
inputs = image_processor(images=image, return_tensors="pt")

# Run inference
with torch.no_grad():
    outputs = model(**inputs)

# Post-process results
target_sizes = torch.tensor([image.size[::-1]])  # (height, width)
results = image_processor.post_process_object_detection(
    outputs, 
    target_sizes=target_sizes, 
    threshold=0.5
)[0]

# Print predictions
class_names = ["digital", "fisico", "tesoreria"]
for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
    print(f"Class: {class_names[label.item()]}")
    print(f"Confidence: {score.item():.3f}")
    print(f"BBox: {box.tolist()}")

Training Procedure

The model was fine-tuned using the Hugging Face Transformers library with:

• Pre-augmented dataset focusing on challenging cases
• Format-specific augmentation strategies applied during data preparation
• MLflow experiment tracking for reproducibility
• External train/validation split for unbiased evaluation

Limitations and Bias

• Trained specifically on voucher/receipt images
• Performance may vary on images significantly different from training distribution
• Model optimized for 3-class voucher classification task

Citation

If you use this model, please cite:

@misc{rtdetr-v2-voucher-classifier,
  title={RT-DETRv2 Fine-tuned for Voucher Classification},
  author={Your Name},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/jnmrr/rtdetr-v2-voucher-classifier}
}