violence_detection
This repository contains a fine-tuned Vision Transformer (ViT) model for detecting violence in images. The model is trained on the Violence Detection Dataset from Kaggle, which consists of images categorized into violent and non-violent scenes.
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training results
| Training Loss | Epoch | Validation Loss | Accuracy |
|---|---|---|---|
| 0.167900 | 1 | 0.164447 | 0.931034 |
| 0.010100 | 2 | 0.170160 | 0.942529 |
| 0.002200 | 3 | 0.170347 | 0.948276 |
| 0.001400 | 4 | 0.170811 | 0.948276 |
| 0.001100 | 5 | 0.172756 | 0.948276 |
Framework versions
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1