Deepfake Image Detector (Version 2)

Model Performance

  • Test Accuracy: 99.83%
  • Best Validation Accuracy: 99.67%
  • Best Epoch: 3
  • Planned Epochs: 5
  • Actual Epoches Trained: 4 (early stopping applied)

Dataset

  • Training: 4,800 images
  • Validation: 600 images
  • Test: 600 images

Training Details

Training was stopped early at epoch 4 due to early stopping criteria being met. The best model was achieved at epoch 3 with validation accuracy of 99.67%.

Usage

from transformers import ViTForImageClassification, ViTFeatureExtractor

model = ViTForImageClassification.from_pretrained('shivani1511/deepfake-image-detector-new-v2')
feature_extractor = ViTFeatureExtractor.from_pretrained('shivani1511/deepfake-image-detector-new-v2')
Downloads last month
-
Safetensors
Model size
85.8M params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support