Fine Tuned ViT on Food101 🌮🍣🍕🍣🍝

ViT feature extractor computer vision model to classify images of classes Food101 dataset.

Training Details

This model was fine-tuned on the Food-101 dataset using a pretrained Vision Transformer (ViT) in PyTorch with.

Final Result

• Top-1 Accuracy: 89%
• Total Training Time: 3:26:16
• Test loss: test_loss=1.16616
• Train loss: test_loss=1.83015
• Batch size: 128
• Num epochs: 40
• Hardware: NVIDIA DGX Spark