D

Deit Tiny Patch16 224

Developed by facebook
DeiT is an efficiently trained vision Transformer model, pretrained and fine-tuned on the ImageNet-1k dataset, suitable for image classification tasks.
Downloads 29.04k
Release Time : 3/2/2022

Model Overview

This model is a more efficiently trained Vision Transformer (ViT), pretrained and fine-tuned on the ImageNet-1k dataset in a supervised manner, and can be used for image classification tasks.

Model Features

Efficient Training
Compared to traditional vision Transformers, DeiT achieves more efficient training through attention mechanisms and distillation techniques.
Lightweight
The tiny model has only 5M parameters, making it suitable for resource-constrained environments.
High Accuracy
Achieves a top-1 accuracy of 72.2% on ImageNet-1k, demonstrating excellent performance.

Model Capabilities

Image Classification
Feature Extraction

Use Cases

Computer Vision
Image Classification
Classify images into one of the 1000 ImageNet categories.
Achieves a top-1 accuracy of 72.2% on ImageNet-1k.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase