AIbase
EN
Home
Vit Base Patch16 Clip 224.datacompxl
V

Vit Base Patch16 Clip 224.datacompxl

Developed by timm
A vision Transformer model based on the CLIP architecture, specifically designed for image feature extraction, using ViT-B/16 structure and trained on the DataComp XL dataset
Image Classification
Transformers
Open Source License:Apache-2.0#CLIP visual encoding#Zero-shot image classification#Multimodal pre-training
Downloads 36
Release Time : 12/24/2024

Model Overview

This model is the image encoder part of CLIP (Contrastive Language-Image Pretraining), capable of converting input images into meaningful feature representations suitable for various vision tasks.

Model Features

Large-scale pre-training
Trained on the DataComp XL dataset, which contains large-scale image-text pairs
Efficient image encoding
Utilizes ViT architecture, capable of efficiently processing 224x224 resolution input images
Contrastive learning optimization
Trained with CLIP's contrastive learning objective, resulting in features with better generalization capabilities

Model Capabilities

Image feature extraction
Visual representation learning
Cross-modal alignment (aligned with text feature space)

Use Cases

Computer vision
Image retrieval
Using extracted image features for similar image search
Visual classification
Used as a feature extractor for downstream classification tasks
Multimodal applications
Image-text matching
Collaborating with text encoders to achieve image-text matching tasks
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase