AIbase
EN
Home
Eva02 Base Patch16 Clip 224.merged2b S8b B131k
E

Eva02 Base Patch16 Clip 224.merged2b S8b B131k

Developed by timm
CLIP model based on EVA02 architecture, suitable for zero-shot image classification tasks
Text-to-Image
Safetensors
Open Source License:MIT#Zero-shot Image Classification#CLIP Architecture#Multimodal Understanding
Downloads 29.73k
Release Time : 4/10/2023

Model Overview

This model is a CLIP model based on the EVA02 architecture, specifically designed for zero-shot image classification tasks. It combines visual and language understanding capabilities, enabling classification without training data for specific categories.

Model Features

Zero-shot Learning Capability
Capable of classification without training data for specific categories
Vision-Language Joint Modeling
Simultaneously understands image content and related text descriptions
Efficient Architecture
Improved architecture based on EVA02, balancing performance and efficiency

Model Capabilities

Zero-shot Image Classification
Image-Text Matching
Cross-modal Understanding

Use Cases

Image Classification
Open-domain Image Classification
Classify images of unseen categories
Performs well on various zero-shot classification benchmarks
Content Retrieval
Cross-modal Retrieval
Retrieve images based on text descriptions or generate descriptions based on images
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