AIbase
EN
Home
Eva02 Enormous Patch14 Plus Clip 224.laion2b S9b B144k
E

Eva02 Enormous Patch14 Plus Clip 224.laion2b S9b B144k

Developed by timm
Large-scale vision-language model based on EVA02 architecture, supporting zero-shot image classification tasks
Text-to-Image
Safetensors
Open Source License:MIT#Zero-shot Image Classification#Large-scale Pretraining#Multimodal Understanding
Downloads 12.57k
Release Time : 4/11/2023

Model Overview

This model is a variant of the CLIP architecture, incorporating EVA02's visual encoder for joint representation learning of images and text, particularly excelling in zero-shot image classification tasks

Model Features

Zero-shot Learning Capability
Capable of performing image classification tasks without task-specific fine-tuning
Large-scale Pretraining
Pretrained on the LAION-2B dataset, possessing strong visual-language understanding capabilities
Efficient Visual Encoding
Utilizes EVA02 architecture's visual encoder for efficient image feature extraction

Model Capabilities

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

Use Cases

Content Management
Automatic Image Tagging
Automatically generates descriptive tags for unlabeled images
Enhances content management efficiency and reduces manual labeling costs
E-commerce
Product Categorization
Automatically classifies product images into relevant categories
Supports flexible product categorization without predefined fixed categories
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase