AIbase
EN
Home
Vit L 14 336
V

Vit L 14 336

Developed by asakhare
Large-scale vision-language model based on Vision Transformer architecture, supporting zero-shot image classification tasks
Image Classification
Safetensors
Open Source License:MIT#Zero-shot Image Classification#High-resolution Processing#Multimodal Understanding
Downloads 20
Release Time : 1/4/2024

Model Overview

This model is part of the OpenCLIP project, utilizing the ViT-L/14 architecture with an input resolution of 336x336, focusing on cross-modal vision-language understanding, particularly suitable for zero-shot image classification scenarios.

Model Features

Zero-shot Learning Capability
Performs image classification on new categories without task-specific fine-tuning
High-resolution Processing
Supports input resolution of 336x336 pixels, capturing finer visual features
Cross-modal Understanding
Simultaneously comprehends visual and textual information for image-text matching

Model Capabilities

Zero-shot Image Classification
Image-Text Matching
Visual Feature Extraction

Use Cases

Content Management
Automatic Image Tagging
Automatically generates descriptive tags for unlabeled images
Improves content retrieval efficiency
E-commerce
Product Categorization
Automatically classifies product images into catalog categories
Reduces manual categorization workload
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