Vit Base Patch16 224 In21k Wr
This model is a fine-tuned Vision Transformer based on google/vit-base-patch16-224-in21k on an unknown dataset, primarily used for image classification tasks.
Downloads 21
Release Time : 9/7/2022
Model Overview
This is an image classification model based on the Vision Transformer architecture, fine-tuned on an unknown dataset, suitable for general image recognition tasks.
Model Features
Fine-tuned based on pre-trained model
Fine-tuned on the google/vit-base-patch16-224-in21k pre-trained model, inheriting powerful image feature extraction capabilities
Mixed precision training
Trained using mixed_float16 precision, balancing training speed and model accuracy
Optimizer configuration
Uses AdamWeightDecay optimizer with PolynomialDecay learning rate scheduling, helping to stabilize the training process
Model Capabilities
Image classification
Feature extraction
Use Cases
Computer vision
General image classification
Can be used to classify common objects and scenes
Validation accuracy 57.7%, top-3 accuracy 80.35%
Featured Recommended AI Models
Š 2025AIbase