Open-source Image Classification Model exper_batch_16_e4 - Free Deployment for High-accuracy Image Classification

Exper Batch 16 E4

Developed by sudo-s

An image classification model fine-tuned on the sudo-s/herbier_mesuem1 dataset based on google/vit-base-patch16-224-in21k, achieving an accuracy of 90.59%

Image Classification

Transformers

Open Source License:Apache-2.0 #Plant Image Classification #High Accuracy #ViT Fine-tuning

Downloads 30

Release Time : 6/26/2022

Model Overview

This model is an image classification model based on the Vision Transformer (ViT) architecture, fine-tuned on a specific dataset, primarily used for image classification tasks.

Model Features

High Accuracy

Achieves an accuracy of 90.59% on the evaluation set, demonstrating excellent performance

Based on ViT Architecture

Utilizes the Vision Transformer architecture, which has strong image feature extraction capabilities

Efficient Fine-tuning

Achieves good performance within a relatively small number of training epochs (4 epochs)

Model Capabilities

Image Classification

Feature Extraction

Use Cases

Botanical Research

Plant Specimen Classification

Classify and identify plant specimen images

Accuracy: 90.59%

General Image Classification

General Object Recognition

Can be used for image classification tasks of various objects

🚀 exper_batch_16_e4

This model is a fine - tuned version of google/vit-base-patch16-224-in21k on the sudo - s/herbier_mesuem1 dataset. It can be used for image classification tasks and is generated from a trainer. The evaluation metric used is accuracy.

🚀 Quick Start

This model achieves the following results on the evaluation set:

Loss: 0.3598
Accuracy: 0.9059

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e - 08
lr_scheduler_type: linear
num_epochs: 4
mixed_precision_training: Apex, opt level O1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.7606	0.16	100	3.7839	0.1989
3.1072	0.31	200	3.0251	0.3285
2.4068	0.47	300	2.4380	0.4719
2.0881	0.63	400	2.0489	0.5412
1.6817	0.78	500	1.7968	0.6025
1.342	0.94	600	1.5044	0.6249
0.9343	1.1	700	1.1881	0.7132
0.9552	1.25	800	1.1064	0.7224
0.7265	1.41	900	0.9189	0.7768
0.6732	1.56	1000	0.9227	0.7606
0.5587	1.72	1100	0.7912	0.7903
0.6332	1.88	1200	0.7606	0.7945
0.3188	2.03	1300	0.6535	0.8288
0.3079	2.19	1400	0.5686	0.8577
0.2518	2.35	1500	0.5517	0.8577
0.2	2.5	1600	0.5277	0.8631
0.2032	2.66	1700	0.4841	0.8701
0.1555	2.82	1800	0.4578	0.8793
0.145	2.97	1900	0.4466	0.8755
0.0985	3.13	2000	0.4249	0.8867
0.0955	3.29	2100	0.3977	0.8932
0.0438	3.44	2200	0.3785	0.9036
0.0589	3.6	2300	0.3717	0.9017
0.0709	3.76	2400	0.3609	0.9052
0.0706	3.91	2500	0.3598	0.9059

Framework versions

Transformers 4.19.4
Pytorch 1.5.1
Datasets 2.3.2
Tokenizers 0.12.1

📄 License

This model is licensed under the Apache - 2.0 license.

Property	Details
Model Type	Fine - tuned version of google/vit - base - patch16 - 224 - in21k for image classification
Training Data	sudo - s/herbier_mesuem1 dataset
Metrics	accuracy

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご