Open-source image classification model exper_batch_32_e8 - Free to use and accurately identify image categories

Exper Batch 32 E8

Developed by sudo-s

An image classification model fine-tuned on the herbier_mesuem1 dataset based on google/vit-base-patch16-224-in21k, achieving 91.13% accuracy

Image Classification

Transformers

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

Downloads 31

Release Time : 6/26/2022

Model Overview

This model is an image classification model based on the Vision Transformer architecture, specifically optimized for plant specimen images

Model Features

High Accuracy

Achieves 91.13% classification accuracy on the evaluation set

Efficient Fine-tuning

Efficient fine-tuning based on pre-trained ViT model, requiring only 8 training epochs

Optimized Training

Uses mixed-precision training and linear learning rate scheduler to optimize the training process

Model Capabilities

Plant specimen image classification

High-precision visual recognition

Transfer learning applications

Use Cases

Botanical Research

Automatic Classification of Plant Specimens

Used for digital classification of plant specimens in museums or research institutions

Accuracy rate of 91.13%

Educational Applications

Plant Identification Teaching Tool

Can be used as an auxiliary identification tool for botany teaching

🚀 exper_batch_32_e8

This model is a fine - tuned version of [google/vit - base - patch16 - 224 - in21k](https://huggingface.co/google/vit - base - patch16 - 224 - in21k) on the sudo - s/herbier_mesuem1 dataset. It achieves the following results on the evaluation set:

Loss: 0.3520
Accuracy: 0.9113

🚀 Quick Start

This model is designed for image - classification tasks. You can use it by fine - tuning the pre - trained model on your own dataset.

✨ Features

Based on the pre - trained model google/vit - base - patch16 - 224 - in21k, which has strong generalization ability.
Achieved high accuracy (0.9113) on the evaluation set after fine - tuning.

📦 Installation

There is no specific installation steps provided in the original README.

📚 Documentation

Model Information

Property	Details
Model Type	Fine - tuned image - classification model
Training Data	sudo - s/herbier_mesuem1 dataset

Training Procedure

Training Hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 32
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
num_epochs: 8
mixed_precision_training: Apex, opt level O1

Training Results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.3787	0.31	100	3.3100	0.3566
2.3975	0.62	200	2.3196	0.5717
1.5578	0.94	300	1.6764	0.6461
1.0291	1.25	400	1.1713	0.7463
0.8185	1.56	500	0.9292	0.7953
0.6181	1.88	600	0.7732	0.8169
0.3873	2.19	700	0.6877	0.8277
0.2979	2.5	800	0.6250	0.8404
0.2967	2.81	900	0.6151	0.8365
0.1874	3.12	1000	0.5401	0.8608
0.2232	3.44	1100	0.5032	0.8712
0.1109	3.75	1200	0.4635	0.8774
0.0539	4.06	1300	0.4495	0.8843
0.0668	4.38	1400	0.4273	0.8951
0.0567	4.69	1500	0.4427	0.8867
0.0285	5.0	1600	0.4092	0.8955
0.0473	5.31	1700	0.3720	0.9071
0.0225	5.62	1800	0.3691	0.9063
0.0196	5.94	1900	0.3775	0.9048
0.0173	6.25	2000	0.3641	0.9040
0.0092	6.56	2100	0.3551	0.9090
0.008	6.88	2200	0.3591	0.9125
0.0072	7.19	2300	0.3542	0.9121
0.007	7.5	2400	0.3532	0.9106
0.007	7.81	2500	0.3520	0.9113

Framework Versions

Transformers 4.19.4
Pytorch 1.5.1
Datasets 2.3.2
Tokenizers 0.12.1

📄 License

This project is licensed under the Apache - 2.0 license.

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