Modeversion1_m7_e4 Open-source Image Classification Model - High Accuracy for Quick Image Classification

Modeversion1 M7 E4

Developed by sudo-s

This model is an image classification model fine-tuned on the sudo-s/herbier_mesuem7 dataset based on google/vit-base-patch16-224-in21k, achieving an evaluation accuracy of 97.31%.

Image Classification

Transformers

Open Source License:Apache-2.0 #High-precision Image Classification #Plant Specimen Identification #ViT Fine-tuned Model

Downloads 28

Release Time : 7/23/2022

Model Overview

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

Model Features

High Accuracy

Achieves a classification accuracy of 97.31% on the evaluation set.

ViT-based Architecture

Utilizes the Vision Transformer base architecture, suitable for processing image data.

Fine-tuned Optimization

Specifically fine-tuned on a dedicated plant specimen dataset.

Model Capabilities

Image Classification

Plant Specimen Identification

Use Cases

Botanical Research

Plant Specimen Classification

Automatically classifies images of different plant specimens.

97.31% classification accuracy

Museum Digitization

Specimen Collection Management

Assists museums in the digital classification management of plant specimens.

🚀 modeversion1_m7_e4

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_mesuem7 dataset. It offers high - accuracy image classification capabilities, achieving excellent results on the evaluation set.

🚀 Quick Start

This model is ready to use for image classification tasks right after fine - tuning on the sudo - s/herbier_mesuem7 dataset.

📚 Documentation

Model Performance

It achieves the following results on the evaluation set:

Loss: 0.0902
Accuracy: 0.9731

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: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.073	0.06	100	3.9370	0.1768
3.4186	0.12	200	3.2721	0.2590
2.6745	0.18	300	2.6465	0.3856
2.2806	0.23	400	2.2600	0.4523
1.9275	0.29	500	1.9653	0.5109
1.6958	0.35	600	1.6815	0.6078
1.2797	0.41	700	1.4514	0.6419
1.3772	0.47	800	1.3212	0.6762
1.1765	0.53	900	1.1476	0.7028
1.0152	0.59	1000	1.0357	0.7313
0.7861	0.64	1100	1.0230	0.7184
1.0262	0.7	1200	0.9469	0.7386
0.8905	0.76	1300	0.8184	0.7756
0.6919	0.82	1400	0.8083	0.7711
0.7494	0.88	1500	0.7601	0.7825
0.5078	0.94	1600	0.6884	0.8056
0.7134	1.0	1700	0.6311	0.8160
0.4328	1.06	1800	0.5740	0.8252
0.4971	1.11	1900	0.5856	0.8290
0.5207	1.17	2000	0.6219	0.8167
0.4027	1.23	2100	0.5703	0.8266
0.5605	1.29	2200	0.5217	0.8372
0.2723	1.35	2300	0.4805	0.8565
0.401	1.41	2400	0.4811	0.8490
0.3419	1.47	2500	0.4619	0.8608
0.301	1.52	2600	0.4318	0.8712
0.2872	1.58	2700	0.4698	0.8573
0.2451	1.64	2800	0.4210	0.8729
0.2211	1.7	2900	0.3645	0.8851
0.3145	1.76	3000	0.4139	0.8715
0.2001	1.82	3100	0.3605	0.8864
0.3095	1.88	3200	0.4274	0.8675
0.1915	1.93	3300	0.2910	0.9101
0.2465	1.99	3400	0.2726	0.9103
0.1218	2.05	3500	0.2742	0.9129
0.0752	2.11	3600	0.2572	0.9183
0.1067	2.17	3700	0.2584	0.9203
0.0838	2.23	3800	0.2458	0.9212
0.1106	2.29	3900	0.2412	0.9237
0.092	2.34	4000	0.2232	0.9277
0.1056	2.4	4100	0.2817	0.9077
0.0696	2.46	4200	0.2334	0.9285
0.0444	2.52	4300	0.2142	0.9363
0.1046	2.58	4400	0.2036	0.9352
0.066	2.64	4500	0.2115	0.9365
0.0649	2.7	4600	0.1730	0.9448
0.0513	2.75	4700	0.2148	0.9339
0.0917	2.81	4800	0.1810	0.9438
0.0879	2.87	4900	0.1971	0.9388
0.1052	2.93	5000	0.1602	0.9508
0.0362	2.99	5100	0.1475	0.9556
0.041	3.05	5200	0.1328	0.9585
0.0156	3.11	5300	0.1389	0.9571
0.0047	3.17	5400	0.1224	0.9638
0.0174	3.22	5500	0.1193	0.9651
0.0087	3.28	5600	0.1276	0.9622
0.0084	3.34	5700	0.1134	0.9662
0.0141	3.4	5800	0.1239	0.9631
0.0291	3.46	5900	0.1199	0.9645
0.0049	3.52	6000	0.1103	0.9679
0.0055	3.58	6100	0.1120	0.9662
0.0061	3.63	6200	0.1071	0.9668
0.0054	3.69	6300	0.1032	0.9697
0.0041	3.75	6400	0.0961	0.9711
0.0018	3.81	6500	0.0930	0.9718
0.0032	3.87	6600	0.0918	0.9730
0.0048	3.93	6700	0.0906	0.9732
0.002	3.99	6800	0.0902	0.9731

Framework versions

Transformers 4.20.1
Pytorch 1.12.0
Datasets 2.3.2
Tokenizers 0.12.1

📄 License

This model 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