Vit-mlo-512-breat_composition Open-Source Medical Imaging Model

Vit Mlo 512 Breat Composition

Developed by mm-ai

This is a medical image classification model based on the ViT architecture, specifically designed for breast composition analysis, fine-tuned on the preprocessed 1024 configuration dataset.

Image Classification

Transformers

#Breast Image Classification #Medical Imaging Analysis #ViT Fine-tuned Model

Downloads 35

Release Time : 2/28/2023

Model Overview

This model is used for medical image classification tasks, particularly breast composition analysis. It achieved approximately 57.9% accuracy and 57.5% F1 score on the evaluation set.

Model Features

Medical Image Analysis

A medical image classification model specifically designed for breast composition analysis.

ViT-based Architecture

Utilizes the Vision Transformer architecture, suitable for processing medical images.

Fine-tuning Optimization

Fine-tuned on the preprocessed 1024 configuration dataset to meet specific task requirements.

Model Capabilities

Medical Image Classification

Breast Composition Analysis

Image Feature Extraction

Use Cases

Medical Diagnosis

Breast Composition Classification

Used to analyze the composition of breast images.

Accuracy 57.9%, F1 score 57.5%

🚀 vit-mlo-512-breat_composition

This model is a fine - tuned version on the preprocessed1024_config dataset, aiming to solve image classification tasks and providing corresponding accuracy and F1 score.

🚀 Quick Start

This model is a fine - tuned version of on the preprocessed1024_config dataset. It achieves the following results on the evaluation set:

Loss: 1.3123
Accuracy: {'accuracy': 0.5791457286432161}
F1: {'f1': 0.5749067914290308}

📚 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: 5e - 05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.2679	1.0	796	1.0281	{'accuracy': 0.5062814070351759}	{'f1': 0.38950358034816535}
0.9805	2.0	1592	0.9240	{'accuracy': 0.5672110552763819}	{'f1': 0.5273112700912543}
0.9167	3.0	2388	0.9608	{'accuracy': 0.5477386934673367}	{'f1': 0.45736748568671376}
0.8292	4.0	3184	0.8973	{'accuracy': 0.5891959798994975}	{'f1': 0.5783349603036094}
0.7695	5.0	3980	1.0477	{'accuracy': 0.5571608040201005}	{'f1': 0.5379432393338944}
0.6912	6.0	4776	0.9479	{'accuracy': 0.585427135678392}	{'f1': 0.5766494177636581}
0.61	7.0	5572	1.1280	{'accuracy': 0.5703517587939698}	{'f1': 0.5560158679652624}
0.5591	8.0	6368	1.1866	{'accuracy': 0.5741206030150754}	{'f1': 0.5541999644498281}
0.5021	9.0	7164	1.1537	{'accuracy': 0.582286432160804}	{'f1': 0.566315815243799}
0.4262	10.0	7960	1.3123	{'accuracy': 0.5791457286432161}	{'f1': 0.5749067914290308}

Framework versions

Transformers 4.20.1
Pytorch 1.12.0
Datasets 2.1.0
Tokenizers 0.12.1

Property	Details
Model Type	vit - mlo - 512 - breat_composition
Training Data	preprocessed1024_config
Metrics	Accuracy, F1

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