đ vit-base-patch16-224-in21k_GI_diagnosis
This is a fine - tuned image classification model for GI diagnosis, achieving high accuracy on the evaluation set.
đ Quick Start
This model is a fine - tuned version of [google/vit - base - patch16 - 224 - in21k](https://huggingface.co/google/vit - base - patch16 - 224 - in21k). It achieves the following results on the evaluation set:
- Loss: 0.2538
- Accuracy: 0.9375
- Weighted f1: 0.9365
- Micro f1: 0.9375
- Macro f1: 0.9365
- Weighted recall: 0.9375
- Micro recall: 0.9375
- Macro recall: 0.9375
- Weighted precision: 0.9455
- Micro precision: 0.9375
- Macro precision: 0.9455
⨠Features
Model description
This is a multiclass image classification model of GI diagnosis'.
For more information on how it was created, check out the following link: https://github.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/blob/main/Computer%20Vision/Image%20Classification/Multiclass%20Classification/Diagnoses%20from%20Colonoscopy%20Images/diagnosis_from_colonoscopy_image_ViT.ipynb
Intended uses & limitations
This model is intended to demonstrate my ability to solve a complex problem using technology.
Training and evaluation data
Dataset Source: https://www.kaggle.com/datasets/francismon/curated - colon - dataset - for - deep - learning
đ§ Technical Details
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: 3
Training results
Training Loss |
Epoch |
Step |
Validation Loss |
Accuracy |
Weighted f1 |
Micro f1 |
Macro f1 |
Weighted recall |
Micro recall |
Macro recall |
Weighted precision |
Micro precision |
Macro precision |
1.3805 |
1.0 |
200 |
0.5006 |
0.8638 |
0.8531 |
0.8638 |
0.8531 |
0.8638 |
0.8638 |
0.8638 |
0.9111 |
0.8638 |
0.9111 |
1.3805 |
2.0 |
400 |
0.2538 |
0.9375 |
0.9365 |
0.9375 |
0.9365 |
0.9375 |
0.9375 |
0.9375 |
0.9455 |
0.9375 |
0.9455 |
0.0628 |
3.0 |
600 |
0.5797 |
0.8812 |
0.8740 |
0.8812 |
0.8740 |
0.8812 |
0.8812 |
0.8813 |
0.9157 |
0.8812 |
0.9157 |
Framework versions
- Transformers 4.22.2
- Pytorch 1.12.1
- Datasets 2.5.2
- Tokenizers 0.12.1
đ License
This model is licensed under the Apache 2.0 license.
Property |
Details |
Model Type |
Fine - tuned image classification model for GI diagnosis |
Training Data |
https://www.kaggle.com/datasets/francismon/curated - colon - dataset - for - deep - learning |