|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_keras_callback |
|
- vision_transformer |
|
model-index: |
|
- name: Guldeniz/vit-base-patch16-224-in21k-lung_and_colon |
|
results: [] |
|
language: |
|
- en |
|
metrics: |
|
- accuracy |
|
library_name: transformers |
|
pipeline_tag: image-classification |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information Keras had access to. You should |
|
probably proofread and complete it, then remove this comment. --> |
|
|
|
# Guldeniz/vit-base-patch16-224-in21k-lung_and_colon |
|
|
|
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on Lung and Colon Histopathological Images dataset. This dataset can be reach via [Kaggle](https://www.kaggle.com/datasets/andrewmvd/lung-and-colon-cancer-histopathological-images). |
|
It achieves the following results on the evaluation set: |
|
- Train Loss: 0.0088 |
|
- Train Accuracy: 1.0 |
|
- Train Top-3-accuracy: 1.0 |
|
- Validation Loss: 0.0084 |
|
- Validation Accuracy: 0.9997 |
|
- Validation Top-3-accuracy: 1.0 |
|
- Epoch: 3 |
|
|
|
## Model description |
|
|
|
The vision transformer model, trained by Google, has been fine-tuned using a lung and colon cancer image dataset consisting of a total of 25,000 images across 5 labels. The obtained results are highly promising, and the model demonstrates the ability to predict the following listed labels. |
|
|
|
- colon_aca |
|
|
|
- colon_n |
|
|
|
- lung_aca |
|
|
|
- lung_n |
|
|
|
- lung_scc |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 3325, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01} |
|
- training_precision: float32 |
|
|
|
### Training results |
|
|
|
| Train Loss | Train Accuracy | Train Top-3-accuracy | Validation Loss | Validation Accuracy | Validation Top-3-accuracy | Epoch | |
|
|:----------:|:--------------:|:--------------------:|:---------------:|:-------------------:|:-------------------------:|:-----:| |
|
| 0.1870 | 0.9784 | 0.9985 | 0.0455 | 0.9987 | 1.0 | 0 | |
|
| 0.0345 | 0.9972 | 1.0 | 0.0189 | 0.9995 | 1.0 | 1 | |
|
| 0.0134 | 1.0 | 1.0 | 0.0110 | 0.9997 | 1.0 | 2 | |
|
| 0.0088 | 1.0 | 1.0 | 0.0084 | 0.9997 | 1.0 | 3 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.26.1 |
|
- TensorFlow 2.12.0 |
|
- Datasets 2.10.1 |
|
- Tokenizers 0.13.3 |