|
--- |
|
library_name: transformers |
|
license: apache-2.0 |
|
base_model: google/vit-base-patch16-224 |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- medmnist-v2 |
|
metrics: |
|
- accuracy |
|
- f1 |
|
model-index: |
|
- name: ViT_breastmnist_std_15 |
|
results: |
|
- task: |
|
name: Image Classification |
|
type: image-classification |
|
dataset: |
|
name: medmnist-v2 |
|
type: medmnist-v2 |
|
config: breastmnist |
|
split: validation |
|
args: breastmnist |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.7884615384615384 |
|
- name: F1 |
|
type: f1 |
|
value: 0.6551215917464996 |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# ViT_breastmnist_std_15 |
|
|
|
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the medmnist-v2 dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4504 |
|
- Accuracy: 0.7885 |
|
- F1: 0.6551 |
|
|
|
## 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: 2e-05 |
|
- train_batch_size: 64 |
|
- eval_batch_size: 16 |
|
- 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 | |
|
|:-------------:|:------:|:----:|:---------------:|:--------:|:------:| |
|
| 0.4628 | 0.2597 | 20 | 0.4724 | 0.7821 | 0.5951 | |
|
| 0.3645 | 0.5195 | 40 | 0.3994 | 0.8590 | 0.7786 | |
|
| 0.2744 | 0.7792 | 60 | 0.4429 | 0.8462 | 0.7524 | |
|
| 0.3004 | 1.0390 | 80 | 0.3893 | 0.8590 | 0.7886 | |
|
| 0.2153 | 1.2987 | 100 | 0.4120 | 0.8462 | 0.7641 | |
|
| 0.1593 | 1.5584 | 120 | 0.4542 | 0.8590 | 0.7786 | |
|
| 0.1189 | 1.8182 | 140 | 0.3911 | 0.8718 | 0.8120 | |
|
| 0.1139 | 2.0779 | 160 | 0.4154 | 0.8590 | 0.7886 | |
|
| 0.0707 | 2.3377 | 180 | 0.4517 | 0.8590 | 0.7886 | |
|
| 0.0482 | 2.5974 | 200 | 0.4824 | 0.8718 | 0.8034 | |
|
| 0.0499 | 2.8571 | 220 | 0.4408 | 0.8462 | 0.7743 | |
|
| 0.0195 | 3.1169 | 240 | 0.4874 | 0.8462 | 0.7743 | |
|
| 0.0146 | 3.3766 | 260 | 0.4723 | 0.8718 | 0.8120 | |
|
| 0.0141 | 3.6364 | 280 | 0.5117 | 0.8590 | 0.7886 | |
|
| 0.017 | 3.8961 | 300 | 0.6032 | 0.8462 | 0.7743 | |
|
| 0.0052 | 4.1558 | 320 | 0.5948 | 0.8590 | 0.7886 | |
|
| 0.005 | 4.4156 | 340 | 0.5897 | 0.8590 | 0.7886 | |
|
| 0.0039 | 4.6753 | 360 | 0.5729 | 0.8462 | 0.7743 | |
|
| 0.0088 | 4.9351 | 380 | 0.5623 | 0.8462 | 0.7743 | |
|
| 0.0104 | 5.1948 | 400 | 0.4814 | 0.8718 | 0.8194 | |
|
| 0.0012 | 5.4545 | 420 | 0.5039 | 0.8718 | 0.8194 | |
|
| 0.001 | 5.7143 | 440 | 0.5268 | 0.8718 | 0.8120 | |
|
| 0.001 | 5.9740 | 460 | 0.5435 | 0.8590 | 0.7886 | |
|
| 0.0007 | 6.2338 | 480 | 0.5435 | 0.8462 | 0.7743 | |
|
| 0.0007 | 6.4935 | 500 | 0.5373 | 0.8590 | 0.7974 | |
|
| 0.0006 | 6.7532 | 520 | 0.5745 | 0.8590 | 0.7886 | |
|
| 0.0007 | 7.0130 | 540 | 0.5674 | 0.8462 | 0.7743 | |
|
| 0.0004 | 7.2727 | 560 | 0.5826 | 0.8462 | 0.7743 | |
|
| 0.0006 | 7.5325 | 580 | 0.5663 | 0.8462 | 0.7743 | |
|
| 0.0006 | 7.7922 | 600 | 0.5751 | 0.8462 | 0.7743 | |
|
| 0.0005 | 8.0519 | 620 | 0.5851 | 0.8462 | 0.7743 | |
|
| 0.0004 | 8.3117 | 640 | 0.5782 | 0.8462 | 0.7743 | |
|
| 0.0004 | 8.5714 | 660 | 0.5875 | 0.8462 | 0.7743 | |
|
| 0.0004 | 8.8312 | 680 | 0.5939 | 0.8462 | 0.7743 | |
|
| 0.0004 | 9.0909 | 700 | 0.5934 | 0.8462 | 0.7743 | |
|
| 0.0004 | 9.3506 | 720 | 0.5925 | 0.8462 | 0.7743 | |
|
| 0.0004 | 9.6104 | 740 | 0.5930 | 0.8462 | 0.7743 | |
|
| 0.0004 | 9.8701 | 760 | 0.5945 | 0.8462 | 0.7743 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.45.1 |
|
- Pytorch 2.4.0 |
|
- Datasets 3.0.1 |
|
- Tokenizers 0.20.0 |
|
|
