|
---
|
|
library_name: transformers
|
|
license: apache-2.0
|
|
base_model: google/vit-base-patch16-224-in21k
|
|
tags:
|
|
- generated_from_trainer
|
|
datasets:
|
|
- imagefolder
|
|
metrics:
|
|
- accuracy
|
|
- precision
|
|
- recall
|
|
- f1
|
|
model-index:
|
|
- name: vit-finetune-kidney-stone-Michel_Daudon_-w256_1k_v1-_SEC-pretrain
|
|
results:
|
|
- task:
|
|
name: Image Classification
|
|
type: image-classification
|
|
dataset:
|
|
name: imagefolder
|
|
type: imagefolder
|
|
config: default
|
|
split: test
|
|
args: default
|
|
metrics:
|
|
- name: Accuracy
|
|
type: accuracy
|
|
value: 0.9108333333333334
|
|
- name: Precision
|
|
type: precision
|
|
value: 0.9190361753451352
|
|
- name: Recall
|
|
type: recall
|
|
value: 0.9108333333333334
|
|
- name: F1
|
|
type: f1
|
|
value: 0.9102828889161464
|
|
---
|
|
|
|
<!-- 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-finetune-kidney-stone-Michel_Daudon_-w256_1k_v1-_SEC-pretrain
|
|
|
|
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 imagefolder dataset.
|
|
It achieves the following results on the evaluation set:
|
|
- Loss: 0.3455
|
|
- Accuracy: 0.9108
|
|
- Precision: 0.9190
|
|
- Recall: 0.9108
|
|
- F1: 0.9103
|
|
|
|
## 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: 0.0002
|
|
- train_batch_size: 32
|
|
- eval_batch_size: 8
|
|
- seed: 42
|
|
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
|
- lr_scheduler_type: linear
|
|
- num_epochs: 15
|
|
- mixed_precision_training: Native AMP
|
|
|
|
### Training results
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|
|
|:-------------:|:-------:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
|
|
| 0.1494 | 0.6667 | 100 | 0.6088 | 0.8442 | 0.8766 | 0.8442 | 0.8390 |
|
|
| 0.0665 | 1.3333 | 200 | 0.5533 | 0.8492 | 0.8810 | 0.8492 | 0.8542 |
|
|
| 0.0215 | 2.0 | 300 | 0.3721 | 0.9017 | 0.9082 | 0.9017 | 0.8985 |
|
|
| 0.0101 | 2.6667 | 400 | 0.5347 | 0.8942 | 0.9061 | 0.8942 | 0.8920 |
|
|
| 0.043 | 3.3333 | 500 | 0.7850 | 0.8425 | 0.8592 | 0.8425 | 0.8427 |
|
|
| 0.0641 | 4.0 | 600 | 0.7735 | 0.8583 | 0.8770 | 0.8583 | 0.8574 |
|
|
| 0.0036 | 4.6667 | 700 | 0.7351 | 0.8367 | 0.8623 | 0.8367 | 0.8250 |
|
|
| 0.0039 | 5.3333 | 800 | 0.3455 | 0.9108 | 0.9190 | 0.9108 | 0.9103 |
|
|
| 0.0021 | 6.0 | 900 | 0.5940 | 0.8758 | 0.8985 | 0.8758 | 0.8730 |
|
|
| 0.054 | 6.6667 | 1000 | 0.7463 | 0.8733 | 0.9068 | 0.8733 | 0.8714 |
|
|
| 0.0015 | 7.3333 | 1100 | 0.8915 | 0.8392 | 0.8722 | 0.8392 | 0.8243 |
|
|
| 0.0013 | 8.0 | 1200 | 0.5725 | 0.8917 | 0.8943 | 0.8917 | 0.8909 |
|
|
| 0.0011 | 8.6667 | 1300 | 0.5772 | 0.8933 | 0.8960 | 0.8933 | 0.8926 |
|
|
| 0.001 | 9.3333 | 1400 | 0.5820 | 0.8933 | 0.8956 | 0.8933 | 0.8926 |
|
|
| 0.0009 | 10.0 | 1500 | 0.5859 | 0.8933 | 0.8954 | 0.8933 | 0.8925 |
|
|
| 0.0008 | 10.6667 | 1600 | 0.5901 | 0.8933 | 0.8955 | 0.8933 | 0.8926 |
|
|
| 0.0008 | 11.3333 | 1700 | 0.5938 | 0.8933 | 0.8955 | 0.8933 | 0.8926 |
|
|
| 0.0007 | 12.0 | 1800 | 0.5971 | 0.8933 | 0.8953 | 0.8933 | 0.8925 |
|
|
| 0.0007 | 12.6667 | 1900 | 0.5998 | 0.8933 | 0.8952 | 0.8933 | 0.8926 |
|
|
| 0.0007 | 13.3333 | 2000 | 0.6016 | 0.8933 | 0.8952 | 0.8933 | 0.8926 |
|
|
| 0.0006 | 14.0 | 2100 | 0.6032 | 0.8933 | 0.8952 | 0.8933 | 0.8926 |
|
|
| 0.0006 | 14.6667 | 2200 | 0.6039 | 0.8933 | 0.8952 | 0.8933 | 0.8926 |
|
|
|
|
|
|
### Framework versions
|
|
|
|
- Transformers 4.48.2
|
|
- Pytorch 2.6.0+cu126
|
|
- Datasets 3.2.0
|
|
- Tokenizers 0.21.0
|
|
|
