|
--- |
|
license: apache-2.0 |
|
base_model: google/vit-large-patch32-384 |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- imagefolder |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: vitLarge-p32-384-2e-4-batch_16_epoch_4_classes_24 |
|
results: |
|
- task: |
|
name: Image Classification |
|
type: image-classification |
|
dataset: |
|
name: imagefolder |
|
type: imagefolder |
|
config: default |
|
split: train |
|
args: default |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.978448275862069 |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# vitLarge-p32-384-2e-4-batch_16_epoch_4_classes_24 |
|
|
|
This model is a fine-tuned version of [google/vit-large-patch32-384](https://huggingface.co/google/vit-large-patch32-384) on the imagefolder dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0894 |
|
- Accuracy: 0.9784 |
|
|
|
## 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: 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 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| 0.3286 | 0.07 | 100 | 0.3636 | 0.8764 | |
|
| 0.3196 | 0.14 | 200 | 0.4602 | 0.875 | |
|
| 0.1705 | 0.21 | 300 | 0.2507 | 0.9152 | |
|
| 0.2873 | 0.28 | 400 | 0.2614 | 0.9282 | |
|
| 0.0982 | 0.35 | 500 | 0.2327 | 0.9310 | |
|
| 0.1669 | 0.42 | 600 | 0.3202 | 0.9210 | |
|
| 0.1612 | 0.49 | 700 | 0.5062 | 0.8807 | |
|
| 0.1532 | 0.56 | 800 | 0.2240 | 0.9425 | |
|
| 0.1728 | 0.63 | 900 | 0.1601 | 0.9511 | |
|
| 0.167 | 0.7 | 1000 | 0.3861 | 0.9138 | |
|
| 0.0752 | 0.77 | 1100 | 0.2198 | 0.9483 | |
|
| 0.0423 | 0.84 | 1200 | 0.2114 | 0.9440 | |
|
| 0.0898 | 0.91 | 1300 | 0.1613 | 0.9511 | |
|
| 0.0498 | 0.97 | 1400 | 0.2824 | 0.9382 | |
|
| 0.016 | 1.04 | 1500 | 0.1921 | 0.9569 | |
|
| 0.0079 | 1.11 | 1600 | 0.2548 | 0.9382 | |
|
| 0.008 | 1.18 | 1700 | 0.2186 | 0.9497 | |
|
| 0.049 | 1.25 | 1800 | 0.2018 | 0.9569 | |
|
| 0.0333 | 1.32 | 1900 | 0.1676 | 0.9598 | |
|
| 0.1119 | 1.39 | 2000 | 0.1601 | 0.9583 | |
|
| 0.0134 | 1.46 | 2100 | 0.1157 | 0.9741 | |
|
| 0.0192 | 1.53 | 2200 | 0.1320 | 0.9641 | |
|
| 0.0085 | 1.6 | 2300 | 0.1590 | 0.9641 | |
|
| 0.0384 | 1.67 | 2400 | 0.0973 | 0.9741 | |
|
| 0.0531 | 1.74 | 2500 | 0.1719 | 0.9569 | |
|
| 0.0221 | 1.81 | 2600 | 0.1280 | 0.9741 | |
|
| 0.0006 | 1.88 | 2700 | 0.1895 | 0.9540 | |
|
| 0.0006 | 1.95 | 2800 | 0.1258 | 0.9713 | |
|
| 0.016 | 2.02 | 2900 | 0.1105 | 0.9713 | |
|
| 0.0004 | 2.09 | 3000 | 0.1118 | 0.9684 | |
|
| 0.0001 | 2.16 | 3100 | 0.0936 | 0.9684 | |
|
| 0.0003 | 2.23 | 3200 | 0.0932 | 0.9684 | |
|
| 0.0001 | 2.3 | 3300 | 0.1247 | 0.9713 | |
|
| 0.0004 | 2.37 | 3400 | 0.0897 | 0.9741 | |
|
| 0.0001 | 2.44 | 3500 | 0.0853 | 0.9784 | |
|
| 0.0002 | 2.51 | 3600 | 0.0948 | 0.9770 | |
|
| 0.0002 | 2.58 | 3700 | 0.0957 | 0.9770 | |
|
| 0.0043 | 2.65 | 3800 | 0.0868 | 0.9756 | |
|
| 0.0001 | 2.72 | 3900 | 0.0904 | 0.9741 | |
|
| 0.0011 | 2.79 | 4000 | 0.0881 | 0.9770 | |
|
| 0.0001 | 2.86 | 4100 | 0.0890 | 0.9784 | |
|
| 0.0001 | 2.92 | 4200 | 0.0896 | 0.9784 | |
|
| 0.0001 | 2.99 | 4300 | 0.0894 | 0.9784 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.39.3 |
|
- Pytorch 2.1.2 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
