convnext-tiny-upgrade-384-batch-32
This model is a fine-tuned version of facebook/convnextv2-tiny-22k-384 on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2521
- Accuracy: 0.9298
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10
Training results
| Training Loss |
Epoch |
Step |
Validation Loss |
Accuracy |
| 0.9343 |
1.0 |
550 |
0.5732 |
0.8410 |
| 0.6456 |
2.0 |
1100 |
0.4130 |
0.8843 |
| 0.5478 |
3.0 |
1650 |
0.3537 |
0.9026 |
| 0.466 |
4.0 |
2200 |
0.3012 |
0.9181 |
| 0.4619 |
5.0 |
2750 |
0.3031 |
0.9141 |
| 0.4046 |
6.0 |
3300 |
0.2971 |
0.9157 |
| 0.3852 |
7.0 |
3850 |
0.2763 |
0.9205 |
| 0.3346 |
8.0 |
4400 |
0.2712 |
0.9225 |
| 0.3386 |
9.0 |
4950 |
0.2672 |
0.9221 |
| 0.3462 |
10.0 |
5500 |
0.2655 |
0.9245 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2
