convnext-large-224-22k-1k-convnext_bottom
This model is a fine-tuned version of facebook/convnext-large-224-22k-1k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0064
- Accuracy: 0.9981
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 7
- total_train_batch_size: 56
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.0899 | 0.99 | 86 | 0.0290 | 0.9852 |
| 0.0651 | 2.0 | 173 | 0.0217 | 0.9889 |
| 0.0364 | 2.99 | 259 | 0.0170 | 0.9944 |
| 0.0678 | 4.0 | 346 | 0.0135 | 0.9963 |
| 0.0129 | 4.99 | 432 | 0.0120 | 0.9944 |
| 0.0189 | 6.0 | 519 | 0.0095 | 0.9944 |
| 0.0399 | 7.0 | 606 | 0.0098 | 0.9944 |
| 0.029 | 7.99 | 692 | 0.0121 | 0.9963 |
| 0.0153 | 9.0 | 779 | 0.0068 | 0.9981 |
| 0.0252 | 9.93 | 860 | 0.0064 | 0.9981 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3