|
--- |
|
license: apache-2.0 |
|
base_model: facebook/convnextv2-large-22k-224 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: ConvNextV2-large-DogBreed |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# ConvNextV2-large-DogBreed |
|
|
|
This model is a fine-tuned version of [facebook/convnextv2-large-22k-224](https://huggingface.co/facebook/convnextv2-large-22k-224) on dog breed classification dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5469 |
|
- Accuracy: 0.9139 |
|
|
|
## 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: 1e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 32 |
|
- total_train_batch_size: 512 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 50 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| 4.8578 | 1.0 | 13 | 4.6940 | 0.0671 | |
|
| 4.6332 | 1.99 | 26 | 4.4169 | 0.1784 | |
|
| 4.4095 | 2.99 | 39 | 4.1105 | 0.3485 | |
|
| 3.8841 | 3.98 | 52 | 3.7581 | 0.5198 | |
|
| 3.5964 | 4.98 | 65 | 3.3647 | 0.6647 | |
|
| 3.2781 | 5.97 | 78 | 2.9442 | 0.7677 | |
|
| 2.6006 | 6.97 | 91 | 2.5252 | 0.8180 | |
|
| 2.2638 | 7.96 | 104 | 2.1256 | 0.8467 | |
|
| 1.9609 | 8.96 | 117 | 1.7626 | 0.8766 | |
|
| 1.3962 | 9.95 | 130 | 1.4453 | 0.9042 | |
|
| 1.143 | 10.95 | 143 | 1.1818 | 0.9102 | |
|
| 0.9423 | 11.94 | 156 | 0.9697 | 0.9138 | |
|
| 0.7674 | 12.94 | 169 | 0.8097 | 0.9174 | |
|
| 0.5007 | 13.93 | 182 | 0.6922 | 0.9186 | |
|
| 0.4097 | 14.93 | 195 | 0.5999 | 0.9162 | |
|
| 0.3392 | 16.0 | 209 | 0.5174 | 0.9269 | |
|
| 0.2285 | 17.0 | 222 | 0.4685 | 0.9257 | |
|
| 0.184 | 17.99 | 235 | 0.4337 | 0.9210 | |
|
| 0.1587 | 18.99 | 248 | 0.4058 | 0.9257 | |
|
| 0.1112 | 19.98 | 261 | 0.3824 | 0.9222 | |
|
| 0.0967 | 20.98 | 274 | 0.3712 | 0.9150 | |
|
| 0.0838 | 21.97 | 287 | 0.3584 | 0.9186 | |
|
| 0.0665 | 22.97 | 300 | 0.3468 | 0.9174 | |
|
| 0.0589 | 23.96 | 313 | 0.3428 | 0.9186 | |
|
| 0.0551 | 24.96 | 326 | 0.3364 | 0.9186 | |
|
| 0.0512 | 25.95 | 339 | 0.3334 | 0.9162 | |
|
| 0.0441 | 26.95 | 352 | 0.3278 | 0.9210 | |
|
| 0.0428 | 27.94 | 365 | 0.3275 | 0.9150 | |
|
| 0.0387 | 28.94 | 378 | 0.3237 | 0.9210 | |
|
| 0.036 | 29.93 | 391 | 0.3242 | 0.9150 | |
|
| 0.0337 | 30.93 | 404 | 0.3204 | 0.9186 | |
|
| 0.0328 | 32.0 | 418 | 0.3176 | 0.9198 | |
|
| 0.0304 | 33.0 | 431 | 0.3183 | 0.9162 | |
|
| 0.0283 | 33.99 | 444 | 0.3150 | 0.9210 | |
|
| 0.029 | 34.99 | 457 | 0.3168 | 0.9174 | |
|
| 0.0264 | 35.98 | 470 | 0.3146 | 0.9174 | |
|
| 0.0259 | 36.98 | 483 | 0.3162 | 0.9174 | |
|
| 0.0258 | 37.97 | 496 | 0.3126 | 0.9186 | |
|
| 0.0251 | 38.97 | 509 | 0.3131 | 0.9174 | |
|
| 0.0239 | 39.96 | 522 | 0.3145 | 0.9186 | |
|
| 0.0234 | 40.96 | 535 | 0.3120 | 0.9198 | |
|
| 0.023 | 41.95 | 548 | 0.3102 | 0.9198 | |
|
| 0.0226 | 42.95 | 561 | 0.3123 | 0.9198 | |
|
| 0.0222 | 43.94 | 574 | 0.3140 | 0.9186 | |
|
| 0.0225 | 44.94 | 587 | 0.3119 | 0.9186 | |
|
| 0.0215 | 45.93 | 600 | 0.3106 | 0.9198 | |
|
| 0.0209 | 46.93 | 613 | 0.3113 | 0.9198 | |
|
| 0.0212 | 48.0 | 627 | 0.3115 | 0.9198 | |
|
| 0.021 | 49.0 | 640 | 0.3113 | 0.9198 | |
|
| 0.0212 | 49.76 | 650 | 0.3113 | 0.9198 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.34.1 |
|
- Pytorch 2.1.0+cu118 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.14.1 |
|
|
