hushem_40x_deit_base_sgd_0001_fold2
This model is a fine-tuned version of facebook/deit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.2543
- Accuracy: 0.3333
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: 0.0001
- 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: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
1.3992 | 1.0 | 215 | 1.3998 | 0.2667 |
1.3515 | 2.0 | 430 | 1.3921 | 0.3111 |
1.3529 | 3.0 | 645 | 1.3851 | 0.3111 |
1.3241 | 4.0 | 860 | 1.3787 | 0.3111 |
1.3174 | 5.0 | 1075 | 1.3730 | 0.3333 |
1.331 | 6.0 | 1290 | 1.3675 | 0.3333 |
1.3242 | 7.0 | 1505 | 1.3623 | 0.3333 |
1.2565 | 8.0 | 1720 | 1.3574 | 0.3111 |
1.2719 | 9.0 | 1935 | 1.3525 | 0.3333 |
1.2647 | 10.0 | 2150 | 1.3477 | 0.3333 |
1.2293 | 11.0 | 2365 | 1.3430 | 0.3556 |
1.1981 | 12.0 | 2580 | 1.3386 | 0.3556 |
1.2258 | 13.0 | 2795 | 1.3342 | 0.3556 |
1.1901 | 14.0 | 3010 | 1.3299 | 0.3111 |
1.1988 | 15.0 | 3225 | 1.3257 | 0.3333 |
1.2048 | 16.0 | 3440 | 1.3214 | 0.3333 |
1.1848 | 17.0 | 3655 | 1.3173 | 0.3333 |
1.1799 | 18.0 | 3870 | 1.3134 | 0.3333 |
1.1629 | 19.0 | 4085 | 1.3095 | 0.3333 |
1.1482 | 20.0 | 4300 | 1.3058 | 0.3333 |
1.1478 | 21.0 | 4515 | 1.3022 | 0.3333 |
1.1472 | 22.0 | 4730 | 1.2988 | 0.3333 |
1.1143 | 23.0 | 4945 | 1.2955 | 0.3333 |
1.1122 | 24.0 | 5160 | 1.2923 | 0.3333 |
1.1 | 25.0 | 5375 | 1.2893 | 0.3111 |
1.1089 | 26.0 | 5590 | 1.2864 | 0.3111 |
1.1001 | 27.0 | 5805 | 1.2836 | 0.3333 |
1.0858 | 28.0 | 6020 | 1.2810 | 0.3333 |
1.093 | 29.0 | 6235 | 1.2786 | 0.3333 |
1.081 | 30.0 | 6450 | 1.2761 | 0.3333 |
1.0426 | 31.0 | 6665 | 1.2739 | 0.3333 |
1.0757 | 32.0 | 6880 | 1.2718 | 0.3333 |
1.0524 | 33.0 | 7095 | 1.2698 | 0.3333 |
1.0725 | 34.0 | 7310 | 1.2680 | 0.3333 |
1.055 | 35.0 | 7525 | 1.2662 | 0.3333 |
1.0308 | 36.0 | 7740 | 1.2646 | 0.3333 |
1.0978 | 37.0 | 7955 | 1.2631 | 0.3333 |
1.0264 | 38.0 | 8170 | 1.2617 | 0.3333 |
1.0508 | 39.0 | 8385 | 1.2604 | 0.3333 |
1.0208 | 40.0 | 8600 | 1.2593 | 0.3333 |
1.0505 | 41.0 | 8815 | 1.2583 | 0.3333 |
1.0147 | 42.0 | 9030 | 1.2574 | 0.3333 |
1.0355 | 43.0 | 9245 | 1.2566 | 0.3333 |
1.0064 | 44.0 | 9460 | 1.2559 | 0.3333 |
1.019 | 45.0 | 9675 | 1.2554 | 0.3333 |
1.0292 | 46.0 | 9890 | 1.2549 | 0.3333 |
1.0584 | 47.0 | 10105 | 1.2546 | 0.3333 |
1.0425 | 48.0 | 10320 | 1.2544 | 0.3333 |
1.0211 | 49.0 | 10535 | 1.2543 | 0.3333 |
1.0221 | 50.0 | 10750 | 1.2543 | 0.3333 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2
- Downloads last month
- 5
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for hkivancoral/hushem_40x_deit_base_sgd_0001_fold2
Base model
facebook/deit-base-patch16-224