hushem_1x_deit_tiny_rms_lr00001_fold5

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.8230
  • Accuracy: 0.7073

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: 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
No log 1.0 6 1.1737 0.4634
1.1816 2.0 12 0.8675 0.5366
1.1816 3.0 18 0.8079 0.6341
0.5246 4.0 24 0.8632 0.5854
0.2225 5.0 30 0.7815 0.5610
0.2225 6.0 36 0.6787 0.6585
0.0792 7.0 42 0.7052 0.6585
0.0792 8.0 48 0.7120 0.6341
0.029 9.0 54 0.8373 0.6585
0.0096 10.0 60 0.6713 0.7317
0.0096 11.0 66 0.7185 0.7073
0.0045 12.0 72 0.7237 0.6829
0.0045 13.0 78 0.7062 0.6829
0.0033 14.0 84 0.7203 0.7073
0.0025 15.0 90 0.7207 0.7073
0.0025 16.0 96 0.7400 0.7073
0.002 17.0 102 0.7337 0.6829
0.002 18.0 108 0.7527 0.6829
0.0017 19.0 114 0.7553 0.6829
0.0015 20.0 120 0.7631 0.6829
0.0015 21.0 126 0.7684 0.6829
0.0014 22.0 132 0.7730 0.6829
0.0014 23.0 138 0.7803 0.6829
0.0012 24.0 144 0.7869 0.6829
0.0011 25.0 150 0.7854 0.6829
0.0011 26.0 156 0.7958 0.6829
0.001 27.0 162 0.7899 0.6829
0.001 28.0 168 0.7956 0.6829
0.001 29.0 174 0.8038 0.6829
0.0009 30.0 180 0.8059 0.6829
0.0009 31.0 186 0.8121 0.6829
0.0008 32.0 192 0.8137 0.6829
0.0008 33.0 198 0.8161 0.6829
0.0008 34.0 204 0.8136 0.6829
0.0008 35.0 210 0.8158 0.6829
0.0008 36.0 216 0.8175 0.7073
0.0007 37.0 222 0.8190 0.7073
0.0007 38.0 228 0.8213 0.7073
0.0007 39.0 234 0.8222 0.7073
0.0007 40.0 240 0.8227 0.7073
0.0007 41.0 246 0.8228 0.7073
0.0007 42.0 252 0.8230 0.7073
0.0007 43.0 258 0.8230 0.7073
0.0007 44.0 264 0.8230 0.7073
0.0007 45.0 270 0.8230 0.7073
0.0007 46.0 276 0.8230 0.7073
0.0007 47.0 282 0.8230 0.7073
0.0007 48.0 288 0.8230 0.7073
0.0007 49.0 294 0.8230 0.7073
0.0007 50.0 300 0.8230 0.7073

Framework versions

  • Transformers 4.35.0
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1
Downloads last month
19
Safetensors
Model size
5.53M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for hkivancoral/hushem_1x_deit_tiny_rms_lr00001_fold5

Finetuned
(363)
this model

Evaluation results