Edit model card

hushem_1x_deit_tiny_rms_00001_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: 1.1280
  • Accuracy: 0.6585

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 1.2888 0.4390
1.3565 2.0 12 1.0130 0.5366
1.3565 3.0 18 0.9361 0.5366
0.667 4.0 24 0.8831 0.6585
0.2929 5.0 30 0.8739 0.5854
0.2929 6.0 36 0.9329 0.5854
0.1055 7.0 42 0.9159 0.6585
0.1055 8.0 48 1.0700 0.5854
0.04 9.0 54 1.0357 0.5854
0.013 10.0 60 0.9379 0.6585
0.013 11.0 66 0.9964 0.6341
0.0046 12.0 72 1.0009 0.6585
0.0046 13.0 78 0.9889 0.6585
0.0029 14.0 84 1.0074 0.6585
0.0023 15.0 90 1.0258 0.6585
0.0023 16.0 96 1.0330 0.6585
0.0018 17.0 102 1.0391 0.6585
0.0018 18.0 108 1.0476 0.6585
0.0015 19.0 114 1.0552 0.6585
0.0013 20.0 120 1.0615 0.6585
0.0013 21.0 126 1.0642 0.6585
0.0011 22.0 132 1.0600 0.6585
0.0011 23.0 138 1.0791 0.6341
0.001 24.0 144 1.0890 0.6585
0.001 25.0 150 1.0948 0.6585
0.001 26.0 156 1.1067 0.6585
0.0008 27.0 162 1.0949 0.6585
0.0008 28.0 168 1.1017 0.6585
0.0008 29.0 174 1.1094 0.6585
0.0007 30.0 180 1.1105 0.6585
0.0007 31.0 186 1.1156 0.6585
0.0007 32.0 192 1.1158 0.6585
0.0007 33.0 198 1.1174 0.6585
0.0007 34.0 204 1.1167 0.6585
0.0006 35.0 210 1.1206 0.6585
0.0006 36.0 216 1.1224 0.6585
0.0006 37.0 222 1.1230 0.6585
0.0006 38.0 228 1.1253 0.6585
0.0006 39.0 234 1.1272 0.6585
0.0006 40.0 240 1.1276 0.6585
0.0006 41.0 246 1.1278 0.6585
0.0006 42.0 252 1.1280 0.6585
0.0006 43.0 258 1.1280 0.6585
0.0006 44.0 264 1.1280 0.6585
0.0006 45.0 270 1.1280 0.6585
0.0006 46.0 276 1.1280 0.6585
0.0006 47.0 282 1.1280 0.6585
0.0006 48.0 288 1.1280 0.6585
0.0006 49.0 294 1.1280 0.6585
0.0006 50.0 300 1.1280 0.6585

Framework versions

  • Transformers 4.35.0
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1
Downloads last month
11
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_00001_fold5

Finetuned
(363)
this model

Evaluation results