Edit model card

smids_5x_deit_base_rms_001_fold5

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: 0.5070
  • Accuracy: 0.8067

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: 0.001
  • 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
1.0986 1.0 375 1.0985 0.3517
0.9277 2.0 750 0.8963 0.5367
0.8868 3.0 1125 0.8248 0.5567
0.8432 4.0 1500 0.8168 0.5367
0.7756 5.0 1875 0.8234 0.555
0.7692 6.0 2250 0.7291 0.6617
0.7096 7.0 2625 0.7367 0.6533
0.8553 8.0 3000 0.7548 0.63
0.7339 9.0 3375 0.7547 0.6233
0.6755 10.0 3750 0.7150 0.665
0.7434 11.0 4125 0.7034 0.685
0.6966 12.0 4500 0.6998 0.6833
0.6638 13.0 4875 0.8188 0.615
0.7005 14.0 5250 0.6380 0.7233
0.7307 15.0 5625 0.6467 0.7017
0.6252 16.0 6000 0.6189 0.7317
0.6235 17.0 6375 0.5966 0.7267
0.6067 18.0 6750 0.5889 0.7367
0.6586 19.0 7125 0.5888 0.745
0.553 20.0 7500 0.5461 0.7583
0.5457 21.0 7875 0.5458 0.7717
0.535 22.0 8250 0.5661 0.745
0.5802 23.0 8625 0.5673 0.7633
0.585 24.0 9000 0.5456 0.7767
0.5034 25.0 9375 0.5600 0.7517
0.519 26.0 9750 0.5101 0.7767
0.578 27.0 10125 0.5562 0.7517
0.5681 28.0 10500 0.5592 0.7633
0.5613 29.0 10875 0.5207 0.7733
0.4923 30.0 11250 0.5540 0.7683
0.4514 31.0 11625 0.5170 0.795
0.4948 32.0 12000 0.5569 0.775
0.4729 33.0 12375 0.5006 0.7967
0.4583 34.0 12750 0.5008 0.7917
0.4376 35.0 13125 0.4986 0.815
0.3894 36.0 13500 0.5048 0.8033
0.4227 37.0 13875 0.5449 0.7883
0.4237 38.0 14250 0.4850 0.81
0.3609 39.0 14625 0.4881 0.8017
0.4451 40.0 15000 0.5131 0.8067
0.411 41.0 15375 0.5305 0.7983
0.4629 42.0 15750 0.4959 0.8
0.4034 43.0 16125 0.5125 0.8083
0.3681 44.0 16500 0.5034 0.8033
0.4332 45.0 16875 0.4946 0.8017
0.3808 46.0 17250 0.4987 0.8067
0.3828 47.0 17625 0.5113 0.8183
0.2902 48.0 18000 0.5081 0.8
0.3255 49.0 18375 0.5035 0.8083
0.3922 50.0 18750 0.5070 0.8067

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.1.0+cu121
  • Datasets 2.12.0
  • Tokenizers 0.13.2
Downloads last month
14

Finetuned from

Evaluation results