Edit model card

Boya1_3Class_SGD_1-e3_20Epoch_Beit-large-224_fold1

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

  • Loss: 0.5544
  • Accuracy: 0.7654

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: 16
  • eval_batch_size: 16
  • 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: 20

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.9015 1.0 924 0.9341 0.6112
0.8409 2.0 1848 0.7990 0.6780
0.787 3.0 2772 0.7073 0.7116
0.638 4.0 3696 0.6605 0.7269
0.7348 5.0 4620 0.6355 0.7355
0.5951 6.0 5544 0.6177 0.7396
0.6333 7.0 6468 0.6036 0.7456
0.5506 8.0 7392 0.5938 0.7491
0.5249 9.0 8316 0.5879 0.7526
0.7137 10.0 9240 0.5788 0.7570
0.5106 11.0 10164 0.5775 0.7575
0.6297 12.0 11088 0.5685 0.7608
0.5637 13.0 12012 0.5647 0.7613
0.5879 14.0 12936 0.5617 0.7632
0.4454 15.0 13860 0.5588 0.7649
0.522 16.0 14784 0.5607 0.7630
0.551 17.0 15708 0.5549 0.7660
0.6177 18.0 16632 0.5560 0.7641
0.5656 19.0 17556 0.5549 0.7641
0.6413 20.0 18480 0.5544 0.7654

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.0.1
  • Datasets 2.12.0
  • Tokenizers 0.13.2
Downloads last month
0
Inference API
Drag image file here or click to browse from your device
This model can be loaded on Inference API (serverless).

Finetuned from

Evaluation results