vit-base-patch16-224-finetuned-ISIC-dec2024gray

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

• Loss: 0.1549
• Accuracy: 0.9369

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7609	1.0	974	0.1860	0.9231
0.5981	2.0	1948	0.1640	0.9320
0.5217	2.9974	2919	0.1549	0.9369

Testing results

• True positive: 1341

• False positive: 290

• True negative: 14923

• False negative: 750

• Precision: 0.822

• Recall: 0.641

• Accuracy: 0.940

• F1: 0.721

Framework versions

• Transformers 4.47.1
• Pytorch 2.6.0.dev20241225+cu126
• Datasets 3.2.0
• Tokenizers 0.21.0