Wound-classification

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1836
• Accuracy: 0.9575

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1241	1.0	200	0.7452	0.765
0.5854	2.0	400	0.4880	0.835
0.4279	3.0	600	0.5049	0.8375
0.4041	4.0	800	0.3321	0.8975
0.2805	5.0	1000	0.4105	0.895
0.279	6.0	1200	0.4269	0.8825
0.1782	7.0	1400	0.3583	0.905
0.1834	8.0	1600	0.3009	0.925
0.1197	9.0	1800	0.3020	0.93
0.1231	10.0	2000	0.3352	0.9225
0.1273	11.0	2200	0.2908	0.91
0.1019	12.0	2400	0.2528	0.94
0.0951	13.0	2600	0.2989	0.9325
0.0957	14.0	2800	0.3189	0.9325
0.0618	15.0	3000	0.1973	0.9475
0.0583	16.0	3200	0.1836	0.9575

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2