vit-base-patch16-224-in21k-finetuned-footulcer

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

• Loss: 0.0555
• Accuracy: 1.0

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.97	8	0.6026	0.7069
0.6438	1.94	16	0.5132	0.7328
0.4569	2.91	24	0.4402	0.7586
0.3098	4.0	33	0.2934	0.8448
0.2204	4.97	41	0.2969	0.8879
0.2204	5.94	49	0.1356	0.9655
0.1668	6.91	57	0.0659	0.9914
0.1531	8.0	66	0.0555	1.0
0.1096	8.97	74	0.0913	0.9741
0.112	9.94	82	0.0454	0.9914
0.1095	10.91	90	0.0463	0.9914
0.1095	12.0	99	0.0648	0.9914
0.0829	12.97	107	0.0427	0.9914
0.0741	13.94	115	0.0514	0.9914
0.0679	14.55	120	0.0548	0.9914

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2