vit-bmr-augmented

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: 0.5922
• Accuracy: 0.6667
• Precision: 0.6663
• Recall: 0.6667
• F1: 0.6664

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: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	160	0.6951	0.4691	0.2201	0.4691	0.2996
No log	2.0	320	0.6841	0.5309	0.2818	0.5309	0.3682
No log	3.0	480	0.6859	0.5802	0.6735	0.5802	0.5354
0.7037	4.0	640	0.6921	0.5309	0.2818	0.5309	0.3682
0.7037	5.0	800	0.6910	0.5309	0.2818	0.5309	0.3682
0.7037	6.0	960	0.6660	0.7037	0.7180	0.7037	0.6941
0.6949	7.0	1120	0.6931	0.4691	0.2201	0.4691	0.2996
0.6949	8.0	1280	0.7131	0.4691	0.2201	0.4691	0.2996
0.6949	9.0	1440	0.6832	0.5432	0.7545	0.5432	0.3952
0.69	10.0	1600	0.5922	0.6667	0.6663	0.6667	0.6664
0.69	11.0	1760	0.7749	0.6543	0.7011	0.6543	0.6417
0.69	12.0	1920	0.6127	0.6914	0.6919	0.6914	0.6915
0.5014	13.0	2080	0.8927	0.6914	0.7115	0.6914	0.6882
0.5014	14.0	2240	0.8250	0.7284	0.7298	0.7284	0.7286
0.5014	15.0	2400	0.8753	0.7284	0.7298	0.7284	0.7286

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1