tifosiai/vri-alzheirmer-classifier

alz-mri-vit

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: 0.2447
• F1: 0.9086

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1
1.1101	1.0	64	0.9590	0.5549
0.9356	2.0	128	0.8877	0.5940
0.9187	3.0	192	0.9376	0.5272
0.8804	4.0	256	0.8667	0.5962
0.7854	5.0	320	0.7756	0.6720
0.7278	6.0	384	0.7202	0.6860
0.6462	7.0	448	0.7124	0.6898
0.5731	8.0	512	0.6027	0.7553
0.473	9.0	576	0.5520	0.7724
0.4378	10.0	640	0.5550	0.7758
0.4086	11.0	704	0.4366	0.8271
0.36	12.0	768	0.4446	0.8225
0.3217	13.0	832	0.3841	0.8441
0.2941	14.0	896	0.4719	0.8182
0.2679	15.0	960	0.4112	0.8410
0.2565	16.0	1024	0.3698	0.8527
0.2502	17.0	1088	0.3283	0.8810
0.2166	18.0	1152	0.3569	0.8627
0.197	19.0	1216	0.3475	0.8699
0.2004	20.0	1280	0.3171	0.8834
0.1722	21.0	1344	0.2711	0.8998
0.1529	22.0	1408	0.2432	0.9100
0.1495	23.0	1472	0.2950	0.8978
0.1307	24.0	1536	0.2811	0.9034
0.1278	25.0	1600	0.2545	0.9086
0.1175	26.0	1664	0.2561	0.9051
0.1264	27.0	1728	0.2128	0.9186
0.1015	28.0	1792	0.3022	0.9014
0.1077	29.0	1856	0.2403	0.9221
0.0932	30.0	1920	0.2447	0.9086

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1