vit-base-brain-mri

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

• Loss: 1.0577
• Accuracy: 0.5990

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: 0.0003
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	72	0.9986	0.6098
1.098	2.0	144	0.8445	0.7003
0.7895	3.0	216	0.7318	0.7526
0.7895	4.0	288	0.6842	0.7474
0.6629	5.0	360	0.6328	0.7857
0.5966	6.0	432	0.5957	0.8101
0.5546	7.0	504	0.5646	0.8118
0.5546	8.0	576	0.5647	0.8049
0.5113	9.0	648	0.5340	0.8275
0.4882	10.0	720	0.5190	0.8328
0.4882	11.0	792	0.5197	0.8328
0.4789	12.0	864	0.5002	0.8258
0.4582	13.0	936	0.4957	0.8310
0.4426	14.0	1008	0.4821	0.8310
0.4426	15.0	1080	0.4706	0.8467
0.4328	16.0	1152	0.4821	0.8153
0.432	17.0	1224	0.4992	0.8275
0.432	18.0	1296	0.4799	0.8345
0.4196	19.0	1368	0.4838	0.8310
0.4287	20.0	1440	0.4598	0.8659

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.0+cu121
• Tokenizers 0.19.1