resnet-fine_tuned

This model is a fine-tuned version of microsoft/resnet-34 on the Falah/Alzheimer_MRI dataset. It achieves the following results on the evaluation set:

• Loss: 0.1983
• Accuracy: 0.9219

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9041	1.0	80	0.9659	0.5352
0.8743	2.0	160	0.9348	0.5797
0.7723	3.0	240	0.7793	0.6594
0.6864	4.0	320	0.6799	0.7031
0.5347	5.0	400	0.5596	0.7703
0.4282	6.0	480	0.5078	0.7766
0.4315	7.0	560	0.5455	0.7680
0.3747	8.0	640	0.4203	0.8266
0.2977	9.0	720	0.3926	0.8469
0.2252	10.0	800	0.3024	0.8742
0.2675	11.0	880	0.2731	0.8906
0.2136	12.0	960	0.3045	0.875
0.1998	13.0	1040	0.2370	0.9
0.2406	14.0	1120	0.2387	0.9086
0.1873	15.0	1200	0.1983	0.9219

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cpu
• Datasets 2.13.1
• Tokenizers 0.13.3