finetuned-breast_cancer_images

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:

• Train Accuracy: 0.9025
• Train Loss: 0.0125
• Accuracy: 0.9589
• Loss: 0.0175

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 103

Training results

Training Loss	Epoch	Step	Accuracy	Loss	Validation Loss
No log	5.06	200	0.9114	0.0241	0.0172
No log	5.06	200	0.9114	0.0241	0.0172
No log	10.13	400	0.9335	0.0209	0.0157
No log	10.13	400	0.9335	0.0209	0.0157
No log	15.19	600	0.9241	0.0189	0.0172
No log	15.19	600	0.9241	0.0189	0.0172
No log	20.25	800	0.9430	0.0182	0.0161
No log	20.25	800	0.9430	0.0182	0.0161
No log	25.32	1000	0.9399	0.0165	0.0186
No log	25.32	1000	0.9399	0.0165	0.0186
No log	30.38	1200	0.9557	0.0155	0.0137
No log	30.38	1200	0.9557	0.0155	0.0137
No log	35.44	1400	0.9335	0.0136	0.0180
No log	35.44	1400	0.9335	0.0136	0.0180
No log	40.51	1600	0.9525	0.0152	0.0139
No log	40.51	1600	0.9525	0.0152	0.0139
No log	45.57	1800	0.9494	0.0144	0.0149
No log	45.57	1800	0.9494	0.0144	0.0149
No log	50.63	2000	0.9430	0.0153	0.0187
No log	50.63	2000	0.9430	0.0153	0.0187
No log	55.7	2200	0.9557	0.0132	0.0149
No log	55.7	2200	0.9557	0.0132	0.0149
No log	60.76	2400	0.9430	0.0124	0.0149
No log	60.76	2400	0.9430	0.0124	0.0149
No log	65.82	2600	0.9525	0.0134	0.0164
No log	65.82	2600	0.9525	0.0134	0.0164
No log	70.89	2800	0.9557	0.0117	0.0143
No log	70.89	2800	0.9557	0.0117	0.0143
No log	75.95	3000	0.9557	0.0112	0.0166
No log	75.95	3000	0.9557	0.0112	0.0166
No log	81.01	3200	0.9589	0.0118	0.0163
No log	81.01	3200	0.9589	0.0118	0.0163
No log	86.08	3400	0.9494	0.0106	0.0188
No log	86.08	3400	0.9494	0.0106	0.0188
No log	91.14	3600	0.9462	0.0121	0.0186
No log	91.14	3600	0.9462	0.0121	0.0186
No log	96.2	3800	0.9525	0.0129	0.0171
No log	96.2	3800	0.9525	0.0129	0.0171
No log	101.27	4000	0.9589	0.0125	0.0175
No log	101.27	4000	0.9589	0.0125	0.0175

Framework versions

• Transformers 4.30.0
• Pytorch 2.3.0+cu121
• Tokenizers 0.13.3