dinov2-small-finetuned-papsmear

This model is a fine-tuned version of facebook/dinov2-small on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3843
• Accuracy: 0.8603

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• 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
0.846	0.9935	38	1.0217	0.5956
1.0241	1.9869	76	0.8413	0.6544
0.9178	2.9804	114	0.7204	0.7426
0.693	4.0	153	0.5731	0.75
0.7157	4.9935	191	0.5501	0.8162
0.5006	5.9869	229	0.6096	0.7794
0.4576	6.9804	267	0.5535	0.7941
0.467	8.0	306	0.5041	0.8162
0.4378	8.9935	344	0.5771	0.8015
0.2876	9.9869	382	0.4234	0.8456
0.2308	10.9804	420	0.4946	0.8382
0.2312	12.0	459	0.5098	0.8309
0.1625	12.9935	497	0.3813	0.8603
0.1775	13.9869	535	0.3695	0.8529
0.1358	14.9020	570	0.3843	0.8603

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1