deit-tiny-patch16-224-finetuned-papsmear

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3418
• Accuracy: 0.9118

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: 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
0.1493	0.9870	19	0.3965	0.8603
0.181	1.9740	38	0.5619	0.8235
0.2222	2.9610	57	0.4637	0.875
0.1273	4.0	77	0.4504	0.8824
0.2217	4.9870	96	0.4186	0.875
0.1678	5.9740	115	0.3936	0.8676
0.1302	6.9610	134	0.3591	0.9044
0.1494	8.0	154	0.3726	0.9118
0.1235	8.9870	173	0.3986	0.8897
0.1648	9.9740	192	0.3476	0.9265
0.1062	10.9610	211	0.3309	0.8971
0.1098	12.0	231	0.3293	0.9265
0.0684	12.9870	250	0.3472	0.9044
0.0907	13.9740	269	0.3483	0.9044
0.0903	14.8052	285	0.3418	0.9118

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1