deit-small-patch16-224-finetuned-piid

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

• Loss: 0.5615
• Accuracy: 0.7945

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1803	0.98	20	1.0233	0.5753
0.706	2.0	41	0.7299	0.7078
0.6016	2.98	61	0.6877	0.7123
0.4903	4.0	82	0.6139	0.7671
0.4692	4.98	102	0.5667	0.7626
0.374	6.0	123	0.5146	0.8037
0.2995	6.98	143	0.5596	0.7534
0.2905	8.0	164	0.5313	0.7534
0.2612	8.98	184	0.5328	0.7900
0.2499	10.0	205	0.5369	0.7991
0.185	10.98	225	0.5754	0.7808
0.1927	12.0	246	0.5886	0.7717
0.1446	12.98	266	0.5160	0.7991
0.155	14.0	287	0.5353	0.8082
0.1577	14.98	307	0.5848	0.7808
0.1243	16.0	328	0.5572	0.7991
0.1038	16.98	348	0.5859	0.7763
0.1305	18.0	369	0.5752	0.7900
0.0868	18.98	389	0.5616	0.8037
0.1364	19.51	400	0.5615	0.7945

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1