convnext-small-224-finetuned-piid

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

• Loss: 0.5651
• Accuracy: 0.7626

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
1.3405	0.98	20	1.3201	0.4155
1.1715	2.0	41	1.1362	0.5708
0.9231	2.98	61	0.9255	0.6438
0.7128	4.0	82	0.7558	0.6986
0.6204	4.98	102	0.7056	0.7534
0.5322	6.0	123	0.6610	0.7397
0.4403	6.98	143	0.6639	0.7443
0.4388	8.0	164	0.6472	0.7306
0.3901	8.98	184	0.6684	0.7352
0.4202	10.0	205	0.5934	0.7397
0.3784	10.98	225	0.5651	0.7626
0.2973	12.0	246	0.6439	0.7580
0.3614	12.98	266	0.5844	0.7534
0.2795	14.0	287	0.6015	0.7306
0.2825	14.98	307	0.6031	0.7626
0.2364	16.0	328	0.6249	0.7534
0.2162	16.98	348	0.6248	0.7626
0.2455	18.0	369	0.6153	0.7489
0.2314	18.98	389	0.6113	0.7580
0.248	19.51	400	0.6131	0.7580

Framework versions

• Transformers 4.33.3
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3