swin-tiny-patch4-window7-224-finetuned-cp1

This model is a fine-tuned version of Yogesh1p/swin-tiny-patch4-window7-224-finetuned-cp1 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3628
• Accuracy: 0.8667

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.92	3	0.4305	0.8222
No log	1.85	6	0.4510	0.8
No log	2.77	9	0.4329	0.7778
0.3489	4.0	13	0.4965	0.7778
0.3489	4.92	16	0.4247	0.8
0.3489	5.85	19	0.3559	0.8444
0.2872	6.77	22	0.3628	0.8667
0.2872	8.0	26	0.3511	0.8667
0.2872	8.92	29	0.3538	0.8667
0.3071	9.85	32	0.3644	0.8222
0.3071	10.77	35	0.3860	0.8222
0.3071	12.0	39	0.5917	0.8
0.2765	12.92	42	0.5149	0.7778
0.2765	13.85	45	0.5605	0.7556
0.2765	14.77	48	0.4737	0.8
0.2637	16.0	52	0.4109	0.8444
0.2637	16.92	55	0.3584	0.8222
0.2637	17.85	58	0.3771	0.8222
0.2082	18.77	61	0.4406	0.8222
0.2082	20.0	65	0.3592	0.8222
0.2082	20.92	68	0.3302	0.8667
0.1944	21.85	71	0.3206	0.8444
0.1944	22.77	74	0.3215	0.8667
0.1944	24.0	78	0.3209	0.8667
0.2144	24.92	81	0.3327	0.8444
0.2144	25.85	84	0.3506	0.8444
0.2144	26.77	87	0.3545	0.8444
0.1754	27.69	90	0.3494	0.8444

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0