swinv2-tiny-patch4-window8-256-finalterm

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2805
• Accuracy: 0.9

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3578	1.0	10	1.2444	0.475
1.1054	2.0	20	0.9180	0.6531
0.8485	3.0	30	0.6632	0.725
0.674	4.0	40	0.4736	0.7969
0.5968	5.0	50	0.4341	0.8125
0.508	6.0	60	0.5391	0.8187
0.4852	7.0	70	0.3906	0.8344
0.4354	8.0	80	0.3257	0.8656
0.4165	9.0	90	0.3478	0.8656
0.4385	10.0	100	0.3114	0.8781
0.4156	11.0	110	0.3461	0.8781
0.4055	12.0	120	0.3108	0.8844
0.4282	13.0	130	0.2916	0.8875
0.3546	14.0	140	0.2972	0.9
0.3608	15.0	150	0.3428	0.8688
0.369	16.0	160	0.2885	0.8969
0.3525	17.0	170	0.2861	0.9
0.338	18.0	180	0.2832	0.9062
0.3633	19.0	190	0.2797	0.9031
0.3712	20.0	200	0.2805	0.9

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1