swin-base-patch4-window7-224-in22k-finetuned_swinv1-autotags-224

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224-in22k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1186
• Accuracy: 0.9675

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1468	0.99	61	0.8540	0.7503
0.6167	1.99	122	0.3772	0.8904
0.45	2.99	183	0.2963	0.9086
0.2163	3.99	244	0.2172	0.9391
0.209	4.99	305	0.1733	0.9431
0.1558	5.99	366	0.2101	0.9310
0.109	6.99	427	0.1268	0.9655
0.1214	7.99	488	0.1251	0.9706
0.1471	8.99	549	0.1194	0.9665
0.0888	9.99	610	0.1376	0.9574
0.1077	10.99	671	0.1211	0.9614
0.0969	11.99	732	0.1231	0.9695
0.0585	12.99	793	0.1472	0.9553
0.0659	13.99	854	0.1203	0.9655
0.0645	14.99	915	0.1405	0.9614
0.0472	15.99	976	0.1340	0.9604
0.0616	16.99	1037	0.1272	0.9655
0.0609	17.99	1098	0.1121	0.9685
0.0525	18.99	1159	0.1162	0.9685
0.0406	19.99	1220	0.1186	0.9675

Framework versions

• Transformers 4.25.1
• Pytorch 1.10.2+cu113
• Datasets 2.10.1
• Tokenizers 0.13.2