Boya2_SGD_1e3_20Epoch_Swin-large-224_fold1

This model is a fine-tuned version of microsoft/swin-large-patch4-window12-384 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.6332
• Accuracy: 0.4716

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: 0.001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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
2.4897	1.0	914	2.4409	0.2804
2.1506	2.0	1828	2.2233	0.2968
2.0176	3.0	2742	2.1001	0.3311
1.9799	4.0	3656	2.0112	0.3564
1.9403	5.0	4570	1.9459	0.3786
1.9907	6.0	5484	1.8909	0.4003
1.7985	7.0	6398	1.8449	0.4159
1.8712	8.0	7312	1.8057	0.4239
1.7195	9.0	8226	1.7733	0.4348
1.8526	10.0	9140	1.7458	0.4436
1.67	11.0	10054	1.7203	0.4488
1.6061	12.0	10968	1.7023	0.4549
1.6256	13.0	11882	1.6832	0.4582
1.8212	14.0	12796	1.6685	0.4634
1.7157	15.0	13710	1.6584	0.4639
1.6148	16.0	14624	1.6491	0.4661
1.7158	17.0	15538	1.6424	0.4675
1.7391	18.0	16452	1.6370	0.4689
1.8077	19.0	17366	1.6337	0.4716
1.7769	20.0	18280	1.6332	0.4716

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.21.0
• Tokenizers 0.13.2