swin-tiny-patch4-window7-224-finetuned-wuhan

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

• Loss: 0.7953
• Accuracy: 0.4

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• 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
No log	1.0	3	0.7953	0.4
No log	2.0	6	0.9477	0.4
No log	3.0	9	1.0106	0.4
0.5883	4.0	12	1.4170	0.4
0.5883	5.0	15	1.7436	0.4
0.5883	6.0	18	2.5380	0.4
0.241	7.0	21	3.8803	0.4
0.241	8.0	24	2.4040	0.2222
0.241	9.0	27	3.9968	0.4
0.125	10.0	30	3.2731	0.4
0.125	11.0	33	3.2202	0.2222
0.125	12.0	36	4.7008	0.4
0.125	13.0	39	4.5588	0.3556
0.0766	14.0	42	4.5434	0.2444
0.0766	15.0	45	4.9792	0.2667
0.0766	16.0	48	5.4095	0.2667
0.0239	17.0	51	5.8507	0.2222
0.0239	18.0	54	6.1023	0.2222
0.0239	19.0	57	6.1666	0.2222
0.0129	20.0	60	6.1948	0.2222

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.1
• Tokenizers 0.13.3