swin-tiny-patch4-window7-224-finetuned-IDRiD

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

• Loss: 0.8534
• Accuracy: 0.7143

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
No log	1.0	3	1.5419	0.3810
No log	2.0	6	1.3336	0.4048
No log	3.0	9	1.2311	0.5238
1.4831	4.0	12	1.1594	0.5238
1.4831	5.0	15	1.0889	0.5
1.4831	6.0	18	1.0743	0.5
1.1593	7.0	21	1.0466	0.5476
1.1593	8.0	24	1.0160	0.5714
1.1593	9.0	27	0.9660	0.6190
1.008	10.0	30	0.9390	0.6190
1.008	11.0	33	0.9158	0.5952
1.008	12.0	36	0.8734	0.6190
1.008	13.0	39	0.8715	0.6667
0.8963	14.0	42	0.8261	0.6667
0.8963	15.0	45	0.8683	0.6190
0.8963	16.0	48	0.8772	0.6429
0.8404	17.0	51	0.8729	0.6905
0.8404	18.0	54	0.8571	0.7143
0.8404	19.0	57	0.8546	0.7143
0.8263	20.0	60	0.8534	0.7143

Framework versions

• Transformers 4.30.0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.13.3