0.50-200Train-100Test-swinv2-base

This model is a fine-tuned version of microsoft/swinv2-base-patch4-window12-192-22k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1459
• Accuracy: 0.8183

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: 17

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7214	0.9931	36	1.0786	0.6515
0.6184	1.9862	72	0.7491	0.7651
0.357	2.9793	108	0.7632	0.7764
0.2085	4.0	145	0.8125	0.7860
0.1343	4.9931	181	0.7920	0.7974
0.0641	5.9862	217	0.8851	0.7860
0.0515	6.9793	253	1.0784	0.7817
0.041	8.0	290	1.0600	0.7965
0.0338	8.9931	326	1.0860	0.8131
0.013	9.9862	362	1.0956	0.8148
0.016	10.9793	398	1.2115	0.7991
0.0154	12.0	435	1.1470	0.8105
0.011	12.9931	471	1.1045	0.8105
0.0027	13.9862	507	1.1310	0.8096
0.0042	14.9793	543	1.1808	0.8227
0.0016	16.0	580	1.1575	0.8157
0.0007	16.8828	612	1.1459	0.8183

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1