swinv2-tiny-patch4-window8-256-dmae-va-U-40

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1175
• Accuracy: 0.9817

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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9	7	1.3839	0.3028
1.4343	1.94	15	1.3643	0.2844
1.3413	2.97	23	1.3310	0.3578
1.2473	4.0	31	1.1081	0.5138
1.2473	4.9	38	0.8292	0.7064
1.0532	5.94	46	0.7420	0.6239
0.917	6.97	54	0.6345	0.6972
0.7939	8.0	62	0.4898	0.8532
0.7939	8.9	69	0.5918	0.7523
0.7457	9.94	77	0.5271	0.7615
0.6834	10.97	85	0.3296	0.9450
0.5847	12.0	93	0.2883	0.9174
0.5199	12.9	100	0.2896	0.9266
0.5199	13.94	108	0.2859	0.8991
0.4657	14.97	116	0.2515	0.9083
0.4585	16.0	124	0.2261	0.9083
0.3892	16.9	131	0.2142	0.9266
0.3892	17.94	139	0.1788	0.9450
0.3939	18.97	147	0.1948	0.9266
0.3429	20.0	155	0.1685	0.9450
0.3493	20.9	162	0.1986	0.9083
0.3462	21.94	170	0.1540	0.9358
0.3462	22.97	178	0.1449	0.9450
0.3117	24.0	186	0.1379	0.9541
0.3109	24.9	193	0.1423	0.9450
0.2867	25.94	201	0.1451	0.9450
0.2867	26.97	209	0.1154	0.9725
0.293	28.0	217	0.1152	0.9541
0.2782	28.9	224	0.1261	0.9633
0.2744	29.94	232	0.1175	0.9817
0.2711	30.97	240	0.1292	0.9633
0.2711	32.0	248	0.1101	0.9817
0.2652	32.9	255	0.1202	0.9633
0.2218	33.94	263	0.1119	0.9817
0.2899	34.97	271	0.1071	0.9817
0.2899	36.0	279	0.1077	0.9817
0.2143	36.13	280	0.1077	0.9817

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1