swinv2-tiny-patch4-window8-256-dmae-humeda-DAV34

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: 1.3276
• Accuracy: 0.6333

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.8889	6	1.4254	0.45
5.9772	1.8889	12	1.3083	0.45
5.9772	2.8889	18	1.3032	0.4667
5.1141	3.8889	24	1.1521	0.5667
5.1141	4.8889	30	1.2171	0.4667
3.8959	5.8889	36	1.1281	0.4833
3.8959	6.8889	42	0.9969	0.6167
2.9663	7.8889	48	0.9814	0.65
2.9663	8.8889	54	0.8797	0.6333
2.416	9.8889	60	0.8956	0.6667
2.416	10.8889	66	0.8364	0.7
1.9336	11.8889	72	0.9800	0.65
1.9336	12.8889	78	0.8707	0.6833
1.5631	13.8889	84	0.9331	0.6333
1.5631	14.8889	90	1.0113	0.6333
1.1295	15.8889	96	1.0988	0.6167
1.1295	16.8889	102	1.0197	0.6833
1.1454	17.8889	108	1.2170	0.6167
1.1454	18.8889	114	1.1182	0.6667
0.8852	19.8889	120	1.1026	0.6833
0.8852	20.8889	126	1.0868	0.6333
0.7881	21.8889	132	1.1674	0.6333
0.7881	22.8889	138	1.1763	0.6667
0.7913	23.8889	144	1.3433	0.6333
0.7913	24.8889	150	1.1228	0.7
0.667	25.8889	156	1.2828	0.6667
0.667	26.8889	162	1.2373	0.6833
0.6299	27.8889	168	1.2951	0.6667
0.6299	28.8889	174	1.3410	0.6333
0.5409	29.8889	180	1.1852	0.7
0.5409	30.8889	186	1.4286	0.6167
0.6085	31.8889	192	1.2376	0.65
0.6085	32.8889	198	1.2249	0.6667
0.562	33.8889	204	1.3640	0.6333
0.562	34.8889	210	1.4234	0.6333
0.4543	35.8889	216	1.3489	0.6333
0.4543	36.8889	222	1.3273	0.6333
0.4708	37.8889	228	1.3215	0.6333
0.4708	38.8889	234	1.3277	0.6333
0.5217	39.8889	240	1.3276	0.6333

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0