swinv2-tiny-patch4-window8-256-dmae-humeda-DAV67

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: 0.2829
• Accuracy: 0.92

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: 4e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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: 45
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.103	1.0	18	1.0588	0.4
0.9007	2.0	36	0.5651	0.8229
0.497	3.0	54	0.3559	0.8743
0.4178	4.0	72	0.4171	0.7886
0.4303	5.0	90	0.6884	0.7314
0.485	6.0	108	0.3255	0.8629
0.3345	7.0	126	0.2631	0.9029
0.3058	8.0	144	0.3533	0.8343
0.35	9.0	162	0.2853	0.8686
0.2535	10.0	180	0.2529	0.9143
0.21	11.0	198	0.3806	0.84
0.2414	12.0	216	0.2829	0.92
0.1978	13.0	234	0.3011	0.9143
0.1683	14.0	252	0.2486	0.9086
0.2351	15.0	270	0.3612	0.8629
0.264	16.0	288	0.3643	0.88
0.1714	17.0	306	0.2481	0.9086
0.1714	18.0	324	0.3479	0.8914
0.1886	19.0	342	0.2644	0.9029
0.1522	20.0	360	0.2587	0.8971
0.1468	21.0	378	0.2832	0.9029
0.1364	22.0	396	0.2830	0.9029
0.1294	23.0	414	0.2954	0.8914
0.122	24.0	432	0.3801	0.8743
0.1114	25.0	450	0.3375	0.9029
0.12	26.0	468	0.3696	0.8743
0.121	27.0	486	0.3460	0.8857
0.1056	28.0	504	0.3296	0.9029
0.0941	29.0	522	0.3977	0.8971
0.0882	30.0	540	0.3436	0.8914
0.1131	31.0	558	0.3397	0.8914
0.0959	32.0	576	0.3420	0.8971
0.1058	33.0	594	0.3312	0.8971
0.0707	34.0	612	0.3917	0.8857
0.0885	35.0	630	0.3855	0.88
0.0788	36.0	648	0.3519	0.8857
0.112	37.0	666	0.3473	0.8914
0.0588	38.0	684	0.3718	0.9029
0.0963	39.0	702	0.4022	0.88
0.0681	40.0	720	0.3574	0.8971
0.0841	41.0	738	0.3621	0.88
0.0739	42.0	756	0.3782	0.8914
0.0649	42.5217	765	0.3766	0.8914

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1