swinv2-tiny-patch4-window8-256-dmae-humeda-DAV64

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: 0.2871
• Accuracy: 0.9371

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
1.1417	1.0	15	1.0312	0.5429
0.9163	2.0	30	0.7115	0.7943
0.7042	3.0	45	0.4227	0.8343
0.4976	4.0	60	0.3409	0.8686
0.4469	5.0	75	0.3059	0.8971
0.3746	6.0	90	0.2957	0.8914
0.3667	7.0	105	0.3036	0.8686
0.3026	8.0	120	0.2555	0.9086
0.2732	9.0	135	0.3009	0.9029
0.2518	10.0	150	0.2463	0.9029
0.2029	11.0	165	0.2345	0.9086
0.2698	12.0	180	0.2969	0.9029
0.2228	13.0	195	0.2966	0.9086
0.2213	14.0	210	0.4314	0.8514
0.1974	15.0	225	0.3352	0.8971
0.1865	16.0	240	0.2836	0.9086
0.157	17.0	255	0.2436	0.9086
0.1588	18.0	270	0.2618	0.88
0.1533	19.0	285	0.2315	0.92
0.1424	20.0	300	0.2358	0.9143
0.1278	21.0	315	0.2752	0.9086
0.1328	22.0	330	0.2196	0.92
0.124	23.0	345	0.2491	0.9143
0.1116	24.0	360	0.2085	0.92
0.1235	25.0	375	0.2379	0.92
0.1746	26.0	390	0.2749	0.9257
0.1143	27.0	405	0.2813	0.92
0.1256	28.0	420	0.3038	0.9086
0.0773	29.0	435	0.2903	0.9314
0.0981	30.0	450	0.2928	0.92
0.0717	31.0	465	0.3073	0.9086
0.1143	32.0	480	0.3466	0.8914
0.0821	33.0	495	0.3079	0.9257
0.0582	34.0	510	0.2845	0.9314
0.1015	35.0	525	0.3610	0.9029
0.075	36.0	540	0.2871	0.9371
0.0783	37.0	555	0.2788	0.9257
0.0919	38.0	570	0.2926	0.9257
0.0863	39.0	585	0.2911	0.9086
0.0721	40.0	600	0.3360	0.9143
0.0994	41.0	615	0.2865	0.9086
0.0734	42.0	630	0.3155	0.9143
0.0935	43.0	645	0.3129	0.9086
0.0651	44.0	660	0.2924	0.9143
0.0732	45.0	675	0.2961	0.9143

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1