swinv2-tiny-patch4-window8-256-dmae-humeda-DAV62

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: 0.2430
• 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: 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: 60
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	5	1.0856	0.3543
1.0609	2.0	10	1.0192	0.4571
1.0609	3.0	15	0.9256	0.5314
0.9026	4.0	20	0.7543	0.7029
0.9026	5.0	25	0.5856	0.7257
0.6347	6.0	30	0.3937	0.8629
0.6347	7.0	35	0.3700	0.8743
0.4754	8.0	40	0.3407	0.8629
0.4754	9.0	45	0.3551	0.8343
0.4252	10.0	50	0.2908	0.9086
0.4252	11.0	55	0.2852	0.8857
0.3628	12.0	60	0.2536	0.9029
0.3628	13.0	65	0.3290	0.88
0.3201	14.0	70	0.2594	0.9029
0.3201	15.0	75	0.2589	0.9029
0.2689	16.0	80	0.2430	0.92
0.2689	17.0	85	0.2531	0.9029
0.2668	18.0	90	0.2330	0.9143
0.2668	19.0	95	0.2579	0.8971
0.257	20.0	100	0.2669	0.88
0.257	21.0	105	0.2573	0.9086
0.2222	22.0	110	0.2863	0.8914
0.2222	23.0	115	0.2636	0.8971
0.1986	24.0	120	0.2873	0.8914
0.1986	25.0	125	0.2735	0.9143
0.1523	26.0	130	0.3072	0.9086
0.1523	27.0	135	0.2790	0.9143
0.1837	28.0	140	0.2813	0.9086
0.1837	29.0	145	0.2681	0.92
0.1628	30.0	150	0.2787	0.92
0.1628	31.0	155	0.2600	0.8971
0.1329	32.0	160	0.2849	0.9086
0.1329	33.0	165	0.3202	0.8914
0.155	34.0	170	0.2847	0.9086
0.155	35.0	175	0.2864	0.9029
0.1132	36.0	180	0.3315	0.8914
0.1132	37.0	185	0.2811	0.9029
0.1261	38.0	190	0.3471	0.8971
0.1261	39.0	195	0.3317	0.8971
0.1209	40.0	200	0.3337	0.8971
0.1209	41.0	205	0.3602	0.8914
0.0988	42.0	210	0.3385	0.8914
0.0988	43.0	215	0.4189	0.8971
0.0974	44.0	220	0.3559	0.9086
0.0974	45.0	225	0.3465	0.8971
0.1132	46.0	230	0.4038	0.8914
0.1132	47.0	235	0.3557	0.9029
0.0982	48.0	240	0.3471	0.9086
0.0982	49.0	245	0.3406	0.9143
0.0949	50.0	250	0.3637	0.9086
0.0949	51.0	255	0.3595	0.8971
0.0774	52.0	260	0.3669	0.8971
0.0774	53.0	265	0.3684	0.9029
0.0978	54.0	270	0.3868	0.8971
0.0978	55.0	275	0.3875	0.9029
0.0958	56.0	280	0.3529	0.9029
0.0958	57.0	285	0.3467	0.9143
0.08	58.0	290	0.3496	0.9086
0.08	59.0	295	0.3517	0.9086
0.0873	60.0	300	0.3534	0.9086

Framework versions

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