swinv2-tiny-patch4-window8-256-dmae-humeda-DAV76

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: 0.4077
• Accuracy: 0.88

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: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1017	0.9524	15	1.0955	0.3143
0.9212	1.9524	30	0.8475	0.6743
0.7117	2.9524	45	0.6980	0.6171
0.5496	3.9524	60	0.4957	0.8
0.5051	4.9524	75	0.4578	0.7714
0.4331	5.9524	90	0.3767	0.8457
0.4324	6.9524	105	0.4334	0.8229
0.3664	7.9524	120	0.4469	0.7829
0.335	8.9524	135	0.3407	0.8743
0.2977	9.9524	150	0.3569	0.84
0.2978	10.9524	165	0.3858	0.8686
0.2983	11.9524	180	0.3657	0.8571
0.2539	12.9524	195	0.3979	0.8514
0.2215	13.9524	210	0.3755	0.8514
0.2474	14.9524	225	0.4143	0.8457
0.2245	15.9524	240	0.3954	0.8629
0.2427	16.9524	255	0.4063	0.8743
0.2036	17.9524	270	0.4762	0.8343
0.2397	18.9524	285	0.4077	0.88
0.2157	19.9524	300	0.5519	0.8114
0.221	20.9524	315	0.5091	0.8114
0.1799	21.9524	330	0.4301	0.8629
0.1777	22.9524	345	0.4592	0.8743
0.1641	23.9524	360	0.4445	0.8686
0.1582	24.9524	375	0.4807	0.8571
0.1394	25.9524	390	0.4472	0.8743
0.16	26.9524	405	0.5020	0.8743
0.1826	27.9524	420	0.4834	0.8686
0.1648	28.9524	435	0.5368	0.8629
0.155	29.9524	450	0.5284	0.8514
0.1378	30.9524	465	0.4585	0.8743
0.1608	31.9524	480	0.4883	0.8686
0.1435	32.9524	495	0.5400	0.84
0.1444	33.9524	510	0.5379	0.8571
0.1504	34.9524	525	0.5876	0.8629
0.1108	35.9524	540	0.5414	0.8571
0.1392	36.9524	555	0.5801	0.8571
0.1065	37.9524	570	0.5940	0.8629
0.087	38.9524	585	0.6316	0.8571
0.127	39.9524	600	0.6509	0.8571
0.1198	40.9524	615	0.6311	0.8571
0.1255	41.9524	630	0.5793	0.8514
0.1317	42.9524	645	0.5860	0.8343
0.1016	43.9524	660	0.5839	0.8629
0.1249	44.9524	675	0.5763	0.8571
0.0762	45.9524	690	0.5853	0.8629
0.1075	46.9524	705	0.5967	0.8514
0.0792	47.9524	720	0.6012	0.8457
0.1033	48.9524	735	0.5989	0.8457
0.1115	49.9524	750	0.6030	0.8457

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 2.19.0
• Tokenizers 0.21.1