swinv2-tiny-patch4-window8-256-dmae-humeda-DA-V2-c

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: 0.9897
• Accuracy: 0.7308

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use 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: 42

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.8421	4	1.5997	0.2308
No log	1.9474	9	1.5322	0.25
1.5784	2.8421	13	1.4223	0.4808
1.5784	3.9474	18	1.2367	0.5192
1.5784	4.8421	22	1.1217	0.5577
1.3137	5.9474	27	0.9917	0.5192
1.3137	6.8421	31	1.0075	0.5577
0.9943	7.9474	36	0.9245	0.5769
0.9943	8.8421	40	0.8960	0.6154
0.9943	9.9474	45	0.8786	0.5769
0.8055	10.8421	49	0.9115	0.5962
0.8055	11.9474	54	0.9065	0.6538
0.8055	12.8421	58	0.9131	0.5962
0.7128	13.9474	63	0.8910	0.6538
0.7128	14.8421	67	0.9413	0.6538
0.6511	15.9474	72	0.9045	0.7115
0.6511	16.8421	76	0.8853	0.7115
0.6511	17.9474	81	0.9149	0.7115
0.5837	18.8421	85	0.9372	0.7308
0.5837	19.9474	90	0.8941	0.7308
0.5837	20.8421	94	0.8905	0.75
0.5269	21.9474	99	0.8929	0.7308
0.5269	22.8421	103	0.9018	0.7308
0.4738	23.9474	108	0.9197	0.7308
0.4738	24.8421	112	0.9889	0.7308
0.4738	25.9474	117	0.9281	0.7308
0.4632	26.8421	121	0.9315	0.7308
0.4632	27.9474	126	1.0017	0.7115
0.4632	28.8421	130	0.9667	0.7115
0.4494	29.9474	135	0.9357	0.7308
0.4494	30.8421	139	0.9506	0.7308
0.423	31.9474	144	0.9922	0.7115
0.423	32.8421	148	1.0190	0.7115
0.423	33.9474	153	1.0155	0.7115
0.4223	34.8421	157	1.0041	0.7115
0.4223	35.9474	162	0.9904	0.7308
0.4223	36.8421	166	0.9897	0.7308
0.3989	37.3158	168	0.9897	0.7308

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3