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

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:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 5.5e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 64
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 Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9474	9	1.5677	0.25
1.6369	2.0	19	1.5069	0.2885
1.5075	2.9474	28	1.3568	0.4615
1.2712	4.0	38	1.1267	0.5385
1.2712	4.9474	47	1.0141	0.6538
0.9909	6.0	57	1.0726	0.5962
0.8736	6.9474	66	0.9800	0.7115
0.7965	8.0	76	0.8837	0.6731
0.7513	8.9474	85	0.9431	0.6538
0.7513	10.0	95	0.9317	0.6923
0.6997	10.9474	104	0.9871	0.6923
0.5749	12.0	114	0.9986	0.6923
0.5366	12.9474	123	1.1291	0.6346
0.5108	14.0	133	0.9154	0.7115
0.5108	14.9474	142	0.9546	0.6731
0.5113	16.0	152	1.2550	0.6346
0.4594	16.9474	161	1.0490	0.6923
0.458	18.0	171	1.0402	0.7115
0.4068	18.9474	180	1.1223	0.6923
0.4068	20.0	190	1.0961	0.6731
0.3652	20.9474	199	1.1317	0.6731
0.3516	22.0	209	1.1016	0.6923
0.3517	22.9474	218	1.0740	0.6923
0.3428	24.0	228	1.5363	0.5769
0.3428	24.9474	237	1.1124	0.7308
0.3026	26.0	247	1.2449	0.6923
0.3162	26.9474	256	1.1481	0.6923
0.315	28.0	266	1.2540	0.6731
0.315	28.9474	275	1.2073	0.6923
0.2485	30.0	285	1.2773	0.6923
0.2839	30.9474	294	1.4296	0.6538
0.2719	32.0	304	1.2424	0.6923
0.2396	32.9474	313	1.3423	0.75
0.2396	34.0	323	1.2244	0.7115
0.2761	34.9474	332	1.2601	0.7115
0.2411	36.0	342	1.2808	0.7115
0.2566	36.9474	351	1.2784	0.7115
0.2065	38.0	361	1.2426	0.7115
0.2065	38.9474	370	1.2823	0.7115
0.2815	39.7895	378	1.2943	0.7115