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

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: 1.2824
• Accuracy: 0.6346

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 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	0.8696	5	1.5656	0.3462
No log	1.8261	10	1.5187	0.3654
1.5752	2.7826	15	1.4276	0.4615
1.5752	3.9130	21	1.2763	0.4615
1.3428	4.8696	26	1.1021	0.5385
1.3428	5.8261	31	1.0576	0.5385
1.0139	6.7826	36	1.0284	0.5769
1.0139	7.9130	42	1.0838	0.5577
1.0139	8.8696	47	1.0973	0.5192
0.7923	9.8261	52	1.0880	0.5962
0.7923	10.7826	57	1.1609	0.5577
0.7424	11.9130	63	1.0270	0.6731
0.7424	12.8696	68	1.0543	0.6154
0.6486	13.8261	73	0.9941	0.6154
0.6486	14.7826	78	1.2237	0.5577
0.595	15.9130	84	1.0447	0.6154
0.595	16.8696	89	1.2137	0.6538
0.595	17.8261	94	1.0769	0.6538
0.5128	18.7826	99	1.1559	0.6346
0.5128	19.9130	105	1.1879	0.6346
0.4968	20.8696	110	1.1066	0.6154
0.4968	21.8261	115	1.0849	0.6154
0.4385	22.7826	120	1.1078	0.6731
0.4385	23.9130	126	1.2649	0.6538
0.4385	24.8696	131	1.1754	0.6346
0.437	25.8261	136	1.1540	0.6346
0.437	26.7826	141	1.3327	0.6154
0.3852	27.9130	147	1.2256	0.6538
0.3852	28.8696	152	1.2546	0.6154
0.3813	29.8261	157	1.3192	0.6154
0.3813	30.7826	162	1.2768	0.6154
0.3614	31.9130	168	1.2230	0.6154
0.3614	32.8696	173	1.2120	0.6154
0.3614	33.8261	178	1.3150	0.6538
0.3339	34.7826	183	1.3258	0.6538
0.3339	35.9130	189	1.2851	0.6154
0.3294	36.8696	194	1.2712	0.6154
0.3294	37.8261	199	1.2630	0.6154
0.3269	38.7826	204	1.2713	0.6346
0.3269	39.9130	210	1.2824	0.6346

Framework versions

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