swinv2-tiny-patch4-window8-256-dmae-va-U5-42

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: 0.6806
• Accuracy: 0.8333

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 42

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9	7	1.3299	0.4
1.3678	1.94	15	1.2662	0.45
1.3678	2.97	23	1.0959	0.5167
1.2546	4.0	31	0.9759	0.55
1.0271	4.9	38	0.9375	0.5667
1.0271	5.94	46	0.8728	0.6
0.8075	6.97	54	0.7360	0.7167
0.7026	8.0	62	0.8097	0.6667
0.7026	8.9	69	0.7074	0.7
0.5711	9.94	77	0.6913	0.7833
0.5063	10.97	85	0.7462	0.7167
0.5063	12.0	93	0.8509	0.5833
0.4701	12.9	100	0.6895	0.7333
0.3708	13.94	108	0.7593	0.6833
0.3708	14.97	116	0.8622	0.7167
0.3581	16.0	124	0.7504	0.7667
0.3581	16.9	131	0.6694	0.75
0.3342	17.94	139	0.7262	0.7333
0.2979	18.97	147	0.7234	0.7167
0.2979	20.0	155	0.6403	0.7833
0.2919	20.9	162	0.6847	0.7667
0.274	21.94	170	0.6943	0.75
0.274	22.97	178	0.7235	0.7833
0.2434	24.0	186	0.7836	0.75
0.239	24.9	193	0.7199	0.8167
0.239	25.94	201	0.6806	0.8333
0.2184	26.97	209	0.6923	0.8
0.2176	28.0	217	0.7070	0.7833
0.2176	28.9	224	0.6991	0.7667
0.231	29.94	232	0.7043	0.7833
0.1889	30.97	240	0.6575	0.7667
0.1889	32.0	248	0.7521	0.75
0.2033	32.9	255	0.7062	0.7833
0.2033	33.94	263	0.6958	0.8
0.1891	34.97	271	0.7189	0.8
0.1739	36.0	279	0.7457	0.8
0.1739	36.9	286	0.7766	0.7833
0.1949	37.94	294	0.7808	0.7667

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2