swinv2-tiny-patch4-window8-256-DMAE-da2-colab2

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.7389
• Accuracy: 0.7174

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: 1.5e-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
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3473	0.9565	11	1.3763	0.2609
1.2385	1.9348	22	1.2596	0.5
1.1683	2.9130	33	1.2476	0.5
1.0809	3.9783	45	1.1625	0.5435
1.0147	4.9565	56	1.0490	0.5435
0.9755	5.9348	67	0.9155	0.6739
0.89	6.9130	78	0.9343	0.6304
0.832	7.9783	90	0.7973	0.6522
0.7846	8.9565	101	0.8325	0.6087
0.74	9.9348	112	0.8328	0.6087
0.679	10.9130	123	0.7195	0.7174
0.6487	11.9783	135	0.6740	0.7826
0.6774	12.9565	146	0.7157	0.6957
0.5727	13.9348	157	0.6934	0.7174
0.5872	14.9130	168	0.7335	0.6739
0.6116	15.9783	180	0.6892	0.7609
0.5783	16.9565	191	0.6796	0.7174
0.5658	17.9348	202	0.6966	0.7609
0.5447	18.9130	213	0.6708	0.7174
0.5452	19.9783	225	0.7297	0.6957
0.509	20.9565	236	0.7137	0.6522
0.527	21.9348	247	0.7047	0.7174
0.5382	22.9130	258	0.7737	0.6739
0.508	23.9783	270	0.7073	0.7174
0.4579	24.9565	281	0.7355	0.6739
0.492	25.9348	292	0.7152	0.7174
0.4273	26.9130	303	0.7243	0.7174
0.4823	27.9783	315	0.7477	0.6739
0.4659	28.9565	326	0.7193	0.6957
0.4401	29.9348	337	0.7509	0.6957
0.4717	30.9130	348	0.7327	0.6957
0.4139	31.9783	360	0.7041	0.7174
0.4365	32.9565	371	0.7286	0.6957
0.4088	33.9348	382	0.7495	0.7174
0.4505	34.9130	393	0.7399	0.7174
0.4143	35.9783	405	0.7160	0.7174
0.4093	36.9565	416	0.7210	0.7174
0.43	37.9348	427	0.7342	0.7174
0.424	38.9130	438	0.7386	0.7174
0.4169	39.1087	440	0.7389	0.7174

Framework versions

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