smids_5x_deit_tiny_sgd_0001_fold1

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.5289
• Accuracy: 0.7763

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.163	1.0	376	1.1394	0.3756
1.1206	2.0	752	1.0830	0.4240
1.0283	3.0	1128	1.0398	0.4624
1.0304	4.0	1504	1.0025	0.4925
0.9327	5.0	1880	0.9675	0.5159
0.9025	6.0	2256	0.9343	0.5342
0.813	7.0	2632	0.9040	0.5492
0.8373	8.0	3008	0.8750	0.5793
0.8156	9.0	3384	0.8442	0.6093
0.7741	10.0	3760	0.8156	0.6344
0.6787	11.0	4136	0.7888	0.6477
0.6914	12.0	4512	0.7651	0.6561
0.6673	13.0	4888	0.7424	0.6761
0.6471	14.0	5264	0.7219	0.6895
0.5827	15.0	5640	0.7043	0.6978
0.5679	16.0	6016	0.6891	0.7062
0.5387	17.0	6392	0.6730	0.7095
0.5827	18.0	6768	0.6602	0.7145
0.5764	19.0	7144	0.6481	0.7179
0.5667	20.0	7520	0.6375	0.7312
0.5598	21.0	7896	0.6271	0.7329
0.4963	22.0	8272	0.6181	0.7346
0.5399	23.0	8648	0.6097	0.7396
0.6005	24.0	9024	0.6025	0.7429
0.5535	25.0	9400	0.5952	0.7479
0.5292	26.0	9776	0.5886	0.7513
0.4834	27.0	10152	0.5826	0.7529
0.4735	28.0	10528	0.5772	0.7546
0.5034	29.0	10904	0.5722	0.7563
0.4846	30.0	11280	0.5675	0.7579
0.5398	31.0	11656	0.5634	0.7596
0.5623	32.0	12032	0.5594	0.7646
0.5122	33.0	12408	0.5557	0.7646
0.483	34.0	12784	0.5523	0.7663
0.4676	35.0	13160	0.5492	0.7663
0.4655	36.0	13536	0.5464	0.7679
0.4231	37.0	13912	0.5438	0.7663
0.5103	38.0	14288	0.5415	0.7663
0.4626	39.0	14664	0.5394	0.7679
0.4372	40.0	15040	0.5375	0.7679
0.496	41.0	15416	0.5357	0.7713
0.4002	42.0	15792	0.5343	0.7746
0.4506	43.0	16168	0.5329	0.7746
0.472	44.0	16544	0.5318	0.7746
0.467	45.0	16920	0.5309	0.7763
0.4861	46.0	17296	0.5301	0.7746
0.4576	47.0	17672	0.5296	0.7746
0.4597	48.0	18048	0.5292	0.7763
0.4465	49.0	18424	0.5290	0.7763
0.3972	50.0	18800	0.5289	0.7763

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2