smids_5x_deit_small_rms_0001_fold4

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

• Loss: 1.4198
• Accuracy: 0.8783

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
0.2762	1.0	375	0.4287	0.8433
0.1833	2.0	750	0.3850	0.88
0.0712	3.0	1125	0.5130	0.8533
0.1056	4.0	1500	0.5359	0.8617
0.0785	5.0	1875	0.5955	0.8767
0.0284	6.0	2250	0.6660	0.875
0.0398	7.0	2625	0.7191	0.8517
0.0358	8.0	3000	0.7890	0.855
0.0148	9.0	3375	0.8497	0.88
0.0842	10.0	3750	0.8652	0.8583
0.0545	11.0	4125	0.9852	0.8583
0.0347	12.0	4500	0.8898	0.8733
0.0481	13.0	4875	0.8728	0.8417
0.0109	14.0	5250	0.7860	0.8717
0.004	15.0	5625	0.9119	0.87
0.0001	16.0	6000	1.0067	0.8667
0.0208	17.0	6375	0.9891	0.8633
0.0022	18.0	6750	0.9452	0.8733
0.0258	19.0	7125	0.9437	0.865
0.0001	20.0	7500	0.9752	0.8867
0.014	21.0	7875	0.9198	0.8717
0.0001	22.0	8250	1.0263	0.8817
0.0007	23.0	8625	1.0253	0.875
0.0298	24.0	9000	0.9487	0.8717
0.0364	25.0	9375	1.0414	0.8567
0.0168	26.0	9750	0.9402	0.8667
0.0267	27.0	10125	1.0294	0.87
0.0004	28.0	10500	1.1186	0.8617
0.0	29.0	10875	1.0351	0.8733
0.0136	30.0	11250	1.0077	0.8667
0.0403	31.0	11625	0.9771	0.885
0.0141	32.0	12000	0.9784	0.875
0.0079	33.0	12375	0.9784	0.875
0.0198	34.0	12750	1.1841	0.87
0.0247	35.0	13125	1.1305	0.8767
0.0348	36.0	13500	1.1191	0.8733
0.0086	37.0	13875	1.0822	0.8817
0.004	38.0	14250	1.1525	0.8733
0.0	39.0	14625	1.3113	0.87
0.0	40.0	15000	1.2648	0.8767
0.0	41.0	15375	1.3219	0.88
0.0	42.0	15750	1.3106	0.88
0.0	43.0	16125	1.3404	0.8783
0.0	44.0	16500	1.3645	0.8783
0.0	45.0	16875	1.3862	0.875
0.0	46.0	17250	1.4020	0.8767
0.0	47.0	17625	1.4061	0.8783
0.0	48.0	18000	1.4149	0.8783
0.0	49.0	18375	1.4183	0.8783
0.0	50.0	18750	1.4198	0.8783

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2