smids_3x_deit_small_rms_0001_fold3

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: 1.1839
• Accuracy: 0.885

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
0.3285	1.0	225	0.3079	0.8733
0.249	2.0	450	0.2876	0.8933
0.1277	3.0	675	0.3615	0.8983
0.1138	4.0	900	0.3786	0.9
0.0577	5.0	1125	0.5033	0.8833
0.0713	6.0	1350	0.6038	0.88
0.0526	7.0	1575	0.5533	0.8783
0.1027	8.0	1800	0.6305	0.885
0.1119	9.0	2025	0.5230	0.8883
0.0596	10.0	2250	0.4789	0.895
0.0427	11.0	2475	0.6163	0.8883
0.0824	12.0	2700	0.6835	0.8817
0.029	13.0	2925	0.6829	0.8933
0.0703	14.0	3150	0.7193	0.8833
0.0396	15.0	3375	0.6819	0.9
0.0094	16.0	3600	0.6229	0.9083
0.032	17.0	3825	0.7440	0.885
0.0019	18.0	4050	0.6469	0.8933
0.0426	19.0	4275	0.6592	0.8817
0.0085	20.0	4500	0.7146	0.8967
0.0003	21.0	4725	0.7501	0.8933
0.0003	22.0	4950	0.7432	0.885
0.0137	23.0	5175	0.7637	0.8717
0.0017	24.0	5400	0.8802	0.8817
0.0465	25.0	5625	0.7437	0.8883
0.0022	26.0	5850	0.7726	0.8933
0.0063	27.0	6075	0.8349	0.8883
0.0003	28.0	6300	0.7609	0.8933
0.0094	29.0	6525	0.8877	0.8917
0.0308	30.0	6750	0.8912	0.8967
0.0037	31.0	6975	0.9195	0.8817
0.0	32.0	7200	0.8416	0.8967
0.0	33.0	7425	0.8645	0.895
0.0046	34.0	7650	0.9341	0.9
0.0	35.0	7875	0.9327	0.8967
0.0014	36.0	8100	1.0523	0.8733
0.0027	37.0	8325	0.9916	0.8883
0.0	38.0	8550	0.9734	0.8983
0.0	39.0	8775	1.0426	0.89
0.0	40.0	9000	1.0377	0.8917
0.0	41.0	9225	1.0718	0.8867
0.0	42.0	9450	1.1007	0.89
0.0	43.0	9675	1.1332	0.8883
0.0	44.0	9900	1.1398	0.89
0.0	45.0	10125	1.1492	0.89
0.0	46.0	10350	1.1564	0.8883
0.0026	47.0	10575	1.1721	0.885
0.0	48.0	10800	1.1759	0.885
0.0	49.0	11025	1.1812	0.885
0.0	50.0	11250	1.1839	0.885

Framework versions

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