smids_10x_deit_tiny_sgd_00001_fold3

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: 0.9716
• Accuracy: 0.5233

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: 1e-05
• 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
1.3912	1.0	750	1.3088	0.3483
1.282	2.0	1500	1.2523	0.3583
1.2674	3.0	2250	1.2106	0.375
1.2413	4.0	3000	1.1803	0.3833
1.1441	5.0	3750	1.1586	0.39
1.1769	6.0	4500	1.1422	0.3967
1.1041	7.0	5250	1.1291	0.395
1.1052	8.0	6000	1.1182	0.4083
1.1033	9.0	6750	1.1088	0.4133
1.0691	10.0	7500	1.1002	0.42
1.0707	11.0	8250	1.0922	0.4333
1.055	12.0	9000	1.0847	0.4417
1.0312	13.0	9750	1.0777	0.4433
1.0496	14.0	10500	1.0710	0.4417
1.0476	15.0	11250	1.0646	0.45
1.0102	16.0	12000	1.0585	0.455
0.9929	17.0	12750	1.0527	0.4617
1.009	18.0	13500	1.0471	0.47
0.9872	19.0	14250	1.0418	0.4767
0.9829	20.0	15000	1.0367	0.485
0.9531	21.0	15750	1.0318	0.4867
0.9943	22.0	16500	1.0272	0.4917
0.9746	23.0	17250	1.0227	0.4933
0.9878	24.0	18000	1.0185	0.4933
0.9395	25.0	18750	1.0145	0.5017
0.9881	26.0	19500	1.0107	0.505
0.9509	27.0	20250	1.0071	0.51
0.9714	28.0	21000	1.0038	0.5117
0.9762	29.0	21750	1.0005	0.5133
0.9566	30.0	22500	0.9975	0.5133
0.9136	31.0	23250	0.9946	0.5133
0.9478	32.0	24000	0.9920	0.5167
0.9511	33.0	24750	0.9895	0.5167
0.9608	34.0	25500	0.9872	0.5183
0.9345	35.0	26250	0.9850	0.5183
0.941	36.0	27000	0.9831	0.5217
0.9373	37.0	27750	0.9813	0.5217
0.9836	38.0	28500	0.9796	0.5233
0.9294	39.0	29250	0.9782	0.5217
0.9394	40.0	30000	0.9769	0.5233
0.9196	41.0	30750	0.9757	0.5233
0.941	42.0	31500	0.9747	0.5233
0.9468	43.0	32250	0.9739	0.5233
0.9204	44.0	33000	0.9732	0.5233
0.9194	45.0	33750	0.9727	0.5233
0.9493	46.0	34500	0.9722	0.5233
0.9121	47.0	35250	0.9719	0.5233
0.9108	48.0	36000	0.9717	0.5233
0.8965	49.0	36750	0.9716	0.5233
0.896	50.0	37500	0.9716	0.5233

Framework versions

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