smids_10x_deit_small_rms_001_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: 2.4181
• Accuracy: 0.7883

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: 0.001
• 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
0.808	1.0	750	0.7526	0.6333
0.7422	2.0	1500	0.7345	0.58
0.7146	3.0	2250	0.7176	0.6433
0.701	4.0	3000	0.6566	0.6933
0.6799	5.0	3750	0.5734	0.745
0.6007	6.0	4500	0.6368	0.72
0.5519	7.0	5250	0.5547	0.7833
0.6119	8.0	6000	0.5502	0.7667
0.5599	9.0	6750	0.5552	0.75
0.508	10.0	7500	0.5666	0.7383
0.5209	11.0	8250	0.5288	0.7683
0.6053	12.0	9000	0.5408	0.765
0.4938	13.0	9750	0.5449	0.755
0.5012	14.0	10500	0.6211	0.7533
0.4544	15.0	11250	0.6619	0.725
0.4855	16.0	12000	0.5101	0.8083
0.3525	17.0	12750	0.5278	0.7833
0.4312	18.0	13500	0.5051	0.7917
0.4593	19.0	14250	0.5277	0.7867
0.3939	20.0	15000	0.5517	0.7833
0.5185	21.0	15750	0.5418	0.7667
0.424	22.0	16500	0.5465	0.7917
0.3637	23.0	17250	0.5971	0.785
0.4457	24.0	18000	0.5681	0.7933
0.4023	25.0	18750	0.5160	0.805
0.3012	26.0	19500	0.5373	0.8283
0.2933	27.0	20250	0.5885	0.8067
0.3104	28.0	21000	0.6014	0.8017
0.292	29.0	21750	0.6093	0.8033
0.3506	30.0	22500	0.6800	0.7633
0.2777	31.0	23250	0.6858	0.795
0.2396	32.0	24000	0.6442	0.805
0.3316	33.0	24750	0.6440	0.81
0.2958	34.0	25500	0.6532	0.8117
0.2121	35.0	26250	0.7494	0.8083
0.1764	36.0	27000	0.7942	0.7933
0.1963	37.0	27750	0.7817	0.7883
0.1829	38.0	28500	0.8010	0.7917
0.1937	39.0	29250	0.8544	0.795
0.1493	40.0	30000	0.9520	0.7967
0.1419	41.0	30750	0.9695	0.81
0.1784	42.0	31500	1.0763	0.8017
0.1485	43.0	32250	1.1404	0.825
0.0665	44.0	33000	1.3155	0.7933
0.1004	45.0	33750	1.5689	0.7933
0.0917	46.0	34500	1.5920	0.7917
0.0556	47.0	35250	1.8022	0.7967
0.0427	48.0	36000	2.0148	0.8067
0.032	49.0	36750	2.3253	0.795
0.012	50.0	37500	2.4181	0.7883

Framework versions

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