smids_10x_deit_tiny_adamax_0001_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: 0.8519
• Accuracy: 0.8983

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
0.201	1.0	750	0.2389	0.905
0.2264	2.0	1500	0.3492	0.9033
0.086	3.0	2250	0.4323	0.8917
0.1151	4.0	3000	0.4359	0.8867
0.0552	5.0	3750	0.5210	0.8933
0.0373	6.0	4500	0.7270	0.885
0.0559	7.0	5250	0.6401	0.8917
0.0155	8.0	6000	0.8201	0.8883
0.0012	9.0	6750	0.7935	0.9017
0.0025	10.0	7500	0.8189	0.8983
0.0349	11.0	8250	0.9771	0.8933
0.0002	12.0	9000	0.9998	0.8817
0.0235	13.0	9750	0.8826	0.8883
0.003	14.0	10500	0.8505	0.9
0.0	15.0	11250	0.9776	0.895
0.0	16.0	12000	0.8400	0.905
0.0	17.0	12750	0.8401	0.9017
0.0	18.0	13500	0.9571	0.8967
0.0	19.0	14250	0.8971	0.89
0.0	20.0	15000	0.8936	0.8917
0.0005	21.0	15750	0.8479	0.8917
0.0	22.0	16500	0.8022	0.9083
0.0136	23.0	17250	0.7709	0.9083
0.0	24.0	18000	0.8730	0.8983
0.0	25.0	18750	0.9311	0.895
0.0	26.0	19500	0.8622	0.905
0.0	27.0	20250	0.8794	0.9017
0.0	28.0	21000	0.8617	0.9017
0.0	29.0	21750	0.8575	0.905
0.0	30.0	22500	0.8219	0.9033
0.0	31.0	23250	0.8966	0.9017
0.0	32.0	24000	0.8450	0.8967
0.0	33.0	24750	0.8688	0.8933
0.0	34.0	25500	0.8664	0.8933
0.0	35.0	26250	0.8160	0.9017
0.0	36.0	27000	0.8559	0.895
0.0	37.0	27750	0.8682	0.8967
0.0	38.0	28500	0.8657	0.8983
0.0	39.0	29250	0.8354	0.8983
0.0	40.0	30000	0.8506	0.8983
0.0	41.0	30750	0.8341	0.8967
0.0	42.0	31500	0.8506	0.9
0.0	43.0	32250	0.8450	0.8983
0.0	44.0	33000	0.8494	0.9
0.0	45.0	33750	0.8478	0.9
0.0	46.0	34500	0.8491	0.9
0.0	47.0	35250	0.8495	0.9
0.0	48.0	36000	0.8491	0.9
0.0	49.0	36750	0.8507	0.9
0.0	50.0	37500	0.8519	0.8983

Framework versions

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