smids_5x_deit_tiny_rms_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: 1.0235
• Accuracy: 0.9017

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
0.3072	1.0	375	0.3497	0.8733
0.1839	2.0	750	0.4255	0.87
0.1528	3.0	1125	0.4557	0.8567
0.1267	4.0	1500	0.3726	0.89
0.1353	5.0	1875	0.4467	0.8917
0.0943	6.0	2250	0.4927	0.91
0.1102	7.0	2625	0.6801	0.8833
0.1057	8.0	3000	0.6555	0.88
0.032	9.0	3375	0.7410	0.8783
0.0843	10.0	3750	0.8478	0.8667
0.0459	11.0	4125	0.6987	0.8917
0.0092	12.0	4500	0.7040	0.8917
0.0349	13.0	4875	0.7908	0.885
0.0111	14.0	5250	0.7260	0.8983
0.0286	15.0	5625	0.7556	0.89
0.0202	16.0	6000	0.7922	0.885
0.0017	17.0	6375	0.7780	0.89
0.0426	18.0	6750	0.7356	0.9033
0.0036	19.0	7125	0.7906	0.88
0.0088	20.0	7500	0.8591	0.8883
0.014	21.0	7875	0.9590	0.8867
0.0	22.0	8250	0.9929	0.8783
0.0363	23.0	8625	0.9559	0.89
0.0156	24.0	9000	0.9344	0.88
0.0345	25.0	9375	0.8898	0.8917
0.0005	26.0	9750	0.9066	0.9
0.0104	27.0	10125	0.9018	0.8983
0.0026	28.0	10500	0.8354	0.89
0.0098	29.0	10875	1.0679	0.885
0.0077	30.0	11250	0.8084	0.8933
0.007	31.0	11625	0.9761	0.8833
0.0079	32.0	12000	0.8798	0.8867
0.0211	33.0	12375	0.9152	0.8967
0.0205	34.0	12750	0.8595	0.8967
0.0	35.0	13125	0.9123	0.8983
0.0	36.0	13500	1.0918	0.8817
0.0001	37.0	13875	0.9598	0.8917
0.0	38.0	14250	0.9005	0.8933
0.0	39.0	14625	0.9817	0.895
0.003	40.0	15000	1.0214	0.8933
0.0	41.0	15375	1.0132	0.895
0.0012	42.0	15750	1.0443	0.8933
0.0	43.0	16125	1.0086	0.895
0.0	44.0	16500	1.0148	0.895
0.0	45.0	16875	1.0171	0.895
0.0	46.0	17250	1.0091	0.8967
0.0	47.0	17625	1.0118	0.8983
0.0	48.0	18000	1.0184	0.9017
0.0	49.0	18375	1.0213	0.9017
0.0	50.0	18750	1.0235	0.9017

Framework versions

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