smids_5x_deit_tiny_sgd_00001_fold2

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: 1.0573
• Accuracy: 0.4476

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
1.3977	1.0	375	1.3189	0.3428
1.3124	2.0	750	1.2870	0.3411
1.2542	3.0	1125	1.2595	0.3378
1.2046	4.0	1500	1.2361	0.3478
1.2563	5.0	1875	1.2165	0.3544
1.2759	6.0	2250	1.1999	0.3561
1.1771	7.0	2625	1.1858	0.3527
1.1858	8.0	3000	1.1739	0.3710
1.1713	9.0	3375	1.1636	0.3644
1.1774	10.0	3750	1.1549	0.3760
1.1522	11.0	4125	1.1472	0.3760
1.1182	12.0	4500	1.1403	0.3744
1.1161	13.0	4875	1.1344	0.3827
1.1676	14.0	5250	1.1289	0.3827
1.1382	15.0	5625	1.1238	0.3860
1.129	16.0	6000	1.1191	0.3943
1.1144	17.0	6375	1.1146	0.3910
1.1043	18.0	6750	1.1105	0.3894
1.1008	19.0	7125	1.1065	0.3960
1.1097	20.0	7500	1.1028	0.4077
1.1084	21.0	7875	1.0993	0.4093
1.0777	22.0	8250	1.0960	0.4110
1.0857	23.0	8625	1.0928	0.4126
1.096	24.0	9000	1.0898	0.4126
1.1016	25.0	9375	1.0869	0.4176
1.0637	26.0	9750	1.0843	0.4226
1.0804	27.0	10125	1.0817	0.4226
1.0961	28.0	10500	1.0793	0.4226
1.0888	29.0	10875	1.0771	0.4293
1.0508	30.0	11250	1.0750	0.4293
1.0685	31.0	11625	1.0730	0.4326
1.1026	32.0	12000	1.0712	0.4309
1.0612	33.0	12375	1.0694	0.4359
1.0734	34.0	12750	1.0679	0.4393
1.0868	35.0	13125	1.0664	0.4393
1.0597	36.0	13500	1.0650	0.4393
1.0653	37.0	13875	1.0638	0.4409
1.0598	38.0	14250	1.0627	0.4443
1.0773	39.0	14625	1.0617	0.4443
1.0819	40.0	15000	1.0608	0.4443
1.0608	41.0	15375	1.0600	0.4459
1.0652	42.0	15750	1.0594	0.4459
1.04	43.0	16125	1.0588	0.4476
1.0518	44.0	16500	1.0583	0.4476
1.0814	45.0	16875	1.0580	0.4476
1.0536	46.0	17250	1.0577	0.4476
1.0612	47.0	17625	1.0575	0.4476
1.0833	48.0	18000	1.0574	0.4476
1.0816	49.0	18375	1.0573	0.4476
1.0754	50.0	18750	1.0573	0.4476

Framework versions

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