smids_3x_deit_tiny_sgd_00001_fold5

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: 1.1096
• Accuracy: 0.3917

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
1.3197	1.0	225	1.3390	0.3517
1.27	2.0	450	1.3176	0.3517
1.3368	3.0	675	1.2980	0.36
1.2821	4.0	900	1.2802	0.3583
1.2856	5.0	1125	1.2641	0.3633
1.2974	6.0	1350	1.2496	0.3633
1.2807	7.0	1575	1.2366	0.3733
1.2428	8.0	1800	1.2247	0.3717
1.2255	9.0	2025	1.2140	0.375
1.2238	10.0	2250	1.2044	0.3717
1.1757	11.0	2475	1.1957	0.3733
1.2719	12.0	2700	1.1877	0.38
1.2029	13.0	2925	1.1807	0.38
1.2184	14.0	3150	1.1743	0.3733
1.2016	15.0	3375	1.1685	0.3717
1.1832	16.0	3600	1.1634	0.385
1.1841	17.0	3825	1.1587	0.3883
1.1122	18.0	4050	1.1544	0.395
1.1798	19.0	4275	1.1504	0.3917
1.1614	20.0	4500	1.1469	0.4
1.1973	21.0	4725	1.1435	0.4017
1.1713	22.0	4950	1.1404	0.4083
1.1334	23.0	5175	1.1376	0.4067
1.1477	24.0	5400	1.1349	0.4083
1.126	25.0	5625	1.1325	0.4067
1.1346	26.0	5850	1.1303	0.3983
1.111	27.0	6075	1.1282	0.4
1.0942	28.0	6300	1.1263	0.4067
1.1712	29.0	6525	1.1245	0.405
1.1249	30.0	6750	1.1229	0.4083
1.13	31.0	6975	1.1213	0.4067
1.1043	32.0	7200	1.1199	0.405
1.1162	33.0	7425	1.1186	0.4033
1.1041	34.0	7650	1.1174	0.4017
1.131	35.0	7875	1.1163	0.3983
1.1255	36.0	8100	1.1153	0.3983
1.0951	37.0	8325	1.1144	0.3983
1.1402	38.0	8550	1.1135	0.3967
1.0931	39.0	8775	1.1128	0.3967
1.1012	40.0	9000	1.1122	0.3967
1.1339	41.0	9225	1.1116	0.395
1.1212	42.0	9450	1.1111	0.395
1.1447	43.0	9675	1.1107	0.395
1.1467	44.0	9900	1.1104	0.395
1.1224	45.0	10125	1.1101	0.3917
1.104	46.0	10350	1.1099	0.3917
1.0801	47.0	10575	1.1098	0.39
1.1097	48.0	10800	1.1097	0.3917
1.0996	49.0	11025	1.1096	0.3917
1.1198	50.0	11250	1.1096	0.3917

Framework versions

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