smids_3x_deit_tiny_adamax_00001_fold4

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: 1.1449
• Accuracy: 0.8667

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
0.4417	1.0	225	0.4569	0.8083
0.2331	2.0	450	0.3996	0.8383
0.2011	3.0	675	0.3738	0.8467
0.2185	4.0	900	0.3745	0.8633
0.1818	5.0	1125	0.3725	0.875
0.1063	6.0	1350	0.4112	0.8717
0.1356	7.0	1575	0.4194	0.875
0.1607	8.0	1800	0.4543	0.8533
0.1116	9.0	2025	0.4653	0.8683
0.0765	10.0	2250	0.5277	0.8733
0.0939	11.0	2475	0.5314	0.875
0.0447	12.0	2700	0.6029	0.87
0.0241	13.0	2925	0.6080	0.88
0.0181	14.0	3150	0.6484	0.8783
0.0217	15.0	3375	0.7384	0.8733
0.0012	16.0	3600	0.7227	0.8717
0.0028	17.0	3825	0.7801	0.8683
0.0175	18.0	4050	0.7970	0.8667
0.0003	19.0	4275	0.8483	0.8683
0.0027	20.0	4500	0.8736	0.875
0.0021	21.0	4725	0.8846	0.8717
0.0002	22.0	4950	0.8988	0.875
0.0002	23.0	5175	0.9264	0.8683
0.0001	24.0	5400	0.9315	0.8717
0.0001	25.0	5625	0.9598	0.87
0.0001	26.0	5850	0.9683	0.8667
0.0001	27.0	6075	0.9914	0.8667
0.0	28.0	6300	1.0058	0.8683
0.0001	29.0	6525	1.0095	0.8683
0.0	30.0	6750	1.0176	0.865
0.0	31.0	6975	1.0268	0.865
0.0	32.0	7200	1.0316	0.8667
0.0	33.0	7425	1.0500	0.8667
0.0	34.0	7650	1.0562	0.87
0.0	35.0	7875	1.0705	0.8667
0.0	36.0	8100	1.0794	0.8683
0.0	37.0	8325	1.0868	0.8667
0.0	38.0	8550	1.0946	0.865
0.0	39.0	8775	1.1049	0.8667
0.0	40.0	9000	1.1087	0.8667
0.0	41.0	9225	1.1122	0.8667
0.0091	42.0	9450	1.1194	0.8667
0.0	43.0	9675	1.1272	0.8667
0.0077	44.0	9900	1.1280	0.8667
0.0	45.0	10125	1.1353	0.8667
0.0	46.0	10350	1.1386	0.8667
0.0	47.0	10575	1.1418	0.8683
0.0	48.0	10800	1.1440	0.8667
0.0	49.0	11025	1.1434	0.8667
0.0	50.0	11250	1.1449	0.8667

Framework versions

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