smids_3x_deit_tiny_adamax_001_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: 1.2812
• Accuracy: 0.8633

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: 0.001
• 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
0.6232	1.0	225	0.4782	0.8
0.3036	2.0	450	0.4594	0.8367
0.2595	3.0	675	0.4759	0.8233
0.267	4.0	900	0.4797	0.8467
0.194	5.0	1125	0.4465	0.8583
0.1353	6.0	1350	0.5225	0.83
0.1792	7.0	1575	0.4654	0.8533
0.2076	8.0	1800	0.5317	0.8567
0.1775	9.0	2025	0.5872	0.83
0.0889	10.0	2250	0.6356	0.8367
0.1784	11.0	2475	0.6312	0.8533
0.0901	12.0	2700	0.7276	0.8517
0.0533	13.0	2925	0.7284	0.855
0.0494	14.0	3150	1.0522	0.8267
0.1015	15.0	3375	0.9687	0.8483
0.0095	16.0	3600	0.8525	0.8517
0.0374	17.0	3825	0.7948	0.8667
0.0184	18.0	4050	0.9557	0.845
0.0248	19.0	4275	0.8840	0.875
0.0164	20.0	4500	1.1244	0.8383
0.0604	21.0	4725	1.1197	0.85
0.0003	22.0	4950	1.2021	0.8433
0.0239	23.0	5175	1.0238	0.8617
0.0001	24.0	5400	0.9855	0.8617
0.0039	25.0	5625	0.9755	0.8567
0.0002	26.0	5850	1.1897	0.8633
0.0001	27.0	6075	1.2400	0.85
0.0047	28.0	6300	1.1483	0.8567
0.0001	29.0	6525	1.0989	0.8717
0.0	30.0	6750	1.2154	0.875
0.0	31.0	6975	1.1782	0.86
0.0005	32.0	7200	1.1567	0.8733
0.0	33.0	7425	1.2479	0.8667
0.0	34.0	7650	1.2149	0.8617
0.0134	35.0	7875	1.2307	0.8733
0.0	36.0	8100	1.1946	0.8567
0.0	37.0	8325	1.3265	0.865
0.0	38.0	8550	1.2696	0.8583
0.0	39.0	8775	1.2332	0.8617
0.0	40.0	9000	1.2400	0.8633
0.0	41.0	9225	1.2491	0.8617
0.0035	42.0	9450	1.2522	0.8617
0.0	43.0	9675	1.2564	0.8633
0.003	44.0	9900	1.2640	0.8617
0.0	45.0	10125	1.2656	0.8633
0.0	46.0	10350	1.2708	0.8617
0.0	47.0	10575	1.2766	0.8633
0.0	48.0	10800	1.2793	0.8617
0.0	49.0	11025	1.2821	0.8633
0.0	50.0	11250	1.2812	0.8633

Framework versions

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