smids_3x_deit_base_adamax_0001_fold5

This model is a fine-tuned version of facebook/deit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.8955
• Accuracy: 0.895

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
0.3014	1.0	225	0.3213	0.8717
0.0815	2.0	450	0.2900	0.8883
0.0668	3.0	675	0.4157	0.8817
0.0616	4.0	900	0.5373	0.8667
0.0313	5.0	1125	0.6762	0.8783
0.0139	6.0	1350	0.5421	0.8867
0.0228	7.0	1575	0.5956	0.885
0.0003	8.0	1800	0.6382	0.89
0.0191	9.0	2025	0.5798	0.8967
0.0066	10.0	2250	0.6950	0.8817
0.0002	11.0	2475	0.6959	0.8783
0.0105	12.0	2700	0.7866	0.8733
0.0009	13.0	2925	0.7496	0.8833
0.02	14.0	3150	0.7895	0.8917
0.0	15.0	3375	0.7885	0.8867
0.0	16.0	3600	0.8188	0.8783
0.0001	17.0	3825	0.8229	0.895
0.013	18.0	4050	0.8881	0.8867
0.008	19.0	4275	0.8377	0.8933
0.0001	20.0	4500	0.8361	0.8833
0.0	21.0	4725	0.8087	0.89
0.0	22.0	4950	0.8001	0.895
0.0	23.0	5175	0.7918	0.8933
0.0043	24.0	5400	0.8030	0.8917
0.0	25.0	5625	0.8092	0.895
0.0	26.0	5850	0.8142	0.8917
0.0	27.0	6075	0.8480	0.8833
0.0	28.0	6300	0.8065	0.8967
0.0035	29.0	6525	0.8102	0.8933
0.0	30.0	6750	0.8694	0.8933
0.0	31.0	6975	0.8371	0.8917
0.0	32.0	7200	0.8420	0.8933
0.0031	33.0	7425	0.8419	0.8933
0.0	34.0	7650	0.8459	0.8933
0.0028	35.0	7875	0.8578	0.8967
0.0	36.0	8100	0.8632	0.8917
0.0032	37.0	8325	0.8626	0.8933
0.0	38.0	8550	0.8689	0.895
0.0	39.0	8775	0.8751	0.8933
0.0	40.0	9000	0.8742	0.8933
0.0	41.0	9225	0.8784	0.8933
0.0	42.0	9450	0.8817	0.8933
0.0025	43.0	9675	0.8831	0.8933
0.0	44.0	9900	0.8843	0.895
0.0	45.0	10125	0.8846	0.895
0.0	46.0	10350	0.8903	0.895
0.0	47.0	10575	0.8928	0.895
0.0	48.0	10800	0.8941	0.895
0.0	49.0	11025	0.8949	0.895
0.0	50.0	11250	0.8955	0.895

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2