smids_3x_deit_small_sgd_00001_fold4

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

• Loss: 1.0105
• Accuracy: 0.505

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
1.0492	1.0	225	1.0666	0.43
1.0754	2.0	450	1.0640	0.43
1.0504	3.0	675	1.0616	0.4283
1.071	4.0	900	1.0591	0.43
1.052	5.0	1125	1.0568	0.4333
1.0616	6.0	1350	1.0545	0.44
1.0716	7.0	1575	1.0524	0.4433
1.0533	8.0	1800	1.0502	0.4417
1.0683	9.0	2025	1.0482	0.44
1.0375	10.0	2250	1.0461	0.4417
1.0594	11.0	2475	1.0442	0.445
1.0638	12.0	2700	1.0423	0.4467
1.0743	13.0	2925	1.0405	0.45
1.0117	14.0	3150	1.0387	0.4517
1.0604	15.0	3375	1.0370	0.4517
1.0498	16.0	3600	1.0354	0.4567
1.0315	17.0	3825	1.0338	0.46
1.0306	18.0	4050	1.0323	0.465
1.0262	19.0	4275	1.0309	0.4667
1.0262	20.0	4500	1.0294	0.4667
1.0341	21.0	4725	1.0281	0.4683
1.0464	22.0	4950	1.0268	0.4717
1.0098	23.0	5175	1.0255	0.4733
1.029	24.0	5400	1.0243	0.475
1.0091	25.0	5625	1.0231	0.4817
1.017	26.0	5850	1.0221	0.4833
1.0365	27.0	6075	1.0210	0.4883
1.019	28.0	6300	1.0200	0.4883
1.0442	29.0	6525	1.0191	0.4883
1.0415	30.0	6750	1.0182	0.4867
1.0316	31.0	6975	1.0174	0.4883
1.045	32.0	7200	1.0166	0.4883
1.0078	33.0	7425	1.0159	0.49
1.023	34.0	7650	1.0152	0.49
1.0174	35.0	7875	1.0146	0.495
1.0095	36.0	8100	1.0140	0.5
1.0162	37.0	8325	1.0135	0.5
1.0427	38.0	8550	1.0130	0.5
1.0155	39.0	8775	1.0125	0.5033
1.0159	40.0	9000	1.0122	0.505
1.0255	41.0	9225	1.0118	0.505
1.023	42.0	9450	1.0115	0.5067
1.0068	43.0	9675	1.0113	0.505
1.0321	44.0	9900	1.0110	0.505
1.0329	45.0	10125	1.0109	0.505
1.0275	46.0	10350	1.0107	0.505
1.0181	47.0	10575	1.0106	0.505
1.0137	48.0	10800	1.0106	0.505
1.0177	49.0	11025	1.0105	0.505
1.0148	50.0	11250	1.0105	0.505

Framework versions

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