hushem_40x_deit_small_rms_001_fold2

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: 7.1718
• Accuracy: 0.6

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
1.2965	1.0	215	1.2091	0.4667
0.8177	2.0	430	1.2899	0.5333
0.7013	3.0	645	1.2461	0.5111
0.6618	4.0	860	1.7839	0.4444
0.462	5.0	1075	1.6189	0.4889
0.4201	6.0	1290	2.6718	0.4889
0.3471	7.0	1505	2.6469	0.5556
0.3247	8.0	1720	2.5316	0.5778
0.3947	9.0	1935	2.1794	0.5556
0.3688	10.0	2150	2.4409	0.4889
0.3554	11.0	2365	2.7892	0.5556
0.3048	12.0	2580	3.7146	0.5333
0.3088	13.0	2795	2.9395	0.6
0.2595	14.0	3010	2.6649	0.6222
0.2467	15.0	3225	3.0874	0.5778
0.2539	16.0	3440	3.4419	0.6
0.2637	17.0	3655	3.2496	0.6
0.2157	18.0	3870	3.3961	0.6
0.1038	19.0	4085	4.4011	0.6
0.1844	20.0	4300	3.9340	0.5111
0.1977	21.0	4515	4.0238	0.5556
0.1597	22.0	4730	3.8533	0.5778
0.1177	23.0	4945	4.1555	0.5556
0.0976	24.0	5160	3.8653	0.6222
0.106	25.0	5375	3.5209	0.5778
0.0483	26.0	5590	4.5365	0.6222
0.0977	27.0	5805	3.6449	0.6
0.115	28.0	6020	4.3353	0.6
0.0948	29.0	6235	4.1300	0.6222
0.0615	30.0	6450	4.7232	0.6
0.1033	31.0	6665	4.3508	0.5778
0.0812	32.0	6880	4.5803	0.5778
0.056	33.0	7095	4.4372	0.6222
0.0361	34.0	7310	4.9845	0.6444
0.0141	35.0	7525	6.0367	0.5778
0.0009	36.0	7740	5.8383	0.6222
0.0009	37.0	7955	5.7637	0.6222
0.0203	38.0	8170	5.3901	0.6
0.049	39.0	8385	5.5891	0.5778
0.0029	40.0	8600	5.9302	0.5778
0.0002	41.0	8815	5.9565	0.5778
0.0046	42.0	9030	6.4259	0.5778
0.0003	43.0	9245	6.2149	0.6222
0.0	44.0	9460	6.6058	0.6
0.0	45.0	9675	6.5823	0.6
0.0	46.0	9890	6.7170	0.6
0.0	47.0	10105	6.8829	0.6
0.0	48.0	10320	7.0021	0.6
0.0	49.0	10535	7.1301	0.6
0.0	50.0	10750	7.1718	0.6

Framework versions

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