hushem_5x_deit_tiny_adamax_001_fold2

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: 5.5945
• Accuracy: 0.4667

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
1.3843	1.0	27	1.6047	0.2667
1.3126	2.0	54	1.5540	0.2889
1.242	3.0	81	1.7659	0.2889
1.2292	4.0	108	1.5553	0.2667
1.1861	5.0	135	1.3959	0.2889
0.9837	6.0	162	1.3509	0.4889
0.9151	7.0	189	1.3189	0.3778
0.8428	8.0	216	1.2233	0.4444
0.6366	9.0	243	1.6296	0.5556
0.6056	10.0	270	2.1707	0.5333
0.6193	11.0	297	2.2637	0.5333
0.419	12.0	324	2.6815	0.4444
0.4931	13.0	351	1.6615	0.4444
0.4945	14.0	378	2.3385	0.4444
0.4035	15.0	405	2.2795	0.4889
0.2803	16.0	432	2.9683	0.5556
0.3104	17.0	459	3.6455	0.4222
0.2085	18.0	486	3.8509	0.5778
0.3092	19.0	513	2.7832	0.5333
0.2911	20.0	540	3.5970	0.4667
0.1508	21.0	567	2.8494	0.4
0.1872	22.0	594	3.4606	0.5111
0.1634	23.0	621	3.9940	0.5778
0.0282	24.0	648	4.2310	0.4667
0.0886	25.0	675	3.4801	0.4889
0.1057	26.0	702	3.6632	0.4222
0.0893	27.0	729	3.7806	0.5556
0.0108	28.0	756	4.5566	0.5556
0.1095	29.0	783	4.2488	0.5556
0.0164	30.0	810	4.4510	0.5556
0.0017	31.0	837	4.8182	0.5333
0.0006	32.0	864	5.4595	0.5333
0.0034	33.0	891	5.3273	0.5556
0.0036	34.0	918	4.8258	0.4889
0.0002	35.0	945	5.5297	0.5111
0.0007	36.0	972	5.4553	0.4667
0.0081	37.0	999	5.5834	0.5333
0.0001	38.0	1026	5.4059	0.5333
0.0001	39.0	1053	5.7543	0.4667
0.0	40.0	1080	5.5912	0.4667
0.0	41.0	1107	5.5774	0.4667
0.0	42.0	1134	5.5838	0.4667
0.0	43.0	1161	5.5882	0.4667
0.0	44.0	1188	5.5908	0.4667
0.0	45.0	1215	5.5928	0.4667
0.0	46.0	1242	5.5933	0.4667
0.0	47.0	1269	5.5942	0.4667
0.0	48.0	1296	5.5945	0.4667
0.0	49.0	1323	5.5945	0.4667
0.0	50.0	1350	5.5945	0.4667

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0