hushem_40x_deit_tiny_rms_00001_fold1

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: 1.4542
• Accuracy: 0.8444

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
0.0572	1.0	215	0.5452	0.8222
0.0213	2.0	430	0.8514	0.8222
0.0002	3.0	645	1.1716	0.7778
0.0001	4.0	860	1.1956	0.8
0.0	5.0	1075	1.3312	0.7778
0.0	6.0	1290	1.3747	0.8
0.0	7.0	1505	1.5420	0.7778
0.0	8.0	1720	1.5431	0.7778
0.0	9.0	1935	1.6767	0.7778
0.0	10.0	2150	1.6620	0.8
0.0141	11.0	2365	0.7752	0.8667
0.0	12.0	2580	1.2616	0.7556
0.0	13.0	2795	1.1161	0.8667
0.0	14.0	3010	1.1254	0.8444
0.0	15.0	3225	1.1188	0.8889
0.0	16.0	3440	1.1820	0.8889
0.0	17.0	3655	1.2564	0.8889
0.0	18.0	3870	1.3559	0.8889
0.0	19.0	4085	1.4292	0.8667
0.0	20.0	4300	1.5164	0.8667
0.0	21.0	4515	1.5191	0.8667
0.0	22.0	4730	1.4544	0.8667
0.0	23.0	4945	1.4836	0.8667
0.0	24.0	5160	1.5747	0.8222
0.0	25.0	5375	1.5707	0.8222
0.0	26.0	5590	1.5222	0.8222
0.0	27.0	5805	1.4844	0.8667
0.0	28.0	6020	1.4898	0.8667
0.0	29.0	6235	1.5381	0.8444
0.0	30.0	6450	1.5320	0.8222
0.0	31.0	6665	1.5518	0.8222
0.0	32.0	6880	1.4681	0.8667
0.0	33.0	7095	1.5245	0.8444
0.0	34.0	7310	1.4517	0.8667
0.0	35.0	7525	1.4519	0.8667
0.0	36.0	7740	1.4734	0.8667
0.0	37.0	7955	1.5324	0.8444
0.0	38.0	8170	1.4772	0.8444
0.0	39.0	8385	1.4506	0.8444
0.0	40.0	8600	1.4509	0.8444
0.0	41.0	8815	1.5306	0.8444
0.0	42.0	9030	1.4735	0.8444
0.0	43.0	9245	1.4585	0.8444
0.0	44.0	9460	1.4843	0.8444
0.0	45.0	9675	1.4519	0.8444
0.0	46.0	9890	1.4772	0.8444
0.0	47.0	10105	1.4373	0.8444
0.0	48.0	10320	1.4662	0.8444
0.0	49.0	10535	1.4530	0.8444
0.0	50.0	10750	1.4542	0.8444

Framework versions

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