hushem_40x_deit_base_sgd_001_fold1

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: 0.7430
• Accuracy: 0.7556

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
1.2392	1.0	215	1.3895	0.2667
1.1003	2.0	430	1.3294	0.3333
1.0196	3.0	645	1.2624	0.4444
0.8639	4.0	860	1.1946	0.4889
0.731	5.0	1075	1.1313	0.5111
0.6646	6.0	1290	1.0718	0.5556
0.545	7.0	1505	1.0254	0.6
0.4701	8.0	1720	0.9800	0.6444
0.4065	9.0	1935	0.9495	0.6222
0.3851	10.0	2150	0.9148	0.6667
0.3271	11.0	2365	0.8947	0.6667
0.2977	12.0	2580	0.8732	0.6889
0.2671	13.0	2795	0.8416	0.7111
0.2428	14.0	3010	0.8450	0.6889
0.2387	15.0	3225	0.8270	0.7111
0.1988	16.0	3440	0.8218	0.7111
0.1804	17.0	3655	0.8107	0.7333
0.1681	18.0	3870	0.8058	0.7333
0.1475	19.0	4085	0.7968	0.7333
0.1494	20.0	4300	0.7851	0.7556
0.1288	21.0	4515	0.7807	0.7556
0.1265	22.0	4730	0.7751	0.7556
0.1136	23.0	4945	0.7744	0.7556
0.094	24.0	5160	0.7654	0.7556
0.0987	25.0	5375	0.7661	0.7556
0.096	26.0	5590	0.7527	0.7556
0.084	27.0	5805	0.7535	0.7556
0.069	28.0	6020	0.7589	0.7556
0.0764	29.0	6235	0.7612	0.7556
0.067	30.0	6450	0.7558	0.7556
0.0458	31.0	6665	0.7531	0.7333
0.0687	32.0	6880	0.7463	0.7556
0.0414	33.0	7095	0.7445	0.7556
0.0522	34.0	7310	0.7378	0.7556
0.0521	35.0	7525	0.7477	0.7556
0.0458	36.0	7740	0.7370	0.7556
0.0586	37.0	7955	0.7425	0.7556
0.0551	38.0	8170	0.7441	0.7556
0.0389	39.0	8385	0.7437	0.7556
0.0335	40.0	8600	0.7446	0.7556
0.0337	41.0	8815	0.7439	0.7556
0.0431	42.0	9030	0.7421	0.7556
0.0392	43.0	9245	0.7439	0.7556
0.03	44.0	9460	0.7447	0.7556
0.0402	45.0	9675	0.7426	0.7556
0.0313	46.0	9890	0.7416	0.7556
0.0341	47.0	10105	0.7428	0.7556
0.0375	48.0	10320	0.7420	0.7556
0.0432	49.0	10535	0.7428	0.7556
0.0389	50.0	10750	0.7430	0.7556

Framework versions

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