hushem_40x_deit_tiny_rms_001_fold5

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: 4.1758
• Accuracy: 0.7317

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
1.1377	1.0	220	1.4706	0.3171
0.8838	2.0	440	1.1169	0.4634
0.7297	3.0	660	0.9759	0.6098
0.6075	4.0	880	0.6205	0.7317
0.4922	5.0	1100	1.0787	0.6098
0.4263	6.0	1320	0.7206	0.7561
0.1851	7.0	1540	0.8644	0.7317
0.1792	8.0	1760	0.9149	0.7073
0.1954	9.0	1980	0.9202	0.7561
0.1313	10.0	2200	0.8113	0.8049
0.0886	11.0	2420	1.2174	0.7073
0.0848	12.0	2640	1.1156	0.7805
0.1044	13.0	2860	1.2101	0.7073
0.1348	14.0	3080	1.1458	0.7317
0.0626	15.0	3300	1.0690	0.7317
0.0309	16.0	3520	1.3430	0.7073
0.0473	17.0	3740	1.3747	0.7317
0.1091	18.0	3960	1.0321	0.7805
0.1069	19.0	4180	1.7755	0.7073
0.0244	20.0	4400	1.5213	0.7317
0.0331	21.0	4620	1.3935	0.7805
0.1167	22.0	4840	1.2913	0.7561
0.0052	23.0	5060	0.9792	0.8049
0.0198	24.0	5280	0.8211	0.8780
0.0358	25.0	5500	1.1277	0.8049
0.0008	26.0	5720	1.4574	0.7073
0.0001	27.0	5940	1.0665	0.7561
0.0001	28.0	6160	1.1272	0.7805
0.0006	29.0	6380	1.2564	0.7073
0.0011	30.0	6600	1.4453	0.7073
0.0002	31.0	6820	1.9210	0.7805
0.0108	32.0	7040	2.0987	0.7317
0.0	33.0	7260	2.5706	0.7317
0.0	34.0	7480	2.5154	0.7317
0.0	35.0	7700	2.7600	0.7317
0.0	36.0	7920	2.9535	0.7317
0.0	37.0	8140	3.1263	0.7317
0.0	38.0	8360	3.3175	0.7317
0.0	39.0	8580	3.4937	0.7317
0.0	40.0	8800	3.6752	0.7317
0.0	41.0	9020	3.8325	0.7317
0.0	42.0	9240	3.9616	0.7317
0.0	43.0	9460	4.0429	0.7317
0.0	44.0	9680	4.1055	0.7317
0.0	45.0	9900	4.1340	0.7317
0.0	46.0	10120	4.1489	0.7317
0.0	47.0	10340	4.1566	0.7317
0.0	48.0	10560	4.1688	0.7317
0.0	49.0	10780	4.1743	0.7317
0.0	50.0	11000	4.1758	0.7317

Framework versions

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