hushem_5x_deit_base_adamax_00001_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: 1.1362
• Accuracy: 0.6444

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
1.2172	1.0	27	1.2445	0.4
0.8523	2.0	54	1.0947	0.4667
0.5686	3.0	81	1.0185	0.5778
0.4004	4.0	108	0.9768	0.5778
0.2464	5.0	135	0.9587	0.5778
0.1691	6.0	162	0.8886	0.6222
0.1191	7.0	189	0.9107	0.6
0.0619	8.0	216	0.8951	0.6444
0.0336	9.0	243	0.9574	0.6
0.0186	10.0	270	0.9860	0.5778
0.0125	11.0	297	0.9869	0.6
0.0084	12.0	324	1.0113	0.6
0.0076	13.0	351	0.9936	0.6
0.0057	14.0	378	1.0048	0.6
0.0052	15.0	405	1.0120	0.6
0.0044	16.0	432	1.0086	0.6222
0.0038	17.0	459	1.0209	0.6222
0.0036	18.0	486	1.0433	0.6222
0.0032	19.0	513	1.0446	0.6444
0.0029	20.0	540	1.0517	0.6444
0.0025	21.0	567	1.0577	0.6444
0.0023	22.0	594	1.0550	0.6444
0.0022	23.0	621	1.0799	0.6444
0.002	24.0	648	1.0753	0.6444
0.002	25.0	675	1.0830	0.6444
0.002	26.0	702	1.0841	0.6444
0.0018	27.0	729	1.0884	0.6444
0.0017	28.0	756	1.0904	0.6444
0.0017	29.0	783	1.1034	0.6444
0.0016	30.0	810	1.1073	0.6444
0.0015	31.0	837	1.1021	0.6444
0.0015	32.0	864	1.1089	0.6444
0.0014	33.0	891	1.1157	0.6444
0.0014	34.0	918	1.1170	0.6444
0.0013	35.0	945	1.1193	0.6444
0.0012	36.0	972	1.1215	0.6444
0.0013	37.0	999	1.1225	0.6444
0.0012	38.0	1026	1.1226	0.6444
0.0012	39.0	1053	1.1299	0.6444
0.0011	40.0	1080	1.1301	0.6444
0.0012	41.0	1107	1.1312	0.6444
0.0011	42.0	1134	1.1308	0.6444
0.0012	43.0	1161	1.1360	0.6444
0.001	44.0	1188	1.1351	0.6444
0.0011	45.0	1215	1.1359	0.6444
0.0011	46.0	1242	1.1364	0.6444
0.001	47.0	1269	1.1364	0.6444
0.0011	48.0	1296	1.1362	0.6444
0.0011	49.0	1323	1.1362	0.6444
0.0011	50.0	1350	1.1362	0.6444

Framework versions

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