hushem_1x_deit_tiny_adamax_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: 1.1341
• Accuracy: 0.4222

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
No log	1.0	6	1.4260	0.2
1.446	2.0	12	1.3794	0.2889
1.446	3.0	18	1.3570	0.3556
1.184	4.0	24	1.3382	0.3111
1.0671	5.0	30	1.3283	0.3111
1.0671	6.0	36	1.3144	0.2889
0.9249	7.0	42	1.2898	0.3333
0.9249	8.0	48	1.2748	0.3556
0.8443	9.0	54	1.2692	0.3333
0.7477	10.0	60	1.2518	0.3778
0.7477	11.0	66	1.2338	0.4
0.662	12.0	72	1.2193	0.3778
0.662	13.0	78	1.2195	0.4
0.622	14.0	84	1.2039	0.3778
0.5154	15.0	90	1.1949	0.4
0.5154	16.0	96	1.1879	0.4
0.4537	17.0	102	1.1810	0.4
0.4537	18.0	108	1.1670	0.4
0.3859	19.0	114	1.1628	0.4
0.3586	20.0	120	1.1721	0.4
0.3586	21.0	126	1.1698	0.4222
0.3151	22.0	132	1.1603	0.4
0.3151	23.0	138	1.1584	0.4222
0.2881	24.0	144	1.1519	0.4222
0.2498	25.0	150	1.1515	0.4222
0.2498	26.0	156	1.1445	0.4222
0.232	27.0	162	1.1430	0.4222
0.232	28.0	168	1.1452	0.4222
0.2183	29.0	174	1.1406	0.4222
0.1798	30.0	180	1.1348	0.4222
0.1798	31.0	186	1.1304	0.4222
0.1811	32.0	192	1.1281	0.4222
0.1811	33.0	198	1.1317	0.4222
0.1748	34.0	204	1.1302	0.4222
0.1492	35.0	210	1.1303	0.4222
0.1492	36.0	216	1.1319	0.4222
0.1477	37.0	222	1.1328	0.4222
0.1477	38.0	228	1.1366	0.4222
0.1357	39.0	234	1.1362	0.4222
0.1379	40.0	240	1.1351	0.4222
0.1379	41.0	246	1.1344	0.4222
0.1325	42.0	252	1.1341	0.4222
0.1325	43.0	258	1.1341	0.4222
0.1377	44.0	264	1.1341	0.4222
0.1332	45.0	270	1.1341	0.4222
0.1332	46.0	276	1.1341	0.4222
0.1323	47.0	282	1.1341	0.4222
0.1323	48.0	288	1.1341	0.4222
0.1276	49.0	294	1.1341	0.4222
0.1376	50.0	300	1.1341	0.4222

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1