hushem_5x_deit_small_sgd_001_fold5

This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.0219
• Accuracy: 0.5854

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
1.5389	1.0	28	1.4343	0.2439
1.3921	2.0	56	1.3847	0.2683
1.3749	3.0	84	1.3585	0.3659
1.3126	4.0	112	1.3409	0.3659
1.3117	5.0	140	1.3251	0.3659
1.3121	6.0	168	1.3119	0.3415
1.2628	7.0	196	1.2980	0.3415
1.2308	8.0	224	1.2843	0.3659
1.2428	9.0	252	1.2711	0.4146
1.1961	10.0	280	1.2591	0.4146
1.1795	11.0	308	1.2486	0.3902
1.1594	12.0	336	1.2381	0.3902
1.1371	13.0	364	1.2260	0.3902
1.1217	14.0	392	1.2140	0.4390
1.0975	15.0	420	1.2018	0.4634
1.1139	16.0	448	1.1910	0.4878
1.0797	17.0	476	1.1802	0.4878
1.0813	18.0	504	1.1682	0.4878
1.0619	19.0	532	1.1572	0.4878
1.0398	20.0	560	1.1467	0.5122
1.0215	21.0	588	1.1362	0.5122
1.0014	22.0	616	1.1280	0.5366
1.0047	23.0	644	1.1216	0.5610
0.9823	24.0	672	1.1144	0.5610
0.9814	25.0	700	1.1058	0.5610
0.9822	26.0	728	1.0976	0.5610
0.9448	27.0	756	1.0916	0.5366
0.9805	28.0	784	1.0839	0.5366
0.9187	29.0	812	1.0780	0.5366
0.9659	30.0	840	1.0725	0.5366
0.9135	31.0	868	1.0663	0.5610
0.889	32.0	896	1.0628	0.5610
0.9089	33.0	924	1.0587	0.5610
0.9062	34.0	952	1.0524	0.5610
0.9029	35.0	980	1.0479	0.5610
0.8924	36.0	1008	1.0439	0.5854
0.8694	37.0	1036	1.0402	0.5854
0.8578	38.0	1064	1.0365	0.5610
0.8992	39.0	1092	1.0340	0.5854
0.8586	40.0	1120	1.0317	0.5854
0.8737	41.0	1148	1.0296	0.5854
0.8517	42.0	1176	1.0278	0.5854
0.8537	43.0	1204	1.0257	0.5854
0.8642	44.0	1232	1.0243	0.5854
0.871	45.0	1260	1.0234	0.5854
0.8594	46.0	1288	1.0226	0.5854
0.8675	47.0	1316	1.0221	0.5854
0.874	48.0	1344	1.0219	0.5854
0.8459	49.0	1372	1.0219	0.5854
0.8538	50.0	1400	1.0219	0.5854

Framework versions

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