hushem_1x_deit_small_adamax_001_fold1

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: 2.0215
• Accuracy: 0.4667

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
No log	1.0	6	2.2870	0.2444
2.1668	2.0	12	1.4669	0.2444
2.1668	3.0	18	1.4980	0.2444
1.4102	4.0	24	1.4751	0.2444
1.4394	5.0	30	1.4286	0.2444
1.4394	6.0	36	1.6019	0.2444
1.3171	7.0	42	1.7291	0.2222
1.3171	8.0	48	1.5314	0.3556
1.2906	9.0	54	1.7281	0.2667
1.2151	10.0	60	1.6012	0.2444
1.2151	11.0	66	1.5621	0.4444
1.1016	12.0	72	1.5069	0.2
1.1016	13.0	78	1.5452	0.4222
1.1085	14.0	84	1.5457	0.2889
0.9838	15.0	90	1.7131	0.4
0.9838	16.0	96	1.9947	0.2889
1.003	17.0	102	1.7538	0.4222
1.003	18.0	108	1.3632	0.4444
0.846	19.0	114	1.7633	0.4
0.7432	20.0	120	1.5259	0.4222
0.7432	21.0	126	1.6982	0.4
0.8111	22.0	132	1.4722	0.4
0.8111	23.0	138	1.5772	0.4222
0.6268	24.0	144	1.6621	0.4222
0.5956	25.0	150	2.2283	0.4
0.5956	26.0	156	1.5965	0.4667
0.863	27.0	162	2.0067	0.4
0.863	28.0	168	2.2609	0.3778
0.575	29.0	174	1.7339	0.4222
0.3505	30.0	180	1.6059	0.3778
0.3505	31.0	186	1.7578	0.4444
0.3884	32.0	192	1.8785	0.4444
0.3884	33.0	198	1.5952	0.4222
0.3742	34.0	204	1.9834	0.4444
0.3113	35.0	210	1.8134	0.4222
0.3113	36.0	216	2.1491	0.4
0.4478	37.0	222	1.9419	0.4667
0.4478	38.0	228	1.8426	0.4444
0.1746	39.0	234	1.9349	0.4222
0.1737	40.0	240	2.0085	0.4667
0.1737	41.0	246	2.0238	0.4667
0.1448	42.0	252	2.0215	0.4667
0.1448	43.0	258	2.0215	0.4667
0.1495	44.0	264	2.0215	0.4667
0.1326	45.0	270	2.0215	0.4667
0.1326	46.0	276	2.0215	0.4667
0.1487	47.0	282	2.0215	0.4667
0.1487	48.0	288	2.0215	0.4667
0.1112	49.0	294	2.0215	0.4667
0.1501	50.0	300	2.0215	0.4667

Framework versions

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