hushem_1x_deit_small_sgd_001_fold4

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: 1.3002
• Accuracy: 0.3333

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	1.4504	0.2857
1.4996	2.0	12	1.4256	0.2619
1.4996	3.0	18	1.4065	0.3095
1.4661	4.0	24	1.3909	0.3333
1.4137	5.0	30	1.3815	0.3333
1.4137	6.0	36	1.3736	0.3810
1.3923	7.0	42	1.3662	0.3571
1.3923	8.0	48	1.3602	0.3095
1.3511	9.0	54	1.3552	0.3333
1.3471	10.0	60	1.3505	0.3333
1.3471	11.0	66	1.3464	0.3333
1.3212	12.0	72	1.3425	0.3333
1.3212	13.0	78	1.3391	0.3333
1.3151	14.0	84	1.3358	0.3333
1.2949	15.0	90	1.3328	0.3333
1.2949	16.0	96	1.3296	0.3333
1.282	17.0	102	1.3270	0.3333
1.282	18.0	108	1.3243	0.3333
1.2637	19.0	114	1.3223	0.3333
1.2828	20.0	120	1.3203	0.3333
1.2828	21.0	126	1.3182	0.3333
1.2384	22.0	132	1.3165	0.3333
1.2384	23.0	138	1.3149	0.3333
1.2419	24.0	144	1.3133	0.3333
1.2404	25.0	150	1.3117	0.3571
1.2404	26.0	156	1.3102	0.3571
1.2294	27.0	162	1.3091	0.3571
1.2294	28.0	168	1.3080	0.3571
1.2327	29.0	174	1.3070	0.3571
1.2115	30.0	180	1.3061	0.3571
1.2115	31.0	186	1.3052	0.3333
1.2091	32.0	192	1.3043	0.3333
1.2091	33.0	198	1.3036	0.3333
1.2111	34.0	204	1.3028	0.3333
1.2001	35.0	210	1.3022	0.3333
1.2001	36.0	216	1.3016	0.3333
1.2048	37.0	222	1.3012	0.3333
1.2048	38.0	228	1.3009	0.3333
1.1981	39.0	234	1.3006	0.3333
1.1973	40.0	240	1.3004	0.3333
1.1973	41.0	246	1.3003	0.3333
1.2009	42.0	252	1.3002	0.3333
1.2009	43.0	258	1.3002	0.3333
1.1848	44.0	264	1.3002	0.3333
1.2	45.0	270	1.3002	0.3333
1.2	46.0	276	1.3002	0.3333
1.2026	47.0	282	1.3002	0.3333
1.2026	48.0	288	1.3002	0.3333
1.1883	49.0	294	1.3002	0.3333
1.2097	50.0	300	1.3002	0.3333

Framework versions

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