hushem_1x_deit_tiny_rms_001_fold2

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: 1.2949
• Accuracy: 0.3556

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	6.6312	0.2667
4.1013	2.0	12	2.1471	0.2444
4.1013	3.0	18	1.7992	0.2444
1.7936	4.0	24	1.5377	0.2667
1.5908	5.0	30	1.6029	0.2444
1.5908	6.0	36	1.5728	0.2444
1.533	7.0	42	1.6272	0.2444
1.533	8.0	48	1.5192	0.2667
1.4887	9.0	54	1.4382	0.2444
1.4288	10.0	60	1.4387	0.2444
1.4288	11.0	66	1.4770	0.2667
1.422	12.0	72	1.3624	0.2444
1.422	13.0	78	1.4332	0.2667
1.4231	14.0	84	1.4892	0.2444
1.385	15.0	90	1.3102	0.4222
1.385	16.0	96	1.3352	0.3333
1.4799	17.0	102	1.6140	0.3111
1.4799	18.0	108	1.4774	0.2444
1.4126	19.0	114	1.3130	0.3333
1.3511	20.0	120	1.2400	0.4222
1.3511	21.0	126	1.5468	0.2667
1.412	22.0	132	1.4525	0.2667
1.412	23.0	138	1.2484	0.3778
1.3184	24.0	144	1.5741	0.2444
1.3429	25.0	150	1.3487	0.4444
1.3429	26.0	156	1.3203	0.3111
1.2824	27.0	162	1.2257	0.4222
1.2824	28.0	168	1.3520	0.2222
1.2504	29.0	174	1.1717	0.4667
1.235	30.0	180	1.2327	0.3778
1.235	31.0	186	1.3371	0.4
1.2286	32.0	192	1.3224	0.2889
1.2286	33.0	198	1.2295	0.3778
1.168	34.0	204	1.2716	0.3111
1.2345	35.0	210	1.2743	0.3111
1.2345	36.0	216	1.3964	0.3778
1.2057	37.0	222	1.3905	0.3556
1.2057	38.0	228	1.2908	0.3778
1.1197	39.0	234	1.2888	0.3556
1.1518	40.0	240	1.2704	0.4
1.1518	41.0	246	1.3067	0.3556
1.1311	42.0	252	1.2949	0.3556
1.1311	43.0	258	1.2949	0.3556
1.109	44.0	264	1.2949	0.3556
1.1464	45.0	270	1.2949	0.3556
1.1464	46.0	276	1.2949	0.3556
1.0982	47.0	282	1.2949	0.3556
1.0982	48.0	288	1.2949	0.3556
1.1635	49.0	294	1.2949	0.3556
1.1115	50.0	300	1.2949	0.3556

Framework versions

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