hushem_1x_deit_tiny_rms_0001_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: 3.1133
• Accuracy: 0.5556

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: 0.0001
• 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	1.9323	0.2444
2.0865	2.0	12	1.4427	0.2444
2.0865	3.0	18	1.4293	0.2444
1.4431	4.0	24	1.3952	0.4667
1.4003	5.0	30	1.2967	0.4
1.4003	6.0	36	1.4719	0.2444
1.3496	7.0	42	1.3224	0.3556
1.3496	8.0	48	1.4673	0.3778
1.2064	9.0	54	1.4551	0.2667
1.1859	10.0	60	1.3687	0.3111
1.1859	11.0	66	1.2313	0.4444
1.0817	12.0	72	1.1514	0.4444
1.0817	13.0	78	1.1701	0.4444
1.0144	14.0	84	1.2204	0.4222
0.8578	15.0	90	1.1603	0.4889
0.8578	16.0	96	1.0987	0.5111
0.8063	17.0	102	0.9277	0.5111
0.8063	18.0	108	1.2038	0.5333
0.601	19.0	114	0.9886	0.6
0.465	20.0	120	1.5667	0.5111
0.465	21.0	126	1.8238	0.4889
0.2956	22.0	132	1.6043	0.4222
0.2956	23.0	138	1.2746	0.4889
0.3513	24.0	144	1.6389	0.5556
0.2137	25.0	150	1.6350	0.4889
0.2137	26.0	156	1.5926	0.4667
0.191	27.0	162	1.8516	0.4889
0.191	28.0	168	2.3628	0.4889
0.0581	29.0	174	2.3998	0.4889
0.0517	30.0	180	2.3913	0.5333
0.0517	31.0	186	2.7108	0.5556
0.005	32.0	192	2.8104	0.5556
0.005	33.0	198	2.8829	0.5556
0.0008	34.0	204	2.9326	0.5333
0.0006	35.0	210	2.9793	0.5556
0.0006	36.0	216	3.0150	0.5556
0.0005	37.0	222	3.0520	0.5556
0.0005	38.0	228	3.0772	0.5556
0.0004	39.0	234	3.0948	0.5556
0.0004	40.0	240	3.1038	0.5556
0.0004	41.0	246	3.1116	0.5556
0.0004	42.0	252	3.1133	0.5556
0.0004	43.0	258	3.1133	0.5556
0.0004	44.0	264	3.1133	0.5556
0.0004	45.0	270	3.1133	0.5556
0.0004	46.0	276	3.1133	0.5556
0.0004	47.0	282	3.1133	0.5556
0.0004	48.0	288	3.1133	0.5556
0.0004	49.0	294	3.1133	0.5556
0.0004	50.0	300	3.1133	0.5556

Framework versions

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