image_classification

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.6838
• Accuracy: 0.525

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: 5e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	10	1.3274	0.5125
No log	2.0	20	1.3119	0.5188
No log	3.0	30	1.3825	0.4625
No log	4.0	40	1.2916	0.5312
No log	5.0	50	1.2821	0.525
No log	6.0	60	1.2407	0.525
No log	7.0	70	1.3288	0.5125
No log	8.0	80	1.2818	0.525
No log	9.0	90	1.3710	0.4875
No log	10.0	100	1.3298	0.5312
No log	11.0	110	1.3539	0.475
No log	12.0	120	1.4498	0.4688
No log	13.0	130	1.5422	0.4437
No log	14.0	140	1.4870	0.4625
No log	15.0	150	1.4354	0.525
No log	16.0	160	1.4286	0.4938
No log	17.0	170	1.5332	0.4437
No log	18.0	180	1.4164	0.5188
No log	19.0	190	1.5024	0.4625
No log	20.0	200	1.4730	0.5125
No log	21.0	210	1.3083	0.55
No log	22.0	220	1.4468	0.525
No log	23.0	230	1.3198	0.525
No log	24.0	240	1.3530	0.5563
No log	25.0	250	1.4821	0.4938
No log	26.0	260	1.3475	0.5437
No log	27.0	270	1.5152	0.4875
No log	28.0	280	1.4290	0.55
No log	29.0	290	1.5505	0.5
No log	30.0	300	1.5796	0.5062
No log	31.0	310	1.5988	0.5125
No log	32.0	320	1.6272	0.4875
No log	33.0	330	1.4324	0.5437
No log	34.0	340	1.5245	0.5062
No log	35.0	350	1.7228	0.45
No log	36.0	360	1.4861	0.525
No log	37.0	370	1.5317	0.5312
No log	38.0	380	1.7776	0.475
No log	39.0	390	1.5386	0.5563
No log	40.0	400	1.7608	0.475
No log	41.0	410	1.5469	0.55
No log	42.0	420	1.6919	0.4625
No log	43.0	430	1.5814	0.525
No log	44.0	440	1.5877	0.5125
No log	45.0	450	1.6370	0.5188
No log	46.0	460	1.7375	0.5188
No log	47.0	470	1.7004	0.5
No log	48.0	480	1.6309	0.4938
No log	49.0	490	1.5931	0.5437
0.2996	50.0	500	1.7687	0.5062
0.2996	51.0	510	1.5321	0.5188
0.2996	52.0	520	1.8099	0.4688
0.2996	53.0	530	1.5138	0.575
0.2996	54.0	540	1.7569	0.4688
0.2996	55.0	550	1.7451	0.4813
0.2996	56.0	560	1.6871	0.5125
0.2996	57.0	570	1.6471	0.525
0.2996	58.0	580	1.6966	0.525
0.2996	59.0	590	1.7714	0.5
0.2996	60.0	600	1.4985	0.5938
0.2996	61.0	610	1.9804	0.4313
0.2996	62.0	620	1.6116	0.5375
0.2996	63.0	630	1.6056	0.525
0.2996	64.0	640	1.6115	0.5062
0.2996	65.0	650	1.9694	0.4625
0.2996	66.0	660	1.6338	0.5563
0.2996	67.0	670	1.4823	0.5938
0.2996	68.0	680	1.9253	0.5
0.2996	69.0	690	1.9015	0.4813
0.2996	70.0	700	1.5446	0.5687
0.2996	71.0	710	1.9302	0.4938
0.2996	72.0	720	1.6973	0.5375
0.2996	73.0	730	1.8271	0.5
0.2996	74.0	740	1.7559	0.5188
0.2996	75.0	750	1.8127	0.5312
0.2996	76.0	760	1.8096	0.4938
0.2996	77.0	770	1.8460	0.5062
0.2996	78.0	780	1.8853	0.4813
0.2996	79.0	790	1.7706	0.5125
0.2996	80.0	800	1.8129	0.5312
0.2996	81.0	810	1.9488	0.4688
0.2996	82.0	820	1.8817	0.4813
0.2996	83.0	830	1.6759	0.5563
0.2996	84.0	840	1.6884	0.5
0.2996	85.0	850	1.8146	0.4875
0.2996	86.0	860	1.6610	0.55
0.2996	87.0	870	1.8811	0.475
0.2996	88.0	880	1.8964	0.5062
0.2996	89.0	890	1.6848	0.5437
0.2996	90.0	900	1.8642	0.4938
0.2996	91.0	910	1.8819	0.5125
0.2996	92.0	920	1.9193	0.4875
0.2996	93.0	930	1.8110	0.5
0.2996	94.0	940	1.9086	0.4813
0.2996	95.0	950	1.8895	0.4625
0.2996	96.0	960	1.7554	0.5312
0.2996	97.0	970	1.8978	0.5188
0.2996	98.0	980	1.9791	0.4875
0.2996	99.0	990	1.7030	0.5687
0.0883	100.0	1000	1.8398	0.4813

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3