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: 1.3640
• Accuracy: 0.55

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1309	1.0	20	1.3481	0.4938
1.0746	2.0	40	1.3706	0.475
1.0367	3.0	60	1.3161	0.5375
0.9814	4.0	80	1.3837	0.45
0.886	5.0	100	1.3633	0.4875
0.8096	6.0	120	1.3045	0.5125
0.7669	7.0	140	1.3903	0.4938
0.708	8.0	160	1.2867	0.5125
0.6265	9.0	180	1.2244	0.5625
0.6191	10.0	200	1.3461	0.525
0.5598	11.0	220	1.3266	0.5625
0.4667	12.0	240	1.3050	0.5563
0.4613	13.0	260	1.3329	0.5375
0.4268	14.0	280	1.4020	0.5312
0.4256	15.0	300	1.3770	0.5188
0.3727	16.0	320	1.3655	0.5188
0.316	17.0	340	1.3642	0.5188
0.3223	18.0	360	1.2535	0.5938
0.3064	19.0	380	1.4173	0.4875
0.2866	20.0	400	1.3343	0.5625
0.2781	21.0	420	1.5072	0.4813
0.3027	22.0	440	1.5067	0.5125
0.26	23.0	460	1.4456	0.5687
0.2156	24.0	480	1.4825	0.525
0.1908	25.0	500	1.5369	0.5375
0.213	26.0	520	1.5397	0.5188
0.241	27.0	540	1.4804	0.5125
0.1974	28.0	560	1.5786	0.5062
0.225	29.0	580	1.4677	0.5375
0.2459	30.0	600	1.5392	0.5312
0.2146	31.0	620	1.6734	0.4625
0.1891	32.0	640	1.5012	0.55
0.2231	33.0	660	1.6265	0.5
0.1903	34.0	680	1.5405	0.5312
0.1852	35.0	700	1.6295	0.5
0.1768	36.0	720	1.5758	0.5375
0.1486	37.0	740	1.6176	0.5188
0.1814	38.0	760	1.5107	0.5375
0.1642	39.0	780	1.5315	0.55
0.1822	40.0	800	1.6309	0.525
0.1819	41.0	820	1.7033	0.4938
0.1326	42.0	840	1.6107	0.5437
0.1452	43.0	860	1.6219	0.55
0.128	44.0	880	1.4348	0.5813
0.1103	45.0	900	1.6185	0.5687
0.1386	46.0	920	1.5848	0.5312
0.1021	47.0	940	1.6036	0.5563
0.1414	48.0	960	1.5455	0.575
0.1989	49.0	980	1.5955	0.525
0.1458	50.0	1000	1.5511	0.55

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1