image_classification

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

• Loss: 1.5394
• Accuracy: 0.4813

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
8.2876	1.0	10	2.0733	0.1375
8.2701	2.0	20	2.0678	0.15
8.2385	3.0	30	2.0564	0.1875
8.1938	4.0	40	2.0484	0.2188
8.1243	5.0	50	2.0263	0.2437
8.043	6.0	60	2.0065	0.2812
7.9327	7.0	70	1.9940	0.275
7.7842	8.0	80	1.9588	0.3438
7.6389	9.0	90	1.9299	0.3125
7.4825	10.0	100	1.8830	0.4
7.3337	11.0	110	1.8519	0.35
7.1512	12.0	120	1.8171	0.4188
7.0169	13.0	130	1.7624	0.4188
6.8618	14.0	140	1.7341	0.45
6.7244	15.0	150	1.6903	0.45
6.5857	16.0	160	1.6709	0.4688
6.4774	17.0	170	1.6624	0.425
6.3616	18.0	180	1.6314	0.4437
6.2635	19.0	190	1.6173	0.4437
6.1831	20.0	200	1.5929	0.4938
6.1224	21.0	210	1.5841	0.45
6.0711	22.0	220	1.5622	0.4625
5.9769	23.0	230	1.5617	0.5062
5.9176	24.0	240	1.5491	0.4813
5.8776	25.0	250	1.5262	0.5687
5.8347	26.0	260	1.5287	0.4875
5.781	27.0	270	1.5284	0.4625
5.7451	28.0	280	1.5018	0.4875
5.6745	29.0	290	1.5057	0.4875
5.6253	30.0	300	1.5090	0.4938
5.6111	31.0	310	1.5275	0.4688
5.5742	32.0	320	1.5008	0.525
5.5516	33.0	330	1.4795	0.5188
5.4796	34.0	340	1.4834	0.5062
5.4958	35.0	350	1.4916	0.5125
5.4824	36.0	360	1.4925	0.4938
5.4659	37.0	370	1.4847	0.5062
5.4715	38.0	380	1.4670	0.5
5.4735	39.0	390	1.4733	0.525
5.4789	40.0	400	1.4881	0.4813

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Tokenizers 0.21.0