belajar_huggingface

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: 1.3425
• Accuracy: 0.5
• Precision: 0.5111
• Recall: 0.5
• F1: 0.4990

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: 0.00025
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	40	1.7396	0.2938	0.2269	0.2938	0.2006
No log	2.0	80	1.7746	0.3812	0.4438	0.3812	0.3612
No log	3.0	120	1.4630	0.3875	0.3634	0.3875	0.3280
No log	4.0	160	1.4815	0.3812	0.3819	0.3812	0.3604
No log	5.0	200	1.2788	0.475	0.5219	0.475	0.4553
No log	6.0	240	1.2866	0.5312	0.5366	0.5312	0.5311
No log	7.0	280	1.4916	0.4313	0.4654	0.4313	0.4053
No log	8.0	320	1.3428	0.5125	0.5307	0.5125	0.5158
No log	9.0	360	1.4789	0.4188	0.4177	0.4188	0.4054
No log	10.0	400	1.6132	0.4375	0.4619	0.4375	0.4323
No log	11.0	440	1.5168	0.4875	0.5142	0.4875	0.4911
No log	12.0	480	1.4779	0.5312	0.5566	0.5312	0.5323
0.9086	13.0	520	1.5962	0.4813	0.4911	0.4813	0.4798
0.9086	14.0	560	1.5281	0.5188	0.5613	0.5188	0.5220
0.9086	15.0	600	1.5682	0.525	0.5536	0.525	0.5283

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0
• Datasets 2.19.1
• Tokenizers 0.19.1