vit-base-patch16-224-finetuned-ind-17-imbalanced-aadhaarmask-new-parameter

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

• Loss: 0.3278
• Accuracy: 0.8519
• Recall: 0.8519
• F1: 0.8508
• Precision: 0.8576

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: 8
• eval_batch_size: 8
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Recall	F1	Precision
No log	1.0	1175	0.5572	0.8076	0.8076	0.7937	0.8043
No log	2.0	2350	0.4673	0.8284	0.8284	0.8271	0.8347
No log	3.0	3525	0.4109	0.8344	0.8344	0.8301	0.8367
No log	4.0	4700	0.3984	0.8382	0.8382	0.8339	0.8375
No log	5.0	5875	0.3886	0.8412	0.8412	0.8398	0.8467
No log	6.0	7050	0.3520	0.8493	0.8493	0.8481	0.8519
No log	7.0	8225	0.4229	0.8416	0.8416	0.8399	0.8512
No log	8.0	9400	0.3140	0.8612	0.8612	0.8600	0.8656
No log	9.0	10575	0.3399	0.8421	0.8421	0.8403	0.8464
0.4263	10.0	11750	0.3399	0.8476	0.8476	0.8468	0.8536

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.0a0+81ea7a4
• Datasets 2.19.0
• Tokenizers 0.19.1