portrait_cosu_exp5

This model is a fine-tuned version of NekoFi/portrait_cosu_exp4 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1603
• Accuracy: 0.9583
• Precision: 0.9582
• Recall: 0.9583
• F1: 0.9580
• Confusion Matrix: [[50, 1], [2, 19]]

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: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Confusion Matrix
0.3966	0.9756	10	0.3024	0.8611	0.8695	0.8611	0.8498	[[50, 1], [9, 12]]
0.3486	1.9512	20	0.3286	0.8333	0.8458	0.8333	0.8147	[[50, 1], [11, 10]]
0.2347	2.9268	30	0.1769	0.9444	0.9446	0.9444	0.9436	[[50, 1], [3, 18]]
0.2039	3.9024	40	0.1603	0.9583	0.9582	0.9583	0.9580	[[50, 1], [2, 19]]

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1