convnextv2-tiny-1k-224-finetuned-crop-upperfull-mix

This model is a fine-tuned version of facebook/convnextv2-tiny-1k-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6187
• Accuracy: 0.8004
• Precision: 0.8160

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: 2e-05
• train_batch_size: 10
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision
No log	1.0	183	2.1471	0.4693	0.5128
No log	2.0	366	1.4576	0.6579	0.6955
1.9821	3.0	549	1.1372	0.6754	0.7183
1.9821	4.0	732	0.9214	0.7303	0.7659
1.9821	5.0	915	0.7792	0.7478	0.7661
0.8885	6.0	1098	0.7455	0.7654	0.7780
0.8885	7.0	1281	0.6756	0.7873	0.8020
0.8885	8.0	1464	0.6787	0.7807	0.7932
0.5696	9.0	1647	0.6694	0.7982	0.8099
0.5696	10.0	1830	0.6799	0.7741	0.7930
0.4056	11.0	2013	0.6187	0.8004	0.8160
0.4056	12.0	2196	0.6868	0.7675	0.8063
0.4056	13.0	2379	0.7525	0.7544	0.7803
0.2904	14.0	2562	0.6572	0.7895	0.8093

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1