convnext-base-15ep

This model is a fine-tuned version of facebook/convnextv2-base-22k-384 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2376
• Accuracy: 0.9448

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: 3e-05
• 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: cosine
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6099	1.0	1099	0.3668	0.8934
0.5086	2.0	2198	0.2773	0.9276
0.386	3.0	3297	0.2587	0.9324
0.335	4.0	4396	0.2400	0.9348
0.3167	5.0	5495	0.2599	0.9340
0.2703	6.0	6594	0.2440	0.9419
0.2638	7.0	7693	0.2496	0.9408
0.1938	8.0	8792	0.2366	0.9431
0.1789	9.0	9891	0.2353	0.9487
0.1738	10.0	10990	0.2380	0.9499
0.1924	11.0	12089	0.2458	0.9463
0.1628	12.0	13188	0.2434	0.9491
0.1431	13.0	14287	0.2390	0.9499
0.1432	14.0	15386	0.2391	0.9503
0.1297	15.0	16485	0.2384	0.9499

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2