cconvnext-tiny-15ep-1e-4

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

• Loss: 0.2767
• Accuracy: 0.9375

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• 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
0.5838	1.0	550	0.4097	0.8811
0.4565	2.0	1100	0.4269	0.8763
0.3628	3.0	1650	0.3464	0.9002
0.2915	4.0	2200	0.3366	0.9066
0.2655	5.0	2750	0.3387	0.9054
0.2395	6.0	3300	0.3313	0.9125
0.2065	7.0	3850	0.3120	0.9181
0.1503	8.0	4400	0.3065	0.9221
0.1503	9.0	4950	0.2948	0.9276
0.1125	10.0	5500	0.2918	0.9304
0.1057	11.0	6050	0.2954	0.9328
0.0937	12.0	6600	0.2959	0.9336
0.0966	13.0	7150	0.2940	0.9352
0.0735	14.0	7700	0.2916	0.9340
0.0881	15.0	8250	0.2902	0.9356

Framework versions

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