convnext-tiny-224-finetuned-main-gpu-20e-final

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

• Loss: 0.0349
• Accuracy: 0.9875

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6197	1.0	551	0.5899	0.7440
0.3906	2.0	1102	0.3245	0.8717
0.3161	3.0	1653	0.2228	0.9135
0.2323	4.0	2204	0.1481	0.9446
0.2049	5.0	2755	0.1100	0.9589
0.1453	6.0	3306	0.0887	0.9671
0.1786	7.0	3857	0.0796	0.9702
0.1576	8.0	4408	0.0635	0.9767
0.1584	9.0	4959	0.0563	0.9798
0.122	10.0	5510	0.0570	0.9793
0.1138	11.0	6061	0.0526	0.9819
0.1116	12.0	6612	0.0498	0.9832
0.0876	13.0	7163	0.0497	0.9830
0.0956	14.0	7714	0.0403	0.9855
0.0892	15.0	8265	0.0414	0.9855
0.0807	16.0	8816	0.0425	0.9861
0.0959	17.0	9367	0.0397	0.9866
0.0847	18.0	9918	0.0373	0.9874
0.0962	19.0	10469	0.0356	0.9870
0.0731	20.0	11020	0.0349	0.9875

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2