convnextv2-large-1k-224-finetuned-Lesion-Classification-HAM10000-AH-60-20-20-Shuffled

This model is a fine-tuned version of facebook/convnextv2-large-1k-224 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0721
• Accuracy: 0.9869

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: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.9
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8937	1.0	114	1.9040	0.3144
1.7208	2.0	229	1.6891	0.5632
1.3822	3.0	343	1.3554	0.6897
1.1497	4.0	458	1.2437	0.5755
0.8979	5.0	572	0.8548	0.7701
0.6382	6.0	687	0.6359	0.8424
0.583	7.0	801	0.4687	0.8966
0.6295	8.0	916	0.5029	0.8456
0.5367	9.0	1030	0.4742	0.8670
0.5091	10.0	1145	0.3038	0.9212
0.3521	11.0	1259	0.1855	0.9606
0.318	12.0	1374	0.1893	0.9573
0.2725	13.0	1488	0.2292	0.9409
0.2937	14.0	1603	0.0866	0.9836
0.1185	14.93	1710	0.0721	0.9869

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3