MobileNet-V2-Retinopathy

This model is a fine-tuned version of google/mobilenet_v2_1.4_224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2044
• Accuracy: 0.9307

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4403	1.0	113	0.5330	0.7079
0.5538	2.0	227	0.4312	0.7723
0.542	3.0	340	0.5137	0.7426
0.4776	4.0	454	0.4656	0.7723
0.4244	5.0	567	1.0400	0.5990
0.4694	6.0	681	0.5936	0.7228
0.4494	7.0	794	0.4667	0.7822
0.4647	8.0	908	0.2629	0.8960
0.3646	9.0	1021	0.2287	0.8861
0.4827	10.0	1135	1.7967	0.5149
0.3679	11.0	1248	0.4184	0.8267
0.3454	12.0	1362	0.1885	0.9406
0.3562	13.0	1475	0.2798	0.9059
0.3397	14.0	1589	1.6444	0.5891
0.4047	14.93	1695	0.2044	0.9307

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1