resnet-50-finetuned-omar

This model is a fine-tuned version of microsoft/resnet-50 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2645
• Accuracy: 0.9144

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• 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: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0695	1.0	111	1.0576	0.5315
0.971	2.0	223	0.9366	0.5416
0.8121	3.0	334	0.7493	0.7103
0.6861	4.0	446	0.5625	0.8363
0.606	5.0	557	0.4239	0.8816
0.5001	6.0	669	0.3159	0.9219
0.4704	7.0	780	0.3254	0.9118
0.4332	8.0	892	0.2808	0.9194
0.4432	9.0	1003	0.2854	0.9219
0.4768	10.0	1115	0.2782	0.9219
0.4432	11.0	1226	0.2768	0.9320
0.4752	12.0	1338	0.2744	0.9219
0.489	13.0	1449	0.2693	0.9194
0.3743	14.0	1561	0.2715	0.9270
0.417	14.93	1665	0.2645	0.9144

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu117
• Datasets 2.13.0
• Tokenizers 0.13.3