resnet-50-finetuned-omars6

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: 0.4990
• Accuracy: 0.8328

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3594	0.99	92	1.3630	0.5015
1.3214	2.0	185	1.3252	0.5714
1.2633	2.99	277	1.2851	0.6140
1.2693	4.0	370	1.2385	0.6626
1.1902	4.99	462	1.1837	0.6991
1.1421	6.0	555	1.1255	0.7568
1.1979	6.99	647	1.0094	0.8024
0.9431	8.0	740	0.9544	0.8237
0.9627	8.99	832	0.8864	0.8267
0.8556	10.0	925	0.8365	0.8328
0.7792	10.99	1017	0.7762	0.8359
0.7941	12.0	1110	0.7467	0.8359
0.8361	12.99	1202	0.7345	0.8237
0.7757	14.0	1295	0.7228	0.8146
0.6977	14.99	1387	0.6923	0.8267
0.6874	16.0	1480	0.6540	0.8146
0.6887	16.99	1572	0.6276	0.8298
0.7204	18.0	1665	0.5989	0.8267
0.8334	18.99	1757	0.6027	0.8237
0.7654	20.0	1850	0.5699	0.8511
0.7628	20.99	1942	0.5465	0.8389
0.7874	22.0	2035	0.5621	0.8298
0.8149	22.99	2127	0.5474	0.8298
0.7565	24.0	2220	0.5388	0.8480
0.7241	24.99	2312	0.5351	0.8267
0.7894	26.0	2405	0.5327	0.8389
0.7664	26.99	2497	0.5065	0.8450
0.6655	28.0	2590	0.5309	0.8359
0.607	28.99	2682	0.5061	0.8541
0.6462	29.84	2760	0.4990	0.8328

Framework versions

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