resnet-50-LongSleeveCleanedData

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: 0.0889
• Accuracy: 0.9788

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: 7
• total_train_batch_size: 56
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9906	0.99	143	1.0394	0.6134
0.7315	2.0	287	0.6790	0.7631
0.559	3.0	431	0.4735	0.8547
0.4905	4.0	575	0.3148	0.8983
0.3465	5.0	719	0.2225	0.9363
0.3372	6.0	863	0.1839	0.9486
0.3349	7.0	1007	0.1617	0.9587
0.3159	7.99	1150	0.1323	0.9620
0.2805	9.0	1294	0.1660	0.9587
0.2657	10.0	1438	0.1456	0.9531
0.2929	11.0	1582	0.1086	0.9698
0.2763	12.0	1726	0.0886	0.9765
0.2475	13.0	1870	0.1041	0.9732
0.2148	14.0	2014	0.0955	0.9777
0.209	14.99	2157	0.1061	0.9709
0.2408	16.0	2301	0.0784	0.9743
0.222	17.0	2445	0.0839	0.9698
0.208	18.0	2589	0.0873	0.9732
0.2214	19.0	2733	0.0889	0.9788
0.2375	19.88	2860	0.0864	0.9743

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3