jialicheng/unlearn_cifar100_resnet-50_random_label_8_87

87

This model is a fine-tuned version of microsoft/resnet-50 on the cifar100 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6510
• Accuracy: 0.8376
• Dt Accuracy: 0.8376
• Df Accuracy: 0.7877
• Unlearn Overall Accuracy: 0.3634
• Unlearn Time: None

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: 0.0002
• train_batch_size: 128
• eval_batch_size: 256
• seed: 87
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
No log	1.0	391	0.6709	0.865	0.2526	0.2526	None
1.5339	2.0	782	0.6458	0.8678	0.2489	0.2489	None
1.3308	3.0	1173	0.6572	0.8685	0.2477	0.2477	None
1.2467	4.0	1564	0.6599	0.8722	0.2419	0.2419	None
1.2467	5.0	1955	0.6461	0.8745	0.2386	0.2386	None
1.1938	6.0	2346	0.6600	0.8675	0.2493	0.2493	None
1.1321	7.0	2737	0.6547	0.8645	0.2539	0.2539	None
1.09	8.0	3128	0.6469	0.8572	0.2650	0.2650	None
1.0594	9.0	3519	0.6468	0.858	0.2639	0.2639	None
1.0594	10.0	3910	0.6526	0.8462	0.2813	0.2813	None
1.0154	11.0	4301	0.6417	0.8448	0.2836	0.2836	None
0.9958	12.0	4692	0.6507	0.8243	0.3132	0.3132	None
0.9663	13.0	5083	0.6423	0.8303	0.3045	0.3045	None
0.9663	14.0	5474	0.6486	0.8137	0.3279	0.3279	None
0.9523	15.0	5865	0.6490	0.7963	0.3519	0.3519	None
0.9254	16.0	6256	0.6479	0.7953	0.3533	0.3533	None
0.905	17.0	6647	0.6509	0.786	0.3658	0.3658	None
0.8996	18.0	7038	0.6485	0.7943	0.3546	0.3546	None
0.8996	19.0	7429	0.6513	0.7893	0.3613	0.3613	None
0.8853	20.0	7820	0.6510	0.7877	0.3634	0.3634	None

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu118
• Datasets 2.18.0
• Tokenizers 0.15.2