jialicheng/unlearn_cifar100_resnet-50_neggrad_10_42

42

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: -61.2603
• Accuracy: 0.0145
• Dt Accuracy: 0.0145
• Df Accuracy: 0.0132
• Unlearn Overall Accuracy: 0
• 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: 5e-05
• train_batch_size: 128
• eval_batch_size: 256
• seed: 42
• 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	40	-0.6705	0.8614	0.2576	0.2576	None
No log	2.0	80	-1.1856	0.7334	0.4182	0.4182	None
No log	3.0	120	-3.0241	0.4682	0	0	None
No log	4.0	160	-6.8932	0.2358	0	0	None
No log	5.0	200	-11.5720	0.1004	0	0	None
No log	6.0	240	-16.6897	0.0524	0	0	None
No log	7.0	280	-21.7166	0.035	0	0	None
No log	8.0	320	-26.8608	0.0274	0	0	None
No log	9.0	360	-31.0666	0.0212	0	0	None
No log	10.0	400	-38.7816	0.0186	0	0	None
No log	11.0	440	-41.5065	0.017	0	0	None
No log	12.0	480	-45.3635	0.017	0	0	None
-18.2648	13.0	520	-49.5281	0.016	0	0	None
-18.2648	14.0	560	-53.6067	0.0154	0	0	None
-18.2648	15.0	600	-50.5046	0.0158	0	0	None
-18.2648	16.0	640	-55.5001	0.0148	0	0	None
-18.2648	17.0	680	-55.9571	0.0148	0	0	None
-18.2648	18.0	720	-53.0937	0.015	0	0	None
-18.2648	19.0	760	-54.5232	0.0144	0	0	None
-18.2648	20.0	800	-61.2603	0.0132	0	0	None

Framework versions

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