jialicheng/unlearn_cifar100_swin-base_bad_teaching_2_87

87

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224 on the cifar100 dataset. It achieves the following results on the evaluation set:

• Loss: 13.0270
• Accuracy: 0.3061
• Dt Accuracy: 0.3061
• Df Accuracy: 0.194
• 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: 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	16	1.5558	0.693	0	0	None
No log	2.0	32	1.5415	0.686	0	0	None
No log	3.0	48	1.8880	0.661	0	0	None
No log	4.0	64	2.3622	0.618	0	0	None
No log	5.0	80	3.2308	0.529	0	0	None
No log	6.0	96	3.3136	0.541	0	0	None
No log	7.0	112	3.8720	0.495	0	0	None
No log	8.0	128	3.7642	0.509	0	0	None
No log	9.0	144	4.3507	0.468	0	0	None
No log	10.0	160	5.2100	0.417	0	0	None
No log	11.0	176	5.5091	0.397	0	0	None
No log	12.0	192	6.1317	0.352	0	0	None
No log	13.0	208	6.8544	0.327	0	0	None
No log	14.0	224	8.7042	0.29	0	0	None
No log	15.0	240	8.6705	0.278	0	0	None
No log	16.0	256	8.6764	0.257	0	0	None
No log	17.0	272	9.9202	0.241	0	0	None
No log	18.0	288	11.7378	0.206	0	0	None
No log	19.0	304	11.2166	0.212	0	0	None
No log	20.0	320	13.0270	0.194	0	0	None

Framework versions

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