jialicheng/unlearn-so_cifar100_resnet-50_neggrad_10_100

100

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: -17.7160
• Accuracy: 0.0301
• Dt Accuracy: 0.0301
• Df Accuracy: 0.0272
• 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: 100
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
No log	1.0	40	-0.8119	0.8082	0.3317	0.3317	None
No log	2.0	80	-2.4196	0.5504	0.5564	0.5564	None
-2.1794	3.0	120	-5.4562	0.3046	0	0	None
-2.1794	4.0	160	-8.5760	0.1732	0	0	None
-10.2455	5.0	200	-11.6439	0.0924	0	0	None
-10.2455	6.0	240	-13.4916	0.0646	0	0	None
-10.2455	7.0	280	-14.6529	0.0424	0	0	None
-18.2027	8.0	320	-16.7375	0.031	0	0	None
-18.2027	9.0	360	-18.3103	0.0272	0	0	None
-22.3184	10.0	400	-17.7160	0.0272	0	0	None

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu118
• Datasets 2.19.0
• Tokenizers 0.15.2