jialicheng/unlearn_cifar10_resnet-34_random_label_6_87

87

This model is a fine-tuned version of microsoft/resnet-34 on the cifar10 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2651
• Accuracy: 0.9309
• Dt Accuracy: 0.9309
• Df Accuracy: 0.8993
• Unlearn Overall Accuracy: 0.3321
• 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: 0.001
• train_batch_size: 32
• 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
1.3779	1.0	1563	0.9916	0.6727	0	0	None
1.1622	2.0	3126	0.6908	0.7607	0	0	None
1.0288	3.0	4689	0.6064	0.804	0.4368	0.4368	None
0.9691	4.0	6252	0.4823	0.8467	0.3929	0.3929	None
0.8915	5.0	7815	0.4758	0.8473	0.3928	0.3928	None
0.8368	6.0	9378	0.4300	0.8733	0.3623	0.3623	None
0.8074	7.0	10941	0.3932	0.888	0.3444	0.3444	None
0.7613	8.0	12504	0.3719	0.9047	0.3224	0.3224	None
0.7079	9.0	14067	0.3453	0.908	0.3182	0.3182	None
0.703	10.0	15630	0.3360	0.907	0.3202	0.3202	None
0.6619	11.0	17193	0.3514	0.9053	0.3223	0.3223	None
0.6564	12.0	18756	0.3146	0.912	0.3134	0.3134	None
0.6277	13.0	20319	0.2995	0.917	0.3068	0.3068	None
0.6109	14.0	21882	0.2992	0.9103	0.3163	0.3163	None
0.5749	15.0	23445	0.2929	0.904	0.3253	0.3253	None
0.5625	16.0	25008	0.2811	0.9103	0.3166	0.3166	None
0.5417	17.0	26571	0.2806	0.9043	0.3249	0.3249	None
0.5289	18.0	28134	0.2686	0.9077	0.3206	0.3206	None
0.5078	19.0	29697	0.2658	0.899	0.3326	0.3326	None
0.4925	20.0	31260	0.2651	0.8993	0.3321	0.3321	None

Framework versions

• Transformers 4.37.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.15.2