jialicheng/unlearn_ucf101_videomae-large_neggrad_4_42

ucf101_42

This model is a fine-tuned version of MCG-NJU/videomae-large on the ucf101 dataset. It achieves the following results on the evaluation set:

• Loss: -14.9674
• Accuracy: 0.0043
• Test Accuracy: 0.0043
• Df Accuracy: 0.0043
• Unlearn Overall Accuracy: 0.5000
• Unlearn Time: 4011.4122

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: 4
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
No log	1.11	14	-14.9674	0.0043	0.5000	0.5000	-1
No log	2.11	28	-15.9597	0.0043	0.5000	0.5000	-1
No log	2.63	36	-16.0358	0.0043	0.5000	0.5000	-1

Framework versions

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