videomae-base-kinetics-binary-finetuned-xd-violence

This model is a fine-tuned version of MCG-NJU/videomae-base-finetuned-kinetics on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6193
• Accuracy: 0.6697
• F1: 0.5244
• Precision: 0.5685
• Recall: 0.4867
• Specificity: 0.7791
• True Positives: 6559
• True Negatives: 17556
• False Positives: 4978
• False Negatives: 6918

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: 0.0005
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 1684

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Specificity	True Positives	True Negatives	False Positives	False Negatives
0.7027	0.2506	422	0.6753	0.6153	0.4978	0.4867	0.5094	0.6787	6865	15293	7241	6612
0.57	1.2506	844	0.6846	0.6354	0.5330	0.5118	0.5561	0.6828	7494	15386	7148	5983
0.6373	2.2506	1266	0.6396	0.6502	0.5753	0.5273	0.6329	0.6606	8529	14887	7647	4948
0.5108	3.2482	1684	0.6193	0.6697	0.5244	0.5685	0.4867	0.7791	6559	17556	4978	6918

Framework versions

• Transformers 4.51.3
• Pytorch 2.1.0+cu118
• Datasets 3.6.0
• Tokenizers 0.21.1