videomae-base-finetuned-kinetics-finetuned-violence-subset

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: 0.0728
• Accuracy: 0.9850
• F1: 0.9843
• Precision: 0.9857
• Recall: 0.9830
• Roc Auc: 0.9950

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 500

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Roc Auc
0.0972	0.102	51	0.0966	0.9700	0.9688	0.9673	0.9705	0.9958
0.1116	1.102	102	0.1824	0.9421	0.9379	0.9560	0.9274	0.9851
0.0509	2.102	153	0.0887	0.9764	0.9752	0.9799	0.9713	0.9930
0.0371	3.102	204	0.0574	0.9850	0.9843	0.9857	0.9830	0.9974
0.0338	4.102	255	0.0501	0.9850	0.9843	0.9848	0.9839	0.9979
0.0046	5.102	306	0.0954	0.9785	0.9774	0.9828	0.9731	0.9925
0.0302	6.102	357	0.0495	0.9850	0.9843	0.9848	0.9839	0.9990
0.0022	7.102	408	0.0703	0.9871	0.9865	0.9884	0.9848	0.9947
0.0185	8.102	459	0.0779	0.9828	0.9820	0.9839	0.9803	0.9935
0.0088	9.082	500	0.0728	0.9850	0.9843	0.9857	0.9830	0.9950

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1