videomae-base-finetuned-ucf-crimevbinaryv4

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

• Loss: 0.8055
• Accuracy: 0.8571
• Precision: 0.8571
• Recall: 0.8571
• Auc: 0.9328

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: 3e-05
• 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_steps: 500
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	Auc
0.4783	1.0	175	0.5235	0.78	0.7825	0.78	0.8269
0.6121	2.0	350	0.5076	0.816	0.8179	0.816	0.8518
0.3305	3.0	525	0.3910	0.856	0.8660	0.856	0.9222
0.2431	4.0	700	0.8684	0.728	0.8190	0.728	0.8933
0.3337	5.0	875	0.5653	0.808	0.8266	0.808	0.9149
0.1768	6.0	1050	0.4709	0.88	0.8803	0.88	0.9311
0.111	7.0	1225	0.6650	0.848	0.8600	0.848	0.9263
0.1507	8.0	1400	0.6151	0.852	0.8522	0.852	0.9232
0.1771	9.0	1575	0.6924	0.872	0.8729	0.872	0.9280
0.13	10.0	1750	0.6670	0.884	0.8846	0.884	0.9304
0.0062	11.0	1925	0.8110	0.86	0.8621	0.86	0.9198
0.0008	12.0	2100	0.8139	0.876	0.8768	0.876	0.9235
0.0004	13.0	2275	0.8727	0.868	0.8711	0.868	0.9258
0.0005	14.0	2450	1.0055	0.86	0.8658	0.86	0.9250
0.0002	15.0	2625	0.9728	0.864	0.8677	0.864	0.9269

Framework versions

• Transformers 4.47.0.dev0
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3