videomae-base-finetuned-kinetics-0407_final_5sec_bbox_retry

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: 0.5853
• Accuracy: 0.8268

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: 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: 0.05
• training_steps: 64400

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0064	0.0087	560	0.2854	0.9155
0.027	1.0087	1120	0.4700	0.8591
0.0034	2.0087	1680	0.8010	0.8127
0.1029	3.0087	2240	0.7237	0.8145
0.0282	4.0087	2800	1.3460	0.6745
0.028	5.0087	3360	0.4030	0.8855
0.0029	6.0087	3920	0.5687	0.8655
0.0034	7.0087	4480	0.6655	0.8291
0.101	8.0087	5040	0.7841	0.8264
0.3789	9.0087	5600	2.0764	0.6673
0.0908	10.0087	6160	1.0014	0.7827
0.0005	11.0087	6720	0.6949	0.8545
0.2388	12.0087	7280	1.9325	0.6973
0.0004	13.0087	7840	1.0476	0.7864
2.0909	14.0087	8400	0.9193	0.8227
0.0047	15.0087	8960	2.3364	0.6755

Framework versions

• Transformers 4.48.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0