videomae-base-finetuned-kinetics-0211-val

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: 0.8226
• Accuracy: 0.7474

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: 5850

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.1638	0.0333	195	1.6381	0.4142
0.0035	1.0333	390	1.4162	0.4349
0.0206	2.0333	585	3.8252	0.3432
0.0139	3.0333	780	3.2767	0.2456
0.007	4.0333	975	2.7539	0.2959
0.0083	5.0333	1170	2.8000	0.4349
0.0014	6.0333	1365	2.7028	0.4467
0.0054	7.0333	1560	2.0244	0.5030
0.0005	8.0333	1755	3.2895	0.2692
0.0003	9.0333	1950	3.3746	0.2959
0.0001	10.0333	2145	2.3001	0.4556
0.0032	11.0333	2340	3.1056	0.4172
0.1476	12.0333	2535	2.9392	0.4231

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0