videomae-base-finetuned-kinetics-allkisa-crop-background-0307-clip_duration-abnormal38

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: 0.1985
• Accuracy: 0.9329

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: 29400

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0024	0.0100	295	0.2932	0.8709
0.0011	1.0100	590	0.6339	0.8045
0.0028	2.0100	885	0.7747	0.7888
0.0029	3.0100	1180	0.5034	0.8551
1.0298	4.0100	1475	0.5480	0.8150
0.0248	5.0100	1770	0.3638	0.8726
0.0435	6.0100	2065	0.8338	0.7976
0.0001	7.0100	2360	0.8290	0.7749
0.0004	8.0100	2655	0.5562	0.8464
0.0001	9.0100	2950	1.0491	0.8080
0.0002	10.0100	3245	0.7868	0.8290
0.0197	11.0100	3540	0.6573	0.8586
0.0002	12.0100	3835	0.8348	0.8325
0.0001	13.0100	4130	0.6864	0.8464
0.0001	14.0100	4425	0.6752	0.8517
0.0	15.0100	4720	0.8101	0.8464

Framework versions

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