videomae-small-finetuned-ssv2-finetuned-judo

This model is a fine-tuned version of MCG-NJU/videomae-small-finetuned-ssv2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6398
• Accuracy: 0.7568

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.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: 0.1
• training_steps: 190

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0545	0.1053	20	1.0499	0.4324
0.8731	1.1053	40	1.0377	0.4865
0.7354	2.1053	60	0.9788	0.5676
0.6503	3.1053	80	0.9652	0.5946
0.6196	4.1053	100	0.8469	0.5946
0.5233	5.1053	120	0.7983	0.5676
0.4382	6.1053	140	0.7651	0.6757
0.2919	7.1053	160	0.7395	0.7297
0.3249	8.1053	180	0.6420	0.7568
0.2896	9.0526	190	0.6398	0.7568

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0