videomae-base-finetuned-ucf101-subset

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: 0.0986
• Accuracy: 0.9784

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 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: 0.05
• training_steps: 2280

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8511	0.1254	286	1.3736	0.5602
0.7099	1.1254	572	0.5359	0.8397
0.3433	2.1254	858	0.4332	0.8772
0.2015	3.1254	1144	0.2627	0.9203
0.1166	4.1254	1430	0.1257	0.9620
0.0394	5.1254	1716	0.0980	0.9714
0.0092	6.1254	2002	0.0888	0.9766
0.0246	7.1219	2280	0.0739	0.9822

Framework versions

• Transformers 4.52.2
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1