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: 0.1863
• Accuracy: 0.9484

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: cosine
• lr_scheduler_warmup_steps: 15
• training_steps: 185

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.339	0.0811	15	2.2562	0.1429
1.9066	1.0541	30	1.8097	0.3571
1.5338	2.0270	45	1.1491	0.5286
0.6251	2.1081	60	0.8128	0.7286
0.4272	3.0811	75	0.4791	0.8429
0.2135	4.0541	90	0.4702	0.9
0.1482	5.0270	105	0.3444	0.8857
0.1036	5.1081	120	0.2044	0.9
0.0583	6.0811	135	0.2283	0.9286
0.0358	7.0541	150	0.2171	0.9
0.0353	8.0270	165	0.1739	0.9571
0.055	8.1081	180	0.1209	0.9714

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2