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: 0.8216
• Model Preparation Time: 0.0016
• Accuracy: 0.7483

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: polynomial
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 3096

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Accuracy
1.468	0.2503	775	1.6524	0.0016	0.4545
0.9186	1.2503	1550	1.3271	0.0016	0.5798
0.8552	2.2503	2325	1.2284	0.0016	0.6207
0.8815	3.2490	3096	1.2222	0.0016	0.6213

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1
• Tokenizers 0.19.1