videomae-base-finetuned-Cheh2_ucf_light_demo-10epochs

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: 1.6448
• Accuracy: 0.5270

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: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 15010

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.3441	0.1001	1502	2.3923	0.2469
1.6506	1.1001	3004	2.0596	0.3513
1.1963	2.1001	4506	2.0148	0.4035
1.5381	3.1001	6008	1.8796	0.4262
1.5438	4.1001	7510	1.8851	0.4398
1.225	5.1001	9012	1.7874	0.4649
1.038	6.1001	10514	1.7263	0.4852
1.0129	7.1001	12016	1.7192	0.5046
1.0309	8.1001	13518	1.6448	0.5270
0.91	9.0994	15010	1.6655	0.5144

Framework versions

• Transformers 4.43.4
• Pytorch 2.4.1
• Datasets 2.20.0
• Tokenizers 0.19.1