videomae-base-finetuned-kinetics-finetuned-custom-subset

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

• Loss: 0.9783
• Accuracy: 0.7564

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: 4
• eval_batch_size: 4
• 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: 710

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6747	0.1014	72	0.6989	0.7308
0.5585	1.1014	144	0.7504	0.7308
1.0181	2.1014	216	0.5370	0.6795
0.5883	3.1014	288	0.6710	0.7308
0.3918	4.1014	360	0.5803	0.7692
0.3903	5.1014	432	0.5857	0.7436
0.2452	6.1014	504	0.8492	0.7308
0.4107	7.1014	576	0.8475	0.7308
0.099	8.1014	648	0.9360	0.7436
0.0679	9.0873	710	0.9783	0.7564

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.1.1
• Datasets 2.13.2
• Tokenizers 0.19.1