videomae-base-finetuned-kinetics-finetuned-right-hand-conflab-v8

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: 1.6294
• Accuracy: 0.6207

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: 1e-06
• 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: 1755

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0111	0.0672	118	1.7595	0.6127
0.0532	1.0672	236	1.7670	0.6029
0.0153	2.0672	354	1.7648	0.5980
0.0348	3.0672	472	1.7793	0.6029
0.0069	4.0672	590	1.7859	0.6029
0.0615	5.0672	708	1.7920	0.5931
0.0202	6.0672	826	1.7948	0.5882
0.0195	7.0672	944	1.8089	0.5882
0.0165	8.0672	1062	1.8133	0.5931
0.0145	9.0672	1180	1.8067	0.5980
0.0118	10.0672	1298	1.8071	0.6078
0.0759	11.0672	1416	1.8138	0.5980
0.01	12.0672	1534	1.8117	0.6029
0.0088	13.0672	1652	1.8139	0.6127
0.0315	14.0587	1755	1.8149	0.6127

Framework versions

• Transformers 4.41.0
• Pytorch 1.12.0+cu116
• Datasets 2.19.1
• Tokenizers 0.19.1