videomae-finetuned_40

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: 5.6806
• Accuracy: 0.4753

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• training_steps: 52700

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6864	0.05	2635	3.6882	0.3937
0.5651	1.05	5270	4.0977	0.4524
0.4606	2.05	7905	4.5958	0.4149
0.6425	3.05	10540	5.2244	0.4139
0.7045	4.05	13175	4.6517	0.4482
0.4105	5.05	15810	5.2383	0.4234
0.4876	6.05	18445	5.2027	0.4252
0.5886	7.05	21080	5.4343	0.3893
0.2125	8.05	23715	5.8632	0.4322
0.3111	9.05	26350	5.3468	0.4406
0.3185	10.05	28985	5.5490	0.4500
0.2399	11.05	31620	5.5484	0.4614
0.201	12.05	34255	5.5056	0.4612
0.1405	13.05	36890	5.6038	0.4612
0.1989	14.05	39525	5.7589	0.4594
0.1457	15.05	42160	5.8408	0.4644
0.1133	16.05	44795	5.7189	0.4696
0.0586	17.05	47430	5.6806	0.4753
0.0339	18.05	50065	5.7528	0.4745
0.0722	19.05	52700	5.7638	0.4751

Framework versions

• Transformers 4.38.1
• Pytorch 2.0.1+cu118
• Datasets 2.18.0
• Tokenizers 0.15.2