videomae-base-finetuned-masa

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: 0.0031
• Accuracy: 1.0

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: 9e-05
• train_batch_size: 1
• eval_batch_size: 1
• 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: 225

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.601	0.0667	15	1.3659	0.8
1.5638	1.0667	30	1.4763	0.4
1.1535	2.0667	45	1.2764	0.6
0.637	3.0667	60	0.6158	0.8
0.809	4.0667	75	1.1473	0.4
0.6648	5.0667	90	0.6767	0.6
0.4322	6.0667	105	0.8144	0.8
0.3076	7.0667	120	0.1087	1.0
0.0364	8.0667	135	0.1052	1.0
0.0037	9.0667	150	0.0106	1.0
0.0027	10.0667	165	0.0075	1.0
0.0029	11.0667	180	0.0058	1.0
0.0049	12.0667	195	0.0033	1.0
0.0028	13.0667	210	0.0031	1.0
0.0027	14.0667	225	0.0031	1.0

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1