videomae-base-finetuned-ucf101-subset

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: 0.3230
• Accuracy: 0.8968

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: 1200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1382	0.06	75	2.1056	0.2571
0.8185	1.06	150	0.6270	0.8
0.5221	2.06	225	0.4341	0.8
0.1069	3.06	300	0.4390	0.8714
0.0195	4.06	375	0.2938	0.8571
0.0097	5.06	450	0.2114	0.9
0.0076	6.06	525	0.1509	0.9429
0.1686	7.06	600	0.2527	0.9571
0.0679	8.06	675	0.0615	0.9714
0.0024	9.06	750	0.1589	0.9429
0.1946	10.06	825	0.4014	0.9
0.154	11.06	900	0.1862	0.9429
0.0021	12.06	975	0.0683	0.9857
0.0019	13.06	1050	0.0541	0.9857
0.002	14.06	1125	0.0473	0.9857
0.0018	15.06	1200	0.0475	0.9857

Framework versions

• Transformers 4.29.1
• Pytorch 2.0.1+cu117
• Datasets 2.12.0
• Tokenizers 0.13.3