videomae-finetuned_37

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: 0.7263
• Accuracy: 0.8972

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: 56160

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7015	0.05	2809	0.6809	0.8385
0.4496	1.05	5618	0.9956	0.8228
0.5392	2.05	8427	0.9517	0.8287
0.5662	3.05	11236	1.0190	0.8071
0.48	4.05	14045	0.8919	0.8148
0.4773	5.05	16854	0.8265	0.8507
0.4749	6.05	19663	0.7975	0.8612
0.3916	7.05	22472	0.7923	0.8648
0.2919	8.05	25281	0.7495	0.8692
0.2068	9.05	28090	0.7460	0.8794
0.2217	10.05	30899	0.7522	0.8683
0.2886	11.05	33708	1.0231	0.8259
0.2319	12.05	36517	0.6641	0.8808
0.2056	13.05	39326	0.7961	0.8774
0.1969	14.05	42135	0.7620	0.8769
0.0624	15.05	44944	0.7757	0.8733
0.091	16.05	47753	0.7268	0.8962
0.0722	17.05	50562	0.7545	0.8931
0.0633	18.05	53371	0.7487	0.8924
0.0713	19.05	56160	0.7263	0.8972

Framework versions

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