search

OpenGVLab / UniFormerV2

[ICCV2023] UniFormerV2: Spatiotemporal Learning by Arming Image ViTs with Video UniFormer
https://arxiv.org/abs/2211.09552
Apache License 2.0
291 stars 18 forks source link

Finetuning with our data. #51

Closed Ravindu-Yasas-Nagasinghe closed 11 months ago

Ravindu-Yasas-Nagasinghe commented 1 year ago

I want to finetune this model to the dataset CrossTask. ''' wget https://www.di.ens.fr/~dzhukov/crosstask/crosstask_release.zip

wget https://www.di.ens.fr/~dzhukov/crosstask/crosstask_features.zip

wget https://vision.eecs.yorku.ca/WebShare/CrossTask_s3d.zip

unzip '*.zip' '''

Can you please guide me through the process of doing this. Thank you

Andy1621 commented 1 year ago

You can prepare the same annotation files as here, and then simply use the same training scripts.