Does this code work on MSR-VTT dataset?

breadbread1984 / GODIVA

this project implements text2video algorithm introduced in paper [GODIVA: Generating open-doain videos from natural descriptions](https://www.arxiv-vanity.com/papers/2104.14806/)

7 stars 1 forks source link

Does this code work on MSR-VTT dataset? #2

Open wangzishuo029 opened 2 years ago

wangzishuo029 commented 2 years ago

Since the paper has experiment of pre-training on HowTo100M and fine-tuning on MSR-VTT, I'd like to know if I can use this code to train the model on MSR-VTT. Thank you.

breadbread1984 commented 2 years ago

this project is an implement according to my understanding. no guarantee to its authenticity. I want to test my implement on a simple dataset for verification first. The vqvae has been pretrained. the godiva doesn't converge on moving mnist. you are welcome to contribute.