search

ali-vilab / VGen

Official repo for VGen: a holistic video generation ecosystem for video generation building on diffusion models
https://i2vgen-xl.github.io
2.75k stars 243 forks source link

How to train a T2V model from scratch? #91

Open xiefan233 opened 4 months ago

xiefan233 commented 4 months ago

Hi, I want to start training from scratch, how should I set up the yaml file or change the code?

LuthandoMaqondo commented 4 months ago

Hi, were you able to do inference though, because the test inference suggested that you make some changes to the training YAML file. On my end, I'm still struggling to do the inference on what they have given.

From there I think the you can be able top either do a finetunning or retraining job easily.