mlfoundations / open_flamingo

An open-source framework for training large multimodal models.
MIT License
3.74k stars 284 forks source link

The format of dataset of mmc4 and laion? So I can train on other dataset. #304

Open hhy150 opened 4 months ago

hhy150 commented 4 months ago

Can I get the approximate format of the data in the tar files of mmc4 and laion_dataset? I want to train with my own dataset, but I don't know how to adjust the data format to make it run.