I am looking for details and files which handle the implementation of the dataloader. Primarily to understand how it handles the heterogeneity in forms and extensions of different data sources. Any help or info will be appreciated.
It seems that you are looking for the implementation of collate_fn which handles the heterogeneity of the data sources from your own Dataset. You should pass your own collate_fn as the input of Dataloader. See more at https://pytorch.org/docs/stable/data.html
1 Like
Yeah, this is it. I don’t know how I missed it. Thanks. 