Open
Description
In the "lessons learnt from codegen2" paper, it's discussed that data mix of pile and thestarcoder data is a better choice to undertake if enough compute is available, but it's not clear if codegen2 or codegen2.5 (base models not instruct models) were trained with natural language data like ThePile etc. Is there any small model <=7B which is trained on both ThePile and TheStarCoder data?
Metadata
Metadata
Assignees
Labels
No labels