It really depends on the use case. :) I would suggest that a smaller model (7b parameters) is a good starting point to begin adding your own data in to fine tune the model for your specific use case.
I don't know about anyone else but I wasted a lot of time trying to get to the filepath you put qna.yaml in. Would've appreciated some steps of navigating the dirs or creating them, at least to know if I was looking in the right place. Thank you for the great demo and all the work.
Hey - sorry about that. I should have been more clear. It doesn't really matter where you place the file under the taxonomy/knowledge directory. It is more about keeping things organized in a way that you will understand for maintenace reasons.
Really nice walk-through of an InstructLab features and how to train various LLM with it. Appreciate it Grant Shipley!!
Thank you for clear & informative demo.
Thanks alot. Very clear & detail for me.
Great demo.
Really and easy if you got your hands on this :) keep up the great work
Awesome
How do match a use case to model implementation or tweak for its applicability
It really depends on the use case. :) I would suggest that a smaller model (7b parameters) is a good starting point to begin adding your own data in to fine tune the model for your specific use case.
wait, so how do I install this ? how do I install ilab?
I don't know about anyone else but I wasted a lot of time trying to get to the filepath you put qna.yaml in. Would've appreciated some steps of navigating the dirs or creating them, at least to know if I was looking in the right place.
Thank you for the great demo and all the work.
Hey - sorry about that. I should have been more clear. It doesn't really matter where you place the file under the taxonomy/knowledge directory. It is more about keeping things organized in a way that you will understand for maintenace reasons.