llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard

A simple script merge lora to HF

Open FNsi opened this issue 1 year ago • 4 comments

Thanks for #1516 I hope your guys enjoy it.

FNsi avatar May 20 '23 00:05 FNsi

  • Add program and argument descriptions the ArgumentParser

  • Remove the default argument values and set them as required instead (for the model, you could also default it to something like decapoda-research/llama-7b-hf)

  • Rename the file to something more descriptive

  • Improve code formatting (you can use an auto-format tool such as black)

  • Remove unused imports

Thank you for the advice, gonna fix then, but... what do you mean by "Add program and argument descriptions the ArgumentParser" ? Should I write some notes or?

FNsi avatar May 21 '23 00:05 FNsi

Just add a short description of what the program and each argument do (you can see how to that in the documentation).

slaren avatar May 21 '23 00:05 slaren

This works pretty well.

I think there should be a mention somewhere that it adds more required pip packages. I don't know if we should put them in the requirements.txt or not.

SlyEcho avatar May 24 '23 21:05 SlyEcho

No clue, but I assume they would like to open that script while they are working on it🤦

FNsi avatar May 25 '23 00:05 FNsi