simple-llm-finetuner
simple-llm-finetuner copied to clipboard
Add A cli version
You did an amazing job.
But most people wants a cli version to quickly get started and play around.
Here it is.
Cheers.
I will say, I would rather keep the implementations cleaner for learning purposes. This is one of the easiest to grasp and understand repo on how to fine tune a project, and is a great learning example.
If anything, it should be split into 3 files. tuner.py
which has the functions and extras that are the core. web.py
which has the webserver code and calls tuner.py
functions. and cli.py
for handling cli args, and calls tuner.py
.
Obviously as the author you can choose what you want, but for those coming to your repo as one of the few easy examples of fine tuning, it might be better to keep things cleanly laid out instead of lumped into 1 file.
If it's possible to add a separate file that can be invoked to do CLI training, i would love that. I am currently attempting to run training via CLI; I will have a look at this branch.