Using Fine Tuning using Ray submission CLI

[chainlink]

This adds some (very light) docs on using the ray job submission CLI to submit a finetuning job

[sfriedowitz]

This is essentially already stated in the package Readme. I think this is duplicate?

Perhaps just add a note to that section saying that lm buddy does not need to be installed locally when submitting to Ray?

[chainlink]

The difference here is that there is no python driver script required, it's all CLI based. I can add to the Readme though.