Migrate LM Backend from Async OpenAI to Rig

[krypticmouse]

We should convert LM to a trait and with Async OAI frictions popping up it's a good idea to use Rig as backend!

[krypticmouse]

Closes for #37