corrensponds with this pr

guidance-ai/guidance#221

im sorry, but i don't see any code in the folder, just a new sub module.

just fixed the submodule issue and it should show a new extension

several issues found after trying the patch:

1. port 9000 conflicts with ports 9000-9004 used by vscode python extensions, better to use more random port e.g. 9555
2. extension can be loaded only after LLM model is already loaded, otherwise error in GuidanceGenerator.init on line self.bos_token= shared.tokenizer.bos_token, where self.tokenizer is None
3. print_type is not defined but used in several places, triggering runtime error