adding model option for generate-embeddings command

[ukaraoz]

@caufieldjh I want to add a model option to "generate-embeddings" command. Currently, it is calling the default model of curate-gpt from hugging-face (set with default as "all-MiniLM-L6-v2" when the "store" object is created).
Is it where this should be done with get_store? How do you change the model that is associated with a store in curate-gpt?

trowel/src/trowel/utils/embedding_generation_utils.py

Line 14 in 2d6ba38

def _get_curategpt_store(store_type: str, db_path: str):

[caufieldjh]

OK, PR #19 added a --model (-m) option to the generate-embeddings command, so it can take a different embedding model.
There are a few issues with how CurateGPT defines model dimensions and saves them in the stored embeddings - these issues prevent trowel from using arbitrary embedding models for now, but I think #20 may help.