Add check max_steps > convergence_steps for celerite_engine

[andresgur]

tinygp, celerite2 engines already check this

num_iterations = int(max_steps / converge_steps)

        if num_iterations < 1:
            raise ValueError(
                f"max_steps ({max_steps}) must be at least as large as converge_steps ({converge_steps}) to run at least one iteration."
            )

So we just need the same logic in celerite_engine.py.

Maybe some of these functionalities could be moved to the base class gp_engine.py? As they are all the same.

[andresgur]

Noted there''s also a typo in convergenece_steps' parameter name definition in celerite_engine, and because it does not match the name in CeleriteGPEngine is causing problems when calling process_lightcurves_par as it is not recognizing the keyword