[FEATURE][DocBot] Add a Cohere Command model

Is your feature request related to a problem?

At the moment we have a connector that will support us in embedding our documents however we are missing a model for getting generative responses.

What solution would you like?

Since we are already using the Cohere connector I think we should try and use the cohere command-nightly LLM. We can at least evaluate this and ensure that it will meet our needs.

This should be created in the cluster_bootstrap.py file after we create the cohere connector. Unlike the other model this one will not get plugged into an index or search pipelines as it will just be used to generate responses from our commands.

What alternatives have you considered?

We could use an outside OpenSearch connector to build this however I think there are advantages to having it within the cluster.