add --spans-key option for CLI spancat evaluation

[evornov]

Description

The internal evaluate method called by the evaluate/benchmark accuracy CLI commands accepts a spans_key value, but the CLI command itself does not. This PR adds a --spans-key option so that the CLI can be used to evaluate spancat models with span key values other than the default sc.

I also updated the CLI documentation, but I didn't add a "new" tag because I'm not sure what the version would be there.

Types of change

Minor CLI enhancement

Checklist

• I confirm that I have the right to submit this contribution under the project's MIT license.
• I ran the tests, and all new and existing tests passed.
• My changes don't require a change to the documentation, or if they do, I've added all required information.

[adrianeboyd]

Thanks for the useful PR, especially with docs! I'm not sure why this wasn't added much sooner.

Do you mind if I push to your branch directly to rephrase the docs a bit? No problem if you'd rather not, I could do it in a follow-up PR instead.

[evornov]

Yes, you can go ahead and push to my branch. Thanks!