feat: token counting endpoint

[gphorvath]

Type: Feature

Description: Enhance the API to allow users to count tokens for text, so that prompts can be cut off at accurate points for a local LLM.

User Story: As an API user, I want the ability to count tokens for a given text input, so that I can accurately truncate prompts at specific token counts when working with a local Language Model (LLM).

Acceptance Criteria:

• [ ] Implement a new endpoint in the API that accepts a text input and returns the token count.
• [ ] The token counting method should be consistent with the tokenization approach used by the local LLM.
• [ ] The API should handle various input lengths and provide an accurate token count for each input.
• [ ] The API response should include the original text input and the corresponding token count.
• [ ] Provide clear documentation and usage examples for the new token counting feature in the API

[CollectiveUnicorn]

Putting this on hold again, the token count will be returned by the completion API as a part of https://github.com/defenseunicorns/leapfrogai/pull/493 but there won't be another endpoint for the just counting just yet.