Standalone inference pipeline

What does this merge request do and why?

This MR introduce a new pipeline (and cli command) called inference to duo-chat. This new pipeline will only query the models to get all answers, and write the answers to any output sinks.

• The pipeline uses the same config system, and the config for inference is very similar to the existing evaluation config
• Move all functions related to model inference from evaluate.py to this new pipeline, then inside evaluate.py we import those functions to keep the code DRY.

Ref: #210 (closed)

Merge request checklist

• I've ran the affected pipeline(s) to validate that nothing is broken.
• Tests added for new functionality. If not, please raise an issue to follow up.
• Documentation added/updated, if needed.