initial refactor of Reasoning Strategies

[StanChan03]

Purpose

Refactors Reasoning Strategies to be more intuitive. Fixes #198 . Additionally, it fixes some of the inconsistencies with sem_extract particularly its output parsing

Test Plan

Created comprehensive tests to test Reasoning with different operators and a mix of reasoning strategies.

Test Results

(Optional) Documentation Update

TODO: update docs on reasoning strategy

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• [] New feature (non-breaking change which adds functionality)
• [x ] Breaking change (fix or feature that would cause existing functionality to not work as expected)
  --> change to reasoning api
• Documentation update
• Performance improvement
• Refactoring (no functional changes)

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, updating docstrings
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[harshitgupta412]

instead of checking whether prompt_strategy is None, it will be better to just define a DEFAULT_PROMPT_STRATEGY (there are already default args in PromptStrategy, so PromptStrategy() will be enough)
The default argument will then be PromptStrategy() instead of None

[harshitgupta412]

If I understand correctly, this is supposed to basically generate demonstrations using the teacher model for the given data. In that case, it will be better to use sem_ops directly with teacher model instead of bootstrap functions

[harshitgupta412]

this will make sure that things are reusable and prevent double work in case some parameters change. (like adding a new prompt strategy or new prompts or new sem op)