Real-world uses

Created for this library

1. 1.

   A factual QA team labels its evaluation set as ORA so the grader checks for the exact correct answer.

2. 2.

   An LLM evaluation team treats arithmetic prompts as ORA and grades against the exact final value.

3. 3.

   A code-generation team uses ORA-style evaluation on coding benchmarks where correctness is whether tests pass.