Specifying numpy dtype in test functions

[ian-coccimiglio]

This PR contains:

• a new test-case for the benchmark
  • I hereby confirm that NO LLM-based technology (such as github copilot) was used while writing this benchmark
• new dependencies in requirements.txt
  • The environment.yml file was updated using the command conda env export > environment.yml
• new generator-functions allowing to sample from other LLMs
• new samples (sample_....jsonl files)
• new benchmarking results (..._results.jsonl files)
• documentation update
• bug fixes

Related github issue (if relevant): would close #115

Short description:

• I think it'd be best to specify a plausible data-type for images/arrays in the test-check.

How do you think will this influence the benchmark results?

• For Otsu's threshold + positive pixel counting, overall model pass-rate goes from 10/230 to 75/230.
• I haven't done other tasks yet. I would assume it would have a generally positive effect in cases where OpenCV is type-sensitive (applies to at least 2-3 more tests as far as I've seen).

Why do you think it makes sense to merge this PR?

• I'm waiting to see whether we decide to do this method or go for prompt-editing (or both). Not to be merged yet.
• Also it fixes the spelling of the test-case which I'm not sure we want to merge as I can't tell what it might break.

[ian-coccimiglio]

My bad, Github and I don't always cooperate.