Full threadsirnicolaz·Because the SWE benchmarks for LLM coding are done on python code bases, hence you are likely gonna have superior resultsView on HN