InfiCoder-Eval: A Comprehensive Benchmark for Evaluating Code Large Language Models' Question-Answering Capabilities

InfiCoder-Eval is a large-scale benchmark that systematically evaluates the free-form question-answering capabilities of code large language models across 15 programming languages and 5 major areas.