Evaluating large language models trained on code

The article appears to have an empty body, with only the title 'Evaluating large language models trained on code' provided. Without the actual content, no meaningful analysis of LLM evaluation methods or findings can be conducted.