Automated Unit Test Generation via Chain of Thought Prompt and Reinforcement Learning

IntroductionThis is the replication package for the paper "Automated Unit Test Generation via Chain of Thought Prompt and Reinforcement Learning".Organization of the Replication Package<b>checkpoints.zip:</b> fine-tuned models, including TestCTRL, TestCT, TestCT-no-cot, TestCT-intention, TestCT-input, TestCT-ti, CodeBERT-line, CodeT5-line, CodeGPT-line, CodeBERT-branch, CodeT5-branch, and CodeGPT-branch.<b>dataset.zip</b>: Datasets for fine-tuning and reinforcement learning, including the CoT dataset, reward dataset (reward folder), and the dataset for PPO optimization (rl folder).<b>evaluation.zip:</b> scripts for evaluating the generated tests, including CodeBLEU, syntactic correct rate, compilation passing rate, line coverage rate, and branch coverage rate.<b>finetune.zip: </b>scripts and configs for fine-tuning large language models for test generation.<b>generated_test_result.zip:</b> the generated tests.<b>pretrain.zip:</b> pre-trained models, including CodeLlama, CodeBERT, CodeT5, and CodeBERT.<b>CoT_quality.zip: </b>the example of evaluating CoT prompts.