* Adjust the parameters of gpt-40-mini model, custom for QA mode Add chunkSize setting input * Adjust the parameters of gpt-40-mini model