cpp_unit_tests_processed_data_chat_format

The dataset primarily contains message content and role information, divided into training and test sets. The training set has 6,603 samples, and the test set has 826 samples. The total download size is 8,009,476 bytes, and the total size is 39,236,589 bytes. The default configuration stores training files under `data/train-*` and test files under `data/test-*`.