Chinese-Llama-2 is a large language model project that enhances the Llama-2 base model for Chinese language understanding and generation. It employs multiple fine-tuning approaches including LoRA fine-tuning, full-parameter instruction fine-tuning, and secondary pre-training. The project provides model weights, training datasets, and documentation to enable researchers and developers to use or further develop the Chinese-enhanced LLM.