COIG-P: Chinese Preference Dataset for Human Value Alignment

A Chinese preference dataset developed for alignment with human values, as described in the associated research paper. The dataset was created by author m-a-p and was last updated on HuggingFace on 2025-04-15. Its specific scale and content are detailed in the paper 'COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values'.

Use Cases

• Fine-tuning language models for human preference alignment based on the described preference data
• Training reward models for reinforcement learning from human feedback (RLHF) using Chinese language annotations
• Benchmarking model performance on Chinese value-alignment tasks as referenced in the paper

Strengths

• Described as 'high-quality' and 'large-scale' in the associated research paper
• Specifically designed for the task of alignment with human values

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment

Provenance

Source

huggingface

Collection Method

Created for the paper 'COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values'.

Freshness

Last updated 2025-04-15 12:31:56; freshness should be verified