I have carefully read the paper and the instructions in the official GitHub repository. However, I encountered an issue when trying to download the evaluation datasets: the links provided in the README (pointing to Self-RAG’s eval_data) are currently broken, and both the Google Drive links and the Self-RAG data links are inaccessible.
Could anyone please kindly share the evaluation datasets used in your paper (e.g., PopQA, PubHealth, Bio, Arc_challenge, etc.)? If there is an updated or alternative way to obtain them, I would greatly appreciate your guidance.
I have carefully read the paper and the instructions in the official GitHub repository. However, I encountered an issue when trying to download the evaluation datasets: the links provided in the README (pointing to Self-RAG’s eval_data) are currently broken, and both the Google Drive links and the Self-RAG data links are inaccessible.
Could anyone please kindly share the evaluation datasets used in your paper (e.g., PopQA, PubHealth, Bio, Arc_challenge, etc.)? If there is an updated or alternative way to obtain them, I would greatly appreciate your guidance.