README.md

metadata

dataset_info:
  features:
    - name: query_id
      dtype: string
    - name: query
      dtype: string
    - name: positive_passages
      list:
        - name: docid
          dtype: string
        - name: text
          dtype: string
        - name: title
          dtype: string
    - name: negative_passages
      list:
        - name: docid
          dtype: string
        - name: text
          dtype: string
        - name: title
          dtype: string
    - name: subset
      dtype: string
  splits:
    - name: train
      num_bytes: 1778484241
      num_examples: 96167
  download_size: 1026965735
  dataset_size: 1778484241
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
task_categories:
  - text-ranking

This dataset contains data used in the paper Fixing Data That Hurts Performance: Cascading LLMs to Relabel Hard Negatives for Robust Information Retrieval. It focuses on improving the quality of training data for information retrieval models by identifying and relabeling false negatives using large language models.

Github repository: https://github.com/FlagOpen/rlhn