README.md

metadata

dataset_info:
  features:
    - name: query
      dtype: string
    - name: pos
      dtype: string
    - name: neg
      dtype: string
    - name: task_name
      dtype: string
    - name: query_instruct
      dtype: string
    - name: pos_instruct
      dtype: string
    - name: neg_instruct
      dtype: string
  splits:
    - name: train
      num_bytes: 2555303114
      num_examples: 1435000
  download_size: 1231001259
  dataset_size: 2555303114
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

MEDI dataset

This dataset was used in the paper GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning. Refer to https://arxiv.org/abs/2402.16829 for details.

The original dataset comes from the paper "One Embedder, Any Task: Instruction-Finetuned Text Embeddings" (https://arxiv.org/abs/2212.09741), which was used to train the INSTRUCTOR family of models (GitHub: https://github.com/xlang-ai/instructor-embedding).

The code for processing and publishing the raw data to HuggingFace Hub is available at https://github.com/avsolatorio/GISTEmbed.

Citation

GISTEmbed

@article{solatorio2024gistembed,
    title={GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning},
    author={Aivin V. Solatorio},
    journal={arXiv preprint arXiv:2402.16829},
    year={2024},
    URL={https://arxiv.org/abs/2402.16829}
    eprint={2402.16829},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

INSTRUCTOR

@inproceedings{INSTRUCTOR,
  title={One Embedder, Any Task: Instruction-Finetuned Text Embeddings},
  author={Su, Hongjin and Shi, Weijia and Kasai, Jungo and Wang, Yizhong and Hu, Yushi and  Ostendorf, Mari and Yih, Wen-tau and Smith, Noah A. and  Zettlemoyer, Luke and Yu, Tao},
  url={https://arxiv.org/abs/2212.09741},
  year={2022},
}