README.md

metadata

datasets:
  - bs-la/xP3ru
license: bigscience-bloom-rail-1.0
model-index:
  - name: bloomz-7b1
    results:
      - task:
          type: Coreference resolution
        dataset:
          type: Muennighoff/xwinograd
          name: XWinograd (ru)
          config: ru
          split: test
          revision: 9dd5ea5505fad86b7bedad667955577815300cee
        metrics:
          - type: Accuracy
            value: 53.97
      - task:
          type: Natural language inference
        dataset:
          type: xnli
          name: XNLI (ru)
          config: ru
          split: validation
          revision: a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16
        metrics:
          - type: Accuracy
            value: 33.49
      - task:
          type: Sentence completion
        dataset:
          type: Muennighoff/xstory_cloze
          name: XStoryCloze (ru)
          config: ru
          split: validation
          revision: 8bb76e594b68147f1a430e86829d07189622b90d
        metrics:
          - type: Accuracy
            value: 48.64

Model Summary

bloom-7b1 finetuned on Russian multitask data. Hence the same as bloomz-7b1, but with only Russian finetuning data. 4b stands for 4 billion finetuning tokens (same as bloomz-7b1).

Citation

BLOOM+1 - TODO

@misc{muennighoff2022crosslingual,
      title={Crosslingual Generalization through Multitask Finetuning}, 
      author={Niklas Muennighoff and Thomas Wang and Lintang Sutawika and Adam Roberts and Stella Biderman and Teven Le Scao and M Saiful Bari and Sheng Shen and Zheng-Xin Yong and Hailey Schoelkopf and Xiangru Tang and Dragomir Radev and Alham Fikri Aji and Khalid Almubarak and Samuel Albanie and Zaid Alyafeai and Albert Webson and Edward Raff and Colin Raffel},
      year={2022},
      eprint={2211.01786},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}