---
license: apache-2.0
language:
  - ja
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: input
    dtype: string
  - name: instruction
    dtype: string
  - name: output
    dtype: string
  - name: input_ppl
    dtype: int64
  - name: instruction_ppl
    dtype: int64
  - name: output_ppl
    dtype: int64
  - name: full_ppl
    dtype: int64
  splits:
  - name: train
    num_bytes: 60856874
    num_examples: 55359
  download_size: 27216157
  dataset_size: 60856874
---

** 以下のrepositoryのforkです。 ** 

https://huggingface.co/datasets/kunishou/oasst1-89k-ja


instructionとinput、outputにまとめ、kenllmでperplexityのスコアが付与してあります。

perplexityの計算に用いたtokenizerはこちら

https://huggingface.co/if001/sentencepiece_ja


- instruction_ppl: instructionのみのperplexity
- output_ppl: outputのみのperplexity
- full_ppl: instructionとoutputを合わせ、instruction用の文章にしたperplexity