Datasets:

devngho
/

korean-textbooks-edu

dataset_info:
  - config_name: raw
    features:
      - name: text
        dtype: string
      - name: score
        dtype: float64
    splits:
      - name: train
        num_bytes: 18413993666
        num_examples: 4414956
    download_size: 6813327957
    dataset_size: 18413993666
  - config_name: scored_over_2
    features:
      - name: text
        dtype: string
      - name: score
        dtype: float64
    splits:
      - name: train
        num_bytes: 16556122056.735634
        num_examples: 3969511
    download_size: 6199789090
    dataset_size: 16556122056.735634
  - config_name: scored_over_3
    features:
      - name: text
        dtype: string
      - name: score
        dtype: float64
    splits:
      - name: train
        num_bytes: 7237438964.033805
        num_examples: 1735255
    download_size: 2832546560
    dataset_size: 7237438964.033805
configs:
  - config_name: raw
    data_files:
      - split: train
        path: raw/train-*
  - config_name: scored_over_2
    data_files:
      - split: train
        path: scored_over_2/train-*
  - config_name: scored_over_3
    data_files:
      - split: train
        path: scored_over_3/train-*
    default: true
license: mit
source_datasets:
  - maywell/korean_textbooks
task_categories:
  - text-generation
language:
  - ko
size_categories:
  - 1M<n<10M

🇰🇷📚 korean-textbooks-edu

maywell/korean_textbooks의 모든 subset을 devngho/ko_edu_classifier_v2_nlpai-lab_KoE5 모델로 평가한 데이터셋

불러오기

from datasets import load_dataset

ds = load_dataset("devngho/korean-textbooks-edu", name="scored_over_3", split="train")

성능

예정

컴퓨팅

Google Cloud TPU, transformers, JAX, tpuswarm

하드웨어

TPU v4-8 x 4 instances, 약 2시간 소요

이 연구는 Google의 TPU Research Cloud (TRC)의 Cloud TPU 제공으로 수행되었습니다. ⚡

라이선스

원본 데이터셋(maywell/korean_textbooks)은 Apache 2.0 라이선스를 따릅니다. 이 데이터셋은 MIT 라이선스로 배포됩니다.