README.md

metadata

dataset_info:
  features:
    - name: max_stars_repo_path
      dtype: large_string
    - name: max_stars_repo_name
      dtype: large_string
    - name: id
      dtype: large_string
    - name: language
      dtype: large_string
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 2407352255
      num_examples: 206644
  download_size: 894488166
  dataset_size: 2407352255
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Code Textbook

200K+ Synthetic Textbook Samples generated with various Open-Source LLMs including Nous Hermes Mixtral 8x7B, OpenHermes-2.5-Mistral, OpenChat and DeepSeek-Coder.