README.md

metadata

dataset_info:
  features:
    - name: instruction
      dtype: string
    - name: input
      dtype: string
    - name: solvability
      dtype: string
    - name: output
      dtype: string
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 4033508
      num_examples: 314
    - name: test
      num_bytes: 953481
      num_examples: 79
  download_size: 2265604
  dataset_size: 4986989
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*

Instruction Tuning Dataset (Split)

This dataset is a copy of the alvwjy/instruct_dataset. This is the split version of the dataset to training and testing which will be used for measuring the performance of the model.