README.md

metadata

dataset_info:
  features:
    - name: INSTRUCTION
      dtype: string
    - name: RESPONSE
      dtype: string
    - name: SOURCE
      dtype: string
  splits:
    - name: train
      num_bytes: 304315
      num_examples: 591
  download_size: 0
  dataset_size: 304315

Dataset Card for "humaneval-mbpp-testgen-qa"

This dataset contains prompt-reply (question-answer) pairs where the prompt is to create a Python unit tests which tests for the functionality described in a specific docstring. The responses are then the generated unit tests.