README.md

---
license: apache-2.0
task_categories:
- text-generation
tags:
- code
size_categories:
- 10K<n<100K
language:
- en
---

# Dataset description
This dataset consists of sequences of Python code followed by a a docstring explaining its function. It was constructed by concatenating code and text pairs 
from this [dataset](https://huggingface.co/datasets/codeparrot/github-jupyter-text-code-pairs) that were originally code and markdown cells in Jupyter Notebooks.

The content of each example the following:
````
[CODE]
"""
Explanation: [TEXT]
End of explanation
"""
[CODE]
"""
Explanation: [TEXT]
End of explanation
"""
...
````
# How to use it
```python
from datasets import load_dataset

ds = load_dataset("codeparrot/github-jupyter-code-to-text", split="train")
````
````
Dataset({
    features: ['repo_name', 'path', 'license', 'content'],
    num_rows: 47452
})
````