JayeshSharma's picture
Update README.md
0dad044 verified
|
Raw
History Blame Contribute Delete
1.59 kB
metadata
license: other
license_name: psf-2.0
pretty_name: Python Documentation Training Dataset
language:
  - en
tags:
  - python
  - code
  - documentation

Python Official Documentation Training Dataset

An Apache Arrow formatted, tokenized dataset created directly from the official Python Documentation (500+ pages). This dataset is optimized for training and fine-tuning language models on core Python concepts, standard library usage, syntax rules, and official programming guidelines.


Dataset Overview

  • Dataset Name: python-training-dataset
  • Source Material: Official Python Documentation (500+ pages)
  • Format: Apache Arrow (data-00000-of-00001.arrow)
  • License: Python Software Foundation License (psf-2.0)
  • Primary Feature: Pre-tokenized sequence arrays (input_ids)

Dataset Structure

Data Schema

The dataset contains pre-tokenized token ID lists designed for immediate ingestion into transformer-based neural network models:

Feature Data Type Description
input_ids List(int32) Tokenized integer sequence representations derived from Python's official documentation

Quickstart & Loading

You can load this dataset directly using the Hugging Face datasets library:

from datasets import load_dataset

# Load dataset from Hugging Face Hub
dataset = load_dataset("JayeshSharma/python-training-dataset")

# Inspect dataset structure
print(dataset)

# Access a single tokenized sequence
sample = dataset["train"][0]
print("Token IDs sample:", sample["input_ids"][:10])