Duong2005's picture
Upload README.md with huggingface_hub
8913642 verified
|
Raw History Blame Contribute Delete
2.35 kB
metadata
language:
  - vi
license: mit
task_categories:
  - question-answering
  - text-generation
tags:
  - python
  - vietnamese
  - instruction-tuning
  - llm-training
size_categories:
  - n<1K
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*

Bộ Dữ Liệu Huấn Luyện Mô Hình Ngôn Ngữ Hỏi - Đáp Lập Trình Python (Vietnamese Python QA)

Bộ dữ liệu gồm 100 mẫu hỏi - đáp chất lượng cao về ngôn ngữ lập trình Python, được thiết kế chuẩn hóa phục vụ huấn luyện (SFT / Instruction Fine-Tuning) các mô hình ngôn ngữ lớn (LLM).

1. Cấu trúc bộ dữ liệu

  • train: 70 mẫu (70%)
  • validation: 20 mẫu (20%)
  • test: 10 mẫu (10%)

Mỗi mẫu có định dạng:

  • instruction: Câu hỏi về lý thuyết hoặc thực hành Python.
  • input: Ngữ cảnh / mã đầu vào (hoặc rỗng).
  • output: Câu trả lời sư phạm, chính xác, có code Python minh họa.

2. Các chủ đề bao quát

  1. Cú pháp Python (Indentation, Comments, Docstrings, PEP 8)
  2. Biến & Đặt tên (Dynamic Typing, Type Hints, Scopes)
  3. Kiểu dữ liệu & Ép kiểu (int, float, complex, isinstance)
  4. Câu lệnh điều kiện (if-elif-else, ternary, match-case)
  5. Vòng lặp (for, while, break, continue, enumerate, zip)
  6. Hàm (def, *args, **kwargs, lambda, map/filter, generator yield)
  7. List (Slicing, Comprehension, sort, shallow/deep copy)
  8. Tuple (Immutability, Packing, Unpacking, namedtuple)
  9. Dictionary (Keys, Items, Get, Dict Comprehension, Merge)
  10. Set (Unique, Union, Intersection, Difference, frozenset)
  11. Chuỗi (f-strings, split, join, strip, find, replace)
  12. Xử lý ngoại lệ (try-except-else-finally, raise, Custom Exception)
  13. Đọc ghi file (open modes, with context manager, json, csv)
  14. Lập trình hướng đối tượng OOP (Class, Object, self, Inheritance, Encapsulation, Polymorphism, Property)
  15. Thư viện Python phổ biến (math, random, os, sys, re, datetime, requests, numpy, pandas)

3. Cách sử dụng với thư viện datasets

`python from datasets import load_dataset

dataset = load_dataset('Duong2005/python-qa-vietnamese') print(dataset) print(dataset['train'][0]) `