https://github.com/aadisrivastava05/pravachakai-hindi-article-generation

This project is a Hindi Article Generator, developed using natural language processing (NLP) techniques to create contextually relevant and coherent articles in Hindi. The model was trained on a custom dataset collected via web scraping. The model has been fine-tuned on a dataset of Hindi news headlines and articles.
https://github.com/aadisrivastava05/pravachakai-hindi-article-generation

fine-tuning huggingface-transformers llama3 nlp unsloth

Last synced: about 2 months ago
JSON representation

This project is a Hindi Article Generator, developed using natural language processing (NLP) techniques to create contextually relevant and coherent articles in Hindi. The model was trained on a custom dataset collected via web scraping. The model has been fine-tuned on a dataset of Hindi news headlines and articles.

• Host: GitHub
• URL: https://github.com/aadisrivastava05/pravachakai-hindi-article-generation
• Owner: AadiSrivastava05
• License: mit
• Created: 2024-09-05T10:25:46.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2024-09-05T11:21:14.000Z (4 months ago)
• Last Synced: 2024-10-19T22:15:26.111Z (3 months ago)
• Topics: fine-tuning, huggingface-transformers, llama3, nlp, unsloth
• Language: Jupyter Notebook
• Homepage: https://www.kaggle.com/code/aadisrivastava/hindi-article-generator
• Size: 42 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Hindi Article Generator📰

This project is a Hindi Article Generator, developed using natural language processing (NLP) techniques to create contextually relevant and coherent articles in Hindi. The model was trained on a custom dataset collected via web scraping. The model has been fine-tuned on a dataset of Hindi news headlines and articles to generate high-quality, fluent articles for a variety of use cases such as news, blogs, and creative content.

## Direct link to Kaggle Notebook with outputs https://www.kaggle.com/code/aadisrivastava/hindi-article-generator

(I would highly recommend you to check it out, as it has all the cells including the outputs)

# 📁 Dataset

The model is trained on a dataset that I created through web scraping from the BBC Hindi website. The dataset includes a wide variety of articles, ensuring that the generator produces diverse and contextually accurate content.

## Here is the link to the Dataset repo containing the dataset and scraping code:-

https://github.com/AadiSrivastava05/BBC-Hindi-News-Dataset-with-web-scraping-script

## The dataset is also available on Kaggle which you can use directly in your code without downloading:-

https://www.kaggle.com/datasets/aadisrivastava/bbc-hindi-news-articles-dataset-detailed

# 📌 Features

* Generates fluent and contextually accurate Hindi articles.

* Can be used for content creation in media platforms, blogs, and creative writing.

* Built with advanced NLP techniques and fine-tuned for the Hindi language on **llama 3**.

* This same dataset and approach can be used for many other tasks, some of them I am listing below:-

 1) Generating a headline for an article

 2) Classification of an article into different categories

 3) Classification of an article headline into different categories