Abstractive-Text-Summarizer
所属分类：自然语言处理
开发工具：Jupyter Notebook
文件大小：0KB
下载次数：0
上传日期：2023-12-26 03:45:34
上 传 者：
sh-1993
说明： 使用编码器-解码器架构和自我注意机制的抽象文本摘要生成器。
(An Abstrative Text Summarizer using encoder decoder architecture and self attention mechanisms.)
文件列表:
AbstractiveTextSummarizer.ipynb
news.xlsx
# Abstractive Text Summarization
## Overview
This is an abstractive summarization model implemented using a transformer model with encoder decoder architecture, including self attention mechanisms on inshorts news dataset containing news headlines and articles.
## Requirements
- Python 3.x
- NLTK library (`nltk`)
- Google Colab (for the provided Google Drive integration)
## Execution
Upload the ipynb file and dataset to Google Colab and run all cells.
Rouge score is calculated. Results can be improved by changing the hyperparameters (increasing epochs and layers) i.e by performing more training which would require more computing power.
## Acknowledgments
https://arxiv.org/abs/1706.03762
Inshorts Dataset: https://www.kaggle.com/shashichander009/inshorts-news-data
The NLTK library: https://www.nltk.org/
Google Colab: https://colab.research.google.com/
近期下载者：
相关文件：
收藏者：