• Code Workflow :

  ─ Code

    ── lstm_attention.py
  
    _ About_news_dataset.py(EDA)
  
    _ generating_news.py
  
    - llama_news_generation.py
  
    - Llama_streamlit_news_generation_app.py
  
    - news_dataset_cleaning.py
  

  • LSTM + attention Implementation and Llama + aws bedrock:

  • Lstm_Attention.py: Code to handle data loading and initial preprocessing, Robust error handling for CSV readingCleans and filters text. Training and evaluation the LSTM model with Attention.

  • generating_loading_lstm_news_model.py: Code to load the model.pt file and the generate news.

  • About_news_dataset.py: code to load , check description and dataset information

  • llama_news_generation.py : code for generating news article using llama + aws bedrock + prompt engineering

  • Llama_streamlit_news_generation_app.py : streamlit code for news generating article

  • news_dataset_cleaning.py : this code is used for cleaning, preprocessing, handling error and filtering text , the output dataset is use as input for Llama Implementataion

• Setup and Implementation Guide for News Generation System Using Llma and AWS Bedrock

  - connect to ec2 instance
  - setup aws config file -  Add AWS credentials to config.ini
  - create a .env and activate the .env file
  -  install requirement.txt file Package on .env 
  -  Run the first Code- # Run the code - python llama_news_generation.py
  

  . Streamlit Implementation (Local Machine)

  -  Environment Setup:
  
      - Create virtual environment locally- python -m venv myenv : source myenv/bin/activate
      - download requirements.txt     
    -  Install requirements
        pip install -r requirements.txt
        - # Add:
            streamlit==1.24.0
              pandas
              boto3
              python-dotenv
              configparser
  
  
  -  Configuration Setup: make sure you Create  config directory :  aws .cong file
  
  -  Run the app; streamlit run Llama_streamlit_news_generation_app.py