# Lyric_LLM_Project

Aim:

Developing an RAG using the Langchain LLMs and HuggingFaceHub model to perform sentiment analysis on BTS songs lyrics translations and suggest a playlist and set of quotes from the songs according to the user's emotions.

Libraries Required:

• Langchain
• NLTK
• HuggingFaceHub
• Streamlit
• Beautifulsoup
• Pandas
• JSON

Procedure:

1. Scrape data from to obtain all BTS songs and their lyrics with translations
2. Clean the scraped data and extract only the translations into a JSON
3. Create two Langchain LLMs using the HuggingFaceHub model to perform sentimental analysis and obtain
   • emotion for every line translation for every song in the JSON
   • the overall emotion of each song
4. Save the results in a consolidated dataset
5. Using Streamlit obtain user input for any emotions and derive the appropriate playlist and quotes for the user based on the emotions

How to run:

Clone the repository and run the emotion_chatbot.py file using Streamlit

streamlit run emotion_chatbot.py

Output