https://github.com/bibek36/dialogue-summarization-with-generative-ai
Welcome to the Dialogue Summarization with Generative AI project! In this project, your main goal is to perform dialogue summarization using cutting-edge language models and investigate how different input techniques impact the quality of generated summaries.
https://github.com/bibek36/dialogue-summarization-with-generative-ai
datapreprocessing genai jupyter-notebook large-language-models machine-learning prompt-engineering python
Last synced: about 1 month ago
JSON representation
Welcome to the Dialogue Summarization with Generative AI project! In this project, your main goal is to perform dialogue summarization using cutting-edge language models and investigate how different input techniques impact the quality of generated summaries.
- Host: GitHub
- URL: https://github.com/bibek36/dialogue-summarization-with-generative-ai
- Owner: bibek36
- Created: 2024-12-04T11:43:28.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-04T11:47:01.000Z (over 1 year ago)
- Last Synced: 2025-04-07T15:23:56.624Z (about 1 year ago)
- Topics: datapreprocessing, genai, jupyter-notebook, large-language-models, machine-learning, prompt-engineering, python
- Language: Jupyter Notebook
- Homepage:
- Size: 10.7 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Dialogue Summarization with Generative AI
Welcome to the **Dialogue Summarization with Generative AI** project! In this project, your main goal is to perform dialogue summarization using cutting-edge language models and investigate how different input techniques impact the quality of generated summaries.
## Project Overview
In this project, you will embark on a journey to understand and harness the power of generative AI for dialogue summarization. By leveraging large language models, you'll explore how variations in input text influence the model's output. Through the process of prompt engineering, you'll fine-tune the model's behavior to align with your summarization task. Specifically, you will:
1. **Investigate Input-Output Relationship:** Study how the content and structure of input dialogues influence the quality and relevance of the generated summaries.
2. **Prompt Engineering:** Develop and refine techniques to construct prompts that guide the model towards producing desired summarization outputs.
3. **Comparison of Inference Techniques:** Compare and analyze the performance of different inference strategies, including zero-shot, one-shot, and few-shot inferences, to gain insights into their effectiveness for dialogue summarization.
4. **Enhancing Generative Output:** Take the first steps in harnessing prompt engineering to enhance the output quality of large language models, paving the way for improved generative AI applications.
## Project Structure
The project is structured into several key components:
- **Data Collection and Preprocessing:** Gather and preprocess a dataset of dialogues for training and evaluation.
- **Model Selection and Fine-tuning:** Choose a suitable generative AI model and fine-tune it for dialogue summarization.
- **Prompt Engineering:** Develop prompt engineering strategies to guide the model's behavior and align it with the summarization task.
- **Inference and Evaluation:** Implement various inference techniques (zero-shot, one-shot, few-shot) and evaluate their impact on summary quality.
- **Analysis and Insights:** Compare results, analyze the model's behavior, and draw conclusions about the effectiveness of different input techniques and prompt engineering.
## Getting Started
Follow these steps to get started with the project:
1. Copy the link of .ipynb file and open it in colab or your preferred IDE (you can use nbviewer if you want to see the project only).
2. Begin experimenting with different input variations, prompt engineering techniques, and inference strategies.
3. Document your progress, code, and findings in this repository.
## Contribution Guidelines
Contributions to this project are welcome! If you find improvements, fixes, or exciting insights, please feel free to submit a pull request. Before submitting, make sure to review the contribution guidelines outlined in `CONTRIBUTING.md`.
## Acknowledgments
This project is part of the Lab Assignment from the course "Deep Learning" for my Masters in Data Science from IIT Jammu.