Summary of Prompting and Fine-tuning Of Small Llms For Length-controllable Telephone Call Summarization, by David Thulke and Yingbo Gao and Rricha Jalota and Christian Dugast and Hermann Ney
Prompting and Fine-Tuning of Small LLMs for Length-Controllable Telephone Call Summarization
by David Thulke, Yingbo Gao, Rricha Jalota, Christian Dugast, Hermann Ney
First submitted to arxiv on: 24 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents a novel approach to developing a telephone call summarization system using large language models (LLMs). The authors experiment with prompting existing LLMs to generate summaries of phone conversations, creating a tailored synthetic training dataset using stronger frontier models. They focus on generating diverse data and controlling summary length for various use-cases. Evaluation is performed using two state-of-the-art evaluation techniques, demonstrating that the fine-tuned Llama-2-7B-based summarization model performs similarly to GPT-4 in terms of factual accuracy, completeness, and conciseness. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a special system for summarizing phone calls. They use super-smart computers called large language models (LLMs) to make summaries of phone conversations. The LLMs are trained on special data that is made just for this task. The authors want the summaries to be different and flexible, so they can fit different needs. To check how good it is, they used two top-of-the-line methods to test the summaries. They found out that their system works almost as well as a super-powerful one called GPT-4. |
Keywords
» Artificial intelligence » Gpt » Llama » Prompting » Summarization
