LLM-KT

A Versatile Framework for Knowledge Transfer from Large Language Models to Collaborative Filtering.

Installation

1. Clone the repository

2. Create a virtual environment and activate it

3. Install the required dependencies:

   poetry install

Usecase

Step 1. Prepare embedding profiles

Based on a descrptions of user and item in given dataset derive an embeddings of profiles by using LLM.

Basic guideline here

Step 2. Setting up and lunching experiments with knowledge transfering

Use knowledge_transfer_framework to setup and lunch experiments with different models.

Basic guideline here