https://github.com/alokamgnaneswarasai/p5-finetuning
This repository contains the implementation of recommendation systems utilizing Large Language Models (LLMs) for rating prediction and sequential recommendation tasks and review summarization , explanation generation etc
Science Score: 13.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
○.zenodo.json file
-
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (7.3%) to scientific vocabulary
Keywords
Repository
This repository contains the implementation of recommendation systems utilizing Large Language Models (LLMs) for rating prediction and sequential recommendation tasks and review summarization , explanation generation etc
Basic Info
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Topics
Metadata Files
README.md
Recommendation Systems using Large Language Models (LLM)
Table of Contents
Project Overview
This project focuses on implementing recommendation systems using large language models (LLMs). Specifically, we employ the P5 recommendation model for rating prediction and sequential recommendation tasks. Additionally, we explore the transformer-decoder architecture, including fine-tuning with GPT-2 and traditional RNN models.
Models and Architectures
P5 Recommendation Model
The P5 model is utilized for two primary tasks: 1. Rating Prediction: Predicting user ratings for items. 2. Sequential Recommendation: Recommending the next item in a user's sequence of interactions.
Transformer-Decoder Architecture
We experiment with the transformer-decoder architecture, fine-tuning it using GPT-2 for enhanced sequential recommendation performance. This architecture helps in capturing long-term dependencies in user interaction sequences.
Recurrent Neural Network (RNN) Models
Traditional RNN models are also implemented to compare their performance with the transformer-based approaches. RNNs are useful for modeling sequential data, though they may struggle with long-term dependencies compared to transformer models.
Datasets
The models are evaluated on various datasets, including: - MovieLens - Amazon Beauty - Delicious
These datasets provide a diverse set of user-item interactions, enabling comprehensive evaluation of the models' performance.
Owner
- Login: alokamgnaneswarasai
- Kind: user
- Repositories: 1
- Profile: https://github.com/alokamgnaneswarasai
GitHub Events
Total
- Watch event: 1
Last Year
- Watch event: 1
Dependencies
- numpy *
- pandas *
- scikit-learn *
- scipy *