Data

Tools

Indexes

Applications

Experiments

Open Source Science

scholarly-llm-project

https://github.com/xinningcui/scholarly-llm-project

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ○
    Academic publication links
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (7.5%) to scientific vocabulary
Last synced: 9 months ago · JSON representation ·

Repository

Basic Info
  • Host: GitHub
  • Owner: XinningCui
  • Language: Python
  • Default Branch: main
  • Size: 54.7 KB
Statistics
  • Stars: 0
  • Watchers: 1
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created over 1 year ago · Last pushed over 1 year ago
Metadata Files
Readme Citation

README.md

ScholarLLM

Overview

ScholarLLM is a project aimed at developing a domain-specific Large Language Model (LLM) for the computer science (CS) research community. Our goal is to assist researchers in efficiently processing, understanding, and utilizing scientific knowledge.

Objectives

  • Generate a CS-focused LLM to aid researchers in scientific tasks.
  • Explore and compare different approaches for improving model performance.
  • Provide tools and models that can enhance the efficiency of research-related activities.

Methods

We explored three different strategies to achieve our objectives:

  1. Multi-task Fine-tuning

    • Fine-tune a single model to handle multiple tasks simultaneously.

  2. Instruction Tuning

    • Perform instruction-based fine-tuning for individual tasks.

  3. Retrieval-Augmented Generation (RAG)

    • Implement RAG to augment responses with relevant retrieved information from external knowledge sources.

Directory Structure

  • codemultitaskmodel: Code and resources related to multi-task fine-tuning.
  • instruction_tunning: Code and resources for instruction-based tuning.
  • rag: Code and resources for the implementation of RAG.

Usage

The models and scripts provided can be utilized to perform tasks such as document summarization, question answering, and simplification in the CS domain.

Acknowledgment

Insturction fine-tunning relies on Unsloth

Owner

  • Login: XinningCui
  • Kind: user

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: "Cui"
  given-names: "Xinning"
- family-names: "Luo"
  given-names: "Jing"
- family-names: "Abgaryan"
  given-names: "Meri"
title: "Scholar LLM"
version: 2.0.4
date-released: 2025-02-09
url: "https://github.com/XinningCui/Scholarly-LLM-project"

GitHub Events

Total
  • Member event: 2
  • Push event: 8
  • Create event: 3
Last Year
  • Member event: 2
  • Push event: 8
  • Create event: 3