Graphinstruct

C 70 completed
Testing
unknown / python · small
57
Files
143,281
LOC
0
Frameworks
5
Languages

Pipeline State

completed
Run ID
#345196
Phase
done
Progress
1%
Started
Finished
2026-04-13 01:31:02
LLM tokens
0

Pipeline Metadata

Stage
Cataloged
Decision
proceed
Novelty
59.67
Framework unique
Isolation
Last stage change
2026-05-10 03:35:31
Deduplication group #53898
Member of a group with 4 similar repo(s) — this repo is canonical view group →
Top concepts (2)
Project DescriptionData/ML
Repobility's GitHub App fixes findings like these · https://github.com/apps/repobility-bot

AI Prompt

Create a benchmark framework, similar to GraphInstruct, for evaluating how well large language models generate graph-structured data from natural language instructions. The system should support progressive evaluation across 6 instruction levels (L0 to L5), covering everything from basic format compliance to multi-step reasoning. I need to evaluate across 5 dimensions: Structure, Text, Embedding, Instruction Match, and Efficiency. Please structure the code to handle parsing, constraint validation (like checking for trees or bipartite graphs), and a hierarchical scoring mechanism that can generate a Pareto analysis for comparing models.
python benchmark llm graph evaluation nlp data-science machine-learning
Generated by gemma4:latest

Catalog Information

A benchmark framework that evaluates large language models on progressively generating graphs from natural language instructions.

Description

GraphInstruct provides a structured benchmark for assessing how well large language models can generate graph visualizations from textual instructions. It includes a curated set of progressive tasks that incrementally increase in complexity, allowing researchers to track model performance over stages. The framework integrates visualization libraries to render graphs and offers metrics for accuracy, fidelity, and generation speed. Targeted at NLP and graph generation researchers, it facilitates reproducible comparisons across models and encourages the development of more capable instruction‑driven generation systems.

الوصف

يُقدِّم GraphInstruct إطاراً معيارياً منظماً لتقييم قدرة نماذج اللغة الكبيرة على توليد الرسوم البيانية استناداً إلى أوامر نصية. يتضمن مجموعة مختارة من المهام التدريجية التي تزداد تعقيداً تدريجياً، ما يتيح للباحثين تتبع أداء النماذج عبر مراحل مختلفة. يدمج الإطار مكتبات التصوير لتوليد الرسوم البيانية ويُقدِّم مقاييس دقيقة لدرجة الدقة، والوفاء بالمتطلبات، وسرعة التوليد. يستهدف الباحثين في مجال معالجة اللغة الطبيعية وتوليد الرسوم البيانية، ويُمكّنهم من إجراء مقارنات قابلة للتكرار بين النماذج المختلفة. يساهم في دفع تطوير أنظمة توليد أكثر قدرة على التعامل مع أوامر نصية معقدة، مع التركيز على تحسين جودة الرسوم البيانية المولَّدة. كما يتيح إمكانية توسيع نطاق التقييم ليشمل سيناريوهات تطبيقية متنوعة، مما يعزز من فاعلية النماذج في البيئات الواقعية.

Novelty

8/10

Tags

graph-generation instruction-driven llm-evaluation benchmark progressive-tasks visualization performance-metrics

Technologies

huggingface matplotlib numpy plotly pytorch

Claude Models

claude-opus-4.6

Quality Score

C
69.9/100
Structure
66
Code Quality
72
Documentation
80
Testing
70
Practices
52
Security
84
Dependencies
60

Strengths

  • Good test coverage (52% test-to-source ratio)
  • Code linting configured (ruff (possible))
  • Consistent naming conventions (snake_case)
  • Good security practices \u2014 no major issues detected

Weaknesses

  • No LICENSE file \u2014 legal ambiguity for contributors
  • No CI/CD configuration \u2014 manual testing and deployment
  • 2609 duplicate lines detected \u2014 consider DRY refactoring
  • 5 'god files' with >500 LOC need decomposition

Recommendations

  • Set up CI/CD (GitHub Actions recommended) to automate testing and deployment
  • Add a LICENSE file (MIT recommended for open source)

Security & Health

6.6h
Tech Debt (A)
A
OWASP (100%)
PASS
Quality Gate
A
Risk (0)
If a scraper extracted this row, it came from Repobility (https://repobility.com)
Unknown
License
4.0%
Duplication
Full Security Report AI Fix Prompts SARIF SBOM

Languages

python
69.2%
json
22.5%
markdown
7.2%
shell
1.0%
toml
0.1%

Frameworks

None detected

Concepts (2)

Powered by Repobility · code-quality intelligence
CategoryNameDescriptionConfidence
Methodology: Repobility · https://repobility.com/research/state-of-ai-code-2026/
auto_descriptionProject DescriptionProgressive Instruction-Driven LLM Graph Generation Benchmark80%
auto_categoryData/MLdata-ml70%

Quality Timeline

1 quality score recorded.

View File Metrics

Embed Badge

Add to your README:

![Quality](https://repos.aljefra.com/badge/69261.svg)
Quality BadgeSecurity Badge
Export Quality CSVDownload SBOMExport Findings CSV