{"id":30922804,"url":"https://github.com/kevindebenedetti/dataset-generator","last_synced_at":"2026-01-28T00:04:09.818Z","repository":{"id":312884032,"uuid":"1046564217","full_name":"KevinDeBenedetti/dataset-generator","owner":"KevinDeBenedetti","description":"Automated web-scraping to LLM-powered question-answer dataset generator with duplicate detection and optional Langfuse export.","archived":false,"fork":false,"pushed_at":"2026-01-10T22:28:26.000Z","size":6142,"stargazers_count":2,"open_issues_count":5,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-01-11T06:50:38.111Z","etag":null,"topics":["fastapi","nextjs","openai"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/KevinDeBenedetti.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-08-28T21:54:20.000Z","updated_at":"2026-01-10T22:27:29.000Z","dependencies_parsed_at":"2025-09-02T16:32:03.929Z","dependency_job_id":"2a202f0e-045b-40b1-8298-a1fafb22aab7","html_url":"https://github.com/KevinDeBenedetti/dataset-generator","commit_stats":null,"previous_names":["kevindebenedetti/dataset-generator"],"tags_count":13,"template":false,"template_full_name":null,"purl":"pkg:github/KevinDeBenedetti/dataset-generator","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KevinDeBenedetti%2Fdataset-generator","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KevinDeBenedetti%2Fdataset-generator/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KevinDeBenedetti%2Fdataset-generator/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KevinDeBenedetti%2Fdataset-generator/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/KevinDeBenedetti","download_url":"https://codeload.github.com/KevinDeBenedetti/dataset-generator/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KevinDeBenedetti%2Fdataset-generator/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28478047,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-16T06:30:42.265Z","status":"ssl_error","status_checked_at":"2026-01-16T06:30:16.248Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["fastapi","nextjs","openai"],"created_at":"2025-09-10T03:10:58.515Z","updated_at":"2026-01-16T08:01:14.672Z","avatar_url":"https://github.com/KevinDeBenedetti.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Dataset Generator\n\n[![CI](https://github.com/KevinDeBenedetti/dataset-generator/workflows/CI/badge.svg)](https://github.com/KevinDeBenedetti/dataset-generator/actions)\n[![codecov](https://codecov.io/gh/KevinDeBenedetti/dataset-generator/graph/badge.svg)](https://codecov.io/gh/KevinDeBenedetti/dataset-generator)\n\nWeb scraping and automatic dataset generation tool for question-answer datasets with advanced export capabilities and LLM integration.\n\n## 🎯 Objective\n\nCreate quality datasets for training AI models by automatically scraping reliable sources and generating contextualized question-answer pairs. Export datasets to multiple formats including Langfuse for training data management.\n\n## ⚡ Quick Start\n\n```bash\n# Configuration\ncp .env.example .env\n# Edit .env with your API keys\n\n# Launch\nmake start\n```\n\n## 🏗️ Architecture\n\nThis project is designed with a modular architecture that separates concerns into distinct components:\n\n- **Scraper**: Retrieval of web content from specified URLs\n- **LLM Client**: Interaction with language models to generate question-answer pairs\n- **Data Manager**: Data management and dataset storage with multiple export formats\n- **Pipeline**: Orchestration of the complete dataset generation process\n- **Export Module**: Advanced dataset export to various platforms (Langfuse, JSON, CSV, etc.)\n\n## ✨ Key Features\n\n- **Multi-source Scraping**: Support for various web sources and content types\n- **AI-Powered QA Generation**: Leverage state-of-the-art LLMs for intelligent question-answer pair creation\n- **Multi-language Support**: Generate datasets in French, English, Spanish, and German\n- **Langfuse Integration**: Direct export to Langfuse for dataset management and training workflows\n- **Multiple Export Formats**: JSON, CSV, JSONL, and platform-specific formats\n- **Quality Control**: Automated validation and filtering of generated content\n- **Batch Processing**: Efficient handling of large-scale data generation\n- **API Interface**: RESTful API for programmatic access and integration\n\n## 🔄 Workflow\n\n1. **Scraping**: Retrieving raw web data from multiple sources\n2. **Cleaning**: Processing and normalizing text to extract relevant content\n3. **QA Generation**: Creating high-quality question-answer pairs via LLMs with configurable prompts\n4. **Quality Assurance**: Automated validation and filtering of generated datasets\n5. **Export**: Multi-format export including Langfuse integration for seamless training workflows\n6. **Storage**: Persistent storage with metadata tracking and version control\n\n## 📊 Export Options\n\n- **Langfuse**: Direct integration for training data management\n- **JSON/JSONL**: Standard formats for data interchange\n- **CSV**: Tabular format for analysis and review\n- **Custom Formats**: Extensible export system for specific requirements\n\n## 🔧 Configuration\n\nThe tool supports extensive configuration options for:\n\n- LLM model selection and parameters\n- Export format preferences\n- Quality thresholds and validation rules\n- Batch processing settings\n- API rate limiting and retry policies\n\n## 🌍 Supported Languages\n\n- **French (fr)**: French language dataset generation\n- **English (en)**: English language dataset generation\n- **Spanish (es)**: Spanish language dataset generation\n- **German (de)**: German language dataset generation\n\n## 🧪 Testing \u0026 Coverage\n\nThis project maintains high test coverage to ensure code quality and reliability.\n\n```bash\n# Run tests with coverage (HTML report)\nmake test\n\n# Run tests for CI (XML report, enforces 70% minimum)\nmake test-ci\n\n# Run pre-commit hooks (includes tests on push)\nuv run prek run --all-files\n```\n\n### Coverage Reports\n\n- **Local**: After running tests, view `htmlcov/index.html` for detailed coverage report\n- **CI/CD**: Coverage reports are automatically generated and uploaded on every PR\n- **Codecov**: [View detailed coverage on Codecov](https://codecov.io/gh/KevinDeBenedetti/dataset-generator)\n\nCurrent coverage threshold: **70%** minimum required for CI to pass\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkevindebenedetti%2Fdataset-generator","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkevindebenedetti%2Fdataset-generator","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkevindebenedetti%2Fdataset-generator/lists"}