# RAG Pipeline
RAG Pipeline, Retrieval-Augmented Generation (RAG) mimarisi ile metin tabanlı arama ve yanıt üretme işlemlerini kolaylaştıran bir Python kütüphanesidir.
Bu kütüphane ile belgelerinizi indeksleyebilir, FAISS veya diğer vektör veri tabanlarıyla arama yapabilir ve LLM modelleri ile entegre edebilirsiniz.
---
## 📦 Kurulum
PyPI üzerinden:
```bash
pip install rag-pipeline-lib
```
Yerel geliştirme modu (kaynak kodu değiştirip test etmek için):
```bash
git clone https://github.com/kullaniciadi/rag_pipeline.git
cd rag_pipeline
pip install -e .
```
---
## 🚀 Quickstart
```python
from rag_pipeline import RAGPipeline, FixedSizeChunking
from rag_pipeline.document_loaders import PDFLoader
from rag_pipeline.embeddings import OllamaEmbeddings
from rag_pipeline.llms import OllamaLLM
from rag_pipeline.vector_stores import FAISSVectorStore
# Bileşenleri başlat
embedding = OllamaEmbeddings("nomic-embed-text:latest", base_url="http://ollama:11434")
vector_store = FAISSVectorStore(collection_name="my_collection", dimension=768)
llm = OllamaLLM("llama3.2:3b", base_url="http://ollama:11434")
chunking = FixedSizeChunking(chunk_size=450, overlap=100)
# Pipeline oluştur
rag = RAGPipeline(vector_store, embedding, llm, chunking)
# Belgeleri yükle
docs = PDFLoader.load_folder("./documents")
rag.add_documents(docs)
# Sorgu yap
print(rag.query("What are the feature methods used in cattle identification?"))
```
---
## Proje Yapısı
```
rag_pipeline/
│
├── rag_pipeline/
│ ├── __init__.py
│ ├── pipeline.py
│ ├── chunking/
│ ├── document_loaders/
│ ├── embeddings/
│ ├── llms/
│ ├── vector_stores/
│ └── retrievers/
│
├── tests/
│
├── setup.py
├── pyproject.toml
├── LICENSE
└── README.md
```
---
## Lisans
Bu proje **MIT Lisansı** ile lisanslanmıştır.
Tüm detaylar için [LICENSE](LICENSE) dosyasına bakabilirsiniz.
```
MIT License
Copyright (c) 2025 İsim
Permission is hereby granted, free of charge, to any person obtaining a copy...
```
---
## Özellikler
- FAISS ve diğer vektör veritabanı desteği
- Hibrit retrieval desteği (BM25 + vektör arama)
- Modüler yapı
- Geliştirici dostu API
---
## Katkıda Bulunma
1. Bu projeyi forklayın
2. Yeni bir branch oluşturun (`git checkout -b feature/ozellik`)
3. Değişikliklerinizi commit edin (`git commit -m 'Yeni özellik eklendi'`)
4. Branch’inizi push edin (`git push origin feature/ozellik`)
5. Pull Request oluşturun
---
## İletişim
Sorularınız veya önerileriniz için: **ornek@email.com**
Raw data
{
"_id": null,
"home_page": null,
"name": "rag-pipeline-lib",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "rag, retrieval, llm, nlp, vector, bm25, hybrid",
"author": null,
"author_email": "Ad\u0131n\u0131z Soyad\u0131n\u0131z <mail@ornek.com>",
"download_url": "https://files.pythonhosted.org/packages/df/bf/9d5525302032d04d296e73d13bc71c2751b8909bebbb32692d63f48e5637/rag_pipeline_lib-0.1.1.tar.gz",
"platform": null,
"description": "# RAG Pipeline\n\nRAG Pipeline, Retrieval-Augmented Generation (RAG) mimarisi ile metin tabanl\u0131 arama ve yan\u0131t \u00fcretme i\u015flemlerini kolayla\u015ft\u0131ran bir Python k\u00fct\u00fcphanesidir. \nBu k\u00fct\u00fcphane ile belgelerinizi indeksleyebilir, FAISS veya di\u011fer vekt\u00f6r veri tabanlar\u0131yla arama yapabilir ve LLM modelleri ile entegre edebilirsiniz.\n\n---\n\n## \ud83d\udce6 Kurulum\n\nPyPI \u00fczerinden:\n```bash\npip install rag-pipeline-lib\n```\n\nYerel geli\u015ftirme modu (kaynak kodu de\u011fi\u015ftirip test etmek i\u00e7in):\n```bash\ngit clone https://github.com/kullaniciadi/rag_pipeline.git\ncd rag_pipeline\npip install -e .\n```\n\n---\n\n## \ud83d\ude80 Quickstart\n\n```python\nfrom rag_pipeline import RAGPipeline, FixedSizeChunking\nfrom rag_pipeline.document_loaders import PDFLoader\nfrom rag_pipeline.embeddings import OllamaEmbeddings\nfrom rag_pipeline.llms import OllamaLLM\nfrom rag_pipeline.vector_stores import FAISSVectorStore\n\n# Bile\u015fenleri ba\u015flat\nembedding = OllamaEmbeddings(\"nomic-embed-text:latest\", base_url=\"http://ollama:11434\")\nvector_store = FAISSVectorStore(collection_name=\"my_collection\", dimension=768)\nllm = OllamaLLM(\"llama3.2:3b\", base_url=\"http://ollama:11434\")\nchunking = FixedSizeChunking(chunk_size=450, overlap=100)\n\n# Pipeline olu\u015ftur\nrag = RAGPipeline(vector_store, embedding, llm, chunking)\n\n# Belgeleri y\u00fckle\ndocs = PDFLoader.load_folder(\"./documents\")\nrag.add_documents(docs)\n\n# Sorgu yap\nprint(rag.query(\"What are the feature methods used in cattle identification?\"))\n```\n\n---\n\n## Proje Yap\u0131s\u0131\n\n```\nrag_pipeline/\n\u2502\n\u251c\u2500\u2500 rag_pipeline/\n\u2502 \u251c\u2500\u2500 __init__.py\n\u2502 \u251c\u2500\u2500 pipeline.py\n\u2502 \u251c\u2500\u2500 chunking/\n\u2502 \u251c\u2500\u2500 document_loaders/\n\u2502 \u251c\u2500\u2500 embeddings/\n\u2502 \u251c\u2500\u2500 llms/\n\u2502 \u251c\u2500\u2500 vector_stores/\n\u2502 \u2514\u2500\u2500 retrievers/\n\u2502\n\u251c\u2500\u2500 tests/\n\u2502\n\u251c\u2500\u2500 setup.py\n\u251c\u2500\u2500 pyproject.toml\n\u251c\u2500\u2500 LICENSE\n\u2514\u2500\u2500 README.md\n```\n\n---\n\n## Lisans\n\nBu proje **MIT Lisans\u0131** ile lisanslanm\u0131\u015ft\u0131r. \nT\u00fcm detaylar i\u00e7in [LICENSE](LICENSE) dosyas\u0131na bakabilirsiniz.\n\n```\nMIT License\n\nCopyright (c) 2025 \u0130sim\n\nPermission is hereby granted, free of charge, to any person obtaining a copy...\n```\n\n---\n\n## \u00d6zellikler\n\n- FAISS ve di\u011fer vekt\u00f6r veritaban\u0131 deste\u011fi\n- Hibrit retrieval deste\u011fi (BM25 + vekt\u00f6r arama)\n- Mod\u00fcler yap\u0131\n- Geli\u015ftirici dostu API\n\n---\n\n## Katk\u0131da Bulunma\n\n1. Bu projeyi forklay\u0131n\n2. Yeni bir branch olu\u015fturun (`git checkout -b feature/ozellik`)\n3. De\u011fi\u015fikliklerinizi commit edin (`git commit -m 'Yeni \u00f6zellik eklendi'`)\n4. Branch\u2019inizi push edin (`git push origin feature/ozellik`)\n5. Pull Request olu\u015fturun\n\n---\n\n## \u0130leti\u015fim\n\nSorular\u0131n\u0131z veya \u00f6nerileriniz i\u00e7in: **ornek@email.com**\n",
"bugtrack_url": null,
"license": null,
"summary": "Mod\u00fcler, hibrit (dense + BM25) RAG pipeline k\u00fct\u00fcphanesi",
"version": "0.1.1",
"project_urls": null,
"split_keywords": [
"rag",
" retrieval",
" llm",
" nlp",
" vector",
" bm25",
" hybrid"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "921ecfe254d296e833810f811703c920937947e9dd2d84d45054e692fe5c6a3d",
"md5": "4278a83229c791cecada40f4aea046d6",
"sha256": "e5a1b538f247e6dad53bd2f0143b6652e4167ec499eb9e2c66f8db45a378a459"
},
"downloads": -1,
"filename": "rag_pipeline_lib-0.1.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "4278a83229c791cecada40f4aea046d6",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 29633,
"upload_time": "2025-08-13T11:22:59",
"upload_time_iso_8601": "2025-08-13T11:22:59.677563Z",
"url": "https://files.pythonhosted.org/packages/92/1e/cfe254d296e833810f811703c920937947e9dd2d84d45054e692fe5c6a3d/rag_pipeline_lib-0.1.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "dfbf9d5525302032d04d296e73d13bc71c2751b8909bebbb32692d63f48e5637",
"md5": "593e45405c039226db58a0de632337d4",
"sha256": "8255ca68acc0ffea54f9e81c0f324bceb3e70c5f48e2c88f31c8f51563f6efcd"
},
"downloads": -1,
"filename": "rag_pipeline_lib-0.1.1.tar.gz",
"has_sig": false,
"md5_digest": "593e45405c039226db58a0de632337d4",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 18658,
"upload_time": "2025-08-13T11:23:00",
"upload_time_iso_8601": "2025-08-13T11:23:00.994765Z",
"url": "https://files.pythonhosted.org/packages/df/bf/9d5525302032d04d296e73d13bc71c2751b8909bebbb32692d63f48e5637/rag_pipeline_lib-0.1.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-13 11:23:00",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "rag-pipeline-lib"
}
JSON