rag.yml

# TxtAI Rag Application
#
# All-in-one RAG application
#
#   - Text Extraction with Docling, Chunking, and Indexing of documents. Supports PDF, DOCX, Web, XLSX files and more.
#   - Vector database with embeddings generation
#   - gpt-oss-20B LLM
#   - RAG pipeline that joins vector search with the LLM

# Embeddings configuration
writable: True
embeddings:
    content: True

# Text extraction
textractor:
    sections: True
    backend: docling
    headers:
        user-agent: Mozilla/5.0
    minlength: 50
    tuples: True

# RAG pipeline
rag:
    path: unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf
    n_ctx: 20000
    system: You are a friendly assistant
    output: flatten
    template: |
        Answer the following question using the provided context.

        Question:
        {question}

        Context:
        {context}

# Indexing workflow
workflow:
    index:
        tasks:
            - textractor
            - index