zeta-mlx-rag 0.3.2

RAG (Retrieval-Augmented Generation) for Zeta MLX

zeta-mlx-rag

RAG (Retrieval-Augmented Generation) pipeline for Zeta MLX.

Installation

pip install zeta-mlx-rag

Features

• Document Processing: Text chunking and preprocessing
• Embeddings: Integration with zeta-mlx-embedding
• Retrieval: Vector similarity search
• Generation: Context-aware LLM responses

Usage

from zeta_mlx.rag import RAGPipeline

pipeline = RAGPipeline(
    embedding_model="bge-m3",
    llm_model="qwen3-8b"
)

# Add documents
pipeline.add_documents(["doc1.txt", "doc2.txt"])

# Query
response = pipeline.query("What is MLX?")

Links

• GitHub
• Documentation