Open Source RAG with Nomic’s New Embedding Model (and ChromaDB and Ollama)

The context window of OSS LLMs and embedding models has been relatively small vs proprietary models. But, methods to expand context window like RoPE and self-extend are quickly changing this. Nomic has launched a new open source, long context embedding model with 8k token context window (using RoPE), strong performance on several benchmarks (outperforms OpenAI’s ada-002), and support for running locally (as well as via an API). Here, we show how to build a long context RAG app using OSS components from scratch: Nomic’s new 8k context window embeddings and Mistral-instruct 32k context window (via Ollama).

Cookbook – https://github.com/langchain-ai/langchain/blob/master/cookbook/nomic_embedding_rag.ipynb