Free Local RAG Pipeline - Step-by-Step Tutorial

 This video is a step-by-step tutorial to locally install BeyondLLM with Ollama. Beyond LLM offers an all-in-one toolkit for experimentation, evaluation, and deployment of Retrieval-Augmented Generation (RAG) systems.

Code:

conda create -n beyondllm python=3.11

pip install beyondllm

pip install llama-index-embeddings-huggingface

pip install ollama

from beyondllm.source import fit

data = fit(path="mypdf.pdf", dtype="pdf")

from beyondllm.embeddings import HuggingFaceEmbeddings

embed_model = HuggingFaceEmbeddings(model_name="BAAI/bge-small-en-v1.5")

from beyondllm.retrieve import auto_retriever

retriever = auto_retriever(                            

    data=data,

    embed_model=embed_model,

    type="normal",

    top_k=5

 )

 

from beyondllm.llms import OllamaModel

llm = OllamaModel(model="mistral")

user_prompt = "who is fahd mirza?"

system_prompt = "You are an AI assistant...."

from beyondllm import generator

pipeline = generator.Generate(question=user_prompt, system_prompt = system_prompt, llm = llm, retriever=retriever)

print(pipeline.call())

print(pipeline.get_context_relevancy())

print(pipeline.get_answer_relevancy())

print(pipeline.get_groundedness())

print(pipeline.get_rag_triad_evals())