Enhancement: Haystack integration

What features would you like to see added?

Haystack is a great opportunity to use RAG. Haystack has an API and LibreChat is a great UI with all the good implementation of user and document management. Do you see an opportunity to access Haystack with LibreChat using the API? https://haystack.deepset.ai

Haystack offers numerous features that can significantly enhance LibreChat, while LibreChat is the perfect complement for using Haystack with a user-friendly interface.

More details

Here is a small step by step guide to get the idea: If ollama is not already installed and running, visit https://ollama.com for linux installation:

curl -fsSL https://ollama.com/install.sh | sh

Get the Model used in the code (its a small one for testing purposes)

ollama pull mistral

After installing the necessary dependencies I have the following code provided to geht a small glimpse of what I mean.

conda create -n haystack python pip -y
conda activate haystack
pip install haystack-ai datasets ollama-haystack gradio

file: example.py

# Importing required libraries
from datasets import load_dataset
from haystack import Document
from haystack.document_stores.in_memory import InMemoryDocumentStore
from haystack.components.retrievers import InMemoryBM25Retriever
from haystack.components.builders import PromptBuilder
from haystack_integrations.components.generators.ollama import OllamaGenerator
from haystack import Pipeline
import gradio as gr

# Write documents to InMemoryDocumentStore
document_store = InMemoryDocumentStore()
document_store.write_documents([
    Document(content="My name is Jean and I live in Paris."), 
    Document(content="My name is Mark and I live in Berlin."), 
    Document(content="My name is Giorgio and I live in Rome.")
])

# Initialize retriever
retriever = InMemoryBM25Retriever(document_store)

# Define prompt template
template = """
Given the following information, answer the question.

Context:
{% for document in documents %}
    {{ document.content }}
{% endfor %}

Question: {{question}}
Answer:
"""

# Initialize prompt builder
prompt_builder = PromptBuilder(template=template)

# Initialize Ollama generator
generator = OllamaGenerator(
    model="mistral",
    url = "http://localhost:11434/api/generate",
    generation_kwargs={
        "num_predict": 100,
        "temperature": 0.9,
    }
)

# Create and configure pipeline
basic_rag_pipeline = Pipeline()
basic_rag_pipeline.add_component("retriever", retriever)
basic_rag_pipeline.add_component("prompt_builder", prompt_builder)
basic_rag_pipeline.add_component("llm", generator)
basic_rag_pipeline.connect("retriever", "prompt_builder.documents")
basic_rag_pipeline.connect("prompt_builder", "llm")

# Visualize pipeline (optional)
# basic_rag_pipeline.draw("basic-rag-pipeline.png")

# Define function to run pipeline with Gradio
def ask_question(question):
    response = basic_rag_pipeline.run(
        {
            "retriever": {"query": question}, 
            "prompt_builder": {"question": question}
        }
    )
    return response["llm"]["replies"][0]

# Create Gradio interface
gr_interface = gr.Interface(
    fn=ask_question,
    inputs=gr.components.Textbox(lines=2, placeholder="Enter your question here..."),
    outputs="text"
)

gr_interface.launch()

run the script

python example.py

access to the the gradio UI: http://127.0.0.1:7860 and ask a question about the given informations like:

My name is Jean, where do I live?

Instead of gradio it would be very nice using LibreChat as it offers so much more possibilities.

Thanks for reading all of this, you are very nice! And I send some nice greetings to all of you in the community! Cheers! Matthias :)

Which components are impacted by your request?

Endpoints

Pictures

Bildschirmfoto 2024-08-21 um 11 14 50

Code of Conduct

[X] I agree to follow this project's Code of Conduct

danny-avila / rag_api