semantic-router

Semantic Router is a superfast decision-making layer for your LLMs and agents. Rather than waiting for slow LLM generations to make tool-use decisions, we use the magic of semantic vector space to make those decisions — routing our requests using semantic meaning.

Read the Docs

Quickstart

To get started with semantic-router we install it like so:

pip install -qU semantic-router

❗ If wanting to use a fully local version of semantic router you can use HuggingFaceEncoder and LlamaCppLLM (pip install -qU "semantic-router[local]", see here). To use the HybridRouteLayer you must pip install -qU "semantic-router[hybrid]".

We begin by defining a set of Route objects. These are the decision paths that the semantic router can decide to use, let's try two simple routes for now — one for talk on politics and another for chitchat:

from semantic_router import Route

# we could use this as a guide for our chatbot to avoid political conversations
politics = Route(
    name="politics",
    utterances=[
        "isn't politics the best thing ever",
        "why don't you tell me about your political opinions",
        "don't you just love the president",
        "they're going to destroy this country!",
        "they will save the country!",
    ],
)

# this could be used as an indicator to our chatbot to switch to a more
# conversational prompt
chitchat = Route(
    name="chitchat",
    utterances=[
        "how's the weather today?",
        "how are things going?",
        "lovely weather today",
        "the weather is horrendous",
        "let's go to the chippy",
    ],
)

# we place both of our decisions together into single list
routes = [politics, chitchat]

We have our routes ready, now we initialize an embedding / encoder model. We currently support a CohereEncoder and OpenAIEncoder — more encoders will be added soon. To initialize them we do:

import os
from semantic_router.encoders import CohereEncoder, OpenAIEncoder

# for Cohere
os.environ["COHERE_API_KEY"] = "<YOUR_API_KEY>"
encoder = CohereEncoder()

# or for OpenAI
os.environ["OPENAI_API_KEY"] = "<YOUR_API_KEY>"
encoder = OpenAIEncoder()

With our routes and encoder defined we now create a RouteLayer. The route layer handles our semantic decision making.

from semantic_router.routers import SemanticRouter

rl = SemanticRouter(encoder=encoder, routes=routes, auto_sync="local")

We can now use our route layer to make super fast decisions based on user queries. Let's try with two queries that should trigger our route decisions:

rl("don't you love politics?").name

[Out]: 'politics'

Correct decision, let's try another:

rl("how's the weather today?").name

[Out]: 'chitchat'

We get both decisions correct! Now lets try sending an unrelated query:

rl("I'm interested in learning about llama 2").name

[Out]:

In this case, no decision could be made as we had no matches — so our route layer returned None!

Integrations

The encoders of semantic router include easy-to-use integrations with Cohere, OpenAI, Hugging Face, FastEmbed, and more — we even support multi-modality!.

Our utterance vector space also integrates with Pinecone and Qdrant!

📚 Resources

Docs

Online Course

Community

• Dimitrios Manias, Ali Chouman, Abdallah Shami, Semantic Routing for Enhanced Performance of LLM-Assisted Intent-Based 5G Core Network Management and Orchestration, IEEE GlobeCom 2024
• Julian Horsey, Semantic Router superfast decision layer for LLMs and AI agents, Geeky Gadgets
• azhar, Beyond Basic Chatbots: How Semantic Router is Changing the Game, AI Insights @ Medium
• Daniel Avila, Semantic Router: Enhancing Control in LLM Conversations, CodeGPT @ Medium
• Yogendra Sisodia, Stop Chat-GPT From Going Rogue In Production With Semantic Router, Medium
• Aniket Hingane, LLM Apps: Why you Must Know Semantic Router in 2024: Part 1, Medium
• Adrien Sales, 🔀 Semantic Router w. ollama/gemma2 : real life 10ms hotline challenge 🤯
• Adrien Sales, Kaggle Notebook 🔀 Semantic Router: ollama/ gemma2:9b hotline