Simple App with Langchain

This tutorial guides you through writing of your first LLM app using Langchain and Agenta. The objective is to create an app that can produce a persuasive startup pitch, using the startup's name and core idea. By the end of this tutorial, your app will be set up locally and ready for testing and refinement in the playground.

Installation

Run pip install agenta to install the Agenta SDK.

1. Project Initialization

First, create an empty project in a new directory.

mkdir my-first-app; cd my-first-app
agenta init

Follow the prompts to initialize your project. Make sure to select start with an empty project when prompted.

Now create a virtual environment:

python3 -m venv env
source env/bin/activate

Create a requirements.txt file and list the dependencies:

langchain
agenta
openai

Then install them:

pip install -r requirements.txt

Your project is now ready for development.

2. Write a Simple LLM App

Create a file named app.py. This will house a simple script that uses the langchain library to generate a startup pitch:

from langchain.chains import LLMChain
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate


default_prompt = """
Write a short message to an investor pitching this startup:
startup name: {startup_name}
startup idea: {startup_idea}"""

temperature = 0.5

def generate(
    startup_name: str,
    startup_idea: str) -> str:
    llm = OpenAI(temperature=temperature)
    prompt = PromptTemplate(
        input_variables=["startup_name", "startup_idea"],
        template=prompt_template)

    chain = LLMChain(llm=llm, prompt=prompt)
    output = chain.run(startup_name=startup_name,
                       startup_idea=startup_idea)
    return output

print(generate("agenta",
               "the open-source developer-first llmops platform"))

3. Add the Agenta SDK

The Agenta SDK transforms your code into an LLM app deployable in Agenta. The resulting code below illustrates how to integrate the SDK:

import agenta as ag
from langchain.chains import LLMChain
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate

default_prompt = """
    please write a short linkedin message (2 SENTENCES MAX) to an investor pitching the following startup:
    startup name: {startup_name}
    startup idea: {startup_idea}"""

ag.init()
ag.config.register_default(prompt_template=default_prompt, temperature=0.5)

@ag.entrypoint
def generate(
    startup_name: str,
    startup_idea: str,
) -> str:
    llm = OpenAI(temperature=ag.config.temperature)
    prompt = PromptTemplate(
        input_variables=["startup_name", "startup_idea"],
        template=ag.config.prompt_template)

    chain = LLMChain(llm=llm, prompt=prompt)
    output = chain.run(startup_name=startup_name, startup_idea=startup_idea)
    return output

Let's examine how we modified the original code.

ag.init()
ag.config.register_default(prompt_template=ag.TextParam(default_prompt), temperature=ag.FloatParam(0.5))

These two lines initialize agenta, then set a default configuration for the app.

@ag.entrypoint
def generate(
    startup_name: str,
    startup_idea: str,
) -> str:

The decorator @ag.entrypoint signals to agenta that this function is the app's entrypoint. This function will be the one that will be made into an api endpoint, and the one that will be called in the playground.

ag.config.temperature

All parameters for the app can be accessed using ag.config.paramname

Note that later when we modify the configuration in the playground, we will see the parameters modified in the code.

4. API Key Integration

For using OpenAI or other APIs, store your keys in a .env file:

OPENAI_API_KEY=sk-xxxxxxx

5. Run the app in CLI

Use the following command to execute the app via Command Line Interface (CLI):

python app.py agenta "the open-source developer-first llmops platform"

Hi there, I'm excited to tell you about Agenta, the open source Dev First LLMOPS Platform. It's a revolutionary way to simplify and streamline the development process, and I'm confident it will be a game changer in the industry. Let's chat soon!

6. Deploying to Agenta

Our app is still local. To make it available in the playground, add it to Agenta by running:

agenta variant serve app.py

Now you can interact with the app through the API or experiment in the playground.

7. Iterative Improvement in the Playground

Your app is now accessible in the Agenta UI at http://localhost. Interact with your app, experiment with different parameters, and fine-tune your app's behavior to achieve desired results.

tip

Well done! You've created and deployed your first LLM app with Agenta. Continue to experiment and refine your app as needed in the playground.