LangChain installation and setup in Prompt Engineering / GenAI - ML Experiment: Train & Evaluate

Experiment - LangChain installation and setup

Problem:You want to use LangChain to build AI applications that connect language models with external data and tools. Currently, LangChain is not installed or set up on your system.

Current Metrics:No LangChain package installed; no environment configured.

Issue:Without LangChain installed and set up, you cannot build or run AI applications that use its powerful chaining and integration features.

Your Task

Install LangChain and set up a basic environment to verify it works by running a simple example that calls a language model.

Use Python 3.12+ environment

Do not use any deprecated installation methods

Verify installation by running a minimal LangChain script

Hint 1

Hint 2

Hint 3

Solution

Prompt Engineering / GenAI

from langchain.llms import OpenAI

# Step 1: Install LangChain (run in terminal, not in script)
# pip install langchain

# Step 2: Set your OpenAI API key as environment variable
# export OPENAI_API_KEY='your_api_key_here'  # Linux/macOS
# setx OPENAI_API_KEY "your_api_key_here"  # Windows PowerShell

# Step 3: Simple LangChain usage example

# Initialize the OpenAI LLM
llm = OpenAI(temperature=0)

# Generate a simple prompt
response = llm("What is the capital of France?")

print(f"Response from LangChain LLM: {response}")

Corrected import statement to 'from langchain.llms import OpenAI'

Updated installation command to 'pip install langchain' instead of 'langchain-openai'

Results Interpretation

Before: No LangChain installed, no AI chaining possible.

After: LangChain installed and verified by running a simple prompt that returns a correct answer.

Installing and setting up LangChain correctly is the first step to building AI applications that combine language models with external data and tools.

Bonus Experiment

Try using LangChain with a different language model or a local mock model instead of OpenAI.

💡 Hint

Look into LangChain's documentation for other LLM wrappers or create a simple mock class that mimics the LLM interface.