Skip to content

INSTALL_STEPS.md

Latest commit

 

History

History
215 lines (156 loc) · 5.67 KB

INSTALL_STEPS.md

File metadata and controls

215 lines (156 loc) · 5.67 KB

Installation Steps for LangBuilder Platform

Prerequisites

Ensure you have the following installed:

  1. Python 3.11 or 3.12 (recommended) - python --version
    • Note: Python 3.13+ is too new and has compatibility issues with some dependencies. Use Python 3.12 for best compatibility.
    • Recommended: Use pyenv for version management on newer Ubuntu versions (25.04+)
  2. Node.js 20.19.0+ or 22.x.x LTS (recommended) - node --version
    • Recommended: Use nvm (Node Version Manager) for version management
  3. npm 6.0.0+ - npm --version
  4. uv (Python package manager) - Install with: pipx install uv
  5. Git - For cloning the repository

Installation Steps

1. Clone the Repository

git clone <repository-url>
cd LangBuilder

2. Set Up OpenWebUI_CG

cd openwebui

# Create virtual environment
python -m venv .venv

# Activate virtual environment
# On Linux/Mac:
source .venv/bin/activate
# On Windows (Git Bash):
# source .venv/Scripts/activate

# Install backend dependencies
pip install -r backend/requirements.txt

# Install frontend dependencies
npm install --legacy-peer-deps

cd ..

3. Set Up LangBuilder_CG

cd langbuilder

# Create Python virtual environment with uv (REQUIRED)
uv venv

# Install backend dependencies
make install_backend

# Install frontend dependencies
make install_frontend

cd ..

Important Notes:

  • On Windows: Use Git Bash or WSL to run these commands
  • The scripts are cross-platform and work on Windows, Linux, and macOS
  • If uv venv fails on Windows, remove .venv manually first: rm -rf .venv

4. Configure Environment Variables

Review and update the .env file in the root directory with your API keys and configuration.

Starting Services

OPTION A - Start All Services (Recommended):

./start_all.sh

OPTION B - Start OpenWebUI_CG Only:

./openwebui/start_openwebui.sh

OPTION C - Start LangBuilder_CG Only:

./langbuilder/start_langbuilder_stack.sh

OPTION D - Start Services Manually:

# Terminal 1: OpenWebUI Backend
./openwebui/backend/start_openwebui_simple.sh

# Terminal 2: OpenWebUI Frontend
cd openwebui && npm run dev -- --port 5175

# Terminal 3: LangBuilder Backend
cd langbuilder && make backend port=8002

# Terminal 4: LangBuilder Frontend
cd langbuilder && make frontend

Service URLs

Connecting OpenWebUI with LangBuilder

After everything is working correctly, you need to connect OpenWebUI with LangBuilder to use LangBuilder workflows as AI providers.

Step 1: Generate an API Key in LangBuilder

  1. Open LangBuilder Frontend: http://localhost:3000
  2. Navigate to Settings or API Keys section
  3. Click Generate New API Key or Create API Key
  4. Copy the generated API key (save it securely, you'll need it in the next step)

Step 2: Configure LangBuilder as OpenAI Provider in OpenWebUI

  1. Open OpenWebUI Frontend: http://localhost:5175
  2. Click on your username at the bottom left corner
  3. Select Admin Panel
  4. Navigate to Settings → Connections
  5. Find Manage OpenAI API Connections section
  6. Click Add Connection
  7. Fill in the connection details:
    • Name: LangBuilder (or any name you prefer)
    • API Base URL: See table below depending on your setup
    • API Key: Paste the API key you generated in Step 1
  8. Click Save or Add Connection

API Base URL Configuration

Setup Type API Base URL
Docker (using ./start-docker-dev.sh) http://langbuilder-backend:8002/v1
Local/Manual (using ./start_all.sh or manual) http://localhost:8002/v1

Important: When running with Docker, containers communicate through an internal network. Using localhost inside a container refers to the container itself, not your host machine. Use the Docker service name langbuilder-backend instead.

Step 3: Verify the Connection

  1. Go back to the main OpenWebUI chat interface
  2. Start a new conversation
  3. Select the LangBuilder provider from the model dropdown
  4. Send a test message to verify the connection is working

Note: Make sure both LangBuilder backend (port 8002) and OpenWebUI (port 5175) are running before attempting to connect them.

Troubleshooting

Error: "No virtual environment found"

If you get this error when running make backend:

cd langbuilder
uv venv
make install_backend

Error: "Función incorrecta" or "Failed to create virtual environment" (Windows)

On Windows, if uv venv fails to replace an existing virtual environment:

cd langbuilder
# Remove the old virtual environment
rm -rf .venv
# Create a new one
uv venv
make install_backend

Error: "ModuleNotFoundError: No module named 'langbuilder'"

This means the backend dependencies are not installed:

cd langbuilder
uv venv
make install_backend

Verifying Services

Check if all ports are running:

# Using ss (recommended)
ss -tlnp | grep -E ':(3000|8002|8767|5175)'

# Or check each port individually
for port in 3000:LangBuilder-Frontend 8002:LangBuilder-Backend 8767:OpenWebUI-Backend 5175:OpenWebUI-Frontend; do
  p=$(echo $port | cut -d: -f1)
  name=$(echo $port | cut -d: -f2)
  echo -n "$name (port $p): "
  ss -tlnp | grep ":$p " > /dev/null && echo "✓ Running" || echo "✗ Not running"
done

Error: "uv is not installed"

Install uv using pipx:

pipx install uv

Or follow the instructions at: https://github.com/astral-sh/uv