Local AI Models Setup
Ollama Setup for Local AI
To use local Ollama models with our application, you need to configure CORS (Cross-Origin Resource Sharing). This allows our web application to interact with locally running Ollama API.
Setup Instructions:
- Download and install Ollama for your operating system
- Download the CORS configuration script for your operating system (links below)
- Run the script with administrator privileges
- Restart Ollama if necessary
Note:
The script configures Ollama to accept requests from any origin, which is necessary for our application to work.
How to Use:
For Windows: Run the downloaded .bat file as administrator.
For Linux/macOS: Open terminal, navigate to the downloaded file and run:
chmod +x ollama-cors_linux-mac.sh
sudo ./ollama-cors_linux-mac.sh
After successful CORS configuration, you'll be able to use local Ollama models in our application.
Additional Information
Using local Ollama models provides several benefits:
- Enhanced privacy for your data
- Reduced latency when generating responses
- No API limits or usage fees
Recommended models to use with DreamAuto:
ollama pull gemma2:2b- basic model for most tasks