OLLAMA.md

Ollama Turbo Mode Guide

Optimize local model performance with turbo mode.

Demo

# Turbo mode
python examples/50_ollama/turbo_mode_demo.py turbo_cloud

# Baseline
python examples/50_ollama/turbo_mode_demo.py local_preset

Performance

• 20-40% latency improvement (hardware dependent)
• Reduced context window
• Lower temperature
• Optimized batch sizes

See examples/50_ollama/README.md for details.