Back to Help Center
Vision Settings Guide
Configure and optimize Grout's AI vision model for your specific needs.
Model Configuration
Current Model:
ModelLlama 2.2 1TB Vision
Max Tokens500
Temperature0.7
Understanding Parameters
Max Tokens
Controls the maximum length of the model's response. Higher values allow for longer, more detailed explanations.
- Default: 500 tokens
- Range: 100-1000 tokens
- Recommended: 500 for most use cases
Temperature
Controls the randomness in the model's responses. Higher values make the output more creative but less focused.
- Default: 0.7
- Range: 0.0-1.0
- Lower values (0.3-0.5): More focused, factual responses
- Higher values (0.7-0.9): More creative, varied responses
Image Processing Features
Actions
- Crop Image
- Remove Background
- Fit to Canvas
AI Features
- Explain Image
- Extract Data
- Custom Prompt
Filters and Effects
Image Adjustments
- Brightness100%
- Contrast100%
- Saturation100%
- Blur0px
Effects
Grayscale
Sepia
Invert
Best Practices
Optimize for Your Use Case
Adjust temperature based on whether you need precise analysis (lower) or creative insights (higher).
Balance Quality and Speed
Higher max tokens provide more detailed responses but may take longer to generate.
Save Custom Settings
Create and save different configurations for different types of tasks.