How to add more samples to krita ai generation tool

To enhance your experience with the AI generation tool in Krita, you can add more samplers, which are algorithms that influence the quality and style of generated images. Here’s how you can do it:

1. Understand the Role of Samplers

Samplers in Krita’s AI generation tool determine how the AI model refines images during the generation process. Different samplers offer various strengths, such as faster processing times or higher-quality outputs. For instance, samplers like DPM++ 2M Karras are known for quick convergence and stability, while others like Euler Ancestral might introduce more variation, offering unique artistic results. citeturn0search0

2. Locate the Sampler Configuration File

Krita’s AI Diffusion plugin stores sampler settings in a file named samplers.json. To access and modify this file:

• Find the File: Navigate to the presets directory, typically located at <appdata>/krita/ai_diffusion/presets/.
• Open the File: Use a text editor like Notepad to open samplers.json.

3. Add New Sampler Presets

To introduce a new sampler:

• Edit the File: In samplers.json, add a new entry following the existing structure. For example:

  {
      "My Custom Sampler - DPM++ 3M": {
          "sampler": "dpmpp_3m_sde",
          "scheduler": "exponential",
          "steps": 20,
          "minimum_steps": 4,
          "cfg": 7.0
      }
  }



This configuration defines a sampler named “My Custom Sampler – DPM++ 3M” with specific parameters. citeturn0search0

• Save Changes: After adding your desired samplers, save the samplers.json file.

4. Apply the New Sampler Settings

For the new samplers to take effect:

• Restart Krita: Close and reopen Krita to reload the plugin with the updated sampler settings.
• Select the Sampler: In the AI Image Generation panel, your custom samplers should now appear in the sampler selection dropdown.

5. Experiment and Adjust

After adding new samplers, experiment with them to find the best fit for your projects. Adjust parameters like the number of steps and the Classifier-Free Guidance (CFG) scale to fine-tune the image generation process.

Additional Tips

• Backup Your Settings: Before making changes, it’s advisable to back up your current samplers.json file.
• Compatibility: Ensure that the samplers you add are compatible with your current version of Krita and the AI Diffusion plugin.
• Performance Considerations: Some samplers may require more processing power. Monitor your system’s performance and adjust settings accordingly.

By following these steps, you can customize and expand the sampler options in Krita’s AI generation tool, tailoring the image generation process to better suit your creative needs.

For a visual walkthrough on adding samplers to Krita’s AI Diffusion plugin, you might find this video helpful: