🌟 Highly Recommended Settings 🌟

For the best results for the styles you must read the following settings and what you can also do if you wish to help me refine my models with using the model;

1. Prompt

Positive Prompt: plushy style, 3d, score_9, score_8_up, score_7_up, score_6_up

Negative Prompt: loli, child, young, source_cartoon, source_anime

Include these terms to refine the output and avoid unwanted elements.

2. Model Configuration

Checkpoint: VoightKampFF

Best results observed with this checkpoint.

LoRA Strength: 1

Optimal performance within this range.

3. Generation Parameters

Steps: 20+

Increasing steps can enhance output quality.

CFG: 7

Balances creativity and prompt adherence.

Sampler: Euler A

Efficient and high-quality results.

Scheduler: Normal

Consistent results with this scheduler.

🚨 No Illegal Content

You are not allowed to create illegal content that depicts minors or other -18 content that is immoral or against CivitAI TOS.

Q&A:

Q1: What are the beta versions of the loras?

The beta versions are for testing purposes to evaluate the likelihood of success with the dataset. Therefore, it will be a great help to see what your results are in the gallery. It is important to use the recommended checkpoints because testing displayed no influence on other checkpoints.

Q2: How can I help to improve the model?

1. Share Your Results

Upload your generated images to the gallery. This helps in evaluating the effectiveness of the model and provides valuable insights for improvements.

2. Provide Feedback

Leave detailed feedback on your experience using the model. Mention what worked well and any issues you encountered. Your input is essential for refining the model.

3. Stay Updated

Press the green thumbs up to stay tuned for future releases and updates. Your engagement helps in prioritizing new features and improvements.

Thank you for reading! If you want to stay tuned for future releases, press the green thumbs up. 👍