Expand this post for the recommended Anima parameters.Suggested tags for consistent results:<pre><code>colette bg, blonde hair, blue eyes, capelet, choker, long hair, pantyhose, jewelry</code></pre>Recommended settings:<ul><li>DPM++ 2M or Euler a</li><li>Karras</li><li>Sampling Steps: 25-35</li><li>CFG Scale: 4-6</li><li>Model Weight: 0.75-1 (I've had good success with weight 1)</li></ul>I haven't attempted NSFW, but considering the tests that I've preformed, it should pan out.As a personal preference, I generally use ADetailer for the faces of any generation, although this model is fairly consistent with it's accuracy.I have had success with multiple Checkpoints, so anything you want to use, should be good.<hr />Anima Version:<pre><code>Colette_kurba, long hair, blonde hair, blue eyes,</code></pre><pre><code>gold choker, jewelry, red gem, capelet, white dress, blue trim, black pantyhose, white boots,</code></pre>If you want to add her wings or chakrams, then add energy wings and chakrams.I've had mixed results while using quality tags. Here are a few that I've tested and found to work across multiple checkpoints.Generally okay to use with any checkpoint:<pre><code>masterpiece, best quality, score_9, score_8, score_7, absurdres, highres, newest, </code></pre>Best with higher-detail checkpoints:<pre><code>hyper realism, photorealistic,</code></pre>The realism tags improved overall quality while preserving the anime style.The lighting was the most notable difference when using these tags.Recommended settings:Sampling Steps: 15-35CFG Scale: 3.5-4I would love to see what you create with this LoRA. If you have any feedback, I'd be be interested in hearing it.

Colette Brunel - [Tales of Symphonia]

This version is trained on the same dataset as the Illustrious version, but using Anima.

Files with SHA256: c8359c241402f4fdb1937cf75fc26e1e453e51d0f446f0a63cd7bd4ff7ec9327