<h1 id="deskup-anima-edit-general">Deskup Anima Edit General</h1><code>deskup-anima-edit-general.safetensors</code> is a general-purpose image-editing LoRA designed for the Anima model. It was trained on approximately 80,000 image pairs and subsequently merged with <a target="_blank" rel="ugc" href="https://civarchive.com/models/2650553/anima-edit">Anima Edit</a>.The LoRA was trained for:- Pose changes- Background changes- Character changes- General scene modificationThis LoRA is experimental and is unstable. The same prompt may produce inconsistent results, and some edits may require several retries before producing the intended outcome.When changing background, original one might bleed into new image. For background changes I suggest to use single color background as reference.<h2 id="usage">Usage</h2>Use the LoRA through either of these interfaces:- ComfyUI: Install the <a target="_blank" rel="ugc" href="https://github.com/Mirumo0u0/ComfyUI-Cosmos-Reference">ComfyUI-Cosmos-Reference</a> custom node, then begin with the provided <a target="_blank" rel="ugc" href="https://civarchive.com/api/download/models/2976234?fileId=2856079">sample workflow</a>.- Stable Diffusion WebUI: Use the <code>neo</code> branch of <a target="_blank" rel="ugc" href="https://github.com/Haoming02/sd-webui-forge-classic/blob/neo/README.md">sd-webui-forge-classic</a>, which supports Anima image-editing LoRAs.Prompt primarily in natural language. Supplement the prompt with tags when natural-language instructions alone do not produce the desired result.<h2 id="prompting-guide">Prompting guide</h2>The LoRA accepts both descriptive captions and direct edit instructions. For more stable results, describe the desired final image clearly:<pre><code>A character wearing YYY, in pose ZZZ, against background XYZ</code></pre>Short instructions can also work:<pre><code>make character xxx a xyz</code></pre><pre><code>change yyy to xyz</code></pre><pre><code>replace zzz to xyz</code></pre>When replacing something, focus on the desired result or name the broader category instead of emphasizing the old item's exact appearance:<pre><code>Prefer: Replace the outfit with XXX</code></pre><pre><code>Avoid: Replace the white shirt with XXX</code></pre>If an unwanted element remains in the result, add that element to the negative prompt. If the model fails to preserve an important character, object, style, or scene detail, explicitly mention what must remain in the positive prompt.For difficult edits, keep the instruction concise and generate several variations rather than making the prompt increasingly complicated.

Files with SHA256: 92d6fa801ba452344e9005b1300f63fa3d16a024fb222348cd56e76ffcf65446