<h3 id="important:-read-belowstrike-witches-characters-here-9vj4abrhd"><a target="_blank" rel="ugc" href="https://ko-fi.com/ctian"><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/cf614846-63c8-4596-a5fd-f5962d69ae75/width=525/cf614846-63c8-4596-a5fd-f5962d69ae75.jpeg" /></a>IMPORTANT: READ BELOW Strike Witches characters <a target="_blank" rel="ugc" href="https://civarchive.com/collections/2226666">here</a></h3>Trigger word(s): ctiansanya, Clothing: uniform, necktie, strapless vest, long sleeves, black skirt, black pantyhose, Cat: cat ears, cat tail,commission: <a target="_blank" rel="ugc" href="https://ko-fi.com/ctian">https://ko-fi.com/ctian</a> 😼 Example Positive prompt ✅:<pre><code>score_9, score_8_up, score_7_up, source_anime, masterpiece, 1girl, ctiansanya, uniform, necktie, strapless vest, long sleeves, black skirt, black pantyhose, beach, outdoors, cowboy shot, sunny, ocean, scenery, depth of field, looking at viewer, cinematic,</code></pre>Negative prompt ❌:<pre><code></code></pre>Recommended resolution, aspect ratio for sdxl, pony 📷768 x 1344: Vertical (9:16)832 x 1216: Portrait (4:5) **1024 x 1024: Square (1:1)1182 x 886: Photo (4:3)1254 x 836: Landscape (3:2)1365 x 768: Widescreen (16:9)Work flow, procedure 🔧1. Collect dataset (booru, anime screencaps, search)2. Clean up dataset using upscalers, inpainting and photoshop3. Tag dataset using booru <a target="_blank" rel="ugc" href="https://github.com/starik222/BooruDatasetTagManager">github</a>4. Train LoRa in kohya ss locally <a target="_blank" rel="ugc" href="https://github.com/bmaltais/kohya_ss">github</a>5. Test epochs using xyz grid in controlnet to ensure LoRa is not overfitted/underfitted.6. Submit model to CivitAI <a target="_blank" rel="ugc" href="https://civarchive.com/user/ctian">profile</a>do not ask me training parameters, check lora metadata for training parameters

Sanya V. Litvyak サーニャ・V・リトヴャク PONY - Strike Witches ストライクウィッチーズ

fixed captioning and retrained, model will be more stable.

Sanya_updated_ver2_ct.safetensors

v2.9

v1.0

<h3 id="sanya-v.-litvyak-v-pony-strike-witches-v6p21as1y">Sanya V. Litvyak サーニャ・V・リトヴャク PONY - Strike Witches ストライクウィッチーズ</h3>Strike Witches characters <a target="_blank" rel="ugc" href="https://civitai.com/collections/2226666">here</a> kofi commission AkumakushuTrigger word(s): ctiansanya, Clothing: uniform, necktie, strapless vest, long sleeves, black skirt, black pantyhose, Cat: cat ears, cat tail,commission: <a target="_blank" rel="ugc" href="https://ko-fi.com/ctian">https://ko-fi.com/ctian</a> 😼 Example Positive prompt ✅:<pre><code>score_9, score_8_up, score_7_up, source_anime, masterpiece, 1girl, ctiansanya, uniform, necktie, strapless vest, long sleeves, black skirt, black pantyhose, beach, outdoors, cowboy shot, sunny, ocean, scenery, depth of field, looking at viewer, cinematic,</code></pre>Negative prompt ❌:<pre><code></code></pre>Recommended resolution, aspect ratio for sdxl, pony 📷768 x 1344: Vertical (9:16)832 x 1216: Portrait (4:5) **1024 x 1024: Square (1:1)1182 x 886: Photo (4:3)1254 x 836: Landscape (3:2)1365 x 768: Widescreen (16:9)Work flow, procedure 🔧1. Collect dataset (booru, anime screencaps, search)2. Clean up dataset using upscalers, inpainting and photoshop3. Tag dataset using booru <a target="_blank" rel="ugc" href="https://github.com/starik222/BooruDatasetTagManager">github</a>4. Train LoRa in kohya ss locally <a target="_blank" rel="ugc" href="https://github.com/bmaltais/kohya_ss">github</a>5. Test epochs using xyz grid in controlnet to ensure LoRa is not overfitted/underfitted.6. Submit model to CivitAI <a target="_blank" rel="ugc" href="https://civitai.com/user/ctian">profile</a>do not ask me training parameters, check lora metadata for training parameters

&lt;p&gt;fixed captioning and retrained, model will be more stable.&lt;/p&gt;

Files with SHA256: 9b178c870f9384a05c18d53ee0b589e678f2ce03f50af8b8164f222a085f04a9