<h3 id="important:-read-below-w69wav5n7"><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</h3>Trigger word(s): ctianobor0zuki, long hair, Clothing: neck ribbon, collared shirt, skirt, pleated skirt, long sleeves,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, ctianobor0zuki, long hair, medium breasts, neck ribbon, collared shirt, blue skirt, pleated skirt, long sleeves, school uniform, outdoors, sunny, city, depth of field, cowboy shot, looking at viewer, hand on chest,</code></pre>Negative prompt ❌:<pre><code>sketch, 3d, shiny hair,</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<h1 id="faq:-2zkralglt">FAQ:</h1>Q: why not use character name as trigger word? A: I use unique token to prevent other things from being prompted when using triggerword Q: do you have other account in other websites? A: yes, tensor, pixai Q: what external generator do you use? A: a111 mostly, comfyui for img2vid

Kanae Oborozuki 朧月 カナエ PONY - Jimoto no Ijimekko-tachi ni Shikaeshi Shiyou to Shitara, Betsu no Tatakai ga Hajimatta.

Kanae_Oborozuki_pony_ct.safetensors

v1.0

<h3 id="kanae-oborozuki-pony-jimoto-no-ijimekko-tachi-ni-shikaeshi-shiyou-to-shitara-betsu-no-tatakai-ga-hajimatta.-2c522d29w">Kanae Oborozuki 朧月 カナエ PONY - Jimoto no Ijimekko-tachi ni Shikaeshi Shiyou to Shitara, Betsu no Tatakai ga Hajimatta.</h3><h3 id="ko-fi-commissioneduse-word-&quot;greyscale&quot;-if-want-manga-style-jfjuufsh7"><a target="_blank" rel="ugc" href="https://ko-fi.com/ctian"><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/7a25ef69-a98e-4d14-9e8b-8b68ff0c3339/width=525/7a25ef69-a98e-4d14-9e8b-8b68ff0c3339.jpeg" /></a>Ko-fi commissioned use word "greyscale" if want manga style</h3>Trigger word(s): ctianobor0zuki, long hair, Clothing: neck ribbon, collared shirt, skirt, pleated skirt, long sleeves,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, ctianobor0zuki, long hair, medium breasts, neck ribbon, collared shirt, blue skirt, pleated skirt, long sleeves, school uniform, outdoors, sunny, city, depth of field, cowboy shot, looking at viewer, hand on chest,</code></pre>Negative prompt ❌:<pre><code>sketch, 3d, shiny hair,</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<h1 id="faq:-d9ux94yoj">FAQ:</h1>Q: why not use character name as trigger word? A: I use unique token to prevent other things from being prompted when using triggerword Q: do you have other account in other websites? A: yes, tensor, pixai Q: what external generator do you use? A: a111 mostly, comfyui for img2vid

&lt;p&gt;trained from manga chapter 1-38&lt;/p&gt;

Files with SHA256: 112766def2f9b565b7eefa562adb208dd04ae811a8f12731f634d11a0bc85f45