{"id":14781,"date":"2024-07-04T10:25:23","date_gmt":"2024-07-04T02:25:23","guid":{"rendered":"https:\/\/www.1ai.net\/?p=14781"},"modified":"2024-07-04T10:25:23","modified_gmt":"2024-07-04T02:25:23","slug":"stable-diffusion%e6%95%99%e7%a8%8b%ef%bc%8c%e7%94%a83d%e5%a4%a7%e6%a8%a1%e5%9e%8b%e7%bb%98%e5%88%b63d%e9%a3%8e%e6%a0%bc%e7%9a%84%e5%8d%a1%e9%80%9a%e5%8a%a8%e6%bc%ab%e4%ba%ba%e7%89%a9","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/14781.html","title":{"rendered":"Stable Diffusion tutorial, use 3D large models to draw 3D style cartoon anime characters"},"content":{"rendered":"

Today I will share with you a SD1.5 based3D Large Model<\/a>: AWPoly3D. This model is a large 3D model of the AWP series. The characters in this model are trained based on Alibaba's "Duiyou" open source Avatar as a data set. The generated characters have very obvious recognition. It can also be used to make 3D icons and 3D scenes. It is also highly recommended to use this model for training and3D Characters<\/a>LoRA related to IP.<\/p>\n

The current latest version 1.1 is an upgrade, optimization and improvement based on 1.0.<\/p>\n

(1) Optimized support for generating avatars of different races<\/p>\n

(2) The response to prompt words has also been greatly optimized<\/p>\n

(3) The generated scene images are more detailed and realistic than 1.0<\/p>\n

(4) Greatly optimized limb performance, allowing for more complex movements and shapes<\/p>\n

Combined with comfy UI, you can create many interesting special effects workflows. For example, you can use controlnet to convert real photos into 3D avatar special effects.<\/p>\n

Model download address (you can also get the network disk address at the end of the article)<\/strong><\/p>\n

\n

LiblibAI<\/strong>: https:\/\/www.liblib.art\/modelinfo\/0f6d33ae67884e9b87ed399e3272555e<\/p>\n<\/blockquote>\n

The author provides instructions for using this large model on the official website.<\/p>\n

    \n
  • Sampling method: DPM++ 2M Karras<\/li>\n
  • Expression keywords: smile, happy, angry, stick out tongue, etc. (can also be added in ADetailer)<\/li>\n<\/ul>\n

    \n

    Let's actually experience it and see how the pictures produced by using this model look like.<\/p>\n

    Reverse prompt word<\/p>\n

    (worst quality:2),(low quality:2),(normal quality:2),lowres,bad anatomy,bad hands,((monochrome)),((grayscale)) watermark,moles,large breast,big breast ng_deepnegative_v1_75t badhandv4<\/p>\n

    Common parameter settings<\/p>\n

      \n
    • Large model: AWPoly3D V1.1<\/li>\n
    • Sampler: DPM++ 2M Karras<\/li>\n
    • Sampling iteration number: 40<\/li>\n
    • CFG: 7<\/li>\n
    • Image width and height: 512*768<\/li>\n
    • Open ADetailer plug-in, model selection: face_yolov8n<\/li>\n<\/ul>\n

      1. Beauty Portrait<\/strong><\/p>\n

      \n

      Prompt:<\/strong>close up portrait of a super beautiful Chinese girl,headshot,happy and smile expression,brown eyes,dark hair with undercut hairstyle,clean skin,looking at camera,wearing dark shirt,gray background<\/p>\n

      Prompt word<\/strong>: Close-up of a super pretty Chinese girl, headshot, happy and smiling expression, brown eyes, dark hair, short cut hairstyle, clean skin, looking at the camera, wearing a dark shirt, grey background<\/p>\n<\/blockquote>\n

      \"get-173\"<\/div>\n

      2. Male portrait<\/strong><\/p>\n

      \n

      Prompt<\/strong>\uff1aclose up portrait of a super beautiful Chinese boy,headshot,happy and smile expression,brown eyes,dark hair with undercut hairstyle,clean skin,looking at camera,wearing dark shirt,gray background,<\/p>\n

      Prompt word<\/strong>: Close up of a super nice Chinese boy, head shot, happy and smiling expression, brown eyes, dark hair, short cut hairstyle, clean skin, looking at camera, wearing dark shirt, grey background<\/p>\n<\/blockquote>\n

      \"get-174\"<\/div>\n

      3. Elon Musk<\/strong><\/p>\n

      \n

      Prompt<\/strong>\uff1aclose up portrait of Elon Musk,chinese vintage military coat,a red scarf was tied to his chest,looking at camera<\/p>\n

      Prompt word<\/strong>: Close-up of Elon Musk, wearing a retro Chinese military coat and a red scarf (red scarf) tied on his chest, looking at the camera<\/p>\n<\/blockquote>\n

      \"get-175\"<\/div>\n

      4. Girl with a cat<\/strong><\/p>\n

      \n

      Prompt: <\/strong>masterpiece,best quality,official art,ultra high res,8k,a girl holding a cat,surprised expression,white vest,look at viewer,in the room,daytime,soft light<\/p>\n

      Tips:<\/strong>Masterpiece, best quality, official art, ultra high resolution, 8k, a girl holding a cat, surprised expression, white vest, looking at the viewer, in the room, daytime, soft light<\/p>\n<\/blockquote>\n

      \"get-176\"<\/div>\n

      5. Boys and Girls<\/strong><\/p>\n

      \n

      Prompt: <\/strong>masterpiece,best quality,official art,ultra high res,8k,a boy and a girl,smile,back to back,looking at viewer,full body,building background,sky,sunshine<\/p>\n

      Tips:<\/strong>Masterpiece, best quality, official art, ultra high resolution, 8k, a boy and a girl, smiling, back to back, looking at the viewer, full body, building background, sky, sunlight<\/p>\n<\/blockquote>\n

      \"get-177\"<\/div>\n

      6. Little Princess<\/strong><\/p>\n

      \n

      Prompt: <\/strong>close up portrait of 1 girl,smile,wearing princess lace trim and gradient crystal dress and princess crystal crown,upper body,studio light<\/p>\n

      Tips:<\/strong>1 girl portrait close up, smiling, wearing princess lace trim and gradient crystal dress and princess crystal crown, top view, studio lighting<\/p>\n<\/blockquote>\n

      \"get-178\"<\/div>\n

      7. Angry expression<\/strong><\/p>\n

      \n

      Prompt: <\/strong>masterpiece,best quality,1girl,angry,golden ponytail hair,yellow hat,green shirt,looking at viewer,upper body<\/p>\n

      Tips:<\/strong>Masterpiece, best quality, 1 girl, angry, blonde ponytail, yellow hat, green shirt, looking at viewer, upper body<\/p>\n<\/blockquote>\n

      \"get-179\"<\/div>\n

      8. Slam Dunk<\/strong><\/p>\n

      \n

      Prompt: <\/strong>masterpiece,best quality,1boy,solo,guanlangaoshou,black hair,headband,black eyes,sportswear,basketball uniform,retro art style,sleeveless,hands in pockets,looking at viewer<\/p>\n

      Tips:<\/strong>Masterpiece, best quality, 1 boy, solo, close up blue high head, black hair, headband, black eyes, sportswear, basketball uniform, retro art style, sleeveless, hands in pockets, looking at the audience<\/p>\n<\/blockquote>\n

      \"get-180\"<\/div>\n

      9. Fashion Model<\/strong><\/p>\n

      \n

      Prompt: <\/strong>masterpiece,best quality,1girl,fashion model,upper body,multicolored<\/p>\n

      Tips:<\/strong>Masterpiece, best quality, 1 girl, fashion model, topless, multi-colored<\/p>\n<\/blockquote>\n

      \"get-181\"<\/div>\n

      Okay, that\u2019s all for today\u2019s sharing. I hope that what I shared today will be helpful to you.<\/p>\n

      The model is placed in the network disk, and those who are interested can take it!<\/p>\n

      https:\/\/pan.quark.cn\/s\/b3df771404e2<\/p>\n

       <\/p>","protected":false},"excerpt":{"rendered":"

      Today I would like to share with you a 3D large model based on SD1.5: AWPoly3D, which is a 3D large model of AWP series, the character of this model is based on the open source Avatar of Alibaba's \"Pile of Friends\" as a dataset training, and the generated character has a very obvious recognizable. It can also be used to create 3D icons, 3D scenes, and is also highly recommended to use this model to train 3D characters and IP-related LoRA. The latest version 1.1 is based on 1.0 and has been upgraded and optimized. (1) Optimized the support of avatar generation for different ethnicities (2) Response to prompts has been greatly optimized (3) Generation of images with scenes is more detailed and realistic than in 1.0 (4) Significantly optimized<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[144],"tags":[3338,3339,491,197,493],"collection":[302,262],"class_list":{"0":"post-14781","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"hentry","6":"category-baike","7":"tag-3d","9":"tag-ai","10":"tag-stable-diffusion","12":"collection-prompt","13":"collection-stablediffusion"},"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/14781","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=14781"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/14781\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=14781"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=14781"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=14781"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=14781"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}