The artistic style has been ft base the Pony XL model, following the guidelines of the Animagine XL V3 caption rules, by training with the collective works of 55 artists. A detailed inventory is included, although not every term may yield effective results.Sorry, a mistake yesterday,in the final step I merge 5 layers from js2pony. I tested it and the difference isn't significant, so I don't feel like re-uploading.Let's just go with the flow (●'◡'●).<h1 id="heading-41">please add artist name in the list to prompts if you want to get some style.</h1>use prompts like use pony XLartist list in dataset(sort by random?):<ol><li>hoji (hooooooooji1029)</li><li>fuzichoco</li><li>kani biimu</li><li>kfr</li><li>mika pikazo</li><li>scottie (phantom2)</li><li>toosaka asagi</li><li>atdan</li><li>dsmile</li><li>gin00</li><li>happoubi jin</li><li>hiten (hitenkei)</li><li>kazutake hazano</li><li>ke-ta</li><li>kedama milk</li><li>mignon</li><li>misaka 12003-gou</li><li>momoko (momopoco)</li><li>morikura en</li><li>noco (adamas)</li><li>shiratama (shiratamaco)</li><li>torino aqua</li><li>amashiro natsuki</li><li>anmi</li><li>as109</li><li>ask (askzy)</li><li>fangxiang cuoluan</li><li>kaede (yumesaki kaede)</li><li>matanonki</li><li>mochizuki kei</li><li>rin yuu</li><li>soraneko hino</li><li>tiv</li><li>toraishi 666</li><li>toridamono</li><li>ciloranko</li><li>ogipote</li><li>poco (asahi age)</li><li>jima</li><li>miv4t</li><li>noyu (noyu23386566)</li><li>quan (kurisu tina)</li><li>rurudo</li><li>sy4</li><li>hito komoru</li><li>snow is</li><li>weri</li><li>akakura</li><li>binggong asylum</li><li>chen bin</li><li>nekojira</li><li>akizero1510</li><li>hong bai</li><li>da mao banlangen</li></ol>quality_tag_list(in this ft caption):<ol><li>"masterpiece",</li><li>"best quality",</li><li>"great quality",</li><li>"good quality",</li><li>"normal quality",</li><li>"low quality",</li><li>"worst quality"</li></ol>year_tag_lsit(in this ft caption):2005 &lt;= year &lt;= 2010: year_tag = "old"elif year &lt;= 2014: year_tag = "early"elif year &lt;= 2017:year_tag = "mid"elif year &lt;= 2020: year_tag = "recent"elif year &lt;= 2024:year_tag = "newest"dataset from:<a target="_blank" rel="ugc" href="https://huggingface.co/datasets/KBlueLeaf/danbooru2023-webp-2Mpixel">https://huggingface.co/datasets/KBlueLeaf/danbooru2023-webp-2Mpixel</a><a target="_blank" rel="ugc" href="https://huggingface.co/datasets/NebulaeWis/gelbooru_images">https://huggingface.co/datasets/NebulaeWis/gelbooru_images</a>images process and filter :<a target="_blank" rel="ugc" href="https://github.com/deepghs/waifuc/"> https://github.com/deepghs/waifuc/</a>caption &amp; creation dataset :<a target="_blank" rel="ugc" href="https://github.com/KohakuBlueleaf/HakuBooru"> https://github.com/KohakuBlueleaf/HakuBooru</a>other code :me<h2 id="heading-131">License</h2>This model is released under Fair-AI-Public-License-1.0-SD Plz check this website for more information: <a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">Freedom of Development (</a><a target="_blank" rel="ugc" href="http://freedevproject.org">freedevproject.org</a><a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">)</a>

Aurora XL base[pony XL]

auroraXLBasePonyXL_v10.safetensors

v1.0

After extensive testing, it is sure that Pony model cannot accept enough content within the acceptable time limit for training.Aurora is achieve sufficient results after 44k images overtrain,but under overtrain almost lora train on Pony cannot work well on aurora.Tests with 89k and 200k images almost did not produce any effective results.Conversely, only a few hundred datasets yielded good results.In conclusion, aurora can no longer be updated base pony.After all,if a larger training volume is required, why not skip Pony and directly fine-tuning base SDXL 1.0?---------------------------------------The performance of Kohaku XL Delta is over my expectations, and I'm not a Furry.So this model will stop all update base the Pony XL.please wait new kohaku XL.2024-03-03 04:18 am (utc+8)The artistic style has been ft base the Pony XL model, following the guidelines of the Animagine XL V3 caption rules, by training with the collective works of 55 artists. A detailed inventory is included, although not every term may yield effective results.Sorry, a mistake yesterday,in the final step I merge 5 layers from js2pony. I tested it and the difference isn't significant, so I don't feel like re-uploading.Let's just go with the flow (●'◡'●).please add artist name in the list to prompts if you want to get some style.use prompts like use pony XLartist list in dataset(sort by random?):<ol><li>hoji (hooooooooji1029)</li><li>fuzichoco</li><li>kani biimu</li><li>kfr</li><li>mika pikazo</li><li>scottie (phantom2)</li><li>toosaka asagi</li><li>atdan</li><li>dsmile</li><li>gin00</li><li>happoubi jin</li><li>hiten (hitenkei)</li><li>kazutake hazano</li><li>ke-ta</li><li>kedama milk</li><li>mignon</li><li>misaka 12003-gou</li><li>momoko (momopoco)</li><li>morikura en</li><li>noco (adamas)</li><li>shiratama (shiratamaco)</li><li>torino aqua</li><li>amashiro natsuki</li><li>anmi</li><li>as109</li><li>ask (askzy)</li><li>fangxiang cuoluan</li><li>kaede (yumesaki kaede)</li><li>matanonki</li><li>mochizuki kei</li><li>rin yuu</li><li>soraneko hino</li><li>tiv</li><li>toraishi 666</li><li>toridamono</li><li>ciloranko</li><li>ogipote</li><li>poco (asahi age)</li><li>jima</li><li>miv4t</li><li>noyu (noyu23386566)</li><li>quan (kurisu tina)</li><li>rurudo</li><li>sy4</li><li>hito komoru</li><li>snow is</li><li>weri</li><li>akakura</li><li>binggong asylum</li><li>chen bin</li><li>nekojira</li><li>akizero1510</li><li>hong bai</li><li>da mao banlangen</li></ol>quality_tag_list(in this ft caption):<ol><li>"masterpiece",</li><li>"best quality",</li><li>"great quality",</li><li>"good quality",</li><li>"normal quality",</li><li>"low quality",</li><li>"worst quality"</li></ol>year_tag_lsit(in this ft caption):2005 &lt;= year &lt;= 2010: year_tag = "old"elif year &lt;= 2014: year_tag = "early"elif year &lt;= 2017:year_tag = "mid"elif year &lt;= 2020: year_tag = "recent"elif year &lt;= 2024:year_tag = "newest"dataset from:<a target="_blank" rel="ugc" href="https://huggingface.co/datasets/KBlueLeaf/danbooru2023-webp-2Mpixel">https://huggingface.co/datasets/KBlueLeaf/danbooru2023-webp-2Mpixel</a><a target="_blank" rel="ugc" href="https://huggingface.co/datasets/NebulaeWis/gelbooru_images">https://huggingface.co/datasets/NebulaeWis/gelbooru_images</a>images process and filter :<a target="_blank" rel="ugc" href="https://github.com/deepghs/waifuc/">&nbsp;https://github.com/deepghs/waifuc/</a>caption &amp; creation dataset :<a target="_blank" rel="ugc" href="https://github.com/KohakuBlueleaf/HakuBooru">&nbsp;https://github.com/KohakuBlueleaf/HakuBooru</a>other code :meLicenseThis model is released under Fair-AI-Public-License-1.0-SD Plz check this website for more information: <a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">Freedom of Development (</a><a target="_blank" rel="ugc" href="http://freedevproject.org/">freedevproject.org</a><a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">)</a>

Files with SHA256: c78f6fc96ee983bba208690e44f7b35508feda3ef1cb9483af6520440b7bb70a