require <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">comfy-ui custom node (&gt;=0.0.4) : <a target="_blank" rel="ugc" href="https://registry.comfy.org/publishers/ttddrr11/nodes/animadualkvstyle">https://registry.comfy.org/publishers/ttddrr11/nodes/animadualkvstyle</a>should somewhat compatible with turbo and nagtest something to more easily overfit the style into anima model without text conditioningsimilar to <a target="_blank" rel="ugc" href="https://civarchive.com/models/2688130/animastyletest01">previous version</a>, but attention now from the input latent instead of just forcing query to go through static kvhave to balance between <code>--r_payload</code> and <code>--r_route</code>, should use with <code>--gradient_accumulation_steps</code> or higher batch size so the model have more refined direction when train<span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">code based on commit <code>c162e90</code><span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px"> <span style="color:rgb(138, 141, 145);font-family:&quot;Monaspace Neon&quot;, ui-monospace, SFMono-Regular, &quot;SF Mono&quot;, Menlo, Consolas, &quot;Liberation Mono&quot;, monospace;font-size:16px;background-color:rgb(21, 27, 35) !important">of <code>kohya-ss/sd-scripts</code><span style="color:rgb(138, 141, 145);font-family:&quot;Monaspace Neon&quot;, ui-monospace, SFMono-Regular, &quot;SF Mono&quot;, Menlo, Consolas, &quot;Liberation Mono&quot;, monospace;font-size:16px;background-color:rgb(21, 27, 35) !important"> <a target="_blank" rel="ugc" href="https://github.com/TheDucker1/sd-scripts/commit/c162e9039cc3228057b3fc6ae64ce770b87ab9bf"><span style="color:rgb(138, 141, 145);font-family:&quot;Monaspace Neon&quot;, ui-monospace, SFMono-Regular, &quot;SF Mono&quot;, Menlo, Consolas, &quot;Liberation Mono&quot;, monospace;font-size:16px;background-color:rgb(21, 27, 35) !important">repo on <span style="color:rgb(138, 141, 145);font-family:&quot;Monaspace Neon&quot;, ui-monospace, SFMono-Regular, &quot;SF Mono&quot;, Menlo, Consolas, &quot;Liberation Mono&quot;, monospace;background-color:rgb(21, 27, 35) !important">github</a>sometimes the style isn't showing, just bump the strength to 1.3 or something and it should overfit to showhow model were trained ( <code>--r_route</code> / <code>--r_payload</code> )<code>iel</code>: ~300ish magazine + cosplay images, prompt mixed of natural languages and tags, 64/64 (barbie texture at best, not really good) (took me to this to realize using <code>--gradient_accumulation_steps 4</code> helps the training a lot, all other was trained with effective batch size of 1 so it kinda noisy and may affect the prompt adherence)<code>itachi_3dt</code>: mixed of manga colored panel and danbooru images, prompts are tags, ~10ish, 16/16 (overfit little too much)<code>soleilmtfbwy03</code>: danbooru, , prompts are tags, ~10ish, 16/16 (16 was too small to correctly learn mapping the prompt, maybe should have bumped it to 32/* instead)<code>sakuram032</code>: twitter images, , prompts are tags, ~20ish, 16/16 (artstyle change a little too much, losing capabilities too generate good hand, not like the base model was that good anyway, limbs proportion not reached the extreme of original artist)<code>reiji</code>: cover images on danbooru, prompts are tags, ~10ish, 16/16<code>ceey</code>: danbooru,prompts are tags, ~10ish, 16/16<code>daodtt</code>: danbooru, prompts are tags, ~10ish, 8/16 (not to the same effect as the artist, but should be acceptable as a copycat)

anima_style_test02

animaStyleTest02_reiji.safetensors

reiji

Files with SHA256: 46a5b7878e4de865ddce3b0656447dee5ed5b427f29b06700933a39fa5853f88