I created LoRA which is based on the style of the artist Tamotsu Murakami. His paintings are based on the cover for the Kozo Murashita album.<h3 id="heading-257">Recommended settings</h3>The closer the face, the more you will need to set the token weight (0.85-1). I recommend setting it to (0.9-1).If you only want to inherit the style, you can easily set it to 0.6-0.75.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/91cdaca6-0c7a-4c77-a35c-a0a1b7a72129/width=525/91cdaca6-0c7a-4c77-a35c-a0a1b7a72129.jpeg" />In some situations, it may not draw the face the first time, but don't worry.You can use any value of cfg scale<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/d64383e6-3862-419a-876d-8239df12d093/width=525/d64383e6-3862-419a-876d-8239df12d093.jpeg" />This token must be in your prompt - simple art And also this negative - (3d, photo, octane:1.2) Model trained on Deliberate v5, and i recommended use this model. You can found this model on huggingface.If you like my LoRA model, you can like it and rate it if you want to make me feel happy. Also I have Twitter (X) - <a target="_blank" rel="ugc" href="https://twitter.com/CuPYWEyAuYEmXH0">https://twitter.com/CuPYWEyAuYEmXH0</a> and Safebooru - <a target="_blank" rel="ugc" href="https://safebooru.org/index.php?page=artist&amp;s=show&amp;id=149">https://safebooru.org/index.php?page=artist&amp;s=show&amp;id=149</a>

Files with SHA256: 5c624a65a31d291ebc31c141ba968c1bf352e7072bbec307f264ad0e14938acb