<a target="_blank" rel="ugc" href="https://www.patreon.com/ParanoidAmerican"><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/bb608bba-7b1b-4009-869f-5a9b6f7faf50/width=525/bb608bba-7b1b-4009-869f-5a9b6f7faf50.jpeg" /></a>follow me Paranoid American <a target="_blank" rel="ugc" href="https://www.instagram.com/paranoidamerican/">Instagram</a>, <a target="_blank" rel="ugc" href="https://www.youtube.com/@ParanoidAmerican/">Youtube</a>, <a target="_blank" rel="ugc" href="https://www.patreon.com/ParanoidAmerican">Patreon</a> and the <a target="_blank" rel="ugc" href="https://paranoidamerican.com/">website</a>.<h1 id="heading-9">Howard the Duck</h1>Suggested Weight: 1.0 (lower weights start looking more like a regular duck)<h2 id="heading-10">Suggested prompts</h2><ul><li>professional portrait photo of (h0w4rdth3duck:1.0) as the pope, duck bill, light bokeh, sharp focus</li><li>Note, you might want to add "duck bill" to the prompt to avoid getting a human face, along with adding "human nose" to the negative prompt</li></ul><h2 id="heading-11">Training</h2><ul><li>trained on ~10 images</li><li>Learning Rate: 0.005:1000, 0.001:2000, 0.0001:5000, 0.00005</li><li>Total Steps: 15,000</li><li>Batch Size: 1</li><li>Gradient Steps: 1</li></ul>

Files with SHA256: 55b2304df6b64ab28824128169b938176c29713db80de4f90378d96fc7dfd881