Update README.md

README.md

@@ -63,17 +63,16 @@ widget:
   output:
     url: >-
       3104516.jpeg
-
+inference:
+  parameters:
+    width: 900
+    height: 1300
 ---
 
 # Doctor Diffusion's Tarot Card Crafter 
 
 <Gallery />
 
-
-
-
-
 <p>A powerful LoRA trained on modified photos taken by myself of all 78 of a <a target="_blank" rel="ugc" href="https://commons.wikimedia.org/wiki/Category:Rider-Waite_tarot_deck">Rider-Waite</a> tarot deck images originally from 1909, based on Golden Dawn symbolism.<br /><br />This LoRA was trained at 768x1344 but works great at most sizes, I recommend: 1024x1400. <br /><br /><strong>Text will often double or miss letters</strong>, but <strong>text will almost always be added</strong> in the correct locations on the cards. Often very simple to fix in an external photo editor. <strong>On the bright side</strong>, it forces you to fix and alter the image enough to perhaps <strong>claim a level of ownership</strong> on the created images.<br /><br />Will update if I can get more reliable text with training. <br /><br />Works best with simple prompts under five words. Use lesser strength prompts after the words you want to appear to add more control to your card.<br /><br />"<strong>trtcrd artstyle</strong> <em>prompt here</em>" <br />"<strong>trtcrd artstyle</strong> <em>queen of derp</em>"</p><p>"<strong>trtcrd artstyle</strong> <em>king of emeralds [elon musk]"</em> <br /><br /><br /><span style="color:rgb(17, 17, 17)">☕ </span>Like what I do? <span style="color:rgb(17, 17, 17)">☕</span><br /><span style="color:rgb(17, 17, 17)">☕ </span><a target="_blank" rel="ugc" href="https://www.buymeacoffee.com/doctordiffusion">Buy me a coffee or two</a>! <span style="color:rgb(17, 17, 17)">☕</span><br /></p>
 
 ## Image examples for the model: