mrpopsalot
/

sd-speed-list

@@ -78,7 +78,7 @@ tags:
   <div class="section-div">
       <img src="./sd-speeds-v002.png">
       <p>A hand curated list of generation speeds for various hardware and models.</p>
-      <p><b>LAST UPDATED:</b> 2025.08.09</p>
       <p><b>Use the ComfyUI workflow above to start testing.</b></p>
   </div>
   <div class="section-div">
@@ -107,7 +107,7 @@ tags:
     </li>
     <p>Raw gen times are not recorded due to variance due to steps being variable. Instead iterations per second (and the inverse of it) are given since they are independent of steps.</p>
     <p>The given speed value (it/s or s/it) is used, and then extrapolated using the formula <code>1/speed</code> to get the other value. If its under 0.01 then it will be expanded to four digits compared to the usual 2</p>
-    <p>If you can contribute to the list, do so as well. Lets make the most comprehensive, curated list of local Image Gen speeds!</p>
     <div class="subsection-div">
       <h3>Models Used</h3>
       <p>The following are the models used for testing. The models you use can be the same architecture as the tested models</p>
@@ -2035,9 +2035,8 @@ tags:
         <li>Lower your image sizes. SDXL can work coherently down to 384px and SD1.5 can go down to 128px.</li>
         <li>Use addons such as TeaCache</li>
         <li>Use low step LoRAs such as DMD2</li>
-        <li>As a last resort, disable CFG by setting your CFG to 1. This will disable your negative prompt but also increase your speeds drastically. This will also severely affect your output quality. The lack of negative prompt can be circumvented with the use of NAG</li>
         <li>Upgrade your potato with a new GPU if all else fails.</li>
     </ul>
   </div>
-</div>
->

   <div class="section-div">
       <img src="./sd-speeds-v002.png">
       <p>A hand curated list of generation speeds for various hardware and models.</p>
+      <p><b>LAST UPDATED:</b> 2025.10.15</p>
       <p><b>Use the ComfyUI workflow above to start testing.</b></p>
   </div>
   <div class="section-div">
     </li>
     <p>Raw gen times are not recorded due to variance due to steps being variable. Instead iterations per second (and the inverse of it) are given since they are independent of steps.</p>
     <p>The given speed value (it/s or s/it) is used, and then extrapolated using the formula <code>1/speed</code> to get the other value. If its under 0.01 then it will be expanded to four digits compared to the usual 2</p>
+    <p>If you can contribute to the list, do so as well. Lets make the most comprehensive, curated list of local image gen speeds!</p>
     <div class="subsection-div">
       <h3>Models Used</h3>
       <p>The following are the models used for testing. The models you use can be the same architecture as the tested models</p>
         <li>Lower your image sizes. SDXL can work coherently down to 384px and SD1.5 can go down to 128px.</li>
         <li>Use addons such as TeaCache</li>
         <li>Use low step LoRAs such as DMD2</li>
+        <li>As a last resort, disable CFG by setting your CFG to 1. This will disable your negative prompt but also increase your speeds drastically. This will also severely affect your output quality. The lack of negative prompt can be circumvented with the use of NAG or NegPIP</li>
         <li>Upgrade your potato with a new GPU if all else fails.</li>
     </ul>
   </div>
+</div>