<h2 id="find-out-how-to-use-this-workflow-on-youtube:"><span style="color:rgb(250, 176, 5)">Find out how to use this workflow on YouTube:</span></h2><p>Walkthrough of the workflow: </p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/jvzgOuR2KQ8" start="0"></iframe></div><p><br />A breakdown of the system prompts:</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/522VLc6OXps" start="0"></iframe></div><p> </p><h2 id="a-custom-fork-of-comfyui-gguf-is-required"><span style="color:rgb(250, 176, 5)"><strong>A custom fork of ComfyUI-GGUF is required</strong></span></h2><p>I would suggest following these steps:</p><ul><li><p>Stop ComfyUI if it is running.</p></li><li><p>Open your terminal and cd to <span style="color:rgb(64, 192, 87)"><strong>ComfyUI/custom_nodes</strong></span>.</p></li><li><p>Type the following command <code>git clone https://github.com/RealRebelAI/ComfyUI-GGUF_KREA-2.git</code></p></li><li><p>Restart ComfyUI</p></li></ul><p>Once clone the custom node folder structure should look like this:</p><pre><code>📂 ComfyUI/
├── 📂 custom_nodes/
│   ├── 📂 ComfyUI-GGUF_KREA-2/</code></pre><p></p><h2 id="installing-the-lmstudio-prompt-enhancer-custom-node."><span style="color:rgb(64, 192, 87)"><strong>Installing the LMStudio Prompt Enhancer custom node.</strong></span></h2><ul><li><p>Click the "Extensions" button at the top left hand corner of ComfyUI.</p></li><li><p>Search for a Node Pack called <strong>"comfyui-lmstudio-prompt-enhancer "</strong></p></li><li><p>Install "comfyui-lmstudio-prompt-enhancer "</p></li><li><p>Restart ComfyUI and refresh this page.</p></li></ul><p><a target="_blank" rel="ugc" href="https://github.com/AllenCraigBarnard/comfyui-lmstudio-prompt-enhancer"><u>Learn more about this custom node</u></a></p><p>I found 2 models in LMStudio model catalogue that fit the 32GB or less requirement for this workflow they are:</p><ul><li><p><strong>huihui-ornith-1.0-9b-abliterated-mtp</strong> you will want to set llm_max_tokens to 5000.</p></li><li><p><strong>meta-llama-3.1-8b-instruct-abliterated</strong> you will want to set llm_max_tokens to 2500.</p></li></ul><p>🛈 IMPORTANT!!!</p><p>You will need to search for, download, and load the model <strong>huihui-ornith-1.0-9b-abliterated-mtp</strong> or <strong>meta-llama-3.1-8b-instruct-abliterated</strong> in LM studio. You will also need to start the API server and generate an API Key. <a target="_blank" rel="ugc" href="https://github.com/AllenCraigBarnard/comfyui-lmstudio-prompt-enhancer#lm-studio-setup"><u>Learn how to configure LM Studio here.</u></a></p><h3 id="notes-for-using-this-32gb-workflow"><span style="color:rgb(64, 192, 87)"><strong>Notes for using this 32GB workflow</strong></span></h3><ul><li><p>If the image looks flat, check the logs in LM Studio, the last log entry will contain a JSON path x.choices[0].message.content if the content field is empty that means the LLM ran out of reasoning tokens, this can be resolved by increasing max_tokens.</p></li></ul><h2></h2><p></p>

Files with SHA256: 679f9b75bdd0e88c04f2bcf2b1220cc0c750fd57a611f487e7fb27707366eb1c