<h3 id="comfyui-workflow">ComfyUI Workflow</h3><ul><li>required Custom Nodes:<ul><li><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_Eclipse">ComfyUI_Eclipse: Version 2.4.23, </a><a target="_blank" rel="ugc" href="https://registry.comfy.org/publishers/rvage/nodes/ComfyUI_Eclipse">ComfyReg (Nodes 2.0 compat, V3 API)</a></li><li><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_SmartLML">ComfyUI_SmartLML: Version 2.5.0</a>, <a target="_blank" rel="ugc" href="https://registry.comfy.org/publishers/rvage/nodes/ComfyUI_SmartLML">ComfyReg (Nodes 2.0 compat, V3 API)</a></li></ul></li><li>if you need help check the readme files in the repo <a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_SmartLML#">readme</a><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_RvTools/tree/main/Readme"> </a>folder at github.</li><li>llama-cpp-python is needed for the loading_method: GGUF (llama-cpp-python) in the Smart LML v2 Node<ul><li><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_Eclipse/releases/download/wheels/llama_cpp_python-0.3.16-cp312-cp312-win_amd64.whl">https://github.com/r-vage/ComfyUI_Eclipse/releases/download/wheels/llama_cpp_python-0.3.16-cp312-cp312-win_amd64.whl</a></li></ul></li><li>switch the template in the Smart Language Model Loader v2 to download the model. the node detects if the model already exists, it is also not downloading florence2 anymore when models\florence2 already exists, except for models that are not found in this folder.</li></ul><ul><li>load images from folder</li><li>Smart LML v2 loads the selected model for the image description</li><li>ReplaceString v3 removes unwanted parts</li><li>save prompt (.txt, .csv, .json)</li><li>if you create one file e.g. random_prompts you can use this file in the wildcard processor. for that:<ul><li>copy the file to the wildcards folder e.g. comfyui\models\wildcards. reload the page or restart comfyui if that didnt work and load the file from the list.</li></ul></li></ul><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/9f00e16d-f987-4f16-a49f-e2f418a88bd6/original=true/9f00e16d-f987-4f16-a49f-e2f418a88bd6.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e215d37d-7e26-4dbc-8563-b18cbac1922d/original=true/e215d37d-7e26-4dbc-8563-b18cbac1922d.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1f310384-fa3e-45f5-b79a-59504b4ed7a3/original=true/1f310384-fa3e-45f5-b79a-59504b4ed7a3.jpeg" />

Batch Captioning Workflow

<ul><li>update eclipse: 1.2.1<ul><li>smart lml v2 supports multi task mode with up to 4 tasks to reduce the need of nodes in the workflow</li><li>e.g. florence2 -&gt; tags</li><li>e.g. lexi llama -&gt; tags to natural lang -&gt; expand text - short story -&gt; refine &amp; expand prompt</li><li>added gguf caching, to allow keep_model_loaded</li></ul></li><li>reduced the number of smart lml v2 nodes :P</li></ul>

Files with SHA256: 4937829eb8aba0dfff2bbdb9ebe315d180c5b948e9ac408810d20bd2a9a4bde9