<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>updated eclipse 1.0.81<ul><li>fixed random seed (smart lml v2),</li><li>fixed llm instructions for GGUF (llama-cpp-python)<ul><li>fixed think stripping with fallback</li><li>prompt defaults</li></ul></li><li>improved style prompt (detection nl vs tags)</li><li>save prompt (added log_prompt to print directly to console to view the prompts without the need of show any)</li></ul></li><li>added a 2nd Smart LML v2 to expand the replaced description after the use of prompt styler.</li></ul>

Files with SHA256: 6c7cad411053b42e35ab21f574066c83ce02c667766a439d937cc63d509de50d