<h2 id="about-bx9nr69gv">About</h2>This is a workflow that makes use of multiple image-to-text tools and a LLM to produce the final image descriptions for a batch of images in a folder, and write out the corresponding <code>.txt</code> files.This is especially helpful when captioning / describing NSFW images for LoRA training or fine-tuning, therefore the choice of the following 3 VLMs:<ul><li>Florence2, WD1.4 tagger</li><li>JoyCaption alpha 2</li><li>huihui-ai/Qwen2-VL-7B-Instruct-abliterated</li></ul>The LLM part for the final composition of the image description is done via the ollama node. I would say it is one of the easiest way to use a local LLM.You will get amazing results by using a uncensored large model such as the huihui-ai/Llama-3.3-70B-Instruct-abliterated(abliterated / uncensored models shall be used for both Qwen2-VL and the LLM to achieve the best results, nsfw or not).<h2 id="installation-xh4hjhq1d">Installation</h2>Except for the ComfyUI_Qwen2-VL-Instruct and Comfyui_JC2 nodes, install the missing nodes using the ComfyUI manager.ComfyUI_Qwen2-VL-InstructYou will need to use the Qwen2-VL-Instruct node from <a target="_blank" rel="ugc" href="https://github.com/edwios/ComfyUI_Qwen2-VL-Instruct">this fork</a> for this workflow to work:<a target="_blank" rel="ugc" href="https://github.com/edwios/ComfyUI_Qwen2-VL-Instruct">https://github.com/edwios/ComfyUI_Qwen2-VL-Instruct</a>This fork incorporated two major changes: Allows Image input same as the other VLM tools and uses the Mac GPU (mps) with Python 3.12 and up to and including PyTorch 2.6.Comfyui_JC2You may also want to use this <a target="_blank" rel="ugc" href="https://github.com/edwios/Comfyui_JC2">ComfyUI_JC2 fork</a> to utilise the Mac GPU for JoyCaption: Alpha 2.<h2 id="how-to-use-vocww8tr4">How to use</h2>Everything you will need to interact with this workflow is on the Left most.The simplest way to start is to enter the path to the directory containing the images. The results will be written to the same directory and name of the images but with the <code>.txt</code> extension.Optionally, you can do the followings:<ul><li>Change the VLM prompt to have Qwen2-VL to focus on a specific aspect of the image or images</li><li>Change the LLM prompt such as for better reasoning, or if you want it to write the descriptions in a SFW way (use at least a 70b instruct model for this). [No, this is NOT the same as using a 'safe' model.]</li></ul><h2 id="-u12kz0gtz"></h2><h3 id="credits-fmyjdoqd2">Credits</h3>Credits go to all that contributed to make ComfyUI and all these nodes available for all of us.Especially <a target="_blank" rel="ugc" href="https://github.com/IuvenisSapiens/ComfyUI_Qwen2-VL-Instruct">ComfyUI_Qwen-VL-Instruct, </a><a rel="ugc" href="https://github.com/TTPlanetPig/Comfyui_JC2">ComfyUI_JC2, </a><a rel="ugc" href="https://github.com/pythongosssss/ComfyUI-WD14-Tagger">ComfyUI-WD14-Tagger, </a><a rel="ugc" href="https://github.com/stavsap/comfyui-ollama">ComfyUI-Ollama, </a><a rel="ugc" href="https://github.com/kijai/ComfyUI-Florence2">ComfyUI-Florence2</a> and <a rel="ugc" href="https://ollama.com">Ollama</a> for making these amazing machine learning models available on mps, or at least not force it into a Nvidia only solution.

Files with SHA256: 7d73ebb7d5759f58ae188248d92cac1582281abc54be28d7a9c94205e5623dbf