Error: Can't find mmproj file for 'Qwen2.5-VL-7B-Instruct-Q3_K_M.gguf' (matching:'qwen2.5-vl-7b-instruct')! Qwen-Image-Edit will be broken!Rename the mmproj-F16:<h3 id="make-sure-the-downloaded-text-encoder-has-the-correct-file-naming-(whatever_text-encoder.gguf-paired-with-whatever_text-encoder_mmproj-f16.gguf)">Make sure the downloaded text encoder has the correct file naming (whatever_text-encoder.gguf paired with whatever_text-encoder_mmproj-F16.gguf)</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/5359bf23-cc7b-488e-a784-8fdd884ce37e/width=525/5359bf23-cc7b-488e-a784-8fdd884ce37e.jpeg" />Demo: <a target="_blank" rel="ugc" href="https://chat.qwen.ai/">https://chat.qwen.ai/</a> Text encooder FP8: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1864281?modelVersionId=2110105">https://civarchive.com/models/1864281?modelVersionId=2110105</a>Workflow GGUF: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1881851/qwen-iamge-edit-gguf-lightning-4steps-comfyui">https://civarchive.com/models/1881851/qwen-iamge-edit-gguf-lightning-4steps-comfyui</a><hr />This FP8-optimized (e4m3fn) conversion of Qwen-Image-Edit, now fully compatible with the latest Qwen-VL image editing features! Leveraging Alibaba's advanced multimodal AI, this model enables seamless text-guided image manipulation, perfect for precise edits, prompt-based adjustments, and creative refinements. The model is mirrored here for convenience.

Qwen-Image-Edit FP8 & GGUF

Text encoder: Qwen2.5-VL-7B-Instruct-Q3_K_M

qwenImageEditFP8GGUF_txtEnc7BInstructQ3.gguf

txt-enc-7B-Instruct-Q3

mmproj-F16

Q3_K_M.gguf

FP8-e4m3fn

This is a 3-bit quantized GGUF conversion of the <a target="_blank" rel="ugc" href="https://huggingface.co/Qwen/Qwen-Image">Qwen/Qwen-Image</a> model, released by <a target="_blank" rel="ugc" href="https://huggingface.co/city96/Qwen-Image-gguf">City96</a> and mirrored here for convenience. The Q3_K_M variant is optimized for GPUs with at least 10GB VRAM,<ul><li>Tested VRAM Usage:<ul><li>Ubuntu, Firefox (8 tabs): ~8.8 - 9.2 GB VRAM (CFG 3, 20 Steps, uni_pc, normal, 5.4s/it)</li><li>Windows 11, Brave (1 tab), MiniConda, GGUF excluded from Windows Defender: ~9.6 GB VRAM</li></ul>Tips: Offload monitor tasks to an integrated GPU to free up VRAM. Runs smoothly on Linux or Windows</li></ul>

Qwen-Image GGUF  3Q_K_M - 10G VRAM

Text encoder for Qwen-Image GGUF 3Q_K_M

Files with SHA256: accacb52744cadbe58b49f37ebb44e094f8de7ee2b52eb0b51ed9289329ef0c1