Warning: Although these quants work perfectly with ComfyUI - I couldn't get them to work with Forge UI yet. Let me know if this changes. The original non-k quants can be found <a target="_blank" rel="ugc" href="https://civarchive.com/models/705681">HERE</a> which are verified working with Forge UI.[Note: Unzip the download to get the GGUF. Civit doesn't support it natively, hence this workaround]These are the K(_M) quants for HyperFlux 16-steps. The K quants are slightly more precise and performant than non-K quants. HyperFlux is a merge of Flux.D with the 16-step HyperSD LoRA from ByteDance - turned into GGUF. As a result, you get an ultra-memory efficient and fast DEV (CFG sensitive) model that generates fully denoised images with just 16 steps while consuming ~6.2 GB VRAM (for the Q4_K_M quant). Also, the quality is pretty close to the original DEV model with ~30 steps.It can be used in ComfyUI with this <a target="_blank" rel="ugc" href="https://github.com/city96/ComfyUI-GGUF">custom node</a>. But I couldn't get these to work with Forge UI. See <a target="_blank" rel="ugc" href="https://github.com/lllyasviel/stable-diffusion-webui-forge/discussions/1050">https://github.com/lllyasviel/stable-diffusion-webui-forge/discussions/1050</a> for where to download the VAE, clip_l and t5xxl models.<h3 id="advantages-904rv91k3">Advantages</h3><ul><li>Quality similar to the original DEV model, while requiring only ~16 steps.</li><li>Better quality and expressivity than the 8-step HyperFlux in general.</li><li>For the same seed, the output image is pretty similar to the original DEV model, so you can use it to do quick searches and do a final generation with the dev model.</li><li>Sometimes you might even get better results than the DEV model due to serendipity.</li><li>Disadvantage: requires 16 steps.</li></ul><h3 id="which-model-should-i-download-zdxf7iwgj">Which model should I download?</h3>[Current situation: Using the updated Comfy UI (GGUF node) I can run Q6_K on my 11GB 1080ti.]Download the one that fits in your VRAM. The additional inference cost is quite small if the model fits in the GPU. Size order is Q2 &lt; Q3 &lt; Q4 &lt; Q5 &lt; Q6. I wouldn't recommend Q2 and Q3 unless you absolutely cannot fit the model in memory.All the license terms associated with Flux.1 Dev apply.PS: Credit goes to ByteDance for the HyperSD Flux 8-steps LoRA which can be found at <a target="_blank" rel="ugc" href="https://huggingface.co/ByteDance/Hyper-SD/tree/main">https://huggingface.co/ByteDance/Hyper-SD/tree/main</a>

GGUF_K: HyperFlux 16-Steps K_M Quants

ggufKHyperflux16StepsK_q4KM.zip

Files with SHA256: d086ee4d3e0b92b4f70f573c6f452585b4c2c2cbab8a39d090a098947a8084f8