<h2 id="flux-2-klein-9b-release">Flux 2 Klein 9B Release</h2><p>Suggested strength: <strong>0.7-0.8 </strong>(base model)</p><p>The LORA was trained on the base model. From my observations, the model really struggles with anatomy and often times struggles to interpret prompt, but is great at realism.</p><p>I was unable to get it to follow my prompts well using the distilled model. If anyone has suggestions or samples you could provide using the distilled model, please let me know. </p><p></p><p>Qwen image generation model. Suggested strength: <strong>0.6-0.85</strong></p><h1 id="notes">Notes</h1><p>LORA strength -&gt; it seems to work reasonably well at strengths 0.6-0.85, but feel free to experiment. I found that at strengths over 0.85, it significantly affects image composition.</p><p>At lower strengths, it can struggle with side-trained aspects of the lora (such as mismatched footwear)</p><h1 id="workflows">Workflows</h1><p>Workflows attached to sample images.</p><p>The workflows use VAE-utils, that help get rid of artifacts that are common on qwen generated images. Please follow installation tutorial here:</p><p><a target="_blank" rel="ugc" href="https://github.com/spacepxl/ComfyUI-VAE-Utils/tree/main">https://github.com/spacepxl/ComfyUI-VAE-Utils/tree/main</a></p><h2 id="parameters">Parameters</h2><p>If you are looking for a quicker generations (but lower fidelity), you can use the following parameters:</p><ul><li><p>Enable Lightning lora</p></li><li><p>Sampler: res_2s</p></li><li><p>Steps: 10</p></li><li><p>CFG: 1</p></li></ul><p></p>

Files with SHA256: e3ffef9444180fe3eb198d742d53808c9d09cc848194c63c7a13f9b14c6105e8