<h1 id="overview-6b10lzt1k">Overview</h1>This model is developed based on the architecture of FLUX.1-dev, incorporating several structural modifications. It is trained on a custom dataset containing both high-quality anime and realistic images sourced from the internet. The dataset is initially captioned by ChatGPT o3-mini and then manually rechecked to ensure accuracy. Throughout training, the model utilizes FLUX.1-dev as a teacher for guidance.<h1 id="important-a2gars771">Important</h1><ul><li>As this is a self-reconstructed model, I highly value any feedback you can provide. Your insights will not only motivate me but also help me better understand its strengths and weaknesses for future improvements.</li><li>This is an AIO model (All-In-One) that includes all essential components: VAE, CLIP, and T5Encoder. I recommend using the default settings. Additionally, the model utilizes the bf16 data type, ensuring optimal quality in image generation.</li></ul>Currently, the model is not available for use via Civitai Generation. You can visit the following website to use it:<ul><li><a target="_blank" rel="ugc" href="https://tensor.art/models/844097477800918826/AuroraYume-Flux.1-Dev-v1.0">tensor.art</a></li></ul><h1 id="recommended-settings:-ado3tabvl">Recommended Settings:</h1><ul><li>Guidance Scale: 2.0 - 5.0</li><li>Steps: 20 - 30</li><li>Sampler: Euler a / DPM++ 2M</li><li>Schedule: Simple</li></ul><h1 id="notes-eys6tu2pw">Notes</h1><ul><li>No post-processing or LoRA was applied to enhance the example images. They were generated solely using the recommended settings and a custom prompt with my base model.</li><li>The example images were created using prompts from various sources and authors to provide an independent evaluation of the model's performance.</li></ul><h1 id="acknowledgments-mqkgtr55y">Acknowledgments</h1><ul><li>Thanks to <a target="_blank" rel="ugc" href="https://blackforestlabs.ai/">Black Forest Labs</a> for providing an amazing model!</li></ul>If you'd like to support my work, you can do so through <a target="_blank" rel="ugc" href="https://ko-fi.com/duongve">Ko-fi</a>!

AuroraYume Flux.1-Dev

aurorayumeFlux1Dev_v10.safetensors

v1.0

<ol><li>Overview</li></ol>This model is developed based on the architecture of FLUX.1-dev, incorporating several structural modifications. It is trained on a custom dataset containing both high-quality anime and realistic images sourced from the internet. The dataset is initially captioned by ChatGPT o3-mini and then manually rechecked to ensure accuracy. Throughout training, the model utilizes FLUX.1-dev as a teacher for guidance.<ol start="2"><li>Important</li></ol><ul><li>As this is a self-reconstructed model, I highly value any feedback you can provide. Your insights will not only motivate me but also help me better understand its strengths and weaknesses for future improvements.</li><li>This is an AIO model (All-In-One) that includes all essential components: VAE, CLIP, and T5Encoder. I recommend using the default settings. Additionally, the model utilizes the bf16 data type, ensuring optimal quality in image generation.</li></ul><ol start="3"><li>Recommended Settings:</li></ol><ul><li>Guidance Scale: 2.0 - 7.0</li><li>Steps: 20 - 30</li><li>Sampler: Euler a / DPM++ 2M</li><li>Schedule: Simple</li></ul><ol start="4"><li>Notes<ul><li>No post-processing or LoRA was applied to enhance the example images. They were generated solely using the recommended settings and a custom prompt with my base model.</li><li>The example images were created using prompts from various sources and authors to provide an independent evaluation of the model's performance.</li></ul></li><li>AcknowledgmentsThanks to <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://blackforestlabs.ai/">Black Forest Labs</a> for providing an amazing model!If you'd like to support my work, you can do so through <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://ko-fi.com/duongve">Ko-fi</a>!</li></ol>

Files with SHA256: b439333d7275081a1a47dbefbd3e3cc0d6224dc7c5f6bb6cd7cf3ad5a6802a70