<h1 id="dreamjourney_v1-is-here!-16vmzbezc">Dreamjourney_v1 is Here!</h1>This model is made to improve the overall details of generation &amp; color enhancements + more!<h2 id="what's-in-this-model-6g4beqjax">What's in this model?</h2><ul><li>Color Enhancement / Color Grader</li><li>Improves details</li><li>Helps checkpoint models to better follow the prompts</li><li>Supports SD1.5 models</li></ul>Keep the lora strength in between 1 to 1.2 or below for perefect generationEx;- &lt;lora:dreamjourney_v1:1&gt; or dreamjourney:1But you can always play around settings.Keep that in mind, using dreamjourney:1 prompt is specially for fine tuning images like color enhancement or extra color detailing while &lt;lora:dreamjourney_v1:1&gt; is for producing Midjourney like "Surprise me"(not a prompt) kind of images.For &lt;lora:dreamjourney_v1:1&gt; dont go above 1.2 strength. Because it will start to mess up generations. Play in between 0.1 to 1.2. Same goes for the dreamjourney prompt as well.<h3 id="generation-guide:-ku9zu9s4v">Generation Guide:-</h3>realisticVision, absolutereality &amp; zovyaphotoreal were used for testing purpose.<ul><li>Checkpoint = Any SD1.5 checkpoints</li><li>Sampling = DPM++ 2M Karras / DPM++ SDE Karras</li><li>Step = 25</li><li>Guidance / CFG = 3 - 7 (Guidance 5 or less seems to be working fine for most of the generations)</li><li>Size = 512x512 or 720x1080</li></ul>

Files with SHA256: 809be6bf4df34954927a42337916b79a2c9d23d4974a3a4fa1553fd11e264a73