<h3 id="note:-there-are-two-fp8-download-options-for-many-versions.-the-17gb-option-is-an-all-in-one-checkpoint-that-includes-the-text-encoder-and-vae-as-well-so-load-it-like-a-regular-checkpoint.-the-12gb-option-is-just-the-unet-so-load-it-like-any-diffusion-model.">Note: There are two FP8 download options for many versions. The 17GB option is an all-in-one checkpoint that includes the text encoder and vae as well, so load it like a regular checkpoint. The 12GB option is just the unet so load it like any diffusion model.</h3>So many new model architectures, I can't keep up. Krea 2 is definitely going to have a place in my toolbox though. It might be the best current T2I model.For best results, follow the recommended suggestions in the About This Version section.<ul><li>v1.0 seems to be very good at NSFW content but it has the problem of producing unprompted nudity.</li><li>v1.0 Alt has addressed the unprompted nudity issue to some extent but is weaker at NSFW.</li><li>v2.0 Training Base is my test bed for new training.</li><li>v2.0 RC1 is a potential release candidate that builds on Training Base.</li><li>v2.0 Final is based on the original v1.0, abandoning the ALT, Training Base and RC lineages completely.</li><li>v3.0 is just a refinement of v2.0.</li></ul>Beginning with v2.0, there are at least 5 versions available for download:<ul><li>BF16</li><li>FP8</li><li>INT8 convrot</li><li>Q4_0 GGUF</li><li>FP8 All-in-One Checkpoint (text encoder and vae included)</li></ul>

GonzaLomo Krea 2

Added some additional training and also merging back with Krea 2 Raw and Gonzalomo Krea v1.0.Similar settings as v2.0 but an additional step is sometimes needed.Steps: 4-8CFG: 1.0Sampler/Scheduler combos: <ul><li>euler/beta</li><li>res_multistep/simple</li><li>er_sde/beta</li><li>gradient_estimation/simple</li><li>dpmpp_2m/simple</li><li>sa_solver/simple</li></ul>The workflow is embedded in all of my images.

Files with SHA256: 0de29ad3b77ed1ea672e38dadcfc29a01d602616efa4dd739731d0d4fdc7a02b