Made a lightning version of V2-X-Tau, using the top lightning models and suffice to say I'm blown away by it myself.
For anyone who likes this model I highly recommend trying it out, I feel you will be very happy with the results. Happy Diffusing! :)
For best results I found
CFG - 1 - 2 (1.5 I found to be the best)
Steps - 4 - 8
Sampler - DPM++ SDE
Scheduler - Karras
As you can expect I've merged the new JuggernautXL_X_RunDiffusion model into AmalgamationXL. Had done a few merges before hand as well which I hadn't uploaded hence the jump to V2.
Electrifying Speed and Quality
The AmalgamationXL V1.1-Tau-Lightning model represents the latest fusion of top SDXL models and the new SDXL-Lightning models. This iteration takes the robust foundation of AmalgamationXL and infuses it with the unparalleled agility of SDXL-Lightning models, resulting in a version that redefines the limits of image generation. V1.1-Tau-Lightning is designed to cater to creators seeking not just excellence in visual fidelity but also unprecedented speed in bringing their visions to life.
I appreciate if people want to test out the model's capabilities and help run testing
<3 Have at it!
I've made a SD 1.5 Amalgamation Checkpoint as well
https://civitai.com/models/363285/amalgamation-sd15
AmalgamationXL V1.0-Mu
After many merges of some new amazing models, I've finally reached V1.0-Mu.
When I have time I will go through a more detailed breakdown of the merges that led to this process but for now enjoy this amazing model. Improved in all aspects of realism, coherency and prompt adherence.
AmalgamationXL-V0.7-Theta stands as the zenith of recursive model enhancement, a fusion of art and technology that has traversed several transformative phases. Each recursive merge has been a careful curation, integrating the strengths of specialized Stable Diffusion models to form a versatile and powerful tool for image generation.
AmalgamationXL-V0.7-Theta
This model is not just an iteration; it is a gateway to a multiverse of imagery. AmalgamationXL-V0.7-Theta is your artistic partner, capable of rendering images with a level of detail and realism that breathes life into your creative vision. It is a culmination of relentless innovation, standing as a versatile, robust, and adaptive model for any creator seeking to push the boundaries of image synthesis.
This model (V0.7-Theta) seems to have better overall cohesion, realism and prompt adherence.
Generation Guide:
Steps: 20-50
Sampler: DPM++ 2M SDE
Scheduler: Karras
If you like what I do, feel free to have a look at my ko-fi or patreon where I have a bunch of ComfyUI workflows and other Stable Diffusion related things.
I also offer custom AI services and custom model merging on my Patreon
Made a lightning version of the V2-X merge and I am quite blown away by it myself.