"DAMN!" (c) - my first thought when the finetuning ended and I saw the final results. UPD 19.05: Illustrious-based version REBORN is out! It turned out really costly to train, but I feel like the end result is well worth it. If you want to support me or help me cover the costs: https://boosty.to/coyotte/donate or Bitcoin 1BiNMKwVADy68rUrRxCR3mr3TGewWvrtUV. Every donation will go towards developing chekpoints, workflows and loras (this stuff is costly). I have big plans, but little means to fulfill them 💀 Important prompt structure: Positive prompt: masterpiece, best quality, realistic Negative prompt: bad quality, worst quality, low quality For now, you still need these, but as the model will get more training, the need for them will decrease.
RECOMENDED PARAMETERS:
◦ Sampler: DPM++ 2M SDE / DPM++ 3M SDE / DPM++ SDE
◦ 30-40 Steps.
◦ CFG: 4-6
◦ The model was on trained mostly on 1024x1024, 832x1216, 1216x832, 896x1152, 1152x896 resolutions, so it could help the quality if you used these (although that's not necessary - it behaves nicely on other aspect ratios as well).
◦ Highres fix: Upscale by 1.35-1.5, denoising strength ~0.4, upscaler - 4x NMKD Superscale or Siax.
For the distant shots of people you probably wanna use highres.fix and/or adetailer.