Cat Tower is anime style NoobAI XL checkpoint.

VAE is baked in.

v-prediction versions are experimental models.

You need to use the webui that support v-prediction.

ComfyUI (recommended)
Forge Classic (recommended)
reForge
Forge (maybe)
AUTOMATIC1111 (from v1.4, and you need to place the config file in the same folder as the checkpoint)

License

This model adheres to the NoobAI-XL License (Modified Fair AI Public License 1.0-SD) and is available for non-commercial use only. This license prohibits any form of commercialization, including but not limited to monetization or commercial use of the model, derivative models, or model-generated products.

Recommended Setting

Steps: 25-30
CFG scale: 5-7
Sampler (vPred): I recommend using "Euler". Using "Euler a" sometimes generates colored skin and objects such as red, blue and so on.
Sampler (Epsilon-pred): DPM++ 2M Karras, Euler a
Scheduler (vPred) : Normal, Simple ("Automatic" produce noisy images.)
Use ADetailer as needed

Positive Prompt

masterpiece, best quality

Negative Prompt

worst quality, bad quality, low quality, lowres, scan artifacts, jpeg artifacts, sketch, light particles

Merge recipes (ChenkinNoob v1.1)

Merging the difference between ChenkinNoob-XL-V0.2 and stable-diffusion-xl-base-1.0, CyberRealistic XL v7.0 and stable-diffusion-xl-base-1.0 to ChenkinNoob-XL-V0.2 by "Perpendicular Component" and "Add Difference" (Checkpoint A)
Merging Checkpoint A to ChenkinNoob-XL-V0.2 by "SLERP", alpha = 0.5 (Checkpoint B)
Merging Checkpoint B to ChenkinNoob-XL-V0.2 by "Rotate", alignment = 1.0, alpha = 0.0 (Checkpoint C)
Replacing the CLIP of Checkpoint C by ChenkinNoob-XL-V0.2 (Checkpoint D)
Finetuning Checkpoint D with anime style dataset(0.5k) by OFT method, dim = 4, alpha = 1e-3, learning rate = 2.5e-6, 15000 steps. And merging the OFT model to ChenkinNoob-XL-V0.2, ratio = 1.0 (Checkpoint E)
Merging Checkpoint E to Checkpoint D by "SLERP" with block (0,0,0,0,0,0,0,0,0,0,0,0.7,0.7,0.7,0.7,0,0,0,0,0) (Checkpoint F)
Merging Cat Carrier v8.0 to Checkpoint F by "Weighted Sum" with block (0,0,0,0,0,0,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0,0,0,0,0) (Checkpoint G)
Merging the difference between Cat Carrier v8.0 and Raehoshi illust XL v7.1 to Checkpoint G by "Add Difference" with block (0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0) (Checkpoint H)
Merging the difference between LunarPeachMix v2.0 and Illustrious XL v1.1 to Checkpoint H by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.3,0.3,0,0,0,0,0,0,0) (Checkpoint I)
Merging the difference between copycat-RouWei vpred-v0.42 and Rouwei 0.80 vpred to Checkpoint I by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.3,0.3,0,0,0,0,0,0,0) (Checkpoint J)
Merging style LoRA to Checkpoint J (Cat Tower ChenkinNoob v1.1)

Merge recipes (ChenkinNoob v1.0)

Merging the difference between ChenkinNoob-XL-V0.1 and stable-diffusion-xl-base-1.0, CyberRealistic XL v7.0 and stable-diffusion-xl-base-1.0 to ChenkinNoob-XL-V0.1 by "Perpendicular Component" and "Add Difference" (Checkpoint A)
Merging Checkpoint A to ChenkinNoob-XL-V0.1 by "SLERP", alpha = 0.5 (Checkpoint B)
Merging Checkpoint B to ChenkinNoob-XL-V0.1 by "Rotate", alignment = 1.0, alpha = 0.0 (Checkpoint C)
Replacing the CLIP of Checkpoint C by ChenkinNoob-XL-V0.1 (Checkpoint D)
Finetuning Checkpoint D with anime style dataset(0.5k) by OFT method, dim = 4, alpha = 1e-3, learning rate = 2.5e-6, 15000 steps. And merging the OFT model to ChenkinNoob-XL-V0.1, ratio = 1.0 (Checkpoint E)
Merging Checkpoint E to Checkpoint D by "SLERP" with block (0,0,0,0,0,0,0,0,0,0,0,0.7,0.7,0.7,0.7,0,0,0,0,0) (Checkpoint F)
Merging Cat Carrier v8.0 to Checkpoint F by "Weighted Sum" with block (0,0,0,0,0,0,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0,0,0,0,0) (Checkpoint G)
Merging the difference between Cat Carrier v8.0 and Raehoshi illust XL v7.1 to Checkpoint G by "Add Difference" with block (0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0) (Checkpoint H)
Merging the difference between LunarPeachMix v2.0 and Illustrious XL v1.1 to Checkpoint H by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.3,0.3,0,0,0,0,0,0,0) (Checkpoint I)
Merging the difference between copycat-RouWei vpred-v0.42 and Rouwei 0.80 vpred to Checkpoint I by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.3,0.3,0,0,0,0,0,0,0) (Checkpoint J)
Merging style LoRA to Checkpoint J (Cat Tower ChenkinNoob v1.0)

Merge recipes (Obsession vPred v2.0)

Merge recipes (Epsilon-pred v1.4)

Merge recipes (vPred v1.8)

Merging the difference between NoobAI-V-Pred-1.0-Version and stable-diffusion-xl-base-1.0, CyberRealistic XL v5.6 and stable-diffusion-xl-base-1.0 to NoobAI-V-Pred-1.0-Version by "Perpendicular Component" and "Add Difference" (Checkpoint A)
Merging Checkpoint A to NoobAI-V-Pred-1.0-Version by "SLERP", alpha = 0.5 (Checkpoint B)
Merging Checkpoint B to NoobAI-V-Pred-1.0-Version by "Rotate", alignment = 1.0, alpha = 0.0 (Checkpoint C)
Replacing the CLIP of Checkpoint C by NoobAI-V-Pred-1.0-Version (Checkpoint D)
Finetuning Checkpoint D with anime style dataset(0.5k) by OFT method, dim = 4, alpha = 1e-3, learning rate = 2.5e-6, 15000 steps. And merging the OFT model to Checkpoint D, ratio = 1.0 (Checkpoint E)
Merging Cat Carrier v6.0 to Checkpoint E by "Weighted Sum" with block (0,0,0,0,0,0,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0,0,0,0,0) (Checkpoint F)
Merging the difference between LunarPeachMix v2.0 and Illustrious XL v1.1 to Checkpoint F by "Add Difference" with block (0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,0,0,0,0,0) (Checkpoint G)
Merging the difference between JANKU v4.0 and RouWei v0.7 epred to Checkpoint G by "Add Difference" with block (0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,0,0,0,0,0) (Checkpoint H)
Merging the difference between copycat-RouWei vpred-v0.42 and Rouwei 0.80 vpred to Checkpoint H by "Add Difference" with block (0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,0,0,0,0,0) (Checkpoint I)
Merging the Checkpoint I by "Clamp", bounds = Checkpoint F, LunarPeachMix v2.0, JANKU v4.0 and copycat-RouWei vpred-v0.42 (Checkpoint J)
Add v_pred and ztsnr keys to Checkpoint J by noob_v_pencil-XL merge recipes (Cat Tower vPred v1.8)

Merge recipes (Epsilon-pred v1.3)

Merging Cat Tower Epsilon-pred v1.2 and Cat Tower v-pred v1.7 by "DARE-TIES" (Cat Tower Epsilon-pred v1.3)

Merge recipes (vPred v1.7)

Merging the difference between CyberRealistic XL v5 and stable-diffusion-xl-base-1.0 to NoobAI-V-Pred-1.0-Version by "Add Difference" (Checkpoint A)
Replacing the CLIP of Checkpoint A by NoobAI-V-Pred-1.0-Version (Checkpoint B)
Merging Cat Carrier v3.0 to Checkpoint B by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.4,0.4,0.4,0.4,0.4,0.4,0.0,0,0,0,0) (Checkpoint C)
Merging Cat Tower vpred v1.5 to Checkpoint C by "Add Cosine B" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint D)
Merging Cat Bread eps-pred v1.0 to Checkpoint D by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint E)
Merging the difference between Catloaf and ponyDiffusionV6XL to Checkpoint E by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint F)
Merging the difference between copycat-noob Vpred_v1.01 and NoobAI-V-Pred-1.0-Version to Checkpoint F by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint G)
Merging style LoRA to Checkpoint G (Checkpoint H)
Add v_pred and ztsnr keys to Checkpoint H by noob_v_pencil-XL merge recipes (Cat Tower vPred v1.7)

Merge recipes (vPred v1.6)

Merging NoobAI-V-Pred-1.0-Version and IterComp by Noob Vpred itercomp merge's recipes. (Checkpoint A)
Finetuning NoobAI-V-Pred-1.0-Version with anime style dataset(0.5k) by OFT method, dim = 8, alpha = 1e-3, learning rate = 1e-4, 15000 steps
Merging the OFT model to Checkpoint A, ratio = 1.0 (Checkpoint B)
Merging Checkpoiint B to Checkpoint A by "Weighted Sum", alpha = 0.4 (Checkpoint C)
Merging Cat Carrier v3.0 to Checkpoint C by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.6,0.6,0.6,0.4,0.6,0.6,0.0,0,0,0,0) (Checkpoint D)
Merging Checkpoint D and HarmoniqMix_vPred_v1 by "Weighted Sum" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint E)
Merging the difference between Catloaf and ponyDiffusionV6XL to Checkpoint E by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint F)
Merging the difference between copycat-noob Vpred_v1.01 and NoobAI-V-Pred-1.0-Version to Checkpoint F by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint G)
Merging style LoRA to Checkpoint G (Checkpoint H)
Add v_pred and ztsnr keys to Checkpoint H by noob_v_pencil-XL merge recipes (Cat Tower vPred v1.6)

Merge recipes (Epsilon-pred v1.2)

Merging the difference between cyberrealisticXL_v31 and stable-diffusion-xl-base-1.0 to NoobAI-XL Epsilon-pred v1.1 by "Add Difference", alpha=1.0. (Checkpoint A)
Replacing the CLIP of Checkpoint A by NoobAI-XL Epsilon-pred v1.1. (Checkpoint B)
Merging the difference between Cat Carrier v3.0 and Checkpoint B by "Add Difference" with blocks (0.01,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.35,0.35,0.35,0.2,0.35,0.35,0,0,0,0,0). (Checkpoint C)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred v0.75 to Checkpoint C by "Train Difference" with blocks (0,0,0,0,0,0,0,0,0.2,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0). (Checkpoint D)
Merging style LoRA to Checkpoint D (Cat Tower Epsilon-pred v1.2)

Merge recipes (vPred v1.5)

Merging NoobAI-V-Pred-1.0-Version and IterComp by Noob Vpred itercomp merge's recipes. (Checkpoint A)
Finetuning NoobAI-V-Pred-1.0-Version with anime style dataset(0.5k) by OFT method, dim = 8, alpha = 1e-3, learning rate = 1e-4, 15000 steps
Merging the OFT model to Checkpoint A, ratio = 1.0 (Checkpoint B)
Merging Checkpoint B and Cat Tower v1.1 Epsilon-pred by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.4,0.4,0.4,0.4,0.0,0,0,0,0) (Checkpoint C)
Merging Checkpoint C and HarmoniqMix_vPred_v1 by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.0,0,0,0,0) (Checkpoint D)
Merging the difference between Catloaf and ponyDiffusionV6XL to Checkpoint D by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint E)
Merging the difference between copycat-noob Vpred_v0.4 and Noobai-XL-V-pred-0.75S-Version to Checkpoint E by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint F)
Add v_pred and ztsnr keys to Checkpoint F by noob_v_pencil-XL merge recipes (Cat Tower vPred v1.5)

Merge recipes (vPred v1.4)

Merging NoobAI-V-Pred-0.75S-Version and IterComp by Noob Vpred itercomp merge's recipes. (Checkpoint A)
Finetuning NoobAI-V-Pred-0.75S-Version with anime style dataset(0.5k) by OFT method, dim = 8, alpha = 1e-3, learning rate = 1e-4, 15000 steps
Merging the OFT model to Checkpoint A, ratio = 1.0 (Checkpoint B)
Merging Checkpoint B and Cat Tower v1.1 Epsilon-pred by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.4,0.4,0.4,0.4,0.0,0,0,0,0) (Checkpoint C)
Merging Checkpoint C and HarmoniqMix_vPred_v1 by "Weighted Sum" with block (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.0,0,0,0,0) (Checkpoint D)
Merging the difference between Catloaf and ponyDiffusionV6XL to Checkpoint D by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint E)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint E by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint F)
Merging NoobAI-V-Pred-0.75S-Version and Checkpoint F by "Weighted Sum" with block (0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1) (Cat Tower vPred v1.4)

Merge recipes (Epsilon-pred v1.1)

Merging NoobAI-XL Epsilon-pred 1.1-Version and cyberrealisticXL_v31, and replace CLIP by NoobAI-XL Epsilon-pred 1.1-Version. It's a same method of NoobaiCyberFix. (Checkpoint A)
Merging the difference between Cat Carrier v1.0 and Raehoshi illust XL v1.0 to Checkpoint A by "trainDifference" with blocks (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.4,0.2,0.2,0.4,0.6,0.6,0.6,0.2,0.2,0.2,0.2,0.2) (Checkpoint B)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint B by "trainDifference" with blocks (0,0,0,0,0,0,0,0,0.2,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint C)
Merging style LoRA to Checkpoint C (Cat Tower Epsilon-pred v1.1)

Merge recipes (vPred v1.3)

Merging NoobAI-V-Pred-0.65S-Version and IterComp by Noob Vpred itercomp merge's recipes. (Checkpoint A)
Finetuning NoobAI-V-Pred-0.65S-Version with anime style dataset(0.5k) by OFT method, dim = 8, alpha = 1e-3, learning rate = 1e-4, 15000 steps
Merging the OFT model to Checkpoint A, ratio = 1.0 (Checkpoint B)
Merging Checkpoint B and HarmoniqMix_vPred_v1 by "Weighted Sum" with block (0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.5,0.5,0.5,0.5,0.0,0,0,0,0) (Checkpoint C)
Merging the difference between Catloaf and ponyDiffusionV6XL to Checkpoint C by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint D)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint D by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint E)
Merging the difference between ExilluSPO Anime v1 and illustriousXL_smoothftSPO to Checkpoint D by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0.15,0.15,0.15,0.15,0,0,0,0,0) (Checkpoint F)
Merging NoobAI-V-Pred-0.65S-Version + (Checkpoint F - NoobAI-V-Pred-0.65S-Version) (Cat Tower vPred v1.3)

Merge recipes (vPred v1.2)

Merging NoobAI-V-Pred-0.65S-Version and cyberrealisticXL_v31, and replace CLIP by NoobAI-V-Pred-0.65S-Version. It's a same method of NoobaiCyberFix. (Checkpoint A)
Merging NoobAI-V-Pred-0.65S-Version and IterComp by Noob Vpred itercomp merge's recipes. (Checkpoint B)
Merging Checkpoint A * (1 - 0.5) + Checkpoint B * 0.5 (Checkpoint C)
Merging the difference between Cat Carrier v1.0 and Raehoshi illust XL v1.0 to Checkpoint C by "Train Difference" with block (0,0,0,0,0,0,0,0,0.25,0,0,0.25,0.45,0.45,0.45,0,0,0,0,0) (Checkpoint D)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint D by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint E)
Merging style LoRA to Checkpoint E (Checkpoint F)
Merging NoobAI-V-Pred-0.65S-Version + (Checkpoint F - NoobAI-V-Pred-0.65S-Version) with adjust (0,0,0,0,0,2.34,0,-0.95). (Cat Tower vPred v1.2)

Merge recipes (vPred v1.1)

Merging NoobAI-V-Pred-0.6-Version and cyberrealisticXL_v31, and replace CLIP by NoobAI-V-Pred-0.6-Version. It's a same method of NoobaiCyberFix. (Checkpoint A)
Merging the difference between Cat Carrier v1.0 and Raehoshi illust XL v1.0 to Checkpoint A by "Train Difference" with block (0,0,0,0,0,0,0,0,0.3,0,0,0.3,0.5,0.5,0.5,0,0,0,0,0) (Checkpoint B)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint B by "Train Difference" with block (0,0,0,0,0,0,0,0,0,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0) (Checkpoint C)
Merging style LoRA to Checkpoint C (Checkpoint D)
Merging NoobAI-V-Pred-0.6-Version and Checkpoint D by the following calc, NoobAI-V-Pred-0.6-Version + (Checkpoint D - NoobAI-V-Pred-0.6-Version). (Cat Tower vPred v1.1)

Merge recipes (vPred v1)

Merging NoobAI-V-Pred-0.5-Version and cyberrealisticXL_v31, and replace CLIP by NoobAI-V-Pred-0.5-Version. It's a same method of NoobaiCyberFix. (Checkpoint A)
Merging the subtract (HarmoniqMix_vPred_v1 - NoobAI-V-Pred-0.5-Version) to Checkpoint A by "Add Difference" and "Clamp" (Checkpoint B)
Merging the difference between Cat Carrier v1.0 and Raehoshi illust XL v1.0 to Checkpoint B by "Train Difference" with block (0.1,0.1,0.1,0.1,0.1,0.1,0.1,0.2,0.1,0.1,0.2,0.3,0.3,0.3,0.1,0.1,0.1,0.1,0.1) (Checkpoint C)
Merging NoobAI-V-Pred-0.5-Version and Checkpoint C by the following calc, NoobAI-V-Pred-0.5-Version + (Checkpoint C - NoobAI-V-Pred-0.5-Version). (Cat Tower vPred v1)

Merge recipes (Epsilon-pred v1)

Merging the difference between Cat Carrier v1.0 and Raehoshi illust XL v1.0 to NoobaiCyberFix v1.0, NoobaiCyberFix + (Cat Carrier - Raehoshi illust XL) x alpha (0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.4,0.2,0.2,0.4,0.6,0.6,0.6,0.2,0.2,0.2,0.2,0.2), mode = Add difference, calcmode = trainDifference (Checkpoint A)
Merging the difference between copycat-noob v0.3test and NoobAI-XL Epsilon-pred 0.75-Version to Checkpoint A, Checkpoint A + (copycat-noob - NoobAI-XL) x alpha (0,0,0,0,0,0,0,0,0.2,0,0,0.2,0.2,0.2,0.2,0,0,0,0,0), mode = Add difference, calcmode = trainDifference (Checkpoint B)
Merging style LoRA to Checkpoint B (Cat Tower Epsilon-pred v1)

Files with SHA256: 3d48f30674a6f139c6d093100ac050c4df4be90fea1bf8275f15f797ff53dce6