[Cat Tower (NoobAI XL checkpoint)](https://civitai.com/models/920709/cat-tower-noobai-xl-checkpoint)

---

Additional versions

• [v1.0](https://pixai.art/model/1810565126476013375)
• [v1.1 (vPred)](https://pixai.art/model/1816356443530795524)
• [v1.2 (vPred)](https://pixai.art/model/1819075028869481161)

---

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 (latest commit, No need to install config file)
• reForge (latest commit, No need to install config file)
• Automatic1111 (dev branch, latest commit. And place the config file in the same folder as the checkpoint. If you change the name of the checkpoint, rename the basename without extension of config file to the same as the changed checkpoint name.)

Recommended Setting

• Steps: 25-30
• CFG scale: 5-7
• Sampler for vPred v1.1, v1.2, v1.3: Euler, Euler a
• Sampler for vPred v1: Euler
• Sampler: DPM++ 2M Karras, Euler a
• Scheduler for 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 (vPred v1.3)

1. Merging [NoobAI-V-Pred-0.65S-Version](https://civitai.com/models/833294?modelVersionId=1093948) and [IterComp](https://civitai.com/models/840857/itercomp) by [Noob Vpred itercomp merge's recipes.](https://civitai.com/models/980544/noob-vpred-itercomp-merge) (Checkpoint A)
2. 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 3.Merging the OFT model to Checkpoint A, ratio = 1.0 (Checkpoint B)
3. Merging Checkpoint B and [HarmoniqMix_vPred_v1](https://huggingface.co/hybskgks28275/checkpoint/tree/main/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)
4. Merging the difference between [Catloaf](https://civitai.com/models/615558) and [ponyDiffusionV6XL](https://civitai.com/models/257749/pony-diffusion-v6-xl?modelVersionId=290640) 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)
5. Merging the difference between [copycat-noob v0.3test](https://civitai.com/models/894218/copycat-noob?modelVersionId=1010358) and [NoobAI-XL Epsilon-pred 0.75-Version](https://civitai.com/models/833294?modelVersionId=998979) 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)
6. Merging the difference between ExilluSPO Anime v1 and [illustriousXL_smoothftSPO](https://civitai.com/models/811067) 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)
7. Merging NoobAI-V-Pred-0.65S-Version + (Checkpoint F - NoobAI-V-Pred-0.65S-Version) (Cat Tower vPred v1.2)

License

This model adheres to the [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/) and is available for non-commercial use only. Commercial use is prohibited.