Noobai XL Merges

🚀 NoobAI-XL-Merges

A collection of various merges based on Laxhar Lab's Illustrious - xl text - to - image model, uploaded for testing purposes.

These models are provided as they are, and results may vary. Users are responsible for all outputs generated using these checkpoints. Other models involved in these merges are:

• comin/IterComp
• CyberRealistic XL

🚀 Quick Start

This project consists of various text - to - image model merges. These models are built on Laxhar Lab's Illustrious - xl model and are uploaded for testing.

✨ Features

• Built on Laxhar Lab's Illustrious - xl text - to - image model.
• Involve multiple other well - known models in the merge process.

📚 Documentation

Methods

Perpendicular merges are carried out via [sd - mecha](https://github.com/ljleb/sd - mecha) using the Python API.

💻 Usage Examples

Basic Usage

import sd_mecha
sd_mecha.set_log_level()


text_encoder_recipe = sd_mecha.model("noobaiXLNAIXL_vPred10Version.safetensors", "sdxl")

unet_recipe = sd_mecha.add_perpendicular(
    sd_mecha.model("noobaiXLNAIXL_vPred10Version.safetensors", "sdxl"),
    sd_mecha.model("cyberrealisticXL_v4.safetensors", "sdxl"),
    sd_mecha.model("sd_xl_base_1.0_0.9vae.safetensors", "sdxl"),
)

recipe = sd_mecha.weighted_sum(
    text_encoder_recipe,
    unet_recipe,
    alpha=(
        sd_mecha.blocks("sdxl", "txt") |
        sd_mecha.blocks("sdxl", "txt2") |
        sd_mecha.default("sdxl", "unet", 1)
    ),
)

merger = sd_mecha.RecipeMerger(
    models_dir=r"C:\path\to\models\directory",
)

merger.merge_and_save(recipe, output="output.safetensors")

Advanced Usage

from safetensors.torch import load_file, save_file
import torch

state_dict = load_file("output.safetensors")
state_dict["v_pred"] = torch.tensor([])
state_dict["ztsnr"] = torch.tensor([])
save_file(state_dict, "noobaiXLNAIXL_vPred10Version-cyberrealistic4-perpendicular.safetensors")

📄 License

This project uses the Fair AI Public License 1.0 - SD.