LoRA XL Models - CivitAI Repository ?
Welcome to the FFusion LoRA extracted models repository on Hugging Face & CivitAI! Here, we present a collection of models extracted using the Low-Rank Adaptation (LoRA) technique to provide a rich dataset for research and further exploration.
? FFusion's Universe of ? Curated LoRA Extractions
Our LoRAs are carefully extracted from a variety of models, allowing you to mix and match styles to create truly unique and artistic fusions. These extracted LoRAs are not a direct copy; they capture the essence of the original model, adding a creative influence "In the style of" or "Influenced by" the original work.
? Research-Focused LoRAs
Please note that all FFusionAI extracted LoRAs are intended for research purposes only and are not licensed for commercial use. We encourage responsible and ethical utilization of these LoRAs to advance the field of AI-powered art creation.
⚠️ License & Usage Disclaimers
Please review the full license agreement before accessing or using the models.
The correct license and permissions can be found at:
https://huggingface.co/FFusion/
https://huggingface.co/FFusion/FFXL400/blob/main/LICENSE.md
"Model Weights: The weights used for the models/loras are provided "as is." FFusion AI and Source Code Bulgaria do not grant any rights for their commercial use. These weights are strictly for testing and experimental purposes.
ORIGIN OF LORAS:
The LORAs and weights provided are extracted from SDXL models (checkpoints).
All licenses, terms, and conditions set forth by the original checkpoint creator must be respected and adhered to."? The models and weights available in this repository are strictly for research and testing purposes, with exceptions noted below. They are not generally intended for commercial use and are dependent on each individual LORA.
? Exception for Commercial Use: The FFusionXL-BASE, FFusion-BaSE, di.FFUSION.ai-v2.1-768-BaSE-alpha, and di.ffusion.ai.Beta512 models are trained by FFusion AI using images for which we hold licenses. Users are advised to primarily use these models for a safer experience. These particular models are allowed for commercial use.
? Disclaimer: FFusion AI, in conjunction with Source Code Bulgaria Ltd and BlackswanTechnologies, does not endorse or guarantee the content produced by the weights in each LORA. There's potential for generating NSFW or offensive content. Collectively, we expressly disclaim responsibility for the outcomes and content produced by these weights.
? Acknowledgement: The FFusionXL-BASE model model is a uniquely developed version by FFusion AI. Rights to this and associated modifications belong to FFusion AI and Source Code Bulgaria Ltd. Ensure adherence to both this license and any conditions set by Stability AI Ltd for referenced models.
Unparalleled Customization
Unlike conventional models that limit LoRA strength to a narrow range, FFusionAI provides unparalleled flexibility. Adjust LoRA strength from 0.2 for subtle effects to 2.2 for intense transformations. This extended range ensures that you have the tools to achieve the perfect stylistic blend, regardless of the base model or desired outcome.
? Recommended Strength Settings for FF100+ ?
? Visuals: Boost up to 2.2 for vibrant and striking detail.
? Fusing Loras: Maintain 0.3 - 1.0 for seamless and safe integration with up to 6 FF Lora from FF100 above.
? Main Base Model for Testing:
? UPDATE: 10/22/23 ?
? Introducing our upcoming batch of LoRAs numbered FF.100 to FF.176!
? Optimized new Size: ~200 - 400MB (depending on original models training and weights)
?️ New Naming: Optimized experience on Hugging Face for faster inference and testing.
pipe = DiffusionPipeline.from_pretrained("FFusion/FFusionXL-BASE", torch_dtype=torch.float16).to("cuda")
lora_model_id = "FFusion/400GB-LoraXL"
lora_filename = "FF.101.safetensors"
pipe.load_lora_weights(lora_model_id, weight_name=lora_filename)CivitAI naming format stays the same.
Loading from CivitAI in diffusers
Todo: ? Sync CivitAI Repo: Up to date to FF98
Latest FF60-FF98
The following models served as the foundation for our extractions: For those on the quest for ideal models to drive their inference tasks, we especially recommend: FFusionXL-BASE - Our signature base model, meticulously trained with licensed images. FFXL400 Combined LoRA Model ? - A galactic blend of power and precision in the world of LoRA models. Rest assured, our LoRAs, even at weight 1.0, maintain compatibility with most of the current SDXL models. Variants: Each base model was extracted into 4-5 distinct variants. Extraction Depth: The models uploaded here contain approximately 70% of extracted data. These extractions yield a dataset size of around 400 GB. Precision: We experimented with both Differences Measurement: Singular Value Decomposition (SVD) was utilized to measure differences between the original and the tuned models. A threshold of 1e-3 was commonly used, although in some cases, 1e-5 and 1e-2 were tested. Demonstration Parameters: For our demonstration, we employed Most SDXL models in this collection are not traditionally "trained." Instead, they are merged from previous SDXL 0.9 versions or created using other methods with the help of Comfy UI. An important note for users: all models saved with Comfy add an extra key These extracted models are intended for research and testing. They can be particularly useful for: Investigating the potential of merging multiple LoRAs. Weighting experiments with 1-5 LoRAs simultaneously. Exploring the differences and similarities between LoRAs extracted from different base models. Welcome to the technical guide for using the FFusion LoRA extracted models. This document will walk you through the steps required to fuse LoRA parameters, load checkpoints, and perform inference. To merge the LoRA parameters with the original parameters of the underlying model(s), leading to a potential speedup in inference latency: To revert the effects of To control the influence of the LoRA parameters on the outputs: Here's how to load and utilize our FFusion models: After loading the desired model, you can perform inference as follows: You can choose any of the models from our repository on Hugging Face or the upcoming repository on CivitAI. Here's a list of available models with Based on the extraction process, we observed the following differences in the text encoder across various models: A huge shoutout to the community for their continued support and feedback. Together, we are pushing the boundaries of what's possible with machine learning! We would also like to acknowledge and give credit to the following projects and authors: ComfyUI: We've used and modified portions of ComfyUI for our work. kohya-ss/sd-scripts and bmaltais: Our work also incorporates modifications from kohya-ss/sd-scripts. lora-inspector: We've benefited from the lora-inspector project. KohakuBlueleaf: Special mention to KohakuBlueleaf for their invaluable contributions. It outputs: The total storage capacity of each scanned drive or directory. The space occupied by The free space available. A neat bar chart visualizing the above data. You can easily install Clone the repository: Navigate to the cloned directory and install: Run the tool without any arguments to scan all drives: Or, specify a particular directory or drive to scan: Proudly maintained by Source Code Bulgaria Ltd & Black Swan Technologies. ? Email Us: [email protected] - For inquiries or support. ? Locations: Sofia | Istanbul | London Connect with Us: Our Websites: FFai.0011.Bluepencilxl_V021.lora 名称: FFai.0011.Bluepencilxl_V021.lora.safetensors 大小 (KB): 631051 类型: Model Pickle 扫描结果: Success Pickle 扫描信息: No Pickle imports 病毒扫描结果: Success 
? Base Models
? Recommended Models for Inference
? Extraction Details
float32 and float64 for optimal extraction results."conv_dim": 256 and "conv_alpha": 256.
⚙️ Technical Notes
text_model.encoder.text_model.embeddings.position_ids. We've made necessary adjustments to ensure compatibility with the current scripts from Kohoya.? Use Cases
FFusion LoRA Extracted Models - How to Use Guide ?
Fusing LoRA Parameters ?
pipe.fuse_lora()
Unfusing LoRA Parameters ⛓️
fuse_lora():pipe.unfuse_lora()
Working with Different LoRA Scales ?️
pipe.fuse_lora(lora_scale=0.5)
Working with FFusion Models ?
from diffusers import DiffusionPipeline
import torch
pipeline_id = "FFusion/FFusionXL-BASE"
pipe = DiffusionPipeline.from_pretrained(pipeline_id, torch_dtype=torch.float16)
pipe.enable_model_cpu_offload()
lora_model_id = "FFusion/400GB-LoraXL"
lora_filename = "FFai.0038.Realitycheckxl_Alpha11.lora.safetensors"
pipe.load_lora_weights(lora_model_id, weight_name=lora_filename)
prompt = "papercut sonic"
image = pipe(prompt=prompt, num_inference_steps=20, generator=torch.manual_seed(0)).images[0]
image
Running Inference ?️
generator = torch.manual_seed(0)
images_fusion = pipe(
"masterpiece, best quality, mountain", output_type="np", generator=generator, num_inference_steps=25
).images
Library of Available LoRA Models ?
lora_model_id = "FFusion/400GB-LoraXL":lora_filename =
- FFai.0001.4Guofeng4xl_V1125d.lora_Dim64.safetensors
- FFai.0002.4Guofeng4xl_V1125d.lora_Dim8.safetensors
- FFai.0003.4Guofeng4xl_V1125d.loraa.safetensors
- FFai.0004.Ambiencesdxl_A1.lora.safetensors
- FFai.0005.Ambiencesdxl_A1.lora_8.safetensors
- FFai.0006.Angrasdxl10_V22.lora.safetensors
- FFai.0007.Animaginexl_V10.lora.safetensors
- FFai.0008.Animeartdiffusionxl_Alpha3.lora.safetensors
- FFai.0009.Astreapixiexlanime_V16.lora.safetensors
- FFai.0010.Bluepencilxl_V010.lora.safetensors
- FFai.0011.Bluepencilxl_V021.lora.safetensors
- FFai.0012.Breakdomainxl_V03d.lora.safetensors
- FFai.0013.Canvasxl_Bfloat16v002.lora.safetensors
- FFai.0014.Cherrypickerxl_V20.lora.safetensors
- FFai.0015.Copaxtimelessxlsdxl1_V44.lora.safetensors
- FFai.0016.Counterfeitxl-Ffusionai-Alpha-Vae.lora.safetensors
- FFai.0017.Counterfeitxl_V10.lora.safetensors
- FFai.0018.Crystalclearxl_Ccxl.lora.safetensors
- FFai.0019.Deepbluexl_V006.lora.safetensors
- FFai.0020.Dream-Ffusion-Shaper.lora.safetensors
- FFai.0021.Dreamshaperxl10_Alpha2xl10.lora.safetensors
- FFai.0022.Duchaitenaiartsdxl_V10.lora.safetensors
- FFai.0023.Dynavisionxlallinonestylized_Beta0371bakedvae.lora.safetensors
- FFai.0024.Dynavisionxlallinonestylized_Beta0411bakedvae.lora.safetensors
- FFai.0025.Fantasticcharacters_V55.lora.safetensors
- FFai.0026.Fenrisxl_V55.lora.safetensors
- FFai.0027.Fudukimix_V10.lora.safetensors
- FFai.0028.Infinianimexl_V16.lora.safetensors
- FFai.0029.Juggernautxl_Version1.lora_1.safetensors
- FFai.0030.Lahmysterioussdxl_V330.lora.safetensors
- FFai.0031.Mbbxlultimate_V10rc.lora.safetensors
- FFai.0032.Miamodelsfwnsfwsdxl_V30.lora.safetensors
- FFai.0033.Morphxl_V10.lora.safetensors
- FFai.0034.Nightvisionxlphotorealisticportrait_Beta0681bakedvae.lora_1.safetensors
- FFai.0035.Osorubeshialphaxl_Z.lora.safetensors
- FFai.0036.Physiogenxl_V04.lora.safetensors
- FFai.0037.Protovisionxlhighfidelity3d_Beta0520bakedvae.lora.safetensors
- FFai.0038.Realitycheckxl_Alpha11.lora.safetensors
- FFai.0039.Realmixxl_V10.lora.safetensors
- FFai.0040.Reproductionsdxl_V31.lora.safetensors
- FFai.0041.Rundiffusionxl_Beta.lora.safetensors
- FFai.0042.Samaritan3dcartoon_V40sdxl.lora.safetensors
- FFai.0043.Sdvn6realxl_Detailface.lora.safetensors
- FFai.0044.Sdvn7realartxl_Beta2.lora.safetensors
- FFai.0045.Sdxl10arienmixxlasian_V10.lora.safetensors
- FFai.0046.Sdxlbasensfwfaces_Sdxlnsfwfaces03.lora.safetensors
- FFai.0047.Sdxlfaetastic_V10.lora.safetensors
- FFai.0048.Sdxlfixedvaefp16remove_Basefxiedvaev2fp16.lora.safetensors
- FFai.0049.Sdxlnijiv4_Sdxlnijiv4.lora.safetensors
- FFai.0050.Sdxlronghua_V11.lora.safetensors
- FFai.0051.Sdxlunstablediffusers_V5unchainedslayer.lora.safetensors
- FFai.0052.Sdxlyamersanimeultra_Yamersanimev2.lora.safetensors
- FFai.0053.Shikianimexl_V10.lora.safetensors
- FFai.0054.Spectrumblendx_V10.lora.safetensors
- FFai.0055.Stablediffusionxl_V30.lora.safetensors
- FFai.0056.Talmendoxlsdxl_V11beta.lora.safetensors
- FFai.0057.Wizard_V10.lora.safetensors
- FFai.0058.Wyvernmix15xl_Xlv11.lora.safetensors
- FFai.0059.Xl13asmodeussfwnsfw_V17bakedvae.lora.safetensors
- FFai.0060.Xl3experimentalsd10xl_V10.lora.safetensors
- FFai.0061.Xl6hephaistossd10xlsfw_V21bakedvaefp16fix.lora.safetensors
- FFai.0062.Xlperfectdesign_V2ultimateartwork.lora.safetensors
- FFai.0063.Xlyamersrealistic_V3.lora.safetensors
- FFai.0064.Xxmix9realisticsdxl_Testv20.lora.safetensors
- FFai.0065.Zavychromaxl_B2.lora.safetensors
? Text Encoder Difference Overview
? Acknowledgements & Citations
HowMuch ???
HowMuch is a Python tool designed to scan your drives (or a specified directory) and report on the total space used by files with specific extensions, mainly .ckpt and .safetensors.
.ckpt and .safetensors files.Installation
From PyPI
HowMuch via pip:pip install howmuch
From Source
git clone https://github.com/1e-2/HowMuch.git
cd HowMuch
pip install .
Usage
howmuch
howmuch --scan C:
? FFusion.ai Contact Information
描述:
训练词语:
