Skip to content

Realistic Image Generation, Realistic trigger works properly, better for photorealistic trigger words, close-up shots, face diffusion, male, female characters.

Notifications You must be signed in to change notification settings

PRITHIVSAKTHIUR/Canopus-Realism

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

title emoji colorFrom colorTo sdk sdk_version app_file pinned license header short_description
CANOPUS REALISM
🧒🏻
red
gray
gradio
4.37.2
app.py
false
creativeml-openrail-m
mini
Canopus Realism | Text to Image LoRA

⭐⭐⭐ Model Repository : https://huggingface.co/prithivMLmods/Canopus-Realism-LoRA ⭐⭐⭐

alt text

Run in Colab / kaggle notebooks

🧲 : https://colab.research.google.com/drive/18dLI-YQPOP74ZQmEn6MhymZHZTzyD677?usp=sharing

Procedure :
Get HF_TOKEN (access token) to access the model.
Get Logged into Huggingface✅

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Spaces: https://huggingface.co/spaces/prithivMLmods/Canopus-Realism

alt text

Clone by :

# Make sure you have git-lfs installed (https://git-lfs.com)
git lfs install

git clone https://huggingface.co/spaces/prithivMLmods/Canopus-Realism

# If you want to clone without large files - just their pointers

GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/spaces/prithivMLmods/Canopus-Realism

🧒🏻Demo Images :

Image 1 Image 2
Image 3 Image 4

Working Demo

alt text

Screenshots

alt text

Requirements.txt

torch diffusers transformers safetensors
accelerate spaces peft pillow

Model Dx

alt text


Canopus-Realism-LoRA

The LoRA for the Realism model was trained based on Stable Diffusion XL, working paired with the RealVisXL_V4.0_Lightning model to obtain better diffusion and enhanced results.

Parameter Description
Space Name Canopus-Realism
Model Repo ⚡Datasets Used⚡
Default Theme Light by Defalut / Can be altered
Description Canopus-Realism provides realistic image generation capabilities with a focus on high-quality outputs.

Training Parameters

Parameter Value Parameter Value
LR Scheduler constant Noise Offset 0.03
Optimizer AdamW Multires Noise Discount 0.1
Network Dim 64 Multires Noise Iterations 10
Network Alpha 32 Repeat 20
Epoch 20 Save Every N Epochs 1

Basic Parameters Passed to the Code

Parameter Value
Prompt Closeup of blonde woman depth of field, bokeh, shallow focus, minimalism, fujifilm xh2s with Canon EF lens, cinematic --ar 85:128 --v 6.0 --style raw
Sampler euler

Download Model

Weights for this model are available in Safetensors format. You can download them from the Files & versions tab.


.

.

.@prithivmlmods - hf

About

Realistic Image Generation, Realistic trigger works properly, better for photorealistic trigger words, close-up shots, face diffusion, male, female characters.

Topics

Resources

Stars

Watchers

Forks