Canopus-LoRA-Flux-UltraRealism-2.0

Prompt
woman in a red jacket, snowy, in the style of hyper-realistic portraiture, caninecore, mountainous vistas, timeless beauty, palewave, iconic, distinctive noses --ar 72:101 --stylize 750 --v 6

The model is still in the training phase. This is not the final version and may contain artifacts and perform poorly in some cases.

Model description

prithivMLmods/Canopus-LoRA-Flux-FaceRealism

Image Processing 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 & Steps	30 & 3.8K+
Epoch	20	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 70 [ Hi-RES ] & More ...............

Trigger words

You should use Ultra realistic to trigger the image generation.

Other Versions

Here’s a table format for the Hugging Face model "prithivMLmods/Canopus-LoRA-Flux-FaceRealism":

Attribute	Details
Model Name	Canopus-LoRA-Flux-FaceRealism
Model ID	`prithivMLmods/Canopus-LoRA-Flux-FaceRealism`
Hugging Face URL	Canopus-LoRA-Flux-FaceRealism
Model Type	LoRA (Low-Rank Adaptation)
Primary Use Case	Face Realism image generation
Supported Framework	Hugging Face Diffusers
Data Type	`bfloat16`, `fp16`, `float32`
Compatible Models	Stable Diffusion, Flux models
Model Author	`prithivMLmods`
LoRA Technique	LoRA for image style transfer with a focus on generating realistic faces
Model Version	Latest
License	Open-Access
Tags	LoRA, Face Realism, Flux, Image Generation

Setting Up

import torch
from pipelines import DiffusionPipeline

base_model = "prithivMLmods/Canopus-LoRA-Flux-UltraRealism-2.0"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "prithivMLmods/Canopus-LoRA-Flux-FaceRealism"
trigger_word = "Ultra realistic"  # Leave trigger_word blank if not used.
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.