Downloads · 30 days
689
99% of all-time downloads
Hoops-McCann/hoi4-portraits-flux2-klein-9b-lora
hoi4-portraits-flux2-klein-9b-lora is a machine learning model from Hoops-McCann. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
The project uses the 2500-step HOI4 style LoRA:
Downloads · 30 days
689
99% of all-time downloads
All-time downloads
696
Public
Repo size
6.1 GB
Likes
1
Public
Click a slice to open those files.
.safetensors4.6 GB · 100%
From the Hugging Face model README
The project uses the 2500-step HOI4 style LoRA:
hoi4_portrait_flux2_klein_9b_lora_000002500.safetensors
It targets the distilled FLUX.2 Klein 9B model and works with the project's full, FP8, and GGUF distilled variants. Use the trigger hoi4_portrait style and load the file at strength 1.00 in LoraLoaderModelOnly.
Place the checkpoint under ComfyUI/models/loras/. Comfy Cloud users can import it through Models → Import, select LoRA, and wait for the filename to appear in the loader dropdown.
Style sampling uses CFG 1.0, FLUX guidance 1.0, Euler, simple scheduling, 4 steps, and denoise 1.00. The source workflow starts with fixed seeds 757254001619850, 629907966167866, and 42; batch starts with fixed seed 757254001619850; text-to-image is randomized by default. Set Control after generate to randomize, increment, or decrement to explore different results.
Source prompt:
make this portrait hoi4_portrait style
Text-to-image prompt:
hoi4_portrait style, a soviet soldier with the head of a brown bear wearing a soviet hat, ears still visible. No military uniform decorations.
Source, processing-only, and batch workflows use the Adonis Base → Post restoration path when enabled; the source and batch graphs keep one red restoration switch. Source starts with three candidates, while batch defaults to one candidate per input and can create more. All image-based outputs keep the input filename stem.
See the Hugging Face model card, workflow guide, and output layout reference for the complete model stack, workflow behavior, and saved-file structure.