modèles/Chroma - v.41-low-step-rl

Chroma - v.41-low-step-rl

8/16/2025

3:30:46 AM

Mots-clés et tags associés

5m dataset,8.9b parameter model,modèle de base,base_model,checkpoint,chroma,flux.1-schnell,lodestone,open-source ai,other,uncensored model,v.41-low-step-rl

Abstract painting of a young woman with long white hair and blue eyes wearing a futuristic black corseted dress made of bones and metal, set against a muted background.

Abstract urban landscape featuring a solitary female silhouette in black, surrounded by vibrant splashes of color and street lamps in a rain-soaked city street.

A fair-skinned fantasy woman with flowing blonde hair and turquoise dress adorned with gold accents reclines among overgrown ruins and yellow daffodils under ethereal light.

Digital illustration of a tall, intricate futuristic building made from metal and glass with cables and wires, standing solitary in a barren desert under a deep blue sky with scattered clouds.

An abstract painting featuring a poised woman with her graceful back, created with broad, bold brushstrokes on a white canvas, set against a soft background of blues, greys, and pinks.

A majestic castle with grand classical columns and arches nestled between towering mountains under a starry sky, illuminated by warm interior light and cool shadows.

Watercolor portrait of a virtual idol with pastel pink bob hair, wearing a semi-transparent orange dress, set against a textured abstract background with gold leaf details.

Oil painting of a medieval castle on a verdant hill with dramatic clouds, a winding path with figures on horseback, and a distant village.

$A hyper-dramatic close-up of a faceless cyborg, with a cracked translucent orb face dissolving into abstract Klimt-inspired green and black geometric swirls. The body is clad in sharp matte black and white ceremonial armor with glowing neon orange veins. The background shows fractured monochrome clouds and red fractals.$

A pale-skinned girl with white braided hair, cyberpunk black horns, red eyes, black lipstick, wearing a kimono, kneeling against a deep red abstract background with dark brushstrokes.

Close-up abstract portrait of a young woman with textured brushstrokes and fragmented geometric overlays dissipating into a muted, surreal background.

Hyper-realistic digital painting of a cyborg girl with pink hair, mechanical joints, and cyberpunk fashion, set against an abstract painted background.

Prompts recommandés

Extreme close-up photograph of a single tiger eye, direct frontal view. The iris is very detailed and the pupil resembling a dark void. The word "CHROMA LOW STEPS RL TRAINING" is across the lower portion of the image in large white stylized letters, with brush strokes. Each strand of the thick fur is highly detailed and distinguishable. Natural lighting to capture authentic eye shine and depth.

Prompts négatifs recommandés

low quality, ugly, unfinished, out of focus

Paramètres recommandés

samplers

DPM++ 2M

steps

cfg

1.3

Conseils

Model training is ongoing, and user feedback is encouraged.

Use the latest model from the Hugging Face repo https://huggingface.co/lodestones/Chroma

Alternative quantization options are available including FP8 Scaled Quant and GGUF Quantized formats.

You can follow live training progress via https://training.lodestone-rock.com/

ComfyUI inference and workflow resources are provided to facilitate integration.

Sponsors du créateur

Shoutout to Fictional.ai for the awesome support — seriously appreciate you helping push open-source AI forward.

Hey everyone!

Chroma is a 8.9B parameter model based on FLUX.1-schnell (technical report coming soon!). It’s fully Apache 2.0 licensed, ensuring that anyone can use, modify, and build on top of it—no corporate gatekeeping.

The model is still training right now, and I’d love to hear your thoughts! Your input and feedback are really appreciated.

What Chroma Aims to Do

Training on a 5M dataset, curated from 20M samples including anime, furry, artistic stuff, and photos.
Fully uncensored, reintroducing missing anatomical concepts.
Built as a reliable open-source option for those who need it.

See the Progress

Hugging Face Repo: https://huggingface.co/lodestones/Chroma (go to this repo for the latest model!)
Hugging Face Debug Repo: https://huggingface.co/lodestones/chroma-debug-development-only
Live AIM Training Logs: https://training.lodestone-rock.com/
ComfyUI Inference node [WIP]: https://github.com/lodestone-rock/flux-mod
ComfyUI workflow: https://huggingface.co/lodestones/Chroma/resolve/main/simple_workflow.json
Training code!: https://github.com/lodestone-rock/flow

Quantization options

Alternative option: FP8 Scaled Quant (Format used by ComfyUI with possible inference speed increase)
Alternative option: GGUF Quantized (You will need to install ComfyUI-GGUF custom node)