This repository contains both the scaled and non-scaled FP8 weights of Krea 2 Raw.

The non-scaled FP8 weights are fully compatible with lora trainers.

The scaled FP8 weights are for apps like ComfyUI.

The GGUFs are fully compatible with ComfyUI if you are using the updated GGUF loader.

There are no merges or bakes, this is the raw base model.

sample images and workflows are on the way

Downloads last month
77
GGUF
Model size
13B params
Architecture
krea2
Hardware compatibility
Log In to add your hardware

4-bit

6-bit

8-bit

16-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for ChrisColeTech/krea-2-raw-FP8

Base model

krea/Krea-2-Raw
Quantized
(11)
this model