Krea 2 Turbo BF16 converted to W4A4 using Vince Merge Lab for Forge. Tested on Forge Neo, low VRAM friendly, and ideal for LoRA testing. Clean base model with no merged LoRAs, preserving the original Krea 2 Turbo behavior as closely as possible.
# Krea2 Turbo BF16 to W4A4
Lightweight Krea 2 Turbo conversion for Forge / Forge Neo and lower-VRAM setups.
Krea2 Turbo BF16 to W4A4 is a community-made quantized conversion of the original Krea 2 Turbo BF16 model.
This version was converted by Vince AI using the Vince Merge Lab extension for Forge, with the goal of reducing memory usage and making Krea 2 Turbo more accessible on smaller hardware configurations while preserving the original model’s visual behavior as much as possible.
## Key Features
- W4A4 quantized conversion
- Created with Vince Merge Lab for Forge
- Tested and working in Forge / Forge Neo
- More suitable for lower-VRAM and modest hardware setups
- Based on the original Krea 2 Turbo BF16 model
## Model Information
- Model name: krea2_turbo_bf16_to_W4A4
- Original model: Krea 2 Turbo BF16
- Conversion type: W4A4
- Format: SafeTensors (.safetensors)
- Converted by: Vince AI
- Tool used: Vince Merge Lab extension for Forge
- Version: v1.0
## Compatibility
This model has been successfully tested in Forge / Forge Neo.
It is intended as a more memory-efficient alternative to the original BF16 version and may be better suited for users with smaller GPU configurations.
Actual VRAM usage, generation speed, and overall performance may vary depending on hardware, software version, resolution, and generation settings.
## Credits and License
Original model developed by Krea.ai.
W4A4 conversion by Vince AI using Vince Merge Lab for Forge.
This is an unofficial community conversion and is not affiliated with, endorsed by, or sponsored by Krea.ai.
Original model:
https://huggingface.co/krea/Krea-2-Turbo
Official licensing information:
https://www.krea.ai/krea-2-licensing
Distributed under the applicable Krea 2 Community License terms.
