Update README.md
Browse files
README.md
CHANGED
|
@@ -18,11 +18,13 @@ library_name: diffusers
|
|
| 18 |
|
| 19 |
This repository contains Nunchaku-quantized (SVDQ) versions of [CenKreChro](https://civitai.com/models/1836040/cenkrechro), a text-to-image model based on Chroma and Flux Krea merged by [TiwazM](https://civitai.com/user/TiwazM)
|
| 20 |
|
| 21 |
-
|
| 22 |
Model Files
|
| 23 |
- [svdq-int4_r32-CenKreChro.safetensors](https://huggingface.co/spooknik/CenKreChro-SVDQ/blob/main/svdq-int4_r32-CenKreChro.safetensors): SVDQuant INT4 (rank 32) CenKreChro model. For users with non-Blackwell GPUs (pre-50-series).
|
| 24 |
- svdq-int4_r32-qwen-image.safetensors: SVDQuant NVFP4 (rank 32) CenKreChro model. For users with Blackwell GPUs (50-series).
|
| 25 |
|
|
|
|
|
|
|
|
|
|
| 26 |
### Quality Evaluation
|
| 27 |
|
| 28 |
Below is the quality and similarity evaluated with 256 samples from MJHQ-30K dataset. 256 is a very low sample size, but was done to increase the workflow speed.
|
|
|
|
| 18 |
|
| 19 |
This repository contains Nunchaku-quantized (SVDQ) versions of [CenKreChro](https://civitai.com/models/1836040/cenkrechro), a text-to-image model based on Chroma and Flux Krea merged by [TiwazM](https://civitai.com/user/TiwazM)
|
| 20 |
|
|
|
|
| 21 |
Model Files
|
| 22 |
- [svdq-int4_r32-CenKreChro.safetensors](https://huggingface.co/spooknik/CenKreChro-SVDQ/blob/main/svdq-int4_r32-CenKreChro.safetensors): SVDQuant INT4 (rank 32) CenKreChro model. For users with non-Blackwell GPUs (pre-50-series).
|
| 23 |
- svdq-int4_r32-qwen-image.safetensors: SVDQuant NVFP4 (rank 32) CenKreChro model. For users with Blackwell GPUs (50-series).
|
| 24 |
|
| 25 |
+
If you find my models useful please consider: <a href='https://ko-fi.com/B0B21MPRDT' target='_blank'><img height='36' style='border:0px;height:36px;' src='https://storage.ko-fi.com/cdn/kofi6.png?v=6' border='0' alt='Buy Me a Coffee at ko-fi.com' /></a>
|
| 26 |
+
|
| 27 |
+
|
| 28 |
### Quality Evaluation
|
| 29 |
|
| 30 |
Below is the quality and similarity evaluated with 256 samples from MJHQ-30K dataset. 256 is a very low sample size, but was done to increase the workflow speed.
|