Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
### Quality Evaluation
|
| 2 |
|
| 3 |
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.
|
|
@@ -7,3 +22,5 @@ Below is the quality and similarity evaluated with 256 samples from MJHQ-30K dat
|
|
| 7 |
| CenKreChro (25 Steps) | BF16 | -- | 135.52 | 0.842 | -- | -- |
|
| 8 |
| | INT W4A4 | SVDQ | 135.50 | 0.795 | 0.255 | 20.64 |
|
| 9 |
| | NVFP4 | SVDQ | | 0.000 | 0.000 | 00.00 |
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
pipeline_tag: text-to-image
|
| 3 |
+
tags:
|
| 4 |
+
- Flux1
|
| 5 |
+
- Chroma
|
| 6 |
+
- Krea
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
This repository contains Nunchaku-quantized (SVDQ) versions of CenKreChro, a text2image model based on Chroma and Flux Krea merged by [TiwazM](https://civitai.com/user/TiwazM)
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
Model Files
|
| 13 |
+
- [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).
|
| 14 |
+
- svdq-int4_r128-qwen-image.safetensors: SVDQuant NVFP4 (rank 32) CenKreChro model. For users with Blackwell GPUs (50-series).
|
| 15 |
+
|
| 16 |
### Quality Evaluation
|
| 17 |
|
| 18 |
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.
|
|
|
|
| 22 |
| CenKreChro (25 Steps) | BF16 | -- | 135.52 | 0.842 | -- | -- |
|
| 23 |
| | INT W4A4 | SVDQ | 135.50 | 0.795 | 0.255 | 20.64 |
|
| 24 |
| | NVFP4 | SVDQ | | 0.000 | 0.000 | 00.00 |
|
| 25 |
+
|
| 26 |
+
|