Update README.md
Browse files
README.md
CHANGED
|
@@ -5,7 +5,23 @@ language:
|
|
| 5 |
- zh
|
| 6 |
library_name: diffusers
|
| 7 |
pipeline_tag: text-to-image
|
|
|
|
|
|
|
| 8 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
<p align="center">
|
| 10 |
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/qwen_image_logo.png" width="400"/>
|
| 11 |
<p>
|
|
|
|
| 5 |
- zh
|
| 6 |
library_name: diffusers
|
| 7 |
pipeline_tag: text-to-image
|
| 8 |
+
base_model:
|
| 9 |
+
- Qwen/Qwen-Image
|
| 10 |
---
|
| 11 |
+
|
| 12 |
+
This is an NF4 quantized model of Qwen-image so it can run on GPUs using 20GB VRAM. You can run it on lower VRAM like 16GB.
|
| 13 |
+
There were other NF4 models but they made the mistake of blindly quantizing all layers in the transformer. This one does not.
|
| 14 |
+
We retain some layers at full precision in order to ensure that we get quality output.
|
| 15 |
+
|
| 16 |
+
You can use the original Qwen-Image parameters as is though I recommend atleast 20 inference steps.
|
| 17 |
+
|
| 18 |
+
This model is available for inference at [JustLab.ai](https://justlab.ai)
|
| 19 |
+
|
| 20 |
+
The original Qwen-Image attributions are included verabtim below.
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
|
| 25 |
<p align="center">
|
| 26 |
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/qwen_image_logo.png" width="400"/>
|
| 27 |
<p>
|