Update README.md
Browse files
README.md
CHANGED
|
@@ -17,6 +17,14 @@ Original model repository: https://huggingface.co/MiniMaxAI/MiniMax-H3
|
|
| 17 |
|
| 18 |
The Qwen3-VL-32B nvfp4_awq quant is converted from: https://huggingface.co/cybermotaz/Qwen3-VL-32B-Instruct-NVFP4
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
Place the files in the following folders:
|
| 21 |
|
| 22 |
```
|
|
|
|
| 17 |
|
| 18 |
The Qwen3-VL-32B nvfp4_awq quant is converted from: https://huggingface.co/cybermotaz/Qwen3-VL-32B-Instruct-NVFP4
|
| 19 |
|
| 20 |
+
- This nvfp4 text encoder does not require Blackwell GPU to use.
|
| 21 |
+
|
| 22 |
+
- For diffusion_model prefer int8_convrot if you are able to use pytorch with cu130.
|
| 23 |
+
|
| 24 |
+
- Fp8_scaled should only be used if you for any reason can not use the int8_convrot.
|
| 25 |
+
|
| 26 |
+
---
|
| 27 |
+
|
| 28 |
Place the files in the following folders:
|
| 29 |
|
| 30 |
```
|