Commit
•
1acc513
1
Parent(s):
447b65f
Update README.md
Browse files
README.md
CHANGED
|
@@ -14,7 +14,7 @@ We can load this anime model < 15GB VRAM if enable_model_cpu_offload is True.
|
|
| 14 |
otherwise, we can load this anime model < 20GB VRAM.
|
| 15 |
We can run this model on RTX 4090 or NVIDIA L4.
|
| 16 |
|
| 17 |
-
|
| 18 |
- diffusers
|
| 19 |
1. Install quanto-optinum.
|
| 20 |
```bash
|
|
@@ -73,7 +73,7 @@ image = pipe(
|
|
| 73 |
image.save("modern-anime-fp8.png")
|
| 74 |
```
|
| 75 |
|
| 76 |
-
|
| 77 |
1. Install quanto-optinum.
|
| 78 |
```bash
|
| 79 |
pip install optimum-quanto
|
|
|
|
| 14 |
otherwise, we can load this anime model < 20GB VRAM.
|
| 15 |
We can run this model on RTX 4090 or NVIDIA L4.
|
| 16 |
|
| 17 |
+
## Usage
|
| 18 |
- diffusers
|
| 19 |
1. Install quanto-optinum.
|
| 20 |
```bash
|
|
|
|
| 73 |
image.save("modern-anime-fp8.png")
|
| 74 |
```
|
| 75 |
|
| 76 |
+
## How to cast fp8
|
| 77 |
1. Install quanto-optinum.
|
| 78 |
```bash
|
| 79 |
pip install optimum-quanto
|