readme update
Browse files
README.md
CHANGED
|
@@ -105,7 +105,7 @@ Go to [ComfyUI-Animemory-Loader](https://github.com/animEEEmpire/ComfyUI-Animemo
|
|
| 105 |
|
| 106 |
3.Diffusers inference.
|
| 107 |
|
| 108 |
-
The pipeline has not been merged yet. Please use the following code to setup the environment.
|
| 109 |
```shell
|
| 110 |
git clone https://github.com/huggingface/diffusers.git
|
| 111 |
cd ..
|
|
@@ -115,7 +115,7 @@ cp diffusers_animemory/* diffusers -r
|
|
| 115 |
cd diffusers
|
| 116 |
pip install .
|
| 117 |
```
|
| 118 |
-
And then, you can use the following code to generate images.
|
| 119 |
|
| 120 |
```python
|
| 121 |
from diffusers import AniMemoryPipeLine
|
|
@@ -137,7 +137,7 @@ images = pipe(prompt=prompt,
|
|
| 137 |
images.save("output.png")
|
| 138 |
```
|
| 139 |
|
| 140 |
-
Use `pipe.enable_sequential_cpu_offload()` to offload the model into CPU for less GPU memory cost (about 14.25 G,
|
| 141 |
compared to 25.67 G if CPU offload is not enabled), but the inference time will increase significantly(5.18s v.s.
|
| 142 |
17.74s on A100 40G).
|
| 143 |
|
|
|
|
| 105 |
|
| 106 |
3.Diffusers inference.
|
| 107 |
|
| 108 |
+
- The pipeline has not been merged yet. Please use the following code to setup the environment.
|
| 109 |
```shell
|
| 110 |
git clone https://github.com/huggingface/diffusers.git
|
| 111 |
cd ..
|
|
|
|
| 115 |
cd diffusers
|
| 116 |
pip install .
|
| 117 |
```
|
| 118 |
+
- And then, you can use the following code to generate images.
|
| 119 |
|
| 120 |
```python
|
| 121 |
from diffusers import AniMemoryPipeLine
|
|
|
|
| 137 |
images.save("output.png")
|
| 138 |
```
|
| 139 |
|
| 140 |
+
- Use `pipe.enable_sequential_cpu_offload()` to offload the model into CPU for less GPU memory cost (about 14.25 G,
|
| 141 |
compared to 25.67 G if CPU offload is not enabled), but the inference time will increase significantly(5.18s v.s.
|
| 142 |
17.74s on A100 40G).
|
| 143 |
|