kenshin commited on
Commit ·
e4e2703
1
Parent(s): f1c37d4
docs: add turbo sampling note
Browse files
README.md
CHANGED
|
@@ -142,6 +142,9 @@ image.save("llada-image-base.png")
|
|
| 142 |
|
| 143 |
Use the Turbo checkpoint for fast generation and editing. Its recommended sampling configuration is **4 steps**.
|
| 144 |
|
|
|
|
|
|
|
|
|
|
| 145 |
```python
|
| 146 |
import torch
|
| 147 |
|
|
|
|
| 142 |
|
| 143 |
Use the Turbo checkpoint for fast generation and editing. Its recommended sampling configuration is **4 steps**.
|
| 144 |
|
| 145 |
+
> [!NOTE]
|
| 146 |
+
> For LLaDA-Image-Turbo inference, you can try setting `stochastic_sampling` to `false` in `scheduler/scheduler_config.json`, which may produce sharper details in some cases.
|
| 147 |
+
|
| 148 |
```python
|
| 149 |
import torch
|
| 150 |
|