Update README.md
Browse files
README.md
CHANGED
|
@@ -31,11 +31,12 @@ LiteVision-v1 is a lightweight, fast Stable Diffusion 1.5-based model optimized
|
|
| 31 |
|
| 32 |
For best results:
|
| 33 |
|
| 34 |
-
num_inference_steps = 6
|
| 35 |
-
guidance_scale = 1.5
|
| 36 |
-
scheduler = LCMScheduler
|
| 37 |
|
| 38 |
Lower guidance gives cleaner results.
|
|
|
|
| 39 |
Higher guidance may introduce instability due to LCM behavior.
|
| 40 |
|
| 41 |
# 🧠 Why LiteVision?
|
|
@@ -46,7 +47,6 @@ Traditional SD 1.5 models require 20–30 steps for good quality. LiteVision ONL
|
|
| 46 |
# 🛠 Usage
|
| 47 |
python
|
| 48 |
```
|
| 49 |
-
Copy code
|
| 50 |
import torch
|
| 51 |
from diffusers import DiffusionPipeline
|
| 52 |
|
|
|
|
| 31 |
|
| 32 |
For best results:
|
| 33 |
|
| 34 |
+
num_inference_steps = 6 <br>
|
| 35 |
+
guidance_scale = 1.5 <br>
|
| 36 |
+
scheduler = LCMScheduler <br>
|
| 37 |
|
| 38 |
Lower guidance gives cleaner results.
|
| 39 |
+
|
| 40 |
Higher guidance may introduce instability due to LCM behavior.
|
| 41 |
|
| 42 |
# 🧠 Why LiteVision?
|
|
|
|
| 47 |
# 🛠 Usage
|
| 48 |
python
|
| 49 |
```
|
|
|
|
| 50 |
import torch
|
| 51 |
from diffusers import DiffusionPipeline
|
| 52 |
|