Update README.md
Browse files
README.md
CHANGED
|
@@ -21,6 +21,15 @@ Like the tile controlnets for SD, this takes a blurred video as the control sign
|
|
| 21 |
|
| 22 |
Training: Videos only, 9-13 frames at 624px area equivalent. 62k training steps, about 3 days on 1x3090.
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
#### More coming soon
|
| 25 |
|
| 26 |
## Technical details
|
|
|
|
| 21 |
|
| 22 |
Training: Videos only, 9-13 frames at 624px area equivalent. 62k training steps, about 3 days on 1x3090.
|
| 23 |
|
| 24 |
+
#### Depth
|
| 25 |
+
|
| 26 |
+
<video width="80%" controls>
|
| 27 |
+
<source src="https://huggingface.co/spacepxl/Wan2.1-control-loras/resolve/main/assets/depth_example.mp4" type="video/mp4">
|
| 28 |
+
Your browser does not support the video tag.
|
| 29 |
+
</video>
|
| 30 |
+
|
| 31 |
+
Structural control via depth map, trained on DepthAnythingV2-small depth inputs. Prompt following can be somewhat degraded with the v0.1 lora, so it's best used with early cutoff and/or reduced weight.
|
| 32 |
+
|
| 33 |
#### More coming soon
|
| 34 |
|
| 35 |
## Technical details
|