Commit
•
e9cce28
1
Parent(s):
6c75376
Update d-adaptation/notes.md
Browse files- d-adaptation/notes.md +1 -0
d-adaptation/notes.md
CHANGED
|
@@ -22,6 +22,7 @@ This was further confirmed by running dim 8 alpha 1 with constant learning sched
|
|
| 22 |
## Dim
|
| 23 |
128 dim shows some local noisy patterns. Reranking the model to a lower dim from 128 doesn't get rid of it. Converting the weights of the last up block in the unet does but also causes a noticable change in the generated character. Obviously you could reduce the last up block by a smaller amount.
|
| 24 |
Lower dims show good performance. Need much larger test to check for accuracy between them.
|
|
|
|
| 25 |
|
| 26 |
## Resolution
|
| 27 |
512 resolution is good so far 700 steps.
|
|
|
|
| 22 |
## Dim
|
| 23 |
128 dim shows some local noisy patterns. Reranking the model to a lower dim from 128 doesn't get rid of it. Converting the weights of the last up block in the unet does but also causes a noticable change in the generated character. Obviously you could reduce the last up block by a smaller amount.
|
| 24 |
Lower dims show good performance. Need much larger test to check for accuracy between them.
|
| 25 |
+
Dim 1 worked! Very good accuracy on the details too. See Amber's pattern on her stomach, it is very clean compared to higher dims. The issue seems to be that this doesn't work too well wiith some style loras. More investigation needed.
|
| 26 |
|
| 27 |
## Resolution
|
| 28 |
512 resolution is good so far 700 steps.
|