Readme for Frankenweights with examples and links
Browse files
README.md
CHANGED
@@ -8,23 +8,38 @@ FRANKENWEGIHTS
|
|
8 |
dunh dunh dunh...
|
9 |
|
10 |
You wouldn't think something like this might work, but it does. I took the text encoder from my "Storytime" model
|
11 |
-
and then blindly pasted it over the text encoder stuff in the SD 1.5 model and now we have FrankenWeights.
|
12 |
-
|
13 |
It's ALIVE!!!:
|
14 |
|
|
|
|
|
|
|
15 |
|
|
|
|
|
|
|
16 |
|
17 |
-
|
18 |
-
the extra awesomesauce done by ByteDance with this Lora, which allows you to greatly reduce the steps required
|
19 |
-
to produce an image --
|
20 |
|
21 |
-
Frakenweights in this context responds best to "normal" CFG ranges, especially with the Lora linked above. I generally
|
22 |
like to make things that will take a huge range of CFG values, but in this case everything hums along together so well
|
23 |
for my purposes that I'm not gonna mess with a good thing. We hack stuff and just try things in these parts, and it
|
24 |
just so happens that the combo of FrankenWeights, that Lora, and the "Superposition" bits make a truly Frank-en Stein
|
25 |
with enough ummph to bring all your wild ideas to life:
|
26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
|
|
|
28 |
|
29 |
So, get your grubby little hands (I'm sure space aliens consider us to be "grubby" and "little") on FrankenWeights today.
|
30 |
I have intentionally left this as a 32 bit model, because more often than not the "burn" in an image can be stretched
|
|
|
8 |
dunh dunh dunh...
|
9 |
|
10 |
You wouldn't think something like this might work, but it does. I took the text encoder from my "Storytime" model
|
11 |
+
and then blindly pasted it over the text encoder stuff in the SD 1.5 model and now we have ![FrankenWeights](/FrankenWeights.safetensors)
|
|
|
12 |
It's ALIVE!!!:
|
13 |
|
14 |
+
![](Examples/FrankenPosition.PNG)
|
15 |
+
|
16 |
+
|
17 |
|
18 |
+
Also included here is a somewhat "refined" version of the "Superposition" ComfyUI workflow called ![Frankenposition](Frankenposition.json) This includes
|
19 |
+
the extra awesomesauce done by ByteDance with their work on ![Hyper-SD](https://huggingface.co/ByteDance/Hyper-SD/blob/main/README.md) and their
|
20 |
+
![Lora](https://huggingface.co/ByteDance/Hyper-SD/blob/main/Hyper-SD15-8steps-CFG-lora.safetensors) for SD1.5 models, which allows you to greatly reduce the steps required to produce an image --
|
21 |
|
22 |
+
![](Examples/StarryWarry2.PNG)
|
|
|
|
|
23 |
|
24 |
+
Frakenweights in this context responds best to "normal" CFG ranges (5-8), especially with the Lora linked above. I generally
|
25 |
like to make things that will take a huge range of CFG values, but in this case everything hums along together so well
|
26 |
for my purposes that I'm not gonna mess with a good thing. We hack stuff and just try things in these parts, and it
|
27 |
just so happens that the combo of FrankenWeights, that Lora, and the "Superposition" bits make a truly Frank-en Stein
|
28 |
with enough ummph to bring all your wild ideas to life:
|
29 |
|
30 |
+
![](Examples/Starrywarry_00263_.png)
|
31 |
+
|
32 |
+
![](Examples/Starrywarry_00020_.png)
|
33 |
+
|
34 |
+
![](Examples/Starrywarry_00062_.png)
|
35 |
+
|
36 |
+
![](Examples/Starrywarry_00166_.png)
|
37 |
+
|
38 |
+
![](Examples/Starrywarry_00226_.png)
|
39 |
+
|
40 |
+
|
41 |
|
42 |
+
(all of the above come from the same prompts and settings)
|
43 |
|
44 |
So, get your grubby little hands (I'm sure space aliens consider us to be "grubby" and "little") on FrankenWeights today.
|
45 |
I have intentionally left this as a 32 bit model, because more often than not the "burn" in an image can be stretched
|