Update README.md
Browse files
README.md
CHANGED
@@ -16,18 +16,7 @@ base_model: black-forest-labs/FLUX.1-dev
|
|
16 |
until the next Diffusers pypi release, please install Diffusers from source and use [this PR](xxxxxx) to be able to use.
|
17 |
TODO: change when new version.
|
18 |
|
19 |
-
|
20 |
-
## Checkpoint
|
21 |
-
|
22 |
-
The training of union controlnet requires a significant amount of computational power.
|
23 |
-
The current release is only an alpha version checkpoint that has not been fully trained.
|
24 |
-
The beta version is in the training process.
|
25 |
-
We have conducted ablation studies that have demonstrated the validity of the code.
|
26 |
-
The open-source release of the alpha version is solely to facilitate the rapid growth of the open-source community and the Flux ecosystem;
|
27 |
-
it is common to encounter bad cases (please accept my apologies).
|
28 |
-
It is worth noting that we have found that even a fully trained Union model may not perform as well as specialized models, such as pose control.
|
29 |
-
However, as training progresses, the performance of the Union model will continue to approach that of specialized models.
|
30 |
-
|
31 |
| Control Mode | Description |
|
32 |
|:------------:|:-----------:|
|
33 |
|0|canny|
|
@@ -38,16 +27,33 @@ However, as training progresses, the performance of the Union model will continu
|
|
38 |
|5|gray|
|
39 |
|6|lq|
|
40 |
|
41 |
-
|
42 |
<img src="./images/image_demo_canny.jpg" width = "800" />
|
|
|
43 |
<img src="./images/image_demo_tile.jpg" width = "800" />
|
|
|
44 |
<img src="./images/image_demo_depth.jpg" width = "800" />
|
|
|
45 |
<img src="./images/image_demo_blur.jpg" width = "800" />
|
|
|
46 |
<img src="./images/image_demo_pose.jpg" width = "800" />
|
|
|
47 |
<img src="./images/image_demo_gray.jpg" width = "800" />
|
|
|
48 |
<img src="./images/image_demo_lq.jpg" width = "800" />
|
49 |
|
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
|
52 |
|
53 |
# Demo
|
|
|
16 |
until the next Diffusers pypi release, please install Diffusers from source and use [this PR](xxxxxx) to be able to use.
|
17 |
TODO: change when new version.
|
18 |
|
19 |
+
## Control Mode
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
| Control Mode | Description |
|
21 |
|:------------:|:-----------:|
|
22 |
|0|canny|
|
|
|
27 |
|5|gray|
|
28 |
|6|lq|
|
29 |
|
30 |
+
### canny
|
31 |
<img src="./images/image_demo_canny.jpg" width = "800" />
|
32 |
+
### tile
|
33 |
<img src="./images/image_demo_tile.jpg" width = "800" />
|
34 |
+
### depth
|
35 |
<img src="./images/image_demo_depth.jpg" width = "800" />
|
36 |
+
### blur
|
37 |
<img src="./images/image_demo_blur.jpg" width = "800" />
|
38 |
+
### pose
|
39 |
<img src="./images/image_demo_pose.jpg" width = "800" />
|
40 |
+
### gray
|
41 |
<img src="./images/image_demo_gray.jpg" width = "800" />
|
42 |
+
### low quality
|
43 |
<img src="./images/image_demo_lq.jpg" width = "800" />
|
44 |
|
45 |
|
46 |
+
## Checkpoint
|
47 |
+
|
48 |
+
The training of union controlnet requires a significant amount of computational power.
|
49 |
+
The current release is only an alpha version checkpoint that has not been fully trained.
|
50 |
+
The beta version is in the training process.
|
51 |
+
We have conducted ablation studies that have demonstrated the validity of the code.
|
52 |
+
The open-source release of the alpha version is solely to facilitate the rapid growth of the open-source community and the Flux ecosystem;
|
53 |
+
it is common to encounter bad cases (please accept my apologies).
|
54 |
+
It is worth noting that we have found that even a fully trained Union model may not perform as well as specialized models, such as pose control.
|
55 |
+
However, as training progresses, the performance of the Union model will continue to approach that of specialized models.
|
56 |
+
|
57 |
|
58 |
|
59 |
# Demo
|