Transformers
English
controlnet
Inference Endpoints
Gerold Meisinger commited on
Commit
1406363
1 Parent(s): 6903d2c
Files changed (1) hide show
  1. README.md +7 -7
README.md CHANGED
@@ -10,10 +10,10 @@ tags:
10
 
11
  Controls image generation by edge maps generated with [Edge Drawing](https://github.com/CihanTopal/ED_Lib). Note that Edge Drawing comes in different flavors: original (_ed_), parameter free (_edpf_), color (_edcolor_).
12
 
13
- * Based on my monologs at [github.com - Edge Drawing](https://github.com/lllyasviel/ControlNet/discussions/318)
14
  * For usage see the model page on [civitai.com - Model](https://civitai.com/models/149740).
15
- * To generate edpf maps you can use [this space](https://huggingface.co/spaces/GeroldMeisinger/edpf) or [this script gitlab.com](https://gitlab.com/-/snippets/3601881).
16
- * For evaluation see the corresponding .zip with images at the files.
17
  * To run your own evaluations you can use [this script at gitlab.com](https://gitlab.com/-/snippets/3602096).
18
 
19
  **Edge Drawing Parameter Free**
@@ -113,7 +113,9 @@ edges = ed.detectEdges(image)
113
  edge_map = ed.getEdgeImage(edges)
114
  ```
115
 
116
- 45000 steps => looks good. resuming with left-right flipped images. released as **version 0.1 on civitai**.
 
 
117
 
118
  **Experiment 3.1 - 2023-09-24 - control-edgedrawing-cv480edpf-drop0-fp16-checkpoint-90000**
119
 
@@ -176,9 +178,7 @@ TODO: Why did I end up with less images after I added more images? fastdup sudde
176
 
177
  **Experiment 6.1 - control-edgedrawing-cv480edpf-rect-fp16-batch32-checkpoint-6696**
178
 
179
- see experiment 6.0. restarted from 0 with `--train_batch_size=2 --gradient_accumulation_steps=16`.
180
-
181
- 1 epoch = 107122 * 2 / 32 = 6696 steps per epoch
182
 
183
  **Experiment 6.2 - control-edgedrawing-cv480edpf-rect-fp16-batch32-drop50-checkpoint-6696**
184
 
 
10
 
11
  Controls image generation by edge maps generated with [Edge Drawing](https://github.com/CihanTopal/ED_Lib). Note that Edge Drawing comes in different flavors: original (_ed_), parameter free (_edpf_), color (_edcolor_).
12
 
13
+ * Based on my monologs at [github.com - Edge Drawing](https://github.com/lllyasviel/ControlNet/discussions/318) with detailed report and evaluations.
14
  * For usage see the model page on [civitai.com - Model](https://civitai.com/models/149740).
15
+ * To generate edpf maps you can use [this space](https://huggingface.co/spaces/GeroldMeisinger/edpf) or [this script at gitlab.com](https://gitlab.com/-/snippets/3601881).
16
+ * For evaluation images see the corresponding .zip's at "files".
17
  * To run your own evaluations you can use [this script at gitlab.com](https://gitlab.com/-/snippets/3602096).
18
 
19
  **Edge Drawing Parameter Free**
 
113
  edge_map = ed.getEdgeImage(edges)
114
  ```
115
 
116
+ 45000 steps => looks good. released as **version 0.1 on civitai**.
117
+
118
+ resuming with left-right flipped images.
119
 
120
  **Experiment 3.1 - 2023-09-24 - control-edgedrawing-cv480edpf-drop0-fp16-checkpoint-90000**
121
 
 
178
 
179
  **Experiment 6.1 - control-edgedrawing-cv480edpf-rect-fp16-batch32-checkpoint-6696**
180
 
181
+ see experiment 6.0. restarted from 0 with `--train_batch_size=2 --gradient_accumulation_steps=16`. 1 epoch = 107122 * 2 / 32 = 6696 steps per epoch => released as **version 0.2 on civitai**.
 
 
182
 
183
  **Experiment 6.2 - control-edgedrawing-cv480edpf-rect-fp16-batch32-drop50-checkpoint-6696**
184