Update README.md
Browse files
README.md
CHANGED
@@ -100,11 +100,11 @@ Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 5, Seed: 3442031040, Size: 512x
|
|
100 |
|
101 |
### Recipe
|
102 |
|
103 |
-
**Step 1
|
104 |
|
105 |
You can try replacing animefull-latest with Anything-V3.0 or your preferred model. However, I cannot confirm if this will yield better results and it requires you to experiment with it on your own.
|
106 |
|
107 |
-
**Step 2
|
108 |
|
109 |
| Model A | Model B | base_alpha | Weight | Merge Name |
|
110 |
| ----------- | ---------- | ---------- | ------------------------------------------------- | --------------- |
|
@@ -114,7 +114,7 @@ This is the step of the sfw version. The steps for the nsfw version are as follo
|
|
114 |
|
115 |
LoRAs related to real people should be merged into Chilloutmix or other photo-realistic models that you like, while LoRAs related to anime should be merged into X-mix-V2.0-base. Which LoRAs to use depends on your preference.
|
116 |
|
117 |
-
**Step 3
|
118 |
|
119 |
| Model A | Model B | base_alpha | Weight | Merge Name |
|
120 |
| -------------- | --------------- | ---------- | ------------------------------------------------------- | -------------- |
|
@@ -207,13 +207,13 @@ Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 5, Seed: 2548407675, Size: 512x
|
|
207 |
|
208 |
If you need the embedding used in examples, click them:
|
209 |
|
210 |
-
- **EasyNegative
|
211 |
|
212 |
-
- **bad-artist
|
213 |
|
214 |
-
- **bad_prompt_version2
|
215 |
|
216 |
-
- **Deep Negative V1.x
|
217 |
|
218 |
You can consider whether to use them according to your preferences.
|
219 |
|
|
|
100 |
|
101 |
### Recipe
|
102 |
|
103 |
+
**Step 1:** animefull-latest (model) + pastelmix-lora (lora) + ligneClaireStyleCogecha (lora) = pastel-Cpgecha
|
104 |
|
105 |
You can try replacing animefull-latest with Anything-V3.0 or your preferred model. However, I cannot confirm if this will yield better results and it requires you to experiment with it on your own.
|
106 |
|
107 |
+
**Step 2:** MBW: Chilloutmix + X-mix-V1.0
|
108 |
|
109 |
| Model A | Model B | base_alpha | Weight | Merge Name |
|
110 |
| ----------- | ---------- | ---------- | ------------------------------------------------- | --------------- |
|
|
|
114 |
|
115 |
LoRAs related to real people should be merged into Chilloutmix or other photo-realistic models that you like, while LoRAs related to anime should be merged into X-mix-V2.0-base. Which LoRAs to use depends on your preference.
|
116 |
|
117 |
+
**Step 3:** MBW: pastel-Cogecha + X-mix-V2.0-base
|
118 |
|
119 |
| Model A | Model B | base_alpha | Weight | Merge Name |
|
120 |
| -------------- | --------------- | ---------- | ------------------------------------------------------- | -------------- |
|
|
|
207 |
|
208 |
If you need the embedding used in examples, click them:
|
209 |
|
210 |
+
- **EasyNegative:** [embed/EasyNegative · Hugging Face](https://huggingface.co/embed/EasyNegative)
|
211 |
|
212 |
+
- **bad-artist:** [nick-x-hacker/bad-artist · Hugging Face](https://huggingface.co/nick-x-hacker/bad-artist)
|
213 |
|
214 |
+
- **bad_prompt_version2:** [embed/bad_prompt · Hugging Face](https://huggingface.co/embed/bad_prompt)
|
215 |
|
216 |
+
- **Deep Negative V1.x:** [Deep Negative V1.x | Stable Diffusion TextualInversion | Civitai](https://civitai.com/models/4629/deep-negative-v1x)
|
217 |
|
218 |
You can consider whether to use them according to your preferences.
|
219 |
|