Update noob_v_pencil-XL-v2.0.1.md
Browse files- noob_v_pencil-XL-v2.0.1.md +49 -0
noob_v_pencil-XL-v2.0.1.md
CHANGED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Tools
|
| 2 |
+
- [ComfyUI](https://github.com/comfyanonymous/ComfyUI)
|
| 3 |
+
- [sd-mecha](https://github.com/ljleb/sd-mecha)
|
| 4 |
+
- [sd-mecha for Comfyui](https://github.com/ljleb/comfy-mecha)
|
| 5 |
+
|
| 6 |
+
# Checkpoints
|
| 7 |
+
- [noob_v_pencil-XL-v2.0.0](https://huggingface.co/bluepen5805/noob_v_pencil-XL)
|
| 8 |
+
- [noob_v_pencil-XL-v1.2.0](https://huggingface.co/bluepen5805/noob_v_pencil-XL)
|
| 9 |
+
- [NoobAI-XL (NAI-XL) V-Pred-1.0-Version](https://civitai.com/models/833294?modelVersionId=1190596)
|
| 10 |
+
- [NoobAI-XL (NAI-XL) Epsilon-pred 1.0-version](https://civitai.com/models/833294?modelVersionId=1022833)
|
| 11 |
+
- [Illustrious XL v1.1](https://huggingface.co/OnomaAIResearch/Illustrious-XL-v1.1)
|
| 12 |
+
- [Illustrious XL v0.1](https://huggingface.co/OnomaAIResearch/Illustrious-xl-early-release-v0)
|
| 13 |
+
|
| 14 |
+
# Recipe
|
| 15 |
+
|
| 16 |
+
1. Merge using `sd-mecha for ComfyUI`. ( below sd-mecha format recipe )
|
| 17 |
+
```
|
| 18 |
+
model "noob_v_pencil-XL-v2.0.0.safetensors" "sdxl" "base"
|
| 19 |
+
model "noob_v_pencil-XL-v1.2.0.safetensors" "sdxl" "base"
|
| 20 |
+
model "noobaiXLNAIXL_vPred10Version.safetensors" "sdxl" "base"
|
| 21 |
+
merge "subtract" &1 &2
|
| 22 |
+
merge "add_difference" &0 &3 alpha=1.5
|
| 23 |
+
model "OnomaAI\Illustrious-XL-v1.1.safetensors" "sdxl" "base"
|
| 24 |
+
model "noobaiXLNAIXL_epsilonPred10Version.safetensors" "sdxl" "base"
|
| 25 |
+
model "Illustrious-XL-v0.1.safetensors" "sdxl" "base"
|
| 26 |
+
merge "train_difference" &5 &6 &7 alpha=0.5
|
| 27 |
+
merge "train_difference" &8 &2 &6 alpha=1.5
|
| 28 |
+
merge "subtract" &9 &2
|
| 29 |
+
merge "add_difference" &4 &10 alpha=0.2
|
| 30 |
+
merge "add_cosine_a" &11 &0 alpha=0.25
|
| 31 |
+
dict sdxl_unet_default=1.0 sdxl_txt2_default=0.0 sdxl_txt_default=0.0
|
| 32 |
+
merge "weighted_sum" &0 &12 alpha=&13
|
| 33 |
+
```
|
| 34 |
+
|
| 35 |
+
2. Add `v_pred` and `ztsnr` keys.
|
| 36 |
+
|
| 37 |
+
```python
|
| 38 |
+
BASE_MODEL_NAME = "noob_v_pencil-XL-v2.0.1"
|
| 39 |
+
|
| 40 |
+
tensors = {}
|
| 41 |
+
with safe_open(BASE_MODEL_NAME + "-base.safetensors", framework="pt", device="cpu") as f:
|
| 42 |
+
for key in f.keys():
|
| 43 |
+
tensors[key] = f.get_tensor(key)
|
| 44 |
+
|
| 45 |
+
tensors["v_pred"] = torch.tensor([0.0])
|
| 46 |
+
tensors["ztsnr"] = torch.tensor([0.0])
|
| 47 |
+
|
| 48 |
+
save_file(tensors, BASE_MODEL_NAME + ".safetensors")
|
| 49 |
+
```
|