Rename Z-Image-Base/Z-Image-Rank128-Kintsugi/INFO to Z-Image-Base/Z-Image-Rank128-Kintsugi/INFO.md
Browse files
Z-Image-Base/Z-Image-Rank128-Kintsugi/INFO
DELETED
|
@@ -1,45 +0,0 @@
|
|
| 1 |
-
Kintsugi became closely associated with ceramic vessels used for chanoyu (Japanese tea ceremony).
|
| 2 |
-
One theory is that kintsugi may have originated when Japanese shลgun Ashikaga Yoshimasa sent a damaged
|
| 3 |
-
Chinese tea bowl back to China for repairs in the late 15th century.
|
| 4 |
-
When it was returned, repaired with simple metal staples,
|
| 5 |
-
it may have prompted Japanese craftsmen to look for a more aesthetically pleasing means of repair.
|
| 6 |
-
On the other hand, according to Bakลhan Saลki (record of tea-bowl with a 'large-locust' clamp)
|
| 7 |
-
such "ugliness" was considered inspirational and Zen-like, as it connoted beauty in broken things.
|
| 8 |
-
The bowl thus became highly valued due to the large metal staples, which looked like a locust,
|
| 9 |
-
and the bowl was named 'bakลhan ("large-locust clamp").
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
๐ฆ Full Resumable Training Package (Weights + Optimizer State + Config)
|
| 13 |
-
|
| 14 |
-
This folder contains the full training state (LoRA weights, optimizer state, and .yaml configuration)
|
| 15 |
-
|
| 16 |
-
allowing you to resume or extend training directly from step 2000.
|
| 17 |
-
|
| 18 |
-
Place the folder with the optimizer and checkpoint in the output folder "D:\Your path.\ai-toolkit\output\r128zkitsu"
|
| 19 |
-
|
| 20 |
-
Run the file r128zkitsu.yaml from any directory on your PC.
|
| 21 |
-
|
| 22 |
-
############################################################################################################
|
| 23 |
-
|
| 24 |
-
The training tool is available in the repository on my GitHub:
|
| 25 |
-
|
| 26 |
-
This tool completely eliminates intermediate image generation steps,
|
| 27 |
-
|
| 28 |
-
and the lightning-fast training speed will pleasantly surprise you.
|
| 29 |
-
|
| 30 |
-
https://github.com/OrakulStudio/AI-Toolkit-Windows11
|
| 31 |
-
|
| 32 |
-
MODEL INSPECTION: r128zkitsu.safetensors
|
| 33 |
-
โญโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโฎ
|
| 34 |
-
โ Parameter โ Value โ
|
| 35 |
-
โโโโโโโโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโค
|
| 36 |
-
โ File โ r128zkitsu.safetensors โ
|
| 37 |
-
โ Disk size โ 0.63 GB โ
|
| 38 |
-
โ Total layers / tensors โ 720 โ
|
| 39 |
-
โ Exact parameter count โ 340,132,080 โ
|
| 40 |
-
โ Parameters (billions) โ 0.340132 B โ
|
| 41 |
-
โ Base model โ zimage โ
|
| 42 |
-
โ Trigger words โ r128zkitsu โ
|
| 43 |
-
โ Scan time โ 0.9539 sec โ
|
| 44 |
-
โฐโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโฏ
|
| 45 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Z-Image-Base/Z-Image-Rank128-Kintsugi/INFO.md
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
>[!IMPORTANT]
|
| 2 |
+
>Kintsugi became closely associated with ceramic vessels used for chanoyu (Japanese tea ceremony).
|
| 3 |
+
>One theory is that kintsugi may have originated when Japanese shลgun Ashikaga Yoshimasa sent a damaged
|
| 4 |
+
>Chinese tea bowl back to China for repairs in the late 15th century.
|
| 5 |
+
>When it was returned, repaired with simple metal staples,
|
| 6 |
+
>it may have prompted Japanese craftsmen to look for a more aesthetically pleasing means of repair.
|
| 7 |
+
>On the other hand, according to Bakลhan Saลki (record of tea-bowl with a 'large-locust' clamp)
|
| 8 |
+
>such "ugliness" was considered inspirational and Zen-like, as it connoted beauty in broken things.
|
| 9 |
+
>The bowl thus became highly valued due to the large metal staples, which looked like a locust,
|
| 10 |
+
>and the bowl was named 'bakลhan ("large-locust clamp").
|
| 11 |
+
|
| 12 |
+
>[!IMPORTANT]
|
| 13 |
+
>๐ฆ Full Resumable Training Package (Weights + Optimizer State + Config)
|
| 14 |
+
>
|
| 15 |
+
>This folder contains the full training state (LoRA weights, optimizer state, and .yaml configuration)
|
| 16 |
+
>
|
| 17 |
+
>allowing you to resume or extend training directly from step 2000.
|
| 18 |
+
>
|
| 19 |
+
>Place the folder with the optimizer and checkpoint in the output folder "D:\Your path.\ai-toolkit\output\r128zkitsu"
|
| 20 |
+
>
|
| 21 |
+
>Run the file r128zkitsu.yaml from any directory on your PC.
|
| 22 |
+
>
|
| 23 |
+
>############################################################################################################
|
| 24 |
+
>
|
| 25 |
+
>The training tool is available in the repository on my GitHub:
|
| 26 |
+
>
|
| 27 |
+
>This tool completely eliminates intermediate image generation steps,
|
| 28 |
+
>
|
| 29 |
+
>and the lightning-fast training speed will pleasantly surprise you.
|
| 30 |
+
|
| 31 |
+
## [AI-Toolkit-Windows11](https://github.com/OrakulStudio/AI-Toolkit-Windows11)
|
| 32 |
+
|
| 33 |
+
>[!IMPORTANT]
|
| 34 |
+
> Parameter Value
|
| 35 |
+
> File r128zkitsu.safetensors
|
| 36 |
+
> Disk size 0.63 GB
|
| 37 |
+
> Total layers / tensors 720
|
| 38 |
+
> Exact parameter count 340,132,080
|
| 39 |
+
> Parameters (billions) 0.340132 B
|
| 40 |
+
> Base model zimage
|
| 41 |
+
> Trigger words r128zkitsu
|
| 42 |
+
> Scan time 0.9539 sec
|
| 43 |
+
|
| 44 |
+
### [Kintsugi Art & Interior Concept | Z-IMAGE LoRA (Rank 128)](https://civitai.com/models/2975100/kintsugi-art-and-interior-concept-or-z-image-lora-rank-128)
|