Update README.md

README.md

@@ -45,12 +45,20 @@ pipeline_tag: text-to-image
 ## Training Configuration & Tools
 
 - **Moss Image**
-  - Training Tools: [Onetrainer (Koratahiu‘s Fork)](https://github.com/Koratahiu/OneTrainer/tree/moss-image)
+  - Training Tools: [Onetrainer (Koratahiu‘s Fork)](https://github.com/Koratahiu/OneTrainer/tree/moss-image) / [Fizgig](https://github.com/shootthesound/Fizgig)
   - Dataset: Mostly 50+ images (averaging 100+)
   - Tagging Style: Mixed format (Booru tags as prefixes combined with natural language)
   - Training  Res: 512x512
   - Repeats per Image: 50+
-  - Training Parameters: Default settings from Onetrainer
+  - Training Parameters: Default settings from Onetrainer / Fizgig
+    - Onetrainer tweak:
+      - Attention:`torch CuDNN` (overriding the default SDPA)
+    - Fizgig tweak:
+      - Presets: `Moss-Image Fast(rank 16, 0.25MP)`
+      - Pictrue Size: 0.26 MP ≈ 512 res
+  - Training Speed (Benchmark: 39 imgs, 0.26MP, r16):
+    - OneTrainer: 910 steps, 1.12 s/it (Batch 3) → ~0.37s/img (17m 43s)
+    - Fizgig: 2730 steps, 1.15 it/s (Grad Accum 3) → ~0.87s/img (36m 40s)
   - Showcase & Card Image Parameters:
     - Model:
       -